2022 - Williams - Formulating Poorly Water Soluble Drugs

AAPS Advances in the Pharmaceutical Sciences Series 50
Robert O. Williams III

Daniel A. Davis Jr.
Dave A. Miller Editors
Formulating
Poorly Water
Soluble Drugs
Third Edition
AAPS Advances in the Pharmaceutical
Sciences Series
Volume 50
Series Editor
Yvonne Perrie, Strathclyde Institute of Pharmacy and Biomedical Sciences
University of Strathclyde, Glasgow, UK
The AAPS Advances in the Pharmaceutical Sciences Series, published in
partnership with the American Association of Pharmaceutical Scientists, is
designed to deliver volumes authored by opinion leaders and authorities from
around the globe, addressing innovations in drug research and development,
and best practice for scientists and industry professionals in the pharma and
biotech industries. Indexed in
Reaxys
SCOPUS Chemical Abstracts Service (CAS) SCImago
EMBASE
More information about this series at http://www.springer.com/series/8825

Robert O. Williams III
Daniel A. Davis Jr. • Dave A. Miller
Editors
Formulating Poorly
Water Soluble Drugs
Third Edition
Editors
Robert O. Williams III Daniel A. Davis Jr.
Division of Pharmaceutics Division of Pharmaceutics
The University of Texas at Austin College of Pharmacy
Austin, TX, USA The University of Texas at Austin
Austin, TX, USA
Dave A. Miller
DisperSol Technologies, LLC
Georgetown, TX, USA
ISSN 2210-7371 ISSN 2210-738X (electronic)

AAPS Advances in the Pharmaceutical Sciences Series
ISBN 978-3-030-88718-6 ISBN 978-3-030-88719-3 (eBook)
https://doi.org/10.1007/978-3-030-88719-3
# American Association of Pharmaceutical Scientists 2022

AAPS Dual imprint and the AAPS holds the copyright
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher,
whether the whole or part of the material is concerned, specifically the rights of translation,
reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any
other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publishers, the authors, and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publishers nor
the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publishers
remain neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
We wish to sincerely thank our friends and colleagues who have
helped with this third edition of this book covering a most
important topic of formulating poorly water-soluble drugs.
Without your insight, wisdom, expertise, time and enthusiasm,
this book would not have been possible.
To my ever loving and supportive family, Jill, Rory, and Maddi,
for your continued patience, encouragement, and sense of
humor throughout the writing of this third edition and editing
process. I love you all!
Bill
To my wife, Sydney. Thank you for always being there with your
unconditional love and support!
Danny
To Allison, Westley, Gwendolyn, and Lyla for your constant love

and support.
Dave
Preface
High-throughput screening (HTS) methodologies for lead identification in

drug discovery were developed in the 1980s to enable the utilization of
advances in genomics and combinatorial chemistry. Since their advent, HTS
methodologies have developed rapidly and have been widely adopted in the
pharmaceutical industry. Consequently, the number of potential drug
candidates identified by HTS has steadily increased over the decades. The
HTS approach tends to identify leads with high-molecular weight and
lipophilicity, and, consequently, poor water solubility. As more and more
leads are identified by HTS, poorly water-soluble drug candidates are
emerging from drug discovery with greater frequency. The problem of poor
solubility has therefore become pervasive in the pharmaceutical industry
recently, with percentages of poorly water-soluble compounds in develop-
ment pipelines reaching as high as 80–90% depending on the therapeutic area.
Drug dissolution is a necessary step to achieve systemic exposure that
ultimately leads to binding at the biological target to elicit the therapeutic
effect. Poor water solubility hinders dissolution and therefore limits drug
concentration at the target site, often to an extent that the therapeutic effect
is not achieved. This can be overcome by increasing the dose; however, it may
also lead to highly variable absorption that can be detrimental to the safety and
efficacy profile of the treatment. In these cases, solubility enhancement is
required to improve exposure, reduce variability, and, ultimately, improve the
drug therapy. It is therefore understood that in modern pharmaceutical devel-
opment, solubility-enhancement technologies are becoming critical to render-
ing viable medicines from the growing number of insoluble drug candidates.
A pharmaceutical scientist’s approach toward solubility enhancement of a
poorly water-soluble molecule typically includes detailed characterization of
the compounds physiochemical properties, solid-state modifications,
advanced formulation design, nonconventional process technologies,
advanced analytical characterization, and specialized product performance
analysis techniques. The scientist must also be aware of the unique regulatory
considerations pertaining to the nonconventional approaches often utilized for
poorly water-soluble drugs. One faced with the challenge of developing a
drug product from a poorly soluble compound must possess at minimum a
working knowledge of each of the above-mentioned facets and detailed
knowledge of most. In light of the magnitude of the growing solubility
problem to drug development, this is a significant burden especially when
vii
viii Preface
considering that knowledge in most of these areas is relatively new and

continues to develop. There are numerous literature resources available to
pharmaceutical scientists to educate and provide guidance toward
formulations development with poorly water-soluble drugs; however, a sin-
gle, comprehensive reference is lacking. Furthermore, without access to a vast
journal library, the detailed methods used to implement these approaches are
not available. The objective of this book is therefore to consolidate within a
single text the most current knowledge, practical methods, and regulatory
considerations pertaining to formulations development with poorly water-
soluble molecules.
The volume begins with an analysis of the various challenges faced in the
delivery of poorly water-soluble molecules according to the route of adminis-
tration, that is, oral, parenteral, and pulmonary. This chapter provides under-
standing of the formulation strategies that one should employ depending on
the intended route of administration. Chapter 2 covers analytical techniques
most pertinent to poorly water-soluble drugs with regard to preformulation,
formulation characterization, and in vitro performance assessment. Solid-state
approaches to overcoming solubility limitations are discussed in Chap. 3. This
chapter presents an in-depth review of the solubility benefits obtained via
conversion of drug crystals to salts, cocrystals, metastable polymorphs, and
amorphous forms. When such solid-state approaches are not viable, particle-
size reduction of the stable crystalline form is perhaps the next most straight-
forward option. In Chap. 4, mechanical particle-size reduction technologies
are described, providing a comprehensive discussion of traditional and
advanced milling techniques commonly used to increase surface area and
improve dissolution rates.
Oftentimes, modification of the API form is not possible and particle-size
reduction fails to appreciably increase the dissolution rate owing to the
inherent solubility limitation of the stable crystalline polymorph. In these
cases, a noncrystalline approach is necessary; perhaps the most straightfor-
ward noncrystalline approach is a solution-based formulation. Solution-based
approaches are covered by Chaps. 5, 6, and 7 where liquid formulation
technologies for poorly water-soluble drugs are presented. Chapter 5 provides
a review of solution systems for oral delivery whereby the molecule is
dissolved in a suitable nonaqueous vehicle. The chapter discusses the various
vehicles available for such systems as well as options for conversion to a final
dosage form. Chapter 6 reviews techniques for overcoming compound solu-
bility challenges in developing liquid formulations for parenteral administra-
tion, which is of particular relevance as the number and complexity of cancer
therapeutics continue to increase. Advanced liquid formulations for oral
delivery, self-emulsifying systems, are discussed in Chap. 7. These systems
are advancements over traditional solution formulations in that the formula-
tion droplet size formed on contact with GI fluids can be controlled through
rational formulation design. Controlling droplet size to the micro- or nanome-
ter scales has been shown to produce significant enhancements in drug
absorption.
In many cases, poorly water-soluble compounds also exhibit limited solu-
bility in vehicles suitable for oral liquid formulations. In these cases (assuming
Preface ix
all other previously mentioned options are not viable), an amorphous formu-
lation approach is often necessary. The design of amorphous formulations
presents numerous challenges, which much of the latter half of this book
(Chaps. 8, 9, 10, 11 and 12) aims to address. These chapters describe the
importance of appropriate preformulation studies, formulation design, process
selection, as well as considerations specific to the selected process technology.
In Chap. 8, a structured, rational approach toward the development of
optimized amorphous solid dispersion formulations is presented. Specific
emphasis is given to critical preformulation studies, identification of the best
excipient carrier system, optimization of drug loading, and process technol-
ogy selection. Chapter 9 provides a comprehensive guide to the application of
hot-melt extrusion technology for the formulation of poorly water-soluble
drugs. This chapter provides a detailed overview of the process technology as
well as formulation design considerations specific to hot-melt extrusion
applications. Spray drying is the subject of Chap. 10, again emphasizing the
process technology and formulation development specific to spray drying.
Particular focus is given to the development of amorphous spray-dried
dispersions owing to its industrial relevance to the production of viable
products containing poorly water-soluble drugs. Chapter 11 teaches cryogenic
technologies whereby nanostructured particles and amorphous solid
dispersions are formed by rapid freezing technologies. The chapter discusses
different cryogenic process technologies, formulation design considerations,
and downstream processing options. Precipitation technologies for the pro-
duction of engineered particles and solid dispersions are covered in Chap. 12.
Various solvent/antisolvent techniques are discussed along with formulation
design principles, particle recovery techniques, and key process design
considerations.
Emerging technologies relevant to the formulation of poorly water-soluble
drugs are discussed in Chap. 13. These are technologies that have begun to
appear in the literature and elsewhere in recent years, which exhibit promise
but have yet to mature. Finally, in Chap. 14, regulatory considerations specific
to drug products of poorly water-soluble compounds are presented. It is the
aim of this chapter to educate formulation scientists regarding unique regu-
latory aspects to consider for solubility-enhancement approaches, that is,
solid-state modifications, particle-size reduction, lipid/solution formulations,
and amorphous solid dispersions. This chapter also provides a unique review
of case studies for marketed products that employ these solubility-
enhancement approaches, highlighting the principal regulatory concerns for
each case.
This volume is intended to provide the reader with a breadth of understand-
ing regarding the many challenges faced with the formulation of poorly water-
soluble drugs as well as in-depth knowledge in the critical areas of develop-
ment with these compounds. Further, this book is designed to provide practi-
cal guidance for overcoming formulation challenges toward the end goal of
improving drug therapies with poorly water-soluble drugs. Enhancing solu-
bility via formulation intervention is a unique opportunity in which formula-
tion scientists can enable drug therapies by creating viable medicines from
seemingly undeliverable molecules. With the ever-increasing number of
x Preface
poorly water-soluble compounds entering development, the role of the for-

mulation scientist is growing in importance. Also, knowledge of the advanced
analytical, formulation, and process technologies as well as specific regulatory
considerations related to the formulation of these compounds is increasing in
value. Ideally, this book will serve as a useful tool in the education of current
and future generations of scientists, and in this context contribute toward
providing patients with new and better medicines.
The editors sincerely thank all contributors for their dedication toward
achieving the vision of this book. It is thanks only to your knowledge and
efforts that it was accomplished.
Austin, TX, USA Robert O. Williams III

Austin, TX, USA Daniel A. Davis Jr.
Georgetown, TX, USA Dave A. Miller
Contents
1 Route-Specific Challenges in the Delivery

of Poorly Water-Soluble Drugs . . . . . . . . . . . . . . . . . . . . . . . 1
Zachary Warnken, Hugh D. C. Smyth,
and Robert O. Williams III
2 Optimizing the Formulation of Poorly
Water-Soluble Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Xiangyu Ma, Daniel Ellenberger,
Kevin P. O’Donnell, and Robert O. Williams III
3 Solid-State Techniques for Improving Solubility . . . . . . . . . . 103
Miguel O. Jara, Justin R. Hughey, Siyuan Huang,
4 Mechanical Particle-Size Reduction Techniques . . . . . . . . . . . 141
Javier O. Morales, Alan B. Watts,
and Jason T. McConville
5 Cosolvent and Complexation Systems . . . . . . . . . . . . . . . . . . 179
Junhuang Jiang and Robert O. Williams III
6 Injectable Formulations of Poorly
Water-Soluble Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Hannah L. O’Mary and Zhengrong Cui
7 Lipid-Based Formulations . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Daniel A. Davis Jr., Han-Hsuan Peng,
8 Structured Development Approach
for Amorphous Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Susanne Page, Reto Maurer, Nicole Wyttenbach,
and Felix Ditzinger
9 Melt Extrusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Stephen A. Thompson, Daniel A. Davis Jr.,
and James C. DiNunzio, Charlie Martin,
Robert O. Williams III, and Feng Zhang
xi
xii Contents
10 Spray-Drying Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

Dave A. Miller, Daniel Ellenberger, Tiago Porfirio,
and Marco Gil
11 Pharmaceutical Cryogenic Technologies . . . . . . . . . . . . . . . . 453
Sawittree Sahakijpijarn, Chaeho Moon,
12 Precipitation Technologies for Nanoparticle Production . . . . . 529
Tuangrat Praphawatvet and Robert O. Williams III
13 Emerging Technologies to Increase the Bioavailability
of Poorly Water-Soluble Drugs . . . . . . . . . . . . . . . . . . . . . . . 599
Daniel A. Davis Jr., Rishi Thakkar, Mohammed
Maniruzzaman, Dave A. Miller, and Robert O. Williams III
14 Scientific and Regulatory Considerations for Development
and Commercialization of Poorly Water-Soluble Drugs . . . . . 651
Zedong Dong and Hasmukh Patel
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675
Route-Specific Challenges
in the Delivery of Poorly Water-Soluble 1
Drugs
Zachary Warnken, Hugh D. C. Smyth, and Robert O. Williams III
Abstract site, and the excipients commonly used in

Poor aqueous solubility of new chemical these formulations. Successful formulation
entities presents various challenges in the design of poorly soluble drugs’ intended alter-
development of effective drug delivery native routes of administration may be hin-
systems for various delivery routes. Poorly dered by the limited number of excipients
soluble drugs that are delivered orally may generally recognized as safe for this route of
commonly result in low bioavailability and delivery and the anatomical and physiological
are often subject to considerable food effects. clearance mechanisms found in these tissues.
In addition, poorly soluble drugs intended for In summary, this chapter reviews the specific
parenteral delivery may also have to be challenges faced in the delivery of poorly
solubilized with large amounts of cosolvents water-soluble drugs via oral, parenteral, and
and surfactants, oftentimes resulting in adverse mucosal administration.
physiological reactions. Other routes also offer
unique opportunities for this class of drug Keywords
molecules but also their own challenges. Ocu- Oral · Parenteral · Pulmonary administration ·
lar delivery of poorly soluble drugs is chal- Aqueous solubility · Food effects ·
lenging due to the efficient absorption Metabolism · Biopharmaceutics Drug
barriers and clearance mechanisms. Develop- Disposition Classification System (BDDCS)
ment of poorly soluble drugs administered
mucosally through routes such as the nasal
cavity, oral mucosa, and others may be 1.1 Introduction
restricted by the relatively small administered
volume, the geometry of the administration Adequate aqueous solubility of new chemical
entities (NCEs) is one of the key properties
required for successful pharmaceutical formula-
Z. Warnken (*) tion development. Solubility is generally defined
Via Therapeutics LLC, Austin, TX, USA
e-mail: [email protected] as the concentration of the compound in a solu-
tion which is in contact with an excess amount of
H. D. C. Smyth · R. O. Williams III
Division of Molecular Pharmaceutics and Drug Delivery, the solid compound when the concentration and
College of Pharmacy, The University of Texas at Austin, the solid form do not change over time (Sugano
Austin, TX, USA et al. 2007). Solubility is closely related to disso-
e-mail: [email protected]; bill. lution which is a kinetic process that involves the
[email protected]
# The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 1

R. O. Williams III et al. (eds.), Formulating Poorly Water Soluble Drugs, AAPS Advances in the
Pharmaceutical Sciences Series 50, https://doi.org/10.1007/978-3-030-88719-3_1
2 Z. Warnken et al.
detachment of drug molecules from the solid sur- invasive intraocular injections. Anatomical
face and subsequent diffusion across the diffusion features of the eye form barriers for drug absorp-
layer surrounding the solid surface. The relation- tion into the eye. Additionally, clearance
ship of solubility and dissolution rate is described mechanisms on the surface and inside the eye
by the Nernst–Brunner/Noyes–Whitney add challenges to effective drug delivery. Poorly
equation: soluble drugs delivered nasally are limited by the
small deliverable volumes, nasal mucosal irrita-
dM D ∙ A
¼ ∙ ðcs ct Þ, tion, and relatively short retention times for
dt h
absorption. Finally, formulation design of poorly
where dM/dt is the dissolution rate, D the diffu- soluble drugs intended for pulmonary administra-
sion coefficient, A the surface area, h the diffusion tion is limited by the few excipients already in
layer thickness, cs the saturation solubility of the approved products and generally recognized as
drug in the bulk medium, and ct the amount of safe for this route of delivery. This chapter
drug in solution at time t (Noyes and Whitney reviews the specific challenges faced in the deliv-
1897; Nernst 1904). The use of high-throughput ery of poorly water-soluble drugs for oral, paren-
screening and combinatorial chemistry for the teral, and mucosal delivery.
development of NCEs has resulted in an increas-
ingly number of compounds that are
characterized by low aqueous solubility (Lipinski 1.2 Oral Route of Administration
2000). From the Nernst–Brunner/Noyes–
Whitney equation, it is evident that compounds Despite significant advances in pulmonary, trans-
characterized by low solubility (cs) will only dermal, and other sites of drug delivery, the oral
establish a small concentration gradient (cs –ct), route remains the most favored method of admin-
resulting in low dissolution rates. This, in turn, istration for systemic administration. Not only are
causes many problems in vivo when poorly solu- oral drug products conveniently and painlessly
ble drugs are administered via various routes of administered resulting in high acceptability, they
administration. Poorly soluble drugs that are can also be produced in a wide variety of dosage
delivered orally without consideration for forms at comparably low costs, making them
improving solubility will commonly result in attractive for patients and pharmaceutical
low bioavailability and high intersubject companies alike (Sastry et al. 2000; Gabor et al.
variability. Additionally, poorly soluble 2010). In theory, the physiology of the gastroin-
compounds are known to have a higher predispo- testinal (GI) tract with its high intestinal surface
sition for interaction with food resulting in high area and rich mucosal vasculature offers the
fast/fed variability (Gu et al. 2007). In order to potential for excellent drug absorption and
make low-solubility drugs available for intrave- accordingly high bioavailability (Lee and Yang
nous administration, they generally have to be 2001). Still, oral bioavailability is often low and
solubilized, for example, by employing large variable as the process of drug absorption from
amounts of cosolvents and surfactants. Problems the GI tract is far more complex and influenced by
often arise when these excipients may not be well physiological factors such as GI motility, pH,
tolerated, potentially causing hemolysis and/or efflux transporters, and presystemic metabolism;
hypersensitivity reactions (Yalkowsky et al. extrinsic factors such as food intake and formula-
1998). In addition, there may be a risk of drug tion design; and critically, the physicochemical
precipitation upon injection due to the subsequent properties of the drug (Levine 1970; Martinez
dilution of the solubilized formulation. and Amidon 2002).
Depending on the intended target tissue, ocular Following oral administration of a solid dos-
delivery may be accomplished utilizing various age form, the drug must first dissolve in the GI
dosage forms, from topical eye drops to more fluids, be absorbed across the intestinal mucosa,
and pass through the liver to reach the systemic
1 Route-Specific Challenges in the Delivery of Poorly Water-Soluble Drugs 3
circulation and exert its pharmacological effect. 1.2.1 Challenges in Oral Delivery
Accordingly, the key properties of potential drug of Poorly Water-Soluble Drugs
candidates defining the extent of oral bioavail-
ability and thus being vital for successful oral Co-administration of oral dosage forms with
product development include aqueous solubility meals generally results in one of the three
and intestinal permeability. Based on these two scenarios: (1) the extent of absorption decreases
crucial parameters, the Biopharmaceutics Classi- which is referred to as a negative food effect;
fication System (BCS) assigns drugs to one of the (2) the extent of absorption increases
four categories: high solubility, high permeability corresponding to a positive food effect; and
(BCS I); low solubility, high permeability (BSC (3) no substantial change in the extent of absorp-
II); high solubility, low permeability (BCS III); tion takes place (Welling 1996). Given the fact
and low solubility and low permeability (BCS IV) that food intake commonly translates into univer-
(Amidon et al. 1995). sal physiological actions, predictions of what sce-
Ideally, a NCE is characterized by high aque- nario will take place may be made based on the
ous solubility and permeability (BCS I); yet, physicochemical properties of the drug (Gu et al.
reported in 2006 by Benet et al., only about 5% 2007). For instance, Fleisher et al. estimated the
of NCEs fulfill this requirement, while approxi- effect of food on the extent of drug absorption
mately 90% of NCEs are considered poorly solu- based on the characteristics of the drug as classi-
ble in combination with either high or low fied by the BCS (Fleisher et al. 1999). Specifi-
permeability (BCS II and IV) (Benet et al. cally, it was suggested that the extent of
2006). This is in part a result of contemporary absorption of a poorly water-soluble, highly per-
approaches used in molecule discovery and syn- meable BCS II drug is most likely increased,
thesis as well as a necessity for molecule while it will remain unchanged for a highly
lipophilicity to interact with current molecular water-soluble and permeable BCS I drug. In
targets (Boyd et al. 2019). Due to the combination fact, the same trend was observed by Gu and
of low permeability and low solubility, BCS IV coworkers, who evaluated the effect of food
compounds are generally troublesome drug intake on the extent of absorption, defined as the
candidates and, therefore, rarely developed and area under the curve of the time–plasma concen-
marketed. BCS II compounds are usually more tration curve (AUC), by analyzing clinical data of
promising candidates since permeability through 90 marketed drug products (Gu et al. 2007). For
the GI mucosa is not a problem. Nevertheless, the majority of products containing a BCS I com-
intestinal absorption is solubility/dissolution pound (67%), no statistically significant differ-
rate-limited, oftentimes resulting in low and ence in the AUC in the fasted and fed state was
erratic oral bioavailability. observed. In contrast, more than 70% of the drug
In addition to oral dosage forms which are products comprising BCS II or BCS IV drugs
ingested for the intention of drug absorption tak- exhibited a positive food effect as indicated by a
ing place in the gastrointestinal tract, there are significant increase in the AUC in the fed state
transmucosal oral dosage forms for absorption compared to the fasted state (Fig. 1.1). Mathias
across the mucosa in the mouth including sublin- et al. further confirmed this effect by studying
gual and buccal products which bypass the first- in vitro–in vivo relationships of 22 new chemical
pass effect. entities (Mathias et al. 2015).
Overall, problems associated with poorly sol- The positive food effect oftentimes encoun-
uble compounds not only revolve around low oral tered with poorly water-soluble drugs can be pri-
bioavailability but also involve high susceptibility marily ascribed to several physiological changes
to factors such as food and metabolism as in the GI environment that ultimately increase
discussed in more detail in the following sections. drug solubility and dissolution. First of all, the
intake of food is known to delay gastric emptying
4 Z. Warnken et al.
Fig. 1.1 Occurrence of

food effects (positive, No food effect
90 Negative food effect
negative, or no effect) in
percent by Positive food effect
Biopharmaceutics 80
Classifications System
(BCS) category (Gu et al. 70
2007). (Adapted with
permission) 60
Percent
50
40
30
20
10
0
BCS I BCS II BCS III BCS IV
which, in turn, is beneficial in terms of absorption from 13 min (fasted state) to 49 min (fed state)
as it increases the time available for drug dissolu- was considered to play a role in bioavailability
tion (Charman et al. 1997). Second, a substantial enhancement.
rise in the gastric and intestinal fluid volume in In the case of weakly acidic or basic drugs,
the fed state offers the potential for increased which in the aqueous GI environment exist in
dissolution rates (Custodio et al. 2008; Tanaka ionized and unionized form, variations in gastro-
et al. 2015). Furthermore, food intake stimulates intestinal pH due to food intake can significantly
the release of bile from the gallbladder into the increase or decrease drug solubility. In healthy
duodenum where its components, primarily bile subjects, the gastric pH in the fasted state typi-
salts, cholesterol, and phospholipids, solubilize cally lies in the range of 1–3 but may temporarily
dietary lipids into mixed micelles (Hofmann and rise to 4–7 after meal intake (Lee and Yang 2001;
Mysels 1987). Similarly, these mixed micelles Dressman et al. 2007). Studies using the
have the ability to incorporate lipophilic drug SmartPill®, a telemetric capsule which can mon-
molecules potentially boosting drug solubility itor pH changes during motility in the gastrointes-
by several orders of magnitude (Dressman et al. tinal track, found that the pH increases to 3.3–5.3
2007). Bile salts may also enhance the dissolution after intake of a high-caloric, high-fat meal
rate of poorly soluble drugs by improved wetting (Koziolek et al. 2015). Since the extent of ioniza-
which is predominantly the case when their con- tion and consequently the solubility of a weakly
centration stays below the critical micelle concen- acidic drug is generally greater at elevated pH,
tration. As an example, a study conducted in food intake may enhance drug dissolution in the
healthy male volunteers found that the oral bio- stomach. In contrast, the extent of ionization of a
availability of danazol, a BCS II drug, was weakly basic drug will be reduced at increased
increased by 400% when administered together gastric pH, potentially resulting in reduced disso-
with a lipid-rich meal (Sunesen et al. 2005). This lution and/or potential precipitation of already
can be attributed to the presence of bile salts and dissolved drug molecules. Changes in the pH in
lecithin in the small intestine allowing for micel- the stomach and the intestines as well as the
lar solubilization of the drug (Anby et al. 2014). transition in pH from the stomach to the intestines
In addition, an increase in gastric emptying time can not only directly affect drug solubility but can
also affect the performance of drug delivery The presence of metabolic enzymes of cyto-
systems. Amorphous solid dispersions, for exam- chrome P 450 (CYP 450) within the endoplasmic
ple, using pH-dependent soluble polymers such reticulum of hepatocytes and intestinal
as hypromellose acetate succinate or enterocytes may significantly decrease oral bio-
hypromellose phthalate can resist drug release in availability of many drugs (Lee and Yang 2001;
the stomach and target release after the transition Paine et al. 2006). Presystemic metabolism of
to the higher pH environment of the intestines drugs is often referred to as first-pass metabolism.
such as with Noxafil® (posaconazole) Smith et al. suggested that this will particularly be
(Monschkle and Wagner 2019). Another potential the case for drugs that are lipophilic and therefore
complication has been reported by Jara et al., easily cross cell membranes, thereby gaining
showing how the acidic environment of the stom- access to CYP enzymes (Smith et al. 1996). Fur-
ach resulted in crystallization of a drug from an ther analysis by Wu and Benet confirmed that
amorphous solid dispersion, negating the highly permeable BCS I and II drugs are primar-
advantages of the drug delivery system if not ily eliminated via metabolism, while poorly per-
protected before reaching the intestines (Jara meable BCS III and IV drugs are mostly
et al. 2021). eliminated unchanged into the urine and bile
Due to their high sensitivity to gastrointestinal (Wu and Benet 2005; Benet 2010). It should be,
changes caused by food intake, poorly soluble however, noted that the low/high permeability
compounds are often associated with extremely characteristics as defined in the BCS reflect the
variable and unpredictable oral bioavailability. differences in access of the drug to metabolic
Especially in the case of drugs that exhibit a enzymes within the cells and not necessarily
narrow therapeutic window, sub-therapeutic or differences in permeability into the cells
toxic concentrations of the drug in the systemic (Custodio et al. 2008).
circulation may easily occur. To prevent either Based on their findings, Wu and Benet pro-
scenario, patients generally have to adhere to posed the Biopharmaceutics Drug Disposition
certain food restrictions, potentially compromis- Classification System (BDDCS) in which drugs
ing patient compliance and quality of life. are categorized in terms of extent of metabolism
It should be noted though that the occurrence and solubility as opposed to permeability and
of food effects may be prevented by the selection solubility used in the BCS (Fig. 1.2). According
of an appropriate formulation design. Several for- to the BDDCS, poorly soluble, highly permeable
mulation approaches that enhance drug solubility BCS II compounds are characterized by extensive
and therefore enable class II drugs to act as class I metabolism defined as 70% metabolism of an
drugs have already been successfully applied to oral dose in vivo in humans.
reduce or eliminate fed/fasted variability (Yasuji The BDDCS also considers the influence of
et al. 2011). These include, among others, active uptake/efflux transporters on drug disposi-
nanoparticulate (Jinno et al. 2006; Sauron et al. tion as shown in Fig. 1.3. Since most BCS II
2006), self-emulsifying (Perlman et al. 2008; compounds are substrates or inhibitors for P-gly-
Woo et al. 2008), and solid dispersion-based coprotein (P-gp), a transmembrane efflux trans-
drug delivery systems (Klein et al. 2007; porter, it is expected that the interplay of P-gp and
Mogalian et al. 2014), all of which will be metabolizing enzymes will notably influence the
addressed in depth in upcoming chapters. extent of metabolic extraction and oral bioavail-
The extent of oral bioavailability is affected ability of BCS II substrates (Custodio et al. 2008).
not only by drug characteristics such as solubility Results from a number of studies aimed at
and gastrointestinal permeability but also by a understanding the interaction of CYP 450
drug molecule’s susceptibility to intestinal and enzymes and P-gp and its effect on compounds
hepatic metabolism and active influx/efflux that are dual substrates suggest that both work
transporters. synergistically to increase presystemic metabo-
lism (Hochman et al. 2000). It is assumed that
6 Z. Warnken et al.
Fig. 1.2 The

Biopharmaceutics Drug
Disposition Classification
System (BDDCS)
(Custodio et al. 2008).
(Reprinted with
permission)
Fig. 1.3 Transporter

effects, following oral
dosing, by
Biopharmaceutics Drug
Disposition Classification
System (BDDCS) class
(Custodio et al. 2008).
(Reprinted with
permission)
exposure of drugs, which are substrates of P-gp, be circumvented with oral dosage forms using
to intestinal CYP 450 enzymes is increased due to transmucosal formulations (Patel et al. 2011). In
repeated cycles of intracellular uptake and efflux. oral transmucosal formulations, drugs are directly
However, the complexity of metabolic enzyme-P- absorbed across the mucosa in the oral cavity into
gp interactions is still only partially understood systemic circulation. An array of dosage forms
(Knight et al. 2006; Mudra et al. 2011). exists to take advantage of this pathway of drug
Initial metabolism resulting from the first pass delivery including gums, tablets, films, patches
effect in the gastrointestinal tract and the liver can and sprays for applying drugs to different regions
Fig. 1.4 Characteristic of different regions of the oral mucosa as it relates to oral transmucosal drug delivery (Patel et al.
2011)
of the oral cavity. Figure 1.4 depicts the different absorption. Some dosage forms such as a
regions of the oral mucosa as it relates to drug mucoadhesive tablets for buccal administration
delivery. As many of these dosage forms are of cannabidiol have shown extended absorption
easily self-administered and do not require times in vivo in part due to the delivery system as
swallowing, they have been used for palliative well as the inherent lipophilicity of the molecule
care treatments such as for breakthrough pain resulted in a reservoir in the tissue (Itin et al.
(Lam et al. 2020). 2020). Other challenges to drug delivery using
While attractive for particular applications, oral transmucosal dosage forms include patient
oral transmucosal delivery of poorly water- condition, as they may not be suitable for patients
soluble drugs comes with its own challenges. that are experiencing nausea and vomiting or also
Firstly, with particular relevance for poorly may be a choking hazard for very young or
water-soluble drugs, is the limited volume avail- elderly patients. As the dosage forms are meant
able for dissolution. Despite reports of up to 2.0 L to remain in the mouth throughout the absorption
of daily salivary secretions, some estimates indi- process, the taste of the drug may also be a com-
cate there is only around 1.1 mL present in the plication when using these types of delivery
mouth for dissolving drug molecules (Patel et al. systems.
2011). Additionally, drug product residence time
can be relatively short depending on the particular
dosage form used for transmucosal delivery.
1.3 Parenteral Route
Although the relative permeability of the sublin-
of Administration
gual mucosa is greater due to it being thinner than
other regions of the oral cavity and not having a
Parenteral administration is commonly defined as
keratinized permeation barrier, delivery by this
the injection of dosage forms by subcutaneous,
mucosa can be complicated by shorter residence
intramuscular, intra-arterial, and intravenous
times of the drug at the site for absorption as it is
(i.v.) routes (Jain 2008). In the case of
in contact with salivary secretions which results
i.v. administration, the drug is directly delivered
in swallowing of the drug rather than direct
to the bloodstream, thereby allowing for rapid
8 Z. Warnken et al.
distribution to highly perfused organs. The con- (Wong et al. 2008). Preventing particle agglom-
sequently rapid onset of pharmacological effect eration, aggregation, or crystal growth by adding
that is achieved by i.v. administration is critical suitable stabilizers is vital as an increase in parti-
for several clinical conditions that require imme- cle size could result in the mechanical blockage of
diate action such as cardiac arrest and anaphylac- small-caliber arterioles and capillaries. The
tic shock (Shi et al. 2009). In addition, choice of stabilizers and generally excipients
i.v. administration is advantageous for drugs for accepted for i.v. administration is, however,
which oral delivery would result in low and rather limited which presents a common chal-
erratic bioavailability due to gastrointestinal deg- lenge for the formulation strategies mentioned.
radation or significant presystemic/first-pass Another consideration is the rapid clearance of
metabolism. Overall, i.v. administration offers nanoparticles following i.v. administration due to
excellent control over the actual dose and rate at opsonization. Formulators should consider the
which the drug is delivered, providing more pre- dynamics of the particle surface chemistry during
dictable pharmacokinetic and pharmacodynamic and after administration.
profiles than obtained after oral administration
(Bhalla 2007).
Since i.v. formulations are directly injected 1.3.1 Challenges in Parenteral
into the bloodstream, they are subject to strict Delivery of Poorly
regulatory requirements regarding their physical Water-Soluble Drugs
and chemical stability as well as their
microbiological characteristics. The latter Poorly soluble weak acids or bases may be
implicates that products intended for solubilized by pH modification of the solution to
i.v. administration must be sterile and free of be administered. Yet, if the drug is characterized
pyrogens (Akers 2014). Additionally, the pH by very low solubility, pH adjustment to extreme
and tonicity of i.v. products should be carefully values might be necessary to achieve the desired
considered to prevent irritation, pain, and hemo- drug concentration in solution (Lee et al. 2003). It
lysis of blood cells. To achieve the highest possi- is recommended, however, that the pH for
ble in vivo tolerability for an i.v. product, it i.v. infusions should be in the range of 2–10 in
should ideally be formulated as an aqueous- order to reduce side effects such as irritation and
based solution that is isotonic and possesses a pain at the injection side (Egger-Heigold 2005).
pH of 7.4. Clearly, this is not feasible for drugs Side effects may occur not only due to extreme
that are characterized by poor aqueous solubility pH values but also due to potential precipitation
at this specific pH. Generally, poorly soluble of the drug upon injection. A change in pH caused
compounds may be solubilized by pH adjustment by dilution in the bloodstream may reduce the
(if the drug molecule is ionizable), the use of solubility of the drug below the solubility limit
organic solvent mixtures or mixed aqueous/ resulting in precipitation. Buffer species as well
organic cosolvents, and cyclodextrin complexa- as buffer strength have been identified as key
tion (Strickley 2004; Bracq et al. 2008). However, factors influencing drug solubility and conse-
these solubilization approaches are associated quently precipitation in pH-adjusted formulations
with drawbacks such as increased toxicity or the (Narazaki et al. 2007). Phenytoin is a weakly
possibility of drug precipitation upon injection acidic drug which is poorly soluble at pH 7.4
and subsequent dilution (Yalkowsky et al. 1998). and has been reported to precipitate after injec-
Alternatively, the drug can be formulated in tion. Addition of a cyclodextrin as a solubilizing
the form of a dispersion of particles which are agent was shown to reduce the risk of precipita-
suspended in aqueous media. The size distribution upon dilution (McDonald and Muzumdar
tion of intravenous suspensions is critical for 1998). It is essential to prevent precipitation as
safety and distribution of particles in vivo and precipitated drug crystals may cause inflamma-
generally restricted to the submicron range tion of the vein wall, also known as phlebitis,
Table 1.1 Detection of hemolysis by in vivo and in vitro methods

In vitro (% hemolysis detected)
Formulation composition In vivo literature Human blood Rabbit blood Dog blood
Normal saline (NS) No 0.0 0.0 0.0
10% EtOH in NS No 0.0 0.0 10.0
30% EtOH in NS No 0.0 0.0 2.5
40% PG in NS Yes 61.0 37.3 29.7
60% PG in water Yes 100.00 96.7 53.4
10% PG + 30% EtOH in NS No 0.0 0.0 0.0
10% EtOH + 20% PG in water No 8.8 0.0 0.3
10% EtOH + 40% PG in water Yes 69.2 52.6 31.5
20% EtOH + 30% PEG 400 in water No 0.0 0.0 3.3
PG propylene glycol, EtOH ethanol; Amin and Dannenfelser (2006). Reprinted with permission
mainly due to mechanical irritation and prolonged and Yalkowsky 1987; Shalel et al. 2002).
drug exposure at the vein wall (Johnson et al. Resulting hemoglobin release into the blood
2003). Besides, precipitation of solubilized drug plasma may induce vascular irritation, phlebitis,
molecules may result in erratic or reduced bio- anemia, kernicterus, and acute renal failure
availability as well as altered pharmacokinetics (Krzyzaniak et al. 1997; Amin and Dannenfelser
(Yalkowsky et al. 1998). For instance, 2006). The hemolytic potential of these additives
precipitated particles in the low micron to submi- has been evaluated in numerous studies
cron range may be taken up by macrophages of (Zaslavsky et al. 1978; Ohnishi and Sagitani
the reticuloendothelial system following 1993; Mottu et al. 2001). Yet, conflicting results
opsonization resulting in a significantly increased have been reported due to different
drug plasma clearance rates (Bittner and methodologies used. Table 1.1 summarizes
Mountfield 2002). Furthermore, dissolution of in vitro hemolysis data for different cosolvent
precipitated drug at later time points may increase systems obtained in rabbit, dog, and human
the terminal half-life as well as the volume of blood compared to human in vivo data acquired
distribution. from the literature (Amin and Dannenfelser
Drugs that are not sufficiently solubilized by 2006). For all vehicles a higher percentage of
pH adjustment or drugs that have no ionizable hemolysis is seen for data obtained with human
groups may be formulated using organic water- blood followed by rabbit and dog blood; yet, the
miscible cosolvents and surfactants. Frequently rank order of different vehicles evaluated is simi-
used cosolvents for i.v. formulations are propyl- lar for the different species evaluated.
ene glycol, ethanol, and polyethylene glycols, Just like solubilization via pH adjustment, sol-
while commonly used surfactants include poly- ubilization by means of cosolvents has the limita-
sorbate 80, Cremophor EL, and Cremophor RH tion of potential drug precipitation (Li and Zhao
60 (Strickley 2004; Bracq et al. 2008). Highly 2007). Figure 1.5 exemplarily depicts the solubil-
lipophilic compounds may even require formula- ity curve of a drug at different cosolvent levels
tion in a nonaqueous, organic vehicle comprising (squares) compared to the drug concentration
only water-miscible solvents and/or surfactants. curve based on dilution (dots). The saturation
These are commonly concentrates which are solubility of the drug in a 50% (v/v) cosolvent
diluted with aqueous media prior to administra- system is 2.4 mg/mL, while the drug is
tion. Overall, the number and concentration of formulated at a concentration of 1.6
organic solvents and surfactants are limited as mg/mL. Upon injection, the concentrations of
they may cause side effects. Organic solvents as the cosolvent and drug will decrease linearly
well as surfactants have been reported to provoke due to dilution in the bloodstream. In contrast,
hemolysis, the rupturing of erythrocytes (Reed drug solubility will decrease exponentially,
10 Z. Warnken et al.
Fig. 1.5 Illustration of 2500

precipitation of a drug
2.4mg/ml
formulated in a 50% (v/v)
cosolvent system (Li and 2000 solubility curve
Zhao 2007). (Reprinted
with permission) 1.6mg/ml
Drug (ug/ml)
1500
1000 dilution curve
500
0
0 10 20 30 40 50
Cosolvent % (v/v)
causing it to fall below the actual drug concentra- is further diluted with 13% ethanol in water for
tion rapidly. This means that the drug is present in injection and saline or dextrose solution before
the supersaturated state where it is susceptible to i.v. administration. Like Taxol®, Taxotere®
precipitation. It has been suggested that the addi- often results in severe side effects, specifically
tion of surfactants to cosolvent formulations, even severe hypersensitivity reactions, mainly due to
in small concentrations (0.05–0.5% w/v), may the presence of polysorbate 80 in the formulation.
prevent precipitation upon i.v. administration The use of surfactants in i.v. formulations may
(Li and Zhao 2007). not only cause hypersensitivity reactions but also
The formulation of i.v. products with some alter drug pharmacokinetics by interfering with
surfactants, especially in high concentrations, distribution processes, transporters, or metabolic
has been associated with acute hypersensitivity enzymes (Egger-Heigold 2005). It has been
reactions characterized by dyspnea, flushing, reported that Cremophor EL modifies the phar-
rash, chest pain, tachycardia, and hypotension macokinetics of several drugs such as etoposide,
(Ten Tije et al. 2003). Paclitaxel, a poorly doxorubicin, and paclitaxel (Ellis et al. 1996;
water-soluble molecule with antineoplastic activ- Webster et al. 1996; Sparreboom et al. 1996). A
ity, was first formulated in form of a nonaqueous study conducted in mice, which received Taxol®
solution for i.v. infusion (Taxol®), in which the (paclitaxel solubilized in Cremophor EL and eth-
drug is solubilized in a mixture of Cremophor EL anol) by i.v. injection at three different dose
and ethanol (Singla et al. 2002). This formulation levels, revealed a nonlinear pharmacokinetic
can cause significant hypersensitivity reactions, behavior of paclitaxel (Sparreboom et al. 1996).
which are primarily attributed to Cremophor EL, In particular, a disproportional increase in cmax
necessitating premedication of patients with and a decrease in the plasma clearance upon dos-
steroids and antihistamines. Complement activa- age escalation were observed. In contrast,
tion due to binding of the hydroxyl-rich surface of i.v. administration of a Cremophor EL-free solu-
Cremophor EL to naturally occurring anti- tion of paclitaxel in the organic solvent
cholesterol antibodies has been proposed as a dimethylacetamide resulted in a cmax that varied
possible underlying mechanism for the occur- proportionally with dosage as well as a dose-
rence of these hypersensitivity reactions (Szebeni independent clearance. Studies in mice with
et al. 1998). Docetaxel, a semi-synthetic analog of Cremophor EL and various other active
paclitaxel, is solubilized with the nonionic surfac- ingredients have confirmed these findings to be
tant polysorbate 80 in its marketed formulation an effect of the surfactant (Liu et al. 2015). The
Taxotere® (Engels et al. 2007). This concentrate same nonlinear pharmacokinetic was also
observed in an in vivo study involving patients i.v. injection is generally rapid and quantitative,
with solid tumors who were treated with different with the main driving force being the dilution in
dose levels of Taxol® (van Zuylen et al. 2001). It the blood stream (Stella et al. 1999). Problems
has been suggested that the Cremophor may however arise for strongly bound drugs with
EL-related nonlinear paclitaxel pharmacokinetics high complex-forming constants where the drug
is caused by entrapment of the drug into does not rapidly dissociate from the complex
Cremophor EL micelles which function as the potentially altering pharmacokinetics.
primary carrier in the systemic circulation leading Finally, suspending the drug in vehicle can
to a disproportionate paclitaxel accumulation in have particular advantages and challenges
the plasma (Sparreboom et al. 1999). depending on the drug to be administered. By
Alternatively, complexation of poorly water- suspending the drug, it can be possible to avoid
soluble drugs with cyclodextrins has been excipients which may result in unwanted
explored as an approach for i.v. delivery of toxicities. For example, Abraxane®, a nanoparti-
these troublesome compounds. Cyclodextrins cle formulation of paclitaxel using albumin, is an
are cyclic oligosaccharides composed of six, i.v. product which successfully allows paclitaxel
seven, or eight (α-1, 4) linked α-D-glucopyranose dosing without the need for potentially harmful
units corresponding to α-, β-, and γ-cyclodextrins, excipients like the Taxol® product which
respectively (Brewster and Loftsson 2007). They contains Cremophor EL, which is known to
are characterized by a hydrophilic outer surface cause hypersensitivities in some individuals
and a lipophilic inner cavity, which is capable of (Green et al. 2006). Another advantage that
accommodating suitable drug compounds. suspended parenteral products can have over
Cyclodextrins employed for parenteral delivery, solutions can be improved reconstitution times
that is, hydroxypropyl-β-cyclodextrin and especially for medications which need to me
sulfobutylether-β-cyclodextrin, are derivatives of administered rapidly such as Ryanodex®
β-cyclodextrin with increased aqueous solubility (dantrolene sodium) (Schutte et al. 2011). This
and improved in vivo safety profiles (Stella and formulation approach too has its own challenges,
He 2008). Cyclodextrins oftentimes solubilize such as limitations on types and concentrations of
drug molecules as a linear function of their con- excipients to stabilize the suspension and particle
centration. Consequently, dilution of the formu- size limits to avoid being captured in the around
lation in the blood stream upon i.v. administration 6-micron-diameter capillaries of the lungs (Wong
will result in a linear reduction of both drug and et al. 2008).
cyclodextrin concentration. Based on that, drug
precipitation that is oftentimes seen with
cosolvent or pH-adjusted systems is very unlikely 1.4 Ocular Route of Administration
to occur with cyclodextrin-based formulations.
Nevertheless, there can be several shortcomings Drug delivery by the ophthalmic route is
associated with the use of cyclodextrins as means characterized by specialized preparations which
of solubility enhancers. Solubilization by are intended to provide direct contact with the eye
cyclodextrins is not generally applicable to all most often via topical delivery. Currently the
drug molecules. In order to successfully form a most commonly used commercial eye
stable cyclodextrin-drug inclusion complex, the medications are prepared as eye drops, as they
drug molecule needs to have the appropriate size, are relatively easy to administer by patients
shape, and polarity to fit into the central cyclo- (Vandervoort and Ludwig 2007). However,
dextrin cavity (Radi and Eissa 2010). Addition- other ophthalmic dosage forms exist, including
ally, cyclodextrins are excreted in the urine, and gel and ointment-based topicals, intravitreal
accumulation could occur in patients with renal injections, periocular drug delivery preparations,
insufficiency (Stella and He 2008). Drug release and ocular devices. Each of these possesses their
from cyclodextrin inclusion complexes after own advantages and disadvantages for treating
certain diseases of the eye. Ophthalmic than that of topical delivery (0.001% and less)
formulations are targeted for local treatment (Tsuji et al. 1988; Kim et al. 2004; Kaur and
of ocular diseases. By using the ophthalmic Kakkar 2014). Intravitreal injections, injections
route of delivery, therapy can be maximized at directly into the vitreous humor, are the most
the site of action while minimizing systemic direct method of delivering medications to the
exposure, reducing the chances for adverse posterior portion of the eye. Periorbital, but most
events. Drug delivery to the eye is met with its often, intravitreal injections can be used for
own unique challenges which must be overcome treating conditions like age-related macular
to achieve therapeutic delivery which can be reli- degeneration residing in the posterior portion of
ably used by patients. the eye.
The eye comprises two main regions, the ante-
rior and posterior compartments, which are
separated and delineated by the crystalline lens.
1.4.1 Challenges in Ocular Delivery
The layer at the most anterior portion of the eye is
of Poorly Water-Soluble Drugs
the cornea, a window located in front of the lens
that allows light to enter the eye. Eye drops and
For many topically applied drugs to have efficacy,
other topical ophthalmic preparations are
they must permeate across the cornea membrane.
intended for absorption across the cornea into
Transcorneal absorption is the predominate
the aqueous humor, the fluid residing in the ante-
mechanism of entrance for small molecules enter-
rior compartment. This is the site of action for
ing the eye (Urtti 2006). However, absorption
many therapeutic agents, largely including those
into the eye from the external environment is
which lower intraocular pressure for treating
hindered by a number of mechanisms resulting
glaucoma (Weinreb and Khaw 2004). The poste-
in ocular bioavailability which is typically less
rior part of the eye is where the photoreceptors are
than 5% (Urtti 2006). One of these mechanisms
located, allowing visual information to be relayed
is related to the structure of the cornea as depicted
to the brain (Alqawlaq et al. 2012). The chamber
in Fig. 1.6. It consists of five layers which drugs
in the back of the eye is filled with vitreous
must pass through to enter the aqueous humor.
humor. Unlike the aqueous humor, this vitreous
The outermost layer of the cornea is the hydro-
humor media is more gel-like in nature and
phobic stratified squamous epithelium, and
contributes to the orbital structure of the eye
beneath this is Bowman’s membrane. The
(Chowhan et al. 2012). The vitreous humor is
thickest layer of the cornea is a hydrophilic matrix
approximately 4 mL in volume and composed of
located underneath Bowman’s membrane called
98% water along with hyaluronic acid, collagen
the stroma. Following the stroma is Descemet’s
fibrils, and some phagocytic mononuclear cells
membrane then the corneal endothelium, another
(Martens et al. 2013; Sebag 2013). Excluding
hydrophobic layer (Edwards and Prausnitz 1998;
the cornea, the outermost layer of the eye is
Friedman et al. 2007). The complexity of the
made up of the sclera. The sclera is a tough
cornea, transitioning from hydrophobic to hydro-
fibrous layer which is the white of the eye.
philic to hydrophobic layers, makes transcorneal
Drugs which are administered by periorbital
drug transport a challenging route for delivery.
routes may be absorbed through the sclera
Current methods to overcome this barrier include
(Ahmed and Patton 1985). Periorbital routes
increasing the dissolution rate of the drugs and
include peribulbar, subconjunctival, posterior
including excipients for increased permeability
juxtascleral, sub-Tenon, and retrobulbar
(Li et al. 2013; Nagai et al. 2015). Formulating
injections, which administer drugs in contact
poorly soluble drugs as a nanosuspension has
with the sclera for transscleral penetration into
been shown to increase the ocular bioavailability
the vitreous humor and to the retina. The retinal
as well as decrease irritation of the eye (Kim et al.
and vitreal drug bioavailability is about
2011).
0.01–0.1% via these routes, which is much higher
Fig. 1.6 Illustration of the

layers comprising the
cornea membrane (Sharif
et al. 2015). (Reprinted with
Permission)
In addition to the permeability limits for also be utilized to increase contact time for
absorption, topically administered drugs are lim- absorption. However, these formulations are typ-
ited by a relatively short residence time in contact ically more difficult to administer than eye drops
with the cornea. The typical volume of the tear and suffer from greater dose variability (Chowhan
film, the liquid layer coating the rostral surface of et al. 2012).
the eye, is between 5 and 7 μL, but the area as a Many reports have shown that cyclodextrin
whole has a maximum capacity of about 30 μL formulations can achieve effective drug delivery
(Foster and Lee 2013). On average, the of poorly water-soluble drugs administered
administered volume from commercial eye ophthalmically (Kristinsson et al. 1996;
drops is 39 μL, ranging from 25.1 to 56.4 μL Sigurdsson et al. 2005; Jansook et al. 2010;
(Van Santvliet and Ludwig 2004). Volumes Ohira et al. 2015). Cyclodextrins can help
delivered above the maximum capacity of the improve ocular bioavailability by complexing
eye are rapidly cleared, one avenue being through and solubilizing poorly soluble drugs as well as
the nasolacrimal duct which leads to increased by acting as permeation enhancers, increasing the
systemic absorption (Van Santvliet and Ludwig diffusion of drugs across the gel-like inner most
2004). There are several formulation strategies layer of the tear film (Loftsson et al. 2012).
which can be used to help reduce clearance of Jansook et al. formulated dorzolamide as a com-
medications from the ocular surface. Administra- plex with γ-cyclodextrins which formed revers-
tion of eye drops of smaller volume can be as ible mucoadhesive agglomerates in the
efficacious as larger volume doses with the same microparticle range. These suspended particles
concentration solution by reducing the rate at were found to act as a reservoir for sustaining
which the preparation is removed from the site dorzolamide concentrations within the tear film.
of absorption (Petursson et al. 1984). Formulating This resulted in concentrations detectable for up
poorly soluble drugs, for example, acetazolamide to 24 hours after topical administration, while the
or pilocarpine, into eye drops which gel or commercial formulation was shown to have prac-
increase in viscosity after coming into contact tically no drug left in the aqueous humor after
with the eye permits the ease of administration only 8 hours. It has also been reported that
of an eye drop with an increase in residence time cyclodextrins can increase posterior drug delivery
for absorption (Verma et al. 2013; Miyazaki et al. of topically applied medications (Loftsson et al.
2001). Gel and ointment-based formulations can 2007; Loftsson et al. 2008; Jansook et al. 2010;
Ohira et al. 2015). The enhanced posterior deliv- acetonide, lasts for up to 3 years after injection
ery is due to the higher permeability of the con- into the eye which maximizes drug delivery to the
junctiva/sclera membrane compared to that of the retina while minimizing systemic and anterior
cornea (Loftsson et al. 2008). Emulsion drug chamber exposure (Kane et al. 2008).
delivery systems have been reported to increase
drug delivery of poorly soluble drugs (Naveh
et al. 1994; Calvo et al. 1996; Tamilvanan and 1.5 Nasal Route of Administration
Kumar 2011;Ying et al. 2013). Cyclosporine A, a
poorly soluble drug used to treat chronic dry eye Nasal drug delivery can have many potential
disease, is commercially available as Restasis®, a advantages and disadvantages over conventional
viscous emulsion intended for topical eye deliv- oral drug delivery. Nasal drug delivery can be
ery. Restasis® utilizes castor oil as a disperse targeted for treating local and systemic diseases
phase, which is stabilized with polysorbate and, more recently, explored for central nervous
80 and carbomer 1342, to produce an emulsion system (CNS) diseases. Traditionally, nasal deliv-
which is effective and nonirritating to the sensi- ery has been focused on treating local disease
tive eye tissue (Ding et al. 1995; Tamilvanan and such as nasal congestion, nasal allergies, and
Benita 2004). Another barrier recently found to nasal infections (Illum 2003). Systemic delivery
play a role in ocular bioavailability of topically through nasal administration can be advantageous
applied therapeutics is mucus and mucus penetra- for a number of reasons. The relatively high vas-
tion (Popov 2020). cularization and permeability of the nasal respira-
Intravitreal injections can be used to deliver tory epithelium often allow for favorable
medications directly into the vitreous humor of absorption. Additionally, the bioavailability can
the posterior eye. The clearance of medications be increased for drugs which would otherwise
given intravitreally is often magnitudes slower undergo significant presystemic metabolism in
than that for drugs absorbed into the aqueous the liver if given orally.
humor, resulting in half-lives of days as opposed The nasal cavity (Fig. 1.7) is comprised of
to hours. The smaller the particles injected into three main areas, the vestibule and nasal valve
the vitreous humor, the longer the residence time area, the respiratory area, and the olfactory area,
for the particles. For example, Sakurai et al. found each of which is divided into two halves by the
that 50 nm polymeric nanoparticles have nearly nasal septum (Clerico et al. 2003). The nasal
twice the half-life (10.1 days) of similar 2 μm valve area within the vestibule is the narrowest
particles (5.4 days) (Sakurai et al. 2001) when portion of the nasal cavity and is responsible for
administered by intravitreal injection in rabbits. the majority of its airway resistance. The respira-
Due to the relatively long half-life for tory area, posterior to the nasal vestibule, is
medications given intravitreally, dosing regimens comprised of three turbinates. The inferior, mid-
can be extended to monthly and even quarterly dle, and superior turbinates function to produce
administration for some medications (Kuar and turbulent airflow within the nasal cavity. The
Kakkar 2014). Intravitreal injection administra- airflow within the nasal cavity is designed to filter
tion is more technically difficult than topical and condition the air before it reaches the later
delivery to the eye and, therefore, requires the stages of the respiratory system (Thomas 2008).
need of healthcare professionals. They also intro- The olfactory region is located in the uppermost
duce additional risks compared to topical therapy portion of the nasal cavity and is responsible for
such as retinal detachment, which may be irre- our sense of smell. The region is comprised of
versible (Meyer et al. 2011). Intravitreal inserts olfactory neuroepithelium, the only place where
are designed to further improve the pharmacoki- first-order neurons are in contact with the external
netics by controlling drug release and reduce the environment (Lochhead and Thorne 2012).
number of needed injections. Iluvien®, an Nasal drug delivery has been accomplished
intravitreal implant delivering fluocinolone using several methods. One of the oldest methods
Fig. 1.7 Anatomy of the

nasal cavity and sinuses Lateral
from a lateral (top) and
anterior (bottom) view
(Djupesland 2013) (with
permission)
1. Inferior turbinate
2. Middle turbinate
3. Frontal sinus
4. Maxillary sinus
5. Ethrnoid sinus
6. Sphenoid sinus
7. Nasopharynx
8. Nasal valve area
Anterior view
of delivering liquids to the nasal cavity is the use meter-dosed pump sprays. Meter-dosed pump
of drops. Drops are advantageous as they are sprays accurately deliver volumes between
low-cost and relatively straightforward to manu- 25 and 200 μL. The particle size of the drops
facture (Kublik and Vidgren 1998). However, the from pump sprays is a product of the device,
dose from nasal drops is often difficult to control, patient handling, as well as the formulation,
the larger drop volume results in rapid clearance which varies based on the viscosity and surface
compared to sprays, and complex maneuvers can tension of the product (Dayal et al. 2004). Cur-
be required for proper administration by patients rently a few marketed nasal formulations are also
(Hardy et al. 1985). To overcome the available as powders in the United States includ-
disadvantages of nasal drops, most pharmaceuti- ing Onzetra® Xsail® (sumatriptan) and
cal liquids on the market today are delivered by Baqsimi® (glucagon). Powder drug delivery
provides the highest mass of active ingredients for microgram doses for efficacy and are formulated
a given volume, a limiting factor for nasal drug as aqueous suspensions. For drugs requiring
delivery (Kublik and Vidgren 1998). higher doses, formulation scientists may use
excipients and alter the physical characteristics
of the formulation to solubilize the drug to a
1.5.1 Challenges in Nasal Delivery greater extent. A study whose objective was to
of Poorly Water-Soluble Drugs achieve CNS-targeted delivery of olanzapine, a
drug typically requiring milligram doses for effi-
Systemic absorption of drugs delivered nasally cacy with limited solubility in water, was
primarily takes place in the respiratory region formulated in a nanoemulsion to obtain a concen-
due to the high surface area, vascularization, and tration of 8.5 mg/mL. The formulation, in combi-
airflow restriction (Kublik and Viidgren 1998). nation with the targeted delivery, was effective in
However, the narrow geometry of the nasal showing a pharmacodynamic response when
valve makes it challenging for dosage forms to dosed in rats (Kumar et al. 2008). Like other
deposit in this area. Several studies have shown routes of administration, cyclodextrins can be
that a majority of the droplets from meter-dosed used to increase the solubility of poorly water-
pump sprays deposit in the anterior third of the soluble drugs. Additionally, cyclodextrins can act
nasal cavity, which is mostly comprised of the as permeation enhancers to increase the bioavail-
vestibule and nasal valve area (Suman et al. 1999; ability for poorly permeable drugs (Marttin et al.
Cheng et al. 2001; Djupesland et al. 2006; Shah 1998; Kim et al. 2014). Another approach to
et al. 2014). The area which nasal sprays deposit providing larger doses is using powder delivery
is influenced by the geometry of the emitted formulations. Depending on the bulk density of
plume though few studies have linked plume the powder, quantities up to about 50 mg can be
geometry to in vivo deposition patterns in dosed intranasally (Filipović-Grčić and Hafner
patients. Narrower plume geometries are formed 2008). A challenge to utilizing formulation
by modifying the device or increasing the viscos- parameters to enhance nasal drug delivery is the
ity of the formulation. Narrower plume relatively limited list of inactive ingredients that
geometries result in greater deposition to the pos- have been approved in nasal products. Using new
terior portions of the nasal cavity (Foo et al. formulation technologies that require higher
2007). Additionally, to successfully target CNS quantities and new excipients for the nasal route
drug delivery by intranasal administration, drug of delivery requires toxicity studies to assure
deposition needs to reach the olfactory region, safety of the nasal mucosa (FDA Guidance for
requiring novel device designs (Djupesland Industry 2005). The pH of the solution may be
2013). As the neurons in the neuroepithelium of modulated to affect the solubility and permeabil-
the olfactory region are in direct contact with the ity of the poorly water-soluble drugs. Pujara et al.
external environment, drugs can be directly report the nasal mucosa can withstand buffers
transported from the nose to the brain, bypassing with pH range of 3–10 with minimal signs of
the blood–brain barrier (Dhuria et al. 2010). This damage based on nasal epithelium irritation stud-
can be beneficial for drugs which do not typically ies. Additionally, they found the concentration
cross the blood–brain barrier to therapeutic and type of buffer, including the buffer capacity,
concentrations, as well as for drugs which other- play a role in the safety of the formulation to the
wise would cause high systemic adverse effects. nasal mucosa (Pujara et al. 1995).
Due to the small volume limitations for nasal One of the limiting barriers to the bioavailabil-
drug delivery dosages, delivery of poorly water- ity of drugs delivered nasally is the short resi-
soluble drugs in quantities that are sufficient for a dence time due to mucociliary clearance. The
therapeutic response can be challenging. Many of respiratory epithelium of the nasal cavity is
the commercially available poorly soluble equipped with motile cilia that beat at 1000
corticosteroids used nasally only require strokes per minute (Illum 2003). This results in
a mucus flow rate of 8–100 mm/min in the poste- maximized while systemic exposure and
rior regions of the nasal cavity, which is directed associated adverse effects are minimized. The
towards the nasopharynx where it will be pulmonary route of administration also offers
swallowed (Kublik and Vidgren 1998). To several benefits for systemic delivery of drugs
increase the residence time for nasal absorption including a large absorptive surface area, a thin
of drugs after delivery, formulators add viscosity- epithelial barrier, and low metabolic activity
increasing and mucoadhesive agents to the (Patton et al. 2004).
formulations (Chaturvedi et al. 2011). To permit The respiratory system comprises the upper
effective dosing of the formulation while airways, including the nasopharynx, trachea, and
maintaining an increased residence time, Wang large bronchi, and the respiratory region, includ-
et al. prepared an in situ gelling formulation ing the small bronchioles and alveoli (Groneberg
utilizing deacetylated gellan gum. Curcumin was et al. 2003). It is known that the trans-epithelial
formulated as a microemulsion as it is poorly transport of inhaled compounds will differ signif-
soluble in water. The deacetylated gellan gum icantly among these regions. Transport in the
was incorporated into the aqueous phase of the upper airways is generally restricted by its lower
microemulsion to facilitate the in situ gelling surface area, epithelium thickness, and blood flow
action. When the formulation comes into contact as well as rapid clearance through the mucociliary
with the nasal secretions of the nasal mucosa, it escalator. Accordingly, drugs intended for sys-
turns from a liquid into a gel due to the presence temic delivery need to be targeted to the respira-
of ions in the secretions (Wang et al. 2012). Other tory region where high surface area, thinner
products, like Nasacort® AQ, take advantage of epithelium, and rich vascularization offer superior
thixotropic rheological properties in order to have conditions for drug absorption (Groneberg et al.
a low viscosity during actuation. However, these 2003; Patton and Byron 2007).
products have a higher relative viscosity during Several factors in regard to the formulation,
shelf life and after intranasal administration com- such as particle diameter, shape, density, or elec-
pared to during actuation (Kim 2011). Another trical charge, have been shown to influence where
method of utilizing mucoadhesive excipients in and to what extent aerosolized particles deposit in
the formulation intended for nasal delivery is to the lungs (Crowder et al. 2002; Saini et al. 2007).
produce microspheres of drug within the excipi- Particularly, it has been demonstrated that
ent. For example, carbamazepine has been spray- particles with mass median aerodynamic
dried with chitosan to produce microspheres diameters (MMAD) of 1–3 μm preferentially
which provide high bioavailability in sheep deposit in the deep lungs (Heyder et al. 1986;
when compared to carbamazepine given alone. Carvalho et al. 2011). Particles with MMAD
This could be contributed to the mucoadhesive larger than 5 μm primarily deposit in the upper
ability of chitosan; however, in this case, it may airways and near-bronchial branching points
also be due to the higher dissolution rate obtained where they are rapidly cleared, while particles
when formulated as microspheres (Gavini et al. smaller than 1 μm are, to the most part, not
2006). deposited in the airways but rather exhaled after
inspiration.
Formulations for pulmonary delivery are
1.6 Pulmonary Route restricted not only to the appropriate particle
of Administration size range but also to the use of specific and
very few excipients. Generally, excipients
Pulmonary drug delivery may be aimed at intended for use in pulmonary products need to
treating numerous diseases either locally or sys- be either physiologically compatible with lung
temically. Local therapy of conditions such as tissue in terms of pH, tonicity, and immunogenic
asthma or pulmonary infections is advantageous potential or of endogenous nature in order to
in that drug concentrations at the site of action are avoid airway hyper-responsiveness, spasticity,
or inflammation (Tolman and Williams 2009; with non-ozone-depleting propellants such as

Pilcer and Amighi 2010). hydro-fluoroalkanes (HFAs). While HFAs are
Several formulations of poorly soluble drugs currently still acceptable products on the market,
for pulmonary delivery have been developed and in 2016 the Montreal Protocol which was respon-
reported in the literature; some of which, like sible for the phase out of CFCs incorporated the
several corticosteroids, have even been marketed. Kigali Amendment to encourage the removal of
Formulation approaches employed mainly HFAs in favor of mechanisms that have lower
include solubilization in nonaqueous solvents global warming impact (Panigone et al. 2020).
and particle size reduction into the submicron In MDI formulations, the drug is either dissolved
range. Formulation development is however or suspended in the propellant(s). In the case of
greatly challenged due to the very limited number solution-based formulations, it is imperative that
of acceptable excipients and the fact that these the drug has sufficient solubility to allow thera-
can only be used in small concentrations in order peutic doses to be delivered in a few actuations
to maintain adequate aerosol performance and (Smyth 2003). The use of cosolvents such as
prevent adverse physiological effects (Smyth ethanol oftentimes enables solubilization of satis-
2006; Mogalian and Myrdal 2007). factory amounts of lipophilic drugs in the propel-
lant or propellant mixture of interest. As an
example, beclomethasone dipropionate, a slightly
1.6.1 Challenges in Pulmonary water-soluble corticosteroid used in the treatment
Delivery of Poorly of asthma, is dissolved in the propellant HFA
Water-Soluble Drugs 134a with the help of ethanol in one of its
marketed products (QVAR®). This cosolvent-
In order to generate and deliver an aerosol of based approach might however not be applicable
appropriate size distribution and reproducible for all drugs. Especially, in the case of drugs that
dose to the lungs, different devices such as are very poorly soluble or require a large deliv-
metered dose inhalers (MDIs), nebulizers, and ered dose, great amounts of ethanol might be
dry powder inhalers (DPIs) have to be employed needed. This may be problematic in terms of
(Labiris and Dolovich 2003). Depending on the aerosol performance as it has been demonstrated
delivery device and the properties of the active that increased ethanol concentrations can consid-
pharmaceutical ingredient, inhalation products erably affect aerosol characteristics. A study
will be formulated with different types of evaluating solubility and product performance of
excipients, i.e., to ensure effective aerosolization beclomethasone dipropionate in various blends of
performance, to improve physical or chemical HFA 134a and ethanol showed that with increas-
stability of the API, or in the case of poorly ing ethanol concentrations, the solubility of the
soluble drugs to enhance solubility/dissolution. drug in the propellant was almost linearly
MDIs emit an aerosol driven by a single pro- increased. However, product performance was
pellant or a blend of various propellants upon greatly reduced at ethanol concentrations above
activation of an appropriate valve system. Gener- 10% (w/w) as illustrated by a decrease in the
ally, propellants are subject to strict selection respirable deposition (Fig. 1.8; Gupta et al. 2003).
criteria with the key requirements being: benign Besides its negative impact on MDI aerosoli-
toxicology, suitable boiling point, solvent capac- zation performance, ethanol may also cause some
ity, and density, as well as nonflammability irritation of the lung tissue (Coon et al. 1970).
(Noakes 2002). Since chlorofluorocarbons However, pulmonary tolerance testing with
(CFCs) exhibit all of these desirable propellant 150 μL of a 10% ethanol solution conducted in
characteristics, they had long been the propellants a rat model over 4 days found only limited cellu-
of choice. However, due to their harmful effects lar reactions, including minimal hypertrophy of
on the ozone layer, it is required that goblet cells in the lungs and trachea, minimal
pharmaceutical aerosols have to be reformulated
Fig. 1.8 “Effective 1.4 60

solubility” (product of the
Respirable Deposition (%recovered)

drug solubility multiplied
1.2
by the respirable deposition 50
at a given ethanol level) as a
function of ethanol 1
concentration (% w/w). 40
BDP (%w/w)
Also shown in the graph are BDP Solubility
0.8
the beclomethasone
dipropionate solubility (% Effective Solubility 30
w/w) and respirable 0.6 Respirable Deposition
deposition as a function of
ethanol concentration (% 20
0.4
w/w) (Gupta et al. 2003).
(Reprinted with
10
permission) 0.2
0 0
0 5 10 15 20
EtOH (%w/w)
hyperplasia in the tracheal epithelial, and slight will show enhanced dissolution rates may be
lung inflammation (Montharu et al. 2010). employed.
Alternatively, to solution-based MDI Solubilization and subsequent nebulization of
formulations, suspension-based systems can be poorly water-soluble drugs in the form of non-
employed. Most poorly water-soluble aqueous solutions have been attempted. This
corticosteroids intended for pulmonary delivery approach is however challenged by the generally
have a sufficiently low solubility in the tradition- lower tolerability of the lungs for solvents such as
ally used CFC propellants and HFA propellants to ethanol (Pilcer and Amighi 2010). Cyclosporine
be formulated in the form of suspensions. To A, a highly lyophilic immunosuppressant, was
efficiently disperse drug particles of preferred initially formulated in ethanol, and the efficacy
sizes between 1 and 3 μm in the propellant sys- and pharmacokinetics of this formulation after
tem, surfactants are commonly employed. Specif- nebulization were evaluated in different animal
ically, surfactants will prevent particle models (Dowling et al. 1990; Blot et al. 1995;
agglomeration which is vital as particle-size Mitruka et al. 1998). Even though inhaled
changes will lead to irregularities in the emitted solutions of cyclosporine resulted in high and
dose. Challenges arise from the fact that the num- sustained drug levels in the lung, local irritation
ber of possible surfactants to choose from is due to the use of ethanol as a formulation vehicle
rather limited, with currently only oleic acid, was observed. As a result, further studies have
sorbitan trioleate (Span 85), polysorbate 80, and been conducted using propylene glycol as the
phospholipids being accepted for inhalation formulation vehicle. Only scant data are available
(Pilcer and Amighi 2010). regarding effects of propylene glycol on the respi-
Nebulizers are used to aerosolize aqueous ratory tract after pulmonary administration. How-
solutions or suspensions by means of compressed ever, a recent 28-day repeat dose inhalation
air (jet nebulizers) or a vibrating mesh (ultrasonic toxicity study in rats suggested its safety for inha-
nebulizers). In the case of poorly water-soluble lation therapy as no adverse respiratory or sys-
drugs, solutions may be obtained by employing temic effects of high doses of propylene glycol
different solubilizing agents such nonaqueous were seen (Wang et al. 2007). Still, the high
solvents or cyclodextrins. Otherwise suspensions viscosity of propylene glycol presents a signifi-
of micronized or even nanosized material which cant challenge for effective aerosol generation via
nebulization, generally requiring longer nebuliza- nanosuspensions which, due to their increased
tion times and resulting in lower output rates surface area, exhibit enhanced dissolution rates
(McCallion et al. 1995; Corcoran 2006). (Jacobs and Müller 2002; Tam et al. 2008; Yang
As cyclodextrins have shown to be valuable et al. 2008). Nanoparticles are also more uni-
excipients in the delivery of poorly soluble drugs formly distributed throughout the carrier droplets
via the oral or parenteral route, great interest has generated by nebulization than microparticles,
been created to extend their use into pulmonary thereby promoting more efficient drug distribu-
delivery as well. Several studies have tion throughout the alveoli (Jacobs and Müller
demonstrated that aqueous solutions containing 2002). Despite these advantageous characteristics
inclusion complexes of poorly water-soluble of nanosuspensions, the formulation of stable
drugs and cyclodextrins can be successfully nanoparticulate suspensions remains challenging
nebulized with droplet sizes suitable for deep as surfactants, required for efficient stabilization,
lung delivery (Tewes et al. 2008; Thi et al. are limited in number and concentration when
2009; Yang et al. 2010). Cyclodextrin containing used for inhalation products (see above).
dry powders have also demonstrated suitable Aqueous nanosuspensions may also be dried,
characteristics for effective lung delivery i.e., employing spray-drying or lyophilization, to
(Drumond et al. 2014; Dufour et al. 2015). be used in DPIs. Delivery of drug nanoparticles
Despite their potential benefits in pulmonary by means of DPI can however problematic since
delivery of poorly soluble drugs, little data the particles are too small for efficient lung depo-
regarding their safety via this route of administra- sition and are mostly exhaled after inspiration
tion are available to date. A short-term toxicity (Duret et al. 2012; Abdelbary et al. 2015).
study in mice assessing bronchoalveolar lavage, El-Gendy et al. formulated agglomerated
lung, and kidney histology as well as bronchial nanoparticles of budesonide by wet milling,
responsiveness after inhalation of 2-hydroxypro- nanoclusters, which behave as larger particles
pyl-β-cyclodextrin, randomly aerodynamically, emitting a large fine particle
methylated-β-cyclodextrin, and γ-cyclodextrin fraction (El-Gendy et al. 2012a, b). “Trojan”
concluded that the cyclodextrins tested were gen- micron-sized particles, which are porous
erally nontoxic at dosages between 20 and structures completely composed of either
100 mM (Evrard et al. 2004). Another study nanoparticles or nanoparticles and suitable carrier
conducted in in vitro cell cultures demonstrated materials such as sugars or phospholipids, have
that the effect of cyclodextrins on respiratory cell been suggested as an alternative (Tsapis et al.
metabolic activity and permeability was concen- 2002). As a result of their larger size, they will
tration-dependent and varied with cyclodextrin more efficiently deposit in the deep lungs, where
type (β or γ), chemical derivatization, and degree they then disassociate to release the nanoparticles.
of substitution (Salem et al. 2009). Especially, The in vivo performance of these types of
methylated cyclodextrins, which have a high particles has yet to be evaluated. Furthermore,
affinity for membrane lipids such as phosphati- inhalation of nanoparticles raises considerable
dylcholine and cholesterol, are able to permeate concern regarding potential toxicity. Various
the airway epithelium. Consequently, it cannot be studies that have evaluated the effects of inhaled
excluded that cyclodextrins have the ability to be ultrafine particles have reported inflammatory
absorbed into the systemic circulation after pul- reactions as well as carcinogenicity (Oberdörster
monary administration (Matilainen et al. 2008). 1997; Renwick et al. 2004). However, most stud-
Alternatively to solutions, suspensions of ies were conducted with insoluble, inorganic
poorly water-soluble drugs may be employed for particles such as TiO2. It is likely that soluble or
inhalation therapy via nebulization. Considerable biodegradable excipients as principally employed
research has focused on the use of for pulmonary drug delivery may behave
differently. It is assumed that inorganic Method Capsule 1

nanoparticles’ high surface area leads to oxidative
Development of a Self-Emulsifying Formula-
stress and calcium changes in alveolar
tion that Reduces the Food Effect for
macrophages and epithelial cells, thereby
Torcetrapib
activating cells for inflammation (Renwick et al.
Based on the method reported by Perlman
2001; Donaldson 2001). Inflammation as well as
et al. (2008)
inflammatory cell-derived oxidants may, in turn,
Objective
induce and stimulate neoplastic transformation of
• To achieve a lower food effect and a higher
epithelial cells (Driscoll 1997).
dose of torcetrapib per unit through the use
of self-emulsifying formulation.
Equipment and Reagents
1.7 Summary • Torcetrapib (C log P 7.45; non-ionizable)
• Miglyol 812 BP (medium-chain
The increasing percentage of NCEs characterized triglycerides)
by poor aqueous solubility presents various • Triacetin (triacetyl glycerin)
challenges in the development of effective drug • Polysorbate 80
delivery systems. Oral delivery of poorly soluble • Capmul MCM (medium-chain mono- and
drugs is often characterized by low bioavailability diglycerides)
as well as increased likelihood for food effects. • Glass vessel
With the advances in various solubilization and • Gelatin softgel shells
particle-size reduction technologies, the obstacles • Rotary-die encapsulation machine
faced may however be addressed and enable suc- Method
cessful delivery of a wide range of drugs via oral • Excipients are added to glass vessel (semi-
delivery. solid excipient Capmul MCM melted prior
One of the main challenges in the delivery of to addition).
poorly soluble drugs via the pulmonary, nasal, • Mixture is stirred until it is homogeneous.
and parenteral routes is the limited number of • Then the desired amount of the drug is
approved excipients. The development of novel added.
excipients may offer new opportunities and • The resulting mixture is stirred at ambient
enable more effective delivery of poorly soluble temperature with occasional scraping of the
drugs. However, the time and costs involved in a vessel walls until a solution is obtained.
complete toxicological evaluation of a novel • Softgels are manufactured on a rotary-die
excipient as well as the possibility of regulatory machine by encapsulation using a shell
delays or even rejection may discourage pharma- prepared from gelatin, glycerin, and water.
ceutical companies from developing new Results
excipients. • The mean droplet size of the emulsion
Overall, significant achievements have already formed by mixing the formulation and
been made in the formulation of poorly soluble water at a ratio of 1:100 by five times gentle
drugs. Several approaches aimed at enhancing inversion was determined to be 257 nm.
solubility/dissolution such as nanosuspensions • Softgel capsules stored for 6 weeks at 5 C/
or cyclodextrin inclusion complexes have shown 75% RH, 30 C/60% RH, and 40 C/75%
promising results in terms of enhanced in vivo RH showed no change in potency as
performance. Still, concerns regarding the safety analyzed by HPLC. Also, no sign of crys-
of certain excipients or delivery systems remain. tallization in the fill under any condition
Therefore, further studies need to be conducted in based on microscopic examination was
the future to make poorly soluble drugs available found.
via various routes of delivery.
• The use of the lipophilic, GRAS cosolvent • Nanosuspension remains physically and
triacetin in the formulation allowed a two- chemically stable for 2 months at room
fold increase in the dose per capsule as temperature.
compared to the formulation used in early • Polyvinyl alcohol was shown to be the best
clinical trials where the drug was dissolved at stabilizing particles from creaming and
in Miglyol 812 BP only. sedimentation as well as resulted in the
• Pharmacokinetic studies in dogs smallest particle size after milling.
demonstrated that the food effect seen • Draize irritation study in rabbits showed no
with Miglyol softgels was reduced from difference to commercial Restasis®
fivefold to threefold with the Miglyol/ formulation.
triacetin/polysorbate 80/Capmul MCM • Schirmer tear test showed lower irritation
formulation. from the nanosuspension compared to com-
mercial formulation.
Method Capsule 2
Method Capsule 3
Development of a Novel Ophthalmic Cyclo-
sporine A-Loaded Nanosuspension Using Development of In Situ Gelling Microemulsion
Top-Down Media Milling Methods for Nose-to-Brain Delivery
Based on the method reported by Kim et al.
(2011) Based on the method reported by Wang et al.
Objective (2012)
• To develop a stable cyclosporine A Objective
nanosuspension which can be delivered to • To develop a curcumin microemulsion for-
the eye and cause low ocular irritation. mulation for nasal delivery which is liquid
Equipment and Reagents when administered but turns to a gel when
• Cyclosporine A in contact with the nasal mucosa.
• Polyvinyl alcohol Equipment and Reagents
• Polyvinyl pyrrolidone • Curcumin
• Hydroxyethyl cellulose • Capryol 90
• Hydroxypropyl cellulose • Solutol HS15
• Hypromellose • Transcutol HP
• 300 μm zirconia beads • Deacetylated gellan gum (DGG)
• Water • Deionized water
• High shear homogenizer • Glass vessels
• Bead mill Method
Method • Capryol 90, Transcutol HP, and Solutol
• About 2.5 g cyclosporine A was added to HS15 were added to vial and stirred.
500 mL water. • Curcumin was dissolved in above solution.
• 5 g surface-stabilizing polymer is added to • In a separate glass vessel, DGG is added to
the mixture. deionized water and heated to 90 C with
• Mixture is homogenized at 7000 rpm for stirring and then cooled to -4 C.
30 minutes and then transferred into the • To the curcumin solution, DGG solution is
bead mill. added dropwise under stirring.
• Mixture is milled with zirconia beads at Results
1000 rpm for 2 hours. • After the addition of DGG, the formulation
Results remains a liquid. However, once added to
• Formation of a nanosuspension with a min- artificial nasal fluid, it forms a gel.
imum particle size of about 530 μm.
• The average droplet size for the resulting applied to a cryogenic solid substrate
curcumin-loaded microemulsion was 54.1 cooled to 70 C.
1.66 nm. • The resultant frozen solids are collected and
• In vitro release studies determined about lyophilized.
52% of curcumin was released from the Results
microemulsion within 5 hours and 81% • Particle-size distributions of lyophilized
released within 24 hours. ITZ powder redispersed in water showed a
• When dosed intranasally to rats, the abso- narrow size range with a D50 and D90
lute bioavailability was higher than seen for (diameter at which the cumulative sample
oral delivery. Brain tissue concentrations volume is under 50 and 90%, respectively)
showed nearly fourfold higher AUC in the of 230 and 540 nm, respectively.
brain after intranasal administration and • ITZ nanoparticles were found to be amor-
was consistent with direct targeting of the phous as indicated by the absence of the
brain from the nose. characteristic crystalline peaks of ITZ and
mannitol.
Method Capsule 4 • ITZ nanoparticles produced supersaturation
levels 27 times the crystalline solubility
High Bioavailability from Nebulized
upon dissolution in simulated lung fluid.
Itraconazole Nanoparticle Dispersions with
• The colloidal dispersion obtained after
Biocompatible Stabilizers
redispersion of the powder in water
Based on the method reported by Yang et al.
demonstrated optimal aerodynamic
(2008)
properties with a fine particle fraction of
Objective
66.96% and a mass median aerodynamic
• To develop an itraconazole nanoparticle
diameter of nebulized droplets of 2.38 μm.
dispersion for pulmonary delivery by nebu-
• An in vivo single-dose 24 h pharmacoki-
lization that does not require the use of
netics study of the nebulized colloidal dis-
synthetic polymers and surfactants to
persion demonstrated substantial lung
achieve high supersaturation values
deposition and systemic absorption with
in vitro and high bioavailability in in vivo.
blood levels reaching a peak of 1.6 μg/mL
serum in 2 h.
• Itraconazole (ITZ)
• Mannitol
Method Capsule 5
• Lecithin
• 1,4-Dioxane Amorphous Cyclosporine Nanodispersions for
• Ultra-rapid freezing (URF) apparatus Enhanced Pulmonary Deposition and
• Lyophilizer Dissolution
Method Based on the method reported by Tam et al.
• Lecithin (118 mg) is dissolved in a mixture (2008)
of 1,4-dioxane and purified water (65/35, Objective
v/v) cosolvent system (200 mL) using a • To nebulize stable amorphous nanoparticle
magnetic stirrer. dispersions of cyclosporine A to achieve
• ITZ (588 mg) and mannitol (294 mg) are high fine particle fractions and high levels
subsequently dissolved in the mixture; this of absorption into the lung epithelium.
provides a dissolved solids ratio of Equipment and Reagents
ITZ/mannitol/lecithin of 1:0.5:0.2 by • Cyclosporine A
weight. • Polysorbate 80
• The solution is rapidly frozen using the • Liquid nitrogen
URF apparatus in which the solution is • Methanol
• Glass vessel Equipment and Reagents

• Temperature-controlled water bath • Budesonide
• Rotary evaporator • Soy lecithin
Method • Span 85
• 15 g of methanol containing 3.2% w/w • Tyloxapol
cyclosporine A is injected into 50 g • Cetyl alcohol
deionized water containing an appropriate • Ultra-Turrax
amount of polysorbate 80 and maintained at • High-pressure homogenizer
3 C by means of a temperature-controlled Method
water bath. • The surfactants are dissolved or dispersed
• Methanol is then separated from the aque- in warm (~40 C) bi-distilled water by using
ous dispersion via vacuum distillation. an Ultra-Turrax.
• If the dry powder is desired, the aqueous • Budesonide is then dispersed in the aque-
dispersion can be frozen dropwise into liq- ous surfactant solution/dispersion by using
uid nitrogen and then be lyophilized. the Ultra-Turrax for 1 min at 9,500 rpm.
Results • The obtained premix is homogenized by
• Static light scattering results demonstrated using a Micron LAB 40 homogenizer at
a cyclosporine-to-polysorbate 80 ratio of 1: two cycles at 150 bar and two cycles at
0.1 produced particles with an average 500 bar as a kind of premilling and then
diameter of 300 nm. 20 homogenization cycles at 1,500 bar to
• The absence of characteristic crystalline obtain the final product.
cyclosporine A peaks in XRD patterns Results
indicated a primarily amorphous character. • A combination of lecithin (0.5%, w/w) and
• Dissolution of the aqueous cyclosporine A tyloxapol (0.2%, w/w) proved to be the
dispersion produced supersaturation values most suitable to stabilize budesonide (1%,
18 times the aqueous equilibrium solubility w/w) nanosuspension.
of the drug. • The mean particle size of this
• The sizes of the aerosolized aqueous nanosuspension was 500 nm as analyzed
droplets (1–4 μm) obtained by nebulization by photon correlation spectroscopy.
were found to be optimal for deep lung • The scale-up of the formulation from 40 to
deposition. 300 mL was successful with similar size
• Nebulization of the dispersion to mice pro- distributions obtained for both batch sizes.
duced therapeutic lung levels and systemic • The mean particle size as determined by
concentrations below toxic limits. photon correlation spectroscopy did not
change after nebulization of the
Method Capsule 6 nanosuspension indicating suitability for
pulmonary delivery.
Production and Characterization of a
• Nanosuspension stored at room tempera-
Budesonide Nanosuspension for Pulmonary
ture was stable in terms of size distribution
Administration
over 1 year period.
Based on the method reported by Jacobs and
Müller (2002)
Method Capsule 7
Objective
• To develop a budesonide nanosuspension Trojan Particles: Large Porous Carriers of
by high-pressure homogenization and to Nanoparticles for Drug Delivery
investigate the aerosolization properties of Based on the method reported by Tsapis et al.
this nanosuspension. (2002)
Objective ultimate physical dimension of the spray-

• To combine the drug release and delivery dried LPNP.
potential of nanoparticle (NP) systems with • In all cases, the LPNPs had a solid deform-
the ease of flow, processing, and aerosoli- able shell, consisting of several layers of
zation potential of large porous particle NPs, and a wrinkled structure indicative of
(LPP) systems by spray-drying solutions a low relative density, making their aerody-
of NPs into large porous NP (LPNP) namic properties highly favorable.
aggregates. • Additional control over LPNPs physical
Equipment and Reagents characteristics is achieved by adding other
• 1,2-Dipalmitoyl-sn-glycero-3- components to the spray-dried solutions,
phosphocholine (DPPC) including sugars, lipids (DPPC, DMPE),
• 1,2-Dimyristoyl-sn-glycero-3- polymers (hydroxypropyl cellulose), and
phosphoethanolamine (DMPE) proteins (BSA).
• Lactose monohydrate
• Hydroxypropyl cellulose
• Bovine serum albumin (BSA)
References
• Ethanol
• Aqueous suspensions of surfactant-free car- Abdelbary AA, Al-mahallawi AM, Abdelrahim ME, Ali
boxylate-modified polystyrene NP (PS-NP) AM. Preparation, optimization, and in vitro simulated
or aqueous suspensions (pH 9) of Nyacol inhalation delivery of carvedilol nanoparticles loaded
9,950 colloidal silica NP on a coarse carrier intended for pulmonary administra-
tion. Int J Nanomedicine. 2015;10:6339–53.
• Spray-dryer Ahmed I, Patton TF. Importance of the noncorneal absorp-
Method tion route in topical ophthalmic drug delivery. IOVS.
• Solutions of NPs were prepared by mixing 1985;26(4):584–7.
ethanol and water (70:30, v/v) to the Akers MJ. Parenteral preparations, Remington: the science
and practice of pharmacy. Baltimore, MD: Lippincott
desired w/v fraction. Williams & Wilkins; 2014.
• If desired, additional excipients may be Alqawlaq S, Huzil JT, Ivanova MV, Foldvari
included, e.g., sugars, lipids such as DPPC M. Challenges in neuroprotective nanomedicine devel-
and DMPE, polymers, or proteins. opment: progress towards noninvasive gene therapy of
glaucoma. Nanomedicine. 2012;7(7):1067–83.
• The resulting mixture is spray-dried under Amidon GL, Lennernäs H, Shah VP, Crison JR. A theo-
the following conditions: the inlet tempera- retical basis for a biopharmaceutic drug classification:
ture is fixed at 110 C, the outlet tempera- the correlation of in vitro drug product dissolution and
ture is about 46 C, a V24 rotary atomizer in vivo bioavailability. Pharm Res. 1995;12(3):
413–20.
spinning at 20,000 rpm is used, and the feed Amin K, Dannenfelser RM. In vitro hemolysis: guidance
rate of the solution is 70 mL/min, the drying for the pharmaceutical scientist. J Pharm Sci. 2006;95
air flow rate is 98 kg/h. (6):1173–6.
• Spray-dried particles are collected with a Anby MU, Williams HD, Feeney O, Edwards GA,
Benameur H, Pouton CW, et al. Non-linear increases
6 in. cyclone. in Danazol exposure with dose in older vs. younger
Results Beagle dogs: The potential role of differences in bile
• The chemical nature of the NPs seemed to salt concentration, thermodynamic activity, and formu-
be of little importance, since LPNPs were lation digestion. Pharm Res. 2014;31(6):1536–52.
Benet LZ. Predicting drug disposition via application of a
also successfully produced with colloidal biopharmaceutics drug disposition classification sys-
silica NPs instead of PS-NPs. tem. Basic Clin Pharmacol Toxicol. 2010;106(3):
• The formation of LPNPs appears to be gen- 162–7.
erally independent of the size of the NPs Benet L, et al. Predicting drug absorption and the effects of
food on oral bioavailability. Bulletin Technique
(25 nm and 170 nm NPs were tested), Gattefosse. 2006;99:9–16.
provided they are much smaller than the
Bhalla S. Parenteral drug delivery. In: Desai A, Lee M, and the particular relevance to dry powder inhaler
editors. Gibaldi’s drug delivery systems in pharmaceu- formulation and development. Pharm Res. 2002;19
tical care. Bethesda, MD: ASHP; 2007. (3):239–45.
Bittner B, Mountfield RJ. Intravenous administration of Custodio JM, Wu C-Y, Benet LZ. Predicting drug dispo-
poorly soluble new drug entities in early drug discov- sition, absorption/elimination/transporter interplay and
ery: the potential impact of formulation on pharmaco- the role of food on drug absorption. Adv Drug Deliv
kinetic parameters. Curr Opin Drug Discov Devel. Rev. 2008;60(6):717–33.
2002;5(1):59–71. Dayal P, Shaik MS, Singh M. Evaluation of different
Blot F, Tavakoli R, Sellam S, Epardeau B, Faurisson F, parameters that affect droplet-size distribution from
Bernard N, et al. Nebulized cyclosporine for preven- nasal sprays using the Malvern Spraytec®. J Pharm
tion of acute pulmonary allograft rejection in the rat: Sci. 2004;93(7):1725–42.
pharmacokinetic and histologic study. J Heart Lung Dhuria SV, Hanson LR, Frey WH. Intranasal delivery to
Transplant. 1995;14(6 Pt 1):1162–72. the central nervous system: Mechanisms and experi-
Boyd BJ, et al. Successful oral delivery of poorly water- mental considerations. J Pharm Sci. 2010;99(4):
soluble drugs both depends on the intraluminal behav- 1654–73.
ior of drugs and of appropriate advanced drug delivery Ding S, Tien WL, Olejnik O. Nonirritating emulsions for
systems. Eur J Pharm Sci. 2019;137:104967. sensitive tissue [Internet]. US5474979 A, 1995
Bracq E, Lahiani-Skiba M, Guerbet M. Ethical Djupesland PG. Nasal drug delivery devices:
observations on the choice of parenteral solvents. characteristics and performance in a clinical perspec-
Drug Dev Ind Pharm. 2008;34(12):1306–10. tive—a review. Drug Deliv Transl Res. 2013;3(1):
Brewster ME, Loftsson T. Cyclodextrins as pharmaceuti- 42–62.
cal solubilizers. Adv Drug Deliv Rev. 2007;59(7): Djupesland PG, Skretting A, Winderen M, Holand
645–66. T. Breath actuated device improves delivery to target
Calvo P, Alonso MJ, Vila-Jato JL, Robinson JR. Improved sites beyond the nasal valve. Laryngoscope. 2006;116
ocular bioavailability of indomethacin by novel ocular (3):466–72.
drug carriers. J Pharm Pharmacol. 1996;48(11): Donaldson K. Ultrafine particles. Occup Environ Med.
1147–52. 2001;58(3):211–6.
Carvalho TC, Peters JI, Williams RO. Influence of particle Dowling RD, Zenati M, Burckart GJ, Yousem SA,
size on regional lung deposition – what evidence is Schaper M, Simmons RL, et al. Aerosolized cyclospor-
there? Int J Pharm. 2011 Mar 15;406(1-2):1–10. ine as single-agent immunotherapy in canine lung
Charman WN, Porter CJ, Mithani S, Dressman allografts. Surgery. 1990;108(2):198–204.
JB. Physiochemical and physiological mechanisms Dressman JB, Vertzoni M, Goumas K, Reppas
for the effects of food on drug absorption: the role of C. Estimating drug solubility in the gastrointestinal
lipids and pH. J Pharm Sci. 1997;86(3):269–82. tract. Adv Drug Deliv Rev. 2007;59(7):591–602.
Chaturvedi M, Kumar M, Pathak K. A review on Driscoll K. Effects of particle exposure and particle-
mucoadhesive polymer used in nasal drug delivery elicited inflammatory cells on mutation in rat alveolar
system. J Adv Pharm Technol Res. 2011;2(4):215–22. epithelial cells. Carcinogenesis. 1997;18(2):423–43.
Cheng YS, Holmes TD, Gao J, Guilmette RA, Li S, Drumond N, Couto AS, Costa A, Cabral-Marques
Surakitbanharn Y, et al. Characterization of nasal H. Study of aerodynamic and release properties of
spray pumps and deposition pattern in a replica of the inhaled particles containing cyclodextrins. J Incl Phe-
human nasal airway. J Aerosol Med. 2001;14(2): nom Macrocycl Chem. 2014;80(1-2):25–30.
267–80. Dufour G, Bigazzi W, Wong N, Boschini F, de Tullio P,
Chowhan M, Lang J, Missel P. Ophthalmic Piel G, et al. Interest of cyclodextrins in spray-dried
preparations. In: Allen Jr LV, Adejare A, Desselle microparticles formulation for sustained pulmonary
SP, Felton LA, editors. Remington: the science and delivery of budesonide. Int J Pharm. 2015;495(2):
practice of pharmacy. 22nd ed. London; Philadelphia: 869–78.
Pharmaceutical Press; 2012. Duret C, Wauthoz N, Sebti T, Vanderbist F, Amighi
Clerico D, To W, Lanza D. Anatomy of the Human Nasal K. New inhalation-optimized itraconazole
Passages. In: Handbook of Olfaction and Gustation nanoparticle-based dry powders for the treatment of
[Internet]. New York: CRC Press; 2003. p. 1–16. invasive pulmonary aspergillosis. Int J Nanomedicine.
Coon RA, Jones RA, Jenkins Jr LJ, Siegel J. Animal 2012;7:5475–89.
inhalation studies on ammonia, ethylene glycol, form- Edwards A, Prausnitz MR. Fiber matrix model of sclera
aldehyde, dimethylamine, and ethanol. Toxicol Appl and corneal stroma for drug delivery to the eye. AIChE
Pharmacol. 1970;16(3):646–655. J. 1998;44(1):214–25.
Corcoran TE. Inhaled delivery of aerosolized cyclospor- Egger-Heigold B. The effect of excipients on pharmacoki-
ine. Adv Drug Deliv Rev. 2006;58(9–10):1119–27. netic parameters of parenteral drugs. PhD Thesis, Uni-
Crowder TM, Rosati JA, Schroeter JD, Hickey AJ, versity of Basel, Faculty of Science, 2005
Martonen TB. Fundamental effects of particle mor-
phology on lung delivery: predictions of Stokes law
El-Gendy N, Selvam P, Soni P, Berkland C. Development 134a-based pressurized metered dose inhalers. J Aero-
of budesonide nanocluster dry powder aerosols: sol Med. 2003;16(2):167–74.
Processing. J Pharm Sci. 2012a;101(9):3425–33. Hardy JG, Lee SW, Wilson CG. Intranasal drug delivery
El-Gendy N, Huang S, Selvam P, Soni P, Berkland by spray and drops. J Pharm Pharmacol. 1985;37(5):
C. Development of budesonide nanocluster dry powder 294–7.
aerosols: Formulation and stability. J Pharm Sci. Heyder J, Gebhart J, Rudolf G, Schiller C, Stahlhofen
2012b;101(9):3445–55. W. Deposition of particles in the human respiratory
Ellis AG, Crinis NA, Webster LK. Inhibition of etoposide tract in the size range 0.005–15 μm. J Aerosol Sci.
elimination in the isolated perfused rat liver by 1986;17(5):811–25.
Cremophor EL and Tween 80. Cancer Chemother Hochman JH, Chiba M, Nishime J, Yamazaki M, Lin
Pharmacol. 1996;38(1):81–7. JH. Influence of P-glycoprotein on the transport and
Engels FK, Mathot RA, Verweij J. Alternative drug metabolism of indinavir in Caco-2 cells expressing
formulations of docetaxel: a review. Anticancer cytochrome P-450 3A4. J Pharmacol Exp Ther.
Drugs. 2007;18(2):95–103. 2000;292(1):310–8.
Evrard B, Bertholet P, Gueders M, Flament M-P, Piel G, Hofmann AF, Mysels KJ. Bile salts as biological
Delattre L, et al. Cyclodextrins as a potential carrier in surfactants. Colloid Surface. 1987;30(1):145–73.
drug nebulization. J Control Release. 2004;96(3): Illum L. Nasal drug delivery—possibilities, problems and
403–10. solutions. J Control Release. 2003;87(1–3):187–98.
FDA Guidance for Industry: Nonclinical studies for the Itin C, et al. Prolonged oral transmucosal delivery of
safety evaluation of pharmaceutical excipients. Food highly lipophilic drug cannabidiol. Int J Pharm.
and Drug Administration, 2005 2020;581:119276.
Filipović-Grčić, Hafner. Nasal powder drug delivery. In: Jacobs C, Müller RH. Production and characterization of a
Gad SC, editor. Pharmaceutical Manufacturing Hand- budesonide nanosuspension for pulmonary administra-
book: Production and Processes. 1 edition. Hoboken, tion. Pharm Res. 2002;19(2):189–94.
N.J: Wiley-Interscience, 2008. p. 668–699 Jain KK. Drug delivery systems – an overview. In: Jain
Fleisher D, Li C, Zhou Y, Pao LH, Karim A. Drug, meal KK, editor. Drug delivery systems. Totowa, NJ:
and formulation interactions influencing drug absorp- Humana Press; 2008.
tion after oral administration clinical implications. Clin Jansook P, Stefánsson E, Thorsteinsdóttir M, Sigurdsson
Pharmacokinet. 1999;36(3):233–54. BB, Kristjánsdóttir SS, Bas JF, et al. Cyclodextrin
Foo MY, Cheng Y-S, Su W-C, Donovan MD. The Influ- solubilization of carbonic anhydrase inhibitor drugs:
ence of Spray Properties on Intranasal Deposition. J formulation of dorzolamide eye drop microparticle
Aerosol Med. 2007;20(4):495–508. suspension. Eur J Pharm Biopharm. 2010;76(2):
Foster JB, Lee WB. The tear film. In: Ocular surface 208–14.
disease: cornea, conjunctiva and tear film: Expert Jara MO, et al. Amorphous solid dispersions and the
Consult – Online and Print, 1st edition. London: contribution of nanoparticles to in vitro dissolution
Saunders, 2013. p. 17–21 and in vivo testing: niclosamide as a case study.
Friedman NJ, Kaiser PK. Ocular anatomy, physiology, Pharmaceutics. 2021;13(1):97.
and embryology. In: Essentials of ophthalmology. Jinno J, Kamada N, Miyake M, Yamada K, Mukai T,
Elsevier; 2007. p. 1–17. Odomi M, et al. Effect of particle size reduction on
Gabor F, Fillafer C, Neutsch L, Ratzinger G, Wirth dissolution and oral absorption of a poorly water-
M. Improving oral delivery. In: Schaefer-Korting M, soluble drug, cilostazol, in beagle dogs. J Control
editor. Drug delivery. Berlin: Springer; 2010. Release. 2006;111(1–2):56–64.
Gavini E, Hegge AB, Rassu G, Sanna V, Testa C, Johnson JLH, He Y, Yalkowsky SH. Prediction of
Pirisino G, et al. Nasal administration of Carbamaze- precipitation-induced phlebitis: a statistical validation
pine using chitosan microspheres: In vitro/in vivo stud- of an in vitro model. J Pharm Sci. 2003;92(8):1574–81.
ies. Int J Pharm. 2006;307(1):9–15. Kane FE, Burdan J, Cutino A, Green KE. IluvienTM: a
Green M, et al. Abraxane®, a novel Cremophor®-free, new sustained delivery technology for posterior eye
albumin-bound particle form of paclitaxel for the treat- disease. Expert Opin Drug Deliv. 2008;5(9):1039–46.
ment of advanced non-small-cell lung cancer. Ann Kaur IP, Kakkar S. Nanotherapy for posterior eye diseases.
Oncol. 2006;17(8):1263–8. J Control Release. 2014;193:100–12.
Groneberg DA, Witt C, Wagner U, Chung KF, Fischer Kim S-I. Aqueous-based pharmaceutical composition
A. Fundamentals of pulmonary drug delivery. Respir [Internet]. US7977045 B2, 2011
Med. 2003;97(4):382–7. Kim H, Robinson MR, Lizak MJ, Tansey G, Lutz RJ,
Gu C-H, Li H, Levons J, Lentz K, Gandhi RB, Yuan P, et al. Controlled drug release from an ocular
Raghavan K, et al. Predicting effect of food on extent implant: an evaluation using dynamic three-
of drug absorption based on physicochemical dimensional magnetic resonance imaging. Invest
properties. Pharm Res. 2007;24(6):1118–30. Ophthalmol Vis Sci. 2004;45(8):2722–31.
Gupta A, Stein SW, Myrdal PB. Balancing ethanol Kim JH, Jang SW, Han SD, Hwang HD, Choi H-G.
cosolvent concentration with product performance in Development of a novel ophthalmic ciclosporin
A-loaded nanosuspension using top-down media mill- for ophthalmic delivery of pilocarpine nitrate. Int J
ing methods. Pharmazie. 2011;66(7):491–5. Pharm. 2013;455(1–2):75–84.
Kim J-E, Cho H-J, Kim D-D. Budesonide/cyclodextrin Lipinski CA. Drug-like properties and the causes of poor
complex-loaded lyophilized microparticles for intrana- solubility and poor permeability. J Pharmacol Toxicol
sal application. Drug Dev Ind Pharm. 2014;40(6): Methods. 2000;44(1):235–49.
743–8. Liu B, Gordon WP, Richmond W, Groessl T, Tuntland
Klein CE, Chiu Y-L, Awni W, Zhu T, Heuser RS, Doan T, T. Use of solubilizers in preclinical formulations:
et al. The tablet formulation of lopinavir/ritonavir Effect of Cremophor EL on the pharmacokinetic
provides similar bioavailability to the soft-gelatin cap- properties on early discovery compounds. Eur J
sule formulation with less pharmacokinetic variability Pharm Sci. 2015;Forthcoming
and diminished food effect. J Acquir Immune Defic Lochhead JJ, Thorne RG. Intranasal delivery of biologics
Syndr. 2007;44(4):401–10. to the central nervous system. Adv Drug Deliv Rev.
Knight B, Troutman M, Thakker DR. Deconvoluting the 2012;64(7):614–28.
effects of P-glycoprotein on intestinal CYP3A: a major Loftsson T, Hreinsdóttir D, Stefánsson E. Cyclodextrin
challenge. Curr Opin Pharmacol. 2006;6(5):528–32. microparticles for drug delivery to the posterior seg-
Koziolek M, Schneider F, Grimm M, Modeβ C, ment of the eye: aqueous dexamethasone eye drops. J
Seekamp A, Roustom T, et al. Intragastric pH and Pharm Pharmacol. 2007;59(5):629–35.
pressure profiles after intake of the high-caloric, high- Loftsson T, Sigurdsson HH, Konrádsdóttir F,
fat meal as used for food effect studies. J Control Gísladóttir S, Jansook P, Stefánsson E. Topical drug
Release. 2015;220(Part A):71–8. delivery to the posterior segment of the eye: anatomical
Kristinsson J, Fridriksdóttir H, Thórisdóttir S, and physiological considerations. Pharmazie. 2008;63
Sigurdardóttir A, Stefánsson E, Loftsson (3):171–9.
T. Dexamethasone-cyclodextrin-polymer Loftsson T, Jansook P, Stefánsson E. Topical drug deliv-
co-complexes in aqueous eye drops. Aqueous humor ery to the eye: dorzolamide. Acta Ophthalmol. 2012;90
pharmacokinetics in humans. Invest Ophthalmol Vis (7):603–8.
Sci. 1996;37(6):1199–203. Martens TF, Vercauteren D, Forier K, Deschout H,
Krzyzaniak JF, Núñez Fa A, Raymond DM, Yalkowsky Remaut K, Paesen R, et al. Measuring the intravitreal
SH. Lysis of human red blood cells. 4. Comparison of mobility of nanomedicines with single-particle track-
in vitro and in vivo hemolysis data. J Pharm Sci. ing microscopy. Nanomedicine. 2013;8(12):1955–68.
1997;86(11):1215–7. Martinez MN, Amidon GL. A mechanistic approach to
Kublik H, Vidgren MT. Nasal delivery systems and their understanding the factors affecting drug absorption: a
effect on deposition and absorption. Adv Drug Deliv review of fundamentals. J Clin Pharmacol. 2002;42(6):
Rev. 1998;29(1–2):157–77. 620–43.
Kumar M, Misra A, Mishra AK, Mishra P, Pathak Marttin E, Verhoef JC, Merkus FW. Efficacy, safety and
K. Mucoadhesive nanoemulsion-based intranasal mechanism of cyclodextrins as absorption enhancers in
drug delivery system of olanzapine for brain targeting. nasal delivery of peptide and protein drugs. J Drug
J Drug Target. 2008;16(10):806–14. Target. 1998;6(1):17–36.
Labiris NR, Dolovich MB. Pulmonary drug delivery Mathias N, Xu Y, Vig B, Kestur U, Saari A, Crison J, et al.
Part II: the role of inhalant delivery devices and drug Food effect in humans: predicting the risk through
formulations in therapeutic effectiveness of aerosolized in vitro dissolution and in vivo pharmacokinetic
medications. Br J Clin Pharmacol. 2003;56(6):600–12. models. AAPS J. 2015;17(4):988–98.
Lam JKW, et al. Transmucosal drug administration as an Matilainen L, Toropainen T, Vihola H, Hirvonen J,
alternative route in palliative and end-of-life care dur- Järvinen T, Jarho P, et al. In vitro toxicity and perme-
ing the COVID-19 pandemic. Adv Drug Deliv Rev. ation of cyclodextrins in Calu-3 cells. J Control
2020;160:234–43. Release. 2008;126(1):10–6.
Lee VHL, Yang JJ. Oral drug delivery. In: Hillery AM, McCallion ON, Taylor KM, Thomas M, Taylor
Lloyd AW, Swarbrick J, editors. Drug delivery and AJ. Nebulization of fluids of different physicochemical
targeting for pharmacists and pharmaceutical properties with air-jet and ultrasonic nebulizers. Pharm
scientists. London: Taylor & Francis; 2001. Res. 1995;12(11):1682–8.
Lee Y-C, Zocharski PD, Samas B. An intravenous formu- McDonald C, Muzumdar PP. Prevention of precipitation
lation decision tree for discovery compound formula- of phenytoin in an infusion fluid by hydroxypropyl
tion development. Int J Pharm. 2003;253(1–2):111–9. beta-cyclodextrin. J Clin Pharm Ther. 1998;23(3):
Levine RR. Factors affecting gastrointestinal absorption of 235–9.
drugs. Am J Dig Dis. 1970;15(2):171–88. Meyer CH, Michels S, Rodrigues EB, Hager A, Mennel S,
Li P, Zhao L. Developing early formulations: practice and Schmidt JC, et al. Incidence of rhegmatogenous retinal
perspective. Int J Pharm. 2007;341(1–2):1–19. detachments after intravitreal antivascular endothelial
Li J, Wu L, Wu W, Wang B, Wang Z, Xin H, et al. A factor injections. Acta Ophthalmol. 2011;89(1):70–5.
potential carrier based on liquid crystal nanoparticles Mitruka SN, Pham SM, Zeevi A, Li S, Cai J, Burckart GJ,
et al. Aerosol cyclosporine prevents acute allograft
rejection in experimental lung transplantation. J Thorac acuity and decrease macular thickness in diabetic mac-
Cardiovasc Surg. 1998;115(1):28–37. ular oedema. Acta Ophthalmol. 2015;93(7):610–5.
Miyazaki S, Suzuki S, Kawasaki N, Endo K, Takahashi A, Ohnishi M, Sagitani H. The effect of nonionic surfactant
Attwood D. In situ gelling xyloglucan formulations for structure on hemolysis. J Am Oil Chem Soc. 1993;70
sustained release ocular delivery of pilocarpine hydro- (7):679–84.
chloride. Int J Pharm. 2001;229(1–2):29–36. Paine MF, Hart HL, Ludington SS, Haining RL, Rettie
Mogalian E, Myrdal PB. Pharmaceutical solvents for pul- AE, Zeldin DC, et al. The human intestinal cytochrome
monary drug delivery. In: Augustijns P, Brewster M, P450 “PIE” abstract. Methods. 2006;34(5):880–6.
editors. Solvent systems and their selection in Panigone S, et al. Environmental impact of inhalers for
pharmaceutics and Biopharmaceutics. New York: respiratory diseases: decreasing the carbon footprint
Springer; 2007. while preserving patient-tailored treatment. BMJ
Mogalian E, Oliyai R, Stefanidis D, Zia V. Solid disper- Open Respir Res. 2020;7(1):e000571.
sion formulation of an antiviral compound [Internet]. Patel VF, et al. Advances in oral transmucosal drug deliv-
US20140212487 A1, 2014 ery. J Control Release. 2011;153(2):106–16.
Monschke M, Wagner KG. Amorphous solid dispersions Patton JS, Byron PR. Inhaling medicines: delivering drugs
of weak bases with pH-dependent soluble polymers to to the body through the lungs. Nat Rev Drug Discov.
overcome limited bioavailability due to gastric pH 2007;6(1):67–74.
variability – an in-vitro approach. Int J Pharm. Patton JS, Fishburn CS, Weers JG. The lungs as a portal of
2019;564:162–70. entry for systemic drug delivery. Proc Am Thorac Soc.
Montharu J, Le Guellec S, Kittel B, Rabemampianina Y, 2004;1(4):338–44.
Guillemain J, Gauthier F, et al. Evaluation of lung Perlman ME, Murdande SB, Gumkowski MJ, Shah TS,
tolerance of ethanol, propylene glycol, and sorbitan Rodricks CM, Thornton-Manning J, et al. Develop-
monooleate as solvents in medical aerosols. J Aerosol ment of a self-emulsifying formulation that reduces
Med Pulm Drug Deliv. 2010;23(1):41–6. the food effect for torcetrapib. Int J Pharm. 2008;351
Mottu F, Stelling M-J, Rüfenacht DA, Doelker (1–2):15–22.
E. Comparative hemolytic activity of undiluted organic Petursson G, Cole R, Hanna C. Treatment of glaucoma
water-miscible solvents for intravenous and intra- using minidrops of clonidine. Arch Ophthalmol.
arterial injection. PDA J Pharm Sci Technol. 2001;55 1984;102(8):1180–1.
(1):16. Pilcer G, Amighi K. Formulation strategy and use of
Mudra DR, Desino KE, Desai PV. In silico, in vitro and in excipients in pulmonary drug delivery. Int J Pharm.
situ models to assess interplay between CYP3A and 2010;392(1–2):1–19.
P-gp. Curr Drug Metab. 2011;12(8):750–73. Popov A. Mucus-penetrating particles and the role of
Nagai N, Yoshioka C, Mano Y, Tnabe W, Ito Y, ocular mucus as a barrier to micro-and
Okamoto N, et al. A nanoparticle formulation of disul- nanosuspensions. J Ocul Pharmacol Ther. 2020;36(6):
firam prolongs corneal residence time of the drug and 366–75.
reduces intraocular pressure. Exp Eye Res. 2015;132: Pujara CP, Shao Z, Duncan MR, Mitra AK. Effects of
115–23. formulation variables on nasal epithelial cell integrity:
Narazaki R, Sanghvi R, Yalkowsky SH. Estimation of Biochemical evaluations. Int J Pharm. 1995;114(2):
drug precipitation upon dilution of pH-controlled 197–203.
formulations. Mol Pharm. 2007;4(4):550–5. Radi AE, Eissa S. Electrochemistry of cyclodextrin inclu-
Naveh N, Muchtar S, Benita S. Pilocarpine incorporated sion complexes of pharmaceutical compounds. Open
into a submicron emulsion vehicle causes an unexpect- Chem Biomed Meth J. 2010;3:74–85.
edly prolonged ocular hypotensive effect in rabbits. J Reed KW, Yalkowsky SH. Lysis of human red blood cells
Ocul Pharmacol Ther. 1994;10(3):509–20. in the presence of various cosolvents III. The relation-
Nernst W. Theorie der Reaktionsgeschwindigkeit in ship between hemolytic potential and structure. J
heterogenen Systemen. Zeitschrift f Physikalische Parenter Sci Technol. 1987;41(1):37–9.
Chemie. 1904;47:52–5. Renwick LC, Donaldson K, Clouter A. Impairment of
Noakes T. Medical aerosol propellants. J Fluor Chem. alveolar macrophage phagocytosis by ultrafine
2002;118(1–2):35–45. particles. Toxicol Appl Pharmacol. 2001;172(2):
Noyes A, Whitney W. The rate of solution of solid 119–27.
substances in their own solutions. J Am Chem Soc. Renwick LC, Brown D, Clouter A, Donaldson
1897;19:930–4. K. Increased inflammation and altered macrophage
Oberdörster G. Pulmonary carcinogenicity of inhaled chemotactic responses caused by two ultrafine particle
particles and the maximum tolerated dose. Environ types. Occup Environ Med. 2004;61(5):442–7.
Health Perspect. 1997;105(Suppl):1347–55. Saini D, Biris AS, Srirama PK, Mazumder MK. Particle
Ohira A, Hara K, Jóhannesson G, Tanito M, Ásgrímsdóttir size and charge distribution analysis of pharmaceutical
GM, Lund SH, et al. Topical dexamethasone aerosols generated by inhalers. Pharm Dev Technol.
γ-cyclodextrin nanoparticle eye drops increase visual 2007;12(1):35–41.
Sakurai E, Ozeki H, Kunou N, Ogura Y. Effect of Particle results from the pharmaceutical vehicle Cremophor
Size of Polymeric Nanospheres on Intravitreal Kinet- EL. Cancer Res. 1996;56(9):2112–5.
ics. Ophthalmic Res. 2001;33(1):31–6. Sparreboom A, van Zuylen L, Brouwer E, Loos WJ, de
Salem LB, Bosquillon C, Dailey LA, Delattre L, Martin Bruijn P, Gelderblom H, et al. Cremophor
GP, Evrard B, et al. Sparing methylation of beta- EL-mediated alteration of paclitaxel distribution in
cyclodextrin mitigates cytotoxicity and permeability human blood: clinical pharmacokinetic implications.
induction in respiratory epithelial cell layers in vitro. Cancer Res. 1999;59(7):1454–7.
J Control Release. 2009;136(2):110–6. Stella VJ, He Q. Cyclodextrins. Toxicol Pathol. 2008;36
Sastry S, Nyshadham J, Fix J. Recent technological (1):30–42.
advances in oral drug delivery - a review. Pharmaceut Stella V, Rao V, Zannou E, Zia V. Mechanisms of drug
Sci Tech Today. 2000;3(4):138–45. release from cyclodextrin complexes. Adv Drug Deliv
Sauron R, Wilkins M, Jessent V, Dubois A, Maillot C, Rev. 1999;36(1):3–16.
Weil A. Absence of a food effect with a 145 mg Strickley RG. Solubilizing excipients in oral and injectable
nanoparticle fenofibrate tablet formulation. Int J Clin formulations. Pharm Res. 2004;21(2):201–30.
Pharmacol Ther. 2006;44(2):64–70. Sugano K, Okazaki A, Sugimoto S, Tavornvipas S,
Schütte JK, et al. Comparison of the therapeutic effective- Omura A, Mano T. Solubility and dissolution profile
ness of a dantrolene sodium solution and a novel assessment in drug discovery. Drug Metab
nanocrystalline suspension of dantrolene sodium in Pharmacokinet. 2007;22(4):225–54.
malignant hyperthermia normal and susceptible pigs. Suman JD, Laube BL, Dalby R. Comparison of nasal
Eur J Anaesthesiol (EJA). 2011;28(4):256–64. deposition and clearance of aerosol generated by a
Sebag J. Vitreous anatomy and pathology. In: Yanoff M, nebulizer and an aqueous spray pump. Pharm Res.
Duker JS, editors. Ophthalmology: Expert consult: 1999;16(10):1648–52.
Online and Print. 4th edition. Saunders, 2013. p. 687–9 Sunesen VH, Vedelsdal R, Kristensen HG, Christrup L,
Shah SA, Dickens CJ, Ward DJ, Banaszek AA, George C, Müllertz A. Effect of liquid volume and food intake on
Horodnik W. Design of experiments to optimize an the absolute bioavailability of danazol, a poorly soluble
in vitro cast to predict human nasal drug deposition. J drug. Eur J Pharm Sci. 2005;24(4):297–303.
Aerosol Med Pulm Drug Deliv. 2014;27(1):21–9. Szebeni J, Muggia FM, Alving CR. Complement activa-
Shalel S, Streichman S, Marmur A. The mechanism of tion by Cremophor EL as a possible contributor to
hemolysis by surfactants: effect of solution composi- hypersensitivity to paclitaxel: an in vitro study. J Natl
tion. J Colloid Interface Sci. 2002;252(1):66–76. Cancer Inst. 1998;90(4):300–6.
Sharif MS, Qahwaji R, Shahamatnia E, Alzubaidi R, Tam JM, McConville JT, Williams RO, Johnston
Ipson S, Brahma A. An efficient intelligent analysis KP. Amorphous cyclosporin nanodispersions for
system for confocal corneal endothelium images. enhanced pulmonary deposition and dissolution. J
Comput Methods Programs Biomed. 2015;122(3): Pharm Sci. 2008;97(11):4915–33.
421–36. Tamilvanan S, Benita S. The potential of lipid emulsion
Shi Y, Porter W, Merdan T, Li LC. Recent advances in for ocular delivery of lipophilic drugs. Eur J Pharm
intravenous delivery of poorly water-soluble Biopharm. 2004;58(2):357–68.
compounds. Expert Opin Drug Deliv. 2009;6(12): Tamilvanan S, Kumar BA. Influence of acetazolamide
1261–82. loading on the (in vitro) performances of non-
Sigurdsson HH, Stefánsson E, Gudmundsdóttir E, phospholipid-based cationic nanosized emulsion in
Eysteinsson T, Thorsteinsdóttir M, Loftsson comparison with phospholipid-based anionic and
T. Cyclodextrin formulation of dorzolamide and its neutral-charged nanosized emulsions. Drug Dev Ind
distribution in the eye after topical administration. J Pharm. 2011;37(9):1003–15.
Control Release. 2005;102(1):255–62. Tanaka Y, Goto T, Kataoka M, Sakuma S, Yamashita
Singla AK, Garg A, Aggarwal D (2002) Paclitaxel and its S. Impact of luminal fluid volume on the drug absorp-
formulations. Int J Pharm 235(1–2):179–192 tion after oral administration: analysis based on in vivo
Smith DA, Jones BC, Walker DK. Design of drugs involv- drug concentration–time profile in the gastrointestinal
ing the concepts and theories of drug metabolism and tract. J Pharm Sci. 2015;104(9):3120–7.
pharmacokinetics. Med Res Rev. 1996;16(3):243–66. Ten Tije AJ, Verweij J, Loos WJ, Sparreboom
Smyth HDC. The influence of formulation variables on the A. Pharmacological effects of formulation vehicles:
performance of alternative propellant-driven metered implications for cancer chemotherapy. Clin
dose inhalers. Adv Drug Deliv Rev. 2003;55(7): Pharmacokinet. 2003;42(7):665–85.
807–28. Tewes F, Brillault J, Couet W, Olivier J-C. Formulation of
Smyth HDC. Excipients for pulmonary formulations. In: rifampicin-cyclodextrin complexes for lung nebuliza-
Katdare A, Chaubal M, editors. Excipient development tion. J Control Release. 2008;129(2):93–9.
for pharmaceutical, biotechnology, and drug delivery Thi THH, Azaroual N, Flament M-P. Characterization and
systems. CRC Press; 2006. p. 225–44. in vitro evaluation of the formoterol/cyclodextrin com-
Sparreboom A, van Tellingen O, Nooijen WJ, Beijnen plex for pulmonary administration by nebulization. Eur
JH. Nonlinear pharmacokinetics of paclitaxel in mice J Pharm Biopharm. 2009;72(1):214–8.
Thomas C, Ahsan F. Nasal delivery of peptide and Welling PG. Effects of food on drug absorption. Annu Rev
nonpeptide drugs. In: Gad SC, editor. Pharmaceutical Nutr. 1996;16:383–415.
manufacturing handbook: production and processes. Wong J, Brugger A, Khare A, Chaubal M,
1st ed. Hoboken, N.J: Wiley-Interscience; 2008. Papadopoulos P, Rabinow B, et al. Suspensions for
p. 591–650. intravenous (IV) injection: a review of development,
Tolman JA, Williams RO. Advances in the pulmonary preclinical and clinical aspects. Adv Drug Deliv Rev.
delivery of poorly water-soluble drugs: influence of 2008;60(8):939–54.
solubilization on pharmacokinetic properties. Drug Woo JS, Song Y-K, Hong J-Y, Lim S-J, Kim C-K.
Dev Ind Pharm. 2009;36(1):1–30. Reduced food-effect and enhanced bioavailability of
Tsapis N, Bennett D, Jackson B, Weitz DA, Edwards a self-microemulsifying formulation of itraconazole in
DA. Trojan particles: large porous carriers of healthy volunteers. Eur J Pharm Sci. 2008;33(2):
nanoparticles for drug delivery. Proc Natl Acad Sci U 159–65.
S A. 2002;99(19):12001–5. Wu C-Y, Benet LZ. Predicting drug disposition via appli-
Tsuji A, Tamai I, Sasaki K. Intraocular penetration kinetics cation of BCS: transport/absorption/elimination inter-
of prednisolone after subconjunctival injection in play and development of a biopharmaceutics drug
rabbits. Ophthalmic Res. 1988;20(1):31–43. disposition classification system. Pharm Res. 2005;22
Urtti A. Challenges and obstacles of ocular pharmacoki- (1):11–23.
netics and drug delivery. Adv Drug Deliv Rev. Yalkowsky SH, Krzyzaniak JF, Ward GH. Formulation-
2006;58(11):1131–5. related problems associated with intravenous drug
Van Santvliet L, Ludwig A. Determinants of eye drop size. delivery. J Pharm Sci. 1998;87(7):787–96.
Surv Ophthalmol. 2004;49(2):197–213. Yang W, Tam J, Miller DA, Zhou J, McConville JT,
Van Zuylen L, Karlsson MO, Verweij J, Brouwer E, de Johnston KP, et al. High bioavailability from nebulized
Bruijn P, Nooter K, et al. Pharmacokinetic modeling of itraconazole nanoparticle dispersions with biocompati-
paclitaxel encapsulation in Cremophor EL micelles. ble stabilizers. Int J Pharm. 2008;361(1–2):177–88.
Cancer Chemother Pharmacol. 2001;47(4):309–18. Yang W, Chow KT, Lang B, Wiederhold NP, Johnston
Vandervoort J, Ludwig A. Ocular drug delivery: KP, Williams RO. In vitro characterization and phar-
nanomedicine applications. Nanomedicine. 2007;2(1): macokinetics in mice following pulmonary delivery of
11–21. itraconazole as cyclodextrin solubilized solution. Eur J
Verma P, Gupta RN, Jha AK, Pandey R. Development, Pharm Sci. 2010;39(5):336–47.
in vitro and in vivo characterization of Eudragit RL Yasuji T, Kondo H, Sako K. The effect of food on the oral
100 nanoparticles for improved ocular bioavailability bioavailability of drugs: a review of current
of acetazolamide. Drug Deliv. 2013;20(7):269–76. developments and pharmaceutical technologies for
Wang T, Noonberg S, Steigerwalt R, Lynch M, Kovelesky pharmacokinetic control. Ther Deliv. 2011;3(1):
RA, Rodríguez CA, et al. Preclinical safety evaluation 81–90.
of inhaled cyclosporine in propylene glycol. J Aerosol Ying L, Tahara K, Takeuchi H. Drug delivery to the ocular
Med. 2007;20(4):417–28. posterior segment using lipid emulsion via eye drop
Wang S, Chen P, Zhang L, Yang C, Zhai G. Formulation administration: Effect of emulsion formulations and
and evaluation of microemulsion-based in situ surface modification. Int J Pharm. 2013;453(2):
ion-sensitive gelling systems for intranasal administra- 329–35.
tion of curcumin. J Drug Target. 2012;20(10):831–40. Zaslavsky BY, Ossipov NN, Rogozhin SV. Action of
Webster LK, Cosson EJ, Stokes KH, Millward MJ. Effect surface-active substances of biological membranes
of the paclitaxel vehicle, Cremophor EL, on the phar- III. Comparison of hemolytic activity of ionic and
macokinetics of doxorubicin and doxorubicinol in nonionic surfactants. Biochim Biophys Acta.
mice. Br J Cancer. 1996;73(4):522–4. 1978;510(1):151–9.
Weinreb RN, Khaw PT. Primary open-angle glaucoma.
The Lancet. 2004;363(9422):1711–20.
Optimizing the Formulation of Poorly
Water-Soluble Drugs 2
Xiangyu Ma, Daniel Ellenberger, Kevin P. O’Donnell,
Abstract Keywords
With as high as 90% of drugs in the develop- Solubility studies · Solid-state
ment phase and 40% of marketed products characterization · Lead formulations · Active
exhibiting poor aqueous solubility, the ability pharmaceutical ingredients (APIs) · In vitro
to successfully develop a poorly water-soluble dissolution · In vivo dose administration ·
drug has become essential. Gaining a detailed General solubility equation (GSE) · Aqueous
understanding of a compound through suspension · pH-Solubility · Flory–Huggins
preformulation studies can be especially chal- theory
lenging for poorly water-soluble drugs limit-
ing their development. Therefore, this chapter
focuses on the application of preformulation 2.1 Introduction
studies essential in understanding a poorly
water-soluble drug, including solubility stud- Reportedly, 90% of drugs in the development
ies, solid-state characterization of the active phase and 40% of marketed products are poorly
ingredient and formulations, cutting-edge ana- water-soluble (Boyd et al. 2019; Kawabata et al.
lytical technologies for molecular-level 2011; Loftsson and Brewster 2010; Jermain et al.
understandings of the formulations, 2018). Many of these active pharmaceutical
applications of machine and deep learning in ingredients (APIs) are designated BCS class II
preformulation and formulation development, compounds (low solubility and high permeabil-
and in vitro and in vivo testing of the lead ity), making their aqueous solubility the limiting
formulations. factor regarding bioavailability (Fahr and Liu
2007). This has led to numerous novel formula-
tion approaches such as particle engineering,
alterations of the API into a salt form,
amorphization of the compound, the use of
X. Ma · R. O. Williams III (*)
Division of Pharmaceutics, College of Pharmacy, surface-active agents or cosolvents, inclusion of
The University of Texas at Austin, Austin, TX, USA polymeric stabilizers for supersaturation, and the
e-mail: [email protected] generation of solid dispersions/solutions, as well
D. Ellenberger as many other novel techniques. Each of these
DisperSol Technologies, LLC, Georgetown, TX, USA techniques focuses on improving the extent or
K. P. O’Donnell rate at which the drug enters solution in an effort
International Flavors & Fragrances, Pharma Solutions, to increase the bioavailability. During this
Midland, MI, USA

34 X. Ma et al.
formulation process, it is important to perform the 1:11ΔS f ðT m 25Þ

correct studies to ensure that the development is log Sw ¼ 1:00 log PC
1364
proceeding in the desired direction. This includes þ 0:54:
tests designed to gain an understanding of the API
alone, as well as experiments necessary in formu- Analysis of multiple compounds led to a pro-
lation optimization. posed universal value of 56.5 J/mol/K for the
In the following sections, numerous analyti- entropy of melting and a reduction of the
cal and experimental techniques will be equation to:
presented, focusing on understanding and
optimizing poorly water-soluble drugs or log Sw ¼ 1:05 log Pc 0:012t m þ 0:87:
formulations thereof during development. This When applying the equation to a new chemical
includes solubility testing, solid-state character- entity, the true entropy of melting should be
ization studies, in vitro dissolution work, and experimentally determined (i.e., by differential
in vivo dose administration. scanning calorimetry) to be more precise.
More recently, Jain and Yalkowsky (2001)
analyzed 580 compounds and further refined the
2.1.1 Solubility Studies above equation such that:
log Sw ¼ 1:031 log Pc 0:0102t m þ 0:679:
As mentioned, approximately 90% of drugs in
company pipelines are poorly water-soluble, This equation is more commonly referred to as
being designated BCS class II compounds (low the general solubility equation (GSE) and can be
solubility and high permeability), making their used as an empirical estimation of the aqueous
aqueous solubility the limiting factor regarding solubility of a compound.
bioavailability; formulation will thus focus on Baena et al. have employed these equations in
improving the extent or rate at which the drug the solubility prediction of acetanilide
enters solution (Fahr and Liu 2007). Additionally, derivatives. All three equations provided accurate
an understanding of a compound’s solubility is predictions of the aqueous solubility of acetani-
required for pharmacological, toxicological, and lide and phenacetin; however, they were not accu-
pharmacokinetic studies as a solution of the drug rate at estimating the aqueous solubility of
is often required to perform such studies (Mashru acetaminophen (Baena et al. 2004). A similar
et al. 2005; Teijeiro and Briñón 2006). Aqueous study by Venkatesh et al. determined the solubil-
solubility determination has proven extremely ity of cosalane by multiple methods. Indeed,
difficult for poorly water-soluble (PWS) drugs. empirical estimation yielded a difference of
This section describes both experimental and pre- three orders of magnitude compared to experi-
dictive (mathematical and advanced machine- mental determination for the PWS drug
deep learning) methods for determining the aque- (Venkatesh et al. 1996).
ous solubility of a drug. Ran et al. analyzed 1026 nonelectrolytes by
the GSE and compared them to their respective
Solubility Prediction experimentally determined aqueous solubilities.
The solubility of a solid compound may be math- As can be seen in Fig. 2.1, empirical calculation
ematically predicted using the equation devel- correlates well with experimental determination
oped by Yalkowsky and Valvani (1980). Here, for many of the compounds analyzed (Radha
solubility (Sw) is defined in terms of the melting et al. 2010). However, these studies demonstrate
temperature (Tm), octanol–water partition coeffi- that while empirical analysis may provide a good
cient (Pc), and the entropy of melting (ΔSf) of the estimation of solubility and should be employed
substance such that: when limited quantities of drug are available, they
are not accurate for all compounds, and
2 Optimizing the Formulation of Poorly Water-Soluble Drugs 35
Fig. 2.1 Calculated

solubility from the GSE
versus experimentally
determined solubility.
N ¼ 1026. Reproduced
with permission from
Elsevier
experimental testing must be performed to vali- filtration steps are critical as they eliminate poten-
date the calculated results. tial seeds for recrystallization and ensure that no
particulate matter is dissolved upon dilution for
analysis, which may provide inaccurate results.
Teijeiro analyzed the solubility of AZT-Iso, a
2.1.2 Experimental Aqueous
derivative of zidovudine, by direct determination
Solubility Determination
in aqueous suspension using excess drug in 4 mL
of water under constant shaking. Again, filtration
The most widely employed experimental method
was applied following equilibration, and the
for solubility determination is by direct determi-
samples were analyzed by UV absorbance
nation in aqueous suspension. In this method,
(Teijeiro and Briñón 2006).
excess drug is placed in a designated volume of
It is crucial that compounds analyzed by direct
water and held at constant temperature. Samples
determination in aqueous suspension be allowed
are taken periodically, and the drug concentration
to reach equilibrium. The aforementioned studies
is determined analytically (i.e., HPLC analysis).
required only 24 h to achieve equilibration; how-
For example, Seedher et al. analyzed the solu-
ever, for many PWS drugs, this time is insuffi-
bility of seven PWS antidiabetic drugs. In their
cient. Venkatesh found that cosalane required
study, excess drug was placed in 5 mL of water in
48 h to reach equilibrium at 25 C. In contrast,
a sealed container. The suspensions were held at
Shaw et al. found that 5 days were required to
25 C and placed under magnetic stirring for 24 h
attain equilibrium for ibuprofen at 37 C (Shaw
at which point equilibrium was obtained. The
et al. 2005). For a new chemical entity, it is
suspensions were then centrifuged, filtered
recommended that samples be taken every 8 h
through a 0.45-μm membrane, and analyzed by
up to 24 h and then every 24 h after for 4 days.
UV absorbance for drug quantification (Seedher
If the determined concentration has not reached a
and Kanojia 2009). Applying the same method,
plateau, the study may be continued for longer
the researchers also screened numerous
durations of time if necessary.
cosolvents for their ability to enhance solubiliza-
tion. It should be noted that the centrifugation and
36 X. Ma et al.
Solubility Measurement on Amorphous 2019a, b) and naproxen sodium (Stukelj et al.

Drugs and Limited Amounts of Materials 2020a, b) in on-chip method for biologically
Using Single Particle Analysis active substances screening (Svanback et al.
As mentioned earlier, an excessive amount of 2015), were assessed using SPA in combination
materials is often needed for experimentally with imaging analysis.
determining the aqueous solubility. However,
the amount available of those potential drugs is
not always guaranteed, especially in the case
2.1.3 pH-Solubility Profiles
where drug candidates are just after the discovery
stage. Additionally, as most of the crystalline
The pH-solubility profile of a PWS drug can be a
drugs exhibit poor water solubility, amorphous
driving factor in its course of development.
drugs have been dominating in the pharmaceuti-
Compounds such as itraconazole show a marked
cal product development. There is a need to
reduction in solubility upon entry into neutral
directly measure the solubility of amorphous
media creating the need for gastric absorption or
materials. However, amorphous drugs often are
supersaturation in the intestine to enhance
not stable in aqueous solutions, having strong
bioavailability.
tendency to crystallize. Therefore, it is necessary
The direct determination in aqueous suspen-
to develop and utilized new tools to enable the
sion method can be adapted to provide
direct and accurate measurement of amorphous
pH-solubility profiles. For example, Wang et al.
drugs.
placed excess sildenafil citrate in 10 mL of
Recently, researchers have demonstrated that
deionized water in multiple 15-mL vials. The
using single particle analysis (SPA), amorphous
pH values of individual vials were then titrated
solubility of diverse drugs can be directly
to a pH of 3–11. These suspensions were then
measured where only a limited amount is avail-
placed at 37 C for 48 h. Following confirmation
able. Stukelj et al. (2019a, b), for the first time,
that no shift in pH had occurred, samples were
illustrated a direct approach to measure the amor-
filtered through 0.2-μm filters, diluted with
phous solubility through combining SPA with
mobile phase, and analyzed by HPLC (Wang
optics and fluidics. Compared to other conven-
et al. 2008). Table 2.1 shows that, indeed, the
tional approaches, including theoretical estima-
solubility of sildenafil drastically decreases
tion using thermal analysis and standardized
above pH 4. Norfloxacin was analyzed in the
supersaturation and precipitation method, the
same manner; however, in this case a solution of
SPA approach enabled the experimental measure-
NaCl was added to each vial in amounts neces-
ment of amorphous solubility for griseofulvin for
sary to achieve constant ionic strength among
only 0.1 mg material available, and the measured
samples (Ahumada et al. 1993). The direct deter-
solubility was comparable to other methods.
mination in aqueous suspension method was also
More importantly, griseofulvin is a fast-
employed for the pH-solubility profile of haloper-
crystallizing drug, and SPA approach was able
idol free base and the corresponding hydrochlo-
to accurately determine the solubility without
ride and mesylate salts to identify the more
using crystallization inhibitors. Furthermore, var-
soluble species (Li et al. 2005). In this iteration,
ious drugs in different conditions, such as indo-
a single vial of each compound was prepared by
methacin in bio-relevant media (Stukelj et al.
adding excess solids to 5 mL of water. The
Table 2.1 pH-solubility profile of sildenafil. Reproduced with permission from Elsevier
pH
3 4 5 6 7 8 9 10 11
Solubility (mg/ml) 6.965 7.077 2.068 0.114 0.025 0.027 0.040 0.103 0.322
S.D. 0.092 0.047 0.042 0.001 0.001 0.001 0.001 0.001 0.058
suspensions were then titrated with HCl or NaOH 2.1.4 Solubility Prediction Using
solutions to the desired pH and allowed to equili- Machine and Deep Learning
brate at 37 C for 24 h. At 24 h, the pH was
confirmed and an aliquot taken for analysis. The In recent years, artificial intelligences including
pH was then adjusted to the next desired level, the machine and deep learning for data analytics have
vials resealed, and the equilibration and sampling become a promising area of research and have
processes repeated. Indeed, such a process allows applications towards prediction, visual inspec-
identification of the more soluble salt in media at tion, and appearance testing. Over the past few
physiologically relevant pH values. years, due to a combination of improvements in
A modified version of the method has been mathematical algorithms, advancements in com-
used to determine the solubility of a synthetic puter hardware, and increasing amounts of digital
recombinant plague antigen (D’Souza et al. data, the applications of machine and deep
2009). The researchers first prepared buffers learning in various fields have grown exponen-
with pH values ranging from 3 to 10. Aliquots tially, which can significantly improve the effi-
of a stock solution of the monomer were then ciency and accuracy of product development,
diluted with one of the buffers to a concentration industrial manufacturing, and analytics processes
of 360 μg/mL. This solution was then placed in a (Lecun et al. 2015). In some cases, machine
dialysis cassette and dialyzed against the respec- learning and deep learning, especially deep
tive diluting buffer for 15 h under refrigeration. learning neural networks, are capable of matching
The dialysate was then filtered and analyzed by or even exceeding human performance and pos-
UV absorbance for solubilized compound. sess the potential to transform real-world
The direct determination in aqueous suspen- applications that rely on domain expertise and
sion method is not without limitations. years of empirical experiences (Lecun et al.
Substances with extremely low solubilities may 2015, Esteva et al. 2017). In the field of pharma-
fall outside of the detection limits for many avail- ceutical sciences, researchers have utilized artifi-
able analytical techniques. Additionally, cial intelligence to successfully predict the
molecules with stability issues may undergo com- performances of various pharmaceutical
plexation or aggregation, yielding inaccurate products. An integrated transfer learning was
results. An alternative to the direct determination developed in combination of multitask learning
in aqueous suspension method was presented by approach, successfully predicting pharmacoki-
Avdeef et al., in which the solubility is deter- netic parameters for small molecule drugs on the
mined through acid–base titration. Using experi- market (Ye et al. 2018). Compared to other
mentally determined p Ka and log P values, the machine learning approaches, Yang et al. (2019)
solubility of 12 generic drugs was determined and demonstrated that deep learning methods can
compared to values obtained by direct determina- achieve the highest accuracy to predict the
tion in aqueous suspension. Indeed, the data in vitro performances of pharmaceutical
correlated well. Uniquely, the method not only formulations. Furthermore, a deep learning
allows for intrinsic solubility determination but model was established for predicting the disinte-
also provides a complete pH-solubility profile for gration time of oral disintegrating tablets, and the
the drugs analyzed (Avdeef et al. 2000). This model achieved 80% accuracy (Han et al. 2018).
method was applied using the pSOL instrument As discussed previously, it is critical to mea-
and corresponding software in determining the sure the solubility of poorly water-soluble drugs
solubility profile of an experimental PWS com- in various media in order to guide formulation
pound NP-647 (Khomane et al. 2011). The disso- development, such as polymorph and salt screen-
lution titration method provided the researchers ing, excipient selections, and dissolution-
with both intrinsic solubility and enhancing strategy. Conducting such experimen-
pH-solubility data. tal measurements of those drug candidates can be
38 X. Ma et al.
time-consuming and laborious. Recent develop- been applied to predict the properties of pharma-
ment in automation has partially facilitated the ceutical compounds and formulations. Louis et al.
solubility measurements. However, due to the (2010) compared the prediction performance of
extensive applications of high-throughput screen- an artificial neural network and a support vector
ing in drug discovery, the efficiency of machine. Those two models were used to predict
synthesizing drug molecules has greatly the intrinsic solubility of 74 generic drugs. With
improved, generating a considerable number of the implementation of multiple linear regression
therapeutically active and low-molecular-weight analysis and cluster analysis, appropriate
drugs with preferable permeability across descriptors and training/testing data sets were
biological membranes but low or no solubility in generated. The tenfold cross validation illustrated
aqueous solution (Keseru and Makara 2009; that the artificial neural network and support vec-
Hauss 2007). Therefore, there is an urging need tor machine outperformed the conventional linear
to measure the solubility of those drug candidates model with higher correlation coefficients of
effectively and rapidly. Fortunately, the advance 0.823 and 0.835, respectively. In addition, by
in artificial intelligence has enabled the accurate analyzing the artificial neural network and
and speedy prediction of the solubility for those conducting a quantitative structure–property rela-
drug candidates. tionship study, the relationship between intrinsic
Huuskonen et al. (1997) built 3 separate neural solubility and four properties of the generic drugs
network models to predict the solubility of 83 dif- was established. A recursive neural network
ferent drugs in three classes (steroids, barbituric approach was developed by Lusci et al. (2013)
acid derivatives, and heterocyclic reverse tran- to predict aqueous solubility of drug-like
scriptase inhibitors). Topological descriptors, molecules. In this study, the recursive neural
including connectivity indices, kappa indices, networks were associated with vertex-centered
and electro-topological state indices, were used acyclic orientations of the molecular graph,
to connect the structural information of drugs which was beneficial to minimize the number of
with their solubility. Each neural network model suitable molecular descriptors, and the represen-
was based on back-propagation and consisted of tative descriptors were self-learned by the model
one hidden layer and five topological indices as from the data. Results showed that the perfor-
input parameters. All three neural network mance of the recursive neural network matched
models were optimized to have a neural network or outperformed other state-of-the-art methods in
structure of 5-3-1 and were fully tested for accu- the case where minimal data was available.
racy, efficiency, and processing time by using the Recently, researchers have developed a
leave-one-out procedure. Cross-validated results machine learning-based quantitative structure
of the three models showed that the squared cor- property relationship approach to predict drug
relation coefficients for steroids, barbituric acid solubility in binary solvent systems, which
derivatives, and heterocyclic reverse transcriptase provided useful information for spray-drying
inhibitors were 0.796, 0.856, and 0.721, respec- and coprecipitation process development (Chinta
tively. The relatively high squared correlation and Rengaswamy 2019). The proposed approach
coefficients demonstrated that neural networks utilized the structural features of the molecules,
were able to predict the aqueous solubility of such as molar refractivity, McGowan volume,
those compounds, and more importantly, neural and topological surface area. By correlating the
networks illustrated their extraordinary prediction log solubility values with the structural features,
capability in the case where only structural infor- the prediction efficacy of the model was up to
mation of those compounds were provided. 0.98, demonstrating the superior accuracy of
Due to the advances in computer sciences and predicting solubility of drug molecules compared
data analytics, more complicated machine to other conventional approaches. Cui et al.
learning and deep learning have been developed, (2020) constructed deeper-net models, consisting
and the combination of more than one model has of 20-layer modified ResNet convolutional neural
network structures. 9943 compounds were used comparison of salt vs. free base) and physical
to train and test the model. In addition, 62 newly mixtures of drugs and excipients or final
developed novel compounds were used to test the formulations (Viegas et al. 2001a, b; Lee et al.
model. Results demonstrated that the established 2011). USP 32/NF 27 identifies two apparatuses
deeper-net model outperformed the state-of-the- for the determination of intrinsic dissolution rates:
art approaches, shallow-net model, and four rotating disk and stationary disk. Regardless of
human experts with prediction accuracy of the unit selected for dissolution analysis, the sur-
85.5%. Boobier et al. (2020) utilized a combined face area of the material to be tested is held
machine learning approach by incorporating constant by generating a compact within the
computational chemistry and six different body of the device. This is done by attaching the
machine models, including artificial neural net- body to a baseplate, filling the body cavity with a
work, support vector machine, random forest, known amount of powder, placing a punch on top
ExtraTrees, Baaging, and Gaussian process. The of the powder surface, and compressing the pow-
developed models significantly improved the pre- der by means of a hydraulic press to a given
diction accuracy compared to benchmarked open- pressure for a predetermined period of time.
access tools. The achieved accuracy of the model Figures 2.2 and 2.3 depict a rotating disk and
was close to the expected level of noise in stationary disk assembly, respectively.
training data.
Besides the prediction of solubility, Li et al. Compact Preparation
(2018) developed a hybrid artificial intelligence The preparation of an adequate compact for anal-
model to model the pKa values of neutral and ysis is crucial in intrinsic dissolution testing to
basic drugs. In this model, an improved particle ensure that a constant surface area is maintained
swarm optimization algorithm was developed throughout the test. The term adequate compact
using the population entropy diversity. By del- refers to a compact which will not disintegrate
icately designing the model structure, the popula- during prolonged exposure to the dissolution
tion entropy was used to guide the prediction media and that has been compressed sufficiently
strategy. In addition, a radial basis function artifi- to remove all air from the powder bed, thereby
cial neural network model was combined with the preventing the formation of capillaries and poten-
particle swarm optimization algorithm, and the tial surface bubbles. The USP states that compres-
combined model was able to predict the pKa sion for 1 min at 15 MPa is sufficient; however,
values of 74 drugs with a squared correlation this is not universal, and the compression pressure
coefficient of 0.9685. Moreover, the proposed and time must be tailored to the formulation at
model was also used to predict the supercritical hand such that upon testing the compact does not
carbon dioxide solubility in polymers with disintegrate. High compression forces and exces-
integrated diffusion theory (Li et al. 2021). The sive times may induce physical changes in a
predicted value was consistent with the experi- powder such as polymorphism or crystallization,
mental results, and a square correlation coeffi- and a compact formed by the selected parameters
cient of 0.9954 was achieved. should be analyzed for such transitions. For
example, a novel quinolinone derivative required
compression at 196 MPa to yield a solid nondisin-
2.1.5 Intrinsic Dissolution tegrating compact. X-ray diffraction revealed that
no transformations took place upon compression,
The intrinsic dissolution rate of an API is the rate thereby validating the compaction procedure and
of dissolution of the compound under constant obtained dissolution results (Kimura et al. 2001).
conditions (i.e., identical surface area, tempera- In a study of 13 different APIs, a compression
ture, agitation rate, pH, and ionic strength of the time of 1 min was sufficient for all materials;
dissolution media for all samples) and is used to however, the compression force required varied
demonstrate equivalency of raw components (i.e., from 1.96 to 19.6 MPa (Zakeri-Milani et al.
40 X. Ma et al.
Fig. 2.2 Rotating disk intrinsic dissolution apparatus diagram. From Viegas et al. (2001a, b)
Dissolution Vessel Rotating

(Flat Bottom) Paddle Shaft
Plastic
Cap
Pressure
Gasket
Plunger
Die
Plunger Paddle
Drug Drug Pellet

Die
Powder
Drug 1 inch
Die
Pellet
Surface
Plate
Gasket
Plastic Cap Plunger
Fig. 2.3 Stationary disk intrinsic dissolution apparatus. Reproduced with permission from Dissolution Technologies
2009). Viegas et al. (2001a, b) applied a force of appropriate method for the study at hand is at
2000 psi for 4–5 min to achieve a satisfactory the discretion of the researcher.
disk. As can be seen, the compression force will Sampling time points will depend upon the
vary greatly between formulations; however, a compound under analysis. Sampling at 5 min
study by Yu and co-workers demonstrated that intervals for 1 h has proven sufficient to obtain a
the compression force applied during compact dissolution profile with a region of linearity.
formation does not influence the intrinsic dissolu- Although no solid mass should be present as the
tion rate of a material, provided the compact does compact is nondisintegrating, filtration of the
not undergo physical alterations (i.e., polymor- withdrawn media through a 0.45-micron filter is
phic transitions or crystallization) or disintegrate recommended. As per the USP 32/NF 27, the
during testing (Yu et al. 2004). amount-versus-time profile is plotted; however,
only the initial linear region is used for intrinsic
dissolution rate calculation. Regions of nonline-
Intrinsic Dissolution Testing
arity are excluded from data analysis. The con-
The abovementioned study by Yu additionally
stant surface area dissolution rate (i.e., slope) is
demonstrated that dissolution volume (provided
then reported in the units of mass/s and the disso-
sink conditions are maintained) and disk distance
lution flux reported as mass/cm2/s. Flux is
from the vessel bottom do not influence the intrin-
obtained by dividing the calculated dissolution
sic dissolution rate of a given compound. How-
rate by the surface area of the compact and is
ever, Viegas showed that for both types of ID
the parameter to be reported with regard to intrin-
apparatuses, the rotation speed or paddle speed
sic dissolution data reporting.
greatly influenced the dissolution rate and there-
fore must be held constant when comparing
formulations. The literature supports values
2.2 Solid-State Characterization
between 50 and 150 rpm, with slower speeds
being used for compacts more prone to disinte-
Solid-state characterization of APIs, excipients,
gration. During dissolution testing, the media is to
and combinations thereof is invaluable during
be heated to physiological temperature of 37 C.
product development. The techniques outlined
The standard apparatus II volume of 900 mL is
in the following sections are utilized to gain an
often used; however, smaller volumes such as
understanding of the API’s physical nature,
250 mL can also be employed when larger
interactions with excipients, and response to
volumes will lead to concentrations below the
processing and environmental conditions. The
limit of detection for the compound (Viegas
following sections do not contain detailed infor-
et al. 2001a, b; Lee et al. 2011).
mation regarding the theory behind each tech-
The media for dissolution testing must be ade-
nique. Rather, a discussion of how the methods
quately degassed prior to experimentation. This
should be utilized during PWS drug formulation
not only prevents potential oxidative degradation
is provided.
of dissolved compound but also prevents poten-
tial bubble formation on the surface of the com-
pact during testing, thereby disrupting the factor
2.2.1 Thermal Analysis
of constant surface area. Multiple methods can be
employed to degas the media, including sparging
Differential Scanning Calorimetry
with an inert gas (i.e., helium or nitrogen) for
Differential scanning calorimetry (DSC) is widely
30 min at 6 psi, stirring heated media under vac-
used for physical characterization of drugs and
uum (41 C, 160 mmHg), sonication (30 min),
formulations thereof including identification of
sonication under vacuum (60 mmHg), and equili-
melting, crystallization, and thermal transition
bration at 37 C for 24 h (Diebold and Dressman
phenomena, as well as the associated changes in
1998; Gao et al. 2006). Selection of the
entropy and enthalpy. DSC has the advantage of
42 X. Ma et al.
requiring only minimal amounts of the volume expansion upon heating be limited in
compounds under study and a rapid timeline to their sample size to prevent spillage from the
completion. During analysis, a sample pan and a pan to the DSC cell surface. Such events can
reference pan are subjected to a heating profile, also be prevented by selection of the correct pan
and the energy and temperatures associated with type. If contamination of the cell surface occurs,
thermal events are assessed. For a detailed under- the system must be cleaned following manufac-
standing of DSC theory, the reader is referred to turer guidelines and recalibrated prior to
Craig and Reading (2007) and Schick (2009). The subsequent analyses.
following sections discuss the application of DSC The heating rate selected for testing must be a
to preformulation studies, including selection of compromise between resolution, sensitivity, and
appropriate parameters, identification of promi- time. While slower heating rates allow greater
nent thermal events, and application in poly- separation of thermal events (i.e., increased reso-
morph and excipient screening. lution), the sensitivity to these events is reduced,
and the run time for each analysis is greatly
Parameter Selection Method increased. Conversely, higher heating rates will
As mentioned earlier, DSC has the advantage of decrease the resolution, as thermal events
requiring only minimal amounts of the separated by only a few degrees will overlap.
compounds under study and a rapid timeline to However, faster heating rates generate greater
completion. Reported sample sizes range from sensitivity to thermal transitions, allowing visual-
2 to 15 mg and can be selected to optimize the ization of minor events while also drastically
response profile. Increasing the sample size will reducing the time for analysis. In fact, a ramp
increase the sensitivity of the system; however, rate of 100 K/min was required in order to
this will also decrease the resolution of the trace. detect weak transitions occurring in a lyophilized
The converse holds true with smaller samples protein sample and corresponding formulations
increasing the resolution at the sacrifice of sensi- thereof (Carpenter et al. 2009). Additionally,
tivity. It should additionally be noted that peak increases in ramp rates will alter the Tmax of the
maximum temperature for thermal events will melting endotherm (Abbas et al. 2008). Ramp
increase with increasing sample size, and thus rates for analysis of pharmaceutical materials of
the sample size must be held nominal throughout 2, 5, or 10 K/min are most prevalent in the litera-
a study. ture. Regardless of the rate selected, all samples in
While multiple pans can be used for sample a given study must be analyzed under the same
analysis (i.e., aluminum crimp top and aluminum heating rate for the purpose of comparison.
hermetic), it is imperative that during preparation It is necessary to calibrate the instrument using
sample contact with the pan be maximized. standards of known temperatures and enthalpies
Improper sample contact can increase thermal of melting under the identical conditions to be
lag within the system, creating inconsistent used during analysis. The reference material
results. The most prevalent pan type for analysis (s) should have melting temperatures near those
is aluminum crimp top pans such as those offered of the analytical samples to ensure accuracy.
by Perkin-Elmer. For materials which may Materials for calibration include zinc, indium,
undergo transitions induced by the pressure aluminum, lead, silver, tin, gallium, and gold
applied during crimping, a noncrimped pan with (Cammenga et al. 1993). Additional calibration
lid can be used, or pans capable of being hermeti- standards include naphthalene, benzyl, benzoic
cally sealed can be employed. Whichever pan acid, and anisic acid, which have proven useful
type is selected, it is absolutely essential that the for calibrating temperatures between the melting
corresponding reference pan be identical to that temperatures of the metals indium and gallium
used for sampling (lid included if utilized), except (Charsley et al. 2006).
only the material being analyzed. Additionally, it Modulated differential scanning calorimetry
is crucial that samples which may undergo (mDSC) provides a linear heating ramp with a
superimposed temperature oscillation yielding a crystalline form, and the most stable crystals
modulation in the heating profile. mDSC provides will melt after all defected crystals are melted.
identical data as that obtained by traditional DSC Those intrinsic crystalline defects create
analysis (i.e., total heat flow). However, Fourier difficulties for interpreting the melting from
transform separation of the DSC signal allows the reversing heat flow because it also includes the
heat-capacity-related component (reversible sig- energy used to melt some recrystallized crystals,
nal) to be isolated. Furthermore, subtraction of the which often shows as a broad endothermic peak
reversing component from the total heat flow in the non-reversing heat flow (Craig and Reading
allows for isolation of the non-reversing compo- 2007). Additionally, the process of transferring
nent of the scan (Aldén et al. 1995; Sauer et al. heat into a sample cannot be instantly completed
2000). This ultimately allows the researcher to especially when the heating rate is relatively high,
better visualize thermal events, as no exothermal and the sample temperature cannot match with the
events are present in the reversing signal. When modulated heating rate (Schawe 2007). This tem-
mDSC is to be used, the underlying heating rate perature discrepancy also occurs when a large
should be selected as outlined above based on the amplitude and short period are set for mDSC.
required resolution-to-sensitivity relationship. So, it is not uncommon that the melting event
However, additional parameters are required, also shows in the non-reversing heat flow because
including the modulation temperature and fre- of its kinetic nature (Hill et al. 1999). Moreover,
quency thereof. Selection of these parameters both heat capacity change and latent heat of
will be based on the applied heating rate as it is fusion will be present in the case of melting, but
desired to have six modulations over the temper- they respond differently to the temperature mod-
ature range of a thermal event. As such, slower ulation. Also, the separation of reversing and
heating rates are often employed (i.e., 2–5 K/min) non-reversing heat flow will change based on
to ensure that these criteria are met. Modulation the selected experimental conditions (Hensel
amplitudes should range between 0.1 and 2 K, as et al. 1996). Table 2.2 summarized the advantages
larger amplitudes may adversely affect data and disadvantages of mDSC and DSC for
(Schawe 1996). characterizing various thermal events, which
Many complex thermal events can be will be discussed in detail in the following
differentiated in mDSC, but it is a special case section.
for melting, which is a latent heat. This means that
the heat is absorbed or released at a constant Thermal Events
temperature. Thus, at the melting temperature of Proper identification of the thermal events occur-
a sample, the amount of energy needed to melt all ring upon heating of a sample can prove rather
crystals is fixed, which means the heat flow of challenging in practice. The most prevalent ther-
melting is linearly correlated with the heating mal event identified by DSC analysis is that of
rate. For example, during the heating process, if melting. Upon reaching the melt temperature of a
heat flow increases by twofold, the time for melt- substance, the sample will remain isothermal until
ing all crystals would decrease by half (Craig and the entire sample has melted, as a length of time is
Reading 2007). Therefore, the heat of fusion is required to overcome the thermal lag across the
reflected in the reversing heat flow. However, sample. During this time, the trace yields an
there are some exceptions. First, no crystals are endothermic event; however, the manner in
perfect; in other words, there are always some which the melting temperature (Tm) is reported
defects in the structure of crystals, especially for based on this event varies among researchers. Tm
polymers (Wunderlich 2005). A combination of has been reported at the initial deviation from
different crystals with different imperfections will baseline, at the temperature of onset of the
result in a case where some of the relatively endothermal peak, the temperature at the peak,
unstable crystals will melt first with less amount or the temperature at which the thermogram
of energy and transform to a more stable returns to baseline. While these values may vary
44 X. Ma et al.
Table 2.2 Advantages and disadvantages of mDSC and conventional DSC

Modulated DSC Conventional DSC
Advantages Separation of complex and overlapping thermal events Suitable for the measurement
Deconvolution of heat capacity and kinetically controlled events of melting
Glass transition: increased limit of detection and sensitivity, Simple experimental settings
trustable interpretation, accurate quantification of amorphous Time–cost efficient
phases, available calculation of relaxation time
Study of phase separation
Melting: separation of crystallization from melting
Crystallization: easy interpretation from non-reversing heat flow
Disadvantages Careful experimental designs Unable to analyze overlapping
Strongly condition-dependent events
Melting: difficult interpretation, not accurate measurement in Lower sensitivity for heat
mDSC capacity measurement
Lower limit of detection of
thermodynamic events
by as little as 3 C for highly pure substances material will vary slightly based on the selected
displaying narrow melting peaks, substances heating rate and frequency for mDSC, with
with low purity or compositions containing mul- increased frequencies increasing the Tg, and,
tiple crystalline forms (i.e., semi-crystalline therefore, should be held constant for all samples
polymers) yield broad melting endotherms, caus- in a given study to ensure accuracy in their com-
ing large differences in Tm based on the report parison (Schawe 1996). As will be discussed in a
method (Brown 2001; Craig and Reading 2007). subsequent section, polymer miscibility is crucial
As peak broadening may occur in the presence of for amorphous systems to ensure homogeneity
various excipients, or the pure substance itself and stability, and, oftentimes, the Tg values of a
may have a broad melting endotherm, the initial system are indicative of miscibility. The theoreti-
deviation, onset, and return to baseline cal glass transition temperature of a system can be
temperatures may vary drastically among samples predicted using the Gordon–Taylor equation
in a single study. However, the peak temperature
w1 T g1 þ Kw2 T g2
is likely to remain relatively unchanged in the T g12 ¼ ,
w1 þ Kw2
absence of formulation component interactions
provided the sample size and heating rate are where Tg1 and Tg2 are the glass transition
held constant, making its selection as the temperatures of raw components 1 and 2, respec-
reporting temperature more reliable (Abbas et al. tively, and w is the weight fraction of the
2008). corresponding components. The value of K, the
Amorphous systems such as solid solutions are ratio of the components free volume, is such that
becoming highly prevalent in the development of
PWS drugs. Of great importance to these poly- ρ1 Δα1
K¼ ,
meric and amorphous systems is the glass transi- ρ2 Δα2
tion temperature (Tg). The nature of the glass
with values for Δα being found in the literature.
transition temperature has been discussed in
For materials in which Δα is not available, the
numerous ways, including relaxation processes
value of K can be estimated using the true
and changes of free volume in a system; however,
densities of components 1 and 2 such that
at its simplest, the glass transition temperature is a
change in the heat capacity of the material and is ρ1 T g1
Kffi :
observed as an endothermic step change in the ρ2 T g2
baseline of the DSC trace. Analysis of the revers-
ing heat flow obtained through mDSC provides Direct comparison of the theoretical (calculated)
the best visualization of the Tg event. The Tg of a Tg to the experimentally observed Tg allows
Fig. 2.4 Reversing heat flow DSC profile of itraconazole and HP55 solid dispersions prepared by ultrarapid freezing.
Reproduced with permission from Elsevier
assessment of drug–polymer or polymer–polymer occurring upon heating without sufficient poly-

miscibility (Gordon and Taylor 1952; Boyer and meric stabilization (Overhoff et al. 2007).
Simha 1973). Recrystallization is an exothermic event, and,
As the thermal event associated with the glass as such, it is not observed on the reversing heat
transition occurs over a broad temperature range flow profile. Examination of the corresponding
(i.e., 10 C), the Tg has been reported in multiple heat flow in the above example reveals
ways, including step onset, midpoint, and end- itraconazole to recrystallize between 110 and
point temperatures. Analogous to Tm, the exact 130 C, depending on the polymer concentration
onset and endpoint temperatures may not be eas- of the system. Recrystallization occurs as a result
ily discernable on the trace. Therefore, calculation of increased molecular mobility and subsequent
of the midpoint via the software associated with molecular rearrangement during heating and can
the unit at hand is the recommended means of be observed in poorly stabilized amorphous
interpreting Tg. Figure 2.4 depicts the reversing systems following the glass transition event but
heat flow of formulations of itraconazole and prior to the melting endotherm. This can also be
hydroxypropylmethyl cellulose phthalate NF observed following polymorphic transitions
(HP55). Here, the glass transition temperature (to be discussed), in which following melting of
can readily be observed as a shift in baseline. As form I the material recrystallizes into form II with
the ratio of drug-to-polymer changes, the Tg subsequent melting of form II, if sufficient time is
shifts, and indeed this shift correlated well with provided for recrystallization (i.e., slow heating
the theoretical Tg predicted by the rates) (Alves et al. 2010). As decomposition
abovementioned Gordon–Taylor equation. It events may also appear as exothermic peaks, dif-
should be noted that at high drug loadings, the ferentiation between the two must be confirmed.
melting endotherm of itraconazole can be This can be accomplished by supplemental
observed. This is attributed to recrystallization thermogravimetric analysis (TGA) which will be
discussed in the coming sections.
46 X. Ma et al.
Polymorphic Transformations product for tableting led to the inclusion of stearic

The presence of an undesired polymorphic form acid in the formulation to prevent sticking. How-
of an API within a formulation may drastically ever, it was found that the addition of stearic acid
alter the performance of the product, as the differ- to the formulation with subsequent drying of the
ent polymorphs of a given API possess different wet granulated mass at elevated temperatures
physical properties. An understanding of these induced a transformation to form I, thereby dras-
polymorphic forms with regard to their preva- tically reducing dissolution rates. This was con-
lence and transformations will assist the formula- firmed via DSC analysis of granulated mass
tion scientist in optimizing the formulation. As compared to thermograms of physical mixtures.
physical properties are different for each poly- Indeed, the physical mixture of form II in the
morph of a given API, the Tm values of each presence of stearic acid displayed melting
polymorph can readily be used for their identifi- endotherms for both polymorphic forms,
cation with exothermic events indicating species indicating partial conversion to the undesirable
conversion (Bergese et al. 2003). For example, form (Wang et al. 2010).
DSC analysis of polymorph II of rifampicin
exhibits an endothermic peak at 193.9 C due to Drug–Excipient Interactions
melting. A subsequent exothermic recrystalliza- It should be immediately noted that many of the
tion is observed at 209.4 C as polymorph II API–excipient interactions observed in DSC anal-
recrystallizes into form I (Alves et al. 2010). ysis, as well as the polymorphic transformations
When attempting to identify the polymorphs pres- previously discussed, occur at temperatures dras-
ent in a system, it is imperative that the appropri- tically higher than those to be experienced in real-
ate heating rate be selected for analysis to ensure world application, with the exception of thermal
that the energy applied by the system does not processing techniques such as hot-melt extrusion.
induce a transformation leading to improper iden- This must be considered when analyzing results
tification within a formulation. The poorly water- to ensure that constituents are not incorrectly
soluble drug nateglinide polymorphs B and H deemed incompatible.
have Tm’s of 128 and 138 C, respectively. In Determination of the stability of an API in the
order to properly identify the two forms when presence of excipients requires analysis of the raw
analyzed together, a heating rate of 10 K/min components, a physical blend of the components,
was employed. Indeed, at rates faster than 10 K/ and a processed sample of the components. The
min, the peaks overlapped hindering identifica- physical mixture should be prepared in a manner
tion, while slower heating rates allowed sufficient that does not introduce heat, moisture, or force
time for transformation of form B to H (Bruni into the system. Gentle blending of the powders
et al. 2011). It was additionally shown in this in the desired ratio in a mortar with a spatula at
example that the grinding process (mortar/pestle) ambient temperatures is recommended. The
for preparation of the mixtures of the polymorphs processed sample may be as simple as a
induced the B-to-H transformation. co-ground mixture (i.e., grinding via mortar pes-
Transformations may occur as a result of tle for a minimum of 15 min) or may be a final
processing conditions during production. As will formulation such a as a tablet or hot-melt
be discussed in the next section, an understanding extrudate that has been ground for analysis. Anal-
of API stability in the presence of the excipients ysis of the raw individual components will allow
to be used for the formulation is vital to ensure identification of thermal events associated with
product performance. This includes assessment of the isolated material such as the melting, recrys-
process-induced polymorphic transformations. tallization, and glass transition (if applicable)
For example, form II of an experimental com- temperatures outlined above. Subsequent analysis
pound (drug Z) displays desirable dissolution of physical blends and processed material can
characteristics over form I. Development of the then be inspected for shifts, disappearances or
appearances of thermal events, or variations in
the enthalpy values thereof (Mura et al. 1998a). In or inclusion complexation within a material such
order to promote interactions, concentrations of as cyclodextrins (Cappello et al. 2007). Polymers
excipients are often employed which far exceed identified as being most highly miscible should be
the amount to be present in the final formulation. analyzed by TGA for verification that the absent
For example, a 1:1 physical mixture of melting endotherm is not due to degradation of
carbamazepine-to-stearic acid (typically present the API (Hughey et al. 2010). Flory–Huggins
in tablets at the 0.5–2% level) demonstrates a theory is a useful tool for drug–polymer
polymorphic transition that is unlikely to occur miscibility predictions and will be discussed in a
in a final formulation containing minimal subsequent section.
amounts of the excipient (Joshi et al. 2002). In Upon formation of a solid dispersion, the drug
contrast, the transition described above for drug Z and polymer are intimately mixed such that the
occurred using the actual amount of stearic acid to system, as described above, will have a single
be employed during the granulation process, intermediate Tg. Miscible systems, therefore,
indicating a strong incompatibility. A 1:1 physi- should only display this intermediate value, and
cal blend of picotamide with various excipients the appearance of multiple Tg’s is indicative of
demonstrated that ascorbic and tartaric acids incomplete mixing or drug- or polymer-rich
resulted in degradation of the drug upon heating. domains. Although the presence of a single exper-
The same effect was also observed for processed imental Tg intermediate to that of the polymer and
samples; however, as mentioned above, the API alone is a strong indicator of a homogeneous
temperatures at which this interaction took place system, it is necessary to perform supplemental
were drastically higher than those a final formula- studies to ensure that such a product has truly
tion would experience necessitating confirmation. been obtained. For example, Raman mapping of
Indeed, analysis of freshly prepared samples and solid dispersions has demonstrated that a system
stored samples at room temperature by X-ray displaying a single Tg in fact contained drug-rich
diffraction (XRD) revealed that no interaction domains, ultimately leading to recrystallization-
occurred and the DSC data were falsely related stability issues upon storage (Qian et al.
identifying an incompatibility (Mura et al. 2010). The PWS drug irbesartan has been
1998a, b). prepared into solid dispersions with tartaric acid,
The indication of drug–excipient interactions mannitol, PVP, and HPMC in an effort to increase
may not be detrimental to the formulation, but aqueous solubility. mDSC analysis of the raw
rather a desired interaction such as polymeric component (4–6 mg sample) revealed a melting
miscibility. DSC has been widely used in deter- endotherm at approximately 185 C. Preparation
mining drug–polymer interactions, with emphasis of a quench-cooled sample allowed for analysis
on drug miscibility in the polymer during formu- of amorphous irbesartan in the absence of any
lation of solid dispersions. Indeed, identification excipients which showed a single Tg at 71.7 C
of the polymer in which the drug is most miscible with no subsequent recrystallization event. DSC
will aid in the formulation of a single-phase sys- analysis revealed a single Tg for irbesartan–
tem (Mora et al. 2006). Prior to the analysis of tartaric acid dispersions indicating good
formulations, physical blends of the drug and miscibility. However, irbesartan–mannitol
polymer are analyzed by DSC to isolate polymers formulations presented two Tg values
displaying the highest miscibility with the API. A corresponding to the Tg’s of mannitol and
negative shift in the melting endotherm (i.e., irbesartan, suggesting an immiscible system.
melting point depression) indicates miscibility at Similar results were obtained for both PVP and
the ratio tested. A complete disappearance of the HPMC, indicating the drug is not miscible in the
melting endotherm indicates the absence of crys- selected carriers and that reformulation is neces-
talline API and solubilization of the API within sary in order to achieve a true solid solution
polymeric carrier (Konno et al. 2008). Such an (Chawla and Bansal 2008).
event also demonstrates amorphization of the API
48 X. Ma et al.
Flory–Huggins regions C and D are metastable, and regions E

As noted previously, drug–polymer miscibility is and F are unstable regions (Tian et al. 2012).
a critical parameter for homogeneity and stability There are two primary methods for χ determi-
in amorphous dispersions. Flory–Huggins theory nation: solubility parameter differences and melt-
has been applied to determine miscibility infor- ing point depression evaluation. However,
mation in polymer–polymer and polymer–solvent solubility parameter differences tend to be inac-
systems and, more recently, in drug–polymer curate in systems that have significant drug–poly-
systems. The Flory–Huggins model utilizes a mer interactions, which is true for many drug–
drug–polymer interaction parameter, χ, to calcu- polymer systems. For these systems, melting
late the free energy of mixing for the system using point depression is the preferred approach. DSC
the following equation: is utilized to measure the melting point onset
(Zhao et al. 2011), melting temperature (Marsac
ΔGM
¼ ndrug ln Φdrug þ npolymer ln Φpolymer et al. 2009; Lin and Huang 2010), or endpoint
RT
(Tian et al. 2012). Studies should be performed to
þ ndrug Φpolymer χ determine which of these values provides the
most reproducible results for the system of inter-
where n is the moles of the drug or polymer, Φ is
est. Once melting point depressions have been
the volume fraction of the drug or polymer, ΔGM
determined, the drug–polymer interaction param-
is the free energy of mixing for the system, R is
eter can be calculated utilizing the following
the ideal constant, and T is the temperature of
rearranged equation (Marsac et al. 2006):
interest (Marsac et al. 2006). χ can be further

used to generate spinodal (boundary between 1 1 ΔH fus
pure ln Φdrug
unstable and metastable regions) and binodal T mix
M TM R
(boundary between metastable and stable regions)
1
curves for the system to predict the stable, meta- 1 Φ
m polymer
stable, and unstable regions for solid dispersions
of the system (Huang et al. 2015). A stable system ¼ χΦ2polymer
is one in which the components will remain
where TM values are the melting points of the
single-phase while metastable/unstable systems
mixture or pure drug, T pure
M R is the ideal gas
will tend to phase separate. A phase separated
constant, ΔHfus is the heat of fusion for the pure
system will possess polymer-rich and drug-rich
drug, m is a constant for the relative size of the
regions. The drug-rich region may no longer be
polymer to the drug, and the Φ values are volume
thermodynamically stable ultimately leading to
fraction of drug or polymer. If a plot of the left-
recrystallization of the drug. Of particular con-
hand side of the equation vs. the Φ2 value for the
cern, absorbed moisture can drastically reduce
polymer demonstrates a linear relationship, then
drug–polymer interaction leading to phase sepa-
the slope of the best fit line is equivalent to χ.
ration and recrystallization (Marsac et al. 2010;
More negative χ values suggest miscibility,
Purohit and Taylor 2015). However, systems with
whereas more positive χ values tend to suggest
strong drug–polymer interactions, reduced hygro-
immiscibility. It should be noted that there are
scopicity, or less hydrophobic active ingredients
limitations to the melting point depression
may resist phase separation even in the presence
approach. The components of the system must
of high moisture content (Rumondor et al. 2011).
be chemically stable over the range of evaluation.
Kinetics plays an important role in the phase
Additionally, the components must have ample
separation with metastable systems often
physical interaction for reliable determination of
remaining absent of crystals during
melting point depression. This is particularly
pharmaceutically relevant timelines. Figure 2.5
important for systems where the polymer Tg is
depicts an example phase diagram from Flory–
near or above the depressed melting point of the
Huggins theory. Regions A and B are stable,
Fig. 2.5 Phase diagram for

a drug–polymer system
from Flory–Huggins
theory. Reproduced with
permission from American
Chemical Society
drug as the system will be unfavorable for mixing drug–polymer mixture prior to analysis to obtain
of polymer with molten drug. Optimization of solubility equilibrium temperature and ultimately
experimental parameters can be used to overcome the drug–polymer interaction parameters (Sun
this, such as reducing DSC scan rate or reducing et al. 2010).
particle size (Marsac et al. 2006).
The drug–polymer interaction parameter has
Thermogravimetric Analysis
been utilized in polymer screening to determine
A secondary thermal analytical technique which
ideal polymer candidates and drug loading. A
can stand alone or be complementary to DSC is
Flory–Huggins study of carbamazepine with var-
TGA. Similar to DSC a small quantity of sample
ious polymers was utilized to screen polymers,
(sample sizes range identically to DSC) is placed
predict drug load, and determine initial
in a metal crucible and exposed to a heating
processing parameters for hot-melt extrusion
profile under an inert atmosphere during analysis.
(HME). Interestingly, the AFFINISOL™ HPMC
However, as the name implies, TGA provides the
HME polymers under investigation had increas-
researcher with variations in sample mass as a
ing χ values with increasing temperature that
function of temperature, rather than the deviations
suggested that the drug and polymer were less
in heat capacity. Examples of these reactions
miscible at higher temperatures (Huang et al.
include the evolution of bound solvents, such as
2015). Similarly, this approach has been used to
dehydration, or thermal decomposition of the
generate a complete phase diagram for a
product.
felodipine-poly-acrylic acid system (Lin and
Parameter selection (i.e., temperature range
Huang 2010). An alternative approach has also
and heating rate) depends upon the intention of
been developed to determine the equilibrium
the data. If TGA is to be used complementary to
room temperature solubility of drug in polymer.
DSC, the identical temperature range and heating
The equilibrium solubility value can be useful as
rate employed in for DSC must be selected for
it would provide the room temperature stable
TGA. This will allow for superimposition of the
system boundary (Bellantone et al. 2012).
thermal profiles and a direct comparison of ther-
Another alternative approach is to anneal the
mal events. If superimposition of the data is not
50 X. Ma et al.
performed, a unique heating profile may be weight loss at 208 C (Radha et al. 2010). TGA
assigned regardless of that used for DSC. This profiles overlaid with DSC profiles of rifampicin
includes isothermal analysis at elevated tempera- demonstrated that the exothermic events
ture in the study of reaction rates. evidenced in DSC analysis were indeed decom-
position events (Alves et al. 2010).
Thermal Decomposition
TGA has been broadly used in the study of ther- Excipient Interactions
mal decomposition; however, it must be noted Similar to DSC, TGA allows one to study the
that such studies are only appropriate for thermal compatibility of an API with multiple
materials which exhibit a weight loss upon excipients. Such information is invaluable for
decomposition. A study of numerous pharmaceu- formulations intended for thermal production pro-
tical excipients provided information of their ther- cesses such as hot-melt extrusion (HME). Indeed,
mal stability. While many of the excipients drug–excipient compatibility at elevated
demonstrated a single-step weight loss upon temperatures may be drastically reduced com-
reaching their respective thermal instability pared to ambient conditions. TGA analysis of
point, it is possible to observe multiple hydrocortisone in the presence of the polymers
mechanisms concurrently. Lactose monohydrate HPMC E3 and PVPVA 64 demonstrated that
exhibits multiple-step weight loss under TGA. while all materials experienced decomposition at
The first step, observed between 106 and approximately 200 C, the presence of the poly-
164 C, can be attributed to the dehydration of mer did not induce degradation of the API at
the bound water molecule. The second step from lower temperatures (DiNunzio et al. 2008). Con-
216 to 339 C is attributed to thermal decomposi- trarily, TGA analysis of an experimental com-
tion (Filho et al. 2009). The thermal stability of pound (ROA) in the presence of Eudragit®
drugs can also be assessed as was demonstrated L100-55 demonstrated a significantly higher
by obidoxime chloride. Indeed, the decomposi- weight loss than predicted based on the data
tion occurred over a wide range of temperatures, obtained from the individual components
beginning at 118 C and continuing to 328 C. indicating API–excipient incompatibility
TGA revealed that the drug underwent a 10% (Fig. 2.6). Similar results were observed in the
Fig. 2.6 API thermal 110.0

decomposition in the
presence of Eudragit L100- 100.0
55
Eudragit®L100-55
90.0
Weight Remaining (%)
ROA : Eudragit®L100-55 1:2

80.0
Actual
70.0
ROA : Eudragit®L100-55 1:2
60.0 Predicted
50.0
40.0
ROA
30.0
50 100 150 200 250 300 350
Temperature (°C)
ROA–HPMCAS blends (Hughey et al. 2010). In dσ dγ

σ þ τM ¼η
the abovementioned example of obidoxime chlo- dt dt
ride, the drug in the presence of multiple tableting
where σ is the stress, τM is the relaxation time and
excipients was also assessed by TGA. Indeed, it
equal to Gη , t is the time, η is the viscosity, γ is the
was shown that none of the tested excipients
strain, and G is the modulus of the spring.
altered the decomposition profile of the API.
The Maxwell model explicitly describes the
viscoelastic behavior of pharmaceutical
Dynamic Mechanical Analysis (DMA) polymers. When the stress is applied, unlike an
Maxwell unit (Fig. 2.7) has been widely used to ideal solid, there is a nonlinear correlation
understand the rheological behavior of pharma- between the stress or deformation and time. The
ceutical polymers. Most polymers possess both relaxation of polymers occurs when the viscous
elasticity and viscosity, and a simple analogy liquid in the dashpot releases the stress of the
combining a spring and a dashpot can simulate spring and gradually removes the deformation of
the complicated rheological behavior of polymers the spring until the equilibrium is reached. Relax-
individually. In the Maxwell unit, the spring and ation of polymers is caused by the flow of the
dashpot are connected in series, while they can liquid and can be quantified based on the equation
also be arranged in parallel, which is the Voigt above using a dynamic oscillatory method (Jones
model (Ferry 1980). et al. 2003).
As demonstrated in Fig. 2.7, an elastic spring There are several methods that can be used to
is placed into a dashpot containing a purely vis- characterize the rheological properties of
cous liquid. By imposing stress on this system, polymers, such as oscillation/non-oscillation and
both the spring and dashpot will have the same temperature ramping/isothermal setups. Dynamic
amount of stress, but the strain of the spring and mechanical thermal analysis is a non-sample-
dashpot will be different. The equation below destructive technique to measure the resultant
serves as the principle of theologically mechani- strain that is originated from the applied
cal analysis of polymers through analyzing the oscillatory stress, determining the function of
stress and strain of each component individually the strain versus frequency or temperature. Vari-
and the whole system. ous temperature profiles (temperature ramping or
isothermal) can be set up using DMA, evaluating
the effects of temperature on the rheological
properties of polymers (Davis 1971). DMA is a
convenient and accurate technique to characterize
Fig. 2.7 Illustration of the viscoelastic properties of polymers.
viscoelasticity using the When using DMA to analyze polymers, it is
Maxwell model important to ensure that the rheology experiment
is conducted within the linear viscoelastic region
of the polymer, which means that the deformation
of the sample needs to be linearly correlated with
the applied oscillatory stress at any given time.
Also, viscosity and rigidity should be indepen-
dent and unaffected by the changes of the strain.
DMA enables the investigation of molecular
structures and correlates the properties obtained
from the linear viscoelastic region with the per-
formance of the drug formulations.
Similar to mDSC, a sinusoidal perturbation,
which is referred to as oscillatory stress, can be
52 X. Ma et al.
employed on the linear-changing stress. For an viscosity as a function of shear rate, which
ideal solid, the following equation defines the facilitates the characterization of the materials
rheological response of the object: that cannot be measured under shear conditions
σ σ0 (Aho et al. 2016). Overall, analyzing the rheolog-
γ¼ ¼ cos ðωt Þ ¼ γ 0 cos ðωt Þ ical properties of drug–polymer formulations
G G
provides insights of molecular mobility and
where σ 0 is the amplitude of the stress and ω is the serves as an alternative method to measure the
frequency of the oscillation. glass transition temperature (Andronis and
For a viscous liquid (Newtonian), the follow- Zografi 1997; Abiad et al. 2010).
ing equation is defined:

π
γ ¼ γ 0 cos ωt Þ 2.2.2 Fourier Transform Infrared
2
Spectroscopy and Raman
The above equation indicates that a Newtonian Spectroscopy
liquid exhibits a phase lag (δ) of 90 compared to
the applied stress. For viscoelastic pharmaceutical Fourier transform infrared spectroscopy (FTIR)
polymers, the phase lag is between 0 and 90 . allows one to gain information regarding the
Due to the presence of the phase lag, two different molecular confirmation of a material based on
responses, elastic and viscous components, can be characteristic molecular vibrations that absorb in
deconvoluted and separated by the oscillation the infrared region. There are multiple variations
experiment. Additionally, DMA can measure of FTIR analysis, including diffuse reflectance
storage modulus (G0), loss modulus (G00), com- (DRIFT) and attenuated total reflectance (ATR).
plex modulus (G), damping factor (tanδ), and Each has its own advantages, and selection of the
complex viscosity (η) (Aho et al. 2016). Com- appropriate unit will be at the discretion of the
plex modulus is the ratio of the stress amplitude to researcher; however, the focus of the sections
the strain amplitude. Because of the presence of a below will be on ATR–FTIR as it requires mini-
phase lag between strain and stress, the storage mal alteration of the sample prior to analysis. As a
modulus only reflects the portion of the strain that complementary tool to FTIR, Raman spectros-
is in phase with the stress. While the storage copy detects the polarizability changes of the
modulus represents the energy that is stored in a molecules during the vibrational motion (Ferraro
sample, loss modulus is the portion of the strain 2003). Raman spectroscopy shoots a beam of a
that is 90 out of phase from the stress. Since the laser to a molecule. During this process, inelastic
viscous deformation is not recovered upon the collisions of the laser beam and molecules will
removal of the stress, the energy causing the result in a gain or loss of energy of the scattering
deformation is transformed to heat due to friction, light from the molecule (Ekins et al. 2008). When
which is referred to as dissipation. Thus, loss the molecule is at the ground state, the inelastic
modulus characterizes the viscous properties of collision transfers some energy of the incident
polymers. Also, damping factor is the ratio of loss light to the molecule, becoming the excited state
modulus to storage modulus, describing the abil- (Fig. 2.8). This less intense scattering light
ity of a polymer to transform the applied mechan- (Stokes’ scattering) is measured by the detector
ical energy into heat. In general, amorphous glass and is generally used to characterize drugs and
and crystals have a damping factor of 0.2–3 and polymers.
0.01–0.1, respectively (Davis 1971). Complex With the addition of multivariate curve resolu-
viscosity is the ratio of complex modulus to angu- tion and PCA, Raman spectroscopy is able to
lar frequency, which can be useful to understand quantitatively measure the crystallinity of various
the rheological properties of polymers in an oscil- formulations. Lust et al. (2015) measured the
lation mode. Due to the Cox–Merz relationship, crystallinity of piroxicam ASDs during storage
complex viscosity can be correlated with shear using Raman spectroscopy. Results demonstrated
Fig. 2.8 Illustration of

elastic and inelastic
interactions between
electrons and an atom.
(Adapted with permission
Ekins 2008)
that Raman spectra were sensitive and accurate in traditional FTIR analysis, the material to be
detecting and measuring crystallinity. The combi- analyzed is blended with a diluent such as potas-
nation of FTIR and Raman mapping was used to sium bromide (KBr) and compressed into a trans-
study the drug release mechanisms of ASDs dur- parent pellet. This requires manual grinding of the
ing dissolution. The FTIR and Raman imaging sample prior to blending or concurrently with the
data revealed the distribution and diffusion rates KBr, making preparation of hard or elastic
of each individual component in the tablet matrix. materials problematic. Drug concentration within
Furthermore, Raman imaging detailed the local the pellet should be low, approximately 1%
composition and was able to capture various w/w. Once compressed, the samples should be
phase transitions on the surface of the tablets, analyzed immediately, and it is recommended
where it was in contact with dissolution media that a co-ground powder not be allowed to stand
(Puncochova et al. 2017). Tres et al. (2014) devel- for long periods of time prior to pelletization for
oped real-time Raman spectroscopy imaging in the purposes of stability. Products containing caf-
combination of multiple linear regression to feine were analyzed by blending the pharmaceu-
investigate the dissolution mechanism of tablets tical powder at a concentration of 1% w/w with
consisting of ASDs. Raman spectral analysis was KBr. Samples of this blend were then immedi-
used to study the homogeneity of the ASDs in the ately compressed into a thin pellet for analysis,
tablets containing 95% copovidone. Results while portions were stored for later analysis.
demonstrated that a homogenous ASD without Uniquely, the caffeine/KBr pellet proved to be
phase separation resulted in a consistent dissolu- stable once prepared for greater than 3 months,
tion profile. However, for tablets with only 50% displaying an identical concentration of caffeine
copovidone, the recrystallization of felodipine on day 90 as on the day of preparation with no
was observed after the polymer dissolved in alterations to the absorption spectrum. In contrast,
water using Raman spectroscopy. With the help storage of the prepared, uncompressed caffeine/
of real-time Raman imaging technique, the KBr powder revealed instability with pellets
recrystallization rate of felodipine was found to prepared 30 days post-blending, yielding drasti-
vary at different locations of the tablets. cally reduced caffeine concentrations (Baucells
et al. 1993). The preparation of biologic samples
in this manner is not recommended. In fact, KBr
Sample Preparation
compression has proven to lead to protein
Sample preparation will be dependent upon the
unfolding as well as protein aggregation and
type of FTIR analysis to be performed. For
54 X. Ma et al.
loss of enzymatic activity (Chan et al. 1996; therefore be determined on a case-to-case basis. It
Wolkers and Oldenhof 2005). Similar to tradi- should be noted that collection of the background
tional FTIR, diffuse reflectance IR (DRIFT) spectrum must be performed without the pressure
requires the use of a diluent in sample prepara- applicator engaged for powder samples.
tion. KBr or a similar species such as KCl has
been used during analysis. However, unlike tradi- Polymorph Screening
tional FTIR, the sample does not require mechan- As mentioned above, ATR–FTIR requires no
ical compaction into a pellet. DRIFT does require sample preparation prior to analysis save minor
that the blend be of uniform particle size for compression to ensure contact with the IRE. As
accuracy, creating difficulties for materials that such, the potential for process-induced polymor-
are difficult to triturate and materials that may phism during sample preparation is limited,
be chemically altered by the grinding process. making the technique advantageous in polymorph
Additionally, samples such as the caffeine screening. Analysis of the fingerprint region of
discussed above require immediate analysis to the IR spectrum allows for identification of char-
ensure stability of the physical blend (Hartauer acteristic absorption bands which are unique to
et al. 1992; Sitar Curin et al. 1997). each polymorphic form. When a given poly-
Attenuated total reflectance FTIR (ATR– morph is analyzed alone, spectral differences are
FTIR) employs a crystal surface upon which the readily apparent due to broadening of bands or
material to be analyzed is placed. For information the appearance and disappearance of absorption
regarding ATR–FTIR theory, the reader is bands. Additionally, it has been shown that the
referred to Buffeteau et al. (1996). ATR–FTIR summated absorption spectra of blends of
has the unique advantage of minimal to no sample polymorphs allow for identification of the species
preparation prior to analysis. Powder blends or present. In fact, a study of the three ganciclovir
ground formulations (i.e., ground extrudates) do polymorphs demonstrated that mathematical
not require the addition of supplemental materials addition of the pure polymorphic spectra matched
such as KBr. Additionally, it is possible to place that of experimental blends of the polymorphs,
an unground sample directly on the crystal for thus allowing their accurate identification in
analysis, a method useful in assessing homogene- formulations (Salari and Young 1998).
ity at the surface of a product. The sample to be
analyzed is placed directly on the internal reflec- Excipient Interactions
tion element (IRE), a high-refractive-index crys- Appearance of new absorption bands, broadening
tal such as diamond, and pressure is applied top of bands, or alterations in intensity are the pri-
down using a pressure applicator fit to the ATR– mary events associated with excipient interaction
FTIR in use. The crucial element in ATR–FTIR (Abbas et al. 2008). For an experimental com-
sample preparation is ensuring an even and repro- pound BG 637, superimposition of the IR spectra
ducible contact between the sample and internal of the drug obtained by DRIFT analysis over
reflection element. It has been shown that API–excipient blends yielded no alterations in
improper contact will yield inconsistencies in the spectra, indicating no interactions with the
absorption band intensities. This includes insuffi- tableting excipients tested. This result was con-
cient contact (i.e., too little pressure) as well as firmed via DSC and X-ray diffraction.
too strong contact with the latter, capable of dam- ATR–FTIR analysis of omeprazole sodium
aging both the sample and the crystal used as the isomers and isomer–mannitol blends was
IRE (Buffeteau et al. 1996; Salari and Young performed not only on raw powders but also on
1998). The pressure devices supplied by compacts thereof. While the spectra of the pow-
manufacturers are graduated nominally, such as der mixtures did not display any differences, disk
a scale of 1–5 as that on the Foundation Series samples (1 cm diameter, 2 mm thickness)
ATR attachments from Thermo Fisher Scientific. prepared via compression at 7 tons for 5 min
Selection of the correct application pressure must proved unique, indicating interaction with
mannitol. Samples of the R-isomer compact Parameter Selection

displayed two distinct peaks corresponding to During sample analysis, the operator will be
amino and imino group stretching (3425 and required to input a number of parameters for
3318 cm–1, respectively), while the S-isomer each scan, including the scan range in degrees
compact solely displayed stretching on a 2Θ scale, the step size (degrees per step),
corresponding to the amino group (Agatonovic- and the count time for each step (often referred to
Kustrin et al. 2008). Such a method allows analy- as dwell time). Assigning the proper parameters is
sis of powder blends used for tableting as well as crucial to ensure that adequate peak shape is
the final formulation in solid form. obtained while minimizing processing time. For
the first scan of a new API, it is necessary to
analyze a very broad range, such as 5–120 on
the 2Θ scale. This will allow for identification of
2.2.3 X-Ray Diffraction
the characteristic crystalline peaks, and
subsequent analyses should be shortened to
XRD is the measurement of the intensity of
include only the footprint region.
X-rays scattered by electrons bound to atoms
Twenty data points per peak are desired to
and the corresponding phase shifts that occur as
ensure adequate peak shape. In order to obtain
a result of the position of the atom. For a detailed
this value, the researcher must alter the step size
explanation of XRD theory, the reader is referred
of each scan. For highly crystalline materials, a
to Dinnebier and Billinge (2008). XRD is typi-
step size of 0.02 is adequate to meet this require-
cally a nondestructive test (i.e., the analyzed
ment. For materials exhibiting broad peaks, this
material can be recovered) and, as will be
value can be increased to shorten the run time
discussed, is highly useful for determining
while still maintaining the 20 data points per
differences in crystal structure (i.e., polymorphs),
peak. In order to optimize the step size, the pow-
drug–excipient interactions, and identifying
der should be analyzed over a very narrow range,
amorphous systems.
such as 2–5 2Θ, in a region containing a charac-
Prior to analysis, it is necessary to calibrate
teristic crystalline peak(s) over a range of step
and optimize the device to be used for testing.
sizes. Cameron and Armstrong analyzed quartz
This can be done by use of various reference
from 67 to 69 2Θ and varied the step size from
standards such as those offered by the National
0.1 to 0.01 2Θ in order to find the optimum step.
Institute of Standards and Technology (NIST).
Identification of the step size in which the
Standard Reference Material 674b consists of
diffractogram decays to baseline following the
four oxide powders to be used as internal
peak was selected as optimum (Cameron and
standards or calibrators for an XRD unit: ZnO,
Armstrong 1988). If upon return to baseline the
TiO2, Cr2O3, and CeO2. Alternatively, standards
inception of a new peak is immediate, a shorter
of materials with well-understood diffraction
step size should be selected to increase resolution.
patterns can be used such as alumina, mica, or
Oftentimes, the step size and dwell time are
silicon pellets.
considered together as a scan rate. For example,
One important consideration with XRD is the
analysis of tenofovir disoproxil fumarate (TDF)
limit of detection for trace crystallinity. This limit
was reported as being conducted from 4 to 40 at
has been reported at values ranging between
a rate of 4 /min (Lee et al. 2010). Prevalent rates
1 and 5% w/w (Nunes et al. 2005; Rumondor
in the literature are between 0.5 and 4 /min.
et al. 2011). Radiation via synchrotron source
The final parameter that must be determined is
has begun to be utilized that has a demonstrated
the dwell time. While a longer dwell time will
LOD of 0.2% for analysis when levels of trace
increase the signal-to-noise ratio and improve
crystallinity are critical (Nunes et al. 2005).
counting statistics, it will also drastically increase
the duration of a run. This is especially true for
56 X. Ma et al.
methods utilizing small step sizes. Therefore, in and a step size of 0.014767 and a 2-s dwell time
order to ensure timely analysis, the shortest dwell (Kirk and Blatchford 2007). In another work, a
time, which provides a strong signal of all char- variable temperature cell was utilized, and the
acteristic peaks without significant interference XRD pattern of mebendazole was analyzed as a
from the baseline, should be selected. This param- function of temperature to assess polymorphic
eter will be most influential for systems lacking transformations that occur due to temperature
strong crystallinity, such as pharmaceutical variation (de Villiers et al. 2005). Indeed, it was
systems containing a polymer, in which case lon- seen that at temperatures above 180 C, a trans-
ger dwell times will aid peak structure. Dwell formation to the more thermodynamically stable
times of 1–5 s can be employed, with times of polymorph occurred.
1–3 s being most prevalent in practice. Olanzapine can crystallize into 25 different
crystalline structures of which 7 are active
Polymorph Screening pharmaceutically. Identification and quantifica-
As mentioned earlier, XRD analysis reveals phase tion of each polymorph are necessary throughout
shifts that occur as a result of atomic position the development cycle of a formulation
within a material. Therefore, alterations in crystal containing such a compound to ensure that the
structure that arise as a physical result of poly- final product is of acceptable quality. Tiwari et al.
morphism can often be detected by XRD. This examined two polymorphs of the compound by
can be observed as a shift in a major characteristic XRD. Prior to analysis, the unit was calibrated
peak or the appearance or disappearance of peaks with a silicon pellet. The scan was then optimized
in the diffraction pattern. XRD analysis of the by varying the dwell time and step size such that
three polymorphs of TDF reveals that the diffrac- the maximum number of identifiable peaks was
tion patterns for forms B and I are similar, making obtained. As can be seen in Table 2.2, a dwell
their definite distinction difficult. However, form time of 5 s with a step size of 0.05 produced the
A contains multiple peaks not present in the dif- greatest number of identifiable peaks in the
fraction patterns of the other polymorphs as well shortest amount of time. In order to accurately
as the absence of one characteristic peak near 30 , quantify the amount of a given polymorph present
making its identification absolute (Fig. 2.9). Kirk in a mixture, each form must be analyzed as a
et al. demonstrated that three of four generated pure sample. The highest-intensity peak is then
lactose polymorphs crystallize with monoclinic selected, and the intensity of the pure polymorph
unit cells by XRD employing a range of 5–40 is set to 100%. The ratio of intensities for the
Fig. 2.9 XRD patterns of

three polymorphic forms of
TDF. Forms B and I are
similar, while Form A is
unique. Reproduced with
permission from American
Chemical Society
Intensity
TDF A
TDF B
TDF I
4 9 14 19 24 29 34 39
selected peak in the mixture compared to the pure physical mixtures identical in drug/polymer ratio
polymorph provides the percent present in the to that used in processing must be analyzed to
mixture. This, however, cannot be employed in ensure that the pattern is validated. Tobyn et al.
the case of olanzapine, as the highest-intensity analyzed an experimental compound as a solid
peaks overlap for multiple forms (Tiwari et al. dispersion with PVP at various drug-to-polymer
2007). However, screening of mannitol ratios. As can be seen in Fig. 2.10, the bulk drug
demonstrated that even with similar polymorphic product is highly crystalline with many character-
forms, coupling sample rotation with particle-size istic peaks (parameters: 2–60 2Θ, 2 /min scan
reduction increases the ability to differentiate the rate, calibration with mica and alumina reference
forms present. In the case of mannitol, it allowed standards). At a 50:50 ratio of drug-to-PVP, the
identification of the individual components down formulation produced an amorphous halo,
to approximately the 1% level (Campbell Roberts indicating amorphization of the drug. This was
et al., 2002). confirmed using a dry blend of the materials at the
same ratio. While a decrease in peak intensity can
Excipient Interactions be observed for the 50:50 physical mixture, the
Changes in the XRD pattern may occur as a result peaks are still present, indicating that processing
of drug–excipient interactions. Such alterations indeed rendered the drug amorphous. Addition-
include the conversion to a unique polymorphic ally, it can be seen that at higher drug loading
form or amorphous to crystalline transitions. For levels, the process did not generate or stabilize the
example, a sample of pure β-form carbamazepine drug in the amorphous form (Tobyn et al. 2009).
displays a characteristic peak at 13 , while the
α-form displays this peak as well as an additional Pair Distribution Function (PDF)
peak at 8.8 2Θ. In a study of physical mixtures of The pair distribution function (PDF) is a recent
pure β-form carbamazepine in combination with application of XRD to further characterize amor-
various tableting excipients, it was seen that the phous drug systems by identification of the amor-
characteristic peak at 13 was present in all phous phase and quantifying the relationship
samples. However, following storage at 55 C between ordered and disordered systems (Bates
for 3 weeks, a combination of carbamazepine et al. 2007 and Newman et al. 2008). Convention-
and stearic acid displayed the additional peak at ally, a crystalline polymorph has an identification
8.8 , indicating a conversion to the α-form. This fingerprint captured by XRD. However, the amor-
was attributed to partial solubilization of carba- phous form lacks long-range structure, and thus a
mazepine in steric acid at elevated temperatures fingerprint cannot be captured by typical means.
followed by recrystallization upon cooling (Joshi The PDF approach describes the probability of
et al. 2002). Indeed, stability studies, to be finding any two atoms at various interatomic
discussed later, often rely on alterations in XRD distances. Powder diffraction data is collected
patterns as an indication of formulation stability. with low noise over the momentum transfer
An amorphous material will yield an XRD range, Q. Greater information and reproducibility
pattern termed a “halo” which is a gradual rise can be ascertained over a broader Q range. Q is
and fall of the baseline with no distinct peaks. As related to the Bragg angle, θ, by the formula
many processing techniques focus on rendering Q ¼ 4π sinθ/λ where λ is the wavelength of the
the drug amorphous, XRD can be utilized to incident radiation. It follows that a shorter wave-
identify whether or not a drug or formulation is length source would yield a broader Q range so a
amorphous. This can be complicated when poly- shorter wavelength molybdenum source with a
meric systems are used, as most polymers used in Qmax of ~16 Å1 is strongly preferred to a typi-
pharmaceutical applications are amorphous in cal copper source with Qmax of ~8 Å1. The
nature. At low drug loadings, the amorphous sig- synchrotron source mentioned previously can
nal from the polymer may overshadow or mask measure a range up to a Qmax of 45 Å1. Fig-
the crystalline pattern of the drug. As such, ure 2.11 shows an example of the effects of a
58 X. Ma et al.
Fig. 2.10 XRD patterns of

experimental compound
BMS-488043 bulk
material, physical blend
with PVP, and solid
dispersions thereof. 50:50 API:PVP Dry Blend
Reproduced with
permission from John
Arbitrary Counts
Wiley and Sons
50:50 API:PVP Dispersion
80:20 API:PVP Dispersion
Crystalline BMS-488043
0 5 10 15 20 25 30 35
2 Theta
Fig. 2.11 Representative

PDFs of amorphous
carbamazepine generated
with Qmax values of
20 Å1 (a) and 2.8 Å1 (b).
Reproduced with
permission from Springer
Qmax of 20 Å1 and 2.8 Å1 with the latter DSC and typical XRD pattern. Application of
having poor reproducibility of the amorphous PDF showed only small random fluctuation in
fingerprint (Dykhne et al. 2011). the 4–20 Å region, indicative of a phase separated
Phase separated systems can pose a problem system. It was suggested that a solid
for conventional solid-state characterization nanosuspension had formed (Newman et al.
methods when the Tgs of the components overlap 2008). This form of analysis can be utilized as a
or when the phase separated domains are suffi- miscibility tool for screening polymers. A PDF
ciently small. In these instances, a single Tg can study of 30 and 70% poly(acrylic acid) (PAA) or
be detected by DSC, and the XRD scan will also poly(vinyl pyrrolidone) (PVP) with felodipine
be amorphous. A trehalose–dextran dispersion demonstrated that PAA was immiscible with
was analyzed and found to have a single Tg by felodipine at these concentrations but miscible
Fig. 2.12 A, two instrument designs; B, height measuring mechanisms. (Adapted with permission from Eaton and West
2010)
with PVP. The PAA samples had little to no dominant length scales, which can provide useful
differences between the measures and calculated information for understanding the materials. As
(from individual components) diffractograms, shown in Fig. 2.12, there are two AFM designs,
while the PVP samples showed large differences indirectly measuring the height changes. In AFM,
between the calculated and measured the sharp tip is attached to a flexible
diffractograms (Rumondor et al. 2009). microcantilever through a spring, which is bend-
able under applied forces. When the tip touches
the sample surface and moves around, the tip will
2.2.4 Atomic Force Microscopy (AFM) press or pull the spring, causing the upward or
downward bend of the cantilever. This small
Atomic force microscopy is a powerful tool to change can be detected by the laser disturbance
analyze the surface of a subject in three between the laser projector and photodetector, as
dimensions at a nanoscopic scale. AFM has a shown in Fig. 2.12b. There is another movement
wide range of applications to various materials due to the lateral force, which torques the tip,
regardless of their opaqueness or conductivity leading to a twist of the cantilever. This can be
(Binnig et al. 1986). Compared to conventional measured on the photodetector via the horizontal
light microscopy, AFM does not require light, movement of the laser. In order to obtain less
and it measures the changes of height when a noisy data, another type of AFM is developed
sharp-solid force-tip touches the sample. By (Fig. 2.12a right side), where the laser is fixed,
analyzing those changes, different colors can be and the cantilever is scanned. An extra lens
assigned to different heights. Therefore, the between them keeps the laser light being focused
images collected from AFM contain invaluable on the cantilever. This mechanism of action
information about various properties of the sam- utilizes the feedback from the cantilever to offset
ple (Sarid 1994). AFM can be operated under variations in height, keeping the pressing force
various conditions, such as in air, liquid, or vac- constant (Haugstad 2012).
uum. It has been widely adopted for many scien- AFM is extremely sensitive to detect small
tific fields (e.g., hard and soft materials science, changes in the molecular structure (Bhushan and
nanotechnology, biology, pharmaceuticals sci- Fuchs 2007). AFM measures four types of
ence). The highest lateral resolution of AFM can mechanical properties of the sample (height, stiff-
reach one nanometer, which offers extreme ness, adhesion force, and friction) and is used as a
details of the sample (Sarid 1994). In conjunction tool to differentiate sample due to its extreme
with the Fourier analysis, AFM can identify sensitivity. The combination of those mechanical
repeating components (crystalline lattices), properties enables the comprehensive under-
molecular or atomic crystal structure, and standing of the materials. For instance, an
60 X. Ma et al.
Fig. 2.13 Tip-sample demonstration of a sample in a force-curve cycle. (Reproduced with permission from Eaton and
West 2010)
amorphous drug can act as a plasticizer, softening surface of the microspheres. AFM results
the mixture, while a crystalline drug is more rigid. discerned the structural and compositional hetero-
The difference in stiffness can be used to distin- geneity in different drug-loading samples.
guish amorphous drug from crystalline drug and Matthias et al. used AFM to investigate the sta-
gives the information about crystallinity and bility of amorphous fractured films and success-
miscibility of the materials (Liu et al. 2017). As fully quantified the de-mixing phenomenon by a
illustrated in Fig. 2.13, when the tip is far away phase separation analysis. This study
from the sample (step 1), the force between the tip demonstrated the capability of AFM to achieve
and the sample is zero. By moving the tip to the a high-resolution analysis of the miscibility of
surface of the sample, an attractive force is ASDs. However, these experiments are time-
detected, causing the tip jump to contact the sam- consuming, which limits the application of AFM
ple (step 2). The tip is continuously pressed until (Lauer et al. 2011).
the maximum approach point is reached, and the
sample exerts a repulsive force to the tip because
of the deformation. After the tip returns to the
2.2.5 Specific Surface Area
zero-indentation point, the subsequent adhesive
force will pull the tip back to the sample, causing
Many formulation approaches have been applied
a negative force until it reaches the maximum
to overcome aqueous solubility issues including,
pulling force of the cantilever. After that, the tip
but not limited to, the addition of solubilizing
will jump back to the initial state.
agents (i.e., Cremophor EL), complexing agents
mDSC and AFM were used together to study
(i.e., cyclodextrins), or the generation of a salt
the phase behavior and drug distribution of ter-
form of the API. As an alternative to formulating
nary amorphous solid dispersions (ASDs)
the API in one of these novel ways, many
containing drug and polymers (Meeus et al.
researchers are focusing on particle engineering
2015). Time-of-flight secondary ion mass spec-
techniques which decrease particle size or gener-
trometry (ToF-SIMS) and AFM revealed the
ate a porous system as a means of increasing
differences in the drug distribution of three
aqueous solubility.
formulations, showing that a PVP-rich phase
Substantial particle-size reduction, or the for-
was formed under the surface of the
mation of a porous system, will yield a significant
microspheres, and the drug existed as a glass
increase in surface area for the powder
solution. In addition, a poly(lactic-co-glycolic
(Fig. 2.14). The Noyes–Whitney equation
acid) (PLGA)-rich phase was captured on the
describes the dissolution velocity as:
Fig. 2.14 Surface area Surface enlargement factor

enlargement as a result of
particle-size reduction.
From Junghanns and
Mueller (2008) × 10 × 1000
100 µm 10 µm 100 nm
dW DAðC s CÞ detailed summary of the theory behind BET iso-

¼
dt L therm analysis, the reader is referred to
Condon (2006).
Here, dW/dt is the dissolution rate, A is the
surface area, C is the API concentration in the Sample Preparation
bulk dissolution media, Cs is the concentration of As vials for BET analysis are often extremely
API in the diffusion layer around the solid, D is narrow at the opening, it is imperative that care
the diffusion coefficient, and L is the diffusion be taken not to damage or alter the sample during
layer thickness. It can be observed that an loading. This is especially true for porous
increase in the surface area value, A, will increase materials which may be collapsed if pressure is
the dissolution velocity (Junghanns and Mueller applied and brittle systems which may fracture
2008; Lenhardt et al. 2008). Furthermore, a sub- during loading. Fracture may yield inaccurate
stantial decrease in particle size will cause an higher surface areas, while collapse results in
increase in dissolution pressure resulting in substantially reduced values.
higher saturation solubilities (Mosharraf et al.
1999). Degassing The removal of gases and vapors
Measurement of the surface area may therefore physically adsorbed onto the surface of the pow-
offer insight into the relative dissolution rate of an der being tested, termed degassing, is essential
engineered or processed powder, serving as a prior to determining the specific surface area
screening process between batches or production (SSA, i.e., m2/g) of any sample. Failure to do so
methods. may result in a reduction in the calculated surface
area or a high variability in obtained SSA values
(USP 32/NF 27 General Chapter 846). It has been
BET Surface Area Analysis suggested that an adsorbed impurity will not alter
Surface area (SA) analysis of pharmaceutical
the BET surface area calculation, provided its
powders is often assessed via gas adsorption
boiling point is at least three times higher than
methods. In these methods, a nonreactive gas is
the adsorption temperature (Joy 1953). However,
adsorbed to the surface of the material at or near
such impurities may influence the calculated heat
the boiling point of the liquid adsorptive at a
of adsorption (C-value). Therefore, in order to
single or multiple pressures. The quantity of gas
ensure accurate, reproducible surface area
molecules required to form a monolayer of
measurements, proper sample preparation must
adsorbed gas on the powder surface is deter-
be employed.
mined, and using the average diameter of a single
molecule, the SA is determined (Condon 2006).
Due to factors such as temperature sensitivity,
This is done via mathematical modeling of the
variations in surface energies, particle sizes, and
adsorption isotherms (amount
porosity, no single method may be applied uni-
adsorbed vs. adsorptive pressure), with the most
versally that guarantees complete removal of
established model being BET analysis. For a
adsorbed gases and vapors to the powder surface;
62 X. Ma et al.
however, two methods are often applied: vacuum Sing et al. 1985). Allen (1997) states outgassing
pumping and purging by use of an inert gas is complete if, following 15 min of isolation from
(Lowell and Shields 1991). the vacuum, no pressure increase is observed
When applying a vacuum, a pressure of 105 upon reintroduction to vacuum, while Igwe
Torr has been stated as sufficient in outgassing deems completion as maintaining a pressure
procedures. The application of elevated between 104 and 105 mmHg for a substantial
temperatures will increase the rate at which period of time following isolation from the
impurities/contaminants leave the powder sur- pumping line and cold trap (Igwe 1991; Allen
face, reducing the time required to hold the vac- 1997).
uum (Fagerlund 1973). However, heat-liable As mentioned earlier, purging with an inert gas
products must be monitored with care, and glass may be used to clean the sample. In this case, the
vessels used in most BET equipment have a effluent gas may be monitored via mass spectros-
threshold of 400 C (Igwe 1991). Additionally, copy or a thermal conductivity detector which
elevated temperatures may cause rapid water may detect impurities in the range of a few parts
loss from the sample surface, resulting in altered per million (ppm). Once the level of contaminants
morphology or sample collapse. Samples that are and foreign materials is below detection, adequate
sensitive to heat should be degassed at ambient removal of the adsorbed material has been
temperatures under vacuum for long periods of achieved (Sing et al. 1985; Lowell and Shields
time, such as 12–24 h, in order to reach constant 1991).
weight (Lowell and Shields 1991; Engstrom et al. Again, no single method can be employed
2007). universally to properly degas a sample prior to
Alternatively, flushing the powder sample analysis. Understanding the morphology (i.e.,
with an inert gas may also be used to clean the porosity) and thermal sensitivity of a powder
surface. This purge gas must be of extremely high will aid in determining degassing time and tem-
purity and dry so as not to introduce contaminants perature as a more porous network will require
or induce moisture-related phenomena such as increased times. However, once a sample has
crystallization. Heat may also be provided in been run under a given set of parameters, it is
this method as well to aid in desorption (Sing recommended that a second sample be run using
et al. 1985). For samples in which heat may not an extended degassing time (i.e., 1.5–2 the orig-
be applied, multiple absorption–desorption cycles inal time). If the results are statistically signifi-
may be employed to clean the surface with three cantly different, the first results may be discarded
to six cycles providing sufficient cleanliness for and the second set of parameters selected. Addi-
reproducible measurements (Dios Lopez- tionally, it is recommended that powders be
Gonzalez et al. 1955). This may also be done analyzed at least in triplicate using three individ-
using gas blends such as nitrogen (10% v/v) in ually prepared samples, especially for powders
helium (Pendharkar et al. 1990). with a large particle-size distribution.
Regardless of the chosen degassing method,
the sample must be monitored by some means to Sample Analysis
ensure degassing is complete. With regard to Prior to any sample analysis, the BET unit to be
vacuum pumping, literature often describes the used should be properly calibrated for accurate
method of ensuring all adsorbed gases and vapors results. This can be performed using commercial
are removed as monitoring of the system pres- reference standards such as silica, garnet, or kao-
sure. Webb and Orr (1997) state that isolated linite (Antila and Yliruusi 1991). These can most
samples displaying a pressure rise of <103 readily be purchased from the manufacturer of the
Torr/minute are indicative of adequate degassing. equipment in use, such as the alumina pellets
Similarly, Gregg and Sing (1982) state a pressure offered by Quantachrome and range in SSA
of 104 Torr as sufficient, while Sing states values from below 1 to greater than 150 m2/g.
~10 MPa is satisfactory (Gregg and Sing 1982; Standard selection will depend upon the samples
to be analyzed, as the standard SSA should be submergence of the sample; failure to do so will
near that of the samples. For example, Swinkels result in inaccurate results. Data reduction can
et al. (1994) used a kaolinite reference standard then be performed and the results reported as
with a specific surface area of 16.2 m2/g to verify m2/g material.
that no drift had occurred prior to sample analysis
of the experimental samples (SSA 26–32 m2/g).
Following calibration, sample analysis can be 2.2.6 Solid-State Nuclear Magnetic
performed. As mentioned earlier, care must be Resonance (ssNMR)
taken when adding the sample to the analysis
bulb, and the weight of the added powder must Solid-state nuclear magnetic resonance (ssNMR)
be known. The analysis bulb should be weighed is a characterization tool that can be used to
empty on an analytical balance, the powder then provide detailed structural information for amor-
added, and the final weight of the bulb measured phous dispersions. Measurements of dipolar cor-
and recorded. Subtraction of the bulb weight from relation, spin diffusion, and relaxation
the final weight of the filled bulb allows accurate measurements can be utilized to assess recrystal-
determination of the true amount of sample lization rates, phase separation (Guo et al. 2013),
added. The amount of powder added will depend polymorphic form identification (Dempah et al.
on density, as most commercial BET units require 2013), and solubility of a drug in a carrier.
a minimum surface area rather than minimum Recrystallization rate information for solid
weight. For example, the Nova® Series offered dispersions can be determined from 1H NMR
by Quantachrome requires a minimum of 0.01 m2 relaxation times. A study of amorphous nifedi-
for accurate analysis. While sample sizes may be pine, phenobarbital, and flopropione by this tech-
low for high-surface-area powders such as those nique showed a good correlation between
produced by spray freezing into liquid (200 mg relaxation times, molecular mobility, and recrys-
(Rogers et al. 2002)), higher amounts are neces- tallization rates. Nifedipine had a shorter relaxa-
sary for lower-surface-area materials (5 g tion time suggesting higher molecular mobility
(Swinkels et al. 1994)). Sample degassing can compared to phenobarbital and flopropione. The
then be performed utilizing the appropriate stability study of these samples resulted in sub-
method from those outlined above. Post- stantial recrystallization of the nifedipine mate-
degassing, the weight of the sample should rial, while no recrystallization was observed for
again be taken to ensure accuracy when calculat- the other two amorphous drugs during the same
ing SSA as some weight will be lost during the time frame (Aso et al. 2000). In another instance,
degassing process (Sing et al. 1985). Analysis is the temperature dependence of molecular mobil-
then carried out with nitrogen often used as the ity was studied for a nifedipine–PVP system. The
adsorptive gas; however, for powders exhibiting relaxation times of the dispersion were obtained
extremely low SSA values, it is necessary to use as a function of temperature. A dramatic decline
an adsorptive gas of lower vapor pressure. In such in the relaxation time was observed starting at
cases, krypton can be employed. It must be noted about 20 C below the Tg of the dispersion.
that differences in SA values will be obtained Thus, temperatures above this point have
based on the gas used for the study as the size of increased molecular mobility and could serve as
the gas molecules forming the monolayer will an explanation for recrystallization observed in a
differ resulting in differences in pore penetration number of studies (Yuan et al. 2013).
(Sandell 1993). ssNMR investigation into associations
During analysis, the sample is repeatedly between amorphous drug and polymer can reveal
lowered into a liquid nitrogen bath. While this details about the molecular interactions as well as
process is automated by today’s units, the phase separation of a dispersion. Molecular
researcher is still required to maintain the level interactions of interest include stabilizing hydro-
of liquid nitrogen in the Dewar to ensure proper gen bond interactions between the drug and
64 X. Ma et al.
polymer in the dispersion. An initial ssNMR contain fluorine atoms. This allows selective and
study by 13C-1H cross-polarization of drug sub- sensitive analysis by 19F-NMR. For example,
stance BMS-488043 with PVP was initially inclu- flufenamic acid, a fluorine-containing drug sub-
sive for molecular interactions. However, stance, was prepared as solid dispersions with
subsequent analysis by 1H magic angle spinning PVP and HPMC. Molecular mobility information
(MAS) NMR showed that the NH molecules of was determined, and through correlation to crys-
the drug substance had strong hydrogen bonding tallization tendency it was found that HPMC
with the carbonyl group on the PVP chains ulti- samples had faster crystallization than PVP
mately leading to increased stability in the amor- samples (Aso et al. 2009). Similarly, ezetimibe,
phous state (Tobyn et al. 2009). A similar study another fluorine-containing drug substance, was
was conducted for ketoprofen and polyethylene used as a dispersion with mesoporous silica in
oxide (PEO). Strong hydrogen bond interaction one study. A 19F CP-MAS was found to be spe-
between the components was responsible for the cific and sensitive for F–Si interactions between
high miscibility of the system (Schachter et al. the drug and polymer (Vogt et al. 2013).
1
2004). 2D H-13C cross-polarization Another interesting application of ssNMR was
heteronuclear correlation (CP-HETCOR) can be to assess gabapentin polymorphism and its impact
employed directly to prove formation of a glass on chemical stability. Initially, 13C-ssNMR data
solution. A trial was conducted using acetamino- was collected for samples of gabapentin under
phen and indomethacin as amorphous dispersions differing milling conditions to identify the vari-
with various polymers. CP-HETCOR was able to ous polymorphic forms present. A previously
verify that the dispersion was present as a glass unknown polymorph of gabapentin was discov-
solution (Pham et al. 2010). 15N ssNMR has been ered during this analysis. Samples were then
used to identify the ionic interaction between analyzed for relaxation time by 1H-NMR. The
lapatinib and HPMCP, which serves as the mech- samples with the shortest relaxation times (and
anism of action for HPMCP to inhibit the crystal- greatest mobility) also had the largest amount of
lization of amorphous lapatinib, especially at lactam degradant present. It was concluded that
elevated drug loadings. Additionally, by maximizing the chemical stability of gabapentin
analyzing the ssNMR results, the optimal drug– is aided by control of the polymorphic form dur-
polymer ratio was chosen at 40% where solid ing the milling process (Dempah et al. 2013).
solution was achieved (Song et al. 2015). A 2D Amorphous lactose was similarly studied, and it
double-quantum NMR was deployed with the was found that cryoground samples had shorter
ultrafast MAS technique to investigate the molec- relaxation times than amorphous samples
ular interactions between clofazimine and prepared by lyophilization and spray-drying.
HPMCP. With the aid of computational chemis- This serves as another example of process meth-
try, the proton transfer between clofazimine and odology choice having impact on mobility and
carboxylic acid groups in HPMCP has been potentially stability of a material (Lubach et al.
identified. By conducting principal component 2007).
analysis, ssNMR spectra were interpreted statisti- Conventional ssNMR approaches are limited
cally, and the optimal drug loading was deter- by sensitivity and time of analysis. However,
mined for clofazimine-HPMCP ASDs (Nie et al. recent advances have been made into the applica-
2016). tion of dynamic nuclear polarization (DNP) to
Drug substances containing fluorine atoms ssNMR. Under cryogenic conditions, electron
provide unique advantages for analysis by spin polarization from a polarizing agent is trans-
ssNMR. NMR methods for fluorine are desirable ferred to nuclear spins of the compound(s) of
as it has high sensitivity due to its natural abun- interest through the use of high-frequency micro-
dance and high resonance frequency. Addition- wave sources. The increased nuclear spin
ally, there are a number of drug substances amplifies the signal output of the ssNMR experi-
containing fluorine atoms, but few excipients ment. At present, the DNP approach can enhance
the sensitivity by a factor of up to 102 and accel- application of ssNMR was in a study to investi-
erate the acquisition time by a factor of up to 104 gate the enthalpy relaxation time and spin–lattice
for a multidimensional experiment. These relaxation time. Results illustrated that there was
enhancements enable the application of ssNMR no observable crystallization at temperatures that
to be used for rapid studies and high-throughput were 20–30 C lower than the glass transition
screening of biological samples (Griffin and temperature of the drug. At any temperature
Prisner 2010). Applications of DNP have been below the glass transition temperature, nifedipine
used to study surfactant molecules in mesopores showed a smaller T1ρ than the others, indicating
of silica nanoparticles, interactions involving end the faster crystallization rate of nifedipine either
groups of polymer chains (previously unobserv- above or below its glass transition temperature
able by ssNMR), characterization of indometha- (Aso et al. 2000). Moreover, researchers used
cin and diflunisal–PVP ASDs, and determination 13C-NMR to study the effect of moisture on
of domain size distribution of amorphous molecular mobility (Aso et al. 2002). The molec-
cetirizine in tablets (Paudel et al. 2014). Addition- ular mobility of ASDs was measured in the pres-
ally, a solvent-free sample preparation ence of moisture. The T1ρ of the neat drug
incorporated biradical bis-TEMPO terephthalate, decreased with moisture, while for the drug in
a polarizing agent, with ortho-terphenyl, a model an ASD with polymers, the T1ρ of the drug did
compound, by melt-mixing. The experimental not significantly increase. Results indicated that
conditions were modified such that samples the drug in the ASD was stabilized by polymers
containing both amorphous and crystalline through molecular interactions. Studies of poly-
model were analyzed by DNP-ssNMR with signal mer miscibility illustrated that ssNMR is a pow-
enhancements of 58 and 36, respectively. How- erful tool for miscibility characterization.
ever, the authors highlight that separation of the However, it has not been widely adopted because
polarizing agent from the sample is an issue and of its poor cost-effectiveness. The miscibility of
an area of improvement for DNP (Ong et al. nifedipine and PVP was studied using ssNMR. T1
2013). DNP has also been applied to polymorph and T1ρ were measured to confirm the miscibility
analysis of theophylline. However, minimal gains between nifedipine and PVP. The domain size of
were observed with only form II showing signifi- nifedipine was calculated, which was comparable
cant enhancement, likely due to methyl group to that of PVP in ASDs (Yuan et al. 2013).
rotation. Thus, DNP may not provide the same
enhancement benefit for all sample systems
(Pinon et al. 2015).
2.2.7 Residual Solvent Analysis
Solid-state 1H NMR was applied to compare
the mobility of indomethacin and nifedipine as
Many synthesis processes as well as formulation
well as their ASDs. The relaxation behavior of
production processes used for PWS drugs such as
nifedipine ASDs was found to be correlated with
spray-drying, roto-evaporation, precipitation
the environmental temperature, which explained
methods, and freeze-drying methods require the
the poor stability of nifedipine ASDs compared to
use of various organic and inorganic solvents.
indomethacin ASDs (Forster et al. 2003). How-
The physicochemical properties of the material
ever, a higher molecular mobility may not relate
in use such as crystallinity have proven to influ-
to a higher crystallization rate. By measuring the
ence the amount of residual solvent levels within
enthalpy relaxation time of various pharmaceuti-
the product (Witschi and Doelker 1997). As many
cal glasses, no strong correlation was found
of these solvents are highly toxic, it is imperative
between the relaxation parameters and the crys-
that they be present only at extremely low levels,
tallization tendency of the molecules. It is
if at all, in the final product so as to reduce the risk
concluded that molecular mobility alone is not
to the end user. The following sections discuss
sufficient to predict nucleation and crystal growth
residual solvent acceptance limits as well as
rates of ASDs (Tombari et al. 2008). Another
66 X. Ma et al.
methods for determination of residual solvents ionization detection (GC-FID) as the preferred
present in final products. method of detection and quantitation. Addition-
ally, it is specified that a 0.32 mm 30 m fused-
silica column coated with a 1.8-μm layer of phase
Residual Solvent Guidelines
G43 be used for analysis. This may be substituted
The primary concern for residual solvents in phar-
with a 0.53 mm 30 m wide-bore column coated
maceutical products is organic volatile impurities
with a 3.0-μm layer of phase G43. Carrier gasses
(OVIs). ICH guideline Q3C (R4) from February
of nitrogen or helium are acceptable. It is stated
2009 categorizes OVIs into three classes: class
that the column temperature be maintained at
1 includes solvents that are known or suspected
40 C for 20 min and then raised at 10 C/min
human carcinogens as well as environmental
to 240 C. This temperature is then held for
hazards; class 1 solvents should be avoided if
20 min. Injection port and detector temperatures
possible. Class 2 solvents include solvents that
of 140 C and 250 C are specified, respectively.
are nongenotoxic animal carcinogens or possible
Additional methods such as GC-mass spectros-
causative agents of irreversible toxicity (i.e., neu-
copy have been employed; however, due to the
rotoxicity) as well as significant but reversible
low molecular weights of many of the listed
toxicities; class 2 solvents are to be limited in
solvents, such methods have proven difficult and
final formulations. Class 3 solvents are solvents
only qualitative with quantitation limited (Mulli-
with low toxic potential to man such that no
gan and McCauley 1995; Pavon et al. 2006).
specific exposure limit is required; class 3 solvents
The USP provides reference standard solutions
have permitted daily exposure levels of 50 mg or
for class 1 and 2 solvents (USP Class 1 Residual
more per day. Table 2.3 outlines prevalent class
Solvents Mixture RS and USP Residual Solvents
1, 2, and 3 solvents used in pharmaceutical
Class 2–Mixture A RS) which can be used for
preparations. It should be noted that these residual
system optimization and method development.
solvent guidelines do not apply to new chemical
Headspace analysis is the preferred sampling
entities, excipients, or drug products during clini-
technique for GC analysis. In headspace analysis,
cal research stages of development (Dwivedi
the sample for analysis is placed in a sealed vial
2002).
and equilibrated at elevated temperatures. The
temperatures must be high enough to allow for
Analytical Determination of Residual evolution of the bound solvent without combus-
Solvent Levels tion or damage to the material under analysis,
If only class 3 solvents are used during produc- especially for materials which evolve gasses
tion, nonspecific methods such as loss on drying upon decomposition. An aliquot of the evolved
may be used for quantitation of residual solvent. gas phase is then injected for analysis. During
This includes LOD apparatus analysis as well as sample preparation, it is often necessary to grind
TGA described above. When class 1 or 2 solvents the powder such that the internal area of the
have been used, their residual levels must be formulation is exposed. This process, however,
quantitated by a specific means. The USP 32/NF can lead to solvent evolution, thereby leading to
27 specifies gas chromatography with flame-
Table 2.3 Optimization of scan parameters for XRD analysis. Reproduced with permission from Elsevier
Combination Step time (s) Step size ( ) Scan rate ( 2θ/min) Recording time (min) No. of identifiable peaks
A 0.5 0.025 3 12.33 1
B 0.5 0.0125 1.5 24.67 1
C 1 0.0125 0.75 49.33 2
D 5 0.05 0.6 61.66 4
E 5 0.0125 0.15 246.66 4
inaccurately low results, and care must be taken changes may arise as loss of potency, the forma-
during preparation to prevent this. tion of impurities, alterations in drug form, mois-
A novel liquid-phase extraction method was ture uptake, or changes in dosage form
developed by Liu and Jiang (2007). Using the performance (i.e., dissolution profile, color, hard-
PWS drug adefovir dipivoxil, solutions were ness, leakage, brittleness, and pellicle formation)
prepared by adding 100 mg of API to a 5-mL and can be monitored by the solid-state character-
headspace vial. The ionic liquid 1-butyl-3- ization methods described in previous sections or
methylimidazolium tetrafluoroborate (bminBF4) alternative analytical methods. Formulations that
was then added to each vial (2 mL) and the may uptake significant amounts of moisture
material dissolved. The vials were then crimp should be analyzed for moisture content by loss
sealed and placed in an autosampler for analysis. on drying (LOD) studies or Karl Fischer titration
GC-FID was employed for analysis of all (Williams et al. 2004). Karl Fischer titration of
samples. Indeed, all six analyzed solvents could PWS drugs can be performed by dissolving a
be identified, making the method ideal for known amount of the compound in anhydrous
materials which are PWS and poorly soluble in methanol and analyzing the solution. Alterna-
organic solvents often used for liquid-phase tively, the powder for analysis can be added
extraction such as DMSO and toluene. directly to the titration vessel, provided it
dissolves in the media therein (Rogers et al.
2002). Thermally liable substances are likely to
2.3 Stability Testing be damaged by standard LOD testing (typically
105–130 C). For such a material, LOD studies
The development of PWS drugs often relies on should be conducted at ambient temperatures in
the use of an altered drug form such as a less the presence of strong desiccants (Bizzi et al.
stable but more soluble polymorph or the amor- 2011). For all of these studies, a deviation of 5%
phous form of the compound. While each of these from initial value following storage is considered
solutions may enhance aqueous solubility, they significant. While dissolution-testing parameters
bring with them numerous stability issues such as should be based on USP General Chapter 711,
conversion to a more stable polymorph, recrystal- solid-state characterization methods must be
lization of the amorphous material, or reduced optimized for the formulation at hand as
chemical stability. Indeed, conversion to a differ- described above to ensure adequate assessment
ent polymorph may yield drastic reductions in of stability.
solubility or complete loss of therapeutic activity. Validation or development of analytical
These conversions can be exasperated in the pres- methods capable of detecting impurities or degra-
ence of excipients or by the environmental dation products is essential for new formulations
conditions of the region for distribution (Bott or new chemical entities. Oftentimes, such
and Oliveira 2007). As such, it is necessary to method development will be performed concur-
gain a detailed understanding of the stability of rently with the chemical stability studies outlined
the drug itself as well as all potential formulations below. Selective methods include but are not
throughout the development cycle. limited to high-pressure liquid chromatography
(HPLC), thin-layer chromatography (TLC), gas
chromatography (GC), liquid chromatography
2.3.1 Stability Monitoring mass spectroscopy (LCMS), gas chromatography
mass spectroscopy (GCMS), and Raman spec-
For stability studies of an API or formulation such troscopy. Reflectance spectroscopy has proven
as those outlined in the following sections, useful for determining changes in the appearance
samples are to be taken at designated time points and color of solid dosage forms. This process may
throughout the study and the formulations also be sufficiently selective in isolated cases to
analyzed for any significant changes. These be used for degradation quantitation (Stark et al.
68 X. Ma et al.
1996). The method selected must be capable of which also contained water, the vials sealed and
separating the drug from degradation products as placed at 5, 50, or 100 C. Samples were removed
well as from excipients used during formulation. at predetermined time points over a 75-h period
HPLC is perhaps the most widely employed ana- and analyzed by XRD and DRIFT–IR for
lytical method for stability-indicating assays. transformations. Indeed, it was found that in the
Belal et al. developed a reverse-phase HPLC absence of moisture, even elevated temperatures
method for quantification of quetiapine as well of 100 C did not induce crystal transformations.
as its two degradation products, quetiapine However, the samples containing drug suspended
N-oxide and quetiapine lactam. Chromatographic in water at elevated temperatures demonstrated
conditions included a mobile phase composed of partial conversion of form B to C after 25 h.
acetonitrile and 0.02 M phosphate buffer (50:50) Additionally, it was found that suspensions of
at pH of 5.5 with a flow rate of 1 mL/min and form C stored at 100 C converted entirely to
detection at 254. Separation was achieved using a form A within 50 h (Brits et al. 2010). A lack of
250 mm 4.6 mm i.d., 5-μm particle-size Zorbax understanding of such phenomenon can lead to
SB-Phenyl column. This method proved suffi- processing difficulties during development. For
cient for quantification of drugs and degradation example, dissociation of the salt form of an API
products in both tablets and human plasma to the free base can occur upon moisture uptake or
samples. For additional examples of stability- introduction to the system. Using in-line Raman
indicating assay method development, especially spectroscopy, it was shown that the extended
regarding HPLC, the reader is referred to Cuiping exposure to moisture during the wet granulation
et al. (1993), Stanisz and Kania (2006), Ahuja and process employed prior to tableting of an experi-
Rasmussen (2007), Belal et al. (2008), and mental compound led to significant dissociation
Corrandini and Phillips (2011). of the hydrochloride salt yielding the free base
(Williams et al. 2004).
The ICH has set forth multiple guidelines
2.3.2 Chemical Stability regarding stability testing of APIs. Photostability
studies must expose samples to an overall illumi-
Prior to studying formulations of a given com- nation of 1.2 million lux hours and 200 W
pound, it is imperative that the raw drug h/m2 of near-UV energy. Control samples
components stability be assessed. While the protected by aluminum foil or a similar means
abovementioned analytical techniques such as must be placed next to the exposed samples to
DSC and TGA can be used to identify degrada- ensure authenticity (ICH guideline Q1B). Follow-
tion or transitions at elevated temperatures, the ing the ICH guidelines, the photostability of
influence of moisture is absent from these studies. fexofenadine hydrochloride was assessed via
As such, suspensions of the PWS drug under exposure to 1,800 W h/m2 of UV light and
study should be prepared at multiple pH values demonstrated a reduction in potency of 43%
similarly to the solubility studies previously (Bhalekar et al., 2010). The ICH also outlines
described and analyzed for degradation, poly- stress-testing requirements for determination of
meric conversions, recrystallization of amor- potential degradation pathways, including
phous material, and the dissociation of salt acidic/alkali degradation, oxidative degradation,
forms of APIs. For a detailed understanding of and thermal degradation (ICH guideline Q1A).
potential reaction pathways and kinetics thereof, Such studies can be conducted on drug powders,
the reader is referred to Waterman and Adami suspensions, or solutions. Valsartan and
(2005). For example, the two metastable amlodipine were studied for these four degrada-
polymorphs of mebendazole were studied to tion pathways in the following manner: 1 mL of
understand the interconversion between polymor- 0.1 M HCl (acidic degradation), 0.1 M NaOH
phic forms in the following manner: 500 mg of (alkali degradation), or 3% H2O2 added to 9 mL
the polymorphs was placed in glass vials, half of of drug solution (final concentration: 16 μg/mL)
and allowed to stand at ambient conditions for temperatures as shown above (Bott and Oliveira
24 h. A solution of identical concentration in 2007).
water was placed at 50 C and allowed to stand The ICH has set guidelines (ICH Q1A (R2))
for 24 h. Following the 24-h period, samples were for stability testing of formulations relative to the
analyzed by a validated HPLC method for degra- intended storage conditions upon distribution as
dation products. Results demonstrate that well as geographical region for distribution. This
valsartan degraded at elevated temperatures and includes products intended for storage at room
in the presence of hydrogen peroxide, while temperature, in a refrigerator, or in a freezer.
amlodipine was degraded under all conditions Table 2.4 outlines the storage conditions to be
(Chitlange et al. 2008). Ezetimibe was placed used for short term, long term, and accelerated
under similar stress conditions; however, in stability tests based on the intended storage loca-
order to achieve a desired 0.5-mg/mL solution, tion of the final product. Long-term studies must
30% acetonitrile was added to each vial under be conducted on three primary batches if an NDA
study (final solution 70:30 H2O/ACN). In this is to be filed and must cover a 12 month period. If
case, the acidic (0.1 M HCl), neutral (water), the proposed shelf life is greater than 12 months,
and alkali (0.1 M NaOH) solutions were heated the stability study must be continued for the dura-
to 80 C and held for 8 h to reduce the duration of tion of the expected shelf life. Accelerated
the study. This was repeated at 40 C as well. conditions are often employed during
Oxidative studies were carried out under 3 and preformulation studies as the timeline is signifi-
20% hydrogen peroxide solutions at room tem- cantly shortened; however, such a study is sup-
perature for 24 h. Photodegradation was assessed plemental to and does not replace the long-term
in water and 1 M HCl by exposing the solutions to storage data. Indeed, the elevated conditions of
sunlight for 2 days (60,000–70,000 lx). In this accelerated stability testing can induce changes
study, the bulk drug powder was also studied for which may not occur under normal storage. If any
thermal liability via exposure to dry heat: 50 C significant change occurs over the 6-month stor-
for 45 days and 60 C for 7 days. These studies age period at accelerated conditions, an interme-
revealed the primary degradation pathway to be diate study must be conducted as well (Gad 2008)
alkaline hydrolysis (Singh et al. 2006). Knowl- (Table 2.5).
edge of the potential degradation pathways of an Tests should be carried out in the container
API will greatly aid the formulation scientist identical to that intended for distribution. This
ensuring processes which may induce degrada- has led to a revision of the proposed conditions
tion are avoided. Studies such as these should be for stability testing to include conditions for semi-
carried out concurrently with or prior to solubility permeable containers. In this case, testing
studies to ensure that the solubility values temperatures are identical to those provided in
obtained are not skewed by degradation. Table 2.4; however, relative humidity values are
altered such that long-term storage is carried out
at 40% RH (35% RH if 30 C is selected for
testing), intermediate testing is set to 65% RH,
2.3.3 Stability Testing Conditions
and accelerated conditions employ no more than
25% RH.
Stability testing of potential formulations of an
For systems that are amorphous, the Tg of the
API must be performed to understand the syner-
formulation must be strongly considered when
gistic influence of environmental conditions and
selecting testing conditions. Indeed, accelerated
time on the dosage form. Indeed, a stable API can
conditions may employ temperatures near or
become unstable in the presence of another mate-
above the Tg of a product, yielding alterations in
rial as exemplified in the carbamazepine/stearic
physical form and complicating data interpreta-
acid example provided in the section regarding
tion. Improper selection of temperatures at or
XRD or in the presence of moisture or elevated
above the Tg of a system may lead to under-
70 X. Ma et al.
Table 2.4 Residual solvent levels in pharmaceutical products. Reproduced from ICH guideline Q3C (R4)
Solvent Concentration limit (PPM)
Class 1 solvents
Benzene 2
Carbon tetrachloride 4
1,2-Dichloroethane 5
1,1-Dichloroethene 8
1,1,1-Trichloroethane 1500
Class 2 solvents
Acetonitrile 410
Chlorobenzene 360
Chloroform 60
Cyclohexane 380
1,2-Dichloroethene 1870
Dichloromethane 600
1,2-Dimethoxyethane 100
N,N-Dimethylacetamide 1090
N,N-Dimethylformamide 880
1,4-Dioxane 380
2-Ethoxyethanol 160
Ethylene glycol 620
Formamide 220
Hexane 290
Methanol 3000
2-Methoxyethanol 50
Methylbutyl ketone 50
Methylcyclohexane 1180
N-Methylpyrrolidone 530
Nitromethane 50
Pyridine 200
Sulfolane 160
Tetrahydrofuran 720
Tetralin 100
Toluene 890
1,1,2-Trichloroethene 80
Xylene 2170
Class 3 solvents
Acetic acid Heptane
Acetone Isobutyl acetate
Anisole Isopropyl acetate
1-Butanol 3-Methyl-1-butanol
Butyl acetate Methyl ethyl ketone
Tert-butylmethyl ether Methyl isobutyl ketone
Cumene 2-Methyl-1-propanol
Dimethyl sulfoxide Pentane
Ethanol 1-Pentanol
Ethyl acetate 1-Propanol
Ethyl ether 2-Propanol
Ethyl formate Propyl acetate
Formic acid
Table 2.5 Stability testing conditions for products with various intended storage conditions
Stability study Minimum time period covered by data at
type Stability storage conditions submission (months)
Marketed API intended for room temperature storage conditions
Long term 25 2 C/60% RH 5% RH or 30 2 C/65% 12
RH 5% RH
Intermediate 30 2 C/65% RH 5% RH 6
Accelerated 40 2 C/75% RH 5% RH 6
Marketed API intended for storage in refrigerator
Long term 5 3 C 12
Accelerated 25 2 C/60% RH 5% RH 6
Marketed API intended for storage in freezer
Long term 20 C 5 C 12
prediction of the shelf life of a product (Duddu of solid dispersion powder into 30-mL high-den-
and Weller 1996). Amorphous systems which sity polyethylene bottles which were then induc-
have a high Tg however do not necessitate the tion sealed. Samples were placed under
selection special conditions and can be studied accelerated storage conditions (40 C, 75% RH).
under the accelerated conditions outlined in At time intervals of 1, 3, and 6 months, samples
Table 2.4 (Lakshman et al. 2008). were removed and tested for crystallinity. The
Following the development of indomethacin samples were allowed to equilibrate to room tem-
into coprecipitates and solid dispersions with perature for 24 h prior to analysis. It was shown
Eudragit® polymers with subsequent tableting of that formulations containing a plasticizer
the formulations, stability studies were conducted (Tg ¼ 54 C) exhibited recrystallization under
at –20, 4, 37, 45, and 55 C with relative humidity accelerated conditions, while formulations
values of 11, 51, and 91%. Tablet samples were lacking the plasticizer (Tg ¼ 101 C) did not
removed from storage at intervals of 1 month over (DiNunzio et al. 2010a).
a 6-month period and tested for transformations
by XRD and DSC as well as for changes in
dissolution profiles using rotating basket dissolu-
2.3.4 Predicting Stability Using
tion apparatus. Indeed, storage at extremely low
Machine and Deep Learning
temperatures as well as elevated temperatures in a
humid environment slowed down dissolution
As discussed previously, machine and deep
rates of the drug (Khan et al. 2000). Solid-state
learning exhibits better performance in predicting
characterization demonstrated that this was likely
properties that are intrinsically complex. Report-
due to an increase in crystallinity of the product.
edly, the US Food and Drug Administration
A similar study utilized 10 tablets per container
approved 19 commercial ASD products as of
with 2 containers per storage condition (4, 25,
2017 (Jermain et al. 2018). ASD has continued
37, 45, 55 C, and 37 C with 11, 51, and 91%
to gain popularity in the pharmaceutical industry
RH). Nine tablets were removed at time points of
as a primary formulation design strategy to over-
0.5, 1, 3, 6, 9, and 12 months and characterized by
come the poor aqueous solubility of BCS II drugs
XRD and DSC and tested for dissolution
(Kawabata et al. 2011). However, due to the high
properties (Goskonda et al. 1998). Tablets how-
thermodynamic energy state of the amorphous
ever may not be the intended final formulation as
drug molecules in ASD formulations, ASDs are
is the case with solid dispersions which can be
more likely to be affected by external environ-
filled into capsules. In this instance, the formula-
ment or internal interaction, which may impair
tion should be assessed for stability in powder
the stability and performance of the formulations.
form. For example, DiNunzio et al. placed 2 g
Conventional experimental approaches to test the
72 X. Ma et al.
stability of ASDs are time-consuming and unpre- ball wet milling and high-pressure homogeniza-
dictable. Therefore, predicting the stability of tion methods, but relatively poor prediction accu-
ASDs in advance using machine and deep racy for antisolvent precipitation approach.
learning can be extremely helpful for formulation Importantly, the contribution of the influence
scientists to capture early warning signs in order factors was ranked by the machine learning,
to develop the optimum formulations. revealing that milling time, cycle index, and the
Han et al. (2018) developed a novel prediction concentration of stabilizer were critical to the
model using machine learning techniques. particle size and poly-dispersity index of
8 machine learning approaches were tested nanocrystals.
using 646 stability data points, covering 20 molec-
ular descriptors. Results demonstrated that the
random forest model achieved the highest predic- 2.4 Dissolution Testing
tion accuracy of 82.5%. Furthermore, through
analyzing the trained random forest model, the As mentioned earlier, BSC Class II APIs are
contribution of each input parameter was limited in their bioavailability based on their dis-
revealed, which provided the useful information solution rate or extent. Therefore, many
for scientists to better understand the critical formulators attempt to overcome this barrier by
attributes that affected the stability of ASD generating systems capable of achieving solubili-
formulations. Additionally, the machine learning zation significantly higher than the intrinsic solu-
model was verified by the physical stability bility of the compound (i.e., supersaturation). The
experiment of an ASD, and the molecular ability to achieve supersaturation in physiologi-
interactions were investigated by molecular cally relevant media can indeed be a strong indi-
modeling technique. The study illustrated the cator of enhancement of bioavailability.
capability of machine and deep learning for accu- However, this can only truly be confirmed by
rate prediction of ASD stability. transitioning optimized formulations into animal
Kapourani et al. (2020) developed an approach models.
by incorporating artificial neural networks Dissolution studies are often carried out under
(Fig. 2.15) and ATR–FTIR spectroscopy to sink conditions; however, for many PWS drugs,
simultaneously distinguish and quantify crystal- such conditions cannot be met due to the
linity and the neat amorphous drug located within extremely high volumes of dissolution media
the drug-rich amorphous zones in an ASD sys- that would be required. As such, studies using
tem. Results illustrated that the artificial neural sink conditions will not be discussed; rather, the
network significantly improved the prediction focus of the following sections will be on super-
accuracy compared to conventional approaches, saturation studies and in vivo studies of optimized
such as regression modeling, partial least square, formulations
and principal component regressions.
Recently, He et al. (2020) utilized light gradi-
ent boosting machine learning model to success- 2.4.1 Dissolution Studies
fully predict the particle size and poly-dispersity
index of nanocrystals, which had significant Sample Handling
impact on the stability of nanoparticles. During dissolution testing, samples will be taken
910 nanocrystal size data and 341 poly-dispersity at various time points to assess drug release from
index data from 3 preparation methods (ball wet and supersaturation ability of the formulation.
milling, high-pressure homogenization, and Proper handling of these samples ensures accu-
antisolvent precipitation methods) were collected racy in the assay used to assess drug in solution.
to construct the machine learning model. Results Filtration is a widely used technique for removing
demonstrated that the machine learning undissolved materials, both drug and insoluble
performed well for the nanocrystals produced by excipients, from the analytical sample. This is
Fig. 2.15 Artificial neural

network architecture for the
determination of the
crystalline (a) and the
amorphous (b) ritonavir
content within the prepared
ritonavir–Soluplus® ASDs.
Kapourani et al. 2020)
done by placing a withdrawn dissolution sample Prior to use in dissolution studies, it is impor-
in a syringe and passing through an attached tant to assess filter membrane compatibility with
syringe filter. Filter sizes often used in literature the API and sampling solution. Indeed, drug
are 0.45 μm, 0.22 μm, and 0.1 μm, with the adsorption to the filter membrane can lead to
syringe size selected based on the withdrawn inaccurately low results. Additionally, membrane
sample size. A wide variety of filter membrane degradation in the presence of solvents used may
materials exist for this application, including but allow the passage of undissolved drug into the
not limited to nylon, hydrophilic and hydrophobic analytical sample that may dissolve upon dilu-
polytetrafluoroethylene (PTFE), polyvinylidene tion, yielding inaccurately high results. To vali-
fluoride (PVDF), and cellulosics such as cellulose date a filter, an API stock solution of known
acetate (Okazaki et al. 2008; Ferrari et al. 2009; concentration is prepared in the media to be
Tho et al. 2010; Xia et al. 2010). used during dissolution testing. Aliquots of the
74 X. Ma et al.
stock solution are then filtered and diluted and various pore sizes were assessed, and the 24-h
compared to a diluted unfiltered sample. If the fenofibrate concentrations obtained via dissolu-
filtered sample displays 98–102% recovery of tion studies were compared to the 24-h solubility
the API, the filter is considered acceptable study data. It was shown that pore sizes of 0.45 or
(Fortunato 2005). Such a method was employed 0.2 μm yielded concentrations higher than those
for dissolution studies of the PWSD lamotrigine. obtained during solubility studies, while pore
A stock solution of lamotrigine was prepared at a sizes of 0.1 μm or less generated concentrations
concentration of 11 μg/mL in 0.1M hydrochloric similar to those of the solubility studies. Indeed,
acid. The solution was sonicated for 15 min in a the colloidal fenofibrate was too fine to be
sonication bath to ensure complete dissolution of withheld by filters of larger pore sizes which,
the API. A 4-mL aliquot of the stock solution was upon sample dilution with solvents, resulted in
then diluted with 0.1M HCl to a final volume of apparent supersaturation (Juenemann et al. 2011).
20 mL in a volumetric flask as an unfiltered If adequate recovery cannot be achieved via
control. The remaining solution was passed filtration, samples may be centrifuged instead and
through the filter under study and a 4-mL aliquot the supernatant sampled for analysis. In the
of the filtrate diluted to 20 mL with 0.1M HCl. abovementioned study regarding lamotrigine, a
UV spectroscopy of the filtered and nonfiltered stock solution was centrifuged at 3000 rpm for
samples revealed less than 2% variability between 10 min. A 4-mL aliquot of the supernatant was
the solutions, indicating the filter was acceptable then diluted to 20 mL in dissolution media. The
(Martins et al. 2010). Although not a PWS API, a remaining supernatant was filtered and a 4-mL
study of three potential filters and three potential aliquot of the filtrate diluted to 20 mL in dissolu-
dissolution media (0.01, 0.1M HCl, and pH 6.8 tion media. Samples were analyzed via UV spec-
phosphate buffer) to be used during dissolution troscopy and revealed less than 2% variability.
testing of citalopram was conducted in a similar This confirms the acceptance of the filter to be
manner. The analysis revealed that the quantita- used as well as the potential for centrifugation in
tive and 0.45-μm nylon filters were acceptable for place of filtration. While centrifugation may be
use during dissolution studies, while the 3-μm used as a reference during filter analysis, it can
filter under study was not (Menegola et al. also be used for sample purification in place of
2007). This method was also used in filter evalu- filtration. This can be employed when an accept-
ation for dissolution studies on the BSC Class IV able filter cannot be identified due to membrane
compound entacapone. In this iteration, 50 mL of interactions (Fortunato 2005). To ensure adequate
stock solution was prepared in the dissolution sedimentation of all particulates present, high-
medium at a concentration of 44.44 μg/mL. Com- speed centrifugation is employed. For example,
plete dissolution was ensured by sonication for centrifugation at 14,000 rpm for 10 min was used
30 min in a sonicating bath. Samples were then for dissolution sample preparation for capsule
filtered through quantitative and nylon filters and dissolution studies of an experimental compound
compared to an unfiltered diluted sample. It was (Zhao et al. 2009). Dissolution studies performed
found that both filters were within the limits of directly in the centrifuge tube (discussed below)
acceptance for use during the studies. Addition- have been subjected to rotation at 13,000 g for
ally, it was shown that the quantity of methanol to 1 min prior to sampling to ensure adequate sedi-
be used for dilution did not alter the filter func- mentation (Friesen et al. 2008).
tionality (Paim et al. 2010).
In addition to membrane filter-type selection, Excipient Screening for Supersaturation
the appropriate pore size must be selected based Maintenance Ability
on the formulation at hand. Following solubility The ability to achieve supersaturation is highly
studies of fenofibrate in bio-relevant media, promising with regard to increasing bioavailabil-
Juenemann et al. analyzed the dissolution of ity. However, the ability to maintain supersatura-
nanosized formulations of the API. Filters of tion can prove even more beneficial. APIs with
pH-dependent solubilities may rapidly precipitate subsequently used in the production of solid
from solution upon transition from the stomach to dispersions.
the intestine. Therefore, prevention of this formu-
lation collapse may allow greater time for absorp- Supersaturation Dissolution Studies
tion and increased bioavailability. A novel Following production of lead formulations, their
approach in the screening of excipients capable ability to achieve supersaturation must be
of maintaining supersaturation has been presented assessed. This can be accomplished by adding
by Vandecruys et al. (2007). In this method, vari- an amount of formulation to a dissolution vessel
ous excipients of interest were dissolved at a level such that the amount added contains an excess
of 2.5% w/v in 10 mL of the corresponding media amount of drug relative to the intrinsic solubility;
of interest: 0.01H HCl, USP pH 4.5 buffer, USP thereby, upon complete dissolution, a theoretical
pH 6.8 buffer, or water. Separately, various drugs level of supersaturation is achieved. For example,
of interest were dissolved at a concentration of in a study of HME itraconazole formulations,
50–100 mg/mL in N,N-dimethylformamide 180 mg of milled extrudate (containing 60 mg
(DMF) or dimethylacetamide (DMA). For a itraconazole) was added to each dissolution ves-
given test, the desired dissolution media sel. Assuming complete dissolution, this amount
containing the excipient to be studied was placed corresponds to 80 μg/mL representing 20 super-
under magnetic stirring and equilibrated to 37 C. saturation (intrinsic solubility under acidic
The selected API solution was then added conditions: 4 μg/mL) (Miller et al. 2008). A simi-
dropwise to this stirring media until a precipitate lar work employs itraconazole at a level equiva-
was just noticeable visually. Samples were then lent to 10 equilibrium solubility (DiNunzio
taken at 5, 30, 60, and 120 min following the et al. 2010a, b, c). In a separate work, 7 mg of
completion of drug addition, filtered, and tacrolimus formulations was added to 100-mL
analyzed for drug content via UV spectroscopy. small-volume dissolution vessels containing
Application of this method to excipient blends acidic media corresponding to 28 the equilib-
may allow for an even deeper understanding of rium solubility of the native API (Overhoff et al.
the formulation requirements for the maintenance 2008). Using identical conditions, the ability to
of supersaturation. maintain supersaturation upon pH transitions was
The above method was adapted in the screen- assessed by transitioning the pH of the dissolution
ing of various HPMC grades for their ability to media to 6.8 via addition of appropriate amounts
supersaturate itraconazole in neutral media of 0.2M Na3PO4.
(intrinsic solubility 1–5 ng/mL). In this itera- Such studies are not limited to supersaturation
tion, 75 mg of the excipient was dissolved in 1 L of GI media. Using simulated lung fluid
of pH 6.8 phosphate buffer. Itraconazole was containing 0.02% dipalmitoylphosphati-
separately dissolved in 1,4-dioxane at a concen- dylcholine and small-volume (100 mL) dissolu-
tration of 18.75 mg/mL. Following equilibration tion vessels, 100 μg of itraconazole in a colloidal
of the dissolution media to 37 C, a 2-mL aliquot dispersion corresponding to 100 equilibrium
of the itraconazole solution was added to the solubility was added in the assessment of a
dissolution vessel, and samples were withdrawn nanosized formulation for inhalation. In fact, it
at 5, 10, 15, 30, 45, 60, 90, 120, 180, 240, and was shown that the formulation was capable of
1440 min. Withdrawn samples were immediately achieving 27 supersaturation versus crystalline
passed through a 0.2-μm filter and diluted 1:1 itraconazole (Yang et al. 2008).
with mobile phase (70:30:0.05 acetonitrile/ Of crucial importance to each of these studies
water/diethanolamine) and subsequently is the prevention of precipitation in the withdrawn
analyzed for itraconazole concentration via samples, which can lead to inaccurately low
HPLC (DiNunzio et al. 2010a, b, c). Application results, as well as accidental solubilization of
of the method identified HPMCAS grades as the withdrawn particulates yielding positive
most promising, and these polymers were deviations. To prevent these occurrences,
76 X. Ma et al.
withdrawn samples must be treated immediately Alternative Dissolution Studies

upon withdrawal as discussed above and diluted The use of bio-relevant dissolution media such as
with an appropriate organic solvent, such as the simulated lung, gastric, and intestinal fluids may
mobile phase employed in the aforementioned provide stronger in vitro–in vivo correlations and
screening studies. Analysis of the diluted samples more accurate predictions of formulation perfor-
via a specific method such as HPLC can be mance. However, the use of such fluids
utilized on the diluted samples and the deter- introduces new molecular variables into solution.
mined concentration adjusted for dilution via For example, fasted-state-simulated gastric fluid
minor calculations. comprises sodium taurocholate, lecithin, pepsin,
Additional methods not employing a dissolu- sodium chloride, and hydrochloric acid, while the
tion apparatus have also been used in supersatu- simulated lung fluid previously mentioned
ration dissolution testing. Curatolo et al. contains 0.02% dipalmitoylphosphatidylcholine
developed two novel approaches for supersatura- (Vertzoni et al. 2005). A detailed discussion
tion testing. In the first method, 7.5 mg of a including instructions on the preparation of vari-
formulation was placed in an empty disposable ous simulated intestinal fluids can be found in
10-mL syringe. A 20-gauge needle was attached Jantratid et al. (2008). Two concerns are raised
and used to draw 10 mL of model fasted duodenal during the use of such dissolution media. First,
fluid at 37 C into the syringe. The needle was upon dissolution multiple species may form in
then removed, a 13-mm, 0.45-μm filter was solution. Friesen et al. describes seven potential
attached, and the syringe shaken for 30 s. species based on size that may form upon disso-
6 drops were then expelled as waste followed by lution of a polymeric formulation in such media.
collection of 13 drops as a sample. The plunger These include free/solvated drug, drug in bile-salt
was then pulled back to introduce an air bubble micelles, free/solvated polymer, polymer
and promote mixing, and the syringe with colloids, amorphous drug/polymer
attached filter was placed on a rotating wheel in nanostructures, small aggregates of amorphous
a temperature-controlled box held at 37 C to drug/polymer nanostructures, and large
mix. Sampling was then repeated in the identical precipitates (Friesen et al. 2008). As the free
manner at various time points. All samples were drug is the primary absorbed species and of
diluted with mobile phase to prevent chief concern during dissolution studies, such
precipitation. systems require unique handling during drug
In a second method by the same group, 1.8 mg analysis to ensure that there is no interference
of the formulation was placed into an empty from undesired components. Filtration, as
microcentrifuge vial and 1.8 mL of model fasted described above, has proven sufficient in
duodenal fluid added. The tube was then vortex separating undissolved drug from simulated
mixed for 60 s and then allowed to stand for 6 min fasted-state intestinal fluids and certain gastric
without disruption. Following the equilibration fluids. However, more complex media may
period, the sample was centrifuged at 13,000 g require additional steps. For example, a study in
for 60 s. A sample of the supernatant was then milk-based media required preliminary filtration
taken. Following sampling, the vial was vortexed of undigested samples through a 5.0-μm nylon
for 30 s to resuspend the material and allowed to filter followed by dilution with acetonitrile (pro-
stand for a designated period of time. At the next tein precipitation), centrifugation at 4000 rpm for
set time point, the material was again centrifuged 10 min at 8 C, and final filtration of the superna-
and sampled as described previously (Curatolo tant through 0.45-μm regenerated cellulose filters.
et al. 2009). All samples were diluted 1:1 with Similarly, digested samples were first centrifuged
mobile phase upon withdrawal to prevent precip- under the above conditions to separate the aque-
itation. No filtration was applied as centrifugation ous phase from not only the undissolved drug but
removed all particulate matter from the sampling also the digested and lipid phases. Aqueous phase
media.
samples were then diluted with acetonitrile and which provide an absorptive sink via a
recentrifuged with subsequent filtration through partitioning approach (McAllister 2010). In such
the aforementioned 0.45-μm regenerated cellu- studies, the dissolution media is composed of a
lose filter (Fotaki et al. 2005). It should also be bio-relevant aqueous phase as well as an organic
noted that lipid-based formulations require centri- phase such as octanol. The dissolution rate will
fugation to remove nonemulsified lipid droplets therefore dictate the amount of drug available in
prior to aqueous phase sampling (Jantratid et al. solution for partitioning into the organic phase.
2008). Advantages of biphasic dissolution studies
The implementation of dialysis methods for include prevention of accumulation of the drug
dissolution studies minimizes the need for sample in the aqueous phase which may lead to crashing
handling. As dialysis membranes with specified out phenomena and no post-sampling handling to
molecular weight cutoffs can be used such that eliminate undesired species as the partition pro-
only solubilized drug is able to pass through, cess acts as a filter. In a study by Shi et al., a USP
subsequent filtration or centrifugation is not Apparatus II vessel was filled with 250 mL of
required. It is recommended that filtration still 80 mM phosphate buffer (pH 6.8) and 200 mL of
be carried out on samples, however, to remove octanol. A standard USP II paddle was modified
any potential external contaminates such as dust with a secondary paddle to allow agitation of both
which may enter the sampling media during the aqueous and organic phases. The formulation
lengthy studies. Dialysis dissolution has been to be tested was placed in an external flow
used on micellar solutions of paclitaxel (molecu- through cell and the aqueous dissolution media
lar weight cutoff of 3500) by loading the mem- was circulated through the cell by means of a
brane bag with the solution and placing it in piston pump and Teflon tubing (Fig. 2.16).
20 mL of phosphate-buffered saline (pH 7.4). Preoptimized parameters of 75-rpm paddle
The media was held at 37 C and agitated by an speed and 30-mL/min pump flow rate were used
orbital shaker at 160 rpm. Sampling was during testing as unpublished data demonstrated
performed by complete dissolution media that they provide the strongest in vitro–in vivo
removal and replacement with fresh PBS. Disso- correlations. Samples were taken from both the
lution was allowed to continue for 30 days (Yang aqueous and organic phases at predetermined
et al. 2009). Similarly, 5 mL of a suspension of time points (Shi et al. 2010). A similar study has
chitosan nanoparticles in pH 7.4 buffer was been performed in a USP Apparatus II dissolution
placed in a dialysis bag and placed in pH 7.4 vessel not incorporating the external flow through
buffer reception media. Agitation by orbital shak- cell. Here, 500 mL of a sodium dihydrate
ing (110 rpm) was applied, and samples were phosphate-buffered aqueous phase and 100 mL
taken at predetermined time points by complete of n-octanol were placed in a dissolution vessel
media replacement (Syam et al. 2010). It should held at 37 C and the formulation to be studied
be noted that dialysis methods may require equil- placed directly in the dissolution vessel.
ibration periods depending on the formulation. Uniquely, an automated pH titration and
Indeed, drug nanoparticles placed in a dialysis controlling device was placed in the vessel to
bag exhibited significantly lower dissolution apply a pH gradient over the duration of the test.
rates compared to basket and paddle methods During analysis, the pH was initially held at two
over the initial 60-min time period (Heng et al. for 1 h mimicking gastric conditions. The media
2008). was then adjusted to pH 5.5 within 5 min to
BSC Class II compounds are characterized by simulate gastric emptying into the upper intestine.
a higher degree of permeability than dissolution At 5 h, the pH was adjusted to 6.8 where it
rate, making the dissolution rate the limiting fac- remained for the duration of the study to simulate
tor in bioavailability. In order to more accurately transit through the lower GI. Sampling was
depict the in vivo performance of such drugs, performed on both aqueous and organic phases
biphasic dissolution studies have been developed at predetermined time points. The solubility of the
78 X. Ma et al.
Fig. 2.16 Diagram of a Media USP II with dual paddle

biphasic dissolution test
apparatus. From Shi et al.
(2010). Reproduced with
permission from the
American Chemical
Society
Drug
Organic layer
product
Pump
Aqueous layer
Returned
Media
compounds tested was significantly higher in acid that the artificial neural network model using the
than either basic media or n-octanol. Therefore, near-infrared/Raman and compression force data
dissolution profile analysis reveals a significant outperformed the partial least regression model
amount of drug in solution while acidic and achieved higher accuracy. The proposed
conditions were held. However, following pH machine learning approach was proved to be
adjustment, the aqueous solubility dropped mark- more time-efficient for predicting the dissolution
edly. The presence of the organic phase allowed of extended-release formulations. In another
maintenance of sink conditions and prevented study conducted by Galata et al. (2021), spectro-
precipitation within the aqueous phase (Heigoldt scopic measurements, process data, and critical
et al. 2010). material attributes were used to predict dissolu-
tion profile of sustained-release tablets. Three
Predicting Dissolution Performance Using machine learning methods were used and com-
Machine and Deep Learning pared, including artificial neural networks, sup-
As discussed earlier, machine and deep learning port vector machines, and ensemble of regression
can significantly improve the accuracy and effi- tress. Artificial neural network (Fig. 2.17) showed
ciency of predicting the properties of drug the highest prediction accuracy compared to the
formulations. As one of the crucial performances other two models.
for the pharmaceutical products, dissolution of Yang et al. (2019) developed a deep learning
the poorly water-soluble drugs is critical to ensure model using an improved maximum dissimilarity
the acceptable bioavailability and efficacy. How- algorithm with the small group filter (MD-FIS) to
ever, conducting the dissolution experiments for predict the dissolution of pharmaceutical
various formulations can be time-consuming. formulations. In this study, two types of dosage
Therefore, the application of machine and deep forms were selected as model systems. Results
learning in predicting dissolution profiles offers illustrated that the MD-FIS model outperformed
great advantages to formulation scientists for the other six state-of-the-art machine learning
rapid product development. models with accuracy above 80%. The successful
Galata et al. (2019) implemented the artificial development of the MD-FIS showed great poten-
neural network model to predict the dissolution tial in the implementation of quality-by-design
profile of oral tablets combined with near-infrared concepts throughout the drug development pro-
and Raman spectroscopy. Drotaverine was used cess. The proposed MD-FIS model would signif-
as a model drug and was formulated as an icantly shorten the drug product development
extended-release formulation. 37 different timeline and minimize the material usage
processing settings were used to produce tablets (Fig. 2.18).
with various dissolution profiles. Results showed
Fig. 2.17 Structure of artificial neural network for predicting the dissolution profiles. (Adapted with permission from
Galata et al. 2021)
Fig. 2.18 Illustration of MD-FIS structure
2.4.2 In Vivo Testing However, these studies are only indicative of

performance, not absolute, as the physiological
The above-outlined in vitro testing allows the environment is far more complex than the
researcher to select lead formulations based on conditions used in the laboratory. Although
their performance in dissolution testing. simulated intestinal fluids such as the model
80 X. Ma et al.
fasted duodenal fluid and simulated lung fluid the fluid layer at the cell surface decreases from
mentioned above allow closer approximation to 8 μm to approximately 70 nm in direct correlation
in vivo conditions, in order to truly understand the with the decrease in cell thickness. These factors,
ability of a formulation to provide enhanced bio- coupled with the lack of digestive enzymes and
availability of a poorly water-soluble compound, bypass of first-pass metabolism, make pulmonary
animal models must be employed. delivery an attractive route of delivery for PWS
Unfortunately, there is no single surrogate ani- drugs.
mal species acceptable for all in vivo testing In an effort to allow passive inhalation of a
requiring selection of the appropriate model by drug aerosol, a whole-body chamber was
the researcher. Selection of the appropriate animal designed which allowed up to 14 mice to be
model will depend on numerous factors; how- dosed simultaneously. The solution or dispersion
ever, in all cases, the smallest viable model to be administered could be nebulized directly
should be used. Some factors influencing the into the chamber exposing the unrestrained
choice of model should include the dose size, animals to the dose. Restraint during dosing
route of administration, required surgeries, the may lead to physiological changes such as altered
number of blood draws, and whether or not a temperature regulation and can adversely affect
crossover design is to be used. Rat models are an in vivo study. Therefore, elimination of the
often used for single- and multiple-oral-dose restraint may regulate rodent physiology, reduc-
pharmacokinetic studies as multiple blood draws ing variability in received dose. The whole-body
can be taken without significant risk of anemia. apparatus was employed in the study of nebulized
Rat models, however, are inappropriate for cross- itraconazole nanoparticle dispersions. Using
over studies as their growth rate is too rapid to 10 mice, the formulation was first assessed for
allow for proper comparison of bioavailability dose uniformity, and it was shown that whole-
between formulations. In such a case, a larger body exposure yielded a variation of 13% relative
model such as a canine model may be required. standard deviation (Fig. 2.19). An in vivo study
Regardless of the model chosen, the protocol to was then conducted to determine the lung con-
be carried out must be approved by an institu- centration and residence time of the nebulized
tional review board (IRB) and be in accordance itraconazole. 3 groups of 14 mice each were
with the Institutional Animal Care and Use Com- exposed to the aerosol generated by a 20 mg/mL
mittee (IACUC; if applicable) or similar dispersion for 20 min. Following dosing, two
governing body at the location of the study. The mice were sacrificed at time points of 0.5, 1, 2,
following examples describe experimental 4, 6, 10, and 24 h, the lungs harvested,
protocols for both oral and inhalation studies of itraconazole extracted, and the concentrations
PWS drugs including information on dosing normalized for weight (i.e., μg/g lung tissue). As
methods and sampling protocols, focusing on can be seen in Fig. 2.20, significant drug levels
rodent models. Extraction methods for blood could be not only achieved but also maintained
and tissue analysis will not be presented. following administration via inhalation
(McConville et al. 2005, 2006). A similar study
Administration via Inhalation utilizing the same apparatus demonstrated
The surface area of the distal airway is approxi- enhanced bioavailability of itraconazole amor-
mately 102 m2, while the conducting airways is a phous nanodispersions, with both lung and
mere 2–3 m2, allowing for much greater contact blood (cardiac puncture) samples being taken at
with the inspired gas or therapeutic aerosol each sacrifice time point (Yang et al. 2008).
(Patton 1996). Additionally, the thickness of the While restraint-free whole-body dosing may
cell layer which makes up the respiratory region provide the advantages of dosing many animals
is progressively reduced from approximately at once and minimizing stress to the animal, one
60 m in the upper airway, to submicron thickness must consider the fact that these units provide
in the alveoli (Patton and Byron 2007). Similarly, whole-body exposure. As such, there is possible
Fig. 2.19 Dose variability for mice exposed to nebulized itraconazole via whole-body dosing chamber (McConville
et al. 2006)
Fig. 2.20 Itraconazole 20

lung concentration versus
ITZ Lung Concentration (µg/g)
time following whole-body 17.5

exposure (McConville et al. 15
2006)
12.5
10
7.5 CIP-D
AIP1-D
5 AIP2-D
2.5
0
0 5 10 15 20 25
Time (Hours)
drug absorption through dermal and ocular routes nose-only dosing chambers have been developed
as well as the oral route as the animals may for rodent models. These include commercial
consume drug deposited on their fur during tower units in which rodents are attached perpen-
grooming. This last fact may not be influential dicular to the airflow. The generated aerosol flows
when looking strictly at lung deposition; how- vertically downward across the subjects’ noses,
ever, it will be crucial for biodistribution studies and the restrained animals passively breathe it
as the oral dose may be selectively cleared by a in. A custom-built laboratory scale unit was
specific organ (i.e., kidney/liver) (Phalen et al. used in a pharmacokinetic study of inhaled amor-
1984; Roy et al. 2003). phous colloidal dispersions of itraconazole.
To limit whole-body exposure and prevent Precatheterized (jugular vein) Sprague–Dawley
potential oral, dermal, and ocular exposure, rats were restrained using nose-only restraints
82 X. Ma et al.
(Battelle) and connected to a horizontal chamber the lungs of an animal. Using an attached syringe
such that only the nose of each rodent was inside to generate airflow, the powder under study can
the chamber. The dispersion was then nebulized be expelled intratracheally for deep lung deposi-
with a circulating fan, providing a constant air- tion. Such a device was used on anesthetized rats
flow of 1 L/min to propel the generated aerosol (via diethyl ether) to deliver 5 mg of PLGA
horizontally through the system, and the attached nanocomposite particles loaded with TAS-103, a
rats were allowed to passively inhale the aerosol PWS antitumor agent. Following dosing, blood
for 10 min. Animals were then returned to their was collected at 0.5, 1, 2, 4, and 8 h through the
cages, and blood draws were taken through the retro-orbital plexus region. Withdrawn blood was
jugular vein catheter at 0, 0.5, 1, 1.5, 2, 2.5, 3, 4, then centrifuged at low temperature (5 C) for
6, 8, 12, and 24 h post-dosing. Collected samples 10 min and a sample of the supernatant collected
were placed in preheparinized tubes to prevent for subsequent plasma concentration analysis.
clotting. Lungs were harvested from two rats Lung tissue was also harvested for deposition
immediately after each dosing to assess lung analysis (Tomoda et al. 2009). Another study
deposition, with all others being sacrificed and examined pulmonary delivery utilizing the same
lungs harvested at 24 h (Yang et al. 2010). It device. Using a dose of 10 mg/kg, scutellarin was
should be noted that prior to any dosing, the loaded into the tip of the DPI, attached to a
animals must be acclimated to the restraints over syringe and inserted into the trachea of the
a period of days to prevent any significant physi- anesthetized rat. Depression of the syringe
ological changes during dosing resulting from plunger produced the aerosol for delivery directly
stress. to the lungs. This pulmonary powder formulation
While nose-only dosing chambers have was compared directly to intratracheal adminis-
proven effective for assessment of pulmonary tration of a solution using a Penn-Century Model
formulations, deposition can occur in the upper IA-1B with a volume of 1 mL/kg. In both cases,
airways as well as around the mouth. While this is the animals were held in the upright position for
likely to be only minor external deposition, there 1 min following removal of the delivery device to
is still a risk of an oral dose as a result of ensure deposition (Liu et al. 2008). It should be
grooming and potential poor deposition due to noted that the dose administered in the above
rodent activity within the restraining device. To studies is much greater than a dose receivable
avoid deposition in the upper airway or oral cav- by a murine model. To accommodate for the
ity, a dry powder insufflator (DPI), such as that necessary reduction in dose, 200–300 μg of pow-
offered by Penn-Century (Fig. 2.21), can be used der was loaded into polyethylene tubing (1 cm in
to administer a dry powder formulation directly to length) by dipping the tubing directly in the dry
Fig. 2.21 Penn-Century

Model DP-4 M dry powder
insufflator with attached
syringe
powder. This loaded tube was then inserted in the size 9 gelatin capsules from Torpac. In order to
hole leading from the insufflator chamber to the ensure adequate GI media was present for disso-
cannula, and a syringe was attached. Two hun- lution of the capsule and that no esophageal stick-
dred and fifty microliters of air was used during ing had occurred, 0.4 mL of water was immediate
dosing to disperse the powder. If necessary, addi- given via oral gavage following capsule adminis-
tional puffs of air at identical 250-μL increments tration. As these rats were precatheterized, blood
were used to disperse the material from the device draws could easily be taken at 0.5, 1, 1.5, 2, 3,
(Morello et al. 2009). Additional studies using 4, 6, and 24 h post-dose. At each sampling time
DPIs have proven successful in additional animal point, the withdrawn blood was replaced with
models including guinea pigs (Sung et al. 2009) heparinized normal saline (6 units/μL) (Overhoff
and beagle dogs (Surendrakumar et al. 2003). et al. 2008). In a study of the PWS drug letrozole,
a rat model was selected utilizing both male and
Oral Administration female rats. The drug was then administered in
The oral route of administration is highly desired one of three ways: oral gavage, orally via a
for an end formulation due to high patient com- Torpac size 9 capsule, or IV infusion via the
pliance and ease of administration. However, the ophthalmic venous plexus. For capsule adminis-
gastrointestinal tract presents numerous obstacles tration, an amount of formulation equivalent to
preventing successful oral delivery, including 1.0-mg API was placed in the capsule and
environments of varied pH, reactive enzymes, administered orally. Similar to the previously
and biological clearance systems. These factors discussed study, a bolus of water (500 μL) was
combined with limited aqueous solubility make provided following administration of the capsule.
oral delivery of PWS drugs especially challeng- Oral gavage was performed by generating a sus-
ing (O’Donnell and Williams 2011). As discussed pension of the native compound in a 12.5% aque-
above, many formulators are attempting to gener- ous ethanol media. Sonication provided adequate
ate systems capable of supersaturating the GI dispersion of the material immediately prior to
milieu in an effort to overcome bioavailability dosing. The test formulations for oral gavage
issues associated with poor aqueous solubility. were made into solutions using the same 12.5%
While these systems may prove promising during aqueous ethanol media. IV dose solutions were
in vitro supersaturation studies, success upon prepared into 1-mg/mL solutions, filtered through
in vivo application must be determined via animal a 0.45-μm filter, and immediately dosed. In order
modeling. to generate a comparable IV dose of the native
As PWS drugs are often not formulated as compound, the API was first dissolved in DMSO,
solutions, administration to the animal model via and a 300-μL aliquot of this stock solution was
the oral route relies upon the dosing of powder diluted to 10 mL using saline to yield a 0.297-mg/
formulations. This can be accomplished using a mL solution. This was then filtered and dosed.
size 9 capsule and corresponding dosing syringe Contrary to the precatheterized rodents from the
such as those offered by Torpac and Capsugel. previous study, tail vein collection of blood
Assuming a powder density of 1 g/cc, these samples was performed using minicapillary
capsules hold up to 25 mg of powder and dissolve blood collection tubes pretreated with EDTA
within the rodent’s stomach within 10 min. These dipotassium salt. Samples were stored at
capsules have been utilized for low-density amor- 80 C until analysis (Wempe et al. 2007). Fig-
phous solid dispersions of tacrolimus prepared by ure 2.22 depicts the pharmacokinetic profiles of
ultrarapid freezing. A nominal dose of 5 mg/kg the IV versus capsule-administered API in the
was dosed to male Sprague–Dawley rats (300 g presence and absence of hydroxybute-
rats; approximately 1.5 mgs/capsule) using the nyl-β-cyclodextrin (HBenβCD). As can be seen,
84 X. Ma et al.
Fig. 2.22 Pharmacokinetics of IV (left) and capsule administered letrozole. Closed symbols: API in absence of
HBenβCD. Open symbols: API in presence of HBenβCD. Reproduced with permission from John Wiley and Sons
a drastic difference in bioavailability was of the formulations immediately prior to dosing

observed for male versus female rats depicting such that a 400-μL aliquot contained 9-mg API.
the importance of not only proper species selec- This dispersion was then provided via oral gavage
tion but also gender selection thereof. and blood draws taken through a jugular vein
For PWS drugs in which a solution cannot be catheter. Indeed, these formulations provided a
prepared for oral gavage and a capsule is not threefold increase in itraconazole absorption
capable of holding a sufficient dose, a suspension (Miller et al. 2008). In a similar study, engineered
or dispersion can be prepared as was done with particles of itraconazole were dispersed in
the native letrozole. As previously mentioned, the deionized water at a concentration of 4 mg/mL
bioavailability of itraconazole is significantly hin- and dosed by oral gavage. Unique to this study is
dered by its poor aqueous solubility, especially the comparison of the engineered material to the
upon pH transition to neutral media. Therefore, marketed multiparticulate capsule formulation,
supersaturating the neutral media of the GI fol- Sporanox. As the marketed capsules are far too
lowing stomach emptying may enhance bioavail- large to be dosed to a rodent, ten capsules were
ability. Formulations of itraconazole with opened and the potency per pellet determined. An
polymeric stabilizers were prepared by HME appropriate number of pellets to reach a dose of
and dosed to male Sprague–Dawley rats at a 15 mg/kg were then placed in a size 9 capsule and
nominal dose of 30 mg/kg. Dosing was dosed with the corresponding syringe. Indeed, the
performed by preparing an aqueous dispersion novel particle formulation outperformed the
600.0
Plasma Concentration (mg/ml) 500.0
400.0
300.0
200.0
100.0
0.0
0 6 12 18 24
Time (hr)
Fig. 2.23 In vivo plasma profile of itraconazole dosed to rats via oral gavage of dispersion (square) and adapted
marketed formulation in size 9 capsules (diamond). N ¼ 6. Reproduced with permission from American Chemical
Society
marketed product providing a significant Method Capsule 1

improvement in bioavailability (Fig. 2.23)
pH-Solubility Profile by Direct Determination
(DiNunzio et al. 2008). If a method is to be used
in Aqueous Suspension
in which the drug is dispersed in a liquid carrier, it
Based on the method reported by Li et al.
is important to understand the maximum volume
(2005)
that can be dosed to the selected species without
Objective
interrupting GI function. For example, doses
To determine the pH-solubility profile of halo-
should be kept below 4 mL/kg in rat models to
peridol and the corresponding hydrochlo-
prevent spontaneous release through the pyloric
ride and mesylate salts.
sphincter (Alban et al. 2001).
Haloperidol free base
Haloperidol hydrochloride
2.5 Conclusions Haloperidol mesylate
Deionized water
In order to optimize the formulation of a poorly Hydrochloric acid solution
water-soluble drug, it is imperative to gain an Sodium hydroxide solution
understanding of the physical and chemical 10-mL sealable vials
nature of the compound. This can be accom- Water bath or environmental shaker capable of
plished through preformulation studies, including maintaining 37 C
solubility screenings, solid-state characterization, 0.45-μm Acrodisc filters with attached syringe
dissolution testing, and in vivo studies. Proper UV spectrophotometer (250-nm detection
application of these methods as described above wavelength)
will aid the formulation scientist in directing the Method
development of a poorly water-soluble compound Add 5 mL of water to each 10-mL vial, 9 total,
in the most promising direction. 3 for each compound.
86 X. Ma et al.
Place excess solids (one compound per vial) in Objective

the prefilled vials such that a suspension To determine compatibility of picotamide in
results. the presence of excipients commonly used
Titrate each vial to pH 1 via HCl and NaOH in tableting formulations.
solution addition. Equipment and Reagents
Equilibrate the vials at 37 C for 24 h. Note: Picotamide recrystallized from water–ethanol
agitation recommended during equilibra- 8:1
tion if possible. Excipients: PVP K30, PVPXL, tartaric acid,
Following equilibration, verify that no pH shift ascorbic acid, hydroxypropylmethyl cellu-
has occurred. lose, hydroxyethyl cellulose, sodium
Remove an aliquot of the suspension and filter carboxymethyl cellulose, microcrystalline
through a 0.45-μm (or smaller) filter. cellulose, Veegum F, Arabic gum,
Dilute sample with suitable organic solvent cornstarch
(i.e., acetonitrile) to obtain concentrations Mortar pestle
in the working linear range of the Microbalance (mg scale)
spectrophotometer. Differential scanning calorimeter
Analyze samples on UV spectrophotometer at Aluminum pans with perforated lids
250 nm. Method
Titrate each vial with NaOH and HCl solutions Sieve each material, and obtain the 75–150-μ
to pH 2. m fraction for analysis.
Equilibrate vials for 24 h under identical Prepare individual physical mixtures of the
conditions. API and each excipient (1:1) in a mortar
Confirm no pH shift upon equilibration. using a spatula to gently blend the
Repeat sampling and analysis procedure as components.
above. Prepare co-ground mixtures of each drug–
Continue titration, equilibration, sampling pro- excipient combination by grinding an ali-
cedure through pH range of 1–13, or quot of the corresponding physical mixture
desired regions thereof. in a mortar with a pestle for 10 min.
Results Prepare kneaded mixtures of each drug–excip-
Plotting the solubility (ug/mL) versus pH of ient combination by slurring an aliquot of
the free base and its HCl salt revealed a the corresponding physical mixture with
significant drop in solubility of both ethanol (1–2 mL) and grinding in a mortar
compounds above pH 5. Additionally, the with pestle to obtain a paste. Dry under
plot demonstrates no significant difference vacuum in a desiccator at room temperature
in the aqueous solubility between the free to a constant weight.
base and HCl salt through pH 7. Place aliquots of physical mixtures at 60 C for
The pH versus solubility plot for the mesylate drug–excipient storage stability analysis.
salt of haloperidol revealed significantly Analyze each individual component, physical
higher solubilities of the API through pH mixture, co-ground mixture, kneaded mix-
5. Similar to the free base and HCl salt, ture, and stability sample in the following
solubility of the API decreased significantly manner: Place an aluminum pan (without
above pH 5. lid) on the microbalance and zero the scale.
Fill the pan with approximately 5–10 mg of
Method Capsule 2 the material for analysis. Place the filled
pan back on the balance, and record the
Analysis of Drug–Excipient Interactions via
exact weight of filling. Place a lid in the
filled pan. Place the pan in the DSC appa-
Based on the method reported by Mura et al.
ratus on the sample side. A reference pan
(1998a) Thermochimica Acta
with lid should be placed on the reference Method

cell. Perform scan at 10 K/min from 30 to Pass polymorph I through multiple sieves (i.e.,
200 C. BSS #80, 120, and 240) collecting aliquots
Results of each sieve fraction.
Picotamide monohydrate exhibits an endother- Optimization of scan rate – scan a 5% w/w
mic event at 123.0 2.4 C. mixture of polymorph I in polymorph II
Maintenance of the anhydrous state for physi- over the range of 3–40 2Θ under the fol-
cal mixtures and co-ground mixtures with lowing conditions, and identify the
microcrystalline cellulose, cornstarch, parameters capable of providing the
Methocel, and Ethocel indicates greatest number of identified peaks in the
compatibility. least amount of time:
Hydration of the API in physical mixtures with Step time of 0.5 s, step size of 0.025 .
Veegum and Arabic gum indicates Step time of 0.5 s, step size of 0.0125 .
incompatibility. Step time of 1 s, step size of 0.0125 .
Fresh and stored physical blends with PVP Step time of 5 s, step size of 0.05 .
exhibited no interactions. The presence of Step time of 5 s, step size of 0.0125 .
the dehydration peak at 124 C for ground Using the selected optimized scan rate, ana-
mixtures with PVP-XL indicates incompat- lyze the aliquots of each sieve fraction to
ibility. The absence of the picotamide melt- assess the influence of particle size.
ing endotherm for co-ground mixtures with Results
PVP-K30 indicates amorphization of the A step time of 5 s and step sizes of 0.05 and
API and dissolution into the polymeric 0.0125 allowed identification of four dis-
carrier. tinct peaks.
The presence of acidic excipients yielded Lower step times of 1 s and 0.5 s allowed
broadening and downshift of thermal identification of only two and one peaks,
effects followed by exothermic decomposi- respectively, regardless of step size.
tion for all samples indicating strong The step size of 0.05 was selected as it drasti-
incompatibility. cally reduced the scan time from
246.66 min to 61.66 min.
Method Capsule 3 Particle size significantly influenced the num-
ber of identifiable peaks.
X-Ray Diffraction Parameter Optimization
Sieve fraction BSS #120/240 significantly
Based on the method reported by Tiwari et al.
improved resolution compared to larger
(2007)
particle-size fractions.
Objective
To optimize the scan parameters for X-ray
Method Capsule 4
diffraction analysis including step size and
dwell time as well as to assess the influence Accelerated Stability Monitoring of Amor-
of particle size. phous Solid Dispersions
Equipment and Reagents Based on the method reported by DiNunzio
Olanzapine polymorphs I and II et al. (2010a, b, c)
X-ray diffractometer Objective
Poly(methyl methacrylate) sample holder or To assess formulation stability against recrys-
equivalent tallization upon storage for amorphous
Glass microscope slide or similar itraconazole compositions produced by
hot-melt extrusion and KinetiSol
dispersing.
88 X. Ma et al.
Equipment and Reagents formulations at a single time point for com-

High-density polyethylene (HDPE) bottles or parison. Do this for each time point.
similar with induction sealing capability Results
Oven capable of maintaining 40 1 C Analysis of bulk itraconazole revealed numer-
Saturated salt solution capable of maintaining ous characteristic crystalline peaks between
75% relative humidity within 40 C oven 10 and 35 2Θ.
X-ray diffractometer (XRD) XRD diffraction patterns of formulations
Sample holder for XRD immediately post-production exhibit amor-
Method phous halos and lack any characteristic
Analyze bulk itraconazole and excipients indi- itraconazole peaks.
vidually by XRD over the scan range of Materials produced by KinetiSol dispersing
5–50 2Θ with a step size of 0.05 and a which contained no plasticizer exhibited
dwell time of 3 s. Identify major character- no peak growth over time when stored at
istic peaks to be used in subsequent accelerated conditions.
analysis. Formulations produced by hot-melt extrusion
Analyze physical mixtures of drug and exhibited gradual growth of characteristic
excipients at ratios identical to those used itraconazole peaks, indicating recrystalliza-
for the final formulation by XRD tion upon storage.
employing the same parameters as above.
Identify the major characteristic peaks of Method Capsule 5
itraconazole present in physical mixtures.
Miscibility Prediction by Flory–Huggins
Following formulation production, immedi-
Based on the method reported by Tian et al.
ately obtain XRD profiles for each product
(2012)
using identical scan parameters as before.
Objective
Identify major characteristic peaks of
To predict miscibility of various polymers
itraconazole and corresponding intensities
with a drug substance of interest.
if present.
Place 2 g of a single formulation into a 30-mL
Differential scanning calorimeter (DSC) with
HDPE bottle.
nitrogen purge gas
Prepare three bottles for each formulation for
DSC pans with lids
each of the three time points.
Mortar and pestle
Induction seal each bottle.
Ball mill apparatus
Verify induction seal robustness prior to plac-
Method
ing on stability.
Create drug and polymer mixtures at different
Place all samples in 40 C 75% RH oven.
compositions by first mixing in a mortar
At 1 month remove three samples of each
and pestle and then in the ball mill.
formulation for analysis.
Load samples into individual DSC pans, seal
Allow samples to equilibrate to room temper-
with lids, and poke a hole to allow for water
ature for 24 h.
venting.
Open containers and analyze powders individ-
Analyze samples for melting point of the
ually by XRD using identical parameters as
active drug substance utilizing DSC
described above for characteristic crystal-
parameters appropriate for the drug sub-
line peaks of itraconazole.
stance of interest. 1 C/min is a suitable
Repeat sample removal, equilibration, and
starting scan rate.
analysis at 3 and 6 months.
Generate a plot, and use a linear fit to deter-
Generate a plot for the XRD data of intensity
mine the χ value.
versus angle (degrees 2Θ) for all
Generate a plot of free energy of mixing as a Method

function of drug volume fraction Using the analytical balance, weigh the empty
Generate a phase diagram with spinodal and sample cells, and record the weights.
binodal curves as well Tg as a function of Within the dry box, add powder sample to the
drug weight fraction. BET sample cells. As the device has two
Results cells which can be analyzed simulta-
Drug–polymer interaction parameters, χ, were neously, all powders for analysis are to be
determined as +0.37 for HPMCAS– analyzed in duplicate.
felodipine and 0.37 for Soluplus– Using the analytical balance, record the weight
felodipine indicating weak interactions for of the full sample cell.
the HPMCAS system and miscibility for Attach filled sample cells to degassing station
the Soluplus system. ports.
Free energy of mixing plot revealed that both Engage vacuum.
systems were miscible for all drug loadings Allow samples to degas under vacuum for
at processing temperatures greater than or 12 h.
equal to 140 C. Fill liquid nitrogen Dewar with liquid nitrogen
Free energy of mixing plot also demonstrated to the maximum fill level.
immiscibility of some drug loadings in Repressurize the system, remove sample cells,
HPMCAS systems at/or below 120 C. and, using the analytical balance, immedi-
Additionally, free energy of mixing showed all ately record the weights of the degassed
drug loadings were miscible in Soluplus at samples. Calculate sample weight as:
temperatures greater than or equal to degassed sample cell weight–empty sample
100 C. cell weight.
Phase diagram for the drug–polymer system Attach sample cells to analysis ports of the
constructed to predict miscibility/stability BET apparatus. Verify the level of liquid
of HPMCAS/solubility systems as a func- nitrogen in the Dewar is sufficient for
tion of drug load and temperature. analysis.
Using nitrogen as the adsorptive gas, analyze
Method Capsule 6 the powder samples over the relative pres-
sure range of 0.05–0.30, and use the BET
BET-Specific Surface Area Determination for
equation to fit the adsorption data.
a High-Surface-Area Heat-Liable Material
Results
Based on the method reported by Engstrom
Surface areas of powders produced by spray
et al. (2007)
freezing into liquid ranged from 13 to
Objective
134 m2/g.
To determine the specific surface area of pro-
Increasing the feed concentration decreased
tein powders produced by spray freezing
the specific surface area of the powders.
into liquid.
Increasing the feed concentration decreased
the submicron particle content.
Protein powder produced by spray freezing
Increasing the droplet size during spray freeze-
into liquid
drying resulted in lower specific surface
Quantachrome Nova 2000 BET apparatus
areas.
including sample cells
Dry box
Method Capsule 7
Liquid nitrogen
Nitrogen gas: high purity, dry ssNMR for Polymorphism Analysis
Analytical balance Based on the method reported by Dempah
et al. (2013)
90 X. Ma et al.
Objective and 61 s with thermal stress) which

To identify the physical forms generated dur- correlated with medium stability
ing milling of gabapentin API including the (0.01–0.1% lactam formation).
primary degradation product, gabapentin Milled samples with HPC had short relaxation
lactam. times (11 s without thermal stress and 14 s
Equipment and Reagents with thermal stress) which correlated with
Gabapentin (forms I, II, and III) low stability (>0.4% lactam formation).
Gabapentin lactam In the presence of HPC, form II converted to
Hydroxypropylcellulose (HPC) the metastable form III.
3-Methylglutaric acid
Pulviserette 7, Planetary Micro Mill Method Capsule 8
Four hardened steel balls (15 mm)
Supersaturation Dissolution Studies Using the
Apparatus capable of maintaining samples at
Syringe/Filter and Microcentrifuge
50 C/0% RH
Methods
ssNMR, Bruker Advance 300
Based on the method reported by Curatolo
7 mm-zirconia rotor
et al. (2009)
Method
Objective
Mill a portion of gabapentin (form II) for
To determine the ability of HPMCAS in initi-
45 min with and without 6.5% HPC in
ation and maintenance of supersaturation of
planetary mill.
an experimental compound.
Thermally stress a portion of each milled sam-
ple in 50 C/0% RH for 24 h.
Experimental compound CMPD 2
Pack each sample under ambient conditions
HPMCAS-MF
into zirconia rotor.
10-mL syringes
Operate at 13C frequency of ~75 MHz.
Model fasted duodenal fluid preheated to
Standardize ssNMR system with
37 C
3-methylglutaric acid at a reference angle
Oven capable of maintaining 37 C
of 18.84 ppm.
Wheel capable of rotating syringe in horizontal
Acquire sample spectra using cross-
position at 50 rpm
polarization and magic angle spinning
20-gauge hypodermic needles
(CP/MAS), SPINAL-64 decoupling, and
13-mm, 0.45-μm polyvinylidene difluoride
total sideband suppression.
syringe filters
Use contact time of 1 ms, MAS frequency of
Test tubes
4.0 kHz, and 1H decoupling field of
Polypropylene microcentrifuge tubes
70–80 kHz.
Microcentrifuge
Measure 1H relaxation times by saturation
Vortex mixer
recovery.
Small-volume pipette (i.e., 10–100 μL)
Results
Diluting solution: 60:40 1.7 wt.% ammonium
ssNMR identification spectra for each of the
ascorbate/acetonitrile
physical forms of gabapentin (I, II, and III)
HPLC: Phenomenex ultracarb ODS 20 analyti-
and gabapentin lactam.
cal column, PDA detection at 215 nm
Unmilled samples had long relaxation times
Method
(134 s for neat form II and 130 s for form
Syringe/filter method.
II with HPC) which correlated with high
Accurately weigh 7.5 mg of 67% CMPD 2/
stability (<0.01% lactam formation).
HPMCAS-MF formulation, and add to an
Milled samples without HPC had medium
empty 10-mL syringe with attached
relaxation times (41s without thermal stress
20-guage needle.
Draw 10 mL of model fasted duodenal fluid The 67% compound 2 solid dispersion formu-
preheated to 37 C into the syringe via the lation with HPMCAS resulted in supersatu-
attached needle. ration of the test medium.
Replace attached needle with 13-mm syringe Maximum concentrations of approximately
filter. 130 μg/mL were achieved.
Shake syringe vigorously for 30 s. HPMCAS-MF initiated and maintained super-
Expel six drops of the solution as waste. Col- saturation of compound 2 to a greater extent
lect drops 7–19 as a sample. than HPC, PVAP, or the crystalline
Draw syringe plunger back to generate an air bulk drug.
bubble. Both methods proved successful in achieving
Place syringe on rotating wheel (50 rpm) in the supersaturation.
37 C oven.
Dilute sample 1:1 with diluting solvent. Method Capsule 9
Repeat sampling procedure at t ¼ 5, 10, 20,
Biphasic Dissolution Testing Utilizing an
40, 90, 180 min.
External Flow through Cell
Analyze samples on HPLC to quantify CMPD
Based on the method reported by Shi et al.
2.
(2010)
Microcentrifuge method
Objective
In a 37 C controlled-temperature box, weigh
• To examine the dissolution profiles of three
1.8 mg of formulation into a
celecoxib formulations using a biphasic
microcentrifuge tube.
dissolution-testing method incorporating
Add 1.8 mL of model fasted duodenal fluid to
an organic phase for drug partitioning.
the tube.
Close tube, start timer, and vortex mix for 60 s.
• Octanol
Transfer tube to microcentrifuge, allow to
• Sodium phosphate monobasic
stand for 6 min, and then centrifuge at
monohydrate
13,000 g for 60 s.
• Sodium hydroxide
At the 10-min mark on the timer, remove a
• Gelatin capsules
25-μL sample from the supernatant via a
• Commercial Celebrex capsules (200-mg
pipette. Immediately dilute 1:1 with dilut-
dose strength)
ing solution.
• Extracted celecoxib (extracted from
Resuspend the material via vortex mixing for
Celebrex capsules via ethanol and
30 s.
subsequent evaporation method)
Place tube back in centrifuge. Allow tube to
• USP Dissolution Apparatus II
stand undisturbed until the next sampling
• USP IV flow through cell
time point.
• Piston pump
At each sampling time point, centrifuge the
• Teflon tubing
tube for 60 s, remove supernatant sample,
• Modified apparatus II paddle incorporating
and resuspend as described. Dilute sample
a second paddle capable of agitating the
1:1 with diluting solution.
organic layer during dissolution testing
Analyze all samples via HPLC to quantify
• HPLC (mobile phase 55:45 v/v acetonitrile/
CMPD 2.
ammonium acetate solution)
Plot the dissolved drug concentration versus
Method
time dissolution profile for the API.
• Place 250 mL of 80 mM phosphate buffer
Results
(pH 6.8) in the dissolution vessel.
Compound 2 has an aqueous solubility of
• Add 200 mL of octanol to the dissolution
1 μg/mL.
vessel.
92 X. Ma et al.
• Prior to beginning the study, saturate the exhibited no correlation with in vivo AUC
aqueous phase with octanol and vice versa values.
by agitating the mixture for 30 min. • Analysis of the octanol phase from biphasic
• Allow all media to equilibrate to studies revealed a rank-order correlation to
37 0.2 C. in vivo results.
• Place the Teflon tubing (inlet and outlet)
such that the ends are well within the aque- Method Capsule 10
ous phase of the dissolution media.
Supersaturation Dissolution Testing of Amor-
• Set the paddle speed to 75 rpm.
phous Compositions of a Poorly Water-
• Place the formulation for analysis into the
Soluble Drug
flow through cell.
Based on the method reported by DiNunzio
• Set the pump flow rate to 30 mL/min.
et al. (2008)
• At time points of 15, 30, 45, 60, 75, 90, and
Objective
120 min, remove a 1-mL sample from the
To assess the supersaturation extent and dura-
aqueous phase and a 100-μL sample from
tion of amorphous compositions of
the organic phase. Do not replace the
itraconazole using concentration-enhancing
media.
polymers.
• Immediately centrifuge the aqueous phase
samples at 14,000 rpm for 6 min. Collect
Itraconazole; bulk drug and formulations
the supernatant for HPLC analysis.
prepared via thin film freezing
• Immediately dilute the organic phase
Commercial itraconazole capsules (Sporanox)
samples 100-fold with HPLC mobile phase.
Size 9 porcine gelatin capsules
• Quantitative celecoxib via HPLC analysis;
USP Dissolution Apparatus II with
adjustment for dilution mathematically.
autosampler
Results
0.1N hydrochloric acid
• Corresponding single-phase studies (aque-
0.2M Na3PO4 solution
ous; USP Apparatus II) under sink
0.2-μm PTFE membrane, 13 mm Acrodisc
conditions were not discriminatory for for-
syringe filters
mulation performance.
5-mL syringes
• Nonsink two-phase dissolution revealed the
HPLC System (5 μm C18(2) 100 Å,
same formulation performance rank order
150 mm 4.6 mm column, flow rate of
in the aqueous phase.
1 mL/min, detection at 263 nm)
• Analysis of the octanol phase reveals the
HPLC mobile phase (70:30:0.05 acetonitrile/
self-emulsifying drug delivery system
water/diethanolamine)
(SEDDS) outperformed the solution and
HPLC vials
capsule formulations.
Vortex mixer
• SEDDS provided a higher amount of free
Method
drug compared to the solution formulation
Preheat 735 mL of 0.1N HCl in each dissolu-
in which the drug was associated as surfac-
tion vessel to 37 C.
tant micelles.
Accurately weigh an amount of formulation
• SEDDS formulation supersaturated the
equivalent to 37.5 mg of itraconazole.
aqueous media under nonsink-biphasic dis-
This corresponds to approximately 10
solution conditions, allowing greater
the equilibrium solubility in 0.1N HCl.
partitioning into the octanol phase.
Pre-wet the weighed powder with 15 mL of
• Aqueous phase AUC values from both sin-
heated 0.1N HCl (37 C).
gle and biphasic dissolution testing
Add the pre-wetted powder slurry to the disso-
lution vessel.
After 2 h, add 250 mL of 0.2 M Na3PO4 to BCG vaccine spray-dried powder

each vessel to bring the pH to 6.8. Polyethylene tubing (1.19-mm diameter) cut
At time points of 60, 120, 130, 150, 180, 240, into lengths of approximately 1 cm
300, 360, and 1440 min, a 5-mL sample is Microbalance
to be taken via the autosampler without Penn-Century dry powder insufflator model
replacement of the withdrawn media. DP-4 M
Immediately filter samples following with- Penn-Century model AP-1 air pump
draw through a 0.2-μm PTFE membrane. Rodent work stand
Immediately dilute filtrate 1:1 with mobile Lidocaine applicator
phase and vortex mix. Transfer to an Mouse-sized speculum
HPLC vial. Plastic incisor loop
Analyze all samples by HPLC to quantify Otoscope set
itraconazole adjusting for the volume Ketamine HCl, xylazine, and acepromazine
change and dilution mathematically. Yohimbine solution (0.04 mg/mL)
Results Syringe and 21-gauge needle
Sporanox pellets were able to rapidly and Cage maintained at 37 C
extensively supersaturate the acidic media. Method
Following the pH transition, the drug rap- Prepare mixture of ketamine/xylazine/
idly precipitated. acepromazine by mixing 2-mL ketamine
Thin film freezing formulations resulted in at 100 mg/mL, 0.4 mL of xylazine at
significantly reduced acidic media 100 mg/mL, and 0.6 mL of acepromazine
concentrations as the polymers used were at 10 mg/mL with 7 mL of sterile
enteric. phosphate–buffered saline.
Following pH change, cellulose acetate Anesthetize the mouse via an intraperitoneal
phthalate and polyvinyl acetate phthalate (i.p.) injection (100 μL/20 g body weight)
formulations showed the greatest supersat- of ketamine/xylazine/acepromazine
uration in neutral media. mixture.
Higher ratios of itraconazole/polymer yielded Load 200–300 μg of powder into the 1-cm
lower degrees of supersaturation. segment of polyethylene tubing by dipping
Cellulose acetate phthalate formulations the end vertically into the powder bed 4–6
provided longer half-life values for drug in times.
solution, indicating strong concentration- Place the filled polyethylene tube into the hole
enhancing properties. within the insufflator chamber.
Attach base of insufflator to the cannula.
Method Capsule 11 Place the mouse in the supine position on the
mouse work station, and place the incisor
Pulmonary Delivery to the Murine Model via
loop and lateral supports in position.
Dry Powder Insufflation
Raise the stand to 60 .
Based on the method reported by Morello et al.
Using the otoscope, obtain a clear view of the
(2009)
trachea.
Objective
Using the speculum to guide the applicator,
To achieve successful pulmonary administra-
apply lidocaine to the arytenoid cartilage.
tion of the tuberculosis vaccine (BCG) to
Re-obtain visual focus of the trachea with the
the murine model by dry powder
otoscope.
insufflation.
Insert the cannula of the insufflator device into
the tracheal opening.
Female BALB/c mice (6–12 weeks old,
Remove the otoscope, and attach the air pump.
18–24 g)
94 X. Ma et al.
Depress plunger on air pump 4–5 times. Mon- Mini-capillary tubes containing EDTA
itor rise and fall of the upper chest as con- dipotassium salt
firmation of proper insufflation. Method
Remove mouse from the work stand, and give Weigh 5.2 mg of letrozole (uncomplexed) into
an i.p. injection of 0.11 mg/kg yohimbine a sterile glass vial.
solution (0.04 mg/mL). Immediately before dosing, dilute with 12.5%
Provide 300-μL 0.9% saline subcutaneously to ethanol in water to a concentration of 1 mg/
aid recovery. mL using sonication to disperse.
Move mouse to a 37 C cage until awake. Weigh 626.2-mg complexed letrozole–
Results HBenβCD into a glass vial.
Application of this method allowed rapid Immediately before dosing, dilute with 12.5%
administration of the compound relative to ethanol in water to generate a 1 mg/mL
similar methods. solution.
90% of loaded powder was delivered to the With the gavage needle attached to the
lungs. syringe, draw in the desired dose of the
91% of the loaded dose reached the lungs. suspension or solution to be administered.
Minor pulmonary damage occurred as a result Holding the rat vertically while supporting the
of the procedure; however, mice were hind legs against the body, insert the
asymptomatic and congestion/hemorrhage gavage needle extending it into the stom-
resolved over time. ach, and expel the dose. Immediately
The method developed can be modified for use remove the gavage needle from the animal.
with larger animals by adjusting the dose Repeat dosing for all rats.
and air volumes employed. At time points of 0.45, 1.5, 2, 2.6, 2.9, 3.6, 5.1,
6.2, 8.3, 13.9, 24.8, and 36.3 h, collect
Method Capsule 12 125-μL blood samples through the tail
vein directly into mini-capillary tubes
Oral Drug Delivery of a Poorly Water-Soluble
containing EDTA dipotassium salt.
Drug to the Rat Model
Immediately following blood sample collec-
Based on the method reported by Wempe et al.
tion, cap the tube, and place on dry ice.
(2007)
Keep samples frozen at 80 C until sam-
Objective
ple preparation for analysis.
To determine the bioavailability of letrozole
Results
complexed with hydroxybute-
Letrozole was eliminated from the blood of
nyl-β-cyclodextrin (HBenβCD) compared
male rats within 36 h following dosing via
to that of the native API via an oral dose
oral gavage.
administered as a suspension or solution to
The pharmacokinetics of letrozole are strongly
the rat model.
gender-dependent.
Dosing of letrozole–HBenβCD yielded a two-
Letrozole, complexed to HBenβCD and
fold increase in the AUC compared to a
unprocessed
suspension of the uncomplexed drug.
Male and female Sprague–Dawley rats
Oral dosing of the complexed formulation
(260–294 and 218–262 g, respectively)
increased the Cmax from 87 to 140 ng/mL.
1-mL syringe with 0.01-mL graduations
The Tmax was decreased from 8.4 to 6.3 h by
Sterile glass vials
the HBenβCD formulation.
Ethanol
Solubility limits the rate and extent of absorp-
Sterile water
tion in male rats while only limiting the rate
Sonicator
of absorption in female rats.
Oral gavage needle (16 gauge)
Complexation of letrozole with HBenβCD by 13C-NMR relaxation time. Chem Pharm Bull.
improved oral absorption in male rats 2002;50(6):822–6.
Aso Y, Yoshioka S, Miyazaki T, Kawanishi T. Feasibility
while maximizing absorption in of 19F-NMR for assessing the molecular mobility of
female rats. flufenamic acid in solid dispersions. Chem Pharm
Bulln. 2009;57(1):61–4.
Avdeef A, Berger CM, et al. pH-metric solubility. 2. Cor-
relation between the acid-base titration and the satura-
References tion shake-flask solubility-pH methods. Pharm Res.
2000;17(1):85–9.
Baena Y, Barbosa HJ, Jorge A, Martinez F. Estimation of
Abbas D, Kaloustian J, et al. DSC and physico-chemical
the aqueous solubility of some acetanilide derivatives
properties of a substituted pyridoquinoline and its
from octanol-water partition coefficients and entropies
interaction study with excipients. J Therm Anal
of fusion. Acta Farmaceutica Bonaerense. 2004;23(1):
Calorim. 2008;93(2):353–60.
33–8.
Abiad MG, Campanella OH, Carvajal MT. Assessment of
Bates S, Kelly RC, Ivanisevic I, Schields P, Zografi G,
thermal transitions by dynamic mechanical analysis
Newman AW. Assessment of defects and amorphous
(DMA) using a novel disposable powder holder.
structure produced in raffinose pentahydrate upon
Pharmaceutics. 2010;2(2):78–90.
dehydration. J Pharm Sci. 2007;96(5):1418–33.
Agatonovic-Kustrin S, Markovic N, et al. Compatibility
Baucells M, Ferrer N, Gomez P, Lacort G, Roura
studies between mannitol and omeprazole sodium
M. Determination of caffeine in solid pharmaceutical
isomers. J Pharm Biomed Anal. 2008;48(2):356–60.
samples by FTIR spectroscopy. Mikrochim Acta.
Aho J, Hvidt S, Baldursdottir S. Rheology in pharmaceu-
1993;112:87–98.
tical sciences. In: Analytical techniques in the pharma-
Belal F, Elbrashy A, Eid M, Nasr JJ. Stability-indicating
ceutical sciences. Cham: Springer; 2016. p. 719–50.
HPLC method for the determination of quetiapine:
Ahuja S, Rasmussen H. HPLC method development for
application to tablets and human plasma. J Liq
pharmaceuticals. Boston: Elsevier; 2007.
Chromatogr Relat Technol. 2008;31:1283–98.
Ahumada AA, Seeck J, Allemandi D, Manzo RH. The
Bellantone RA, Patel P, Sandhu H, Choi DS, Singhal D,
pH/solubility profile of norfloxacin. STP Pharm Sci.
Chokshi H, Malick AW, Shah N. A method to predict
1993;3(3):250–3.
the equilibrium solubility of drugs in solid polymers
Alban L, Dahl AK, Hansen AK, Hejgarrd KC, Jensen AL,
near room temperature using thermal analysis. J Pharm
Kragh M, Thomsen P, Steensgaard P. The welfare
Sci. 2012;101:4549–58.
impact of increased gavaging doses in rats. Anim Wel-
Bergese P, Bontempi E, et al. Microstructural investigation
fare. 2001;10(3):303–14.
of nimesulide-crospovidone composites by X-ray dif-
Aldén M, Wulff M, et al. Influence of selected variables on
fraction and thermal analysis. Compos Sci Technol.
heat of fusion determinations by oscillating DSC.
2003;63(8):1197–201.
Thermochim Acta. 1995;265:89–102.
Bhalekar MR, Shete TK, et al. Solid state
Allen T. Particle size measurement, Surface area and pore
photodegradation study of fexofenadine hydrochlo-
size determinations, vol. 2. London: Chapman &
ride. Anal Lett. 2010;43(3):406–16.
Hall; 1997.
Bhushan B, Fuchs H. Applied scanning probe methods
Alves R, TVDS R, Cides da Silva LC, Storpirtis S,
II. Berlin: Springer; 2007.
Mercuri LP, Matos JDR. Thermal behavior and
Binnig G, Quate CF, Gerber C. Atomic force microscope.
decomposition kinetics of rifampicin polymorphs
Phys Rev Lett. 1986;56(9):930.
under isothermal and non-isothermal conditions. Braz
Bizzi CA, Barin JS, Hermes AL, Mortari SR, Flores
J Pharm Sci. 2010;46(2):343–51.
EMM. A fast microwave-assisted procedure for loss
Andronis V, Zografi G. Molecular mobility of supercooled
on drying determination in saccharides. J Braz Chem
amorphous indomethacin, determined by dynamic
Soc. 2011;22(2):376–81.
mechanical analysis. Pharm Res. 1997;14(4):410–4.
Boobier S, Hose DRJ, Blacker AJ, Nguyen BN. Machine
Antila R, Yliruusi J. Measurement of specific surface area
learning with physicochemical relationships: solubility
of pharmaceutical powders by the BET method; effect
prediction in organic solvents and water. Nat Commun.
of drying time and drying temperature. Acta Pharm
2020;11(1)
Nord. 1991;3(1):15–8.
Bott RF, Oliveira WP. Storage conditions for stability
Aso Y, Yoshioka S, Kojima S. Relationship between the
testing of pharmaceuticals in hot and humid regions.
crystallization rates of amorphous nifedipine, pheno-
Drug Dev Ind Pharm. 2007;33:393–401.
barbital, and flopropione, and their molecular mobility
Boyd BJ, Bergström CA, Vinarov Z, Kuentz M,
as measured by their enthalpy relaxation and 1H NMR
Brouwers J, Augustijns P, Brandl M, Bernkop-
relaxation times. J Pharm Sci. 2000;89(3):408–16.
Schnürch A, Shrestha N, Préat V. Successful oral
Aso Y, Yoshioka S, Zhang J, Zografi G. Effect of water on
delivery of poorly water-soluble drugs both depends
the molecular mobility of sucrose and poly (vinylpyr-
on the intraluminal behavior of drugs and of
rolidone) in a colyophilized formulation as measured
96 X. Ma et al.
appropriate advanced drug delivery systems. Eur J Corrandini D, Phillips TM. Handbook of HPLC. Boca
Pharm Sci. 2019;137:104967. Raton: Taylor & Francis; 2011.
Boyer RF, Simha R. Relation between expansion Craig DQM, Reading M. Thermal analysis of
coefficients and glass temperature: a reply. J Polym pharmaceuticals. Boca Raton: CRC Press; 2007.
Sci Polym Lett Ed. 1973;11(1):33–44. Cui Q, Lu S, Ni B, Zeng X, Tan Y, Chen YD, Zhao
Brits M, Liebenberg W, de Villiers MM. Characterization H. Improved prediction of aqueous solubility of novel
of polymorphic transformations that decrease the sta- compounds by going deeper with deep learning. Front
bility of tablets containing the WHO essential drug Oncol. 2020;10
mebendazole. J Pharm Sci. 2010;99(3):1138–51. Cuiping C, Xingrong L, Rujin W. High-performance liq-
Brown ME. Introduction to thermal analysis. Dordrecht: uid chromatographic method for the determination of
Kluwer; 2001. norfloxacin glutamage and glucuronate in solid and
Bruni G, Berbenni V, et al. Determination of the liquid dosage forms and its application to stability
nateglinide polymorphic purity through DSC. J testing. J Pharm Biomed Anal. 1993;11(8):717–21.
Pharm Biomed Anal. 2011;54(5):1196–9. Curatolo W, Nightingale JA, Herbig SM. Utility of
Buffeteau T, Desbat B, Eyquem D. Attenuated total reflec- hydroxypropylmethylcellulose acetate succinate
tion fourier transform infrared microspectroscopy: the- (HPMCAS) for initiation and maintenance of drug
ory and application to polymer samples. Vib Spectrosc. supersaturation in the GI Milieu. Pharm Res. 2009;26
1996;11:29–36. (6):1419–31.
Cameron D, Armstrong EE. Optimization of stepsize in D’Souza AJM, Ford BM, et al. Biophysical characteriza-
X-ray powder diffractogram collection. Powder Diffr. tion and formulation of F1–V, a recombinant plague
1988;3(1):32–8. antigen. J Pharm Sci. 2009;98(8):2592–602.
Cammenga HK, Eysel W, et al. The temperature calibra- Davis SS. Viscoeldstic properties of pharmaceutical
tion of scanning calorimeters: part 2. calibration semisolids III: Nondestructive oscillatory testing. J
substances. Thermochim Acta. 1993;219:333–42. Pharm Sci. 1971;60(9):1351–6.
Campbell Roberts SN, Williams AC, et al. Quantitative de Villiers MM, Terblanche RJ, Liebenberg W,
analysis of mannitol polymorphs. X-ray powder Swanepoel E, Dekker TG, Song M. Variable-
diffractometry–exploring preferred orientation effects. temperature X-ray powder diffraction analysis of the
J Pharm Biomed Anal. 2002;28(6):1149–59. crystal transformation of the pharmaceutically pre-
Cappello B, di Maio C, Iervolino M, Miro A. Combined ferred polymorph C of mebendazole. J Pharm Biomed
effect of hydroxypropyl methylcellulose and Anal. 2005;38:435–41.
hydroxypropyl-beta-cyclodextrin on physicochemical Dempah KE, Barich DH, Kaushal AM, Zong Z, Desai SD,
and dissolution properties of celecoxib. J Incl Phenom Suryanarayanan R, Kirsch L, Munson EJ. Investigating
Macrocycl Chem. 2007;59:237–44. gabapentin polymorphism using solid-state NMR
Carpenter J, Katayama D, Liu L, Chonkaew W, Menard spectroscopy. AAPS PharmSciTech. 2013;14(1):
K. Measurement of Tg in lyophilized protein and pro- 19–28.
tein excipient mixtures by dynamic mechanical analy- Diebold SM, Dressman JB. Dissolved oxygen as a mea-
sis. J Therm Anal Calorim. 2009;95(3):881–4. sure for de- and reaeration of aqueous media for disso-
Chan HK, Ongpipattanakul B, Au-Yeung J. Aggregation lution testing. Dissolution Technol. 1998;5(3):13–6.
of rhDNase occured during the compression of KBr Dinnebier RE, Billinge S. Powder diffraction: theory and
pellets used for FTIR spectroscopy. Pharm Res. practice. Cambridge: Royal Society of
1996;13(2):238–42. Chemistry; 2008.
Charsley EL, Laye PG, et al. DSC studies on organic DiNunzio JC, Miller DA, Yang W, McGinity JW,
melting point temperature standards. Thermochim Williams RO III. Amorphous compositions using con-
Acta. 2006;446(1–2):29–32. centration enhancing polymers for improved bioavail-
Chawla G, Bansal AK. Improved dissolution of a poorly ability of itraconazole. Mol Pharm. 2008;5(6):968–80.
water soluble drug in solid dispersion with polymeric DiNunzio JC, Brough C, et al. Fusion production of solid
and non-polymeric hydrophilic additives. Acta Pharm. dispersions containing a heat-sensitive active ingredi-
2008;58:257–74. ent by hot melt extrusion and Kinetisol® dispersing.
Chinta S, Rengaswamy R. Machine learning derived Eur J Pharm Biopharm. 2010a;74(2):340–51.
Quantitative Structure Property Relationship (QSPR) DiNunzio JC, Brough C, Miller DA, Williams ROIII,
to predict drug solubility in binary solvent systems. Ind McGinity JW. Applications of KinetiSol dispersing
Eng Chem Res. 2019;58(8):3082–92. for the production of plasticizer free amorphous solid
Chitlange SS, Bagri K, Sakarkar DM. Stability indicationg dispersions. Eur J Pharm Sci. 2010b;40(3):179–87.
RP-HPLC method for simultaneous ESTIMATION of DiNunzio JC, Hughey JR, Brough C, Miller DA, Williams
valsartan and amlodipine in capsule formulation. Asian RO III, McGinity JW. Production of advanced solid
J Res Chem. 2008;1(1):15–8. diserpsions for enhanced bioavailability of
Condon JB. Surface area and porosity determinations by itraconazole using kinetisol dispersing. Drug Dev Ind
physisorption: measurements and theory. Amsterdam: Pharm. 2010c;36(9):1064–78.
Elsevier; 2006.
Dios Lopez-Gonzalez J, Carpenter FG, Deitz acetate succinate-based spray-dried dispersions: an

VR. Adsorption of nitrogen on carbon adsorbents at overview. Mol Pharm. 2008;5(6):1003–19.
low pressures between 69 and 90 K. J Res Natl Bur Gad SC. Pharmaceutical manufacturing handbook:
Stand (US). 1955;55:11–8. regulations and quality. Hoboken: Wiley; 2008.
Duddu SP, Weller K. Importance of glass transition tem- Galata DL, Farkas A, Könyves Z, Mészáros LA, Szabó E,
perature in accelerated stability testing of amorphous Csontos I, Pálos A, Marosi G, Nagy ZK, Nagy B. Fast,
solids: case study using a lyophilized aspirin formula- spectroscopy-based prediction of in vitro dissolution
tion. J Pharm Sci. 1996;85(3):345–7. profile of extended release tablets using artificial neural
Dwivedi AM. Residual solvent analysis in networks. Pharmaceutics. 2019;11(8):400.
pharmaceuticals. Pharm Technol November. 2002: Galata DL, Könyves Z, Nagy B, Novák M, Mészáros LA,
42–6. Szabó E, Farkas A, Marosi G, Nagy ZK. Real-time
Dykhne T, Taylor R, Florence A, Billinge SJ. Data release testing of dissolution based on surrogate
requirements for the reliable use of atomic pair distri- models developed by machine learning algorithms
bution functions in amorphous pharmaceutical finger- using NIR spectra, compression force and particle
printing. Pharm Res. 2011;28(5):1041–8. size distribution as input data. Int J Pharm. 2021;597:
Eaton P, West P. Atomic force microscopy. Oxford: 120338.
Oxford University Press; 2010. Gao Z, Moore TW, et al. Effects of deaeration methods on
Ekins S. Pharmaceutical applications of Raman spectros- dissolution testing in aqueous media: a study using a
copy. Chichester: Wiley; 2008. total dissolved gas pressure meter. J Pharm Sci.
Engstrom JD, Simpson DT, Lai ES, Williams RO III, 2006;95(7):1606–13.
Johnston KP. Morphology of protein particles pro- Gordon M, Taylor JS. Ideal copolymers and the second-
duced by spray freezing of concentrated solutions. order transitions of synthetic rubbers. I. Noncrystalline
Eur J Pharm Biopharm. 2007;64:149. copolymers. J Appl Chem. 1952;2:493–500.
Esteva A, Kuprel B, Novoa RA, Ko J, Swetter SM, Blau Goskonda VR, Reddy IK, Durrant MJ, Wilber W, Khan
HM, Thrun S. Dermatologist-level classification of MA. Solid-state stability assessment of controlled
skin cancer with deep neural networks. Nature. release tablets containing Carbopol 971P. J Control
2017;542(7639):115–8. Release. 1998;54:87–93.
Fagerlund G. Determination of specific surface by the BET Gregg SJ, Sing KSW. Adsorption surface area and poros-
method. Materiaux et Constructions. 1973;6(33): ity. New York: Academic; 1982.
239–45. Griffin RG, Prisner TF. High field dynamic nuclear
Fahr A, Liu X. Drug delivery strategies for poorly water- polarization – the renaissance. Phys Chem Chem
soluble drugs. Exp Opin Drug Deliv. 2007;4(4): Phys. 2010;12(22):5737–40.
403–16. Guo Y, Shalaev E, Smith S. Physical stability of pharma-
Ferrari PC, Oliveira GF, Chibebe FCS, Evangelista RC. In ceutical formulations: solid-state characterization of
vitro characterization of coevaporates containing amorphous dispersions. Trends Analyt Chem.
chitosan for colonic drug delivery. Carbohydr Polym. 2013;49:137–44.
2009;78:557–63. Han R, Yang Y, Li X, Ouyang D. Predicting oral
Ferraro JR. Introductory raman spectroscopy. Amsterdam: disintegrating tablet formulations by neural network
Academic Press; 2003. techniques. Asian J Pharm Sci. 2018;13(4):336–42.
Ferry JD. Viscoelastic properties of polymers. New York: Hartauer KJ, Miller ES, et al. Diffuse reflectance infrared
Wiley; 1980. fourier transform spectroscopy for the quantitative
Filho R, Franco P, et al. Stability studies on nifedipine analysis of mixtures of polymorphs. Int J Pharm.
tablets using thermogravimetry and differential scan- 1992;85(1–3):163–74.
ning calorimetry. J Therm Anal Calorim. 2009;97(1): Haugstad G. Atomic force microscopy: understanding
343–7. basic modes and advanced applications. Hoboken:
Forster A, Apperley D, Hempenstall J, Lancaster R, Rades Wiley; 2012.
T. Investigation of the physical stability of amorphous Hauss DJ. Oral lipid-based formulations. Adv Drug Deliv
drug and drug/polymer melts using variable tempera- Rev. 2007;59(7):667–76.
ture solid state NMR. Pharmazie. 2003;58(10):761–2. He Y, Ye Z, Liu X, Wei Z, Qiu F, Li H-F, Zheng Y,
Fortunato D. Dissolution method development for imme- Ouyang D. Can machine learning predict drug
diate release solid oral dosage forms. Dissolution nanocrystals? J Control Release. 2020;322:274–85.
Technol. 2005; Heigoldt U, Sommer F, et al. Predicting in vivo absorption
Fotaki N, Symillides M, et al. Canine versus in vitro data behavior of oral modified release dosage forms
for predicting input profiles of l-sulpiride after oral containing pH-dependent poorly soluble drugs using
administration. Eur J Pharm Sci. 2005;26(3–4): a novel pH-adjusted biphasic in vitro dissolution test.
324–33. Eur J Pharm Biopharm. 2010;76(1):105–11.
Friesen DT, Shanker R, Crew M, Smithey DT, Curatolo Heng D, Cutler D, et al. What is a suitable dissolution
WJ, Nightingale JAS. Hydroxypropyl methylcellulose method for drug nanoparticles? Pharm Res. 2008;25
(7):1696–701.
98 X. Ma et al.
Hensel A, Dobbertin J, Schawe J, Boller A, Schick Kawabata Y, Wada K, Nakatani M, Yamada S, Onoue
C. Temperature modulated calorimetry and dielectric S. Formulation design for poorly water-soluble drugs
spectroscopy in the glass transition region of polymers. based on biopharmaceutics classification system: Basic
J Therm Anal Calorim. 1996;46(3–4):935–54. approaches and practical applications. Int J Pharm.
Hill VL, Craig DQ, Feely LC. The effects of experimental 2011;420:1–10.
parameters and calibration on MTDSC data. Int J Keserü GM, Makara GM. The influence of lead discovery
Pharm. 1999;192(1):21–32. strategies on the properties of drug candidates. Nat Rev
Huang S, O’Donnell KP, Keen JM, Rickard MA, Drug Discov. 2009;8(3):203.
McGinity JW, Williams RO III. A new extrudable Khan MA, Karnachi AA, Agarwal V, Vaithiyalingam SR,
form of Hypromellose: AFFINISOL™ HPMC HME. Nazzal S, Reddy IK. Stability characterization of con-
AAPS PharmSciTech. 2015:1–14. trolled release coprecipitates and solid dispersions. J
Hughey JR, DiNunzio JC, Bennett RC, Brough C, Miller Control Release. 2000;63:1–6.
DA, Hua M, Williams RO III, McGinity Khomane K, Kumar L, et al. NP-647, a novel TRH ana-
JW. Dissolution enhancement of a drug exhibiting logue: investigating physicochemical parameters criti-
thermal and acidic decomposition characteristics by cal for its oral and parenteral delivery. Int J Pharm.
fusion processing: a comparative study to hot melt 2011;406(1–2):21–30.
extrusion and kinetisol dispersing. AAPS Kimura N, Fukui H, Takagaki H, Yonemochi E, Terada
PharmSciTech. 2010;11(2):760–74. K. Characterization of polymorphs of a novel
Huuskonen J, Salo M, Taskinen J. Neural network quinolinone derivative, TA-270 (4-Hydroxy-1-
modeling for estimation of the aqueous solubility of methyl-3-octyloxy-7-sinapinoylamino-2(1 H)-
structurally related drugs. J Pharm Sci. 1997;86(4): quinolinone). Chem Pharm Bull. 2001;49(10):1321–5.
450–4. Kirk JH, Blatchford CG. Lactose: a definitive guide to
Igwe GJI. Powder technology and multiphase systems: gas polymorph determination. Int J Pharm. 2007;334:
permeametry and surface area measurement. 103–14.
Chichester: Ellis Horwood Limited; 1991. Konno H, Handa T, Alonzo DE, Taylor LS. Effect of
Jain N, Yalkowsky SH. Estimation of the aqueous polymer type on the dissolution profile of amorphous
solubility I: application to organic nonelectrolytes. J solid dispersions containing felodipine. Eur J Pharm
Pharm Sci. 2001;90(2):234–52. Biopharm. 2008;70:493–9.
Jantratid E, Janssen N, et al. Dissolution media simulating Lakshman JP, Cao Y, Kowalski J, Serajuddin ATM.
conditions in the proximal human gastrointestinal tract: Application of melt extrusion in the development of a
an update. Pharm Res. 2008;25(7):1663–76. physically and chemically stable high-energy amor-
Jermain SV, Brough C, Williams RO III. Amorphous solid phous solid dispersion of a poorly water-soluble drug.
dispersions and nanocrystal technologies for poorly Mol Pharm. 2008;5(6):994–1002.
water-soluble drug delivery–an update. Int J Pharm. Lauer ME, Grassmann O, Siam M, Tardio J, Jacob L,
2018;535:379–92. Page S, et al. Atomic force microscopy-based screen-
Jones DS, Brown AF, David Woolfson A. Solute and ing of drug-excipient miscibility and stability of solid
solvent effects on the thermorheological properties of dispersions. Pharm Res. 2011;28(3):572–84.
poly (oxyethylene)–poly (oxypropylene) block Lecun Y, Bengio Y, Hinton G. Deep learning. Nature.
copolymers: implications for pharmaceutical dosage 2015;521(7553):436–44.
form design. J Appl Polym Sci. 2003;87(6):1016–26. Lee EH, Smith DT, Fanwick PE, Byrn
Joshi BV, Patil VB, Pokharkar VB. Compatibility studies SR. Characterization and anisotropic lattice expan-
between carbamazepine and tablet excipients using sion/contraction of polymorphs of tenofovir disoproxil
thermal and non-thermal methods. Drug Dev Ind fumarate. Crys Growth Des. 2010;10:2314–22.
Pharm. 2002;28(6):687–94. Lee HG, Zhang GGZ, et al. Cocrystal intrinsic dissolution
Joy AS. Methods and techniques for the determination of behavior using a rotating disk. J Pharm Sci. 2011;100
specific surface by gas adsorption. Vacuum. 1953;3: (5):1736–44.
254. Lenhardt T, Vergnault G, Grenier P, Scherer D, Langguth
Juenemann D, Jantratid E, et al. Biorelevant in vitro disso- P. Evaluation of nanosuspensions for absorption
lution testing of products containing micronized or enhancement of poorly soluble drugs: in vitro transport
nanosized fenofibrate with a view to predicting plasma studies across intestinal epithelial monolayers. AAPS
profiles. Eur J Pharm Biopharm. 2011;77(2):257–64. J. 2008;10(3):435–8.
Junghanns JAH, Mueller RH. Nanocrystal technology, Li S, Wong S, et al. Investigation of solubility and disso-
drug delivery and clinical applications. Int J Nanomed. lution of a free base and two different salt forms as a
2008;3(3):295–309. function of pH. Pharm Res. 2005;22(4):628–35.
Kapourani A, Valkanioti V, Kontogiannopoulos KN, Li M, Zhang H, Chen B, Wu Y, Guan L. Prediction of pKa
Barmpalexis P. Determination of the physical state of values for neutral and basic drugs based on hybrid
a drug in amorphous solid dispersions using artificial artificial intelligence methods. Sci Rep. 2018;8(1)
neural networks and ATR-FTIR spectroscopy. Int J Li K, Wang P, Tang L, Shi R, Su Y, Zhao J. Stability and
Pharm X. 2020;2:100064. NMR chemical shift of amorphous precursors of
methane hydrate: insights from dispersion-corrected calorimeter. J Therm Anal Calorim. 2005;82(1):
density functional theory calculations combined with 167–70.
machine learning. J Phys Chem B. 2021;125(1): McAllister M. Dynamic dissolution: a step closer to pre-
431–41. dictive dissolution testing? Mol Pharm. 2010;7(5):
Lin D, Huang Y. A thermal analysis method to predict the 1374–87.
complete phase diagram of drug–polymer solid McConville JT, Williams RO III, Carvalho TC, Iberg AN,
dispersions. Int J Pharm. 2010;399(1):109–15. Johnston KP, Talbert RL, Burgess D, Peters JI. Design
Liu F-h, Jiang Y. Room temperature ionic liquid as matrix and evaluation of a restraint-free small animal inhala-
medium for the determination of residual solvents in tion dosing chamber. Drug Dev Ind Pharm. 2005;1:35–
pharmaceuticals by static headspace gas chromatogra- 42.
phy. J Chromatogr A. 2007;1167(1):116–9. McConville JT, Overhoff KA, Sinswat P, Vaughn JM, Frei
Liu XB, Ye JX, Quan LH, Liu CY, Deng XL, Yang M, BL, Burgess DS, Talbert RL, Peters JI, Johnston KP,
Liao YH. Pulmonary delivery of scutellarin solution Williams RO III. Targeted high lung concentrations of
and mucoadhesive particles in rats. Eur J Pharm itraconazole using nebulized dispersions in a murine
Biopharm. 2008;70:845–52. model. Pharm Res. 2006;23(5):901–11.
Liu X, Feng X, Williams RO, Zhang F. Characterization of Meeus J, Scurr DJ, Chen X, Amssoms K, Davies MC,
amorphous solid dispersions. J Pharm Investig. 2017: Roberts CJ, et al. Combination of (M) DSC and surface
1–23. analysis to study the phase behaviour and drug distri-
Loftsson T, Brewster ME. Pharmaceutical applications of bution of ternary solid dispersions. Pharm Res.
cyclodextrins: basic science and product development. 2015;32(4):1407–16.
J Pharm Pharmacol. 2010;62:1607–21. Menegola J, Steppe M, et al. Dissolution test for
Louis B, Agrawal VK, Khadikar PV. Prediction of intrin- citalopram in tablets and comparison of in vitro disso-
sic solubility of generic drugs using MLR, ANN and lution profiles. Eur J Pharm Biopharm. 2007;67(2):
SVM analyses. Eur J Med Chem. 2010;45(9):4018–25. 524–30.
Lowell S, Shields JE. Powder surface area and porosity. Miller DA, DiNunzio JC, Yang W, McGinity JW,
powder technology series. B. Scarlett. New York: Williams RO III. Enhanced in vivo absorption of
Chapman & Hall; 1991. p. 155–73. itraconazole via stabilization of supersaturation
Lubach JW, Xu D, Segmuller BE, Munson followign acidic-to-neutral pH transition. Drug Dev
EJ. Investigation of the effects of pharmaceutical Ind Pharm. 2008;34:890–902.
processing upon solid-state NMR relaxation times Mora PC, Cirri M, et al. Differential scanning calorimetry
and implications to solid-state formulation stability. J as a screening technique in compatibility studies of
Pharm Sci. 2007;96(4):777–87. DHEA extended release formulations. J Pharm
Lusci A, Pollastri G, Baldi P. Deep architectures and deep Biomed Anal. 2006;42(1):3–10.
learning in chemoinformatics: the prediction of aque- Morello M, Krone CL, Dickerson S, Howerth E,
ous solubility for drug-like molecules. J Chem Inf Germishuizen WA, Wong YL, Edwards D, Bloom
Model. 2013;53(7):1563–75. BR, Hondalus MK. Dry-powder pulmonary insuffla-
Lust A, Strachan CJ, Veski P, Aaltonen J, Heinämäki J, tion in the moues for application to vaccine or drug
Yliruusi J, et al. Amorphous solid dispersions of studies. Tuberculosis. 2009;89:371–7.
piroxicam and Soluplus®: qualitative and quantitative Mosharraf M, Sebhatu T, Nystrom C. The effects of disor-
analysis of piroxicam recrystallization during storage. dered structure on the solubility and dissolution rates of
Int J Pharm. 2015;486(1–2):306–14. some hydrophilic, sparingly soluble drugs. Int J Pharm.
Marsac PJ, Shamblin SL, Taylor LS. Theoretical and prac- 1999;177:29–51.
tical approaches for prediction of drug–polymer Mulligan KJ, McCauley H. Factors that influence the
miscibility and solubility. Pharm Res. 2006;23(10): determination of residual solvents in pharmaceuticals
2417–26. by automated static headspace sampling coupled to
Marsac PJ, Li T, Taylor LS. Estimation of drug–polymer capillary GCMS. J Chromatogr Sci. 1995;33:49–54.
miscibility and solubility in amorphous solid Mura P, Bettinetti GP, et al. Differential scanning calorim-
dispersions using experimentally determined interac- etry in compatibility testing of picotamide with phar-
tion parameters. Pharm Res. 2009;26(1):139–51. maceutical excipients. Thermochim Acta. 1998a;321
Marsac PJ, Rumondor AC, Nivens DE, Kestur US, (1–2):59–65.
Stanciu L, Taylor LS. Effect of temperature and mois- Mura P, Faucci MT, et al. Compatibility study between
ture on the miscibility of amorphous dispersions of ibuproxam and pharmaceutical excipients using differ-
felodipine and poly (vinyl pyrrolidone). J Pharm Sci. ential scanning calorimetry, hot-stage microscopy and
2010;99(1):169–85. scanning electron microscopy. J Pharm Biomed Anal.
Martins MT, Paim CS, et al. Development of a dissolution 1998b;18(1–2):151–63.
test for lamotrigine in tablet form using an ultraviolet Newman A, Engers D, Bates S, Ivanisevic I, Kelly RC,
method. Braz J Pharm Sci. 2010;46:179–86. Zografi G. Characterization of amorphous API: poly-
Mashru RC, Sutariya VB, et al. Characterization of solid mer mixtures using X-ray powder diffraction. J Pharm
dispersions of rofecoxib using differential scanning Sci. 2008;97(11):4840–56.
100 X. Ma et al.
Nie H, Su Y, Zhang M, Song Y, Leone A, Taylor LS, et al. Pinon AC, Rossini AJ, Widdifield CM, Gajan D, Emsley
Solid-state spectroscopic investigation of molecular L. Polymorphs of theophylline characterized by DNP
interactions between clofazimine and hypromellose enhanced solid-state NMR. Mol Pharm. 2015;12(11):
phthalate in amorphous solid dispersions. Mol Pharm. 4146–53.
2016;13(11):3964–75. Punčochová K, Ewing AV, Gajdošová M, Pekárek T,
Nunes C, Mahendrasingam A, Suryanarayanan Beránek J, Kazarian SG, et al. The combined use of
R. Quantification of crystallinity in substantially amor- imaging approaches to assess drug release from multi-
phous materials by synchrotron X-ray powder diffrac- component solid dispersions. Pharm Res. 2017;34(5):
tometry. Pharm Res. 2005;22(11):1942–53. 990–1001.
O’Donnell KP, Williams RO III. Nanoparticulate systems Purohit HS, Taylor LS. Phase separation kinetics in amor-
for oral drug delivery to the colon. Int J Nanotechnol. phous solid dispersions upon exposure to water. Mol
2011;8(1/2):4–20. Pharm. 2015;12(5):1623–35.
Okazaki A, Mano T, Sugano K. Theoretical dissolution Qian F, Huang J, Zhu Q, Haddadin R, Gawel J, Garmise R,
model of poly-disperse drug particles in biorelevant Hussain M. Is a distinctive single Tg a reliable indicator
media. J Pharm Sci. 2008;97(5):1843–52. for the homogeneity of amorphous solid dispersion?
Ong TC, Mak-Jurkauskas ML, Walish JJ, Michaelis VK, Int J Pharm. 2010;395:232–5.
Corzilius B, Smith AA, Clausen AM, Cheetham JC, Radha S, Gutch PK, et al. Thermal analysis of interactions
Swager TM, Griffin RG. Solvent-free dynamic nuclear between an oxime and excipients in some binary
polarization of amorphous and crystalline ortho- mixtures by differential scanning calorimetry and
terphenyl. J Phys Chem B. 2013;117(10):3040–6. thermagravimetric analysis. J Pharm Res. 2010;3(3):
Overhoff KA, Moreno A, Miller DA, Johnston KP, 590–5.
Williams RO III. Solid dispersions of itraconazole Rogers TL, Nelsen AC, Hu J, Brown JN, Sarkari M,
and enteric polymers made by ultra-rapid freezing. Int Young TJ, Johnston KP, Williams RO III. A novel
J Pharm. 2007;336:122–32. particle engineering technology to enhance dissolution
Overhoff KA, McConville JT, Yang W, Johnston KP, of poorly water soluble drugs: spray-freezing into liq-
Peters JI, Williams RO III. Effect of stabilizer on the uid. Eur J Pharm Biopharm. 2002;54:271–80.
maximum degree and extent of supersaturation and Roy CJ, Hale M, Hartings JM, Pitt L, Duniho S. Impact of
oral absorption of tacrolimus made by ultra-rapid inhalation exposure modality and particle size on the
freezing. Pharm Res. 2008;25(1):167–75. respiratory deposition of ricin in BALB/c mice. Inhal
Paim CS, Martins MT, Malesuik MD, Steppe M. LC Toxicol. 2003;15(6):619–38.
determination of entacapone in tablets: in vitro disso- Rumondor AC, Ivanisevic I, Bates S, Alonzo DE, Taylor
lution studies. J Chromatogr Sci. 2010;48:755–9. LS. Evaluation of drug-polymer miscibility in amor-
Patton JS. Mechanisms of macromolecule absorption by phous solid dispersion systems. Pharm Res. 2009;26
the lungs. Adv Drug Deliv Rev. 1996;19:3–36. (11):2523–34.
Patton JS, Byron PR. Inhaling medicines: delivering drugs Rumondor AC, Wikström H, Van Eerdenbrugh B, Taylor
to the body through the lungs. Nat Rev. 2007;6:67–74. LS. Understanding the tendency of amorphous solid
Paudel A, Geppi M, Van den Mooter G. Structural and dispersions to undergo amorphous–amorphous phase
dynamic properties of amorphous solid dispersions: the separation in the presence of absorbed moisture. AAPS
role of solid-state nuclear magnetic resonance spectros- PharmSciTech. 2011;12(4):1209–19.
copy and relaxometry. J Pharm Sci. 2014;103(9): Salari A, Young RE. Application of attenuated total reflec-
2635–62. tance FTIR spectroscopy to the analysis of mixtures of
Pavon JLP, del Nogal Sanchez M, Pinto CG, Laespada pharmaceutical polymorphs. Int J Pharm. 1998;163:
MEF, Cordero BM. Use of mass spectrometry methods 157–66.
as a strategy for detection and determination of residual Sandell E. Industrial aspects of pharmaceutics. Stockholm:
solvents in pharmaceutical products. Anal Chem. Swedish Pharmaceutical; 1993.
2006;78:4901–8. Sarid D. Scanning force microscopy: with applications to
Pendharkar CM, Jhawar RJ, Rutledge JM, Hause W, Grim electric, magnetic, and atomic forces, Demand, vol.
WM, Harwood RJ. Influence of the specific surface 5. New York: Oxford University Press; 1994.
area of selected raw materials on the granulation pro- Sauer BB, Kampert WG, et al. Temperature modulated
cess using an instrumented mixer. Pharm Technol. DSC studies of melting and recrystallization in
1990;14(4):44–53. polymers exhibiting multiple endotherms. Polymer.
Phalen RF, Mannix RC, Drew RT. Inhalation exposure 2000;41(3):1099–108.
methodology. Environ Health Perspect. 1984;56:23– Schachter DM, Xiong J, Tirol GC. Solid state NMR per-
34. spective of drug–polymer solid solutions: a model
Pham TN, Watson SA, Edwards AJ, Chavda M, Clawson system based on poly (ethylene oxide). Int J Pharm.
JS, Strohmeier M, Vogt FG. Analysis of amorphous 2004;281(1):89–101.
solid dispersions using 2D solid-state NMR and 1H T Schawe JEK. Modulated temperature DSC measurements:
1 relaxation measurements. Mol Pharm. 2010;7(5): the influence of the experimental conditions.
1667–91. Thermochim Acta. 1996;271:127–40.
Schawe J. Analysis of melting processes using TOPEM®. comparison of solubilities of indomethacin and nifedi-
UserCom. 2007;25:13–7. pine in PVP, PVP/VA, and PVAc. J Pharm Sci.
Schick C. Differential scanning calorimetry (DSC) of 2010;99(9):4023–31.
semicrystalline polymers. Anal Bioanal Chem. Sung JC, Padilla DJ, Garcia-Contreras L, VerBerkmoes
2009;395(6):1589–611. JL, Durbin D, Peloquin CA, Elbert KJ, Hickey AJ,
Seedher N, Kanojia M. Co-solvent solubilization of some Edwards DA. Formulation and pharmacokinetics of
poorly-soluble antidiabetic drugs. Pharm Dev Technol. self-assembled rifampicin nanoparticle systems for
2009;14(2):185–92. pulmonary delivery. Pharm Res. 2009;26(8):1847–55.
Shaw LR, Irwin WJ, et al. The effect of selected water- Surendrakumar K, Martyn GP, Hodgers ECM, Jansen M,
soluble excipients on the dissolution of paracetamol Blair JA. Sustained release of insulin from sodium
and ibuprofen. Drug Dev Ind Pharm. 2005;31(6): hyaluronate based dry powder formulation after pul-
515–25. monary delivery in beagle dogs. J Control Release.
Shi Y, Gao P, et al. Application of a biphasic test for 2003;91:385–94.
characterization of in vitro drug release of immediate Susto GA, McLoone S. Slow release drug dissolution
release formulations of celecoxib and its relevance to profile prediction in pharmaceutical manufacturing: a
in vivo absorption. Mol Pharm. 2010;7(5):1458–65. multivariate and machine learning approach. 2015.
Sing KSW, Everett DH, Haul RAW, Moscou L, Pierotti Swinkels D, Bristow N, Gale L. Effect of sample prepara-
RA, Rouquerol J, Siemieniewska T. Reporting tion on the BET surface area of EMD. Belgium: Prog-
physisorption data for gas/solid systems with special ress in Batteries & Battery Materials; 1994.
reference to the determination of surface area and Syam P, Sundaramoorthy K, Vetrichel VT. Design of
porosity. Pure Appl Chem. 1985;57(4):603–19. biodegradable polymer nanoparticles for oral drug
Singh S, Singh B, et al. Stress degradation studies on delivery of stavudine: in vitro dissolution studies and
ezetimibe and development of a validated stability- characterization. Int J Pharm Technol. 2010;3(1):
indicating HPLC assay. J Pharm Biomed Anal. 1360–72.
2006;41(3):1037–40. Teijeiro SA, Briñón MC. 30 -azido-30 -deoxy-50 -o-isonicoti-
Sitar Curin A, Greman M, Vrecer F, Kotar-Jordan B, noylthymidine: a novel antiretroviral analog of zidovu-
Sustar B. Study of crystal modifications of dine. III. Solubility studies. Nucleosides Nucleotides
lansoprazole using FT-IR spectroscopy, solid-state Nucleic Acids. 2006;25(2):191–202.
NMR spectroscopy and FT-Raman spectroscopy. Tho I, Liepold B, et al. Formation of nano/micro-
Farm Vestn. 1997;48:290–1. dispersions with improved dissolution properties
Song Y, Yang X, Chen X, Nie H, Byrn S, Lubach upon dispersion of ritonavir melt extrudate in aqueous
JW. Investigation of Drug–excipient interactions in media. Eur J Pharm Sci. 2010;40(1):25–32.
lapatinib amorphous solid dispersions using solid- Tian Y, Booth J, Meehan E, Jones DS, Li S, Andrews
state NMR spectroscopy. Mol Pharm. 2015;12(3): GP. Construction of drug–polymer thermodynamic
857–66. phase diagrams using Flory–Huggins interaction the-
Stanisz B, Kania L. Validation of HPLC method for deter- ory: identifying the relevance of temperature and drug
mination of atorvastatin in tablets and for monitoring weight fraction to phase separation within solid
stability in solid phase. Acta Pol Pharm. 2006;63(6): dispersions. Mol Pharm. 2012;10(1):236–48.
471–6. Tiwari M, Chawla G, et al. Quantification of olanzapine
Stark G, Fawcett JP, Tucker IG, Weatherall polymorphs using powder X-ray diffraction technique.
IL. instrumental evaluation of color of solid dosage J Pharm Biomed Anal. 2007;43(3):865–72.
forms during stability testing. Int J Pharm. 1996;143: Tobyn M, Brown J, et al. Amorphous drug–PVP
93–100. dispersions: application of theoretical, thermal and
Štukelj J, Svanbäck S, Kristl J, Strachan CJ, Yliruusi spectroscopic analytical techniques to the study of a
J. Image-based investigation: biorelevant solubility of molecule with intermolecular bonds in both the crys-
α and γ indomethacin. Anal Chem. 2019a;91(6): talline and pure amorphous state. J Pharm Sci. 2009;98
3997–4003. (9):3456–68.
Štukelj J, Svanbäck S, Agopov M, Löbmann K, Strachan Tombari E, Ferrari C, Johari G, Shanker RM. Calorimetric
CJ, Rades T, Yliruusi J. Direct measurement of amor- relaxation in pharmaceutical molecular glasses and its
phous solubility. Anal Chem. 2019b;91(11):7411–7. utility in understanding their stability against crystalli-
Štukelj J, Agopov M, Yliruusi J, Strachan CJ, Svanbäck zation. J Phys Chem B. 2008;112(35):10806–14.
S. Image-based dissolution analysis for tracking the Tomoda K, Ohkoshi T, Hirota K, Sonavane GS,
surface stability of amorphous powders. ADMET Nakajima T, Terada H, Komuro M, Kitazato K,
DMPK. 2020a;8(4):401–9. Makino K. Preparation and properties of inhalable
Štukelj J, Agopov M, Yliruusi J, Strachan CJ, Svanbäck nanocomposite particles for treatment of lung cancer.
S. Machine-vision-enabled salt dissolution analysis. Colloids Surf B Biointerfaces. 2009;71:177–82.
Anal Chem. 2020b;92(14):9730–8. Tres F, Treacher K, Booth J, Hughes LP, Wren SA, Aylott
Sun YE, Tao J, Zhang GG, Yu L. Solubilities of crystalline JW, et al. Real time Raman imaging to understand
drugs in polymers: an improved analytical method and
102 X. Ma et al.
dissolution performance of amorphous solid Wolkers WF, Oldenhof H. In situ FTIR assessment of
dispersions. J Control Release. 2014;188:53–60. dried Lactobacillus bulgaricus: KBr disk formation
Vandecruys R, Peeters J, Verreck G, Brewster ME. Use of affects physical properties. Spectroscopy. 2005;19:
a screening method to determine excipients which 89–99.
optimize the extent and stability of supersaturated Wunderlich B. Thermal analysis of polymeric materials.
drug solutions and application of this system to solid Berlin: Springer; 2005.
formulation design. Int J Pharm. 2007;342:168–75. Xia D, Cui F, Piao H, Cun D, Piao H, Jiang Y, Ouyang M,
Venkatesh S, Li J, et al. Intrinsic solubility estimation and Quan P. Effect of crystal size on the in vitro dissolution
pH-solubility behavior of cosalane (NSC 658586), an and oral absorption of nitrendipine in rats. Pharm Res.
extremely hydrophobic diprotic acid. Pharm Res. 2010;27:1965–76.
1996;13(10):1453–9. Yalkowsky SH, Valvani SC. Solubility and partitioning I:
Vertzoni M, Dressman J, Butler J, Hempenstall J, Reppas solubility of nonelectrolytes in water. J Pharm Sci.
C. Simulation of fasting gastric conditions and its 1980;69(8):912–22.
importance for the in vivo dissolution of lipophilic Yang W, Tam J, Miller DA, Zhou J, McConville JT,
compounds. Eur J Pharm Biopharm. 2005;60:413–7. Johnston KP, Williams RO III. High bioavailability
Viegas TX, Curatella RU, Van Winkle LL, Brinker from nebulized itraconazole nanoparticle dispersions
G. Intrinsic drug dissolution testing using the station- with biocompatible stabilizers. Int J Pharm.
ary disk system. Dissolution Technol. 2001a;8(3): 2008;361:177–88.
19–23. Yang L, Wu X, et al. Novel biodegradable polylactide/
Viegas TX, Curatella RU, Van Winkle LL, Brinker poly(ethylene glycol) micelles prepared by direct dis-
G. Measurement of intrinsic drug dissolution rates solution method for controlled delivery of anticancer
using two types of apparatus. Pharm Technol. drugs. Pharm Res. 2009;26(10):2332–42.
2001b;25(6):44–53. Yang W, Johnston KP, Williams RO II. Comparison of
Vogt FG, Roberts-Skilton K, Kennedy-Gabb SA. A solid- bioavailability of amorphous versus crystalline
state NMR study of amorphous ezetimibe dispersions itraconazole nanoparticles via pulmonary administra-
in mesoporous silica. Pharm Res. 2013;30(9):2315–31. tion in rats. Eur J Pharm Biopharm. 2010;75:33–41.
Wang Y, Chow MSS, et al. Mechanistic analysis of Yang Y, Ye Z, Su Y, Zhao Q, Li X, Ouyang D. Deep
pH-dependent solubility and trans-membrane perme- learning for in vitro prediction of pharmaceutical
ability of amphoteric compounds: application to sil- formulations. Acta Pharm Sin B. 2019;9(1):177–85.
denafil. Int J Pharm. 2008;352(1–2):217–24. Ye Z, Yang Y, Li X, Cao D, Ouyang D. An integrated
Wang J, Davidovich M, et al. Solid-state interactions of a transfer learning and multitask learning approach for
drug substance and excipients and their impact on pharmacokinetic parameter prediction. Mol Pharm.
tablet dissolution: a thermal–mechanical facilitated 2018;16(2):533–41.
process-induced transformation or PIT. J Pharm Sci. Yu LX, Carlin AS, et al. Feasibility studies of utilizing
2010;99(9):3849–62. disk intrinsic dissolution rate to classify drugs. Int J
Waterman KC, Adami RC. Accelerated aging: prediction Pharm. 2004;270(1–2):221–7.
of chemical stability of pharmaceuticals. Int J Pharm. Yuan X, Sperger D, Munson EJ. Investigating miscibility
2005;293(1–2):101–25. and molecular mobility of nifedipine-PVP amorphous
Webb PA, Orr C. Analytical methods in fine particle solid dispersions using solid-state NMR spectroscopy.
technology. Norcross: Micromeritics Instrument Molr Pharm. 2013;11(1):329–37.
Corporation; 1997. Zakeri-Milani P, Barzegar-Jalali M, Azimi M, Valizadeh
Wempe MF, Buchanan CM, et al. Pharmacokinetics of H. Biopharmaceutical classification of drugs using
letrozole in male and female rats: influence of com- intrinsic dissolution rate (IDR) and rat intestinal per-
plexation with hydroxybutenyl-β-cyclodextrin. J meability. Eur J Pharm Biopharm. 2009;73:102–6.
Pharm Pharmacol. 2007;59(6):795–802. Zhao C, Furukawa S, Ohki Y. A novel collagenase-
Williams AC, Cooper VB, Thomas L, Griffith LJ, Petts assisted extraction of active pharmaceutical ingredients
CR, Booth SW. Evaluation of drug physical form from gelatin products for quantitative analysis by high
during granulation, tabletting and storage. Int J performance liquid chromatography. J Chromatogr
Pharm. 2004;275:29–39. A. 2009;1216:4524–8.
Witschi C, Doelker E. Residual solvents in pharmaceutical Zhao Y, Inbar P, Chokshi HP, Malick AW, Choi
products: acceptable limits, influences on physico- DS. Prediction of the thermal phase diagram of amor-
chemical properties, analytical methods and phous solid dispersions by Flory–Huggins theory. J
documented values. Eur J Pharm Biopharm. 1997;43 Pharm Sci. 2011;100(8):3196–207.
(3):215–42.
Solid-State Techniques for Improving
Solubility 3
Miguel O. Jara, Justin R. Hughey, Siyuan Huang,
Abstract Keywords
Poor aqueous solubility of a drug substance Solid-state chemistry · Crystal structure · Salt
can often be attributed to strong intermolecular formation · Polymorphs · Amorphous solids ·
forces within its crystal lattice, which, in turn, Cocrystals
prevent molecules from escaping in solution.
Through the use of solid-state chemistry, it is
possible to modify the crystal structure in such 3.1 Introduction
a way that mitigates intermolecular forces, thus
improving aqueous solubility and increasing The number of newly developed chemical entities
rates of dissolution. Solid-state techniques exhibiting poor water solubility has increased
utilized for solubility enhancement include the dramatically in recent years (Lipinski et al.
formation of salts, polymorphic or amorphous 2001; Lipinski 2002). In many cases, this intrinsic
forms, and cocrystals. Each technique has spe- property results in poor or erratic dissolution in
cific advantages and, in some cases, biological fluids and, consequently, poor bio-
disadvantages that may prevent its successful availability. Improving aqueous solubility of
use. The purpose of this chapter is to describe these compounds, even temporarily, can have a
each of the methods, allowing the reader to gain significant impact on in vivo performance.
an understanding of solid-state modifications Aqueous solubility of a drug substance is pri-
available for solubility enhancement. marily a function of its lipophilicity and intermo-
lecular forces within the crystal lattice (Jain and
Yalkowsky 2001; Jain et al. 2006). Therefore,
M. O. Jara · R. O. Williams III (*)
Molecular Pharmaceutics and Drug Delivery Division, solubility enhancement techniques typically
College of Pharmacy, The University of Texas at Austin, focus on addressing these two properties indepen-
Austin, TX, USA dently. In the case of high lipophilicity,
e-mail: [email protected]; bill.williams@austin. techniques such as solubilization in co-solvents,
utexas.edu
micelle formation, and complexation are often
J. R. Hughey employed (Loftsson and Brewster 1996;
ThermoFisher Scientific, High Point, NC, USA
e-mail: justin.hughey@thermofisher.com Dannenfelser et al. 2004; Torchilin 2007). Simi-
larly, when poor aqueous solubility is due to
S. Huang
Syntheic Molecule Design and Development, Lilly strong intermolecular forces within the crystal
Research Laboratories, Eli Lilly and Company, lattice, solid-state modification can be utilized.
Indianapolis, IN, USA Solid-state modification can be classified as
e-mail: [email protected]
104 M. O. Jara et al.
methods that modify supramolecular arrays of the may behave quite differently due to variable
same components by means of forming physical, chemical, and thermodynamic
polymorphs or amorphous solids and methods properties (Pudipeddi and Serajuddin 2005). Sub-
that change molecular components of the crystal stitution of a more suitable salt for a suboptimal
network by means of salt, cocrystal, or solvate salt in the drug development process can help
formation (Rodríguez-Spong et al. 2004). avoid numerous challenges. For example, Gross
This chapter discusses the concepts and et al. reported that the selected salt form of a
applications of solid-state technologies to developmental compound (NBI-75043) was a
improve the aqueous solubility, dissolution solvate, hygroscopic, and exhibited a low melting
rates, and thus bioavailability of poorly water- point (Gross et al. 2007). The researchers were
soluble drug substances. Specifically, these able to develop a more suitable salt through a
methods include salt formation, polymorphs, relatively simple screening process. Therefore, it
amorphous solids, and cocrystals. is critical to be as thorough as possible at the early
stages of development such that the optimal salt is
initially chosen. Sanphui et al. showed that the
3.2 Pharmaceutical Salts hydrochloride salt of voriconazole has stronger
ionic interactions, which increases the hardness
When an acid and base are combined, the anion of of the material from 366.9 2.8 to 870.0
the acid and the cation of the base react to form a 6.0 MPa and its elastic modulus from 3.79
salt. Therefore, a drug substance classified as a 0.17 to 19.41 0.13 GPa and improves its
weak acid or a weak base may be combined with mechanical properties for milling and tableting
a suitable base or acid, respectively, to form a (Sanphui et al., 2015).
pharmaceutical salt (Corrigan 2006). In doing In terms of salt formation, one of the most
so, physicochemical properties of the drug sub- important parameters of the drug substance is
stance, such as solubility, hygroscopicity, stabil- the pKa values of ionizable groups (Bastin et al.
ity, and processability, are manipulated and, in 2000). Once these values are known, potential
many cases, optimized (Berge et al. 1977). Often- salt-forming agents can be selected based on
times, the primary goal of pharmaceutical salt known pKa values of salt-forming agents, or
formation is to increase aqueous solubility and counterions (Gould 1986). In order to ensure suf-
thus bioavailability of the drug substance (Gould ficient proton transfer from the acidic to the basic
1986). The ability to circumvent undesirable species, the difference in pKa values between an
characteristics of the parent drug such as poor ionizable group and that of the counterion should
aqueous solubility without adversely affecting be greater than 3 units (Bowker and Stahl 2008).
its pharmacological activity makes the technique However, a difference of 2 units has been
particularly appealing. Interest in the formation of reported in the literature as being acceptable
pharmaceutical salts has increased significantly (Wells 1988; Tong and Whitesell 1998; Black
over the last 60 years, and the technique has et al. 2007).
become relatively common (Serajuddin 2007). Overall distributions of anions utilized to form
The decision to utilize pharmaceutical salts is pharmaceutical salts are illustrated in Fig. 3.1. It
one that should be made at the very early stages of is clear that chloride is the most common anion
pharmaceutical drug development. Historically, utilized to form salts of weakly basic drugs. This
salts were often chosen based on their ease of is primarily due to the low pKa value of
preparation and raw material costs without taking hydrochloric acid and its ability to readily form
into consideration important physicochemical salts with weak bases. However, loss of volatile
properties such as physical and chemical stability, hydrochloric acid is a stability concern with very
hygroscopicity, and pH-dependent dissolution weak bases (Lee and Hoff 2002). Other strong
rates (Serajuddin 2007). Salts of a single-drug acid counterions such as mesylate and sulfate
substance, prepared from different counterions, readily form salts due to low pKa values, with
3 Solid-State Techniques for Improving Solubility 105
Acetate (17)
Besylate (4)
Bromide (24)
Chloride (279)
Citrate (14)
Fumarate (9)
Gluconate (2) Chloride
Bromide
Iodide (5)
Isethionate (2) Besylate
Lactate (7)
Acetate
Malate (2)
Maleate (22) only used
Mesylate (22) once
Methylsulfate (2) Tosylate
Tartrate
Napsylate (2)
Nitrate (9)
Pamoate (4) Sulfate
Phosphate (14)
Citrate
Succinate (6) Succinate
Fumarate
Sulfate (39) Phosphate
Gluconate
Tartrate (20) Pamoate
Iodide
Tosylate (2) Nitrate
Isethionate Malate
Only used once (16) Mesylate Napsylate
Lactate Maleate Methylsulfate
Fig. 3.1 Overall distribution of anions used to form salts of weak bases in the Orange Book. (From Paulekuhn et al.
(2007))
Silver Procaine
Benzathine (1) Potassium
Calcium (12) Piperazine
Cholinate (1) Meglumine
Diethanolamine (1) Magnesium
Lysine
Diethylamine (1) Diethylamine
Lysine (1) Diethanolamine
Magnesium (2) Cholinate
Meglumine (5) Calcium
Piperazine (1)
Potassium (11) Benzathine
Procaine (1) Zinc
Sodium
Silver (1) Tromethamine
Sodium (131)
Tromethamine (3)
Zinc (2)
Fig. 3.2 Overall distribution of cations used to form salts of weak acids in the Orange Book. (From Paulekuhn et al.
(2007))
mesylate salt forms becoming increasingly com- acids is shown in Fig. 3.2. Sodium is by far the
mon (Elder et al. 2010). The overall distribution most commonly used cation. However, sodium
of cations utilized for forming salts with weak salts often have a tendency to form hydrates and
exhibit hygroscopicity, so control of water con- solids generation, solids detection, and sample
tent during their synthesis is important (Lee and analysis (Morissette et al. 2004).
Hoff 2002). Most commonly, the sample preparation step
consists of transferring a known amount of drug
substance into multi-well plates followed by the
3.2.1 Pharmaceutical Salt Selection addition of selected counterions and crystallizing
solvent (Kumar et al. 2007). Various methods
Unfortunately, there are no known methods for have been developed to prompt crystallization,
predicting the influence of a particular counterion including evaporation of the solvent, thermal
or class of counterions on the behavior of a drug cooling, and anti-solvent addition (Morissette
substance (O’Connor and Corrigan 2001). While et al. 2004). However, the precipitation process
qualitative rules of thumb exist, they are often- can be slow and, depending on the mode of crys-
times found to be unreliable. It is therefore a tallization, may take days to occur, as shown in
requirement to empirically determine the salt Fig. 3.3. The combinatorial approach allows one
that exhibits essential and desirable to evaluate not only a large number of
characteristics. In previous years, salt screening compositions but also crystallization conditions.
involved manually mixing the ionizable drug sub- Sample analysis of the resulting solid is most
stance with selected counterions to form salt commonly conducted using a tiered approach
precipitates which were then fully characterized such that essential criteria are met prior to under-
by a variety of techniques (Kumar et al. 2007). taking more time-consuming analyses, as shown
While effective, this method provides a limited in Fig. 3.4 (Kumar et al. 2007). Solids formed in
amount of data due to time and material the wells are normally first characterized by
constraints. Raman spectroscopy and X-ray powder diffrac-
A number of automated high-throughput tion. Samples identified as being feasible salts are
techniques have been described in the literature then analyzed by techniques to determine if
for the screening of pharmaceutical salts which essential criteria are met. These include hygro-
limit the amount of drug required while providing scopicity, solubility, crystal shape, particle size
a significant amount of data (Morris et al. 1994; distribution, residual solvent, and stability
Bastin et al. 2000; Morissette et al. 2004; Ware analyses.
and Lu 2004; Kumar et al. 2007). Such high- After preliminary identification of viable salts
throughput techniques generally comprise four through screening, production of larger quantities
major functional elements: sample preparation, may be carried out to evaluate the suitability of
Fig. 3.3 Typical rate of 12

appearance of solids during
a thermally driven high-
10
throughput crystallization
experiment. (From Gardner
Wells Precipitated (%)
et al. (2004)) 8
0 2 4 6 8 10
Incubation Time (Days)
Salt synthesis and Data analysis and

Solubility platform interpretation
crystallization
High throughout stage

Preterably crystalline to Solubility dictated by desired Comparative analytical profiling
enhance stability pharmaceutical profile
Informatics platform
Hygroscopicity Selected salt

Design of Stability platform
platform form(s)
experiment
Preterably low hygroscopicity to High chemical and solid-

reduce variability state stability
Scale up stage
Process control and Comprehensive
Optimal salt form
feasibility analysis profiling
Forwarded for Investigational No polymorphic variability, ease of synthesis, handling and Biopharmaceutical
New Drug (IND) application formulation development characterization
Fig. 3.4 Stages of high-throughput salt selection showing criteria, and the cuboidal box shows desirable criteria of
an interface with the scale-up stage of salt screening. salt selection with the preferable parameters depicted in
Stages depicted with a double-lined box indicate essential the underlying shaded areas. (From Kumar et al. (2007))
the synthesis process and ultimately the prepared researchers point out that the solvent can affect
pharmaceutical salt. Care should be taken to the ΔpKa value. This is illustrated in Fig. 3.5b,
ensure that polymorphic forms are not formed where the solvent was changed to methanol
during the scale-up process. which resulted in a dramatic shift in the pKa of
acetic acid. In this solvent system, there is no pH
range in which both ion species exist which
3.2.2 Solubility Enhancement suggests that salt formation would not occur.
Results of ephedrine crystallization showed
The primary purpose of forming a pharmaceutical that strong acids (pKa < 2) gave salts from water
salt, in most cases, is to enhance the solubility of and methanol, as one would expect. Weak acids
the compound and thus bioavailability. In a recent (pKa > 3) gave salts from water, but not from
study, salts of ephedrine from a variety of methanol which was due to the upward shift in
counterions were prepared (Black et al. 2007). acid pKa. Properties of the ephedrine salts formed
This study highlighted the fact that one of the are detailed in Table 3.1. The researchers
most important molecular properties for the demonstrated that the solubility of ephedrine
design of a salt screen is the acid and base disso- could be enhanced by as much as 16-fold by
ciation constants (pKa). The authors illustrated forming pharmaceutical salts. Further inspection
this concept by showing speciation diagrams for of Table 3.1 demonstrates the wide range of
ephedrine and acetic acid in water, as illustrated solubilities that can be obtained by various salts.
in Fig. 3.5a. In the pH range of 6–8, both species Bastin et al. utilized a high-throughput screen-
are predominantly present as ions, which favors ing technique in an effort to identify a suitable salt
salt formation (ΔpKa 5). However, the for a weak base (RPR 200765) that exhibited poor
Fig. 3.5 Speciation a 1

diagrams for ephedrine and
acetic acid in (a) water–
ephedrine pKa ¼ 9.74, 0.8
acetic acid pKa ¼ 4.76, and
Molar Fraction (%)

(b) methanol–ephedrine
pKa ¼ 8.74, acetic acid 0.6
pKa ¼ 9.71. (Adapted from [HA]
Black et al. (2007)) [A]
0.4 [HB]
[B+]
0.2
0
0 2 4 6 8 10 12 14
pH
b 1
0.8
Molar Fraction (%)
0.6
[HA]
[A]
0.4 [HB]
[B+]
0.2
0
0 2 4 6 8 10 12 14
pH
aqueous solubility (10 μg/mL) (Bastin et al. haloperidol, a weak base (pKa 8.0) with an
2000). The researchers demonstrated the success- intrinsic solubility of 2.5 μg/mL, and two of its
ful formation stable salts with hydrochloride, salts (hydrochloride and mesylate) were
hydrobromide, methanesulfonate (mesylate), and evaluated (Li et al. 2005). While the pH solubil-
camphorsulfonate counterions. Results of this ity profiles of the free base and its hydrochloride
study, summarized in Table 3.2, demonstrated salt were very similar, the mesylate salt
that the mesylate salt provided a significant exhibited much greater solubility in the pH
improvement in solubility over the free base. range of 2–5, as shown in Table 3.3. Maximum
While the mesylate salt form was found to be a solubility values for the hydrochloride and
monohydrate, it was non-hygroscopic and lost mesylate salts were found to be 4.2 mg/mL and
moisture only at very low humidity (<10% rela- 30.4 mg/mL, respectively. The researchers also
tive humidity). Furthermore, the mesylate salt showed that the dissolution rate of the mesylate
exhibited good flowability. Based on the results salt was much higher than that of the hydrochlo-
of the study, the mesylate salt was found to be ride salt or free base, except at very low pH
suitable for a solid dose formulation. values (pH < 2).
Li et al. conducted a comprehensive study in Engel et al. conducted a study to determine the
which the solubility and dissolution rates of most suitable salt for a weak base (LY333531)
Table 3.1 The measured and calculated physical properties of ephedrine salts
Salt Molecular weight (g/mol) Ephedrine solubility (mol/L)
Free base 165.23 0.345
Hemihydrate 174.24 0.643
Acetate 225.28 3.751
Adipate 311.37 5.299
Maleate monohydrate 299.32 3.792
Malonate (2:1) 434.52 3.392
Glycolate 241.28 1.828
L-malate 299.32 1.774
L-tartrate monohydrate 333.33 1.761
L-tartrate trihydrate (2:1) 534.68 3.146
HCl 201.69 1.601
Nitrate 228.25 5.77
DHP 263.22 1.417
Bisulfate 263.31 0.896
Besylate 323.4 0.26
Edisylate (2:1) 520.65 1.152
Esylate 275.36 1.493
Mesylate 261.33 2.246
Tosylate 337.42 0.323
Adapted from Black et al. (2007)
Table 3.2 Comparison of the physicochemical properties of RPR200765 salt forms

Salt Molecular weight (g/mol) Solubility at 25 C (mg/mL)
Free base NA 0.01
Mesylate 566.61 39.00
Camphorsulfonate 684.79 19.95
Hydrochloride 524.98 16.68
Hydrobromide 569.43 3.29
Adapted from Bastin et al. (2000)
which exhibited very low aqueous solubility concentrations (Cmax) of LY333531 and its
(1 μg/mL) (Engel et al. 2000). Without utilizing active metabolite, LY338522, over that of the
a high-throughput screening technique, the hydrochloride salt. The plasma concentration
researchers evaluated seven counterions for salt versus time plot for this study is shown in
formation: hydrochloride, sulfate, mesylate, suc- Fig. 3.6.
cinate, tartrate, acetate, and phosphate. Of the The preparation of pharmaceutical salts is an
salts prepared, only the mesylate and hydrochlo- effective method to enhance solubility and bio-
ride salts exhibited the essential criteria and were availability of drug substances exhibiting poor
thus analyzed further. Aqueous solubility of the aqueous solubility. A large number of
mesylate and hydrochloride salts was found to counterions are available for this purpose and
be 0.5 and 0.1 mg/mL (as LY333531), respec- are capable of producing salts with significantly
tively, which was a substantial improvement different physical properties. Thus, as previ-
over the free base. An in vivo study in male ously discussed, selection of the optimal salt
beagle dogs demonstrated that the mesylate salt form early in the development process is
exhibited a 250% increase in plasma critical.
Table 3.3 Classification of common pharmaceutical salts

pH of dissolution medium Dissolution rate (mg/min) Solubility at bulk pH (mg/mL)
Free base
1.1 0.032 0.79
2.0 0.246 3.41
3.1 0.061 4.19
5.0 0.002 2.47
HCl salt
1.1 0.025 0.79
1.5 0.062 2.50
2.0 0.155 3.41
3.1 0.292 4.16
5.0 0.291 2.47
7.0 0.157 0.02
Mesylate salt
1.1 0.033 0.65
1.7 0.115 20.76
2.0 0.865 25.06
3.1 2.037 28.45
5.0 1.962 30.44
7.0 1.760 0.002
Adapted from Li et al. (2005)
Fig. 3.6 Mean plasma 3500

concentrations of
LY333531 and LY 338522 3000
in male beagle dogs orally LY33353 (HCI)
administered with 2500 LY33353 (Mesylate)
ng/mL Plasma
LY333531 HCl and LY338522 (HCI)

LY333531 mesylate (20 mg 2000 LY338522 (Mesylate)
LY333531/kg). (From
Engel et al. (2000)) 1500
1000
500
0
0 8 16 24 32 40 48 56 64 72 80 88 96
Hours
solubility of drug substances with poor aqueous

3.3 Polymorphs
solubility include crystalline polymorphs,
and Amorphous Forms
solvates, desolvates, and amorphous forms.
Polymorphs have been studied extensively
The ability of metastable forms of drug
since the early reports of their existence (Aguiar
substances to enhance solubility over their ther-
et al. 1967; Aguiar and Zelmer 1969; Haleblian
modynamically stable counterparts has been well
and McCrone 1969). These substances are
documented in a number of reviews (Huang and
defined as having identical chemical
Tong 2004; Mao et al. 2005; Pudipeddi and
compositions but differ in internal structure,
Serajuddin 2005). Metastable systems utilized
including unit cell dimensions and crystal
for the purpose of increasing the apparent
packing which can affect pharmaceutical perfor- hydrochloride, rosuvastatin calcium, and
mance (Byrn et al. 1999b). These effects include zafirlukast (Wyttenbach and Kuentz 2017).
solubility, rates of dissolution, bioavailability,
processability, physical stability, and chemical
stability. A large number of drug substances, 3.3.1 Polymorph Preparation
whether they are neutral, free acids, free bases,
or pharmaceutical salts, are capable of exhibiting Polymorphic forms present a number of
polymorphism (Bowker and Stahl 2008). While challenges to the pharmaceutical scientist.
some compounds do not exhibit polymorphism, Polymorphs can be discovered in later stages of
other may have multiple polymorphs. For exam- drug development or, in a worst-case scenario,
ple, progesterone is known to exist as five poly- after the drug product is marketed. Both situations
morphic forms, all of which have different can have a profound impact in terms of time and
physical and chemical properties (Byrn et al. cost, meaning that a comprehensive search for
1999b). The polymorphic form exhibiting the polymorphic forms is justified (Yu et al. 1998).
highest density and melting point is considered In a classic example of polymorphism, ritonavir,
the most thermodynamically stable form. an HIV protease inhibitor, was found to exhibit a
Pseudopolymorphs, also known as hydrates or more thermodynamically stable polymorph
solvates, contain molecules from the crystalliza- 2 years after its market entry (Bauer et al. 2001).
tion solvent within the crystal lattice (Bechtloff This previously unknown polymorph was found
et al. 2001). These solvent molecules are often to be approximately 50% less soluble in the for-
located in discrete crystal sites and bound within mulation vehicle, resulting in a substantial
the lattice (e.g., hydrogen bonding) (Yu et al. decrease in dissolution rate and ultimately with-
1998). It is possible to have both stoichiometric drawal of the drug product from the market. A
and nonstoichiometric forms of these substances, new formulation incorporating the more stable
with the stoichiometric form being most common. polymorph was developed and marketed
Solvates, which contain a solvent other than (Chemburkar et al. 2000). In another example, a
water, are mostly nonpractical from a pharmaceu- polymorph of carbamazepine was reported in
tical standpoint due to potential toxicity. While 2002, two decades after identification of its
hydrates are normally more stable than their other three polymorphic forms (Lang et al. 2002).
anhydrous counterparts, they rarely provide any The process of discovering polymorphs
solubility advantage and, in most cases, exhibit a requires extensive experimentation that is nor-
lower solubility. However, hydrates and solvates mally accomplished by high-throughput screen-
may be desolvated into relatively thermodynami- ing techniques very similar to those used in
cally unstable forms which can provide solubility pharmaceutical salt screening. Variables such as
enhancement (Yu et al. 1998). supersaturation, agitation rate, cooling rate, sol-
Amorphous forms of drug substances are yet vent composition, temperature, seed crystals,
another physical modification of the solid state additives, and impurities are all known to affect
which provide the greatest solubility potential. crystallization and should be investigated (Allesø
They differ from their crystalline polymorph et al. 2008; Alvarez et al. 2009). By evaluating as
counterparts by having low packing efficiency many variables as possible, the likelihood of
and lack of long-range crystalline order identifying a large majority or all of the
(Yu 2001). This results in a very high free energy polymorphs is high. Morissette et al. (2003)
and, thus, the capability to provide a substantial utilized a high-throughput screening study to
improvement in solubility (Hancock and Zografi evaluate polymorphic forms of ritonavir.
1997). Amorphous materials are also known as Utilizing less than 2 g of drug substance, the
“glasses” in the literature (Alhalaweh et al. 2014). researchers conducted 2,000 experiments. In
Examples of these formulations in the market are addition to the two known polymorphs, three
cefuroxime axetil, nelfinavir mesylate, quinapril new forms were discovered: a metastable
Table 3.4 Methods of manufacturing molecularly disordered (amorphous) pharmaceutical materials

From Method Examples
Crystal Disruption/energy input Milling or grinding
Compression or decompression
Dehydration or desolvation
Irradiation
Reaction
Solution Solvent removal Spray-drying
Lyophilization (bulk or coupled with thin-film freezing
Precipitation
Polymerization
Reaction
Liquid Cooling or energy removal Melt quenching
Nucleation suppression
Polymerization
Reaction
Vapor Cooling or energy removal Sublimation
Reaction
Adapted from Hancock (2002)
polymorph, a crystalline solvate, and a disruption, most techniques begin with

nonstoichiometric hydrate. This example solubilizing or melting the drug substance
demonstrates the clear advantage of performing followed by solvent removal or quenching,
extensive studies early in drug development. respectively, at rates that kinetically avoid recrys-
However, even thorough solution-based studies tallization (Hancock 2002). The result is a meta-
may not be sufficient. Peterson et al. (2002) stable substance with no long-range crystalline
conducted a high-throughput polymorphism order (Table 3.4).
screening study on acetaminophen in an attempt
to generate the three known forms (forms I, II,
Glass-Forming Ability
and III). Of 7,776 crystallization trials, only
Not every compound is a good candidate for
723 resulted in precipitates, and of those,
amorphization because it depends on its intrinsic
29 were form II, with the remainder being form
tendency to recrystallize. The concept of glass-
I. Melt crystallization was required to generate
forming ability (GFA) refers to how favorable the
form III, demonstrating the need to consider alter-
vitrification of a melted material is on cooling
native crystallization techniques. While compre-
(Baird et al. 2010). It classifies how challenging
hensive studies cannot guarantee that all
the generation of an amorphous drug (neat drug)
polymorphic forms will be identified, they cer-
from the crystalline form. This same concept can
tainly provide a level of assurance that additional
be used to predict the recrystallization tendency in
forms will not be easily formed.
solid state and even in solution (Van Eerdenbrugh
et al. 2014).
Usually, compounds with molecular weights
3.3.2 Amorphous Form Preparation higher than 300 g/mol are good glass formers, and
those with molecular weights under 200 g/mol
Amorphous solids are generally more easily tend to be poor glass formers. Between those
prepared than polymorphic forms which can ranges, molecular descriptors such as the presence
require specific crystallization conditions. of aromatic rings and amides had great relevance.
Techniques that may be utilized for the prepara- These structures increase the intermolecular inter-
tion of amorphous solids are outlined in action in the crystal lattice, enhancing the ten-
Table 3.4. With the exception of crystal dency to recrystallize (Alhalaweh et al. 2014).
Determination of Glass-Forming Ability

The method proposed by Baird et al. (2010) is
straightforward and robust to determine the GFA
by using a heating/cooling/heating cycle utilizing
differential scanning calorimetry (Fig. 3.7) (Baird
et al. 2010). The sample is heated at 10 C/min to
10 C above its melting point, holding the temper-
ature for 3 min. Thereafter, the sample is cooled at
20 C/min to -75 C. Then, the sample is reheated
again until 10 C above its melting point at 10 C/
min. The classes are assigned depending on the
shape of the thermogram. If recrystallization is
observed in the cooling cycle, the drug is classi-
fied as a poor glass former (GFA Class I). Modest
glass formers (GFA Class II) are those drugs that
stay amorphous in the cooling cycle and then
recrystallize in the second heating. Finally, the
good glass formers (GFA Class III) are those
that do not recrystallize after the cooling and
reheating step, maintaining their amorphous
state during the entire experiment.
Other authors have further expanded on the
GFA classification and have proposed
improvements. The recrystallization of some
drugs depends on the employed temperature
ramp, especially the cooling step. Blaabjerg
et al. (2016) proposed another classification
based on the rates employed during cooling
(Blaabjerg et al. 2016). According to them,
Class I drugs are those that even at a cooling
rate of 750 C/min cannot avoid recrystallization.
Class II are those drugs that require “modest”
cooling rates between 10 and 20 C/min to
avoid recrystallization and obtain an amorphous
material. Finally, Class III comprises those drugs
that can maintain their amorphous state with
cooling rates close to 1 C/min.
The Importance of Glass-Forming Ability

Fig. 3.7 Differential scanning calorimetry thermograms
Van Eerdenbrugh et al. developed a classification
showing the crystallization behavior of the three classes.
for the recrystallization of amorphous drugs in (a) Class I crystallization upon cooling. (b) Class II crys-
aqueous environments by using the solvent shift tallization in the second heat. (c) In Class III, no crystalli-
method and synchrotron-based experiments. In zation was observed. (From Alhalaweh et al. 2014;
reprinted with permission from Amjad Alhalaweh,
this classification, Classes I, II, and III are fast,
Ahmad Alzghoul, Waseem Kaialy, Denny Mahlin, and
intermediate, and slow crystallizers, respectively. Christel A. S. Bergström. Molecular Pharmaceutics 2014
Class I compounds crystallize in less than 150 s 11 (9), 3123–3132. Copyright (2014) American Chemical
(rapid), Class II between 150 s and before 1 h Society. (https://pubs.acs.org/doi/10.1021/mp500303a).
Further permissions related to the material excerpted
(intermediate), and Class III after 1 h (slow).
should be directed to the ACS)
Notably, this proposed classification of crystalli- ΔGIII ¼ ΔH III TΔSIII , ð3:3Þ

zation is frequently in agreement with the GFA
classification (Van Eerdenbrugh et al. 2014). It where ΔH is a measure of crystal lattice energy
can be concluded that knowing the GFA of a drug differences and ΔS is the difference in disorder
provides an idea of its behavior in solution. This and lattice vibrations between the two forms
classification is critical to estimate if an amor- (Rodríguez-Spong et al. 2004). Utilizing the rela-
phous compound can increase bioavailability tionship in (3.3), it is possible to evaluate relative
when exposed to the gastrointestinal tract, due to differences in Gibbs free energy values.
the time needed to disperse, dissolve, and being A diagram showing Gibbs free energy versus
absorbed without recrystallizing. Attempts to pre- temperature provides complete and quantitative
pare an amorphous drug using a Class I com- information about the relative stability of various
pound should be avoided without the use of a metastable forms and can be obtained through
stabilizing additive because of concerns related thermal and solution-based techniques
to solid-state stability and in vivo performance (Yu 1995; Yu et al. 1998). The schematic in
(Wyttenbach and Kuentz 2017). Fig. 3.8 demonstrates Gibbs free energy as a
Alhalaweh et al. stored GFA Class II and III function of temperature for a hypothetical
compounds for 12 h at 20 K below and above single-component system consisting of multiple
their Tg. Class III compounds remained amor- metastable states. It is clear that form C exhibits
phous in both conditions. On the other hand, the lowest free energy at all temperatures and is
most of the Class II compounds underwent thus the most thermodynamically stable form.
recrystallization at 20 K above their Tg Similarly, the amorphous form exhibits a Gibbs
(Alhalaweh et al. 2015). This further free energy that is higher than the crystalline
demonstrates the relevance of GFA as a predictor polymorph states, which can be attributed to a
for solid-state stability and storage conditions. higher enthalpy and entropy, as shown in (3.3)
(Hancock and Zografi 1997). The amorphous
form lacks a crystal lattice and thus does not
exhibit a melting point, but rather a glass transi-
3.3.3 Thermodynamics of Metastable
tion temperature (Tg). From a kinetics perspec-
Solids
tive, an amorphous material may exist in a
number of states with different properties (e.g.,
Metastable solids exhibit excess enthalpy,
Tg, relaxation time, etc.) that are dependent on its
entropy, and, thus, free energy when compared
mode of preparation (Shalaev and Zografi 2002).
to the most thermodynamically stable form
Further inspection of Fig. 3.8 demonstrates the
(Yu 2001). Since thermodynamic stability of a
behavior of two different types of polymorph
solid is a function of both enthalpy (H ) and
systems: monotropic and enantiotropic. Forms A
entropy (S) at constant temperature (T ) and pres-
and C exhibit monotropic polymorphism in that
sure (P), it is important to evaluate the Gibbs free
form C is more stable than form A at all
energy, G, for each system studied (Grant and
temperatures below their melting points. Con-
Higuchi 1990). For form I of a single-component
versely, forms A and B exhibit enantiotropic
system,
behavior in that there is a transition temperature,
GI ¼ H I TSI , ð3:1Þ below that of the melting point, in which the
thermodynamic stability is reversed.
and for form II, Due to their high free energy, metastable
GII ¼ H II TSII : ð3:2Þ solids have the potential to convert to a more
metastable polymorph or to the most thermody-
Subtraction of (3.1) from (3.2), we can obtain namically stable state through nucleation and
the difference in Gibbs free energy between the growth. The conversion process is kinetic in
forms: nature, and if the rate is slow relative to the
Fig. 3.8 Schematic Gibbs

free energy curves for a
hypothetical single-
component system that
exhibits crystalline and
amorphous phase
transitions. Abbreviations
indicate glass transition
temperatures, transition
temperatures, and melting
temperatures for each of the
systems described. (From
Rodríguez-Spong et al.
(2004))
pharmaceutical relevant time scale, the metasta- 1934). This is accomplished by recognizing that
ble form may be utilized in a drug product free energy is related to the activity of a com-
(Dannenfelser et al. 2004). Hancock et al. pound through the following definitions:
evaluated the physical stability of amorphous
GI ¼ RT ln αI , ð3:4Þ
solids and through thermal relaxation studies
determined that if the glass transition temperature GII ¼ RT ln αII , ð3:5Þ
of the material is at least 50 C higher than the
storage temperature, crystallization due to molec- where αI and αII are activities of the respective
ular mobility is not a concern (kinetically stable) forms. Activities are a reflection of “escaping
over the pharmaceutical time scale (Hancock tendencies” and are thus proportional to solubil-
et al. 1995). Adherence to this may be subject to ity, σ (Gupta et al. 2004). By substituting the
the glass-forming ability (GFA) classification of solubility terms and subtracting (3.4) from (3.5),
the drug. It should be noted that chemical stability we can obtain
of metastable forms is a concern, primarily due to
σ
decreased molecular packing, potential for mois- ΔGIII ¼ RT ln II : ð3:6Þ
ture absorption, and increased molecular mobility σI
(Xiang and Anderson 2004). The most thermody-
Hence, the solubility ratio between two forms
namically stable polymorphic form is often cho-
is shown to be proportional to Gibbs free energy.
sen for development, which, in most cases, is the
For a single-component system, as shown in
least-soluble form (Singhal and Curatolo 2004).
Fig. 3.8, the amorphous form exhibits the highest
However, in some cases the advantages of
free energy and thus is the most soluble form.
enhanced solubility, dissolution rates, and bio-
However, the inherent thermodynamic high-
availability outweigh potential disadvantages,
energy unstable state leading to relaxation, nucle-
and a metastable form is developed.
ation, and crystallization during storage or transit
In addition to relative stability, Gibbs free
through the gastrointestinal tract is one of the
energy values can be utilized to estimate relative
most important challenges associated with the
solubility between solid forms (Parks et al. 1928,
application of the amorphous form (Fig. 3.9).
Fig. 3.9 Schematic

representation of the
energetics associated with
nucleation and crystal
growth from the amorphous
state. (From Zografi and
Newman, 2017)
Table 3.5 Predicted solubility ratios for drug compounds

Compound Forms Solubility ratioa Comment
Indomethacin α-crystal/γ-crystal 1.1–1.2 45 C
amorphous/γ-crystal 38–301 5 C
25–104 25 C
16–41 45 C
Carbamazepine III-crystal/I-crystal 1.7–2.1 2 C
1.7–2.0 12 C
1.6–2.0 17 C
1.6–1.9 26 C
1.6–1.8 40 C
1.5–1.7 58 C
Chloramphenicol palmitate A-crystal/B-crystal 3.6 30 C
Iopanoic acid II-crystal/I-crystal 2.3–2.8 37 C
Mefenamic acid I-crystal/II-crystal 1.5 30 C
Glibenclamide Amorphous/crystal 112–1652 23 C
Glucose Amorphous/crystal 16–53 20 C
Griseofulvin Amorphous/crystal 38–441 21 C
Hydrochlorothiazide Amorphous/crystal 21–113 37 C
Iopanoic acid Amorphous/I-crystal 12–19 37 C
Polythiazide Amorphous/crystal 48–455 37 C
Adapted from Hancock and Parks (2000)
The range of values reflects the use of different ΔCp values for the calculations
a
3.3.4 Solubility and Bioavailability obtained from the pharmaceutical literature for
Enhancement all drugs, with the exception of indomethacin
which was determined experimentally.
Hancock et al. utilized the relationship detailed in Polymorphs of indomethacin were prepared by
(3.6) to predict the relative solubility of 11 differ- precipitation from a saturated methanol solution
ent drugs with polymorphic or amorphous forms with water, while amorphous material was
(Hancock and Parks 2000). Through thermal- prepared by melt quenching with liquid nitrogen.
based techniques, the researchers estimated ΔGIII Predicted solubility ratios for each of the
at constant temperature and were able to solve for compounds and their respective metastable
σ II forms are summarized in Table 3.5. The
the activity (solubility) ratio, σ I . Data were
Table 3.6 Experimental solubility ratios for drug compounds

Compound Forms Solubility ratioa Comment
Indomethacin α-Crystal/γ-crystal 1.1–1.2 45 C, water
Amorphous/γ-crystal 38–301 5 C, water
25–104 25 C, water
16–41 45 C, water
Carbamazepine III-crystal/I-crystal 1.7–2.1 2 C, 2-propanol
1.7–2.0 12 C, 2-propanol
1.6–2.0 17 C, 2-propanol
1.6–1.9 26 C, 2-propanol
1.6–1.8 40 C, 2-propanol
1.5–1.7 58 C, 2-propanol
Chloramphenicol palmitate A-crystal/B-crystal 3.6 30 C, 35% t-butanol (aq.)
Iopanoic acid II-crystal/I-crystal 2.3–2.8 37 C, phosphate buffer (aq.)
Mefenamic acid I-crystal/II-crystal 1.5 30 C, dodecyl alcohol
Glibenclamide Amorphous/crystal 112–1652 23 C, buffer (aq.)
Glucose Amorphous/crystal 16–53 20 C, methanol
20 C, ethanol
20 C, isopropanol
Griseofulvin Amorphous/crystal 38–441 21 C, water
Hydrochlorothiazide Amorphous/crystal 21–113 37 C, HCl & PVP (aq.)
Iopanoic acid Amorphous/I-crystal 19-Dec 37 C, phosphate buffer (aq.)
Polythiazide Amorphous/crystal 48–455 37 C, HCl & PVP (aq.)
The range of values reflects the use of different ΔCp values for the calculations
a
Adapted from Hancock and Parks (2000)
magnitude of the predicted solubility advantage conducted by Pudipeddi and Serajuddin (2005).
for higher-energy polymorphs ranged from 1.1- In total, the authors evaluated 55 compounds
to 3.6-fold. However, predicted solubility ratios which resulted in 81 solubility ratios due to the
for amorphous drug forms were significantly existence of multiple polymorphic forms. Over-
higher and ranged from 12- to 1652-fold due to all, the average solubility ratio for the polymorphs
a higher free energy in the amorphous state. The evaluated was 1.7 (excluding the premafloxacin
relatively low-solubility ratios for polymorphs outlier). The solubility ratios from the literature
can be attributed to small differences in free were in agreement with and even included the
energy, while amorphous systems are capable of data presented by Hancock and Parks (2000).
much higher free energy values (Blagden et al. These values are summarized in Fig. 3.11. Addi-
2007). tionally, the authors evaluated 23 anhydrate/
Experimental solubilities were found to be hydrate solubility ratios and found that, in gen-
significantly less than the predicted values, as eral, the values were less than about 2. However,
shown in Table 3.6. However, all higher-energy there were cases in which the solubility ratio was
polymorphic forms provided some degree of sol- significantly higher than this value. The
ubility enhancement. The large discrepancy anhydrate/hydrate values are summarized in
between predicted and experimental solubility Fig. 3.12.
values for the amorphous substances was Although relative improvements in solubility
attributed to a strong driving force for recrystalli- are modest between different polymorphs or
zation in the dissolution media, as illustrated in pseudopolymorphs, for poorly water-soluble
Fig. 3.10. drugs that exhibit dissolution rate-limiting
A comprehensive literature review on the absorption, this difference may provide a signifi-
actual solubility ratio between polymorphs was cant increase in therapeutic activity.
a 3.0 b 3.0
5°C 25°C
Aqueous solubility (mg/100ml)

2.5 2.5
2.0 2.0
1.5 1.5
amorphous
1.0 1.0 amorphous
0.5 γ-crystal 0.5

γ-crystal
0.0 0.0
c 3.0
45°C
2.5
2.0
amorphous
1.5
1.0
α & γ-crystals
0.5
0.0
0 20 40 60 80 100 120
Time (minutes)
Fig. 3.10 Experimental aqueous solubility profiles for amorphous and crystalline indomethacin at (a) 5 C, (b) 25 C,
and (c) 45 C. (From Hancock and Parks (2000))
Fig. 3.11 Solubility ratios 5

for polymorphs (n ¼ 81).
The data do not include the 4
premafloxacin (I/III) ratio
Solubility Ratio
which was found to be 23.1.

3
(From Pudipeddi and
Serajuddin (2005))
2
0
0 20 40 60 80
Compound Number
In a study conducted by Kobayashi et al., the bioavailability was investigated (Kobayashi

effect of crystalline carbamazepine polymorphs et al. 2000). Calculated aqueous solubility values
on solubility, dissolution rate, and oral of form I, form III, and the dihydrate were found
Fig. 3.12 Anhydrate/ 25

hydrate solubility ratios
(n ¼ 23). Compound 6 is
20
(anhydrate/hydrate)
expressed as hydrate/
Solubility Ratio
anhydrate due to an
anomaly. (From Pudipeddi 15
and Serajuddin (2005))
10
0
0 5 10 15 20 25
Compound #
relatively rapidly. However, form III exhibited

sustained supersaturation, which is a desirable
characteristic for poorly water-soluble
compounds in order to achieve increased expo-
sure. Each polymorphic form was evaluated for
in vivo drug absorption using a crossover tech-
nique in male beagle dogs. The in vivo perfor-
mance of the polymorphic forms was compared
to a solubilized amorphous formulation,
representing 100% bioavailability. The study
demonstrated that at high dose, form I provided
the greatest Cmax and AUC0–12h values when
compared to the other two forms. Bioavailability
of form I, form III, and the dihydrate, relative to
the amorphous solubilized formulation, was
found to be 68.7%, 47.8%, and 33%, respec-
tively. This result is consistent with the probable
Fig. 3.13 Dissolution patterns of carbamazepine
polymorphs and dihydrate at 37 C in pH 1.2 media: conversion of form III to the dihydrate form in
square, form I; circle, form III; and triangle, dihydrate. situ. The plasma concentration versus time curve
(From Kobayashi et al. (2000)) for this arm of the study is shown in Fig. 3.14.
This result indicated that while there was a very
to be 460.2 μg/mL, 501.9 μg/mL, and 311.1 μg/ small difference in measured solubility between
mL, respectively. In agreement with the ranges form I and form III, the ability of form I to remain
outlined by Pudipeddi and Serajuddin, the solu- in a supersaturated state for an extended period of
bility ratio between the low- and high-energy time allowed for improved oral bioavailability
polymorphic states ranged from 1.5 to 1.6 for (Fig. 3.14).
form I and form III, respectively (Pudipeddi and In a recent study conducted by Kim et al., the
Serajuddin 2005). Dissolution profiles of the oral bioavailability of amorphous atorvastatin
three drug substances are illustrated in Fig. 3.13. hemi-calcium prepared by various techniques
Forms I and III exhibited a transient dissolution was evaluated (Kim et al. 2008). Specifically,
rate improvement, ultimately converting to the the researchers evaluated spray-drying and super-
more stable form in solution. While form I critical anti-solvent (SAS) processes against
initially provided the greatest solubility, it unprocessed crystalline material with low aque-
converted to the more stable dihydrate form ous solubility (142.2 μg/mL). Material processed
Fig. 3.14 Plasma

concentration–time curves
of carbamazepine
polymorphs and dihydrate
after oral administration to
dogs (n ¼ 4; meanS.
E.). Dose: 200 mg/body.
Diamond, solution; square,
form I; circle, form III; and
triangle, dihydrate. (From
Kobayashi et al. (2000))
100
80
% Dissolved
60
40
20
0
0 5 10 15 20 25 30
Time (min)
Fig. 3.15 Powder dissolution profiles of unprocessed solution ( filled triangle), spray-dried amorphous
atorvastatin particles ( filled square), SAS-processed atorvastatin calcium from an acetone solution (empty cir-
amorphous atorvastatin calcium precipitation from ace- cle), and spray-dried amorphous atorvastatin calcium from
tone ( filled circle), SAS-processed amorphous a tetrahydrofuran solution (empty square) (n ¼ 3),
atorvastatin calcium precipitated from a tetrahydrofuran (meanS.D.). (From Kim et al. (2008))
by these processes, which utilized acetone or particle size ranges from 68.7 to 95.7 nm and
tetrahydrofuran as the solvent, exhibited aqueous 3.62 to 7.31 μm, respectively. Powder dissolution
solubilities ranging from 467.1 to 483.2 μg/mL. analysis revealed that amorphous material
Amorphous material prepared from the SAS and provided significant improvements in dissolution
spray-drying processes were found to have rate, as illustrated in Fig. 3.15. Amorphous
particles prepared by the SAS processing method amorphous form can have a marked impact on
were found to have a faster rate of dissolution bioavailability.
than those prepared by spray-drying, consistent Cefuroxime axetil was the first orally delivered
with its small particle size. Unprocessed and second-generation cephalosporin reported by
amorphous materials were evaluated for in vivo Glaxo Group Ltd. (GlaxoSmithKline). The origi-
drug absorption in male rats. The hypothesis nal processes for the preparation of the axetil ester
tested with this study was that not only would produced the material either in relatively impure
amorphous materials provide enhanced absorp- amorphous form or in the form of pure crystalline
tion but that material prepared by SAS, due to material. The pure crystalline form of cefuroxime
its small particle size and large surface area, axetil was initially developed for commercial
would provide the highest absorption. From this evaluation; however, it did not exhibit adequate
study, it was determined that amorphous oral bioavailability (Crisp and Clayton 1985).
materials provided a significant increase in drug Moreover, side effects were observed due to
absorption, with SAS prepared material providing poor absorption of the antibiotic in the gastroin-
the greatest improvement. The amorphous form testinal tract (Somani et al. 2001). Murshedkav
prepared by SAS with acetone showed a threefold prepared amorphous cefuroxime axetil using
improvement in the AUC0–8h, a fourfold micro-fluidization and solvent evaporation
improvement in Cmax, and a twofold improve- methods and confirmed a twofold solubility
ment in Tmax. Similarly, spray-dried amorphous enhancement of amorphous cefuroxime axetil
material also provided a significant improvement over the crystalline form through a solubility
over crystalline drug. The plasma concentration study performed in 0.07 N HCl, as shown in
versus time curve from this study is shown in Fig. 3.17 (Murshedkar 2002). Oszczapowicz
Fig. 3.16. It is clear from this study that the et al. further proved that the amorphous form
2500
Atorvastatin Plasma Concentration (ng/ml)
2000
1500
1000
500
0
0 60 120 180 240 300 360 420 480
Time (min)
Fig. 3.16 Plasma concentration–time curves of unpro- tetrahydrofuran solution ( filled triangle), spray-dried
cessed atorvastatin particles ( filled square), amorphous atorvastatin calcium from an acetone solution
SAS-processed amorphous atorvastatin calcium precipita- (empty circle), and spray-dried amorphous atorvastatin
tion from acetone ( filled circle), SAS-processed amor- calcium from a tetrahydrofuran solution (empty square)
phous atorvastatin calcium precipitated from a (n ¼ 5), (meanS.D.). (From Kim et al. (2008))
Fig. 3.17 Solubility profile of untreated ( filled square), micro-fluidized (empty square), and solvent-evaporated (empty
triangle) cefuroxime axetil in 0.07 N HCl (meanS.D.). (From Murshedkar (2002))
exhibited higher bioavailability and adequate sta- the case of amorphous solids, which are normally
bility upon storage (Oszczapowicz et al. 1994). In stabilized by excipients or formulated as solid
order to obtain amorphous cefuroxime axetil in a dispersion systems prior to being incorporated
highly pure form, Crisp et al. studied several into solid dosage forms (Leuner and Dressman
solvent removing methods including spray- 2000).
drying, roller drying, solvent precipitation, and One example of polymer stabilization is the
freeze-drying. Spray-drying was chosen to for- case study of olanzapine, which is an antipsy-
mulate the commercial product of cefuroxime chotic drug with more than 60 crystalline
axetil, CeftinTM, given that highly pure amor- structures, considering anhydrates, solvates,
phous drug with a yield of about 70% was salts, and cocrystals (Tang et al. 2021). Most of
achieved through this method (Crisp et al. 1989, them share the same packing motif known as the
Crisp et al. 1991, Zimmerman 1991). The com- SC0 dimer (Fig. 3.18) (Askin et al. 2019). Among
mercial formulation of CeftinTM showed a bio- the anhydrates, olanzapine form I is the most
availability of 36% in fasted state and 52% in fed stable polymorph, with the lowest free energy
state with low incidence of side effects being (Luo et al. 2019). In fact, the polymorphs of
reported. olanzapine are particularly challenging to isolate.
The use of metastable solids has been shown to Testa et al. used solid-state NMR and synchrotron
be an effective method to enhance both solubility powder X-ray to analyze several samples of
and bioavailability of drug substances. While olanzapine raw materials from three different
most polymorphic forms generally provide rela- manufacturers. All of them were a mixture of
tively small improvements in solubility, the polymorphs I, II, and III, as shown in Fig. 3.19
improvement can be significant from a pharma- (Testa et al. 2019).
cological standpoint. However, these systems are According to Testa et al., the crystallization of
inherently unstable, and extreme care should be pure olanzapine form II is almost impossible
taken to ensure that the transformation into an (Testa et al. 2019). To date, olanzapine form III
undesirable form does not occur during has not been isolated as a single crystal (Tang
processing (e.g., granulation, drying, tableting, et al. 2021). In 2013, Bhardwaj et al. predicted
etc.) or storage. This is particularly difficult in other crystal forms of olanzapine without the SC0
Fig. 3.18 (a) Chemical structure of olanzapine and (b) D. Gonçalves, Min Zhao, Derek A. Tocher, Gareth
SC0 centrosymmetric dimer packing motif present in R. Williams, Simon Gaisford, and Duncan Q. M. Craig
almost all anhydrous and solvated crystalline forms of Crystal Growth & Design 2019 19 (5), 2751-2757). Copy-
olanzapine. (Reprinted with permission from Sean Askin, right (2019) American Chemical Society)
Jeremy K. Cockcroft, Louise S. Price, Andrea
Fig. 3.19 Magnification of

the experimental
synchrotron X-ray
diffraction of samples of
olanzapine raw materials
and calculated X-ray
patterns of forms I and
II. The reflections of form
III are indicated with (*).
(Adapted from Testa et al.
(2019))
dimer that were thermodynamically possible of olanzapine PVP at 30% drug loading. They
(Bhardwaj et al. 2013). Years later, Askin et al. crystallized the drug in the polymeric matrix by
reported for the first time the generation of heating at 140 C for 1 h. Thereafter, the material
olanzapine form IV (Askin et al. 2019). They was poured in water to dissolve PVP and
prepared this polymorph by recrystallizing centrifuged to isolate the olanzapine crystals that
olanzapine from an amorphous solid dispersion were identified using DSC-XRD simultaneously.
Table 3.7 Summary of the olanzapine polymorphs that crystallized in each polymer dispersion (*)
Dispersion Form I Form II Form III Form IV
PVP X*
HPMCAS X X*
PLA X X X*
PLGA X X X*
X signifies that the form was observed in the dispersion, and X* signifies that this was the predominant form following
crystallization
Adapted with permission from Sean Askin, Andrea D. Gonçalves, Min Zhao, Gareth R. Williams, Simon Gaisford, and
Duncan Q. M. Craig. Molecular Pharmaceutics 2020 17 (11), 4364–4374. Copyright (2020) American Chemical Society
In this case, the polymer, PVP, allowed the crys- difference being the physical state of individual
tallization and worked as a template that avoided components (Morissette et al. 2004), that is, the
the formation of the SC0 dimers (Askin et al. crystal may be classified as a solvate if one of the
2019). Moreover, olanzapine form IV has a components is a liquid and a cocrystal if both of
lower lattice energy than forms II and III (Askin the components are solid.
et al. 2019). Lately, Askin et al. repeated the same While solvates are capable of providing disso-
experiment using different polymers to prepare lution enhancement, they are of little pharmaceu-
amorphous solid dispersions. Interestingly, the tical value due to inclusion of organic solvents
resulting polymorph depended on the selected that are often not biocompatible. Furthermore,
polymer, as shown in Table 3.7. In other words, solvates are susceptible to desolvation which
the interactions between olanzapine and a specific can potentially lead to recrystallization of a less-
polymer determined the resulting polymorph by soluble polymorphic form. Compared to solvates,
acting as a template for crystallization (Askin cocrystals exhibit a higher-degree physical stabil-
et al. 2020). ity and may incorporate innocuous cocrystal
As mentioned before, to date, olanzapine form formers, the number of which is numerous
III has not been prepared pure in the pure form (Schultheiss and Newman 2009). While cocrystal
and usually crystallizes in a mixture with form II systems have been studied for many years, their
(Testa et al. 2019). Recently, Tang et al. proposed ability to improve solubility and bioavailability of
that olanzapine form IV could be used to generate poorly water-soluble drug substances has only
form III at temperatures higher than 200 K using recently been fully realized (Good and
quantum mechanical methods (Tang et al. 2021). Rodríguez-Hornedo 2009).
3.4 Pharmaceutical Cocrystals 3.4.1 Cocrystal Preparation
Pharmaceutical cocrystals have emerged as an In general, the preparation of pharmaceutical

alternative to salts and metastable solids to cocrystals can be separated into two major steps:
enhance solubility, dissolution, and bioavailabil- design and selection of cocrystal formers and
ity. Cocrystals are defined as a crystalline material process and scale-up of manufacturing. Selection
comprised of at least two molecular species held of a number of complementary cocrystal formers
together by noncovalent interactions (Byrn et al. for a particular drug molecule is a critical aspect
1999a). In addition to their ability to form with of cocrystal design and screening. A suitable
neutral drug substances, cocrystals may be cocrystal former in the context of pharmaceutical
formed with acidic, basic, or salt forms of drug cocrystals must be inherently safe for human
substances (Schultheiss and Newman 2009). administration and should interact with the drug
Morissette et al. described cocrystals as being substance to form a stable cocrystal. The process
structurally similar to solvates, with the primary of determining an optimal cocrystal former is
Fig. 3.20 Common supramolecular synthons – interactions between functional groups on different molecules in crystals
that can sometimes be used in a predictive fashion to engineer crystal and cocrystal structures. (From Berry et al.)
generally best accomplished by selection of use of the Cambridge Structural Database (CSD)
potential cocrystal formers followed by an empir- (Trask et al. 2005a). Hierarchies of supramolecu-
ical study based upon drug substances’ and lar synthons that can occur between functional
cocrystal formers’ supramolecular compatibility, groups (e.g., carboxylic acids, amides, alcohols,
polarity, lattice energy, or molecular electrostatic etc.) have also been published by researchers in
potential surfaces (Abramov et al. 2012, Fábián the field of crystal engineering (Fleischman et al.
2009, Karamertzanis et al. 2009, Musumeci et al. 2003; Bis et al. 2007; Shan and Zaworotko 2008;
2011). High-throughput screening techniques Shattock et al. 2008). The synthons are patterns of
may also be utilized to cover a wider range of repeating intermolecular interactions in the crys-
cocrystal formers, solvents, and crystallizing talline structure of cocrystals (e.g. carboxylic acid
conditions (Morissette et al. 2004). Compatibility dimer, π-π stacking, amide hydrogen bonding,
between a cocrystal former and a drug substance etc.). Some examples of synthon patterns are
is highly dependent on the chemical structure of shown in Fig. 3.20 (Berry and Steed 2017).
both compounds. Therefore, the first step in Specific knowledge of noncovalent
designing a cocrystal is to gain a detailed under- interactions between different types of functional
standing of the functional groups present in the groups, known as heterosynthons, allows one to
drug substance as this will facilitate the selection apply this information to a specific crystal engi-
of a proper cocrystal former (Shan and Zaworotko neering application.
2008). Statistical analysis of known interactions In a recent review, Shan and Zaworotko
between functional groups is accomplished with provided carboxylic acid–aromatic nitrogen,
Raw data:
570 entries both COOH and aromatic nitrogen
433 (76%) entries COOH---Narom supramolecular heterosynthon
30 (5%) entries COOH supramolecular hemosynthon
Refined data:
121 entries COOH and Narom exclusively on hydrocarbon skeleton
117 (97%) hits COOH---Narom supramolecular heterosynthon
6 (7%) hits COOH supramolecular homosynthon
Fig. 3.21 A comparison of the Cambridge Structural that contain both functional groups rather than just
Database statistics associated with the carboxylic acid– cocrystals, and the “refined data” refers to compounds
carboxylic acid supramolecular homosynthon versus the that do not contain additional hydrogen-bond donors or
carboxylic acid–aromatic nitrogen supramolecular acceptors. (From Shan and Zaworotko (2008))
heterosynthon. The “raw data” refers to all compounds
Fig. 3.22 Representation

of the homosynthon
between two
carbamazepine molecules
in carbamazepine–
saccharin form I and the
heterosynthon between a
carbamazepine and a
saccharin molecule in
carbamazepine–saccharin
form II. (From Porter et al.
(2008))
carboxylic acid–amide, and alcohol–pyridine as polymer heteronuclei (Porter III et al. 2008). The
examples of heterosynthons that appear to facili- widely known carbamazepine–saccharin cocrystal
tate the formation of cocrystals (Shan and (form I) was shown to incorporate a homosynthon
Zaworotko 2008). The carboxylic acid–aromatic between carbamazepine molecules (Fig. 3.22),
nitrogen heterosynthon and associated statistical while the newly discovered form (Form II)
data obtained from the CSD are illustrated in contained a heterosynthon between carbamazepine
Fig. 3.21. The data for this particular and saccharin (Fig. 3.22). In form I of the cocrystal,
heterosynthon demonstrate that it is much more molecules were found to pack in such a way that
likely to occur than the carboxylic acid–carbox- the homosynthon formed between two inversion-
ylic acid homosynthon. related carbamazepine carboxamide groups, as
Porter et al. discovered a polymorphic form of shown in Fig. 3.23a. As demonstrated in
the carbamazepine–saccharin cocrystal in a study Fig. 3.23b, the primary feature of form II is differ-
evaluating crystallization in the presence of ent packing due to heterosynthon interactions.
Fig. 3.23 Molecular

packing for (a)
form I and (b)
form II. (From Porter et al.
(2008))
Selection of a suitable cocrystal former will be (Chiarella et al. 2007, Childs et al. 2008,
highly dependent on the specific drug substance Rodríguez-Hornedo et al. 2006), mechanical–
and its associated functional groups that allow for chemical grinding (Delori et al. 2012, James
noncovalent interactions. However, selection et al. 2012), and melt crystallization (Berry et al.
should adhere to substances that have been previ- 2008, Lu et al. 2008). Among those techniques,
ously shown to exhibit some degree of biocom- cocrystals are mainly prepared through solution
patibility. This limits selection to those crystallization techniques similar to those used in
compounds that appear in the Priority-based the preparation of pharmaceutical salts (e.g., sol-
Assessment of Food Additives (PAFA) including vent evaporation, temperature cooling, and anti-
compounds classified as generally recognized as solvent addition) (Zhang et al. 2007). In addition
safe (GRAS), the Everything Added to Food in to the three major techniques mentioned above,
the United States (EAFUS) database, and the other methods have been described in the litera-
Inactive Ingredient Database available from the ture for the preparation of cocrystals including
Food and Drug Administration. sonic-slurrying, fast evaporation, super-critical
Having selected proper cocrystal formers, fluids, wet/dry compression (Bag et al. 2011,
preparation of cocrystals can be achieved by Bag and Reddy 2012, Childs et al., 2008, Lu
three general approaches: solution crystallization et al., 2008), and, to a much lesser extent,
freeze-drying, melt crystallization, and sublima- (Medina et al. 2010). Besides the dry grinding
tion (Eddleston et al. 2013, Medina et al. 2010). process, mechanical–chemical grinding can also
However, from a pharmaceutical manufacturing be achieved with a small amount of solvent to
standpoint, solution crystallization and solid-state enhance kinetics and facilitate the formation of
grinding represent the two methods that are most cocrystals, so that cocrystal formation occurs by
easily scalable for manufacturing of pharmaceuti- reaction in a small liquid phase and/or via an
cal cocrystals. amorphous phase (Trask et al. 2004, 2005a, b).
The solution crystallization method requires While many cocrystals can be prepared by
careful selection of solvents utilized such that all solvent crystallization and solid-state grinding,
components (e.g., drug substance and cocrystal some can only be prepared by a specific tech-
former) completely dissolve without interfering nique. Etter and Adsmond described systems
with the interactions necessary for cocrystal for- that could only be prepared by co-grinding
mation (Morissette et al. 2004). The drug sub- (Etter and Adsmond 1990). Conversely, Etter
stance and cocrystal former should also have et al. described another situation where a
similar solubility in the chosen solvent such that cocrystal could be formed by solution crystalliza-
a cocrystal precipitates prior to any individual tion but not by solid-state grinding (Etter et al.
component (Blagden et al. 2007). Because of 1993). However, oftentimes the cocrystal
this, the major disadvantage of solution crystalli- obtained is independent of the processing tech-
zation method is that individual components may nique utilized (Shan and Zaworotko 2008).
crystallize separately during solvent evaporation There are emerging techniques used in the
or cooling. preparation of cocrystals, such as resonant acous-
In the solid state, stoichiometric amounts of tic wet granulation, spray-drying, and hot-melt
the drug substance and cocrystal former can be extrusion (HME) (Grossjohann et al. 2015;
applied with mechanical–chemical grinding force Walsh et al. 2018; Tanaka et al. 2021).
in order to form cocrystals (Etter et al. 1993, In particular, HME is a large-scale
Kuroda et al. 2002). The driving force of manufacturing method that can be used in the
cocrystal formation is often the stronger hydrogen preparation of cocrystals. One limitation of
bonds in the cocrystal than those present in the HME is that it usually requires not only the drug
crystals of either pure component (Etter et al. and conformer but also a matrix carrier (e.g.,
1993). Medina et al. recently described the utility sugars, polymer) when aiming to a final pharma-
of twin-screw extrusion as a method to prepare ceutical product, as shown in Fig. 3.24.
cocrystals without the use of additional solvent Depending on the carrier, and sometimes the
Fig. 3.24 Schematic illustration of cocrystallization mech- Tao Yu, Yiwei Tian, Colette Lagan, David S. Jones, and
anism, inside the extruder barrel, from a noncomplementary Gavin P. Andrews Molecular Pharmaceutics 2018 15 (9),
carrier excipient. (Reprinted with permission from Shu Li, 3741–3754. Copyright (2018) American Chemical Society)
temperature, the final extrudate can be an amor- (Miller et al. 2008). Due to its poor water solubil-
phous solid dispersion instead of a cocrystal sus- ity characteristics, itraconazole is marketed as a
pension (Li et al. 2016; Walsh et al. 2018). The solid dispersion in which the drug is molecularly
amorphization of cocrystal also happens in other dispersed in a hypromellose-based matrix.
manufacturing methods as spray freeze-drying Remenar et al. utilized a cocrystal approach in
and spray-drying (Tanaka et al. 2021). an attempt to improve solubility and dissolution
Li et al. used the FloryHuggins theory and rates of this compound (Remenar et al. 2003).
phase diagrams to increase the process yield of Succinic acid, malic acid, and tartaric acid were
ibuprofen–isonicotinamide cocrystals using successfully utilized as cocrystal formers in a
HME (Li et al. 2018). They discussed that melting solvent crystallization process. Through hydro-
and mechanochemical methods are not as efficient gen bonding, each cocrystal contained two
as solvent-based methods in the preparation of molecules of itraconazole and one cocrystal for-
cocrystals. However, when adding a carrier that mer, as shown in Fig. 3.25. Dissolution in 0.1 N
melts at processing temperatures, the melted car- HCl was studied to assess relative differences of
rier can provide a medium for molecular collisions Sporanox®, crystalline itraconazole-free base,
and cocrystallization, similar to a solvent from the and cocrystals of itraconazole, with results
solvent-based methods (Li et al. 2018). shown in Fig. 3.26. The cocrystal formed with
malic acid provided a dissolution profile similar
to that of Sporanox® beads, with solubility
3.4.2 Solubility of Cocrystals improving by a factor of 20-fold over the free
base.
Itraconazole is a weak base that exhibits poor Hickey et al. prepared cocrystals of carbamaz-
solubility in both acidic and neutral environments epine, a poorly water-soluble drug substance, and
Fig. 3.25 Cocrystal of two

itraconazole molecules and
a single cocrystal former.
(From Remenar et al.
(2003))
8×10–4
[Itraconazole] (M)
0 100 200 300 400

Time (min)
Fig. 3.26 Itraconazole dissolution profiles of Sporanox® (upside down triangle), L-tartaric acid (circle), and
beads (square), crystalline itraconazole-free base (dia- succinic acid (triangle). Dissolution media: 0.1N HCl at
mond), and cocrystals of itraconazole with L-malic acid 25 C. (From Remenar et al. (2003))
evaluated their performance against the marketed aqueous solubility and therefore bioavailability
drug product, Tegretol® (Hickey et al. 2007). for ionizable drugs (anionic, cationic, and zwit-
Utilizing a solvent crystallization process, the terionic), the cocrystal approach is a particularly
researchers prepared cocrystals of carbamazepine attractive alternative for non-ionizable drugs, or
with saccharin (form I) which exhibited the crys- compounds with pKa values in a range where
tal packing that was previously discussed stable salt formation is limited. Rajput et al.
(Fig. 3.23a). To compare the performance of car- recently discovered 13 new solid forms of
bamazepine cocrystals to Tegretol®, an in vivo etravirine in the process of polymorph and
study was conducted in fasted beagle dogs. The cocrystal/salt screening to improve the solubility
cocrystal formulation exhibited a higher AUC of this anti-HIV drug (Rajput et al. 2013). The
and Cmax value than Tegretol® with a comparable in vitro solubility study indicated that etravirine
Tmax. The plasma concentration versus time plot salts, formed with strong acids (hydrochloric
for this study is shown in Fig. 3.27. acid, methanesulfonic acid, benzenesulfonic
McNamara et al. showed that cocrystals of a acid, and p-toluenesulfonic acid) because of the
development candidate drug could be formed at a low basicity of etravirine, showed only 1.5 times
1:1 ratio with glutaric acid (McNamara et al. improvement in solubility with respect to the
2006). The cocrystal exhibited an 18-fold parent API and that too with poor attendant sta-
improvement in solubility over the crystalline bility. On the other hand, the selected benzenetri-
form of the drug substance. The in vivo perfor- carboxylic acid cocrystal hydrate form of
mance of the cocrystals was compared to bulk etravirine has shown up to sixfold solubility
drug substance in beagle dogs. From this study, enhancement.
it was determined that at low dose (5 mg/kg) and The formation of cocrystals allows for the
high dose (10 mg/kg), the cocrystal formulation opportunity to readily modify solid-state
exhibited the highest Cmax and AUC0–36h when properties of a drug substance, resulting in
compared to the bulk drug substance. The plasma forms that can have improved therapeutic effec-
concentration versus time plots for the low-dose tiveness. The cocrystal approach is rapidly
and high-dose study are shown in Fig. 3.28. gaining interest and has the potential to largely
Although salt formation may be the simplest impact the successful preparation poorly water-
and most cost-effective strategy to enhance soluble drug substances. It is important to
Fig. 3.27 Average

plasma–time curves of
carbamazepine 3000
concentration (S.E.) from C0-crystal 1
Plasma concentration (ng/mL)
a crossover experiment in Tegretol®

fasted beagle dogs (n ¼ 4)
given oral doses of 200 mg
2000
of the active drug as
Tegretol® tablets and
cocrystal. (From Hickey
et al. (2007)) 1000
0 2 4 6 8 10 12
t (hours)
Fig. 3.28 Dog plasma a 160

concentration with time for
(a) 5 mg/kg dosing of drug
140
substance (closed circles)
and cocrystal (open circles)
120
and (b) 50 mg/kg dosing of
Plasma Conc. (ng/mL)

drug substance (closed
circles) and cocrystal (open 100
circles). (Adapted from
McNamara et al. (2006)) 80
60
40
20
0
0 10 20 30 40
Time (hours)
b 400
300
Plasma Conc. (ng/mL)
200
100
0
0 10 20 30 40
Time (hours)
mention that cocrystals can have different poly- shapes for pulmonary delivery, which is difficult
morphic forms, as discussed previously in this to achieved using solvent evaporation or grinding
chapter. For instance, Grossjohann et al. obtained methods. In the study, niclosamide–nicotinamide
a polymorphic form of a sulfadimidine/4- 1:1 cocrystals obtained by spray-drying and sol-
aminosalicylic acid 1:1 cocrystal when using vent evaporation methods were contrasted.
spray-drying (form II). This was a different poly- Spray-dried cocrystals were spherical and
morphic form of the cocrystal than the obtained exhibited unimodal size distribution. On the
after liquid-assisted milling (form I) (Grossjohann other hand, needle-shaped cocrystals were
et al. 2015). The manufacturing method plays a obtained by solvent evaporation. The spherical
significant role in the final purpose of the shape and uniformity achieved by spray-drying
cocrystals. Ray et al. used spray dry to obtain improved the powder’s aerodynamic properties
particles with sizes around 1 to 5 μm and regular for inhalation (Ray et al. 2020).
Fig. 3.29 Crystal packing diagrams of (a) theophylline monohydrate and (b) THAOXA cocrystal. The values in the
figure indicate the distance in A between two atoms. (From Tanaka et al.)
Tanaka et al. used spray-freeze-drying (SFD) water. The distance between theophylline–water
to prepare powders for dry powder inhalation (hydrate form) was 2.053 A and theophylline–
(DPI). SFD process can produce highly porous oxalic acid was 1.758 A (Fig. 3.29), meaning
powder particles with optimized aerodynamic that the interactions with the conformer are stron-
properties for DPI. However, the high powder ger than interaction with water (Tanaka et al.
surface area and the trace amount of amorphous 2020).
domain generated during SFD process caused In pharmaceutical manufacturing, relevant
stability issues. Chemically, these powders are properties of bulk powders include elasticity,
especially susceptible to heat and moisture; phys- plasticity, and fragmentation, among others
ically, the powder particles can aggregate and (Yousef and Vangala 2019). In an interesting
resulted in less optimized deposition patterns of study, Krishna et al. prepared different cocrystals
the inhalers. In addition, Tanaka et al. used the of 6-chloro-2,4-dinitroaniline and vanillin
drug theophylline, which is particularly challeng- isomers, focusing on their mechanical properties
ing due to its transition to a hydrate form in the for tableting. They showed that the synthons
presence of humid environments. SFD was used determined the formation of the layer packing,
for the preparation of theophylline–oxalic acid which was responsible for the mechanical
cocrystals. These cocrystals had an increased properties of the materials. Depending on the 3D
moisture absorption resistance, which can be packing, the crystals were brittle, plastic, or elas-
exemplified by comparing the intermolecular tic. They concluded that bendable-type crystals
distances between drug–conformer and drug– have better properties for tableting than the
shearing or brittle types because their slip planes Equipment and Reagents
provide higher plasticity (Krishna et al. 2015). • Rifaximin
In addition, several papers have shown that • Sintered glass Gooch filter
cocrystals can improve the physical properties • Vacuum oven
for tableting, milling, and other common pharma- Method
ceutical processes (Sun and Hou 2008; Sanphui • β-form. Collect wet rifaximin in a sintered
et al. 2015). In the manufacturing of tablets, glass Gooch filter and wash with water. Dry
powders with low plastic deformation recover the powder in a vacuum oven at room tem-
their volume after removing the pressure from perature to the residual water content rang-
the punches after compaction. Excessive elastic ing from 6% to about 40% (w/w).
recovery can cause tablet capping or delamina- • δ-form. Collect wet rifaximin in a sintered
tion, especially during high-speed tableting pro- glass Gooch filter and wash with water. Dry
cess. The cocrystals can improve these properties the powder in a vacuum oven at 65 C to the
by creating flat slip planes. Crystals with flat residual water content ranging from 4 to 6%
layers connected by hydrogen bonds have, in (w/w).
general, better crystal plasticity for this kind of • α-form. Collect wet rifaximin in a sintered
processes (Sun and Hou 2008). glass Gooch filter and wash with water.
Dry the powder in a vacuum oven at
65 C to the residual water content less
than 3% (w/w).
3.5 Summary
• γ-form. Collect wet rifaximin in a sintered
glass Gooch filter. Dry the powder in a
Solid-state modifications address the strong inter-
vacuum oven at 65 C to the residual
molecular interactions responsible for poor aque-
water content less than 4% (w/w).
ous solubility by taking advantage of properties
• ε-form. Dry the δ-form 3 hours in a vacuum
inherent to specific drug substances. In doing so,
oven at 65 C.
solubility can be enhanced by a large degree,
Results
allowing for improved clinical relevance. The
• X-ray powder diffraction, solid-state 13C
solid-state techniques that have been described
NMR, and HATR-IR spectroscopy
in the literature include the formation of salts,
identified and characterized the five distinct
metastable solids, and cocrystals. Each technique
crystal forms of rifaximin.
has specific advantages that make it amenable to
• The in vitro solubility study indicated a
certain applications. Due to the nature of each
great solubility difference range from 0.07
technique, selection of optimal systems is often
to 6.00 absorbance unit.
empirical and requires the use of efficient screen-
• After oral administration of 100 mg/kg in
ing techniques. However, if applied properly,
the beagle dogs, the Cmax values of the
solid-state techniques are capable of improving
crystal forms ranged from 1.1 to 1085.31
the clinical relevance of many poorly water-
ng/ml, and the AUC0-24 h values ranged
soluble drug substances.
from 10 to 4795 ng*h/ml.
Method Capsule 1
Method Capsule 2
Preparation of Polymorphism: Solvent
Preparation of Amorphous Solids: Spray-
Evaporation
Drying
Based on the method reported by Viscomi
Based on the method reported by Kim et al.
et al. (2008)
(2008)
Objective
Objective
• To obtain different crystal forms of a drug
• To obtain an amorphous drug substance by
substance and evaluate the solubility and
the spray-drying technique.
impact on bioavailability.
Equipment and Reagents • Laboratory-scale supercritical anti-solvent

• Atorvastatin calcium processor
• Acetone or tetrahydrofuran • Carbon dioxide (CO2)
• Laboratory-scale spray dryer Method
Method • Deliver CO2 into particle formation vessel
• Dissolve atorvastatin calcium in acetone or equilibrated at 40 C until the pressure
tetrahydrofuran at a concentration of reaches 12 MPa.
100 mg/mL. • Dissolve atorvastatin calcium in acetone or
• Set the drying air flow rate to 0.70 m3/min. tetrahydrofuran at a concentration of
• Set the inlet temperature of the spray dryer 100 mg/mL.
to 70 C. • The drug solution and supercritical CO2
• Feed the solution into the spray dryer at were co-injected through a two-flow nozzle
3 mL/min with an atomization pressure of at 0.5 g/min and 45 g/min, respectively.
10 kPa. • After the drug solution was exhausted,
• Ensure that the outlet temperature is in the fresh CO2 was cycled into the vessel to
range of 62–65 C. remove residual solvent at 45 g/min.
• Collect the dried powder. Results
Results • Laser diffraction analysis revealed that the
• Laser diffraction analysis revealed that the particle size of SAS-processed atorvastatin
particle size of atorvastatin calcium spray- particles from acetone and tetrahydrofuran
dried from acetone and tetrahydrofuran was was 68.7 15.8 nm and 95.7 12.2 nm,
3.62 0.15 μm and 7.31 0.21 μm, respectively. Particle size of the unpro-
respectively. cessed material was 3.83 0.08 μm.
• BET-specific surface area analysis • BET-specific surface area analysis
demonstrated that the surface area of demonstrated that the surface area of SAS
atorvastatin particles spray-dried from ace- processed atorvastatin particles from ace-
tone and tetrahydrofuran was 3.69 0.06 tone and tetrahydrofuran was
m2/g and 0.95 0.03 m2/g, respectively. 120.35 1.40 m2/g and 79.78 0.93
The specific surface area of unprocessed m2/g, respectively. The specific surface
material was 14.56 0.17 m2/g. area of unprocessed material was
• Differential scanning calorimetry and X-ray 14.56 0.17 m2/g.
diffraction analyses indicated that all spray- • Differential scanning calorimetry and X-ray
dried material lacked crystalline character diffraction results indicated that all spray-
and was amorphous. dried material lacked crystalline character
and was amorphous.
Method Capsule 3
Method Capsule 4
Preparation of Amorphous Solids: Supercriti-
cal Anti-solvent Processing Preparation of Amorphous Solids: Melt
Based on the method reported by Kim et al. Quenching
(2008) Based on the method reported by Hancock and
Objective Parks (2000)
• To obtain an amorphous drug substance by Objective
supercritical anti-solvent (SAS) • To obtain an amorphous drug substance by
processing. melt quenching.
Equipment and Reagents Equipment and Reagents
• Atorvastatin calcium • Indomethacin
• Acetone or tetrahydrofuran • Liquid nitrogen
Method Objective
• Heat indomethacin to a temperature that • To obtain a cocrystal of carbamazepine and
induces melting (160 C). saccharin by temperature-induced
• Quench the melt with liquid nitrogen such precipitation.
that indomethacin solidifies. Equipment and Reagents
Results • Carbamazepine
• Differential scanning calorimetry and X-ray • Saccharin
diffraction results indicated that melt- • Ethanol
quenched material lacked crystalline char- • Methanol
acter and was amorphous. • Water-jacketed glass crystallization vessel
Method
Method Capsule 5 • Anhydrous carbamazepine (0.089 mol) and
saccharin (0.089 mol) were combined in the
Preparation of Amorphous Solids: Freeze-
crystallization vessel.
Drying
• Solids were dissolved in 280 mL of a 62.5/
Based on the method reported by Crisp (1991)
37.5% (v/v) ethanol/methanol, mixed, and
Objective
heated to 70 C for 1 h under reflux.
• To obtain an amorphous drug substance by
• Temperature was decreased in 10 C
freeze-drying.
increments while stirring to induce
precipitation.
• Cefuroxime axetil R- and S-isomers
• Following equilibration at 30 C, solids
• 1,4-Dioxane or t-butanol
were isolated using a Buchner funnel and
• Lyophilizer
rinsed with cold ethanol.
• Vacuum oven
• The resulting powder was air-dried.
Method
Results
• Dissolve cefuroxime axetil 1:1 mixture of
• A product yield of 76% was obtained.
the R- and S-isomers of cefuroxime axetil
• Microscopy studies showed that the particle
in 1,4-Dioxane at a concentration of
size of the crystals was between 500 and
100 mg/mL.
1,000 μm.
• Freeze the drug solution at 25 C for
• X-ray powder diffraction and differential
20 hours; the solvent was then removed
scanning calorimetry studies demonstrated
by freeze-drying at 30 mTorr for 2 days.
that a single polymorphic cocrystal form
• Collect the product pass through 40 mesh
was prepared.
sieve.
• Dry the sieved material in vacuum oven at
Method Capsule 7
50 C for 20 hours to remove residual
solvent. Preparation of Cocrystals: Seed-Induced
Results Precipitation
• The infrared (Nujol) spectrum and micro- Based on the method reported by McNamara
scopic examination results indicated that et al. (2006)
freeze-drying material was amorphous. Objective
• To obtain cocrystals of a new drug candi-
Method Capsule 6 date (compound 1) with glutaric acid by
seed-induced precipitation.
Preparation of Cocrystals: Temperature-
Induced Precipitation
• Compound 1
Based on the method reported by Hickey et al.
• Glutaric acid
(2007)
• Chloroform Method
• Water-jacketed glass crystallization vessel • Equimolar amounts of caffeine and glutaric
Method acid were combined in a stainless steel
• Compound 1 (8.431 mmol) and glutaric grinding jar.
acid (8.410 mmol) were dissolved in boil- • Optional: add four drops of either a
ing chloroform with stirring. non-polar solvent or a polar solvent.
• The solution was concentrated by continued • Grind the materials together.
boiling until the volume was 50 mL. • Allow any residual solvent to evaporate.
• Cocrystal seeds (generated in thermal Results
experiments) were introduced into the hot • X-ray diffraction analysis demonstrated
solution. that cocrystals were formed when the mate-
• After crystallization began, the solution rial was prepared in the absence of solvent
was cooled over a 15 min period. and with non-polar or polar solvents.
• Approximately 100 mL of cyclohexane was • Form I of the cocrystal was found to form
added and the solution was cooled on ice when no solvent or a non-polar solvent
for 30 min. was used.
• The cocrystal was isolated by filtration. • Form II was predominantly formed when a
• The resulting powder was air-dried. polar solvent was used.
Results
• A product yield of 92% was obtained.
• The volumetric median diameter Dv(0.5)
References
and that of the 90th percentile Dv
(90) were found to be 49 μm and 131 μm, Abramov YA, Loschen C, Klamt A. Rational coformer or
respectively. solvent selection for pharmaceutical cocrystallization
• X-ray diffraction analysis showed a unique or desolvation. J Pharm Sci. 2012;101:3687–97.
pattern that was distinguished from com- Aguiar AJ, Zelmer JE. Dissolution behavior of
polymorphs of chloramphenicol palmitate and
pound 1 and glutaric acid. mefenamic acid. J Pharm Sci. 1969;58:983–7.
• Differential scanning calorimetry Aguiar AJ, Krc J, Kinkel AW, Samyn JC. Effect of poly-
experiments indicated that the cocrystal morphism on the absorption of chloramphenicol from
exhibited a melting point different from chloramphenicol palmitate. J Pharm Sci. 1967;56:847–
53.
that of compound 1 or glutaric acid. Alhalaweh A, Alzghoul A, Kaialy W, Mahlin D,
Bergström CAS. Computational predictions of glass-
Method Capsule 8 forming ability and crystallization tendency of drug
molecules. Mol Pharm. 2014;11:3123–32.
Preparation of Cocrystals: Grinding Alhalaweh A, Alzghoul A, Mahlin D, Bergström CAS.
Based on the method reported by Trask et al. Physical stability of drugs after storage above and
below the glass transition temperature: relationship to
(2004) glass-forming ability. Int J Pharm. 2015;495:312–7.
Objective Allesø M, van den Berg F, Cornett C, Jørgensen FS,
• To obtain cocrystals of caffeine and glutaric Halling Sørensen B, de Diego HL, Hovgaard L,
acid by a solid-state grinding technique. Aaltonen J, Rantanen J. Solvent diversity in poly-
morph screening. J Pharm Sci. 2008;97:2145–59.
Equipment and Reagents Alvarez AJ, Singh A, Myerson AS. Polymorph screening:
• Anhydrous caffeine comparing a semi-automated approach with a high
• Glutaric acid throughput method. Cryst Growth Des. 2009;9:4181–
• Optional non-polar solvents: n-hexane, 8.
Askin S, Cockcroft JK, Price LS, Gonçalves AD, Zhao M,
cyclohexane, or heptane Tocher DA, Williams GR, Gaisford S, Craig DQM.
• Optional polar solvents: chloroform, Olanzapine Form IV: discovery of a new polymorphic
dichloromethane, acetonitrile, and water form enabled by computed crystal energy landscapes.
• Ball grinder/mill Cryst Growth Des. 2019;19:2751–7.
Askin S, Gonçalves AD, Zhao M, Williams GR, Byrn SR, Pfeiffer RR, Stowell JG. Drugs as molecular
Gaisford S, Craig DQM. A simultaneous differential solids. Solid-state chemistry of drugs. West Lafayette:
scanning calorimetry–X-ray diffraction study of SSCI; 1999a. p. 143–241.
olanzapine crystallization from amorphous solid Byrn SR, Pfeiffer RR, Stowell JG. Polymorphs. Solid-state
dispersions. Mol Pharm. 2020;17:4364–74. chemistry of drugs. West Lafayette: SSCI;
Bag PP, Reddy CM. Screening and selective preparation 1999b. p. 143–241.
of polymorphs by fast evaporation method: a case Chemburkar SR, Bauer J, Deming K, Spiwek H, Patel K,
study of aspirin, anthranilic acid, and niflumic acid. Morris J, Henry R, Spanton S, Dziki W, Porter W,
Cryst Growth Des. 2012;12:2740–3. Quick J, Bauer P, Donaubauer J, Narayanan BA,
Bag PP, Patni M, Malla Reddy C. A kinetically controlled Soldani M, Riley D, McFarland K. Dealing with the
crystallization process for identifying new cocrystal impact of ritonavir polymorphs on the late stages of
forms: fast evaporation of solvent from solutions to bulk drug process development. Org Process Res Dev.
dryness. Cryst Eng Comm. 2011;13:5650–2. 2000;4:413–7.
Baird JA, Van Eerdenbrugh B, Taylor LS. A classification Chiarella RA, Davey RJ, Peterson ML. Making Cocrystals
system to assess the crystallization tendency of organic the utility of ternary phase diagrams. Cryst Growth
molecules from undercooled melts. J Pharm Sci. Des. 2007;7:1223–6.
2010;99:3787–806. Childs SL, Rodriguez-Hornedo N, Reddy LS,
Bastin RJ, Bowker MJ, Slater BJ. Salt selection and Jayasankar A, Maheshwari C, McCausland
optimisation procedures for pharmaceutical new chem- L. Screening strategies based on solubility and solution
ical entities. Org Process Res Dev. 2000;4:427–35. composition generate pharmaceutically acceptable
Bauer J, Spanton S, Henry R, Quick J, Dziki W, Porter W, cocrystals of carbamazepine. CrstEngComm.
Morris J. Ritonavir: an extraordinary example of confor- 2008;10:856–64.
mational polymorphism. Pharm Res. 2001;18:859–66. Corrigan OI. Salt forms: pharmaceutical aspects. In:
Bechtloff B, Nordhoff S, Ulrich J. Pseudopolymorphs in Swarbrick J, editor. Encyclopedia of pharmaceutical
industrial use. Cryst Res Tech. 2001;36:1315–28. technology. New York: Informa Healthcare; 2006.
Berge SM, Bighley LD, Monkhouse DC. Pharmaceutical p. 3177–87.
salts. J Pharm Sci. 1977;66:1–19. Crisp HA, Clayton JC. Increased absorption via the gas-
Berry DJ, Steed JW. Pharmaceutical cocrystals, salts and trointestinal tract, allowing for oral or rectal adminis-
multicomponent systems; intermolecular interactions tration, 1985. US Patent 4,562,181
and property based design. Adv Drug Deliv Rev. Crisp HA, Clayton JC Elliott LG, Wilson EM. Preparation
2017;117:3–24. of a highly pure, substantially amorphous form of
Berry DJ, Seaton CC, Clegg W, Harrington RW, Coles SJ, cefuroxime axetil, 1989. US Patent 4,820,833
Horton PN. Applying hot-stage microscopy to Crisp HA, Clayton JC, Elliott LG, Wilson EM. Process for
cocrystal screening: a study of nicotinamide with preparing cefuroxime axetil, 1991. US Patent
seven active pharmaceutical ingredients. Cryst Growth 5,013,833
Des. 2008;8:1697–712. Dannenfelser R-M, He H, Joshi Y, Bateman S, Serajuddin
Bhardwaj RM, Price LS, Price SL, Reutzel-Edens SM, ATM. Development of clinical dosage forms for a
Miller GJ, Oswald IDH, Johnston BF, Florence poorly water soluble drug I: application of polyethyl-
AJ. Exploring the experimental and computed crystal ene glycol–polysorbate 80 solid dispersion carrier sys-
energy landscape of olanzapine. Cryst Growth Des. tem. J Pharm Sci. 2004;93:1165–75.
2013;13:1602–17. Delori A, Friscic T, Jones W. The role of mechanochem-
Bis JA, Vishweshwar P, Weyna D, Zaworotko istry and supramolecular design in the development of
MJ. Hierarchy of supramolecular synthons: persistent pharmaceutical materials. CrstEngComm. 2012;14:
hydroxyl pyridine hydrogen bonds in cocrystals that 2350–62.
contain a cyano acceptor. Mol Pharm. 2007;4:401–16. Eddleston MD, Patel B, Day GM, Jones
Blaabjerg LI, Lindenberg E, Löbmann K, Grohganz H, W. Cocrystallization by freeze-drying: preparation of
Rades T. Glass forming ability of amorphous drugs novel multicomponent crystal forms. Cryst Growth
investigated by continuous cooling and isothermal Des. 2013;13:4599–606.
transformation. Mol Pharm. 2016;13:3318–25. Elder DP, Delaney E, Teasdale A, Eyley S, Reif VD,
Black SN, Collier EA, Davey RJ, Roberts RJ. Structure, Jacq K, Facchine KL, Oestrich RS, Sandra P, David
solubility, screening, and synthesis of molecular salts. J F. The utility of sulfonate salts in drug development. J
Pharm Sci. 2007;96:1053–68. Pharm Sci. 2010;99:2948–61.
Blagden N, Md M, Gavan PT, York P. Crystal engineering Engel GL, Farid NA, Faul MM, Richardson LA,
of active pharmaceutical ingredients to improve solu- Winneroski LL. Salt form selection and characteriza-
bility and dissolution rates. Adv Drug Deliv Rev. tion of LY333531 mesylate monohydrate. Int J Pharm.
2007;59:617–30. 2000;198:239–47.
Bowker MJ, Stahl PH. Preparation of water soluble Etter MC, Adsmond DA. The use of cocrystallization as a
compounds through salt formation. In: Wermuth CG, method of studying hydrogen bond preferences of
editor. The practice of medicinal chemistry. New York: 2-aminopyrimidine. J Chem Soc Chem Commun.
Academic; 2008. p. 749. 1990:589–91.
Etter MC, Reutzel SM, Choo CG. Self-organization of with marketed product. Eur J Pharm Biopharm.
adenine and thymine in the solid state. J Am Chem 2007;67:112–9.
Soc. 1993;115:4411–2. Huang L-F, Tong W-Q. Impact of solid state properties on
Fábián L. Cambridge structural database analysis of developability assessment of drug candidates. Adv
molecular complementarity in cocrystals. Cryst Drug Deliv Rev. 2004;56:321–34.
Growth Des. 2009;9:1436–43. Jain N, Yalkowsky SH. Estimation of the aqueous
Fleischman SG, Kuduva SS, McMahon JA, Moulton B, solubility I: Application to organic nonelectrolytes. J
Bailey Walsh RD, Rodríguez-Hornedo N, Zaworotko Pharm Sci. 2001;90:234–52.
MJ. Crystal engineering of the composition of pharma- Jain N, Yang G, Machatha SG, Yalkowsky SH. Estimation
ceutical phases: multiple-component crystalline solids of the aqueous solubility of weak electrolytes. Int J
involving carbamazepine. Cryst Growth Des. 2003;3: Pharm. 2006;319:169–71.
909–19. James SL, Adams CJ, Bolm C, Braga D, Collier P, Friscic
Gardner CR, Almarsson O, Chen H, Morissette S, T. Mechanochemistry: opportunities for new and
Peterson M, Zhang Z, Wang S, Lemmo A, Gonzalez- cleaner synthesis. Chem Soc Rev. 2012;41:413–47.
Zugasti J, Monagle J, Marchionna J, Ellis S, Karamertzanis PG, Kazantsev AV, Issa N, Welch GWA,
McNulty C, Johnson A, Levinson D, Cima Adjiman CS, Pantelides CC. Can the formation of
M. Application of high throughput technologies to pharmaceutical cocrystals be computationally
drug substance and drug product development. predicted? 2. Crystal structure prediction. J Chem The-
Comput Chem Eng. 2004;28:943–53. ory & Comput. 2009;5:1432–48.
Good DJ, Rodríguez-Hornedo N. Solubility advantage of Kim J-S, Kim M-S, Park HJ, Jin S-J, Lee S, Hwang S-J.
pharmaceutical cocrystals. Cryst Growth Des. 2009;9: Physicochemical properties and oral bioavailability of
2252–64. amorphous atorvastatin hemi-calcium using spray-
Gould PL. Salt selection for basic drugs. Int J Pharm. drying and SAS process. Int J Pharm. 2008;359:211–9.
1986;33:201–17. Kobayashi Y, Ito S, Itai S, Yamamoto K. Physicochemical
Grant DJW, Higuchi T. Solubility, intermolecular forces, properties and bioavailability of carbamazepine
and thermodynamics. In: Saunders WH, editor. Solu- polymorphs and dihydrate. Int J Pharm. 2000;193:
bility behavior of organic compounds, vol. XXI. 137–46.
New York: Wiley-Interscience; 1990. p. 12–88. Krishna GR, Shi L, Bag PP, Sun CC, Reddy
Gross TD, Schaab K, Ouellette M, Zook S, Reddy JP, CM. Correlation among crystal structure, mechanical
Shurtleff A, Sacaan AI, Alebic-Kolbah T, Bozigian behavior, and tabletability in the cocrystals of vanillin
H. An approach to early-phase salt selection: Applica- isomers. Cryst Growth Des. 2015;15:1827–32.
tion to NBI-75043. Org Process Res Dev. 2007;11: Kumar L, Amin A, Bansal AK. An overview of automated
365–77. systems relevant in pharmaceutical salt screening.
Grossjohann C, Serrano DR, Paluch KJ, O’Connell P, Drug Discov Today. 2007;12:1046–53.
Vella-Zarb L, Manesiotis P, McCabe T, Tajber L, Kuroda R, Imai Y, Tajima N. Generation of a cocrystal
Corrigan OI, Healy AM. Polymorphism in phase with novel coloristic properties via solid state
sulfadimidine/4-aminosalicylic acid cocrystals: solid- grinding procedures. Chem Commun. 2002:2848–9.
state characterization and physicochemical properties. Lang M, Kampf JW, Matzger AJ. Form IV of carbamaze-
J Pharm Sci. 2015;104:1385–98. pine. J Pharm Sci. 2002;91:1186–90.
Gupta P, Chawla G, Bansal AK. Physical stability and Lee S, Hoff C. Large-scale aspects of salt formation:
solubility advantage from amorphous celecoxib: the processing of intermediates and final products. In:
role of thermodynamic quantities and molecular mobil- Stahl PH, Wermuth CG, editors. Pharmaceutical salts:
ity. Mol Pharm. 2004;1:406–13. properties, selection, and use. New York: Wiley-VCH;
Haleblian J, McCrone W. Pharmaceutical applications of 2002. p. 191–220.
polymorphism. J Pharm Sci. 1969;58:911–29. Leuner C, Dressman J. Improving drug solubility for oral
Hancock BC. Disordered drug delivery: destiny, dynamics delivery using solid dispersions. Eur J Pharm
and the Deborah number. J Pharm Pharmacol. Biopharm. 2000;50:47–60.
2002;54:737–46. Li S, Wong S, Sethia S, Almoazen H, Joshi YM,
Hancock BC, Parks M. What is the true solubility advan- Serajuddin ATM. Investigation of solubility and disso-
tage of the different forms? Pharm Res. 2000;17:397– lution of a free base and two different salt forms as a
404. function of pH. Pharm Res. 2005;22:628–35.
Hancock BC, Zografi G. Characteristics and significance Li S, Yu T, Tian Y, McCoy CP, Jones DS, Andrews
of the amorphous state in pharmaceutical systems. J GP. Mechanochemical synthesis of pharmaceutical
Pharm Sci. 1997;86:1–12. cocrystal suspensions via hot melt extrusion: feasibility
Hancock BC, Shamblin SL, Zografi G. Molecular mobility studies and physicochemical characterization. Mol
of amorphous pharmaceutical solids below their glass Pharm. 2016;13:3054–68.
transition temperatures. Pharm Res. 1995;12:799–806. Li S, Yu T, Tian Y, Lagan C, Jones DS, Andrews
Hickey MB, Peterson ML, Scoppettuolo LA, Morrisette GP. Mechanochemical synthesis of pharmaceutical
SL, Vetter A, Guzmán H, Remenar JF, Zhang Z, Tawa cocrystal suspensions via hot melt extrusion: enhanc-
MD, Haley S, Zaworotko MJ, Almarsson Ö. Perfor- ing cocrystal yield. Mol Pharm. 2018;15:3741–54.
mance comparison of a cocrystal of carbamazepine
Lipinski CA. Poor aqueous solubility: industry wide prob- properties of various forms of the 1-acetoxyethyl ester
lem in drug discovery. Am Pharmaceut Rev. 2002;5: of cefuroxime. Acta Pol Pharm. 1994;52:397–401.
82–5. Parks GS, Huffman HM, Cattoir FR. Studies on glass II:
Lipinski CA, Lombardo F, Dominy BW, Feeney the transition between the glassy and liquid states in
PJ. Experimental and computational approaches to the case of glucose. J Phys Chem. 1928;32:1366–79.
estimate solubility and permeability in drug discovery Parks GS, Snyder LJ, Cattoir FR. Studies on glass
and development settings. Adv Drug Deliv Rev. XI. Some thermodynamic relations of glassy and
2001;46:3–26. alpha-crystalline glucose. J Phys Chem. 1934;2:595–8.
Loftsson T, Brewster ME. Pharmaceutical applications of Paulekuhn GS, Dressman JB, Saal C. Trends in active
cyclodextrins. 1. Drug solubilization and stabilization. pharmaceutical ingredient salt selection based on anal-
J Pharm Sci. 1996;85:1017–25. ysis of the orange book database. J Med Chem.
Lu E, Rodríguez-Hornedo N, Suryanarayanan R. A rapid 2007;50:6665–72.
thermal method for cocrystal screening. Peterson ML, Morissette SL, McNulty C, Goldsweig A,
CrstEngComm. 2008;10:665–8. Shaw P, LeQuesne M, Monagle J, Encina N, March-
Luo H, Hao X, Gong Y, Zhou J, He X, Li J. Rational ionna J, Johnson A, Gonzalez-Zugasti J, Lemmo AV,
crystal polymorph design of olanzapine. Cryst Growth Ellis SJ, Cima MJ, Almarsson Ö. Iterative high-
Des. 2019;19:2388–95. throughput polymorphism studies on acetaminophen
Mao C, Pinal R, Morris KR. A quantitative model to and an experimentally derived structure for form III. J
evaluate solubility relationship of polymorphs from Am Chem Soc. 2002;124:10958–9.
their thermal properties. Pharm Res. 2005;22:1149–57. Porter WW III, Elie SC, Matzger AJ. Polymorphism in
McNamara D, Childs S, Giordano J, Iarriccio A, Cassidy J, carbamazepine cocrystals. Crys Growth Des. 2008;8:
Shet M, Mannion R, O’Donnell E, Park A. Use of a 14–6.
glutaric acid cocrystal to improve oral bioavailability Pudipeddi M, Serajuddin ATM. Trends in solubility of
of a low solubility API. Pharm Res. 2006;23:1888–97. polymorphs. J Pharm Sci. 2005;94:929–39.
Medina C, Daurio D, Nagapudi K, Alvarez-Nunez Rajput L, Sanphui P, Desiraju GR. New solid forms of the
F. Manufacture of pharmaceutical cocrystals using anti-HIV drug etravirine: salts, cocrystals, and solubil-
twin screw extrusion: a solvent-less and scalable pro- ity. Cryst Growth Des. 2013;13:3681–90.
cess. J Pharm Sci. 2010;99:1693–6. Ray E, Vaghasiya K, Sharma A, Shukla R, Khan R,
Miller DA, DiNunzio JC, Yang W, McGinity JW, Kumar A, Verma RK. Autophagy-inducing inhalable
Williams RO III. Targeted intestinal delivery of super- cocrystal formulation of niclosamide-nicotinamide for
saturated itraconazole for improved oral absorption. lung cancer therapy. AAPS Pharm Sci Tech. 2020;21:
Pharm Res. 2008;25:1450–9. 260.
Morissette SL, Soukasene S, Levinson D, Cima MJ, Remenar JF, Morissette SL, Peterson ML, Moulton B,
Almarsson Ö. Elucidation of crystal form diversity of MacPhee JM, Guzmán HR, Almarsson Ö. Crystal
the HIV protease inhibitor ritonavir by high- engineering of novel cocrystals of a triazole drug
throughput crystallization. Proc Natl Acad Sci U S with 1,4-dicarboxylic acids. J Am Chem Soc.
A. 2003;100:2180. 2003;125:8456–7.
Morissette SL, Almarsson Ö, Peterson ML, Remenar JF, Rodríguez-Hornedo N, Nehm SJ, Seefeldt KF, Pagán-
Read MJ, Lemmo AV, Ellis S, Cima MJ, Gardner Torres Y, Falkiewicz CJ. Reaction crystallization of
CR. High-throughput crystallization: polymorphs, pharmaceutical molecular complexes. Mol Pharm.
salts, cocrystals and solvates of pharmaceutical solids. 2006;3:362–7.
Adv Drug Deliv Rev. 2004;56:275–300. Rodríguez-Spong B, Price CP, Jayasankar A, Matzger AJ,
Morris KR, Fakes MG, Thakur AB, Newman AW, Singh Rodríguez-Hornedo N. General principles of pharma-
AK, Venit JJ, Spagnuolo CJ, Serajuddin A. An ceutical solid polymorphism: a supramolecular per-
integrated approach to the selection of optimal salt spective. Adv Drug Deliv Rev. 2004;56:241–74.
form for a new drug candidate. Int J Pharm. Sanphui P, Mishra MK, Ramamurty U, Desiraju
1994;105:209–17. GR. Tuning mechanical properties of pharmaceutical
Murshedkar T. Effect of crystalline to amorphous crystals with multicomponent crystals: voriconazole as
conversions on solubility of cefuroxime axetil. Univer- a case study. Mol Pharm. 2015;12:889–97.
sity of Rhode Island; 2002. Schultheiss N, Newman A. Pharmaceutical cocrystals and
Musumeci D, Hunter CA, Prohens R, Scuderi S, McCabe their physicochemical properties. Crys Growth Des.
JF. Virtual cocrystal screening. Chem Sci. 2011;2: 2009;9:2950–67.
883–90. Serajuddin ATM. Salt formation to improve drug solubil-
O’Connor KM, Corrigan OI. Preparation and ity. Adv Drug Deliv Rev. 2007;59:603–16.
characterisation of a range of diclofenac salts. Int J Shalaev E, Zografi G. The concept of “structure” in amor-
Pharm. 2001;226:163–79. phous solids from the perspective of the pharmaceuti-
Oszczapowicz I, Małafiej E, Szelachowska M, cal sciences. Amorphous Food & Pharmaceut Syst.
Horoszewicz-Małafiej A, Kuklewicz C, Sierańska 2002;281:11–30.
E. Esters of cephalosporins. Part II. Differences in the
Shan N, Zaworotko MJ. The role of cocrystals in pharma- Van Eerdenbrugh B, Raina S, Hsieh Y-L, Augustijns P,
ceutical science. Drug Discov Today. 2008;13:440–6. Taylor LS. Classification of the crystallization behav-
Shattock TR, Arora KK, Vishweshwar P, Zaworotko ior of amorphous active pharmaceutical ingredients in
MJ. Hierarchy of supramolecular synthons: persistent aqueous environments. Pharm Res. 2014;31:969–82.
carboxylic acid-pyridine hydrogen bonds in cocrystals Viscomi G, Campana M, Barbanti M, Grepioni F,
that also contain a hydroxyl moiety. Cryst Growth Des. Polito M, Confortini D. Crystal forms of rifaximin
2008;8:4533–45. and their effect on pharmaceutical properties.
Singhal D, Curatolo W. Drug polymorphism and dosage CrstEngComm. 2008;10:1074–81.
form design – a practical perspective. Adv Drug Deliv Walsh D, Serrano DR, Worku ZA, Madi AM,
Rev. 2004;56:335–47. O’Connell P, Twamley B, Healy AM. Engineering of
Somani JK, Bhushan I, Sen H. Bioavailable oral dosage pharmaceutical cocrystals in an excipient matrix: spray
form of cefuroxime axetil, 2001. US Patent drying versus hot melt extrusion. Int J Pharm.
6,323,193; 2018;551:241–56.
Sun CC, Hou H. Improving mechanical properties of Ware E, Lu DR. An automated approach to salt selection
caffeine and methyl gallate crystals by cocrystal- for new unique trazodone salts. Pharm Res. 2004;21:
lization. Cryst Growth Des. 2008;8:1575–9. 177–84.
Tanaka R, Hattori Y, Otsuka M, Ashizawa K. Application Wells JI. Pharmaceutical preformulation: the physico-
of spray freeze drying to theophylline-oxalic acid chemical properties of drug substances. Chichester:
cocrystal engineering for inhaled dry powder technol- Ellis Horwood; 1988.
ogy. Drug Dev Ind Pharm. 2020;46:179–87. Wyttenbach N, Kuentz M. Glass-forming ability of
Tanaka R, Osotprasit S, Peerapattana J, Ashizawa K, compounds in marketed amorphous drug products.
Hattori Y, Otsuka M. Complete cocrystal formation Eur J Pharm Biopharm. 2017;112:204–8.
during resonant acoustic wet granulation: effect of Xiang T-X, Anderson BD. A molecular dynamics simula-
granulation liquids. Pharmaceutics. 2021;13:56. tion of reactant mobility in an amorphous formulation
Tang J, Han Y, Ali I, Luo H, Nowak A, Li J. Stability and of a peptide in poly(vinylpyrrolidone). J Pharm Sci.
phase transition investigation of olanzapine 2004;93:855–76.
polymorphs. Chem Phys Lett. 2021;767:138384. Yousef MAE, Vangala VR. Pharmaceutical cocrystals:
Testa CG, Prado LD, Costa RN, Costa ML, Linck YG, molecules, crystals, formulations, medicines. Cryst
Monti GA, Cuffini SL, Rocha HVA. Challenging iden- Growth Des. 2019;19:7420–38.
tification of polymorphic mixture: Polymorphs I, II and Yu L. Inferring thermodynamic stability relationship of
III in olanzapine raw materials. Int J Pharm. 2019;556: polymorphs from melting data. J Pharm Sci. 1995;84:
125–35. 966–74.
Tong WQT, Whitesell G. In situ salt screening-a useful Yu L. Amorphous pharmaceutical solids: preparation,
technique for discovery support and preformulation characterization and stabilization. Adv Drug Deliv
studies. Pharm Dev Technol. 1998;3:215–23. Rev. 2001;48:27–42.
Torchilin V. Micellar nanocarriers: pharmaceutical Yu L, Reutzel SM, Stephenson GA. Physical characteriza-
perspectives. Pharm Res. 2007;24:1–16. tion of polymorphic drugs: an integrated characteriza-
Trask AV, Motherwell WDS, Jones W. Solvent-drop tion strategy. Pharmaceut Sci Tech Today. 1998;1:
grinding: green polymorph control of cocrystallisation. 118–27.
Chem Commun. 2004;7:890–1. Zhang GGZ, Henry RF, Borchardt TB, Lou X. Efficient
Trask AV, Motherwell WDS, Jones W. Pharmaceutical cocrystal screening using solution-mediated phase
cocrystallization: engineering a remedy for caffeine transformation. J Pharm Sci. 2007;96:990–5.
hydration. Cryst Growth Des. 2005a;5:1013–21. Zimmerman AW. Treating autism and other developmen-
Trask AV, van de Streek J, Motherwell WDS, Jones tal disorders in children with NMDA receptor
W. Achieving polymorphic and stoichiometric diver- antagonists, 1991. US Patent 4,994,467
sity in cocrystal formation: importance of solid-state Zografi G, Newman A. Interrelationships between struc-
grinding, powder x-ray structure determination, and ture and the properties of amorphous solids of pharma-
seeding. Cryst Growth Des. 2005b;5:2233–41. ceutical interest. J Pharm Sci. 2017;106:5–27.
Mechanical Particle-Size Reduction
Techniques 4
Javier O. Morales, Alan B. Watts, and Jason T. McConville
Abstract homogenization, and microfluidization have

With the increasing number of new drug been developed. More recently, combination
candidates, the number of those that have technologies of both top-down and bottom-up
poor aqueous solubility is also on the rise. To approaches have gained much interest. All
overcome this limitation, a common formula- these different strategies are reviewed and
tion approach has been to decrease drug parti- discussed in this updated chapter. This current
cle size. This strategy results in increased third edition chapter is a revision and update of
surface area, the potential to increase satura- the original authors’ work form the first and
tion solubility, and decreased diffusional dis- second editions.
tance, all of which lead to an increase in the
extent and the rate of dissolution. Mechanical Keywords
techniques to decrease the particle size of Dry milling · Wet milling · Nanocrystal · High-
solids are generally classified in three pressure homogenization · Cryomilling ·
categories: dry milling, wet milling, and Nanoedge technology
high-pressure homogenization. In order to pro-
duce particles in the submicron (nano) range
and further increase solubility, techniques such 4.1 Introduction
as wet-media milling, piston-gap
The percentage of newly discovered drugs that
have poor water solubility has been trending
J. O. Morales (*)
Department of Pharmaceutical Science and Technology, upward, and products with market approval in
School of Chemical and Pharmaceutical Sciences, the pharmaceutical industry include approxi-
University of Chile, Santiago, Chile mately 40% poorly water-soluble compounds
Advanced Center for Chronic Diseases (ACCDiS), and nearly 90% of new drugs in the discovery
Santiago, Chile pipeline (Kalepu and Nekkanti 2015; Lipinski
Center of New Drugs for Hypertension (CENDHY), 2000). Of the total of drugs in the pipeline, up to
Santiago, Chile 60% are derived directly from synthesis (Gribbon
e-mail: [email protected] and Andreas 2005; Lipinski et al. 1997). Aside
A. B. Watts from the most notable limitation, i.e., poor bio-
Catalent Pharma Solutions, Somerset, NJ, USA availability, these compounds are hard to formu-
J. T. McConville late due to a number of other factors such as fed
Department of Pharmaceutical Sciences, University of versus fasted bioavailability variation, lack of
New Mexico, Albuquerque, NM, USA
142 J. O. Morales et al.
dose–response proportionality, suboptimal dos- mechanisms that will be explained in this chapter.
ing, use of problematic excipients (such as An increase in solubility is one of the reasons for
cosolvents), use of extreme basic or acid reducing the particle size of pharmaceutical
conditions to optimize solubilization, uncontrol- powders; and particle-size reduction is applied
lable precipitation after dosing, and inconve- in several fields within the pharmaceutical
nience of the dosage form (Merisko-Liversidge sciences, such as pulmonary drug delivery, solid
and Liversidge 2008). Over the years, a number oral dosages, and powder handling.
of different strategies have been developed in It is known in the literature that pulmonary
order to overcome these limitations. Some of drug delivery by dry powder inhalation is an
these efforts include creating salts, prodrugs, and administration route where particle-size reduction
screening for more soluble analogs. However, is required to reach the target region of the lung.
these efforts have achieved limited success, and The aerodynamic diameter of a particle should be
the need of novel strategies has been the driving in the critical range of 0.5–5 μm (Clark and Shire
force for the development of technologies to 2000) to reach appropriate regions of the deep
improve the bioavailability-related outcomes of lung for maximum absorption (Hassan and Lau
poorly water-soluble drug molecules. Extensive 2010). This size range can be achieved by many
research into the development of novel of the various milling technologies (Chow et al.
techniques to engineer micron-scale particles to 2007).
the nanoscale range has translated into various Additionally, many pharmaceutical processes
proprietary technologies. Technologies that have involving powders can be improved by
been commercialized to decrease drug particle homogenizing the particle size of the drug and
size into the nanosize range for marketed products excipients. Solids produced by uncontrolled crys-
include the NanoCrystal® technology developed tallization or precipitation processes can have a
by Elan Drug Technologies (now Alkermes), broad size distribution that can result in poor flow
which is a high-energy media milling technique properties or tendency to segregate. Blending,
(Elan Drug Technologies—Commercialized compressibility, flow/suspension behavior, and
Products 2010), and the Insoluble Drug Delivery compaction performance can all vary with a het-
Microparticle technology (IDD-P®) by erogeneous particle-size distribution (Fisher
SkyePharma (SkyePharma—Insoluble Drug 2006). Furthermore, the particle-size distribution
Delivery Platform 2010). Other marketed of very potent drugs, requiring low content in the
products have been developed using media mill- final dosage form, is usually homogenized before
ing and high-pressure homogenization (Fontana blending with excipients in order to provide ade-
et al. 2018). quate content uniformity (Clement and Purutyan
Nonspecific formulation approaches are appli- 2002; Rohrs et al. 2006).
cable to almost any drug molecule (apart from a Bulk active pharmaceutical ingredient (API)
few exceptions). Micronization has been for production may often be limited by specified
many years the most widely used nonspecific particle-size profiles to meet the need of formula-
approach, which consists of converting relatively tion and pharmaceutical processes. The ability to
coarse drug particles to micrometer-sized control the particle size during early stage API
particles with a mean diameter in the range of development and to predict the operating
approximately 2–5 μm and a corresponding size conditions that will produce that particle size
distribution approximately between 0.1 and reproducibly at commercial scale is very impor-
20 μm (Muller et al. 2006). Nonetheless, many tant. Furthermore, a final milling process may be
novel drugs are so poorly soluble that a further required in order to narrow the particle-size dis-
decrease in particle size is often required in order tribution across batches so that better flow and
to obtain acceptable solubility. Particle-size handling properties are produced, or match an
reduction techniques increase the surface area, API particle size more closely with the combined
leading to an increase in the solubility by
4 Mechanical Particle-Size Reduction Techniques 143
excipient(s) particle size to minimize the potential of the dissolving solid, C the concentration of
for segregation during blending (Fisher 2006). solute in solution, and Cs the saturation solubility.
This translates into three main factors that can
affect the dissolution rate: surface area, saturation
4.2 Rationale Behind solubility, and diffusional distance.
the Reduction of Particle Size The Kelvin equation describes the increase in
vapor pressure as a function of the curvature of
As stated earlier, a decrease in particle size to the liquid droplets, which ultimately translates into an
few micron range and down to the nanosize range increase in the saturation solubility as seen in
can potentially increase the extent and rate of (4.2) (Simonelli et al. 1970):
solubility of drugs. This is of high relevance for
Pr 2γV m
poorly water-soluble compounds since this fea- n ¼ , ð4:2Þ
P1 rRT
ture is their main limitation. Decreasing the parti-
cle size to the micron range substantially Where Pr is the vapor/dissolution pressure
increases the exposed surface area of a given with particle radius r; P1 is the vapor/dissolution
amount of powder. The micronized powder can pressure with infinite particle size; γ is the inter-
then be further engineered, and its particle size facial tension; Vm is the molar volume; R is the
further decreased to the nanosize range, when the gas constant; T is the absolute temperature; and
surface area increases sharply as can be depicted r is the particle radius. When the particles are
in Fig. 4.1 (Merisko-Liversidge and Liversidge suspended in a saturated solution, it can be
2008). assumed that the Pr/P1 ratio can be approximated
This has a direct effect on the dissolution rate to the ratio of the respective activities of small and
according to the Noyes–Whitney equation as large particles. Furthermore, if the activity
depicted in (4.1) (Noyes and Whitney 1897): coefficients of both particles are equal, the
activities can be replaced by their respective
dm DAs ðC s C Þ
¼ ¼ kAs ðCs C Þ, ð4:1Þ solubilities. Similar to the sharp increase in sur-
dt hH
face area observed for very small particle size,
Where m is the undissolved solid particle mass reducing the particle size below the size threshold
of solids, t the time, hH the diffusion boundary of 1–2 μm leads to a distinct increase in the
layer thickness, D the diffusion coefficient, k the dissolution pressure, thus shifting the solubility
intrinsic dissolution constant, As the surface area equilibrium toward an increased saturation
Fig. 4.1 The plot

demonstrates the increase in
surface area obtained when
solids are fractured from the
micrometer-size range to
the nanometer-size range
(Merisko-Liversidge and
Liversidge 2008)
solubility. This is expressed in the Ostwald– the flat surface. In addition to the above state-
Freundlich (4.3) (Simonelli et al. 1970): ment, this can be extracted from the Prandtl equa-
C s,r 2γV m tion, where individual particles will have a very
ln ¼ , ð4:3Þ small surface exposed in the direction of the flow,
C s,1 rRT
which will ultimately decrease the distance that
Where Cs,r and Cs,1 are the solubilities of a the molecules need to diffuse going into the bulk
particle of radius r and of a very large particle of the solution.
(or an approximately flat surface with very low The equations presented here are clear evi-
dissolution pressure), respectively. dence that reducing the particle size results in
An additional factor that enhances the dissolu- larger surface area, elevated saturation solubility,
tion velocity is a decrease in the diffusion dis- a decrease in the diffusional distance for drug
tance at very small particle size as is described in molecules, and therefore a faster rate of dissolu-
the Prandtl boundary layer (4.4) (Bisrat and tion as summarized in Fig. 4.2.
Nyström 1988), which translates into an increase Additionally, high-energy processes, such as
in the concentration gradient (Cs C): those used in top-down methods for particle-size
1=2 reduction, can be associated with transitions from
L
hH ¼ k 1=2 , ð4:4Þ the crystalline state to an amorphous state (Muller
V et al. 2003). Transformations to the amorphous
Where hH is the diffusion boundary layer state or to different polymorphs have also been
thickness, L is the length of the surface in the observed during high-energy input processes,
direction of flow, k denotes a constant, and v is such as tableting (Chan and Doelker 1985;
the relative velocity of the flowing liquid against Koivisto et al. 2006; Zhang et al. 2004). High-
Fig. 4.2 Changes in properties when decreasing the par- saturation solubility. Bottom figure: Decreasing the parti-
ticle size from the micro- to the nanorange. Top insert: cle size results in an increase in surface area, being pro-
Decreasing the particle size increases the particle size and nounced below 1 μm and very pronounced below 100 nm
therefore increases the dissolution pressure, increasing the (Shegokar and Müller 2010)
pressure homogenization is a commonly used drugs. These techniques can be divided into two
top-down process to obtain micro- and main groups: milling and high-pressure
nanoparticles and is one such high-energy pro- homogenization.
cess; the drug particles are exposed to a power
density of up to 1013 W/m3 (Muller et al. 2003).
This high energy breaks down the drug particles 4.3 Milling
into microparticles and into nanoparticles, and
can also induce the change to an increased amor- Milling processes can be divided into dry or wet
phous fraction or to completely amorphous milling depending on the media in which the
particles (Böhm et al. 1998; Dong and Feng powders are milled, namely gas or liquid. In
2007; Jacobs et al. 2000; Ward and Schultz both cases, particle-size reduction occurs by col-
1995) (Fig. 4.3). It is known in the literature that lision of particles with the surfaces of the equip-
the amorphous state yields a higher saturation ment as well as with each other. The collision
solubility than that exhibited by the crystalline events involve compression, impact and attrition,
structure of the same drug (Agrawal et al. 2004; and cutting or shear as the main mechanisms for
Hancock and Parks 2000; Murdande et al. 2010). particle-size reduction (Clement and Purutyan
Considering the solubility enhancement inherent 2002; Friedrich 2001; Spencer and Dalder
in the decrease in particle size discussed earlier, 1997). Additionally, the wet-milling process
amorphous nanoparticles may exhibit very high also involves liquid shear forces or cavitation,
saturation solubility compared to the crystalline ultimately resulting in particle-size reduction
form prior to processing. (Rabinow 2004; Sharma et al. 2009) (Table 4.1).
Furthermore, decreasing the particle size of The density of solids in the mill is an important
drugs can increase the stability of formulations. parameter directly affecting breakage
An example of this can be found in mechanisms. Density has a direct relationship
nanosuspension formulations of paclitaxel. Pacli- with the number of particle–particle or particle–
taxel is a highly water-sensitive molecule that wall collisions as well as the force of the
when dissolved in water degrades to an extent of collisions (Bentham et al. 2004). Feed rate to the
80% within 25 minutes (Liversidge and Wei mill and mill residence time also directly affect
2003). It has been found that the production of the solid particle concentration of solids in the
an aqueous nanosuspension, with the use of sur- mill. Thus, an understanding of the particle den-
face stabilizers, can increase the stability over a sity in the mill is crucial since it can impact the
period of 4 years when stored at 4 C, i.e., more milling rate and efficiency, both of which are
than 99% of the drug remains intact after recovery important for scale-up purposes. For example,
(Troester 2004). research in Thailand demonstrated that by
The following sections will discuss the theory increasing concentrations from 30 to 55% solids
and applications of various top-down techniques (by volume), the mill net power increased with
used in the field for decreasing the particle size of
Fig. 4.3 Schematic

representation of a
crystalline surface before
and after mechanical
comminution. After Micronization
milling, the surface that has
been exposed to collisions
is disrupted and may exhibit
amorphous domains (Ward
and Schultz 1995)
Crystalline Solid Surface Partially Amorphous
Solid Surface
Table 4.1 Comparison of dry and wet milling

Dry milling Wet milling
Medium Air Liquid
Medium property Compressible Noncompressible
Force/energy transfer Particle–particle collision Particle–particle collision
Particle–equipment Particle–equipment collision, liquid shear forces, and
collision cavitation
Inertial dampening of Weak to none Strong
forces
increased powder filling, and then decreased after instead of surface area together with distribution
an optimum value was achieved since they are so closely related (Clement and
(Tangsathitkulchai 2003). Purutyan 2002).
Due to the nature of the milling process, the Even though understanding the equipment
milled powders produced always exhibit a range capabilities and the mechanism of particle-size
of particle sizes or a characteristic particle-size reduction is critical in selecting the appropriate
distribution. In many cases, this range of particle mill, specific powder properties such as hardness,
sizes obeys a log normal distribution; however, friability, and fracture toughness will affect the
some processed powders can exhibit multimodal milling performance in both dry- and wet-milling
distributions. Therefore, when the goal of milling equipment alike (Chaumeil 1998; Kesisoglou
is to meet a specific unimodal particle-size range, et al. 2007). It is common for milling equipment
a classification step may be needed. Several mills to be described in terms of the particle sizes that
available in the market can perform both can be achieved, and for this purpose, the Mohs
operations (particle-size reduction and classifica- scale is used as an indicator of particle hardness
tion) in sequential steps in one instrument (Fisher and for estimating the mill performance (Clement
2006; Steckel et al. 2006). The influence of the and Purutyan 2002). This scale ranges from a
feed particle size can be diminished when the hardness of 1 for soft materials to a value of
milling equipment includes internal classification 10 for the hardest material (diamond). Addition-
or if a separate classifier is part of the process line. ally, particle morphology or aspect ratio can
Additionally, the solid particle feed rate and the affect milling results in various ways. For exam-
intrinsic density of the particles play a fundamen- ple, it has been shown that β-succinic acid crystals
tal role in the intensity of the interparticle with plate-like morphology are more prone to
collisions; therefore, controlling both factors will crystallinity loss on milling compared to those
determine particle-size distributions and repro- with needle-like morphology (Chikhalia et al.
ducibility in a given milling process. 2006). Therefore, a thorough understanding of
Another parameter that is relevant to consider the properties of the solid, together with the spe-
is the target and achieved surface area during the cific mechanism of particle-size reduction of the
milling process. As stated earlier, decreasing the milling equipment, is of high importance.
particle size of a solid increases the overall sur- One of the most relevant limitations of milling
face area exposed, affecting dissolution perfor- is to actually control and narrow the particle-size
mance. Even so, since the value of surface area distribution; this is normally optimized by
is a scalar quantity representing the totality of the controlling for several parameters in the process.
processed powder, it does not provide a sense of In general, capital costs for installation are high,
the range of particle sizes. Therefore, in the devel- and the operation of the equipment can be labor
opment of controlling parameters during the mill- intensive (Fisher 2006). Additionally, while mill-
ing process, determining the particle-size ing can allow for a reduction of the aspect ratio, it
distribution is crucial for the performance of the is not normally able to control the particle shape.
engineered powder. It is normally recommended Furthermore, some active pharmaceutical
to control for particle-size distribution alone, ingredients may not be able to be processed
through milling due to either compaction sensi- electrostatic agglomeration of milled particles
tivity (losing crystallinity), temperature sensitiv- that may hinder the API dispersion during formu-
ity (melting or changing to a different lation (de Villiers 1995; de Villiers and Tiedt
polymorph), or changing hydration state 1996). Furthermore, the moving parts of some
(Peltonen and Hirvonen 2010; Shoyele and dry mills can generate considerable frictional
Cawthorne 2006). Therefore, depending on the heat. For example, during powder processing in
API characteristics, milling may not always be a pin mill, the internal temperatures can reach
the best choice to reduce the particle size of 40–60 C (Fisher 2006), which can impose a
powders. limitation to certain pharmaceuticals. For these
It is relevant to note that wet mills such as types of powders, cryomilling could be an option
rotor–stator media mills (Netzsch Pumps North and is discussed in further detail later in this
America, LLC, Exton, PA) or dry mills such as jet chapter.
mills (Retsch GmbH, Haan, Germany) are limited In the scale-up process, the flow properties of
in the production of solid particles in the nanome- the dry powder being deposited into the mill
ter scale (Muller et al. 2006). For example, it has could affect uniformity of delivery. Furthermore,
been found that usual range of particles in a jet if the milled or partially milled product is very
mill can be from 0.1 to 20 μm, with only a 10% in cohesive (either inherently or due to electrostatic
the submicron range (Muller et al. 1995). On the charge), it could accumulate, and the overall pro-
other hand, nanoparticles have been obtained by cess yield could be reduced. Removing the prod-
using ball media mills for extended periods of uct periodically from low flow areas (cyclones,
time (Liversidge et al. 1992; Merisko-Liversidge pipe elbows, and bends) can increase the cost and
et al. 1996; Merisko-Liversidge et al. 2003). The extend the processing time. If material were to
following sections focus on several dry- and accumulate in the mill, especially in places that
wet-milling techniques, describing important could block the exit of the mill, this could result in
parameters, particle-size ranges, and applications equipment overheating and eventual failure.
of each technique. Manufacturers of mills normally report the
correlation between particle hardness (on the
Mohs scale) and the extent of size reduction that
4.3.1 Dry Milling could be achieved in a particular mill. A general
lower limit can be described for mills, depending
As mentioned above, particle-size reduction in on the mechanism of action and the amount of
dry mills occurs by pressure, friction, attrition, energy that it can provide for grinding. The fol-
impact, or shearing by particle–particle or lowing sections describe the dry-milling equip-
particle–equipment interactions. Even though a ment used for decreasing the particle size of
variety of mills are available for decreasing the pharmaceutical powders (Clement and Purutyan
particle size, some common issues can be 2002; Fisher 2006; Friedrich 2001) (Table 4.2).
identified. In general, issues related to the While most dry-milling techniques can render
dry-milling process may include powder accumu- particles in micrometer-size range, recent
lation that can impact performance and investigations have developed formulations that
Table 4.2 Types of dry-milling equipment categorized by average particle size achievable and hardness of materials it
can process
Mill type Typical minimum particle size achieved (μm) Maximum hardness (on the Mohs scale)
Cutting mill 150 Soft
Pin/cage mill 10–50 Soft, up to 3
Hammer mill 10–75 Intermediate, up to 6
Jet mill 2 Soft, up to 3
Fluidized bed jet mill 2 Hard, up to 10
Ball mill <1 Hard, up to 8
can achieve drug nanoparticles of lamotrigine by gas exiting the mill (Godet-Morand et al. 2002; de
co-milling with polyvinyl alcohol (Ambrus et al. Vegt et al. 2009). Varying the classifier speed is
2020; Gieszinger et al. 2018). Using a quality by the normal technique used to control the particle-
design (QbD) approach, and after definition of the size distribution of the milled material; however,
design space, selected conditions yielded mean parameters such as grinding gas pressure and total
particle sizes as low as 97 nm (Gieszinger et al. gas flow rate can be modified to achieve the
2018). desired particle-size distribution and processing
times. Changing nozzle diameter is the most com-
Fluidized Bed Jet Milling mon approach for control of these variables. It is
The primary mechanism of particle-size reduction important to note that when the parameters
in the impinging jet or fluidized bed jet mill is described here are fixed, the solid particle feed
particle–particle collisions. When the media to be rate does not normally affect the particle-size
milled is fed to the chamber, it is exposed to distribution, but it does affect the residence time
impinging high-velocity gas jets that allow for and processing times.
the collisions (Fig. 4.4). This type of milling Results will vary depending on the
equipment can process hard materials (Mohs conditions used in the equipment and the
scale hardness of up to 10), which is a prominent properties of the solid (Rasenack and Müller
advantage compared to other air jet mills, such as 2004). However, for typical working pressures
spiral mills (Mohs scale hardness of up to 3.5) between 3 and 10 bar, particle sizes are usually
(Clement and Purutyan 2002). tightly distributed due to the classifier and nor-
Additionally, impinging jet mills feature a sep- mally range between 1 and 10 μm. Typical
arate classifier, e.g., rotating wheel, preventing surface areas achieved range from 2 to 5 m2/g
unmilled solids from exiting the milling chamber and are strictly dependent on the API hardness
until particles have reached a certain particle-size and friability (Fisher 2006).
threshold (Fig. 4.4.). In the case of a rotating One of the main limitations of this type of mill
wheel classifier, the particle size depends on the is the potential for buildup of compressed product
rotation speed of the wheel and the velocity of the in the mill or classifier due to the particle–wall
Fig. 4.4 Scheme of a

fluidized bed air jet mill
(Godet-Morand et al. 2002)
Fig. 4.5 Specific surface created as a function of the factor P/N2 for the fluidized air jet mill. (From Nakach et al. 2004)
and particle–classifier collisions. Accumulation recycling particles back into the mill, there is an
of product in either of these compartments will increase in the mill holdup. The online sizing
change the geometry of the mill and will alter the system evidenced that D50 and D90 varied consid-
performance and reproducibility of the equip- erably during operation, but D10 was more stable
ment. In general, buildup at the exit of the mill (Nakach et al. 2004).
or in the classifier will increase the particle size
over time owing to a reduced performance of the
Spiral Jet “Pancake” Mill
classification system. The material accumulated
Spiral jet or pancake mills have jets of air tangen-
in the mill can become compressed or discolored
tially located around the center of the equipment
or may change into an amorphous form, all of
in order to create a vortex of air. Particle-size
which will affect the product quality if it enters
reduction is a result of particle–particle and
the milled batch stream. Even if intermittent
particle–wall collisions upon feeding of solids
cleaning is introduced in the process to avoid
into the air stream (Fig. 4.6). Normally, gas
these problems, the impact in the overall produc-
velocities are such that a sonic flow is achieved,
tivity could be important.
and only particles that reach a predetermined
Nakach et al. (2004) studied the effect of the
particle size will be able to leave the vortex
air pressure at the grinding nozzles, the solid
through the exit. This internal orientation is
particle feed rate, and the speed of rotation of
advantageous due to the classification that natu-
the turbo selector. In their studies, they found a
rally occurs in the system. Due to centrifugal
dependency of the specific surface of the product
forces, larger particles tend to remain near the
with the air pressure (P) and the speed of rotation
perimeter of the milling chamber until the particle
of the classifier (N ) (Nakach et al. 2004). The
size is reduced, and they are light enough to travel
equation below indicates the relationship between
to the center of the mill due to centripetal forces.
air pressure and the classifier rotational speed,
Particle-size reduction will depend on two
with the forces present in a dry mill.
main variables, namely the equipment and the
P dragforce solid to be milled. Geometric parameters of the
/ : ð4:5Þ
N 2 centrifugalforce equipment such as shape and diameter of the
grinding chamber and shape/type, number, and
It was found that 1–10 kg/h solid particle feed angle of grinding nozzles will determine the per-
rate had little influence on the specific surface of formance of the milling process. Additionally,
the product (Fig. 4.5). That particle size is mostly
controlled by the classifier. Furthermore, by
FEED FEED
GRINDING FINE PARTICLES

NOZZLES
GAS JET
INJECTOR
GAS
GRINDING
GRINDING GAS
GAS
PRODUCT
MILL
CHAMBER
PARTICLES
CUTLET
Fig. 4.6 Above: scheme of a spiral jet “pancake” mill (Midoux et al. 1999). Below: scheme of a spiral jet mill with an
additional classifier online (New Food Pharma Systems Spiral Jet Mill: Introduction 2010)
operating variables such as grinding jet air pres- Finally, by combining the above two
sure, total gas flow rate, and solid particle feed relationships, a third relationship can be defined
rate play a role in the final particle-size distribu- in terms of the solid particle feed rate and the mill
tion achieved (Friedrich 2001; Hoyer et al. 2008; chamber diameter:
Midoux et al. 1999; Schlocker et al. 2006).
Q / D2:80:2 : ð4:8Þ
Geometry Dependence According to a survey using the
Midoux et al. (1999) have described the relation- manufacturer’s product information, the above-
ship between the volumetric flow rate Vn, the derived relationship obtained is in accordance
solid feed rate Q, and the diameter of the mill with what had been described before, namely
chamber D as follows: Q is proportional to D2.3 0.3, as well as to
Smit’s work on waxes describing a
V / D2 : ð4:6Þ
Q proportional to D2.5 0.2 (Smit 1986). Further-
Moreover, the volumetric flow rate and the more, the exponent seems to depend on the
feed rate can be correlated as (Ito 1987): properties of the material being ground (Midoux
et al. 1999).
Q / V 1:40:1
n : ð4:7Þ
Nozzle Dependence with respect to the tangent, Smit’s optimum value

The most common nozzle geometry is the abrupt was equal to 58 (Smit 1986) and Skelton’s was
type, providing sonic velocity at the inlet with an between 52 and 60 (Skelton et al. 1980). None-
exit pressure of about 50% of the initial fluid theless, other operating angles ranging from 63
pressure (Fig. 4.7). A suction that entraps to 67 have been described in the literature
particles from the mill is created after the final (Midoux et al. 1999).
expansion occurs beyond the nozzle inlet. This
phenomenon circulates the gas and induces
Working Conditions
particle–particle collisions. In the Laval-shaped
In addition to the geometry of the chamber and
nozzle, air expansion occurs at the divergent sec-
the nozzles, operating variables can be controlled
tion. This leads to supersonic velocities, increas-
to obtain different outcomes. One of these
ing the air jet action and the velocity of the
variables is grinding pressure, which controls
circulating gas, therefore producing greater
the gas mass flow rate input. As described by
particle–particle collisions and thus increasing
Midoux et al. (1999), if one assumes that the
the production rate and reducing the average par-
nozzles are isentropic, the initial grinding pres-
ticle size (Midoux et al. 1999).
sure P and the pressure at the nozzle throat Pt are
Additionally, regardless of the type of nozzle,
related according to (4.9):
the number of nozzles is an important factor to
consider in the performance of a jet mill. The k
P k 1 2 k1
influence of three different configurations, 3, 6, ¼ 1þ Mt ð4:9Þ
Pt 2
and 12 nozzles, while maintaining the total sec-
tion of the nozzles and varying the solid’s feed Where Mt is the Mach number at the throat,
rates, has been studied (Skelton et al. 1980). It and k is the ratio of specific heats of the gas. With
was found that a greater number of nozzles cre- this, a critical grinding pressure Pc can be defined
ated a more regular pitch circle; furthermore, as follows in (4.10):
thinner jet nozzles led to minor perturbations of
k
the spiral flow in the chamber. Accordingly, the Pc k þ 1 k‐1
¼ : ð4:10Þ
design including 12 nozzles exhibited the best Pt 2
grinding ratio. Moreover, increasing the feed
rates was also found to improve the grinding This critical grinding pressure is equivalent to
ratio. the minimum pressure that yields a sonic flow at
The angle at which the nozzles are oriented the nozzle inlet. Above this value, the gas mass
toward the inner chamber of mill can be flow rate Mg can be expressed by (4.11):
modulated, and their optimal values have been sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
kþ1
reported in the literature. If the angle is measured Mwk 2 k1
M g ¼ PA : ð4:11Þ
RT k þ 1
As can be derived from the (4.11), Mg is

directly proportional to the grinding pressure P,
the throat section A, and the molecular weight Mw
of the fluid employed for grinding. The gas
kinetic energy (Ėk) can be defined by (4.12):
1
E k ¼ M g v2g : ð4:12Þ
2
The concept of specific energy consumption
Fig. 4.7 The (1) abrupt nozzle and (2) Laval-shaped
nozzle. In the scheme, the flow is from left to right
(Esp) has been used in the literature (Kaiser and
(Albus 1964) Nied 1980; Schurr and Zhao 1994; Stairmand
1975) to correlate solid’s feed rate and grinding lower feed rates will result in smaller particles
pressure, which is directly correlated to the gas (Midoux et al. 1999). The literature describes
kinetic energy (defined above). According to that an optimal feed rate can be achieved to
Schurr and Zhao, this is calculated by (4.13) yield the tightest particle-size distributions at a
(Schurr and Zhao 1994): set average particle size (Midoux et al. 1999). In
general, these spiral mills are suitable for rather
E_ k
E sp ¼ : ð4:13Þ soft materials with a Mohs scale hardness of up to
Q 3.5 (Chamayou and Dodds 2007). Nonetheless,
With this value, different mills and different they are widely used in the industry due to two
working conditions can be compared on the same main advantages: (1) there are no moving parts in
system. In a spiral jet mill, the specific surface the equipment, and (2) the Joule–Thomson effect
area (SSA) of the product is related to Esp by a produced by the expansion of the gas passing
power function as follows in (4.14): through the nozzles results in a cooling effect
within the mill. This effect can help to control or
SSA / E xsp : ð4:14Þ suppress any local temperature increases that may
be due to friction consequential of the milling
In their studies, Midoux et al. (1999) process itself (Fisher 2006).
corroborated that the SSA was dependent on the
grinding pressure and the solid’s feed rate and
could compare different systems. It was found Pin Mill
that for the same material, the specific energy A pin mill is a mechanical energy impact mill. Of
required to obtain a certain grinding ratio all dry mills used without a particle classifier, the
diminishes with the mill diameter. Additionally, pin mill achieves the smallest average particle
a different material exhibited a clear Esp transition sizes (Nied 2007). In the pin mill, rotating
around 400 kJ/kg. Below this value, SSA was elements in the equipment allow for particle–par-
proportional to Esp with an exponent of 1.1; how- ticle and particle–mill collisions. A limited inter-
ever, above the critical value of Esp, the exponent nal classification can be achieved if appropriate
decreases sharply as can be seen in Fig. 4.8. elements are selected as milling tools.
In general, high grinding jet pressures result in The milling equipment comprises two disks
smaller particle-size distributions. The solid par- fitted with overlapping pins as depicted on
ticle feed rate is strictly related to the solid particle Fig. 4.9. The pin mill is a type of rotor–stator
concentration in the mill, and thus, higher feed mill; therefore, one of the disks is the stator and
rates will produce coarser particle sizes, while the other one rotates with a high peripheral speed
of up to 150 m/s. An additional modification of
Fig. 4.8 Correlation 10

between Esp versus ΔSSA
DSSA (m2/g)
nitrogen
helium
0,1
10,0 100,0 1000,0 10000,0
Esp (kJ/kg)
the design consists of having two counterrotating tip speed, solid particle feed rate, and gas flow
pin disks, which allow for peripheral speed of up rate through the mill. By optimizing the process,
to 250 m/s (Nied 2007). The solids are fed at a small average particle sizes can be achieved when
controlled rate into the center of the stator by rotor tip speed is maximized, and both solid par-
means of a screw. They are crushed through ticle feed rate and airflow rate are minimized
intermeshing rings of the rotor and stator pins (Muller and Polke 1999). This is illustrated in
and the milled product leaves by centrifugal Fig. 4.10, where it can be seen that D90 decreases
forces to the periphery to be collected or further with increasing tip speed but increases when the
processed. solid particle feed rate is increased. The rotor tip
Besides the physical properties of the solid speed is an adequate way to achieve comparable
material itself, the final average particle size of results for scaling-up purposes (Fisher 2006).
the milled product is determined mainly by rotor An investigation by Nakach et al. (2004) stud-
ied the effect of variables such as the tip speed
(or rotation speed) and the solid particle feed rate.
Additionally, the effect of the type of pin mill was
studied in terms of manufacturer and whether the
equipment was a single- or double-rotor pin mill
(Nakach et al. 2004). As can be seen in Fig. 4.11,
the SSA increases linearly with the square of the
peripheral speed up to 150 m/s for the different
sizes of mills studied (expressed as diameter in
mm).
Further studies with a pin mill attached to a
dynamic selector found that the product quality
was essentially dependent on the performance of
the selector (Fig. 4.12). The selector had 12 blades
and rotated up to 5000 rpm. After the solid was
milled in the pin mill, only fine particles are able
to enter the selector due to the centrifugal force
repulsing the pass of coarser particles, which are
Fig. 4.9 Schematic representation of a pin mill in which brought back to the milling chamber until a suit-
the rotor and the stator are pin disks (Nied 2007) able particle size is achieved. This mode of
Fig. 4.10 Influence of tip Particle size x90

speed and solids feed rate
on reducing the particle µm 20 kg/h
size, expressed as D90 in μm 40 kg/h
(Muller et al. 1999) 200
70 kg/h
150 kg/h
100
0
0 50 100 m/s 150
Rotor tipspeed Vu
Fig. 4.11 Specific surface 0,8

as a function of the square 0,7
of the peripheral speed for
0,6
three different sizes
(expressed as diameter in 0,5
DSsp 0,4
mm) and types of mills
(m2/g)
(Nakach et al. 2004) 0,3
0,2
0,1
0
0 10000 20000 30000 40000 50000 60000
(tip speed)2 (m2s–2)
FORPLEX simple Alpine Simple rotor
rotor 100 mm 160 mm
Alpine Double rotor
250 mm
Fig. 4.12 Relation 10,00

between the specific surface
area and the factor P/N2 in a
pin mill with a dynamic
selector (Nakach et al. 1,00
2004) 1,00E-06 1,00E-05 1,00E-04 1,00E-03
DSsp
(m2/g)
0,10
0,01
2
(Gas flow rate/(Speed selector) (m3h–1rpm–2)
operation is similar to what has been described for temperatures generated in the milling chamber,
fluidized air jet mills, and SSA is dependent on phase transitions may occur and amorphous
the air pressure P and the speed of rotation of the material could be formed during the operation
selector N (see (4.5)). (Fisher 2006). This phenomenon has direct
One of the main concerns with the use of pin consequences for stability of the product and
mills (and any impact mill) is the generation of can affect the performance reproducibility from
heat due to friction (am Ende and Rose 2006). batch to batch.
This is of particular importance for thermally Depending on the properties of the solid, the
labile drugs and materials, with relatively low average particle sizes and the surface areas
phase-transition temperatures (<80–100 C). obtained in impact mills can range from 5 to
Inadequate training and subsequent poor opera- 20 μm and from 1 to 2 m2/g, respectively (Fisher
tion of the mill can result in increasing the tem- 2006).
perature beyond normal ranges. For example, if
the feed rate is higher than the discharge rate, the
Environmental Limitations of Dry Milling
mill can be “choke fed,” such that there is an
Even though the particle-size reduction
accumulation of solids in the mill chamber,
mechanisms involved in the previously described
resulting in heat buildup. As stated above, with
mills may differ, in general, during the process of
other types of milling equipment, due to the high
dry milling, a large amount of small size particles
Table 4.3 Means of preventing and mitigating dust explosions (Eckhoff 2003)
Prevention
Preventing explosive dust
Preventing ignition sources clouds Mitigation
Smoldering combustion in dust, dust flames Inerting by N2, CO2, and rare Partial inerting by inert gas
gases
Other types of open flames Intrinsic inerting Isolation (sectioning)
Hot surfaces Inerting by adding inert dust Venting
Electric sparks and arcs, electrostatic Dust concentration outside Pressure-resistant construction
discharge explosive range
Heat from mechanical impact (metal sparks Automatic suppression
and hot spots) Good housekeeping (dust removal
and cleaning)
are produced, also known as fines, and they need development of explosive mixtures (dust clouds),
to be minimized in order to prevent operator preventing the occurrence of ignition sources, and
exposure and reduce the environmental impact. strategies for mitigation (Table 4.3).
If recovery of the fines is needed in order to A common source of ignition in the industry is
combine the fraction with the batch bulk material, the presence of hot surfaces. Contrary to what has
a pharmaceutical grade dust collector could be been accepted in the past, minimum hot-surface
used (Boundy et al. 2006). ignition temperatures of dust clouds vary signifi-
Additionally, most milled APIs produce a cantly when scaling up. To better estimate the
potentially explosive dust (Fisher 2006; value when scaling up, both the magnitude and
Hamelmann and Schmidt 2003). Certified the geometry of the hot surface in relation to the
laboratories perform studies of minimum ignition dust cloud should be considered (Eckhoff 2005).
energy (MIE), which is the least amount of Other common sources of ignition are the
energy that is required to ignite a dust sample. electric and electrostatic discharges. Switches,
For the purpose of a safe estimation, the concen- failures in electric circuits, and common dis-
tration of dust used is that which will allow for the charge of static electricity are some examples of
minimum energy input for ignition. An MIE these sources. Parameters from the dust cloud that
lower than 20 mJ indicates that low energy can contribute to the ignition are the particle-size/
sources can potentially ignite the dust sample shape distributions, dust moisture content, dust
(Stevenson 2001). For example, static electricity concentration, and the dynamic state of the dust
discharge could be one of the low-energy sources cloud with respect to the spark gap (Eckhoff
that could potentially create the explosion 1994).
(Eckhoff 2003). In cases like these, a thorough To prevent explosive dust clouds, an inert gas,
and careful evaluation of the processing environ- such as nitrogen or carbon dioxide, can be mixed
ment is needed in order to prevent any conditions in the dust to a level where the dust can no longer
that could yield sufficient energy to create a dust ignite in the operating conditions. In contrast with
explosion. In data from Merck and Co., Inc., what has been previously reported in the literature
regarding MIE values (in mJ) for dust samples (Wilén et al. 1999), it has been found that the
(Fisher 2006), it has been found that even though limiting oxygen concentration (LOC) increases
the majority of the samples do not impose a risk with an increase in the initial pressure (in the
for explosion, 20% of the samples have an MIE of range of 5–18 bar) (Schwenzfeuer et al. 2001).
10 mJ or less, requiring special handling It has been reported that the LOC for ignition of
conditions to prevent ignition. dust clouds by electrostatic discharges or metal
There are three main strategies used for dust sparks was significantly higher than that deter-
explosion control, namely preventing the mined in standard tests by using a very strong
pyrotechnical ignition source (Schwenzfeuer A common concern in submicron particle-size

et al. 2001). Nonetheless, decreasing the oxygen reduction techniques is the particle-size change
in the process can impose the risk of suffocation due to dissolution of fine particles and/or growth
on the operator. To overcome this limitation, it on larger particles (Merisko-Liversidge and
has been shown that adding a small volume per- Liversidge 2008). The latter phenomenon,
centage of CO2 to the inert gas mixture can sig- known as Ostwald ripening, can occur with any
nificantly reduce the critical oxygen threshold for material and is accentuated when the solubility is
suffocation (Eckhoff 2005). a function highly dependent on temperature. This
Other operations, such as charging powders can be of particular relevance upon scaling up
into other units, packaging solids from a milling because the milling chamber surface area to
system, or cleaning equipment after operation, can batch volume ratio decreases, which ultimately
create a dust cloud that could lead to explosive influences the heat transfer in the process (Fisher
conditions. Therefore, attention to housekeeping 2006). Additionally, due to significant heating/
and the procedures used for these other operations cooling cycles that could occur during wet mill-
also has importance (Fisher 2006). ing, an annealing effect could potentially be
With an overall industry increase in the manu- induced in the solids (Trasi et al. 2010).
facture of increasingly potent APIs, personnel The system of mechanical sealing of
protection has become particularly important. wet-milling equipment needs to be controlled
By decreasing the particle size and creating fines due to two main issues that can arise: contamina-
in the process, dry milling is an operation that tion of the batch and seal lifetime. For the sealing
increases the risk of exposure to personnel com- system to operate, a seal fluid is normally used for
pared to wet milling (Stein et al. 2010). This lubrication and cooling purposes. Therefore, there
could eventually dictate the choice of milling is always the potential of contamination of the
equipment due to the increased cost of production batch with the seal fluid; thus, the fluid needs to
of dry milling and the protective equipment that be compatible with the solvent and API used in
personnel need to operate the mill. Wet milling in the milling process (Fisher 2006). Additionally,
cases of a very potent API can be a more cost- the sealing system should be thoroughly cleaned
effective choice. to extend the seal lifetime and decrease the
chances of batch contamination.
The particle size achieved in a batch will nor-
mally depend on the type of wet-milling equip-
4.3.2 Wet Milling
ment that is used. However, in general particle
size will be a function of the residence time of the
Wet milling, also known as slurry milling, is a
slurry in the mill (Stenger et al. 2005). This resi-
particle-size reduction process in which the solid
dence time can be controlled by operating the
particles are suspended in a liquid medium. As
system in either single-pass or recycling mode.
such, wet milling has a number of advantages
In either modality, it is important to determine
over dry milling, thermal control over the process
and control the slurry flow rate when considering
being one of the most prominent. Due to the
the scale-up process. Furthermore, depending on
thermal control, heat-labile materials can be
the mill, it should take a predetermined number of
processed through this technique simply by the
passes to achieve a steady particle-size
thermal properties of the liquid in the slurry. If
distribution.
additional cooling is needed, the liquid can be
In the pharmaceutical field, there are two com-
precooled or cooled during the process to control
mon types of wet mills: rotor–stator and media
the temperature. As stated above, this could pre-
(bead) mills. The latter is widely used due to its
vent chemical decomposition, solid phase
capacity to produce particle sizes in the nanoscale
transitions, or melting of the material being milled
range but at the cost of longer milling times (Kipp
(Merisko-Liversidge et al. 2003).
2004). The rotor–stator mill is widely used in the
field for emulsification and homogenization and stator) are the typical design parameters that can
can be used in the wet milling of APIs, achieving be controlled in the development of a rotor–stator
particle sizes in the 20–30 μm range (Atiemo- mill (Lee et al. 2004). For scaling-up purposes in
Obeng and Calabrese 2004; Lee et al. 2004). equipment that has a fixed rotor–stator design to
preserve rotor tip speed, scaling to a larger design
results in milled particle sizes that are comparable
Rotor–Stator Wet Milling
to those obtained in small-scale batches.
Rotor–stator milling equipment are also referred
As with other milling equipment, the solid
to as high-shear devices due to the shear rates
particle concentration in the slurry will typically
generated during the rotation of the mixing ele-
be directly related to the milling efficiency and
ment (rotor), which has a close proximity with the
cycle time. In general, increasing this total solid
static element. Probe rotor–stator mills are useful
particle content will increase the milling rate.
for small-scale development and are normally
Nonetheless, there is a limit to the slurry concen-
used as batch process-type vessels. Even though
tration, in terms of limiting the flow of the media
data from this setup can be used as a reference for
and potentially plugging the milling equipment
scaling-up purposes, the average particle sizes
(Atiemo-Obeng and Calabrese 2004).
might not be comparable upon increasing the
In general, the average particle sizes obtained
batch size. At the scale of tens of grams or
using a rotor–stator mill can be of a few microns,
more, a flow-through unit is the preferable choice,
with a D95 normally not being larger than
as these systems allow the slurry to pass through
40–50 μm (Fisher 2006). The hardness of the
the high-shear patterns in the mill repeatedly,
milled material dictates the cycle time when the
further decreasing particle size and narrowing
goal is to obtain narrow particle-size distribution.
the particle-size distribution (Nied 2007).
A common phenomenon observed with hard,
With toothed probe rotor–stator mills
block-like materials is the potential for multi-
(Fig. 4.13), the mechanism of particle-size reduc-
modal distributions (Lee et al. 2004). With these
tion is believed to occur as a combination of high
materials, breakage is believed to occur mainly at
shear and collision of particles with each other
corners and edges, which translates into “chip-
and the equipment walls (Atiemo-Obeng and
ping” of the hard material resulting in a high
Calabrese 2004). As a consequence, parameters
population of fines and larger particles yielding
that control the high shear of the equipment ulti-
bi- or multimodal particle-size distributions
mately control the performance of the mill in
(Table 4.4).
terms of particle-size distribution. Tip speed
(rotation rate of rotor rotor circumference),
shear rate (tip speed/distance between rotor and Media (Bead) Wet Milling
stator), and shear frequency (rotation rate This type of mill operates by mechanically
number of slots on rotor number of slots in moving media (beads) together with the material
Fig. 4.13 Rotor–stator

geometries: (a) simple
Couette conical geometry
(also known as colloid mill)
and (b) toothed rotor–stator
Table 4.4 Milling parameters that affect the performance of a rotor–stator wet mill (Fisher 2006)
Variable Particle-size outcome
Rotor speed Increasing tip speed increases milling rate and decreases average particle size. Typical range
10–50 m/s (Atiemo-Obeng and Calabrese 2004)
Rotor/stator Decreasing the gap increases shear forces, increasing milling rate, thus decreasing average particle
gap size
Rotor/stator The number of teeth will affect the shear forces, milling rate, and thus the average particle size
design achieved in steady state
Slurry density Higher solids concentration increases milling rate but normally does not translate in a decrease in
average particle size
Temperature Higher temperature can increase product dispersion but may compromise the physical (phase
transitions) or chemical (degradation) integrity of the drug
Residence time Increasing milling time will normally decrease the average particle size
(normally the API) in a liquid. This can be done glass, polytetrafluoroethylene (PTFE), and hard
either by means of a stirrer or by agitating the polystyrene derivatives (Keck and Muller 2006).
container itself. The beads can be made of various In direct comparison with a rotor–stator wet
materials; however, they are normally glass mill, media milling equipment can yield much
spheres, ceramics, metal, or plastic. The bead smaller average particle sizes for the same mate-
densities can range between 2500 and rial. Average particle sizes can range from the
7800 kg/m3, with sizes between 0.1 and 1–2 μm to the nanoscale range, with adequate
>10 mm (Kwade and Schwedes 2007). Closed distribution of particle sizes when proprietary
stirred media mills can be arranged either verti- media or additives in the slurry have been used
cally or horizontally and are normally equipped (Bruno et al. 1996; Liversidge et al. 1992). With
with a thermal jacket for cooling purposes. such a pronounced particle-size decrease, surface
According to the chamber and stirrer geometry, areas in 1–2 μm size batches can be 4–5 m2/g
three types of mills can be distinguished, namely and higher. Furthermore, nanoparticles obtained
a disk stirrer, a medial mill with pin-counter-pin by this milling method can have a surface area of
stirrer, and an annular gap geometry (Fig. 4.14). 10 m2/g or more (Fisher 2006; Muller et al.
The efficiency of the mill is related to a com- 2001).
bination of many variables, such as media speed, The fundamental processes that control
amount of loaded media, slurry concentration, particle-size reduction in this type of equipment
and milling time (Blecher and Schwedes 1996; are the number of stress events and the stress
Blecher et al. 1996; Kwade 2003; Schilde et al. intensity. In general, the average number of stress
2010). Additionally, material properties and the events for each particle, SN, is determined by the
overall mill design are also factors contributing to number of media contacts, NC, the probability
milling performance (Table 4.5). that a particle is caught and sufficiently stressed
The specific characteristics of the milling at a media contact, PS, and the number of product
media will greatly influence the performance of particles inside the mill, NP (Kwade 1999):
the milling process. Among different geometries
N C PS
of the milling media, it has been found that the SN ¼ : ð4:15Þ
NP
spherical geometry is the most effective shape
(Fisher 2006). Manufacturers offer different Kwade further developed a stress model
beads with diameter ranging from 0.05 to depicting the dependence of particle size, specific
130 mm; however, beads 6 mm and smaller are energy, and stress energy of the grinding media
normally used. Beads can also be obtained in (SEGM) in accordance with:
various materials such as (in order of decreasing
density): various grades of steel, zirconia (ZrO2), SE GM ¼ d3GM ∙ ρGM ∙ v2t ð4:16Þ
ZrSiO4, alumina (Al2O3), silica (SiO2), annealed
Fig. 4.14 Different stirrer a Outlet

and grinding chamber A–A A Rotating gap
geometries media (beads)
wet mills (Kwade 1999)
Intlet
A
Water jacket
Disc stirrer
b
A–A A
Pin-counter-pin stirrer
c A
A–A
Annular gap
Annular gap mill
Table 4.5 Milling parameters that affect the performance of a media (bead) wet mill (Fisher 2006)
Variable Particle-size outcome
Bead size Decreasing the bead size normally decreases the steady-state average particle size.
Intrinsic bead Modifying the density of the bead can have different results, based on the material milled. In some
density cases, more dense beads have produced coarser product.
Bead loading A bead charge of 50% normally provides adequate milling efficiency.
Agitator speed Increasing the linear velocity of the agitator tip increases the milling rate.
Temperature Increasing the temperature can promote product dispersion and small average particle sizes;
however, it can compromise the physical and chemical stability of the material and further dissolve
very small particles.
Residence time An increase in the residence time decreases the average particle size achieved.
Where dGM is the grinding media particle size was shown to work on hard inorganic materials
diameter, ρGM is the grinding media density, and (Kwade and Schwedes 2002). More recently,
vt is the tip speed of the stirrer in the mill (Kwade Bitterlich et al. expanded this model to the drugs
2003). Although this model does not consider (organic material) cinnarizine and fenofibrate
process parameters to estimate product quality, it (Bitterlich et al. 2014). While cinnarizine
exhibited a correlation between specific energy mill sizes from the same manufacturer. Key
and stress energy, fenofibrate milling conditions parameters such as bead loading, agitator tip
did not show a similar correlation. The authors speed, stabilizer amount, and linear velocity of
concluded that the employed stress energies were the slurry need to be analyzed during scaling up to
near the optimum region, and thus the influence achieve target particle sizes (Fisher 2006;
of stress energy over the specific energy require- Peltonen 2018).
ment was either very small (cinnarizine) or insig- One issue concerning the operation of this type
nificant (fenofibrate). Furthermore, the authors of mill at high throughput rates is hydraulic pack-
highlighted the potential limitations of the stress ing. This phenomenon is related to the milling
model due to unaccounted effects such as media being concentrated at the mill exit (pack-
agglomeration, ripening, and amorphization during), instead of remaining in the milling chamber
ing media milling (Colombo et al. 2009). during the operating cycle, which translates in
An alternative model has been developed by reduced milling efficiency. Hydraulic packing
Bilgili and Afolabi et al. (Afolabi et al. 2014; can result in an increase in power consumption,
Bilgili and Afolabi 2012) following the work together with an additional heat input to the sam-
earlier developed by Eskin et al (Eskin et al. ple batch (Fisher 2006). The flow rate that triggers
2005a, b). In this work, the authors correlate hydraulic packing has been found to be related to
stirrer speed, bead concentration, and drug load- increasing agitator tip speed, reduced bead load-
ing on the breakage kinetics of drug particles by ing, as well as the sample batch viscosity.
the adapted microhydrodynamic model for the Another limitation found in milling processes
bead–bead collisions. Several performed in media mills is the potential for
microhydrodynamic parameters derived from the shedding of the media and components into the
model provided significant physical insight, yet batch. It has been found that the use of glass beads
the milling intensity factor (F) exhibited a strong yielded glass microparticles in the final product
correlation with the process time constant and (Buchmann et al. 1996; Kipp 2004). Since no
was defined as follows: further purification steps are normally considered
after milling of APIs, it is important during devel-
c2 ð2 cÞ 1 13=10
F¼ θ ð4:17Þ opmental stages of the process to assess the
ð 1 cÞ 3 ε amount of objectionable substances in the batch.
For API manufacture, these levels are normally in
Where c is the volumetric concentration of the
the parts per million range (Fisher 2006); how-
beads, ε is the volumetric drug loading in the drug
ever, they can be as high as 70 ppm, imposing
suspension, and θ is the granular temperature
development limitations in the process (Keck and
(Afolabi et al. 2014). This model was more
Muller 2006). To overcome or limit the extent of
recently used in reducing processing times,
these materials in the final batch, an additional
energy consumption, and optimizing the milling
premilling step that decreases the coarse particle
of griseofulvin and indomethacin to a sub-100 nm
size of the raw product can be performed to
nanosuspension (Li et al. 2015).
reduce the residence time of the batch in the wet
As stated above, the factors determining the
mill, therefore reducing total shedding (Fisher
number of stress events and stress intensity are
2006).
dependent on the design and residence time in the
The wet-milling technology has been further
mill. Scaling up can therefore be complicated by
developed by G. Liversidge and coworkers and
the differences in relative contributions of number
turned into the high-energy media milling
of stress events and stress intensity and if power
NanoCrystal technology by Elan for obtaining
per unit volume for larger mills decreases (Bell
nanoscale nanocrystalline particle distributions
2005). Nonetheless, it has been previously
of API (Liversidge et al. 1992). The nanoparticle
reported that comparable particle sizes, approxi-
dispersions obtained by this technology consist of
mately <150 nm, can be achieved upon scaling up
the API and a surface stabilizer to avoid
aggregation and subsequent particle growth sufficient stabilization by reducing the zeta poten-
(Merisko-Liversidge and Liversidge 2008). To tial, thus preventing particle aggregation. In many
minimize the shedding of materials, the technol- cases, the final choice is a combination of both
ogy makes use of highly cross-linked polystyrene steric and electrostatic stabilization using a com-
beads; however, the extent of erosion is depen- bination of surfactants. There is a wide variety of
dent on a combination of the bead material and surfactants that can be used for the development
the physical characteristics of the drug (i.e., hard- of nanosuspensions meant for oral administration;
ness) and the residence time (Keck and Muller however, for the parenteral route, the choices are
2006). The grinding process developed by limited. Lecithins, Poloxamer 188, Tween
Liversidge et al. has a limitation upon scaling up 80, low-molecular weight polyvinylpyrrolidone
due to the heavy weight that it would impose, (PVP), and sodium glycocholate (combined with
increasing the size of the mill. lecithins) are surfactants that have been accepted
While more research in the field is required to for injection (Muller et al. 2006).
establish general guidelines in wet bead milling, The number of different media mills available
some physical drug properties have been in the market range from laboratory-scale to
identified as key properties that determine mallea- industrial-scale volumes. Upon scaling up, there
bility (Liu et al. 2019). Properties such as solid is need for equipment that can handle large
state (i.e., crystalline vs. amorphous) (Sharma volumes (and weights) of media and product.
et al. 2011), particle morphology (Liu et al. Since enlarging the whole milling unit would
2017, 2018a, b), particle size (Liu et al. 2018b), require a large amount of milling media, an exter-
crystallite size (Liu et al. 2018b), and Young’s nal suspension container has been devised for
modulus (Cerdeira et al. 2011) have all been operating mills with large-scale batches
found to be correlated with milling performance. (Fig. 4.15). With this arrangement, an additional
More specifically, a porous particle morphology container feeds a suspension to the milling system
or a decrease in crystallite size prior to milling, in more discrete quantities, allowing a smaller
obtained through spray drying of various drugs, mill to decrease the particle size of large
has been associated with a further reduction in quantities of feed. The downside of this strategy
size after processing (Liu et al. 2018b). is the increase in milling times (Muller et al.
Particles obtained by media milling that are in 2006).
the range of a few microns to the nanoscale are Generally, it may be understood that the use of
generally required to be stabilized using surface a smaller bead size will yield a smaller resultant
active agents or surfactants to prevent particle particle size. Li and colleagues demonstrated that
growth (Lee et al. 2005; Van Eerdenbrugh et al. decreasing bead sizes down to 50 μm resulted in a
2008). The process normally begins with the for- final nanosuspension mean size as low as 72 nm
mation of a macrosuspension in which the surfac- with polydispersity index values below 0.23 that
tant is added, followed by the wet-milling process had a relative short-term stability within the
itself. The surfactant to be used will be deter- sub-100 nm mean size (Li et al. 2015). In another
mined by a number of factors including the improvement upon bead properties, Funahashi
properties of the solid to be suspended (affinity and coworkers introduced the use of ice beads to
of the solid with the surfactant), physical mecha- produce a particle size reduction comparable to
nism of action (electrostatic vs. steric stabiliza- that observed with zirconia beads that led to simi-
tion), and route of administration of the lar rapid drug dissolution properties (Funahashi
nanosuspension (Berglund et al. 2003; Lee et al. et al. 2019). The use of ice beads as a griding
2005). Steric stabilization can be a more effective media can lead to a contamination-free milling
choice when there is a chance for poor gastroin- process due to the absence of shedding material.
testinal or systemic stability due to excess
electrolytes. Ionic surfactants can also provide
Fig. 4.15 Bead mill DISP Pumping and

ERMAT(use register mark stirring system
symbol) SL for mill base
amounts of 50 mL 50 L.
Schematic representation of
the laboratory mill
Dispermat SL. (Modified
from Schilde et al. 2010)
Milling
system
Cryogenic Milling surface tension (8.85 mN/m at 196 C) and

The term cryomilling has been used in the litera- viscosity (0.158 mPa at 196 C) exhibited by
ture to describe two different processes involving liquid nitrogen might prevent secondary
cryogenic conditions. The first is a grinding tech- coaggregation and undesired particle growth.
nique in which the solids are milled in a slurry Additionally, materials under very low
formed with a cryogenic milling media. The other temperatures (196 C for liquid nitrogen)
configuration consists of using a cryogenic liquid become more brittle and can be milled more effi-
to decrease the temperature of the grinding cham- ciently (Reed 2007).
ber and mill the solids as in dry milling. Nonethe- In the investigation by Niwa, after optimizing
less, both operations rely on a cryogenic for size and amount of zirconia beads and agita-
substance to control the temperature of the pro- tion speed, crystalline nanoparticles were
cess while decreasing by attrition the average obtained (Niwa et al. 2010). The authors found
particle size (Witkin and Lavernia 2006). The that, by increasing the agitation speed of the shaft
direct and obvious advantage of the process is from 550 to 1600 rpm, the average particle size
the suitability for thermally labile compounds; could be decreased. This was attributed to a
however, the removal of the cryogenic liquid higher kinetic energy imparted on the beads,
upon finishing the milling process may impose thus creating high-impact collisions between
certain limitations (Fisher 2006). particles and beads. The bead size in the
An investigation by Niwa et al. (2010) coined 0.1–1 mm diameter range was found not to
the term “ultra cryomilling” and was used for impact the performance of the milling process,
media milling that explicitly uses liquid nitrogen both achieving similar particle-size distributions.
as dispersant (Fig. 4.16). Liquid nitrogen was For the purpose of comparing the performance of
used as the cryogenic liquid due to its poor the novel cryomilling technique, the authors used
solubilizing potency. Furthermore, the low 1 mm beads. Powders processed by the ultra
More recently, the ultra cryomilling process has

Motor
been further modified by using dry ice as milling
media. Not only were the particle sizes achieved
Shaft
comparable to that of the standard process, but
Liquid nitrogen
also the authors found a yield increase due to the
sublimation of both the milling liquid (liquid
Disk nitrogen) and media (dry ice as beads) (Sugimoto
et al. 2012a).
Beads
Vessel 4.4 High-Pressure Homogenization

Techniques
Fig. 4.16 Diagram of the ultra cryomilling apparatus in
liquid nitrogen developed by Niwa et al. The beads, disks, Homogenization is a process by which the
shaft, and inner wall of the vessel were made of zirconia
particle-size distribution of a suspension or an
(Niwa et al. 2010)
emulsion is narrowed or “homogenized,” thus
decreasing the polydispersity of the sample. Par-
cryomilling technique were comprised of 21.5%
ticle breakage is achieved by a combination of
submicron particles, while powders processed by
high shear, turbulence, impact, as well as cavita-
jet milling only achieved 9.07% of submicron
tion in the homogenizer. The type of homoge-
particles (Niwa et al. 2010). The process was
nizer used determines which of these
shown to work with a wide variety of drugs in
mechanisms will be more important for particle-
terms of their physicochemical characteristics
size reduction together with the physical
such as heat sensitivity and water solubility. It
properties of the bulk powder.
was further demonstrated that the crystalline
structure found in the bulk solid remained after
processing. One inherent limitation of the process
though is the aggregation of particles upon liquid 4.4.1 Piston-Gap Homogenizers
nitrogen evaporation. To account for this, the
authors proposed an additional formulation Piston-gap homogenizers work with pure water as
design mainly focused on adding a surfactant to dispersant and rely on cavitation as the driving
increase the wettability of the particles. force for particle diminution. In these
In recent years, this use of ultra cryomilling homogenizers, the suspension (or sometimes a
has seen progress in both the materials used and coarse emulsion) passes through a very thin gap
the processing technology (Sugimoto et al. at very high velocities. Before entering the gap,
2012a, b). Phenytoin has been processed with the suspension is contained in a cylinder with a
ultra cryomilling to render much smaller particle relatively large diameter compared to the width of
sizes and more uniform in size in comparison the gap (Fig. 4.17). This tremendous decrease in
with jet-milled drug, yet the dissolution rate was diameter leads to an immense drop in the static
not improved due to aggregation. To overcome pressure of the liquid. The resulting static pres-
this limitation, Sugimoto et al. added a number of sure is below the vapor pressure of water, leading
polymeric excipients (polyvinylpyrrolidone, to boiling and the creation of gas bubbles that
Eudragit L100, hypromellose, hypromellose implode after leaving the gap, returning to normal
acetate-succinate, microcrystalline cellulose, pressure in the reservoir compartment of the
hydroxypropylcellulose, and carboxymethyl cel- homogenizer. This first technology, known as
lulose) that provided stability to the DissoCubes, considered this phenomenon, cavi-
nanosuspension and resulted in an improvement tation, as a primary factor in particle breakage
of the dissolution rate (Sugimoto et al. 2012b). (Muller et al. 1999).
Fig. 4.17 Basic homogenization principles: piston-gap is due to shear forces, cavitation, and impaction. In
(left) and jet-stream arrangement (right). In the piston- jet-stream homogenizers, the collision of two high-
gap homogenizer, the macrosuspension is forced to pass velocity streams leads to particle breakage mainly by
through a small gap (few microns), and particle breakage impact forces. (Adapted from Muller et al. 2006)
Particle-size decrease is controlled by several been found to achieve particle sizes ranging from
factors: homogenization pressure, number of 500 to 800 nm, depending on the processing
homogenization cycles, and hardness of the drug conditions (Keck and Muller 2006). Finally, the
and stabilizers (Keck and Muller 2006). In gen- nature and concentration of the stabilizer (either
eral, increasing the pressure will translate to a surfactant or polymer) in the formulation, as
decrease in average particle size; however, this opposed to emulsions, have been found to not
relationship is not always linear. It has been found affect the extent of particle-size diminution,
that up to a pressure of 4000 bar, there is a certain although they do have an impact on aggregation
pressure above which particle-size diminution in the long term (Muller et al. 1996).
does not decrease any further (Fichera et al.
2004). The reason for this phenomenon appears
to be that upon crystal breakage and particle-size
4.4.2 Microfluidizer
reduction, the imperfections in particles become
reduced in number per particle, thus decreasing
The mechanism of particle-size reduction for this
the chance for particle breakage. A decrease in
high-pressure homogenization technique is
average particle size may also result from an
through the impaction of jet streams (Dearn
increase in the number of homogenization cycles.
2000). As seen in Fig. 4.18, prehomogenized
In general, after five cycles, further increasing the
liquid is pumped through an interaction chamber
number of cycles does not decrease the batch D50
where impaction occurs. The interaction chamber
(Keck and Muller 2006). Increasing the number
is composed of microchannels that split the liquid
of cycles to 15–20 does have an impact on the
line in two. These two streams are then
homogeneity of the sample and is reflected in a
recombined at high velocities to produce forces
decrease of the batch D95. As in other grinding
of shear, impaction, and cavitation. These forces
techniques, softer material can be processed to
combine to cause particle breakage and size
achieve smaller average particle sizes when
reduction. A detailed description of the operation
processing conditions are kept the same. Pacli-
and examples of the use of the microfluidizer can
taxel, a soft drug, has been reported in the 250 nm
be found in US patent 4,553,254 (Cook and
range after 10–20 homogenization cycles at
Lagace 1985).
1500 bar. A harder drug, azodicarbonamide, has
Fig. 4.18 Flow chart of the particle-size reduction process in the Microfluidizer. (Adapted from Singh and Naini 2006)
A disadvantage of this technology is that the Mueller 2006; Salazar et al. 2014). To this date,
product obtained can have a rather large quantity five combinative strategies have been described
of particles in the larger size range (Illig et al. and are summarized in Table 4.6: NANOEDGE
1996). Furthermore, a high number of passes and nanoedge-like strategies, H 69, H 42, H
have been described in patents, reaching some- 96, and CT.
times even 75 passes and ultimately translating
into long processing times. Additionally, the sys-
tem is more compatible with soft drugs, enabling
4.5.1 NANOEDGE®
nanosuspensions to be achieved during
processing (Keck and Muller 2006; Muller et al.
The NANOEDGE technology belongs to Baxter,
2006). Larger particles often result when harder
and it was the first combinative technology
drugs are processed by this method, reducing the
described back in the early 2000s (Kipp et al.
extent to which solubility can be increased
2002, 2006). The technology combines a first
through mechanical size reduction.
step of microprecipitation based on antisolvent
precipitation of the poorly water-soluble drug.
For this, the drug is first dissolved in a water
4.5 Combination of Top-Down miscible organic solvent and then mixed with an
and Bottom-Up in a Single aqueous solution that triggers the precipitation of
Process for the Manufacture poorly water-soluble drug. The resulting particles
of Drug Nanosuspensions can either result in small size amorphous and/or
semicrystalline particles suitable for the commi-
Due to the limitations found in either wet ball nution step. This microprecipitation step is then
milling or high-pressure homogenization such as followed by a high-pressure homogenization
the need of a micronized suspension as the where particle size is further reduced and
starting material, and the long runtimes required stabilized by either an enhancement of the crys-
for particle-size reduction, new strategies that tallinity or by a rearrangement of the surface
combine top-down and bottom-up approaches stabilizer (J.E. Kipp 2004). Although high-
have been recently described (Moeschwitzer and pressure homogenization is the conventional
Table 4.6 Combinative particle-size reduction techniques and the various drugs that have been used as examples
Smallest
mean
particle
Combinative Size reduction size Administration
technology Pretreatment technique Drugs used achieveda route
NANOEDGE Microprecipitation High-pressure Paclitaxel, nabumetone, 177 nm Intravenous
homogenization prednisolone,
or sonication carbamazepine, itraconazole
Nanoedge- Microprecipitation High-pressure Meloxicam, isradipine, 80 nm Oral
like homogenization 10-hydroxycamptothecin,
or sonication hydrocortisone, ibuprofen,
nitrendipine, all-trans
retinoic acid
H 69 Cavi-precipitation High-pressure Ibuprofen, hydrocortisone 22 nm Oral
homogenization acetate, resveratrol,
omeprazole, prednisolone
H 42 Spray-drying High-pressure Amphotericin B, 172 nm Oral
homogenization glibenclamide,
hydrocortisone acetate,
ibuprofen, resveratrol
H 96 Freeze-drying High-pressure Amphotericin B, 62 nm Oral
homogenization glibenclamide,
cyclosporine A,
hydrocortisone acetate
CT Pearl milling High-pressure Rutin, hesperidin, apigenin 275 nm Topical /oral
homogenization
Full citation details can be found in Salazar et al. (2014)
a
Indicated as the smallest mean particle size of a specific studied drug. Each drug has its own particle-size reduction
properties, including smallest particle size achieved
comminution step, other high-energy processes technology, H 69 involves antisolvent precipita-

such as sonication or microfluidization can be tion and high-pressure homogenization, yet the
used (Kipp et al. 2002, 2006). The remaining precipitation occurs in the presence of cavitation,
organic solvent from the antisolvent step hence cavi-precipitation (Table 4.6). The drug is
represents a substantial limitation for product dissolved in a suitable water-miscible organic
development, and particularly for the injectable solvent which is then incorporated into an aque-
development. The NANOEDGE technology and ous solvent (nonsolvent of the drug) in the high-
nanoedge-like technologies have been used to energy zone of a homogenizer. This results in
formulate a number of poorly water-soluble particle formation, growth, and comminution
drugs (Table 4.6) including anticancer agents occurring simultaneously as the high-pressure
such as placlitaxel (Kipp 2004; Rabinow et al. homogenization process advances through cavi-
2004) and more recently quercetin (Qiao et al. tation, particle collision, and shear forces (Müller
2020). and Möschwitzer 2015; Sinha et al. 2013). The
high-energy homogenization process favors the
crystallized state of nanoparticles, increasing
their stability. The resulting nanosuspension also
4.5.2 smartCrystal Technology
contains a portion of organic solvent which is a
limitation requiring further processing toward the
H 69 Technology
final product. At defined conditions, the H
H 69 technology, along with H 42 and H96, is
69 technology can yield nanosuspensions as
part of the smartCrystal technology family (Keck
small as 22 and 27 nm with rather large
et al. 2008). Similar to the NANOEDGE
polydispersion (polydispersity indices 0.44–0.46) aqueous nanosuspension without further

(Müller and Möschwitzer 2015). processing. The right choice of solvent or solvent
mixtures to solubilize, freeze, and efficaciously
H 42 Technology lyophilization is key in the overall process perfor-
Another application of the smartCrystal technol- mance and depends mainly on the drug (Salazar
ogy is termed H 42 and addresses the organic et al. 2012). As processing occurs at freezing
solvent issue found in technologies above. This temperatures, this technology enables particle-
method combines a particle manufacture step of size reduction of thermolabile drugs. Although
spray drying followed by high-pressure homoge- H 96 has been largely described for high-pressure
nization (Salazar et al. 2013). The drug is first homogenization as the secondary comminution
dissolved in an organic solvent and optionally process, studies have been conducted with
mixed with surfactants or sugars to stabilize the wet-media (bead) milling demonstrating that the
solid during and after the spray drying step. In lyophilized brittle material can be processed into
addition to the advantage of organic solvent a nanosuspensions with both particle-size reduc-
removal, the resulting particles from the spray tion strategies (Salazar et al. 2012). Figure 4.19
drying step are more breakable for the high- summarizes the particle-size reduction potential
pressure homogenization step that follows of the microprecipitation followed by high-
(Moschwitzer 2011). Another advantage of the pressure homogenization H 69, H 42, and H
H 42 technology is the reduction in the degree 96 technologies in comparison with conventional
of crystallinity of the starting powders (spray wet-media (bead) milling and high-pressure
dried), which results in the need for lower number homogenization.
of high-pressure homogenization cycles (reduc-
tion from 20 to 1 cycle) and positive particle-size
performance (Liu et al. 2012; Salazar et al. 2014). 4.5.3 CT Technology
The main limitation of this method of manufac-
ture is associated to the heating required during The Combination Technology (CT) is the only
the drying step, which could limit its use in ther- combination strategy that does not rely on the use
molabile drugs (Moeschwitzer and Mueller of organic solvents and consists of two sequential
2006). top-down processes(Keck et al. 2008; Petersen
2015). This process consists of a low-energy
H 96 Technology pearl milling step followed by a high-pressure
The last addition to the smartCrystal technology homogenization step. The first process reduces
family is H 96 where the process begins with a that size of a macrosuspension to a range between
freeze drying step followed by high-pressure 600 and 1500 nm average size. The high-pressure
homogenization (Moschwitzer and Lemke 2008; homogenization stage further decreases particle
Shegokar and Muller 2010). The process consists size but more importantly provides the product
of dissolving the drug in an organic solvent that is with nanosuspension stability by bringing all
then frozen with liquid nitrogen and subsequently larger particles to a homogeneous small particle
freeze dried. Depending on the freeze drying size (decreasing the impact of Ostwald ripen-
conditions, the resulting material can present ing) (Keck et al. 2008). Although this technol-
varying degrees of crystallinity and yet remain ogy reaches average particle sizes larger than
brittle for the latter high-pressure homogenization the aforementioned strategies, CT is advanta-
step (Salazar et al. 2011). Similar to the H 42 tech- geous due to the lower pressure required,
nology, in the H 69 technology, the organic sol- shorter process length, and improved physical
vent is removed prior to the high-pressure stability of nanosuspensions (Al Shaal et al.
homogenization process, rendering the final 2010).
Fig. 4.19 Particle-size

reduction performance of
standard and combinative
technologies. Six levels of
particle-size reduction
progress are depicted:
premilling (1), 1 High-
pressure homogenization
cycle at 1500 bar/1 hour of
WBM (2), 5 cycles/2 hours
(3), 10 cycles/4 hours (4),
15 cycles/8 hours (5), and
20 cycles/24 hours
(6) (Salazar et al. 2014)
technologies, there are several other methods in

4.6 Conclusion
the pipeline that incorporate mechanical particle-
size reduction using different techniques,
The number of new drugs and drug candidates
demonstrating the growing acceptance of
with poor aqueous solubility is steadily increas-
particle-size reduction as a means of improving
ing, and there is a need to address this solubility
the water solubility for a wide variety of existing
limitation. Decreasing the particle size of a sub-
and new active pharmaceutical ingredients.
stance achieves an increase in surface area by
mass, can increase saturation solubility, and
Method Capsule 1: Particle-Size Reduction by
decreases diffusional distance, each of which ulti-
Spiral Air Jet Milling
mately results in an increase of either or both the
extent and rate of dissolution for a given material. Based on the method reported by Schlocker
Top-down approaches that rely on mechanical et al. (2006).
particle-size reduction can be roughly classified Objective
into milling techniques and high-pressure homog- • To obtain protein-loaded microparticles
enization techniques. Of these top-down with adequate protein stability after
strategies, wet-media milling, piston-gap homog- processing. Horseradish peroxidase was
enization, and microfluidization are the most pop- used as model protein and was
ular. These widely used approaches may yield coprecipitated with either carbomer or a
particle-size distributions in the low micron to poly(methacrylate) prior to grinding.
the nanoscale range. Various technologies have Equipment and Materials
been patented using such top-down approaches. • Carbopol 934P and Eudragit L 100–55 as
Technologies such as NanoCrystal® by Elan and polymer matrix.
IDD-P® by SkyePharma are the only ones that • Horseradish peroxidase as protein model.
yield particles in the nanoscale range and are • Spiral jet mill (Hosokawa Alpine Aeroplex
currently used in marketed products. In addition 50 AS) with a diameter and height of grind-
to these successfully commercialized ing chamber 50 and 4.5 mm, respectively;
standard blowing-out nozzle; number of Objective

nozzles 4; nozzle diameter 0.8 mm; nozzle • To obtain a suspension of PVP surface-
pitch 50 ; and equipped with a temperature modified crystalline danazol nanoparticles
sensor (Lutron, DH-802C, Ming Chuan, to improve bioavailability.
Taiwan). Equipment and Materials
Method • Wet-media mill with a 600 mL cylindrical
• After coprecipitation of the model protein vessel (inner diameter 7.6 cm).
and one of the model polymer matrices, 1 g • 200 mL of zirconia grinding media
of the powder was added in the chamber of (0.85–1.18 mm diameter)
the spiral air jet mill. • Micronized danazol (mean particle size
• Injector air pressure was 7.5 bar. 10 μm).
• Grinding air pressure (GAP) was either 2.5 • PVP K-15.
or 4.5 bar, and the powder was ground for Method
10 minutes. • Zirconia grinding media was added to the
• After milling, the material was collected vessel of the mill.
from the grinding chamber and stored at • Micronized danazol and PVP were added to
18 C. the mill, and water was used as a dispersing
Results agent to an adequate volume.
• After laser diffraction analysis, it was • The vessel was rotated horizontally around
revealed that increasing the GAP in the its axis at 57% of the critical speed.
mill decreased the microparticles’ size. • After 5 days of milling, the slurry was
• Using isopropanol as the nonsolvent in the separated from the grinding media by
coprecipitation of the protein and carbomer sieving.
and increasing the GAP from 2.5 to 4.5 bar Results
decreased D50 from 5.2 to 2.7 μm. By • A sedimentation field flow fractionator was
using petroleum ether as the nonsolvent in used for sizing and, after processing, the
the coprecipitation of the protein and poly number average weight was 84.9 nm and
(methacrylate) and increasing the GAP the weight average weight was 169.1 nm.
from 2.5 to 4.5, the D50 was decreased • Particle size of the batch ranged from 26 to
from 4.5 to 2.8 μm. 340 nm.
• A GAP of 4.5 bar in the case of the poly • X-ray diffraction demonstrated that the
(methacrylate) reduced protein stability crystalline structure of danazol remained
(measured as activity after processing) unchanged after the process.
almost completely, while at 2.5 bar 58% • Dispersions of unmilled danazol and the
activity remained. In the case of carbomer, nanoparticulate slurry were dosed in
increasing GAP from 2.5 to 4.5 bar beagles. The relative bioavailability of dana-
decreased the remaining activity from zol from the nanoparticulate dispersion was
75 to 35%. 15.9-fold higher than from the danazol sus-
• The spiral air jet mill was found to be an pension containing microparticles (10 μm).
effective process for the manufacture of
protein-loaded microparticles with ade- Method Capsule 3: Particle-Size Reduction
quate protein stability after processing. by Ultra Cryomilling
Based on the method reported by Niwa
Method Capsule 2: Particle-Size Reduction
et al. (2010).
by Media Milling
Objective
Based on the method reported by Liversidge • To obtain nanocrystals by a one-step mill-
et al. (1992). ing process for drugs with a wide variety of
physical properties, such as heat sensitivity, Method Capsule 4: Particle-Size Reduction

oxidizability, or water solubility. by Piston-Gap High-Pressure
Equipment and Reagents Homogenization
• Liquid nitrogen.
Based on the method reported by Grau
• Phenytoin.
et al. (2000).
• Batch-type wet mill with a 400-mL capac-
Objective
ity vessel.
• To reproducibly obtain a 1% potency
• Zirconia rotation shaft, disks, and beads.
nanosuspension of 4-[N-(2-hydroxy-2-
• Sieve No. 60 (0.25 mm).
methyl-propyl)-ethanolamino]-2,7-bis
Method
(cis-2,6-dimethylmorpholin-4-yl)-6-phe-
• Similar to a typical wet-milling process, the
nyl-pteridine in water using a piston-gap
drug is suspended in liquid nitrogen, and
high-pressure homogenizer.
the suspension of drug and milling beads is
Equipment and Materials
vigorously agitated.
• Micron Lab 40 (APV Deutschland GmbH,
• Precool the parts of the mill (vessel, shaft,
Germany), 40 mL capacity high-pressure
disks, and beads) with liquid nitrogen.
homogenizer.
• Fill the milling vessel with 550 g of zirconia
• 4-[N-(2-hydroxy-2-methyl-propyl)-
beads 1 mm diameter (approximately
ethanolamino]-2,7-bis(cis-2,6-dimethyl-
150 mL).
morpholin-4-yl)-6-phenyl-pteridine
• Suspend 15 g of phenytoin in liquid nitro-
• Tween 80.
gen and fill the milling chamber to 360 mL.
• Glycerol.
• Start and maintain a controlled agitation at
Method
1600 rpm for 30 minutes, and pour addi-
• The drug is suspended in an aqueous solu-
tional liquid nitrogen to account for loss
tion containing the surfactants.
due to volatilization.
• At room temperature, the suspension is
• Pass the slurry through a sieve to separate
homogenized using a variable pressure
the powder from the beads.
profile.
• Collect the milled powder after all the liq-
• Two cycles are performed at 150 bar.
uid nitrogen evaporates.
• Two more cycles are performed at 500 bar.
Results
• Finally, 10 cycles are performed at 1500 bar.
• After laser diffraction analysis, the particle-
• If needed, at the end of any cycle draw a
size profile revealed a D10, D50, and D90
slurry sample to determine particle size.
of 0.62, 1.84, and 5.27 μm, respectively,
• At the end of the particle-size reduction
with a 27.5% of submicron particles per
process, collect the nanosuspension from
batch.
the high-pressure homogenizer.
• X-ray powder diffraction patterns and DSC
Results
curves indicated that the crystalline struc-
• After the whole homogenization cycle
ture of particles before milling remained
(14 cycles), photon correlation spectros-
after the process.
copy revealed a mean diameter of 502 nm
• The crystalline habit of the particles
and a polydispersity index of 0.390.
remained the same as the initial.
• Laser diffraction data revealed D10, D50,
• Milled particles tend to agglomerate; they
and D90 of 0.31, 0.95, and 2.20 μm,
may require a wetting agent for dispersion.
respectively.
• High reproducibility was found for the mean
size in between batches and at each cycle
number with the parameters utilized in the
high-pressure homogenization process.
Method Capsule 5: Particle-Size Reduction • DSC and powder X-ray diffraction showed
by Cavi-Precipitation Coupled that crystallinity of ibuprofen remained
with High-Pressure Homogenization after the precipitation and the high-pressure
homogenization step as well.
Based on the method reported by Sinha
et al. (2013).
Method Capsule 6: Particle-Size Reduction
Objective
by Microprecipitation Followed by
• To obtain ibuprofen nanocrystals by using
High-Pressure Homogenization
the combined bottom-up and top-down
approach of coupling cavi-precipitation Based on the method reported by Qiao
with high-pressure homogenization. et al. (2020).
Equipment and Materials Objective
• Emulsiflex C5 (Avestin Europe GmbH, • To obtain quercetin nanocrystals by using
Mannheim, Germany) piston-gap homoge- the combined bottom-up and top-down
nizer modified by inserting HPLC tubing to approach of coupling microprecipitation
direct the antisolvent input as close as pos- with high-pressure homogenization.
sible to the homogenization gap. Equipment and Materials
• Ibuprofen. • NS1001L2K high-pressure homogenizer
• Isopropanol. (Niro Soavi S.p.A. Co, Parma, Italy).
• Hydroxypropyl methylcellulose • Quercetin.
(HPMCE5). • mPEG-DCA,
• Sodium dodecyl sulphate (SDS). • DMSO.
Method Method
• The drug (ibuprofen) is dissolved in the • The drug (50 mg quercetin) is dissolved by
solvent phase (isopropanol). sonication in the solvent (organic) phase
• The antisolvent aqueous phase is composed (DMSO).
of 0.1% w/v HPMCE5 and 0.2% w/v SDS. • The antisolvent aqueous phase is obtained
• Set the homogenizer pressure to by dissolving 25 mg of mPEG-DCA as
1200–1300 bar and outside temperature to stabilizer in 50 mL of distilled water.
0–1 C through an iced water jacket. • A presuspension is obtained by
• Set the antisolvent phase (aqueous) to quickly adding the organic in the aqueous
0–1 C and incorporate into the high- phase and stirring at 2000 rpm for
pressure homogenizer. 10 minutes at room temperature.
• Incorporate all the solvent phase (organic • The presuspension is then added to the
solution containing ibuprofen) during the high-pressure homogenizer and processed
first homogenization cycle using an HPLC at various pressure and homogenization
pump as close as possible to the cycles to render a nanosuspension.
homogenization gap. • The nanosuspension is then frozen and
• Continue the homogenization for 10 further freeze dried to obtain a dry powder.
cycles. Results
Results • Large nanoparticles (332 nm) were
• The optimum solvent/antisolvent ratio that achieved at 100 bar and 20 cycles; medium
yielded the smallest particle size (304 nm) nanoparticles (159 nm) were achieved
was found to be 1.65/56.4. using 200–800 bars and 5–25 cycles;
• The ibuprofen concentration that resulted in while small nanoparticles (30 nm) were
the smallest particle size (304 nm) was achieved using a higher presuspension con-
45.44 w/v. centration of 250–1000 bars and
5–25 cycles.
• Quercetin nanoparticles exhibited enhanced Bilgili E, Afolabi A. A combined microhydrodynamics–

antitumor efficacy and improved the polymer adsorption analysis for elucidation of the roles
of stabilizers in wet stirred media milling. Int J Pharm,
immune response of treatment group. 2012;439(1–2), 193–206. https://doi.org/10.1016/j.
ijpharm.2012.09.040.
Acknowledgments The authors thank the support from Bisrat M, Nyström C. Physicochemical aspects of drug
FONDECYT Regular Project 1181689, ANID/PIA/ release VIII. The relation between particle size and
ACT192144, and FONDAP Project 15130011, all granted surface specific dissolution rate in agitated
by Chilean National Agency for Research and Develop- suspensions. Int J Pharm. 1988;47(1–3):223–31.
ment (ANID). https://doi.org/10.1016/0378-5173(88)90235-9.
Bitterlich A, Laabs C, Busmann E, Grandeury A,
Juhnke M, Bunjes H, Kwade A. Challenges in
nanogrinding of active pharmaceutical ingredients.
Chem Eng Technol. 2014;37(5):840–6. https://doi.
References org/10.1002/ceat.201300697.
Blecher L, Schwedes J. Energy distribution and particle
Afolabi A, Akinlabi O, Bilgili E. Impact of process trajectories in a grinding chamber of a stirred ball mill.
parameters on the breakage kinetics of poorly water- Int J Miner Process. 1996;44–45:617–27. https://doi.
soluble drugs during wet stirred media milling: a org/10.1016/0301-7516(95)00070-4.
microhydrodynamic view. Eur J Pharm Sci. 2014;51: Blecher L, Kwade A, Schwedes J. Motion and stress
75–86. https://doi.org/10.1016/j.ejps.2013.09.002. intensity of grinding beads in a stirred media mill.
Agrawal S, Ashokraj Y, Bharatam PV, Pillai O, Part 1: energy density distribution and motion of single
Panchagnula R. Solid-state characterization of rifampi- grinding beads. Powder Tech. 1996;86(1):59–68.
cin samples and its biopharmaceutic relevance. Eur J https://doi.org/10.1016/0032-5910(95)03038-7.
Pharm Sci. 2004;22(2–3):127–44. https://doi.org/10. Böhm BHL, Grau MJ, Hildebrand GE, Thünemann AF,
1016/j.ejps.2004.02.011. Müller RH. Preparation and physical properties of
Al Shaal L, Müller RH, Shegokar R. smartCrystal combi- nanosuspensions (DissoCubes™) of poorly soluble
nation technology--scale up from lab to pilot scale and drugs. Proc Intern Symp Control Rel Bioact Mater.
long term stability. Pharmazie. 2010;65(12):877–84. 1998;25:956–7.
Albus FE. The modern fluid energy mill. Chem Eng Prog. Boundy M, Leith D, Polton T. Method to evaluate the
1964;60(6):102–6. dustiness of pharmaceutical powders. Ann Occup
am Ende D, Rose P. Strategies to achieve particle size of Hyg. 2006;50(5):453–8. https://doi.org/10.1093/
active pharmaceutical ingredients. In: Abdel-Magid A, annhyg/mel004.
Caron S, editors. Fundamentals of early clinical drug Bruno J, Doty B, Gustow E, Illig K, Rajagopalan N,
development: from synthesis design to formulation. Sarpotdar P. Method of grinding pharmaceutical
Hoboken: Wiley-Interscience; 2006. p. 247–67. substances. 1996. Retrieved from http://www.google.
Ambrus R, Gieszinger P, Gáspár R, Sztojkov-Ivanov A, c o m / p a t e n t s / a b o u t ? i d ¼q L Q a A A A A E B A J &
Ducza E, Márki Á, Janáky T, Tömösi F, Kecskeméti G, dq¼5,518,187. Accessed 21 May 1996.
Szabó-Révész P, Bartos C. Investigation of the absorp- Buchmann S, Fischli W, Thiel F, Alex. Aqueous
tion of nanosized lamotrigine containing nasal powder microsuspension, an alternative intravenous formula-
via the nasal cavity. Molecules. 2020;25(5):1065. tion for animal studies. In 42nd Annual Congress of the
https://doi.org/10.3390/molecules25051065. International Association for Pharmaceutical Technol-
Atiemo-Obeng V, Calabrese R. Rotor-stator mixing ogy (APV), Mainz; 1996. p. 124.
devices. In: Paul E, Atiemo-Obeng V, Kresta S, Cerdeira AM, Gander B, Mazzotti M. Role of milling
editors. Handbook of industrial mixing: science and parameters and particle stabilization on nanogrinding
practice. Hoboken: Wiley-Interscience; 2004. of drug substances of similar mechanical properties.
p. 479–505. Chem Eng Technol. 2011;34(9):1427–38.
Bell TA. Challenges in the scale-up of particulate Chamayou A, Dodds JA. Air jet milling. In: Williams JC,
processes–an industrial perspective. Powder Tech. Allen T, editors. Handbook of powder technology, vol.
2005;150(2):60–71. https://doi.org/10.1016/j.powtec. 12. Amsterdam: Elsevier; 2007. p. 421–35.
2004.11.023. Chan HK, Doelker E. Polymorphic transformation of some
Bentham AC, Kwan CC, Boerefijn R, Ghadiri drugs under compression. Drug Dev Ind Pharm.
M. Fluidised-bed jet milling of pharmaceutical 1985;11(2–3):315–32. https://doi.org/10.3109/
powders. Powder Tech. 2004;141(3):233–8. https:// 03639048509056874.
doi.org/10.1016/j.powtec.2004.01.024. Chaumeil J. Micronization: a method of improving the
Berglund K, Przybycien T, Tilton R. Coadsorption of bioavailability of poorly soluble drugs. Methods Find
sodium dodecyl sulfate with hydrophobically modified Exp Clin Pharmacol. 1998;20(3):211–6.
nonionic cellulose polymers. 1. Role of polymer Chikhalia V, Forbes R, Storey R, Ticehurst M. The effect
hydrophobic modification. Langmuir. 2003;19(7): of crystal morphology and mill type on milling induced
2705–13. https://doi.org/10.1021/la026429g.
crystal disorder. Eur J Pharm Sci. 2006;27(1):19–26. Eskin D, Zhupanska O, Hamey R, Moudgil B, Scarlett
https://doi.org/10.1016/j.ejps.2005.08.013. B. Microhydrodynamics of stirred media milling. Pow-
Chow AHL, Tong HHY, Chattopadhyay P, Shekunov der Technol. 2005b;156(2–3):95–102. https://doi.org/
BY. Particle engineering for pulmonary drug delivery. 10.1016/j.powtec.2005.04.004.
Pharm Res. 2007;24(3):411–37. https://doi.org/10. Fichera M, Wissing S, Müller R. Effect of 4000 bar
1007/s11095-006-9174-3. homogenisation pressure on particle diminution in
Clark AR, Shire SJ. Formulation of proteins for pulmonary drug suspensions. (2004. pp 679–680..
drug delivery. In: McNally EJ, editor. Protein formula- Fisher E. Milling of active pharmaceutical ingredients. In:
tion and delivery. New York: M. Dekker; 2000. Swarbrick J, editor. Encyclopedia of pharmaceutical
p. 201–34. technology. 3rd ed. New York: Informa Healthcare;
Clement S, Purutyan H. Narrowing down equipment 2006. p. 2339–51.
choices for particle-size reduction. Chem Eng Prog. Fontana F, Figueiredo P, Zhang P, Hirvonen JT, Liu D,
2002;98(6):50. Santos HA. Production of pure drug nanocrystals and
Colombo I, Grassi G, Grassi M. Drug mechanochemical nano co-crystals by confinement methods. Adv Drug
activation. J Pharm Sci. 2009;98(11):3961–86. Deliv Rev. 2018;131:3–21. https://doi.org/10.1016/j.
Cook EJ, Lagace AP. Apparatus for forming emulsions. addr.2018.05.002.
1985. Retrieved from http://www.google.com/patents/ Friedrich A. Size reduction overview: shear, compression,
about?id¼H6gtAAAAEBAJ&dq¼apparatus+for and impact. Powder Bulk Eng. 2001;15(6):19–25.
+forming+emulsions. Accessed 6 Aug 1985. Funahashi I, Kondo K, Ito Y, Yamada M, Niwa T. Novel
de Vegt O, Vromans H, den Toonder J, van der Voort contamination-free wet milling technique using ice
MK. Influence of flaws and crystal properties on parti- beads for poorly water-soluble compounds. Int J
cle fracture in a jet mill. Powder Tech. 2009;191(1–2): Pharm. 2019;563:413–25. https://doi.org/10.1016/j.
72–7. https://doi.org/10.1016/j.powtec.2008.09.014. ijpharm.2019.04.008.
de Villiers M. Influence of cohesive properties of Gieszinger P, Tomuta I, Casian T, Bartos C, Szab-
micronized drug powders on particle size analysis. J ó-Révész P, Ambrus R. Definition and validation of
Pharm Biomed Anal. 1995;13(3):191–8. https://doi. the design space for co-milled nasal powder containing
org/10.1016/0731-7085(95)01274-O. nanosized lamotrigine. Drug Dev Ind Pharm. 2018;44
de Villiers M, Tiedt L. An analysis of fine grinding and (10):1622–30.
aggregation of poorly soluble drug powders in a vibrat- Godet-Morand L, Chamayou A, Dodds JA. Talc grinding
ing ball mill. Pharmazie. 1996;51(8):564–7. in an opposed air jet mill: start-up, product quality and
Dearn AR. Atovaquone pharmaceutical compositions. production rate optimization. Powder Technol.
2000, January 25. Retrieved from http://www.google. 2002;128(2–3):306–13. https://doi.org/10.1016/
c o m / p a t e n t s / a b o u t ? i d ¼2 i E D A A A A E B A J & S0032-5910(02)00172-9.
dq¼Atovaquone+pharmaceutical+compositions. Grau MJ, Kayser O, Muller RH. Nanosuspensions of
Accessed 25 Jan 2000. poorly soluble drugs: reproducibility of small scale
Dong Y, Feng S. Poly(d, l-lactide-co-glycolide) (PLGA) production. Int J Pharm. 2000;196(2):155–9. https://
nanoparticles prepared by high pressure homogeniza- doi.org/10.1016/S0378-5173(99)00411-1.
tion for paclitaxel chemotherapy. Int J Pharm. Gribbon P, Andreas S. High-throughput drug discovery:
2007;342(1–2):208–14. https://doi.org/10.1016/j. what can we expect from HTS? Drug Discov Today.
ijpharm.2007.04.031. 2005;10(1):17–22. https://doi.org/10.1016/S1359-
Eckhoff R. Dust explosion hazards in the ferro-alloys 6446(04)03275-1.
industry. In: 52nd electric furnace conference. 1994, Hamelmann F, Schmidt E. Methods of estimating the
pp 283–302. dustiness of industrial powders-a review. Kona.
Eckhoff R. Dust explosions – origin, propagation, preven- 2003;21:7–18.
tion, and mitigation: an overview. In: Eckhoff R, edi- Hancock BC, Parks M. What is the true solubility advan-
tor. Dust explosions in the process industries. 3rd tage for amorphous pharmaceuticals? Pharm Res.
ed. New York: Gulf Professional; 2003. p. 1–156. 2000;17(4):397–404.
Eckhoff R. Current status and expected future trends in Hassan S, Lau R. Effect of particle formulation on dry
dust explosion research. J Loss Prev Process Indust. powder inhalation efficiency. Curr Pharm Des.
2005;18(4–6):225–37. https://doi.org/10.1016/j.jlp. 2010;16(21):2377–87.
2005.06.012. Hoyer H, Schlocker W, Krum K, Bernkop-Schnürch
Elan Drug Technologies – Commercialized Products. A. Preparation and evaluation of microparticles from
Retrieved December 9, 2010, from http://www. thiolated polymers via air jet milling. Eur J Pharm
elandrugtechnologies.com/nanocrystal_technology/ Biopharm. 2008;69(2):476–85. https://doi.org/10.
commercialised. Accessed 9 Dec 2010. 1016/j.ejpb.2008.01.009.
Eskin D, Zhupanska O, Hamey R, Moudgil B, Scarlett Illig K, Mueller R, Ostrander K, Swanson J. Use of
B. Microhydrodynamic analysis of nanogrinding in microfluidizer processing for preparation of pharma-
stirred media mills. AICHE J. 2005a;51(5):1346–58. ceutical suspensions. Pharmaceut Tech. 1996;20(10):
https://doi.org/10.1002/aic.10392. 78–88.
Ito H. Scale-up theory of single track jet mill. In: and theoretical relationships between the morphology
Proceedings of the 2nd Korea-Japan powder technol- and breakage of crystals. Am Pharmaceut Rev. 2004;7:
ogy symposium, vol 63. 1987. 120–3.
Jacobs C, Kayser O, Müller RH. Nanosuspensions as a Lee J, Lee S, Choi J, Yoo J, Ahn C. Amphiphilic amino
new approach for the formulation for the poorly solu- acid copolymers as stabilizers for the preparation of
ble drug tarazepide. Int J Pharm. 2000;196(2):161–4. nanocrystal dispersion. Eur J Pharm Sci. 2005;24(5):
https://doi.org/10.1016/S0378-5173(99)00412-3. 441–9. https://doi.org/10.1016/j.ejps.2004.12.010.
Kaiser F, Nied R. Modern jet pulverizers. Aufbereitungs Li M, Yaragudi N, Afolabi A, Dave R, Bilgili
Technik. 1980;10:507–14. E. Sub-100 nm drug particle suspensions prepared via
Kalepu S, Nekkanti V. Insoluble drug delivery strategies: wet milling with low bead contamination through
review of recent advances and business prospects. Acta novel process intensification. Chem Eng Sci.
Pharm Sin B. 2015;5(5):442–53. 2015;130:207–20. https://doi.org/10.1016/j.ces.2015.
Keck C, Muller R. Drug nanocrystals of poorly soluble 03.020.
drugs produced by high pressure homogenisation. Eur Lipinski CA. Drug-like properties and the causes of poor
J Pharm Biopharm. 2006;62(1):3–16. https://doi.org/ solubility and poor permeability. J Pharmacol Toxicol
10.1016/j.ejpb.2005.05.009. Methods. 2000;44(1):235–49. https://doi.org/10.1016/
Keck CM, Kobierski S, Mauludin R, Müller RH. Second S1056-8719(00)00107-6.
generation of drug nanocrystals for delivery of poorly Lipinski CA, Lombardo F, Dominy BW, Feeney
soluble drugs: smartCrystals technology. Dosis. PJ. Experimental and computational approaches to
2008;24(2):124–8. estimate solubility and permeability in drug discovery
Kesisoglou F, Panmai S, Wu Y. Nanosizing – oral formu- and development settings. Adv Drug Deliv Rev.
lation development and biopharmaceutical evaluation. 1997;23(1–3):3–25. https://doi.org/10.1016/S0169-
Adv Drug Deliv Rev. 2007;59(7):631–44. https://doi. 409X(96)00423-1.
org/10.1016/j.addr.2007.05.003. Liu T, Müller RH, Möschwitzer JP. Process optimization
Kipp J. The role of solid nanoparticle technology in the of a novel particle size reduction technology H 42. Tag
parenteral delivery of poorly water-soluble drugs. Int J Der Pharmazie, (Abstract P33), 25. 2012.
Pharm. 2004;284(1–2):109–22. https://doi.org/10. Liu T, Müller RH, Möschwitzer JP. Systematical investi-
1016/j.ijpharm.2004.07.019. gation of a combinative particle size reduction technol-
Kipp J, Wong JT, Doty M, Rebbeck C.. Microprecipitation ogy for production of resveratrol nanosuspensions.
method for preparing submicron suspensions. 2002, AAPS PharmSciTech. 2017;18(5):1683–91.
November 28. Retrieved from http://www.google. Liu T, Müller RH, Möschwitzer JP. Production of drug
com/patents/US20020176935 nanosuspensions: effect of drug physical properties on
Kipp JE, Wong JCT, Doty MJ, Rebbeck nanosizing efficiency. Drug Dev Ind Pharm. 2018a;44
CL. Microprecipitation method for preparing submi- (2):233–42.
cron suspensions. (2006, May 2. Retrieved from http:// Liu T, Müller RH, Möschwitzer JP. Consideration of the
www.google.com/patents/US7037528 solid state for resveratrol nanocrystal production. Pow-
Koivisto M, Heinänen P, Tanninen VP, Lehto V. Depth der Technol. 2018b;332:63–9. https://doi.org/10.1016/
profiling of compression-induced disorders and poly- j.powtec.2018.03.028.
morphic transition on tablet surfaces with grazing inci- Liu T, Yu X, Yin H, Möschwitzer JP. Advanced modifi-
dence x-ray diffraction. Pharm Res. 2006;23(4): cation of drug nanocrystals by using novel fabrication
813–20. https://doi.org/10.1007/s11095-006-9785-8. and downstream approaches for tailor-made drug
Kwade A. Wet comminution in stirred media mills – delivery. Drug Deliv. 2019;26(1):1092–103. https://
research and its practical application. Powder Tech. doi.org/10.1080/10717544.2019.1682721.
1999;105(1–3):14–20. https://doi.org/10.1016/S0032- Liversidge E, Wei L. Stabilization of chemical compounds
5910(99)00113-8. using nanoparticulate formulations. 2003. Retrieved
Kwade A. A stressing model for the description and opti- from http://www.google.com/patents/about?
mization of grinding processes. Chem Eng Tech. id¼lJyGAAAAEBAJ. Accessed 20 Mar 2003.
2003;26(2):199–205. https://doi.org/10.1002/ceat. Liversidge G, Cundy K, Bishop J, Czekai D Surface
200390029. modified drug nanoparticles. 1992. Retrieved from
Kwade A, Schwedes J. Breaking characteristics of differ- http://www.google.com/patents/about?
ent materials and their effect on stress intensity and id¼48caAAAAEBAJ&dq¼5,145,684. Accessed
stress number in stirred media mills. Powder Technol. 8 Sep 1992.
2002;122(2):109–21. Merisko-Liversidge E, Liversidge G. Drug nanoparticles:
Kwade A, Schwedes J. Wet grinding in stirred media formulating poorly water-soluble compounds. Toxicol
mills. In: Salman A, Ghadiri M, Hounslow M, editors. Pathol. 2008;36(1):43–8. https://doi.org/10.1177/
Handbook of powder technology, vol 12, Particle 0192623307310946.
breakage. Amsterdam: Elsevier; 2007. p. 251–382. Merisko-Liversidge E, Sarpotdar P, Bruno J, Hajj S,
Lee I, Variankaval N, Lindemann C, Starbuck C. Rotor- Wei L, Peltier N, Rake J, et al. Formulation and
stator milling of APis-empirical scale-up parameters antitumor activity evaluation of nanocrystalline
suspensions of poorly soluble anticancer drugs. Pharm Nanoparticle technology for drug delivery.
Res. 1996;13(2):272–8. New York: Taylor & Francis; 2006. p. 21–51.
Merisko-Liversidge E, Liversidge G, Cooper Murdande SB, Pikal MJ, Shanker RM, Bogner
E. Nanosizing: a formulation approach for poorly- RH. Solubility advantage of amorphous
water-soluble compounds. Eur J Pharm Sci. 2003;18 pharmaceuticals: I. a thermodynamic analysis. J
(2):113–20. https://doi.org/10.1016/S0928-0987(02) Pharm Sci. 2010;99:1254–64. https://doi.org/10.1002/
00251-8. jps.21903.
Midoux N, Hosek P, Pailleres L, Authelin J. Micronization Nakach M, Authelin J, Chamayou A, Dodds
of pharmaceutical substances in a spiral jet mill. Pow- J. Comparison of various milling technologies for
der Tech. 1999;104(2):113–20. https://doi.org/10. grinding pharmaceutical powders. Miner Engg.
1016/S0032-5910(99)00052-2. 2004;74(Suppl 1):S173–81. https://doi.org/10.1016/j.
Moschwitzer J. Method for producing ultrafine minpro.2004.07.039.
submicronic suspensions. 2011, October 11. Retrieved New Food Pharma Systems Spiral Jet Mill: Introduction.
from http://www.google.com/patents/US8034381 2010. Retrieved December 16, 2010, from http://www.
Moschwitzer J, Lemke A. Method for the gentle produc- labx.com/v2/adsearch/detail3.cfm?adnumb¼406756
tion of ultrafine particle suspensions and ultrafine Nied R. Rotor impact mills. In: Salman A, Ghadiri M,
particles and the use thereof. 2008, August 14. Hounslow M, editors. Handbook of powder technol-
Retrieved from http://www.google.com/patents/ ogy, Particle breakage, vol. 12. Amsterdam: Elsevier;
US20080193520 2007. p. 229–49.
Moschwitzer J, Mueller RH. New method for the effective Niwa T, Nakanishi Y, Danjo K. One-step preparation of
production of ultrafine drug nanocrystals. J Nanosci pharmaceutical nanocrystals using ultra cryo-milling
Nanotechnol. 2006;6(9–10):3145–53. https://doi.org/ technique in liquid nitrogen. Eur J Pharm Sci.
10.1166/jnn.2006.480. 2010;41(1):78–85. https://doi.org/10.1016/j.ejps.
Muller, R. H., & Moschwitzer, J. (2015, October 27). 2010.05.019.
Method and device for producing very fine particles Noyes AA, Whitney WR. The rate of solution of solid
and coating such particles. Retrieved from http://www. substances in their own solutions. J Am Chem Soc.
google.com/patents/US9168498 1897;19(12):930–4. https://doi.org/10.1021/
Muller F, Polke R. From the product and process ja02086a003.
requirements to the milling facility. Powder Tech. Peltonen L. Design space and QbD approach for produc-
1999;105(1–3):2–13. https://doi.org/10.1016/S0032- tion of drug nanocrystals by wet media milling
5910(99)00112-6. techniques. Pharmaceutics. 2018;10(3):104.
Muller R, Peters K, Becker R, Kruss B. Nanosuspensions – Peltonen L, Hirvonen J. Pharmaceutical nanocrystals by
a novel formulation for the iv administration of poorly nanomilling: critical process parameters, particle frac-
soluble drugs. In: 1st World Meeting APGI/APV, pp turing and stabilization methods. J Pharm Pharmacol.
491–492, Budapest. 1995. 2010;62(11):1569–79. https://doi.org/10.1111/j.
Muller R, Peters K, Craig D. Electron microscopic studies 2042-7158.2010.01022.x.
of nanosuspensions – particle shapes as a funktion of Petersen R. Nanocrystals for use in topical cosmetic
drug and surfactant. In: 23rd international symposium formulations and method of production thereof. 2015,
of controlled release of bioactive materials. Japan: August 25. Retrieved from http://www.google.com/
Kyoto; 1996. p. 925–6. patents/US9114077
Muller R, Becker R, Kruss B, Peters K. Pharmaceutical Qiao Y, Cao Y, Yu K, Zong L, Pu X. Preparation and
nanosuspensions for medicament administration as antitumor evaluation of quercetin nanosuspensions
systems with. 1999. Retrieved from http://www. with synergistic efficacy and regulating immunity. Int
google.com/patents/about?id¼TIoXAAAAEBAJ& J Pharm. 2020;589:119830. https://doi.org/10.1016/j.
dq¼5,858,410. Accessed 12 Jan 1999. ijpharm.2020.119830.
Muller RH, Jacobs C, Kayser O. Nanosuspensions as Rabinow B. Nanosuspensions in drug delivery. Nat Rev
particulate drug formulations in therapy: rationale for Drug Discov. 2004;3(9):785–96. https://doi.org/10.
development and what we can expect for the future. 1038/nrd1494.
Adv Drug Deliv Rev. 2001;47(1):3–19. https://doi.org/ Rabinow BE, Gupta P, Wong J, Papadopoulos P, Chaubal
10.1016/S0169-409X(00)00118-6. M. Formulation of water-insoluble antineoplastic
Muller R, Jacobs C, Kayser O. DissoCubes: a novel for- agents as nanosuspensions using NANOEDGE formu-
mulation for poorly soluble and poorly bioavailable lation technology. Cancer Res. 2004;64(7 Suppl):146.
drugs. In: Rathbone M, Hadgraft J, Roberts M, editors. Rasenack N, Müller B. Micron-size drug particles: com-
Modified-release drug delivery technology. 1st mon and novel micronization techniques. Pharm Dev
ed. New York: Marcel Dekker; 2003. p. 135–49. Technol. 2004;9(1):1–13. https://doi.org/10.1081/
Muller R, Moschwitzer J, Bushrab F. Manufacturing of PDT-120027417.
nanoparticles by milling and homogenization Reed R. Trends and advances in cryogenic materials. In:
techniques. In: Gupta R, Kompella U, editors. Timmerhaus K, Reed R, editors. Cryogenic
engineering: fifty years of progress. Dordrecht: Simonelli AP, Mehta SC, Higuchi WI. Inhibition of
Springer; 2007. p. 52–83. sulfathiazole crystal growth by polyvinylpyrrolidone.
Rohrs BR, Amidon GE, Meury RH, Secreast PJ, King J Pharm Sci. 1970;59(5):633–8. https://doi.org/10.
HM, Skoug CJ. Particle size limits to meet USP con- 1002/jps.2600590512.
tent uniformity criteria for tablets and capsules. J Singh S, Naini V. Homogenization and homogenizers. In:
Pharm Sci. 2006;95(5):1049–59. https://doi.org/10. Swarbrick J, editor. Encyclopedia of pharmaceutical
1002/jps.20587. technology. 3rd ed. New York: Informa Healthcare;
Salazar J, Heinzerling O, Müller RH, Möschwitzer 2006. p. 1996–2003.
JP. Process optimization of a novel production method Sinha B, Müller RH, Möschwitzer JP. Systematic investi-
for nanosuspensions using design of experiments gation of the cavi-precipitation process for the produc-
(DoE). Int J Pharm. 2011;420(2):395–403. https://doi. tion of ibuprofen nanocrystals. Int J Pharm. 2013;458
org/10.1016/j.ijpharm.2011.09.003. (2):315–23. https://doi.org/10.1016/j.ijpharm.2013.
Salazar J, Ghanem A, Müller RH, Möschwitzer 10.025.
JP. Nanocrystals: comparison of the size reduction Skelton R, Khayyat A, Temple R. Fluid energy milling–an
effectiveness of a novel combinative method with con- investigation of micronizer performance. Fine particles
ventional top-down approaches. Eur J Pharm processing. 1980;1:113–25.
Biopharm. 2012;81(1):82–90. https://doi.org/10.1016/ SkyePharma–Insoluble Drug Delivery Platform. Retrieved
j.ejpb.2011.12.015. December 9, 2010., from http://www.skyepharma.
Salazar J, Müller RH, Möschwitzer JP. Performance com- com/Technology/Oral_Technology/Particle_Engineer
parison of two novel combinative particle-size-reduc- ing_Technologies/Insoluble_Drug_Delivery_Plat
tion technologies. J Pharm Sci. 2013;102(5):1636–49. form/Default.aspx?id¼80. Accessed 9 Dec 2010.
https://doi.org/10.1002/jps.23475. Smit W. Jetmilling of heat sensitive materials, world congr
Salazar J, Müller RH, Möschwitzer JP. Combinative parti- part technol, part II. Nurnberg: Comminution; 1986.
cle size reduction technologies for the production of Spencer R, Dalder B. Sizing up grinding mills. Chem Eng.
drug nanocrystals. J Pharmaceut. 2014, Article ID 1997;104(4):84–7.
265754. Stairmand CJ. The energy efficiency of milling processes:
Schilde C, Nolte H, Arlt C, Kwade A. Effect of fluid- a review of some fundamental investigations and their
particle-interactions on dispersing nano-particles in application to mill design. Vol. 1975;79:1–17.
epoxy resins using stirred-media-mills and three-roll- Steckel H, Markefka P, teWierik H, Kammelar R. Effect of
mills. Compos Sci Technol. 2010;70(4):657–63. milling and sieving on functionality of dry powder
https://doi.org/10.1016/j.compscitech.2009.12.021. inhalation products. Int J Pharm. 2006;309(1–2):
Schlocker W, Gschließer S, Bernkop-Schnürch 51–9. https://doi.org/10.1016/j.ijpharm.2005.10.043.
A. Evaluation of the potential of air jet milling of Stein J, Fuchs T, Mattern C. Advanced milling and con-
solid protein-poly(acrylate) complexes for microparti- tainment technologies for superfine active pharmaceu-
cle preparation. Eur J Pharm Biopharm. 2006;62(3): tical ingredients. Chem Eng Tech. 2010;33(9):
260–6. https://doi.org/10.1016/j.ejpb.2005.09.001. 1464–70. https://doi.org/10.1002/ceat.200900590.
Schurr GA, Zhao QQ. Fluid mechanic considerations for Stenger F, Mende S, Schwedes J, Peukert W. Nanomilling
fine grinding in a fluid energy mill., 1994, p. 536. in stirred media mills. Chem Eng Sci. 2005;60(16):
Schwenzfeuer K, Glor M, Gitzi A. Relation between igni- 4557–65. https://doi.org/10.1016/j.ces.2005.02.057.
tion energy and limiting oxygen concentration for Stevenson B. Preventing disaster: analyzing your plant’s
powders. 2001. p. 909. dust explosion risks. Powder Bulk Eng. 2001;15(1):
Sharma P, Denny W, Garg S. Effect of wet milling process 19–27.
on the solid state of indomethacin and simvastatin. Int J Sugimoto S, Niwa T, Nakanishi Y, Danjo K. Development
Pharm. 2009;380(1–2):40–8. https://doi.org/10.1016/j. of a novel ultra cryo-milling technique for a poorly
ijpharm.2009.06.029. water-soluble drug using dry ice beads and liquid
Sharma P, Zujovic ZD, Bowmaker GA, Denny WA, Garg nitrogen. Int J Pharm. 2012a;426(1–2):162–9. https://
S. Evaluation of a crystalline nanosuspension: poly- doi.org/10.1016/j.ijpharm.2012.01.007.
morphism, process induced transformation and in vivo Sugimoto S, Niwa T, Nakanishi Y, Danjo K. Novel ultra-
studies. Int J Pharm. 2011;408(1):138–51. https://doi. cryo milling and co-grinding technique in liquid nitro-
org/10.1016/j.ijpharm.2011.01.032. gen to produce dissolution-enhanced nanoparticles for
Shegokar R, Müller RH. Nanocrystals: industrially feasi- poorly water-soluble drugs. Chem Pharm Bull.
ble multifunctional formulation technology for poorly 2012b;60(3):325–33. https://doi.org/10.1248/cpb.
soluble actives. Int J Pharm. 2010;399(1–2):129–39. 60.325.
https://doi.org/10.1016/j.ijpharm.2010.07.044. Tangsathitkulchai C. Effects of slurry concentration and
Shoyele S, Cawthorne S. Particle engineering techniques powder filling on the net mill power of a laboratory ball
for inhaled biopharmaceuticals. Adv Drug Deliv Rev. mill. Powder Tech. 2003;137(3):131–8. https://doi.
2006;58(9–10):1009–29. https://doi.org/10.1016/j. org/10.1016/j.powtec.2003.08.048.
addr.2006.07.010. Trasi NS, Boerrigter SXM, Byrn SR. Investigation of the
milling-induced thermal behavior of crystalline and
amorphous griseofulvin. Pharm Res. 2010;27(7): Wilén C, Moilanen A, Rautalin A, Torrent J, Conde E,
1377–89. https://doi.org/10.1007/s11095-010-0129-3. Lodel R, Carson D, et al. Safe handling of renewable
Troester F. Cremophor-free aqueous paclitaxel fuels and fuel mixtures. Technical Research Centre of
nanosuspension production and chemical Finland (VTT); 1999.
stability. 2004. Witkin D, Lavernia E. Synthesis and mechanical behavior
Van Eerdenbrugh B, Van den Mooter G, Augustijns of nanostructured materials via cryomilling. Prog
P. Top-down production of drug nanocrystals: Mater Sci. 2006;51(1):1–60. https://doi.org/10.1016/j.
Nanosuspension stabilization, miniaturization and pmatsci.2005.04.004.
transformation into solid products. Int J Pharm. Zhang GGZ, Law D, Schmitt EA, Qiu Y. Phase transfor-
2008;364(1):64–75. https://doi.org/10.1016/j.ijpharm. mation considerations during process development and
2008.07.023. manufacture of solid oral dosage forms. Adv Drug
Ward GH, Schultz RK. Process-induced crystallinity Deliv Rev. 2004;56(3):371–90. https://doi.org/10.
changes in albuterol sulfate and its effect on powder 1016/j.addr.2003.10.009.
physical stability. Pharm Res. 1995;12(5):773–9.
Cosolvent and Complexation Systems
5
Junhuang Jiang and Robert O. Williams III
Abstract common manufacturing method. Semisolid

Cosolvent and polyethylene glycol (PEG)- and solid-solubilized formulations that are liq-
based solubilization techniques for the deliv- uid at a higher temperature (50–70 C) can be
ery of poorly soluble drugs are discussed in encapsulated into hard gelatin capsules as mol-
this chapter. The properties of excipients and ten liquids at elevated temperature. Semisolid
the physicochemical principles are presented or solid matrices are formed inside the
for formulating each type of the solubilized capsules when the molten materials are cooled
formulations. Cosolvents are commonly used to ambient temperature. Spray congealing and
in combination with surface-active solubilizers fluidized bed melt granulation are alternative
to increase the solubilizing capacity and to manufacturing processes to convert the
improve the in vivo emulsification of self- solubilized formulations with high melting/
emulsifying formulations. In PEG-based softening points into granules that can be read-
delivery systems, drug is either dispersed as ily processed into capsules or tablets. Pow-
micronized crystalline particles (via the forma- dered solution technology can also be applied
tion of eutectic mixtures) or present in its to transform the solubilized formulation of
amorphous state. Improvement in absorption low-dose drug into free flowing powder by
from a PEG matrix is due to (1) fast dissolution absorbing the formulation into solid carriers.
rate of drug from the dosage forms and In addition, the interest in using cyclodextrins
(2) higher transient solubility of the drug sub- (CDs) for drug solubilization has proven ben-
stance in gastrointestinal tract. eficial for delivery of poorly water-soluble
Various manufacturing techniques to prodrugs through the formation of inclusion
cess the solubilized formulations into oral dos- complexes. This chapter provides an update
age forms are also discussed in this chapter. with studies of drug–CD inclusion complexes,
For the formulations that are liquid under characterization of the complexes, and
ambient conditions, encapsulation into soft examples of commercial products containing
gelatin or hard gelatin capsules is the most CDs. The current authors would like to thank
and acknowledge the previous authors’ sig-
nificant contributions from the first and sec-
ond editions. This current third edition
chapter is a revision and update of the origi-
J. Jiang · R. O. Williams III (*)
nal authors’ work from the first and second
Division of Pharmaceutics, College of Pharmacy,
The University of Texas at Austin, Austin, TX, USA editions.
180 J. Jiang and R. O. Williams III

Keywords ΔH f Tm T
Log X 2 ¼ 2 303RT
2 Tm
Cosolvents · Polyethylene glycol (PEG) ·
Cyclodextrins (CDs) · Solubilized V 2 ϕ21
þ 2 303RT
formulations · Crystalline nonelectrolyte · 2
2
Gastrointestinal (GI) tract · Complexation · δ1 2 þ δ2 2 2W , ð5:1Þ
Soft gel capsules · Hard gelatin capsules
where 1 stands for solvent, while 2 stands for
solute. X 2 is the mole fraction of the solute,
5.1 Introduction ΔHf is the molar enthalpy of melting, R is the
gas constant, Tm is the melting point of the
In solubilized formulations, excipients function solute, T is the absolute temperature (kelvin) at
as solvents to maintain drug molecules in the which the solubility is determined, V2 is the
solution state in the final dosage forms. Applica- molar volume of the liquid solute, Φ1 is the
tion of solubilized formulations for enabling volume fraction of the solvent, and δ is the
delivery of poorly water-soluble drugs has solubility parameter.
attracted substantial research interest for develop- The solubility parameter is defined as the
mental compounds and has been successfully square root of the energy of vaporization per
applied to a number of low-solubility commercial unit volume. It can be calculated using a group
products. This chapter focuses on the following contribution method. W is the interaction energy
three types of formulations: cosolvent and poly- between the solute and solvent. The value of
ethylene glycol-based formulations. Solubilized interaction energy cannot be calculated from fun-
formulations based on polymeric solid dispersion, damental physicochemical properties of the com-
self-emulsifying formulation, and drug- pound. However, W could be expressed as a
cyclodextrins complexation technologies are cov- power series of the solubility parameter of
ered in other chapters of this book. Properties of cosolvent:
commonly used excipients for each type of
W ¼ C0 þ C1 δ1 1 þ C2 δ1 2 þ C 3 δ1 3
formulations are presented in this chapter. Formu-
lation development and selection of the þ C4 δ1 4 þ . . . ð5:2Þ
manufacturing process for the finished dosage
forms are discussed in this chapter as well. As shown in (5.1), low solubility of nonelec-
trolyte can be attributed to two factors: high melt-
ing point of the solute (represented in the first
5.2 Theoretical Modeling term on the right-hand side of the equation) or
of Solubility in Cosolvents low affinity between solute and solvent
(represented in the second term on the right-
The process of solubilizing a crystalline nonelec- hand side of the equation). The maximum solu-
trolyte in a solvent thermodynamically consists of bility occurs when the intermolecular energy is
two steps: dissociation of the solute and mixing the same for solute–solvent and solvent–solvent,
between the solute and solvent molecules. Noting and the second term on the right-hand side of the
that the dissociation process is analogous to melt- equation is equal to zero.
ing, the solubility of a crystalline nonelectrolyte Yalkowsky et al. derived the log linear equa-
solute in a solvent is a function of the melting tion (5.3) that describes the solubility of drug in a
point of the solute and solute–solvent affinity. binary aqueous system (1972):
The extended Hildebrand equation ((5.1), Martin LogS ¼ LogSw þ σF, ð5:3Þ
et al. 1982) has been derived to predict the solu-
bility of drug in pure solvents and binary mixture where S is the solute solubility (molar concentra-
of cosolvents: tion) in water and cosolvent mixture, Sw is the
5 Cosolvent and Complexation Systems 181
solubility in water, F is the volume fraction of experimentally measured. Log linear regression
cosolvent, and σ, the linear regression coefficient, of the solubility data can then be used to predict
represents the solubilizing power of the solubility in cosolvent. The linear regression
cosolvent. This equation works very well when coefficient (slope of log linear regression) is
the water–cosolvent system is more polar than the indicative of the solubilizing power of the
solute. As a general rule of thumb, the solubility cosolvent.
parameter of solute needs to be at least three units Moore (1958) reported that different cosolvent
lower than that of water–cosolvent mixture for the mixtures with the same approximate dielectric
log linear regression to be applicable. requirement (ADR) could solubilize drug to the
The solubility of poorly water-soluble drug in same extent. Approximate dielectric requirement
several different levels of cosolvent is can be calculated using (5.4):
Xn
ðWt:%SolventÞ Dielectric Constanti
ADR ¼ : ð5:4Þ
t¼1
100
When changing to a different cosolvent sys- 100

tem, the Moore equation can be used to identify a MPD ¼
N
good starting percent for the new cosolvent. In X jCalculated Observed j
order to evaluate the correlations among physico- ð5:6Þ
Observed
chemical properties, the Jouyban–Acree model
was widely applied to various pharmaceutical It has been reported that 27 drug compounds
cosolvent systems (Jouyban and Acree Jr 2018). were used to predict solubilities by a trained
The equation predicting the solubility of a solute Jouyban–Acree model, and the result showed
in a water cosolvent at different temperature is that the mean MPD of predicted solubilities was
shown below: found to be 24% (Jouyban 2007).
log X m,T ¼ w1 ∙ log X 1,T þ w2 ∙ log X 2,T

5.3 Oral Absorption of Drug
w ∙w X
2
Substances
þ 1 2∙ Ji ∙ ðw1 w2 Þi
T i¼0
As shown in Fig. 5.1, orally administered drug
ð5:5Þ
can be absorbed in gastrointestinal (GI) tract
where Xm, T, X1, T, and X2, T are the mole fractions through three different pathways: passive
of the drug in the mixed solvent, neat cosolvent, transcellular diffusion, active transcellular trans-
and water at temperature T (in Kelvin), w1 and w2 port, and passive paracellular diffusion. These
are the mass fractions of the pure cosolvent and three absorption mechanisms could occur simul-
water, and Ji is a constant. This model has the taneously. For most drugs, passive transcellular
advantage of describing mole fraction solubility transport through a highly lipophilic GI mem-
of solutes dissolved in solvent mixtures as a func- brane is the primary absorption mechanism.
tion of solvent fraction. The mean percentage Fick’s law of diffusion governs passive
deviations (MPD) were used to evaluate the accu- transcellular transport, where the diffusion rate
racy of the prediction model and are calculated is directly proportional to the concentration gra-
using Eq. (5.6). dient across the membrane. Intrinsic diffusion
Fig. 5.1 GI membrane

transport for drug
absorption (Lobenberg and
Amidon 2000)
coefficient is dependent on the lipophilicity and Drug absorbed in the GI tract can enter either
size of drug molecules. The optimal LogP for blood capillaries or lymphatic capillaries. Most
passive diffusion is 2–9. For ionizable drugs, it drugs are absorbed into the portal vein because
is the unionized species that permeates through the fluid flow rate for blood capillaries is about
GI membrane via passive transcellular pathway. 500 of that for lymphatic capillaries. Lymphatic
Active transport is selective, could take place absorption has several advantages over the portal
against a concentration gradient, and is limited vein absorption. The fraction of drug absorbed
to drugs that are structurally similar to endoge- through lymphatic system is protected from first
nous substances such as vitamins and sugars. pass metabolism. For drug metabolized on first
Active transport can only occur at specific site in pass through the liver, lymphatic transport
GI tract. Passive paracellular diffusion occurs improves oral bioavailability. Surfactants, such
through water-filled intercellular channels. In as Tween 80, Cremophor EL, and vitamin E
human, surface area of intestine available for TPGS that are used in self-emulsifying
paracellular diffusion only accounts for about formulations, are known to improve drug bio-
0.01% of total intestinal surface area. The channel availability by inhibition of P-glycoprotein-
diameter is in the range of 3–10 Å. Therefore, mediated efflux (Collnot et al. 2007) and by inhi-
paracellular diffusion is only applicable to small bition of CYP3A metabolism (Wu and Benet
molecules with a molar mass less than 200 dalton. 2005).
For BCS class II compounds, drug absorption
is limited by the solubility in GI tract, while class
IV compounds are limited by solubility and 5.4 Cosolvent-Based Formulations
absorption constraints. The amount of drug
absorption is proportional to drug concentration Cosolvents are water-miscible organic solvents
time permeability. Drug concentration in GI tract used to improve the solubility of poorly water-
is a function of both solubility and dissolution soluble drugs. They are used to support the pre-
rate. Drugs are dispersed at molecular level in clinical studies in animal models to determine the
the solubilized formulations. When the hydro- pharmacokinetic profiles and toxicity profile of
philic solubilized formulation comes into the con- compounds in the drug discovery stage where
tact with aqueous environment in GI tract, the exposure from a conventional crystalline system
dissolution rate is controlled by the disintegration may be insufficient to elicit a response. It has been
rate of the formulation. As the excipient matrix studied extensively that cosolvency can be used
dissolves, solubilized drug molecules go into the in solid or liquid formulations. The cosolvent
aqueous environment and stay supersaturated formulation can be administrated orally and par-
over a period of time. This fast dissolution and enterally. Commonly used cosolvents for preclin-
transient supersaturation allows for improved ical studies are ethyl alcohol, propylene glycol,
absorption. low-molecular-weight PEG, glycerin,
Table 5.1 Properties of commonly used solvents at 25 C

Solvent Dielectric constant (ε) Solubility parameter (Cal/cm3)1/2 Surface tension (dynes/cm)
Water 80.1 23.50 72.8
Glycerin 42.5 21.10 64.0
Propylene glycol 32.1 11.97 40.1
Ethanol 24.5 12.92 22.3
PEG 400 12.4 11.30 44.5
dimethylsulfoxide, and dimethylacetamide. How- constant of the medium (Zhao and Yalkowsky
ever, only PEG, ethanol, propylene glycol, and 2001).
glycerin demonstrate acceptable safety profile in Cosolvents are also commonly used in oral
long-term use to be included in commercial dos- solutions and soft gelatin capsules for poorly sol-
age forms (Nema et al. 1997). Physical properties uble compounds. The advantage of oral solution
of these four cosolvents are presented in is that it is easier to administer solution to pediat-
Table 5.1. Glycerin and propylene glycol are ric patients and adults having difficulty in
more polar cosolvents than ethanol and liquid swallowing. Sweetening agents are needed for
PEG. To achieve the optimum solubilizing capac- solutions containing propylene glycol which is
ity, cosolvents are often combined to yield a known for its bad taste. In the case of self-
similar polarity to the drug substance. emulsifying formulations, the presence of
cosolvents not only improves the solubility of
drug in the formulation but also facilitates the
5.4.1 Development emulsification in aqueous medium. In combina-
of Cosolvent-Based Solubilized tion with vitamin E TPGS, PEG-400 and propyl-
Formulations ene glycol are present in oral solution and soft
gelatin capsule products of amprenavir
Cosolvents are widely used to prepare (Agenerase®). Self-emulsifying soft gelatin
concentrated drug solutions in parental and oral capsules of cyclosporine (Sandimmune®) contain
solution products. For parental drug delivery, the glycerol, propylene glycol, and ethanol as
concentrated drug solutions are often diluted with cosolvents. Cosolvents facilitate the emulsifica-
saline solution and delivered to patient via IV tion process via “diffusion and stranding” mecha-
infusion. Ethanol and propylene glycol alone or nism (Pouton 1997). Cosolvents dissolve the
in combination are the most common cosolvent water-soluble component into aqueous environ-
systems in injectable products (Nema et al. 1997). ment as a solubilized system. With further migra-
The amount of ethanol or propylene glycol in the tion, cosolvent gets diluted and can no longer
formulation ranging from 0.2% to 80% glycerin maintain a single phase consisting of water, oil,
has been used in injectable at level as high as and cosolvent. As a result, the oil phase separates
80%. PEG 300 and PEG 400 are used in out as fine oil droplets.
injectable products at level as high as 65%. Poly- Recently, novel cosolvent excipients have
ethylene glycol undergoes auto-oxidation emerged, expanding the formulation possibilities
(Hamburger et al. 1975), and the peroxides of cosolvent systems. For example, diethylene
generated can induce drug degradation. There- glycol monomethyl ether (DEGEE), known by
fore, application of PEG as a cosolvent for its trade-name Transcutol®, was previously used
injectable products is limited. Cosolvents not for food and personal-care products. It was found
only need to achieve higher solubility but also to have potential as an effective solubilizer and
need to prevent drug precipitation upon dilution. permeation enhancer for different administration
Cosolvents can also be used to suppress chemical routes (Ha et al. 2020). DEGEE is a polar solvent
degradation of drugs by altering the dielectric that acts as a proton donor. Such characteristics
enable it to have good miscibility with polar and solubilization. Solid dispersion is a drug–polymer
nonpolar solvents, resulting in the dissolving of mixture where hydrophobic drugs are dispersed
hydrophobic drugs. Several studies showed that into hydrophilic polymer matrices. Based on the
DEGEE could effectively improve the solubilities final state of drugs, there are crystalline solid
of small molecule drugs such as carbothioamide, dispersions and amorphous solid dispersions.
glibenclamide, ibrutinib, isatin, and vanillin Fusion and solvent-based are two main
(Ha et al. 2020). Due to the presence of three approaches to prepare solid dispersion. It will be
hydrogen bond donors and one acceptor, discussed in the following sections. In recent
DEGEE can reduce the intermolecular attractions years, several studies show the solubilization effi-
between water molecules and improve the ciency of PEG-based solid dispersions. As a
miscibility with water. According to the FDA’s herbal drug used to rectify hepatic ailments,
inactive ingredient database, the maximum Silybum marianum exhibited a low oral absorp-
amount of DEGEE used for topical and transder- tion rate due to the low-aqueous solubility. A
mal administration is 49.91% (w/w) and 5% solid dispersion consisting of PVP K-30, PEG
(w/w), respectively (Ha et al. 2020). Some clini- 6000, and drug was prepared by solvent-evapora-
cal studies have established this excipient’s safety tion method to address this issue. This drug
and efficacy; however, further research is still exhibited high solubility in such polymer matrix,
needed (Osborne 2011). with an 1150-fold higher solubility than before
(Yousaf et al. 2019). In another study, a mixture
of PEG 4000 and PEG 6000 was combined with
5.5 Polyethylene Glycol-Based glyburide, a hypoglycemic drug with low solubil-
Solid Dispersion ity and dissolution rate. The result showed the
solid dispersion containing two different PEGs
Polyethylene glycol (PEG)-based solid dispersion performed a 12-fold increase in dissolution com-
for poorly water-soluble drugs has attracted a pared to the crystalline bulk of the drug (Betageri
great deal of research interest over the last and Makarla 1995). Classification of drug–PEG
30 years. PEG is an ideal excipient for the prepa- dispersions is presented in Table 5.2 (Craig
ration of hydrophilic solubilized formulation. It 1990). Improved bioavailability from these
has low enthalpy of fusion (2.5 kJ/mol), low systems is attributed to the increase in surface
melting point, and low viscosity in molten state. area of drug substance, solubilization, and
It also has good solvent capacity in its molten improved surface wetting of drug particles in the
state. The relatively high polarity of PEG will microenvironment with concentrated PEG (Chiou
increase the hydrophobic–hydrophilicity and Riegelman 1971).
interactions with drugs, which is a mechanism Figure 5.2 summarizes PEG’s applications
of solubilization. In addition, it is difficult for based on its molecular weight (U. S. Food and
poorly water-soluble drugs to break the lattice of Drug Administration 2021). High-molecular-
water; therefore, cosolvents like PEG can facili- weight PEG has an average molar mass ranging
tate the process of solubilizing by decreasing the from 3350 to 8000. PEGs with such high molec-
polarity of the solvent system (D’souza and ular mass can be used for different purposes, such
Shegokar 2016). PEG in a liquid state (from as ointment bases, cosmetics, lubrication, and
PEG 200 to 600) is typically used as a solubilizer antifouling agents for proteins. These PEGs are
for oral or parenteral delivery. It has also been manufactured by The Dow Chemical Company
applied to other administrations, including oph- and sold under the trade name of Carbowax™
thalmic or otic drops and soft gelatin capsules Sentry™. It is semicrystalline material with a
(D’souza and Shegokar 2016). It has been high degree of crystallinity. With melting points
reported that PEG is widely used in forming ranging from 45 to 65 C, PEGs of high molecu-
solid dispersion/solid solutions to improve lar weight are solid at room temperature. Above
Table 5.2 Classification of drug dispersions in PEG

Classification Properties
Eutectic mixture Negligible solid–solid solubility.
Thermodynamically, the composition is an intimately blended physical mixture of drug crystals
and PEG crystals.
The system is thermodynamically stable.
Solid solution Drug molecules could also be present in the parallel helix unit cell of the crystalline phase of
PEG.
Covalent nature of PEG crystal lattice makes the solubilization of drug molecules in PEG
crystalline phase difficult.
Glass solution Drug molecules are predominately solubilized in the amorphous region of PEG semicrystalline
structure.
The system is not thermodynamically stable. Crystallization of drug can take place when the
composition is exposed to high-humidity and high-temperature environment.
Amorphous Drug and PEG exist in two different phases.
precipitates Drug phase is amorphous.
The system is not thermodynamically stable. Crystallization of drug can take place when the
composition is exposed to high-humidity and high-temperature environment.
Fig. 5.2 Various applications of PEG with different molecular weights
the melting point, PEG becomes a low viscosity 5.5.1 Development of PEG-Based
liquid. At 80 C, viscosity of molten PEG ranges Solubilized Formulations
from 150 (PEG 3350) to 1500 cP (PEG 8000). In
the solid state, PEG forms lamellar structure with PEG is a large molecule with a covalent crystal
alternating crystalline and amorphous regions. lattice. In contrast, most drugs form molecular or
The crystal lattice of PEG consists of two parallel ionic crystals. The lattice mismatch between PEG
helices in a unit cell. and drug makes the formation of a true solution
difficult. When the drug does not interact with weight PEG (PEG 4000) (Unga et al. 2009). Law
PEG strongly, a two-phase solid is formed. The et al. (2003) have developed amorphous PEG
two-phase system could be a simple eutectic mix- 8000-ritonavir solubilized solution. High Tg/Tm
ture, which is thermodynamically stable mixture ratio indicates that ritonavir is a good glass for-
of intimately blended crystalline drug domains mer. Amorphous ritonavir is reasonably stable at
and semicrystalline PEG domains. Enhanced ambient conditions. Solubility of crystalline rito-
drug absorption from eutectic mixture is navir improved in the presence of PEG markedly.
attributed to the large surface area and improved Solubility of ritonavir in the amorphous state is
wetting of micronized or submicron drug crystals. ten times the solubility of crystalline drug. PEG
For formulations of eutectic mixtures, molten was found to have negligible effect on the glass
PEG must have good solubilization capacity for transition temperature of ritonavir. The formula-
the drug substance. In the melt method, tion was moisture sensitive. Crystallization of
processing drug and PEG at the eutectic point amorphous ritonavir was observed when the com-
allows for the low processing temperature to min- position was exposed to high humidity environ-
imize drug degradation. Eutectic mixture formu- ment. However, the formulation was chemically
lation strategy has been successfully applied to and physically stable when the composition was
develop rapidly dissolving mixtures of PEG 8000 protected from the moisture with proper packag-
and fenofibrate, which resulted in a tenfold ing configuration. Even at 30% drug loading,
increase in fenofibrate dissolution (Law et al. ritonavir solid dispersion in PEG 8000 was stable
2003). Differential scanning calorimetry, hot under dry condition for >1.5 years. The presence
stage microscopy, and variable temperature of ritonavir did not affect the PEGs 8000 enthalpy
X-ray diffraction techniques can be used to deter- of fusion. It was concluded that the solid disper-
mine the phase diagram of these systems and sion consisted of two phases: crystalline PEG
were applied to the PEG 8000 and fenofibrate 8000 and amorphous phase comprising a mixture
composition. Crystalline drug and semicrystalline of amorphous ritonavir and amorphous PEG.
PEG phases were identified in samples across all Good stability of the ritonavir-PEG 8000-
drug loadings. The eutectic point was identified to solubilized formulation was attributed to the
be at 20–25% drug loading. The drug–PEG eutec- intrinsic stability of amorphous ritonavir and the
tic crystallization led to the formation of irregular stabilization effect of PEG.
microstructures. Fenofibrate crystal in the micro- Due to the unique crystalline structure and
structure was less than 10 μm. high crystallinity of PEG, formation of interstitial
For drug substances that demonstrate good solid between PEG and drug is not common and
glass-forming properties (Tg/Tm (in kelvin) > 2/ can only occur at low drug loading. When drug is
3) and strong interaction with PEG, a physically present in the crystal lattice of PEG, changes in
stable solid dispersion consisting of amorphous thermal properties of PEG, such as lower melting
drug phase and semicrystalline PEG phase could point and low enthalpy of melting, can be
be successfully prepared. Since the only differ- observed. Interstitial solid solution of clofibrate
ence in the liquid PEG and solid PEG is the in high-molecular-weight PEG has been reported
degree of polymerization, interaction between (Anguiano-Igea et al. 1995). In the presence of
drug and low-molecular-weight liquid PEG is clofibrate, PEG melting point shifted to a lower
indicative of the interaction between drug and temperature, and the broadening of the melting
high-molecular-weight PEG. In the presence of peak was also observed. Drug release rate
small amount of high-molecular-weight PEG, increased with an increase in the molecular
improvement in the solubility of drug substance weight of PEG and PEG/drug ratio.
in aqueous medium is also a good indicator of When drug is dispersed at the molecular level
drug and PEG interaction. Solubility of parabens in PEG matrices, drug molecules can be dissolved
in liquid PEG (PEG 400) has been used to predict either in the amorphous domain or in the helix
the interaction between parabens and high-molar- interstitial space of crystalline domain. It is
believed that most of drug molecules are present storage which limits the effectiveness of
in the amorphous domain of PEG. As pointed quenching. Ternary systems comprising poorly
previously, the amount of drug that could be soluble drug, high-molecular-weight PEG, and
truly dispersed at molecular level in PEG-based stabilizer have been investigated to increase the
solid dispersion is limited since PEG is a highly amorphous content of PEG to improve the drug
crystalline material. The limited commercial suc- loading and to prevent drug from crystallizing
cess with PEG-based solid dispersions is during the storage. Bley et al. (2010) used
attributed to (1) low drug-loading capacity as the Povidone and Copovidone to stabilize nifedipine
result of high crystallinity of PEG and (2) poor in PEG 1500. The stabilized drug/polymer/PEG
chemical stability of drug substances associated dispersion demonstrated more consistent dissolu-
with the system as the result of PEG-induced tion characters, compared to PEG solid
oxidation. Gris-PEG® (griseofulvin in PEG dispersions, which contained a higher amount of
400 and PEG 8000 mixture) and Aprical® (Nifed- crystalline drug. Inclusion of sodium lauryl sul-
ipine in PEG 300 and PEG 6000 mixture) are the fate in naproxen-PEG 4000 solid dispersion fur-
only two market products containing PEG-based ther enhanced the dissolution properties of the
solid dispersion. Gris-PEG® is a eutectic mixture solid dispersion. After 30 months of storage at
of griseofulvin and PEG. Drug is present as ultra- ambient conditions, there was no change in phys-
microsize crystals in Gris-PEG®. Nifedipine is icochemical characteristics and the dissolution
solubilized in Aprical®. Nifedipine is highly sol- properties of solid dispersion (Mura et al. 1999).
uble in both the liquid low-molecular-weight When preparing a new PEG-based formula-
PEG and the melt of high-molecular-weight tion, it is critical to consider the amount of PEG
PEG (Hohne et al. 1990). When high-molecular- used and the limits according to FDA and Inactive
weight PEG is used, nifedipine dissolves easily in Ingredient Guide (IIG). According to this guid-
molten PEG. However, nifedipine crystallizes out ance, different routes of administration may have
of the formulation when molten PEG solidifies at some differences in the limits of the amount of
ambient conditions. The product was successfully PEG used (U. S. Food and Drug Administration
developed when a mixture of liquid PEG and 2021).
solid PEG is used. Incorporation of PEG
300 increases the amorphous content of the PEG
matrices, and no crystallization of nifedipine is 5.5.2 Fusion Method
observed during the storage. for the Preparation of Solid
Pharmaceutical scientists have been studying Dispersion
different formulation and process approaches to
decrease the crystallinity of high-molecular- Drug is dissolved in molten PEG in fusion
weight PEG in order to improve the drug-loading method. The molten mass solidifies when it
capacity of PEG-based drug delivery systems. cools to ambient temperature. Simplicity of the
Polymers and surface-active excipients have manufacturing process is the biggest advantage of
been used to inhibit the crystallization of drug fusion method. Degradation of the drug at the
substances. Strong hydrogen bonding or hydro- elevated processing temperature limits fusion
phobic interactions between these excipients and methods application to thermally stable
drug molecules reduce the mobility of drug compounds. For thermally labile drugs that form
molecules and hinder the molecular packing in a eutectic mixture with PEG, a mixture of PEG
crystal lattice. The same principles could be used and drug at eutectic ratio can be used to reduce the
to reduce the crystallinity of high-molecular- processing temperature since the eutectic mixture
weight PEG. Shock freezing of PEG melt has melts at a temperature much lower than either of
been studied to inhibit crystallization of PEG; the melting points of the individual components.
however, due to the low glass transition tempera- Various techniques, such as pouring the molten
ture (60 C), recrystallization occurs during mass to metallic plate and quenching the molten
mass with liquid nitrogen, have been explored in However, the difficulties associated with organic
the laboratories. Alternatively, the molten mass solvents handling and complete removal of the
can be filled directly into hard shell capsules organic solvents present unique challenges with
(gelatin or hydroxypropyl methylcellulose solvent method.
(HPMC) capsules). Crystallization of PEG is A dispersion of piroxicam and PEG 6000 was
dependent on the temperature of PEG melt. prepared by the evaporation technique. Drug and
Properties of the PEG-based solid dispersion PEG 6000 were dissolved in methanol, stirred,
prepared with the melt method are also known and evaporated to dryness under vacuum in a
to be cooling rate dependent. Rapid quenching rotary evaporator. The dispersion prepared by
during the preparation process, low storage tem- this method created a highly porous drug delivery
perature, and low relative humidity were found to system, which exhibited rapid drug dissolution
prevent crystallization of nimodipine from its (Bouchal et al. 2015).
solid solution in PEG 2000 (Urbanetz and With the extensive research interest in the
Lippold 2005). Recently, the melt method was application of supercritical fluid in pharmaceuti-
used to prepare binary solid systems of poorly cal processing, supercritical CO2 as an alternative
water-soluble drugs, including piroxicam, and solvent to prepare drug–PEG dispersion has been
PEG. Different drug:PEG 6000 ratios were demonstrated. Where drug has sufficient solubil-
mixed and heated until melted, then immersed in ity in supercritical CO2, rapid expansion of super-
an oil bath controlled at the melting point of critical fluid solution (RESS) process is safer and
piroxicam. The study compared the outcomes more environmentally friendly than organic sol-
between the homogeneous liquid that was cooled vent processes. In RESS process, drug and PEG
to room temperature at a slow rate (1 C/min) and solution in supercritical CO2 is passed through a
that which was rapidly solidified in an ice bath. small nozzle and allowed to expand rapidly under
Regarding the binary solid dispersions with high ambient condition. When CO2 was converted to
content of PEG 6000 (90% w/w), it was presumed gas, solid dispersion of very fine particle size and
that these melted solid dispersions exhibited drug uniform size distribution formed. With the rapid
heterogeneity, which the rapid cooling could not flashing of supercritical CO2 and solidification of
prevent. However, FT-IR spectra show some the drug dispersion, PEG processed with RESS
interactions between drug and PEG 6000. The process has higher amorphous contents than that
solid dispersion systems from both techniques processed with traditional solvent-based process
significantly improved the drug dissolution and and fusion process. Higher amorphous content
speed when compared to the crystalline would potentially allow higher drug loading.
piroxicam. These results confirmed that PEG Concentration of drug and PEG, processing tem-
6000 had an influence on reducing degree of perature, and flow rate of the solution through the
drug crystallinity in these solid dispersion expansion nozzle can be controlled to produce
systems (Bouchal et al. 2015). drug–PEG dispersion with different morphology.
Brodin et al. (2003) successfully applied RESS
process to prepare lidocaine and PEG 8000
5.5.3 Solvent Method dispersion.
for the Preparation of PEG For the drug substance with limited solubility
Dispersion in supercritical CO2, a gas antisolvent recrystalli-
zation (GASR) process has been developed. The
In the solvent method, a homogeneous organic GASR process is a more universal process than
solution containing drug and PEG is prepared. RESS process since supercritical CO2 has limited
Processing techniques such as rotary evaporation, solvent capacity for most drug substances. Com-
spray drying, and lyophilization are applied to plete miscibility of organic solvent with supercrit-
remove the organic solvents. Drug degradation ical CO2 is required for GASR process. When
is minimized or avoided in solvent method. organic solution of drug and PEG is mixed with
supercritical CO2, carbon dioxide is dissolved in that the following equations can describe
and expands the organic solvent under moderate (Jambhekar and Breen 2016):
temperature and pressure. The solubilization
power of an organic solvent decreases when
K1:1 D
CO2 is incorporated. When the organic solvent D þ CD , ð5:7Þ
CD
can no longer keep drug and PEG in solution,
nucleation of drug–PEG dispersion starts to where one drug (D) molecule forms a CD com-
form. As more CO2 is mixed in the solution, the plex with a cyclodextrin (CD) molecule.
nuclei continue to grow until the drug–PEG The equilibrium constant K1:1 is known as
completely precipitates out of the solvent. binding or complexation stability.
Moneghini et al. (2001) used supercritical CO2
as the antisolvent to recover carbamazepine–PEG
½CD
4000 solid dispersion from its acetone solution. K1:1 ¼ ð5:8Þ
½C ∙ ½D
The solubility of a drug within a cyclodextrin

5.6 Other Solubilized Systems where they form a complex can be represented by
the equation:
Other formulation strategies to develop
solubilized formulations include drug– K1:1 ∙ S0 ½CD
S ¼ S0 þ ð5:9Þ
cyclodextrins complex, amorphous solid disper- K1:1 ∙ S0 þ 1
sion, and self-emulsifying formulation. Below is
where S refers to the solubility of a drug in the
a brief overview of these three types of
presence of the CD and S0 refers to a drug’s
formulations. The reader is referred to Chaps. 7
intrinsic solubility. Based on this equation, the
and 10 for an in-depth discussion of drug–cyclo-
value of S against [CD] will yield a straight line,
dextrin molecular complexes and self-
and the following equations can calculate the
emulsifying systems. Preparation of amorphous
slope of this line and K1:1:
solid dispersions via spray drying and melt extru-
sion is covered in Chaps. 8 and 9.
S0 ∙ K1:1
slope ¼ ð5:10Þ
S0 ∙ K1:1 þ 1
5.6.1 Drug–Cyclodextrins Complex
slope
Cyclodextrins (CD) are cyclic oligosaccharides K1:1 ¼ ð5:11Þ
derived from starch. CD molecules take the S0 ∙ ð1 slopeÞ
shape of a truncated cone. The hydroxyl groups
Equation 5.10 suggests that the increase of the
of CDs are at the exterior of the cone structure,
solubility of a drug is a function of (1) intrinsic
leaving the interior of the cone a hydrophobic
solubility of the drug, (2) CD concentration, and
environment. Hydroxypropyl and sulfobutylether
(3) binding constant (K1:1). It has been reported
derivatives of CDs have been synthesized to
that K1:1 is between 50 and 2000 M1, and the
improve the solubility of CDs for broader appli-
mean values for α-CD, β-CD, and γ-CD are
cation. CDs have a safe toxicological profile with
129, 490, and 355 M1, respectively (Connons
little oral absorption. CDs that are absorbed are
1996). In addition, the value of the binding con-
excreted unchanged in the urine. CDs have been
stant determines the affinity between drug and
used in formulating poorly water-soluble drugs
CD, which describes the strength of the
EMA 2020.
interaction.
Most CD complexes are 1:1, where one drug
It has been found that some drugs will form
molecule will form a complex with one molecule
dimers, trimers, and other aggregates, while only
of CD. The forming of CDs is a reversible process
individual drugs can form a complex with CD, addition of glucosyltransferase (Davis and
which means dimers and trimers are unable to Brewster 2004). The system of nomenclature for
form CD complexes (Loftsson and Brewster CDs is based on the number of glucose residues in
2004). In such a situation, there is an issue calcu- their structure. For example, glucose hexamer is
lating K1:1 based on the above-mentioned referred to as α-CD, and heptamer and octamer
method. Therefore, a more accurate method to are referred to as β-CD and γ-CD, respectively
determine the complexation efficiency (CE) (Fig. 5.3).
came into use: Due to the 4C1 conformation of glucopyranose
units, secondary hydroxyl groups are placed on
½D∕ ½CD slope
CE ¼ S0 ∙ K1:1 ¼ ¼ ð5:12Þ two edges of the ring, while primary hydroxyl
½CD 1 slope
groups are on the other edge. The ring is
The chemical structure of CD was first characterized in a conical cylinder shape with a
revealed in 1891 by Villiers (Villiers 1891). In cavity in the middle (Szejtli 2004). Figure 5.4
1953, the first patent regarding CD drug depicts the spatial structure of a β-CD and
formulations was issued, which indicated the compares it to that of a hollow cylinder to help
commercial use of CDs. CDs are typically illustrate the orientation of the hydroxyl groups in
generated by the degradation of starch with the relation to the cavity. Other physical and
OH
OH HO O
OH 6 O
5 O OHO
O O HO
HO O O 4 1 OH OOH HO
O OH HO
OH HO HO 3 2 OH O OH O
OH
OH O HO
HO
OH O HO O
O HO OH OH O
O OH
OH HO HO
O OH O
O O O O
HO HO OH OH
O HO
HO OH
HO O OH
O HO OH
OH O HO OH
OH HO O
OH HO O HO O
O OH O OH OHO
O O OH OH HO
O OH HO
O O
HO OH O
O O
HO
HO OH
α-CD β-CD γ-CD
Fig. 5.3 Chemical structures of α-CD (a), β-CD (b), and γ-CD (c) (G.K.E. Scriba 2015)
Fig. 5.4 Spatial structure β-CD. The solid and dashed arrows represent primary and secondary hydroxyl groups,
respectively
Table 5.3 Characteristics of a-CD, b-CD, and g-CD

α-CD β-CD γ-CD
Molecular weight (g/mol) 972 1135 1297
External diameter (Å) 14.6 15.4 17.5
Internal diameter (Å) 4.7–5.3 6.0–6.5 7.5–8.3
Height (Å) 7.9 7.9 7.9
Cavity volume (Å) 174 262 427
Crystal forms Hexagonal plates Monoclinic parallelograms Quadratic prisms
Crystal water (w/w) 10.2 13.2–14.5 8.13–17.7
pK 12.332 12.202 12.081
Water solubilities at ambient condition (w/w) 13% 2% 26%
Fig. 5.5 Schematic representation of the formation of permission from (Szejtli 1998). Copyright (2021) Ameri-
CD–drug inclusion complex. p-xylene is guest molecule. can Chemical Society)
Small circle represents water molecules. (Reprinted with
chemical characteristics of CD can be found solution, the apolar CD cavity is filled with water
through Table 5.3 (Szejtli 1998). molecules which is energetically unfavored
As shown in Table 5.3 above, β-CD was found because of polar–apolar interaction, and will be
to have the lowest water solubility (2%) com- replaced by drug molecules, also known as guest
pared to α-CD (13%) and γ-CD (26%). This is molecules with less polarity (Szejtli 1998).
because of the formation of an internal hydrogen Thermodynamic driving forces for the forma-
bonding network between hydroxyl groups in tion of CD inclusion complexes include:
β-CD (Davis and Brewster 2004). Some studies (1) hydrophobic interaction between the inner
showed that the disruption of hydrogen bonding cavity of CD and drug; (2) release of “high-
through molecular manipulation attributes to the energy water” from the CD cavity; and (3) release
improvement of water solubility. For example, as of conformational strain in a CD-water adduct. As
a derivative of β-CD, hydroxypropyl-β-CD Fig. 5.6 shows, drug and CD can form an inclu-
(HPβCD) exhibited 50-fold higher solubility sion complex in a molar ratio of 1:1 or 1:
(Szente and Szejtli 1999). Even though CD is 2 (Fig. 5.6a, b). The drug can also be partially
water-soluble overall, CD’s unique structure encapsulated by CD and form a complex
divides it into hydrophobic and hydrophilic (Fig. 5.6c, d). This complexation process is in a
parts. The interior side of CD, where drug dynamic equilibrium. However, due to the natural
molecules are encapsulated, is hydrophobic, properties of CDs, they tend to self-assemble and
while the exterior is hydrophilic. Figure 5.5 is aggregate in solution, which will lead to lower
an example of a drug–CD complex. In an aqueous solubilities as shown in Fig. 5.6e and f (Saokham
Fig. 5.6 Formation of a cyclodextrin–drug complex in solution and self-assembly of complex
et al. 2018). In order to solve the solubility issue,

scientists have modified the natural CDs and
replaced the hydroxyl groups with other func-
tional groups. These modified CDs such as
hydroxypropyl β-CD (HPβCD) and 6sulfobutyl
ether β-CD (SBEβCD), also known as CD
derivatives, have been approved by FDA and
widely used (Davis and Brewster 2004).
Phase solubility diagram of the concentration
of dissolved drug as the function of CD concen-
tration is constructed to determine drug to CD
ratio, binding constant, and the complexation effi-
ciency (Connons 1996). Depending on the solu-
bility of formed complex, drug–CD complexation
can be classified into two different categories
(Fig. 5.7). In type A systems, solubility increases
with increase in CD concentration. Type B profile
is indicative of the formulation of drug–CD com-
plex with limited water solubility. Studies have
shown that drug–CD complex itself can also
Fig. 5.7 Phase solubility profile drug in cyclodextrins
aggregate that can further enhance drug solubil-
solution: A and B types (with applicable subtypes)
ity. In most cases, the drug and CD complex is at (Connons 1996)
1:1 ratio and of AL type. Dilution of the solution
will not result in supersaturation, which could have a strong affinity with the CD to achieve the
induce precipitation. However, for a compound desired enhancement in solubility.
with very low intrinsic solubility, the drug must
Complexation Techniques BCS class II drug, and ethylenediamine-β-CD

There have been a variety of techniques used to showed significant improvement in water solubil-
prepare CD-guest inclusion complexes. As for the ity (Heydari et al. 2015).
complexes of drugs and CDs, the methods used
vary from more simplistic techniques to more Solvent Evaporation For this method, the drug
sophisticated ones. Spray drying and freeze dry- and CD are completely dissolved in an organic
ing have become preferred techniques in prepar- solvent. The solvent is evaporated to dryness, and
ing CD complexes. Examples of recently reported the dried powder is collected. Ethanol (90%) was
preparation techniques used to form the inclusion used as the solvent to dissolve fenofibrate and
complexes of poorly water soluble drugs and CDs HPβCD in the solvent-evaporation technique.
are summarized as follows: The solvent was evaporated at 40 C for 48 h.
The complex formation of the resulting product
Grinding Mechanical grinding is one of the sim- was confirmed using various characterization
plest preparation techniques, which does not techniques. Moreover, in particular, PXRD
require the sophisticate equipment, or any toxic indicated that the drug was amorphous after inclu-
solvents that are not environmentally friendly sion into the cavity of HPβCD (Yousaf et al.
(Barzegar-Jalali et al. 2010). Drug–CD complex 2015).
can also be prepared by this method. Telmisartan
was entrapped into β-CD using a solid-state Extrusion Wet and dry extrusion can be applied
grinding method (ball milling). The ratio between to prepare drug–CD inclusion complexes. Drug
drug and β-CD and grinding time was varied. and CD, with or without water, are premixed or
Complexes at a ratio of 1:2 and 1:3 after 30-min mixed in the extruder. Extrusion parameters are
grinding showed faster in vivo pharmacological varied to obtain the optimized complex. A com-
effect than the free drug during in vivo studies parison between melt extrusion and wet extrusion
(Borba et al. 2015). of indomethacin-HPβCD inclusion complexes
was investigated. The in vitro drug release study
Kneading The drug substance is added into a shows that the application of HPβCD in the extru-
CD slurry, and then the mixture is kneaded until sion process improved the dissolution of the
a paste is produced. A small volume of organic poorly water-soluble drug (Yano and
solvent is often used to improve the solubility of Kleinebudde 2010).
drug in the slurry in order to facilitate the drug
inclusion process. For example, the kneaded Spray Drying Drug and CD are firstly dissolved
inclusion complex of famotidine and β-CD also in an appropriate solvent. The solution is then
enhanced the solubility and dissolution rate of the sprayed in the spray dryer. Fenofibrate and
drug (Gupta et al. 2013). A fenofibrate-HPβCD piroxicam were prepared by forming complexes
inclusion complex was prepared by kneading the with CDs by spray drying. The inclusion
drug and HPβCD with a small amount of ethanol complexes were amorphous. The solubility and
as a paste using a mortar and pestle. The resulting dissolution rates of these drugs were dramatically
product exhibited improved dissolution of the enhanced in comparison to the uncomplexed
drug as compared with the uncomplexed drug drugs (Yousaf et al. 2015; Bouchal et al. 2015).
(Yousaf et al. 2015).
Freeze Drying Similar to spray drying, a solu-
Coprecipitation The CD and drug are dissolved tion of drug and CD is prepared prior to freeze
in water or a mixture of water and organic sol- drying, then the solvent is removed by freeze
vent, and the solvent is evaporated. The precipi- drying. The ternary inclusion complexes compris-
tate is collected. An inclusion complex prepared ing HPβCD or hydroxybutenyl-β-cyclodextrin
using a coprecipitation method of nifedipine, a (HBen-β-CD) and Soluplus® prepared by freeze
drying exhibited fast and extensive release of the microscopy (SEM) provides evidence of the phys-
drug in the dissolution media (Taupitz et al. ical mixture and the inclusion complex (Zoppi
2013). et al. 2013; Qiu et al. 2014). Solid-state nuclear
magnetic resonance (solid-state NMR), as well as
For characterization of the actual CD complex computational chemistry, provides a better under-
in the solid state, various techniques have been standing of these complexes (Ogawa et al. 2010;
used, including thermal analytical analysis, X-ray Ferreira et al. 2015).
diffraction, and SEM. Thermal analysis is a Regarding the CD complexes in solution,
widely used approach and includes differential spectroscopic techniques are the most effective
scanning calorimetry (DSC) (Bettinetti et al. techniques, and these include UV-Vis (Ungaro
2002; Al-Marzouqi et al. 2007; Mennini et al. et al. 2011; Pápay et al. 2016), circular dichroism
2014, and Corciova et al. 2015), thermogra- (Song et al. 2011), fluorescence (Iglesias-Garcia
vimetric analysis (TGA) (Meier et al. 2001; Bai et al. 2010; Çelik et al. 2015, and Pápay et al.
et al. 2012, and Garnero et al. 2012), and hot-stage 2016), and NMR (Garnero and Longhi 2007;
microscope (HSM) (Ginés-Dorado et al. 2014; Holm et al. 2015) spectroscopy. Other methods
Trollope et al. 2014). Generally, X-ray diffraction included are electroanalytical techniques (such as
is a technique used to characterize the crystallinity potentiometry) and electro conductivity (Iglesias
of the material. X-ray diffraction methods, in par- 2006)) and the high-performance liquid chroma-
ticular, are used to study the crystal structure of tography (HPLC) (Dodziuk et al. 2011). Other
CD in the inclusion complex (Christoforides et al. techniques used for characterization of the drug
2015). Powder X-ray diffraction determines the and CD complex in aqueous solution include
residual crystallinity of drug in the formulations polarimetry (Lo Meo et al. 2011; Rizzo et al.
or recrystallization of the CD during the process 2014), potentiometric titration, and isothermal
(Al-Marzouqi et al. 2009; Banchero et al. 2013; titration calorimetry (ITC) (Ragab et al. 2015;
Mennini et al. 2014, and Tang et al. 2016). Spec- Holm et al. 2015).
troscopic techniques (e.g., Fourier-transform The success of drug–CD inclusion complexes
infra-red (FT-IR) spectroscopy (Salústio et al. is reflected in numerous approved products avail-
2009; Zhang et al. 2014), attenuated total reflec- able on the worldwide market since about 1976
tance (ATR)-FTIR spectroscopy (Venuti et al. (see Table 5.4) and about 1000 research articles
2014), and Raman IR (Mohan et al. 2012)) are and a large number of patents published every
also used to determine interactions of the drug– year (Del Valle 2004; Loftsson and Duchêne
CD complexes. In addition, scanning electron 2007).
Table 5.4 Examples of the commercial products containing cyclodextrin

Cyclodextrin/ Company/
drug Therapeutic usage Formulation Trade name country Reference
αCD
Alprostadil Treatment of IV solution Caverject dual Pfizer/USA Kurkov and Loftsson
erectile (2013) and Cyclodextrin
dysfunction News (2013)
Cefotiam hexetil Antibiotics Tablet Pansporin T Takeda/ Davis and Brewster (2004)
HCl Japan and Cyclodextrin News
(2013)
Limaprost Vasodilator Tablet Opalmon Ono/Japan Loftsson and Brewster
(2010)
βCD
Aceclofenac Nonsteroidal anti- Tablet Aceclofenac-β- Taj Cyclodextrin News (2013)
inflammatory Cyclodextrin Pharm/India
drug
(continued)
Table 5.4 (continued)

Benexate Treatment of acid- Capsules Ulgut Lonmiel Teikoku/ Davis and Brewster (2004)
related disorders Japan and Cyclodextrin News
(2013)
Betahistine Vertigo Tablet Betahist Geno Cyclodextrin News (2013)
Pharm/India
Cephalosporin Antibiotics Tablet Miact Meiji/Japan Cyclodextrin News (2013)
(E1207)
Cetirizine Antibacterial Chewing Zyrtec Pfizer/USA Kurkov and Loftsson
agent tablet (2013) and Cyclodextrin
News (2013)
Chlordiazepoxide Relieve anxiety Tablet Transillium Gador/ Atwood (2017)
and control Argentina
agitation
Dexamethasone Anti- Ointment, Glymesason Daiichi Davis and Brewster (2004),
inflammatory tablet Sanko/ Kurkov and Loftsson
steroid Japan (2013), and Cyclodextrin
News (2013)
Diphenydramine Antihistamine Tablet Ryndthisol Synthelab/ Cyclodextrin News (2013)
Italy
Ethinylestradiol Birth control Tablet Yasmin Berlex/ Loftsson and Brewster
and drospirenone Germany (2010)
Nicotine Nicotine Sublingual Nicorette GSK/USA Davis and Brewster (2004)
replacement tablet and Kurkov and Loftsson
product (2013)
Lodine Throat Gargling Mena-Gargle Kyushin/ Szejtli 2004
disinfectant Japan
Nimesulide Nonsteroidal anti- Tablet Nimedex Italfarmaco/ Davis and Brewster (2004),
inflammatory Italy Kurkov and Loftsson
drug (2013), and Cyclodextrin
News (2013)
Nitroglycerin Angina symptoms Sublingual Nitropen Nihon Davis and Brewster (2004)
tablet Kayaku/ and Cyclodextrin News
Japan (2013)
Omeprazole Proton pump Tablet Omebeta Betafarm/ Davis and Brewster 2004
inhibitor Germany and Cyclodextrin News
(2013)
Piroxicam Nonsteroidal anti- Tablet, Brexin Novartis, Davis and Brewster (2004),
inflammatory suppository Chiesi/ Kurkov and Loftsson
drug Europe (2013), and Cyclodextrin
News (2013)
Prostaglandin E2 Induction of labor Sublingual Prostarmon E Ono/Japan Frömming and Szejtli
tablet (1993), Davis and Brewster
(2004), and Cyclodextrin
News (2013)
Tiaprofenic acid Nonsteroidal anti- Tablet Surgamyl Roussel- Cyclodextrin News (2013)
inflammatory Maestrelli/
drug Italy
HPβCD
Cisapride Gastroesophageal Suppository Prepulsid Janssen/ Davis and Brewster (2004)
reflux disorders Europe and Cyclodextrin News
(2013)
(continued)

Diclofenac Nonsteroidal anti- IM/IV Dyolect Javelin Cyclodextrin News (2013)
inflammatory solution Pharm
drug
Hydrocortisone Mouth ulcer Mouth Dexocort Actavis/ Davis and Brewster 2004
wash Europe and Cyclodextrin News
(2013)
Indomethacin Nonsteroidal anti- Eye drop Indocid Chauvin/ Davis and Brewster (2004),
inflammatory solution France Kurkov and Loftsson
drug (2013), and Cyclodextrin
News 2013
Itraconazole Antifungal agent Oral and IV Sporanox Janssen/ Davis and Brewster (2004),
solutions Belgium Kurkov and Loftsson
and USA (2013), and Cyclodextrin
News (2013)
Mitomycin Anticancer agent IV infusion MitoExtra Novartis/ Davis and Brewster (2004),
Europe Kurkov and Loftsson
News (2013)
Telavancin Antibacterial IV solution Vibativ Astellas Cyclodextrin News (2013)
agent Pharma/
Europe
SBEβCD
Aripiprazole Antipsychotic IM solution Abilify BMS, Kurkov and Loftsson
drug Otuska/ (2013) and Cyclodextrin
USA, Japan News (2013)
Maropitant citrate Animal health IV solution Cerenia Pfizer/USA Loftsson and Brewster
(2010)
Voriconazole Antifungal agent IV solution Vfend Pfizer/USA Davis and Brewster (2004),
and Europe Kurkov and Loftsson 2013,
and Cyclodextrin News
(2013)
Ziprasidone Antipsychotic IM solution Geodon Pfizer/USA Davis and Brewster 2004,
drug and Europe Kurkov and Loftsson
News (2013)
SBEΥCD
Diclofenac Nonsteroidal anti- Eyedrop Voltaren Novartis/ Davis and Brewster (2004),
sodium inflammatory Ophtha Europe Loftsson et al. (2004), and
drug Kurkov and Loftsson
(2013)
Tc-99 Diagnostic aid, IV solution CardioTec Bracco/ Kurkov and Loftsson
Teboroxime cardiac imaging USA (2013)
RMβCD
17β-estradiol Relieve Nasal spray Aerodiol Servier/ Loftsson and Brewster
menopausal Europe (2010)
symptoms
Chloramphenicol Antibiotics Eye drop Clorocil Oftalder/ Davis and Brewster (2004)
Poland and Cyclodextrin News
(2013)
Adapted from Kurkov and Loftsson 2013 and updated
Note: αCD: α-cyclodextrin, βCD: β-cyclodextrin, HPβCD: 2-hydroxypropyl-β-cyclodextrin, SBEβCD: sulfobutylether
β-cyclodextrin; SBEΥCD: sulfobutylether Υ-cyclodextrin, RMβCD: randomly methylated β-cyclodextrin
Applications of cyclodextrin (CD) complex: HPβCD can significantly modify the mucoadhe-
siveness and drug permeability, resulting in lower
Nasal Drug Delivery Drugs delivered to the toxicity (Luppi et al. 2011).
nasal cavity should pass through the mucosa, a
biological barrier. It is crucial to improve the Ophthalmic Drug Delivery Eye drop solution
solubilities and metabolism rates of drugs (Ban is a widely used administrations for the oph-
et al. 2018. It has been reported that CDs can thalmic delivery system. The requirements for
enhance permeabilities and solubilities of drugs. the excipients used within eye drop
Among the CD derivatives, β-CD dramatically formulations are relatively rigorous, including
increased the nasal mucosa permeability (Yang (1) the therapeutic agent and excipient should
et al. 2004). Several studies have proved the be nonirritating to the ocular surface, (2) non-
safety and efficacy of nasal administration of toxic and well-tolerated, and (3) inert in nature.
CD drug formulation. As a derivate of fentanyl, Several studies have shown that CDs are capa-
carfentanil is recognized as a powerful analgesic ble of enhancing the permeability, solubility,
which exhibits a 100-fold potency than fentanyl. and bioavailability of the drug and effectively
However, due to its lower water solubility reduce drug irritation and discomfort (Tiwari
(0.0259 mg/ml), carfentanil can only be et al. 2010).
administrated through intravenous route. In
order to solve this issue, scientists developed a Pulmonary Drug Delivery It is well known that
complex containing drug and CD in order to pulmonary drug administration is a route for local
improve its solubility and dissolution rate. The treatment of some diseases such as chronic
results from 1H NMR showed that the obstructive pulmonary disease and asthma
carfentanil–CD complex has been prepared suc- (Washington et al. 2000). Lungs exhibit the
cessfully. The drug–CD formulation exhibited benefit of large surface area, high blood flow
elevated bioavailabilities with 2.0-folds and rate, and relatively low enzymatic activity. How-
36.3-folds higher than intramuscular injection ever, due to some drug molecules’ low solubility
and oral administration, respectively. More and dissolution rate, these drugs might be
importantly, according to nasal ciliotoxicity removed through mucociliary clearance and
tests, no significant difference of the cilla move- macrophages (Washington et al. 2000). When
ment was found between complex and free drug, incorporating with CDs, poorly water-soluble
which indicate the safety of this formulation with drugs tend to have high solubilities and perme-
the addition of CD (Yang et al. 2018). Recently, ability, resulting in improved absorption rate
several studies attempt to deliver drugs through (Rajewski and Stella 1996). Several studies
the nose-to-brain pathway. Paliperidone (PLPD), investigating the application of CDs in pulmonary
an FDA-approved drug treated for schizophrenia, drug delivery were reported recently. One study
was incorporated with HPβCD and in situ gel to was designed to formulate a roflumilast dry pow-
demonstrate the potential of nasal delivery. As a der inhaler using HPβCD. Next Generation
result, the formulation containing HPβCD Impactor and transepithelial permeability are
exhibited higher permeability through sheep used to estimate the in vitro deposition properties.
nasal mucosa (Sherje and Londhe 2018). CD The results demonstrate this formulation’s viabil-
complex also shows great potential for treating ity in terms of high fine particle fractions and
Alzheimer’s disease. It is well known that con- improved stability (Suzuki et al. 2018). Another
ventional drug formulation is associated with low study showed that by forming a drug–CD com-
bioavailability and hepatotoxicity. However, a plex, the improved aerodynamic performance
study showed that a novel anti-Alzheimer’s drug was attributed to the porous structure and elevated
formulation containing β-CD, SBE-β-CD, and wettability of such complex (Zhou et al. 2020).
The coronavirus disease 2019 (COVID-19) is a The best-known example of marketed self-
worldwide pandemic. As of January 2021, there emulsifying drug product is cyclosporine formu-
are 86.6 million cases and 1.87 million deaths lation. Earlier formulation contains corn oil, etha-
reported in total. As a broad-spectrum antiviral nol, and labrafil. The formulation disperses into
agent, remdesivir, developed by Gilead Sciences coarse emulsion in aqueous medium. The
Inc., was used to treat this disease for emergency. improved formulation, also known as Neoral™,
It has been reported by European Medicines consists of mixed glycerides as the oil phase,
Agency (EMA) that this drug was administrated Cremophor as emulsifying surfactant, and propyl-
by oral and intravenous (IV) routes (European ene glycol and ethanol as the cosolvents. The
Medicines Agency 2020). However, oral admin- formulation forms a thermodynamically stable
istration is not a suitable delivery method because transparent nanoemulsion when dispersed into
of the first-pass metabolism effect. In addition, aqueous medium. Improved absorption from
according to clinic research reports, remdesivir Neoral™ formulation was attributed to better
administrated by IV was found to have adverse dispersibility of the formulation. Surfactants
impacts, such as acute kidney injury, septic used in self-emulsifying formulations have also
shock, and renal impairment (Grein et al. 2020). been reported to enhance drug absorption by inhi-
In order to solve this pulmonary disease, some bition of P-glycoprotein-mediated efflux (Collnot
researchers developed a remdesivir- et al. 2007) and by inhibition of CYP3A metabo-
SBE-β-CD-based dry powder formulation for lism (Wu and Benet 2005).
inhalation by thin film freezing (TFF). After The predominant mechanism of self-
TFF processing, the drug within the CD carrier emulsification is “diffusion and stranding” driven
exhibits an amorphous form. The optimized for- by osmotic pressure imbalance (Pouton 1997).
mulation presented desirable aerosol performance Construction of ternary surfactant–oil–water
with FPF (Fine Particle Fraction) up to 93% and phase diagram is crucial to understanding the
elevated systemic uptake. This study provides role of each component. Selection of the formula-
valuable opportunities for the treatment of tion should be based not only on the how readily
COVID-19 and to prevent adverse events the formulation self-emulsifies but also on the
(Sahakijpijarn et al. 2020). stability of the emulsion.
5.6.3 Amorphous Solid Dispersion

5.6.2 Self-Emulsifying Formulation
in Polymer Matrices
Self-emulsifying formulations transform into sta-
Amorphous solid dispersion is a single-phase
ble emulsion spontaneously when they are dis-
system consisting of preferably amorphous poly-
persed in aqueous environment. The presence of
mer as the carrier, drug in its high-energy state,
bile salts and digestive products of glycerides in
and other excipients such as processing aid,
the formulation is not required for the emulsifica-
recrystallization inhibitor, and wetting agent.
tion process. Self-emulsifying formulation
Drug molecules are dissolved in polymeric for-
consists of drug, oil, high HLB surfactant, and
mulation matrix. Three-dimensional long-range
cosolvent. Improved drug absorption of poorly
order of drug molecule in crystalline state is
soluble drug from self-emulsifying formulation
absent in amorphous solid dispersion. Hydrogen
is due to the large surface area of the emulsion.
bonding and hydrophobic interactions between
Self-emulsifying formulations also offer the
the drug and polymer are the primary driving
advantage of more rapid absorption onset and
forces for the formation of the solid dispersion
more consistent absorption under different GI
during melt extrusion, inhibition of drug crystal-
conditions such as food effect and pH effect.
lization during storage, and achievement and
sustainment of supersaturation in GI tract.
Fig. 5.8 Microcentrifuge 1000

dissolution data comparing 50% Compound 6:HPMCAS-L SDD
crystalline drug, amorphous
Dissolved Drug Concentration

drug, and amorphous solid 800
dispersion (Friesen
et al. 2008)
600
(μg/mL)
400
Amorphous Compound 6
200
Crystalline Compound 6
0
0 20 40 60 80 100
Time (min)
Copovidone, povidone, hypromellose acetate

5.7 Dosage Form Manufacturing
succinate (HPMCAS), and HPMC have been suc-
of Solubilized Formulations
cessfully used as the carriers in commercial
products.
Various manufacturing techniques process the
In vivo, drug molecules released from the
drug solution in carrier excipients into oral dosage
amorphous dispersion are in four different states:
forms. Selection of the proper process depends on
(1) free drug molecules; (2) amorphous drug/
the physical characteristics of the solubilized
polymer nanostructures; (3) aggregates of amor-
formulations. For solubilized formulations that
phous drug/polymer nanostructures; and (4) amor-
are liquid under room temperature, encapsulation
phous or crystalline precipitates. Supersaturation
in soft gelatin capsule is the most common
of free drug molecules and the rapid release of
manufacturing approach. Absorption of liquid
drug from drug/polymer nanostructures are two
formulation into excipients (powered solution)
contributing factors for achieving higher bioavail-
with large surface area has attracted a great deal
ability. An example of supersaturation and
of research interest. Due to the low liquid-
sustainment of supersaturation of amorphous
carrying capacity of pharmaceutical excipients,
drug alone and a solid dispersion in aqueous
there has been no commercial success with pow-
media is shown in Fig. 5.8 (Friesen et al. 2008).
dered solution technology. For solubilized
Supersaturation up to 100-fold has been reported
formulations that are molten low-viscosity liquids
for an amorphous drug. The solubility ratio
at an elevated temperature and semisolid or solid
(amorphous over crystalline) drug can be theoret-
at room temperature, direct filling into hard gela-
ically calculated using the difference in the heat
tin capsules is a practical choice. In comparison
capacity of crystalline state and amorphous state
with soft gelatin encapsulation, hard gelatin
at the glass transition temperature (Hancock and
encapsulation is simpler and a more efficient
Parks 2000).
manufacturing process. There is also less interac-
Solvent-based spray drying, melt extrusion,
tion between the capsule shell and fill material in
and solvent-based drug layering processes could
hard gelatin encapsulation. Spray congealing and
be used to prepare amorphous drug dispersion in
fluid bed melt granulation are alternative
polymeric carriers. The reader is referred to
manufacturing techniques to process the molten
Chaps. 8 and 9 for an in-depth discussion on
solution at the elevated temperature to solid
spray drying and melt extrusion for solid
particles at room temperature. In spray
dispersion.
congealing, solubilized formulations are manufacture the soft gelatin capsule shells. Glyc-
atomized into fine droplets and transformed into erin, sorbitol, and propylene glycol, individually
spherical solid particles when the molten droplets or in combination, are used as the plasticizer for
are cooled with air in a spray dryer. In fluid bed the capsule shell. Selection of plasticizer is based
melt granulation, solubilized formulation is on the compatibility with fill materials and
sprayed into a powder bed as atomized molten desired properties of the final products (e.g., hard-
droplet. Granules consisting of powder carrier ness and appearance). Propylene glycol is a more
and solidified solubilized formulation are pro- effective plasticizer than glycerin. However, the
duced inside the fluidizing chamber. use of propylene glycol is limited since its
volatiles lead to changes in the mechanical integ-
rity of capsules and too much propylene glycol
5.7.1 Soft Gelatin Capsule interferes with the capsule shell sealing process.
Glycerin is the most common plasticizer for soft
Soft gelatin capsules are the most common dos- gelatin capsules. However, capsules shell
age form used to deliver solubilized formulations. containing too much glycerin can be tacky due
A list of commercial soft gelatin capsules to the hygroscopicity of gelatin. Sorbitol is likely
containing solubilized formulations of poorly sol- to crystallize out of the capsule shells when the
uble compounds is presented in Table 5.5. capsules are stored at low humidity conditions. A
Manufacturing of soft gelatin capsule consists mixture of glycerin and sorbitol is most com-
of the following key steps: gel mass preparation, monly used. For the preparation of gel mass, all
fill material preparation, encapsulation, drying components are mixed together at room tempera-
(primary drying and secondary drying), and ture to hydrate the gelatin. The mixture is then
finishing. Gel mass for the capsule shell typically processed at high temperature (90 C) under vac-
contains gelatin, water, plasticizers, opacifier, col- uum to form molten gel mass, which is kept at
orant, preservative, flavor, and sweetener. Gelatin 60 C until the encapsulation. Fill material must
of different bloom strengths can be used to be kept below 35 C so that the sealing of capsule
Table 5.5 Example compositions of commercial soft gelatin capsule products (adapted from Gullapalli 2010)
Active
Product ingredient Composition of the solubilized formulation
Adalat Nifedipine Peppermint oil, PEG 400
Agenerase Amprenavir PEG 400, TPGS
Procardia Nifedipine Peppermint oil, PEG 400, glycerin
Targretin Bexarotene Polysorbate 20, povidone K-90, BHA, PEG 400
Accutane Isotretinoin Soybean oil, hydrogenated vegetable oil and hydrogenated soybean oil, sodium
EDTA, BHA
Lanoxicap Digoxin Ethyl alcohol, PEG 400
Vesanoid Retinoin Soybean oil, hydrogenated vegetable oil and hydrogenated soybean oil, sodium
EDTA, BHA
Zarontin Ethosuximide PEG 400
Rocaltrol Calcitriol Fractioned triglyceride of coconut oil, BHA, BHT
VePesid Etoposide PEG 400, glycerin, citric acid
Fortase® Saquinavir Medium-chain nono- and diglycerides, povidone, vitamin E
Advil Ibuprofen PEG 600, potassium hydroxide
Acodart Dutasteride Medium-chain nono- and diglycerides, BHA
Nimotop Nimodipine PEG 400, glycerin, peppermint oil
Hytrin Terazosin HCL PEG 400, glycerin, povidone
is not interfered with by the fill material. Soft materials. As the drying process proceeds, the
gelatin capsules are produced on a rotary die excess water partitions back out of the fill
machine, which is fed by two tanks of materials: materials into the capsule shell and out into the
molten gelatin at 60–65 C and fill material at a atmosphere. Water migration during drying is
temperature less than 35 C. Molten gelatin flows illustrated in Fig. 5.9 (Gullapalli 2010). It is
onto the surface of two separate drums, where important to understand the effect of this dynamic
flat, solid ribbons of gelatin are formed. Liquid change in water content on the physical and
fill material is injected into the space between two chemical stability of capsule shell and fill
ribbons. The injection of the liquid formulation materials. Migration of water into the fill
forces the gelatin to expand into die pockets. As materials could cause the precipitation of
the ribbons continue to pass the heated wedge and solubilized drug. Serajuddin et al. (1986) studied
pressed between the die rolls, capsule halves are the effect of water migration on the stability of
sealed together by the application of heat and encapsulated drug solution in PEG 400. After the
pressure (Jimerson 1986). A fill-weight accuracy equilibration of capsule at ambient condition,
of 4% can be easily achieved with rotary die 6.3% water was present in drug solution in
process. Soft gelatin capsules coming off the PEG400. This reduced the solubility by 45%, in
encapsulation machine go through primary and resulting drug crystallization. When Gelucire
secondary drying processes to remove the excess 4414 was incorporated into PEG400 formulation,
water in the gelatin shell. Water content of the water migration was hindered and drug crystalli-
shell is high in order to facilitate the encapsula- zation was not observed during 3 months of
tion process. In primary drying, water leaves the observation. Besides the migration of water,
shell and evaporates into atmosphere. At the same migration of other small molecules, such as the
time, water could also migrate into the fill drug itself, plasticizer in capsule shell, and
Fig. 5.9 Dynamics of water migration during Softgel or long-chain triglycerides (LCT) fill formulation. Yellow
Drying Process Water migration patterns during drying inner core represents fill formulation, and brown exterior
of a softgel product containing a typical (a) PEG 400 fill represents gelatin shell. Numerical values represent
formulation; (b) mixed medium-chain mono-, di-, and approximate percent water content, w/w. (1) Water migra-
triglycerides fill formulation with or without an added tion from shell to fill, (2) water migration from fill to shell,
surfactant(s); and (c) medium-chain triglycerides (MCT) and (3) water migration from shell to environment
cosolvent in fill material, can also take place is observed is considered to be the solubility of
during the primary and secondary drying pro- drug in “soft gelatin compatible vehicle.”
cesses. Migration of these excipients is As discussed previously, hydrophilic
accelerated by the high shell water content and excipients such as low-molecular-weight PEG,
high product temperature during drying. ethanol, and propylene glycol can diffuse into
The biggest challenges with soft gelatin the capsule shell easily, especially during the
capsules are identifying means to prevent or min- drying stage. Plasticizer in the shell also tends to
imize the chemical and the physical interactions migrate into hydrophilic fill materials. The
between the capsule shell and fill material. Devel- amount of these hydrophilic excipients should
opment of the shell formulation is as critical as the be kept at a minimum level. Water content in
development of fill formulation. The chemical the fill material must be kept below 5%. Several
stability of drug solubilized in lipid formulation formulation strategies can be applied to achieve a
must be evaluated by mixing the formulation with stable product containing hydrophilic fill
a small amount of water and representative empty materials. High-bloom, low-viscosity gelatin is
soft capsule shell while monitoring the stability of better suited to encapsulate hydrophilic material,
drug in the mixture. The presence of aldehyde in since the initial water content in this type of
the fill material can result in cross-linking of gelatin shell is lower, and the process for drying
gelatin, which would reduce the drug release this type of gelatin shell is much shorter. Partially
rate during the shelf life of the product. Furfural replacing glycerol with sorbitol as the plasticizer
generated at the elevated temperature from rayon in the capsule shell, and adding small amount of
coil, a common packaging material, is also known glycerol and sorbitol into the fill material can
to cross-link gelatin. If the drug contains aldehyde minimize the migration of plasticizers from the
group, incorporation of succinic acid in the gela- shell to the fill material (Werner 1988). Diffusiv-
tin shell could be used to prevent cross-linking. ity of PEG decreases with increased molecular
Succinylated gelatin is not commonly used for weight. Therefore, better stability can be
pharmaceutical products but has been widely achieved by using liquid PEG with a higher
used in neutraceutical products. Aldehydes can molecular weight (with a minimum molar mass
also be generated from auto-oxidative degrada- of 400).
tion of excipients containing polyoxyethylene Migration of drug into the capsule shell must
groups, such as PEG, Poloxamer, Gelucire, and also be monitored for soft gelatin capsules. In
the likes (Bindra and Stella 1994). When these general, drug migration is less a concern for
excipients are present in fill material for soft gela- solubilized formulations containing poorly
tin capsules, incorporation of antioxidants in the water-soluble drugs, since these hydrophobic
formulation is necessary. BHT, BHA, and vita- drugs have low affinity for hydrophilic capsule
min E are the commonly used antioxidants in soft shells. Armstrong et al. (1984) investigated the
gelatin capsules. drug migration from the fill material (drug solu-
When the fill material is a lipid-based drug tion in isopropyl myristate) into capsule. No cor-
solution, studies must be conducted to determine relation was found between the extent of drug
the solubility of poorly soluble drug in lipid vehi- migration and drug solubility in isopropyl
cle that is in equilibrium with water. The study is myristate. However, there is a direct correlation
performed to simulate the scenario of the migra- between the extent of drug migration and drug
tion and retention water from capsule shell to fill solubility in aqueous medium. Most of the drug
material inside soft gelatin capsule. Solutions of migration took place during the primary drying
poorly soluble drugs in lipid vehicle at several (rotating basket). A smaller degree of migration
different concentrations are prepared. Each solu- took place during the secondary drying (tray dry-
tion is then mixed with small amount of water. ing). There was no additional drug migration in
When the mixtures reach the equilibrium, the the first 6 months of storage following the
highest concentration at which no precipitation manufacturing.
5.7.2 Hard Gelatin Capsules machine used for filling powder and granule can
be easily modified to fill the liquid formulation.
Over the last decade, encapsulating liquid and The composition of hard gelatin capsules for
semisolid compositions in hard gelatin capsules encapsulating liquid and semisolid formulations
is becoming more accepted. Several hard gelatin is the same as that of hard gelatin capsules for
capsules containing solubilized compositions of encapsulating powder and granules. Capsule
poorly soluble drugs have been commercialized. shells contain gelatin as the structural material,
Due to the concern over the leakage of liquid colorant, and TiO2 as the opacifier. Capsule shell
formulation, the fill material for hard gelatin manufacturers have designed gelatin capsules
capsules is generally semisolid or solid at ambient where the body fits tightly into the cap and
condition. Hard gelatin capsules containing ibu- functions as the primary seal barrier in order to
profen solubilized in Gelucire 44/14 is marked in prevent the leakage of liquid formulations prior to
Europe under brand name Solufen®. Hard gelatin sealing or banding operation. In the banding oper-
capsule containing fenofibrate solubilized in ation, concentrated gelatin solution is applied to
Gelucire 44/14 and PEG 20000 mixture is marked the junction of body and cap to form a seal when
in Europe as Fenogal®. The manufacturing pro- the gelatin solution dries up. The gelatin banding
cess for hard gelatin capsules is much simpler process is time consuming and labor intensive.
than that for soft gelatin capsules. A comparison Capsugel Inc. has invented and commercialized
between the soft gelatin capsule and hard gelatin LEMS™ (Liquid Encapsulation by Micro-Spray)
capsule manufacturing processes is presented in banding technology where a water and ethanol (1:
Table 5.6. 1 ratio) mixture is sprayed into the junction of the
Manufacturing of hard gelatin capsules body and cap (Cole et al. 2008). As a result of the
consists of encapsulation and banding processes. capillary effect, the solution is drawn into the gap
With the installation of a liquid dosator, capsule between the body and cap. The gelatin then melts
Table 5.6 Comparison between soft gelatin capsule and hard gelatin capsules
Soft gelatin capsules Hard gelatin capsules
Properties of fill Liquid, liquid suspension Liquid, liquid suspension, semisolid, solid with
materials low melting point
No concern over the leakage of low viscosity Leakage of low viscosity fill materials is a
fill materials concern.
Compatible with hydrophobic fill materials; Compatible with hydrophobic fill materials;
compatible with formulations containing low Hygroscopic excipients such as liquid PEG can
level of hygroscopic excipients such as liquid only be present at a much lower level
PEG
Maximum temperature of fill materials: 40 C Maximum temperature of fill materials: 70 C
Interaction between Major concern Less a concern, especially when the fill material
capsule shell and is semisolid or solid at storage conditions
fill content
Manufacturing Strict requirement on the manufacturing 40–60% relative humidity is preferred
environment environment; 20–30% relative humidity
Manufacturing Capsule shells are made during Capsule shells are prefabricated for
process encapsulation. encapsulation.
Manufacturing process is complex. Only a Manufacturing process is simple.
few companies possess acknowledge,
equipment & facility.
Post encapsulation Lengthy secondary drying (could take several Banding or sealing right after the encapsulation
days) followed with quick drying
to fuse the cap and body together. A gentle heat is cooling rate following encapsulation on crystalli-
required to remove the residual water/ethanol zation of PEG inside gelatin capsule and poten-
solution. Compared with traditional capsule tially the drug release rate from the capsule. Fast
banding, LEMS™ sealing operation is more effi- cooling of PEG-based dispersions has also been
cient and easier to control. reported to result in faster drug release.
Cade et al. (1986) used vitamin A palmitate, To be encapsulated in hard gelatin capsules,
peanut oil, and vitamin E as the model the formulation must be chemically and physi-
compounds to determine the fill-weight unifor- cally compatible with gelatin shell. Chemically,
mity of liquid-filled hard gelatin capsules on a cross-linking of gelatin by aldehydes in the for-
commercial-scale machine. The relative standard mulation is known to cause the decrease in drug
deviation of the fill weight was less than 0.5% at dissolution rate and decrease in bioavailability.
target fill weight of 400 mg. The consistency of For the formulation to be physically compatible
the liquid encapsulation process is highly depen- with hard gelatin capsules, migration of the water
dent on the rheology of the fill materials. The between the fill contents and capsule shells must
viscosity of the fill material should be kept be inhibited or minimized. The water contents of
between 100 and 1000 cP. If the viscosity is too empty gelatin capsules are 14–16%. If too much
low, the fill material might splash in a high-speed water migrates from capsule shell into the fill
filling process and the fill material might seep out material, the shell becomes brittle and is likely
into the overlapping area between the body and to fracture during shipping and handling. As in
cap. If the viscosity is too high, the fill material soft gelatin capsule, the migration of water into
might string at the dosing tip during the fill material could also cause solubilized drug
manufacturing. Splashing, seeping out, and to precipitate in hard gelatin capsule. If too much
stringing due to improper viscosity of the fill water migrates from the fill material to the shell,
materials could “contaminate” the overlapping the shell becomes soft and loses its physical integ-
area between the body and cap and, therefore, rity. Cade and Madit (1996) developed a testing
prevent the proper seal of the capsules (Cole procedure to determine the miscibility between
et al. 2008). Splashing and seeping can be over- the fill material and hard gelatin capsules. Filled
come by incorporating viscosity enhancers. capsules are stored at various relative humidity
Polymers such as povidone and hydroxypropyl conditions (ranging from 2.5% to 60% relative
cellulose, high-melting-point excipients such as humidity), and the weight change in the shell is
hydrogenated oils, and suspending agents such as monitored for 2 weeks. The acceptable weight
silicon dioxide and colloidal clays can be used to change is 2%. Kuentz and Rothlisberger
increase the viscosity of the fill materials. (2002) used texture analysis as a nondestructive
For high-molecular-weight PEG-based test for hard gelatin capsules containing liquid
formulations, the cooling of fill material inside formulations to investigate the effect of water
the capsule could have an effect on drug release migration on the mechanical properties of capsule
properties of the finished product. PEG is a semi- shell. It was concluded that small amounts of
crystalline excipients, and its crystallization is water are often required in the fill material to
dependent on the cooling condition. Higher amor- inhibit water migration between the capsule
phous content due to the fast cooling could result shell and fill material.
in faster drug release. Higher amorphous content Formulations containing high level of hydro-
of PEG as the result of flash cooling was philic cosolvents such as ethanol, glycerin, pro-
identified to be the contributing factor for the pylene glycol, and low-molecular-weight PEG
faster release from PEG matrices (McGinity are not compatible with hard shell capsules. The
et al. 1984). Chatham (1987) used PEG as the capsules become brittle and break easily when
model to demonstrate the effect of processing these hydrophilic solubilizers pull the water
temperature of fill material, the duration of hold- from the capsules during the storage. These
ing the fill material in the molten state, and the formulations are generally encapsulated in soft
Table 5.7 List of excipients compatible with hard gelatin capsule shell
Excipient category Excipient list
Long-chain triglycerides Soybean oil Sesame oil Peanut oil
Olive oil Cottonseed oil Hydrogenated oil
Corn oil Castor oil Safflower oil
Medium-chain triglycerides Medium-chain triglyceride (Miglyol 810, Miglyol 812, Captex 300, Captex
355, Labrafac CC)
Fatty acid and related esters Oleic acid Lauric acid
Propylene glycol ester (Capryol, Lauroglycol)
Polyglycerol ester (Plurol Oleique)
Surfactants Polyoxyl-40 hydrogenated castor oil (Cremophor)
Polyoxylglyceride (Gelucire, Labrasol, Labrafil)
Polyoxyethylene sorbitan monooleate (Tween)
Mono- and di-glyceride (Capmul MCM, Imwitor 191, Imwitor 900, Imwitor
380, Peceol, and Maisine)
Vitamin E TPGS
Poloxamer
Solubilizer PEG (MW 3350)
gelatin capsules. High-molecular-weight PEGs weight PEG (Hohne et al. 1990). Nifedipine is
and Poloxamer are an exception. They are com- highly soluble in both the liquid low-molecular-
patible with hard shell capsules since they form weight PEG and the melt of high-molecular-
solid plugs inside the capsules at ambient weight PEG. Compared against the traditional
conditions. Incorporation of antioxidant in the tablet containing micronized nifedipine, nifedi-
formulation has been demonstrated to further pine solution in PEG 200 has improved bioavail-
improve the stability of hard gelatin capsule ability, higher Cmax, and shorter Tmax. However,
shell containing PEG matrices (Stain and Bindar liquid PEG is not compatible with hard gelatin
2007). Lipophilic excipients such as super-refined capsule shell. When high-molecular-weight PEG
oils and triglycerides of fatty acids are compatible is used, nifedipine dissolves easily in molten
with hard gelatin capsules. A comprehensive PEG. However, nifedipine crystallizes out of the
summary of excipients compatibility of hard gel- formulation when molten PEG solidifies at ambi-
atin capsule shell is presented in Table 5.7. ent conditions. Aprical® is based on the mixture
Most hydrophilic solubilized formulations for of liquid PEG 300 and solid PEG 6000. With the
hard gelatin capsules are based on solid PEG proper PEG 300 and PEG 6000 ratio, the formu-
(MW 4000) and Poloxamer. These lation is compatible with hard gelatin capsule, and
formulations are semisolids or solids at the ambi- no crystallization of nifedipine is observed during
ent temperature and have low melting points the storage.
(<60 C). They can be filled into hard gelatin
capsules as molten liquids. When cooled to ambi-
ent temperature, the molten formulations trans- 5.7.3 Spray Congealing
form into a semisolid or solid plug inside the
capsules. Hard gelatin capsules maintain physical Spray-congealing process, also known as spray
integrity, as long as the temperature of fill cooling or prilling process, is commonly used in
materials is lower than 70 C. PEG-based food and chemical industry to convert a pool of
solubilized nifedipine formulation is molten liquid into solids spheres. The operation
commercialized as hard gelatin capsules in principles and processing parameters for the
Europe under the trade name Aprical®. The for- spray-congealing process is discussed by Killeen
mulation matrix comprises a mixture of low- (1993). The same basic equipment used for spray
molecular-weight PEG and high-molecular- congealing is also used for spray drying. Unlike
spray drying, spray congealing is a solvent or (3) particle size is more uniform. In rotary atomi-
water-free process. In spray-congealing process, zation processes, droplet diameter is proportional
cooling air is passed through the chamber to to liquid feed rate (power of 0.2) and inversely
remove the thermal energy from the atomized proportional to the disk speed (Mackaplow et al.
droplets. In spray-drying processes, hot air is 2006).
passed through the chamber to flash off the sol- Spray congealing has been successfully
vent or water. To undergo spray congealing, solid applied in microencapsulation, drug stability
material is melted and transformed into a enhancement, taste masking, sustained release
low-viscosity liquid. The molten liquid is then delivery, and excipients manufacturing. Yajima
atomized into molten droplets inside a chamber et al. (1999) demonstrated the effectiveness of
with cooling air passing through. The residence taste masking of spray-congealed clarithromycin
time inside the cooling chamber ranges from a wax matrix consisting of glycerol monostearate
few seconds to 1 min. The molten droplets cool and Eudragit E. Spray congealing of hydrophobic
and solidify into spherical particles inside the matrix comprising fatty acid, wax, glyceride, and
cooling chamber. For both spray-drying and emulsifier for masking the taste of pseudoephed-
spray-congealing processes, air cyclone is used rine HCl, dextromethorphan, and cholestyramine
to convey the particles out of the cooling chamber is disclosed by Sharma et al. (1989). Preparation
into a collector. The spray-congealed material can of controlled release of Verapamil hydrochloride
then be compressed along with other excipients wax microparticles via spray congealing has been
into tablets or filled into capsules. reported by Passerini et al. (2003). Microcrystal-
The most critical processing steps in spray line wax and stearyl alcohol were investigated as
congealing are atomization and cooling. Pneu- the sustained release agents. Verapamil hydro-
matic atomizer (two-fluid nozzle) and rotary chloride maintained its crystallinity in spray-
atomizer (centrifugational atomizer) are used to congealed microparticles. At 10% drug loading,
disperse the melt into find droplets. In pneumatic the drug was released over an 8-h period.
atomizer, the kinetic energy from the compressed Compritol® 888 ATO, a free flowing powder of
atomization air breaks the liquid feed into fine glycerol behenate supplied by Gattefosse, is
droplets. The medium particle size of the droplets manufactured via the spray-congealing process.
is typically in the range of 50–300 μm. The higher Preparation of solid solutions of mannitol with
the atomization air volume-to-liquid mass rate various sugars for use as suspending agents or
ratio, the smaller the droplet size. The lower the carrier for drug formulations via spray-
viscosity of the feed material (higher feed mate- congealing process is disclosed in a patent
rial temperature), the smaller the droplet size. (Scott 1967).
Generally speaking, larger particles can be Spray-congealed particles can be blended with
achieved when the process is scaled up in larger excipients and compressed into tablets or filled
size equipment. Longer residence time inside the directly into the capsules. Juppo (2004) applied a
cooling chamber of larger sprayers allows bigger spray-congealing technique to transform the
droplets to cool and solidify sufficiently before solubilized formulation of felodipine into a
they are collected. When a rotary atomizer is multiparticulate-modified release dosage form
used, the molten liquid, introduced to the center with improved drug absorption. The molten
of a rotating wheel, is accelerated to the wheel drug solution maintained at 110 C was atomized
edge by centrifugational forces from the rotating with a pneumatic nozzle using hot atomizing air
wheel. When the liquid is discharged at the edge at 400 C and a pressure of 7 bar. Spray-
of the wheel, it breaks into fine droplets. The congealed particles were 50–100 μm in the diam-
advantages of rotary atomizer over pneumatic eter and 0.95 in roundness.
atomizer are: (1) rotary atomizer can handle Molten formulation is flash-cooled in spray-
materials of higher viscosity; (2) larger particles congealing process. Therefore, polymorphic
can be produced with rotary atomizer; and transformation of the carrier excipients must be
closely monitored. If necessary, curing of the Fluid bed granulator can be modified to per-
spray-congealed particles under elevated temper- form melt granulation. In fluidized bed melt gran-
ature can be applied to allow the conversion to ulation, heating of both the liquid feed line and
more stable form. Emas and Nyqvist (2000) stud- atomization air is required to prevent solidifica-
ied the thermodynamic changes in the thermal tion of the feed material in the line and to ensure
properties of spray-congealed carnauba wax proper atomization of the molten formulation.
microspheres with microcalorimeter. Annealing The spray nozzle needs to be properly positioned
procedures were developed to accelerate the ther- so that the atomized formulation is in the molten
modynamic transition during the storage. Glyc- state in the agglomeration zone of the fluidized
erol tripalmitate crystallized predominately in bed. To achieve uniform and gradual growth of
metastable α form following spray congealing. the melt granules, it is critical to maintain the
The metastable form gradually converted back product bed temperature below the melting point
to stable β form upon storage. The conversion of the solubilized formulation. This would allow
was accelerated when the spray-congealed mate- the growth of the granules to take place in the
rial was heated slightly above the melting point of agglomeration zone only. If the solubilized for-
α form. mulation is predominately in its molten state in
Since acetazolamide is a drug that melts and the fluidized bed, then that could potentially lead
decomposes at high temperature, it is difficult to to the collapse of the bed.
formulate as a molecular dispersion. Spray For the manufacturing of Fenoglide®, drug
congealing under optimal heat treatment was pro- solution in PEG and Poloxamer 188 is kept at
posed as a suitable technique to prepare this drug about 80 C. The temperature of substrate was
in the amorphous state. A mixture of drug found to be the critical processing parameter. In
dissolved in ethanol and molten poloxamer-237 several enabling examples disclosed in the patent,
was sprayed using the spray dryer with an inlet the atomization air was kept at 100 C and the
temperature of 80 C. However, the outer walls of inlet air at ambient temperature (20–25 C) for the
the spray cylinder were chilled with ice packs. controlled agglomeration. When the inlet air tem-
There was no thermal degradation of acetazol- perature was heated to 85 C, the geometric
amide detected in the resulting product. More- weight mean diameter grew to 980 μm even at
over, the test results indicated a drastic increase 25% solubilized solution loading. In contrast,
in solubility and rapid dissolution of the drug geometric weight mean diameter grew only to
(Kulthe and Chaudhari 2014). 450 μm even at 60% solubilized solution loading
under the controlled agglomeration conditions.
Granules containing the solubilized formula-
5.7.4 Fluidized Bed Melt Granulation tion were blended with Avicel PH200 and mag-
nesium stearate, and compressed into tablets. In
Fluidized bed processes have been used for dry- the animal PK study, the percent drug absorbed
ing, granulating, and coating pharmaceutical from the tablets was at least 50% higher than that
products. For solubilized formulation with low from nanosuspension and microemulsion. It was
meting point, fluidized bed processes could be also found that the higher the ratio between the
used to incorporate the formulation into a powder solubilizing agent and drug, the greater the
substrate to produce free flowing granules. absorption.
Fenoglide® is a commercial product containing
solubilized composition of fenofibrate. The prod-
uct is manufactured with a melt granulation pro- 5.7.5 Powered Solution Technology
cess, where a solution of fenofibrate in PEG 6000
and Poloxamer 188 was sprayed onto the Powered solution technology, also known as
fluidized lactose monohydrate powder bed Liquisolid technology, is to absorb drug solution
(Holm et al. 2007). onto porous solid excipients to prepare a flowable
and compressible powder. The powder can be 5.8 Summary

further processed into a capsule or tablet dosage
forms. To date, there has been no commercial Cosolvent and PEG-based solubilization
success with this type of formulation. A mathe- formulations have been demonstrated to be viable
matical model has been established for delivery systems for enhancing bioavailability of
formulating powdered solution composition. poorly water-soluble drugs. Drug bioavailability
The model can be used to calculate the optimum is improved due to fast dissolution rate and tran-
quantity of carrier excipient needed to prepare sient supersaturation of drug. Limited by their
powder with desired flow and compaction drug-loading capacity, PEG-based solubilized
properties (Spireas et al. 1992). Even though a formulations are more suited for delivering
solid dosage form is developed, drug remains in low-dose compounds. Soft gelatin encapsulation,
solubilized liquid state. Colloidal SiO2, micro- hard gelatin encapsulation, spray congealing,
crystalline cellulose, and natural clays are com- fluid bed melt granulation, and absorption into
monly used solid carrier. Large surface of the solid carriers are commonly used techniques
solid carrier and drug being at solubilized state used to prepare the solid dosage forms containing
are two contributing factors for fast drug release solubilized formulations. Physical stability of the
and improved absorption. Drug could precipitate solubilized formulation should be closely moni-
from the absorbed solution as the result of mois- tored. Formulation composition, manufacturing
ture absorption during the storage. process, and packaging configuration must be
Sheth and Jarowski (1990) used powdered properly selected so that poorly soluble drugs
solution to improve the drug-release properties remain solubilized during the manufacturing and
of polythiazide tablets. Solution of polythiazide on the shelf.
in PEG 400 (60 mg/ml) was triturated with amor-
phous silica. The liquid carrying powder mass Method Capsule 1
was then passed through a 40-mesh screen and
blended with microcrystalline cellulose. The Preformulation Support of Solubilized
resulting free-flowing powder was compressed Formulations
into tablet. Almost 90% drug was released at Based on the method reported by Desai and
5 min. In contrast, less than 20% drug was Park (2004)
released from directly compressed tablets Objective
containing drug and microcrystalline cellulose. • To determine the solubility of valdecoxib in
Drug release from polythiazide liquisolid tablets a variety of solid carriers, cosolvents, and
is a function of the ratio between drug solution, surfactants
silica and microcrystalline cellulose, the duration Equipment and Reagents
of mixing, and the order of mixing. The tablets • Valdecoxib
were physically and chemically stable when • Polyethylene Glycol 4000, 6000, 8000
stored at 40 C and 75% relative humidity. • Urea
Hydrochlorothiazide liquisolid tablets were • Mannitol
prepared using PEG200 as the solubilizer and • Tween 20, Tween 80
Avicel® PH 101, and Aerosil and light magne- • Sodium lauryl sulfate
sium carbonate as the carriers (Khaled et al. • Glycerol
2001). The absolute bioavailability of liquisolid • Ethanol
tablets was 15% higher than that of commercial • Methanol
tablets. Inclusion of certain water-soluble • Purified water
additives such as PVP in the drug solution as Method
crystallization inhibitor can further enhance the • Solubility studies with 1, 2, 5, and 10%
bioavailability of the solubilized drug in wt/vol for carrier–water mixtures.
Liquisolid tablets (Spireas 2002).
• Solubility studies with 10, 20, 30, 40, and Method

50% wt/vol for cosolvent–water mixtures. • Compound and excipient stock solutions
• Solubility studies with 0.25, 0.50, 0.75, and were prepared, metered, and dried to using
1.0% wt/vol for surfactant–water mixtures. the TECAN system into 96-well plates.
• Powder X-ray diffraction and scanning • Samples were aged as necessary by
electron microscopy for characterization of protocol.
valdecoxib drug substance. • Microplate dissolution was performed and
Results concentration profiles assessed by HPLC.
• The solubility of valdecoxib increased up to Results
8-, 7-, 6.7-, and 4.2-fold for PEG 4000, • Testing was successfully performed using
PEG 6000, PEG 8000, and urea, respec- drug levels as small as 50 μg per well,
tively. Mannitol provided no solubility allowing for multiple formulation evalua-
benefit. tion within a material sparing design.
• Cosolvent systems through 50% wt/vol • Binary drug–excipient evaluation at multi-
showed a rank-order increase of solubility ple drug loadings revealed that
such that ethanol>methanol>glycerol, JNJ-25894943 and JNJ-3026582 were
which was due to the greater polarity of solubilized at up to 100 mg/g by Vitamin
the mixed solvent system. E TPGS and Incrocas 35, with Vitamin E
• Anionic surfactant (sodium lauryl sulfate) TPGS being the most effective stabilizer
provided greater solubility enhancement over extended durations for both
than nonionic surfactants (Tween) and was compounds.
associated with the micelle interaction • Kinetic solubility data generated by the
between the surfactant and valdecoxib. high-throughput methodology were highly
correlated with conventional solubility
Method Capsule 2 screening.
Miniaturized and Automated Screening of
Method Capsule 3
Liquid and Semisolid Formulations
Based on the method reported by Mansky et al. Preparation and Characterization of
(2007) Posaconazole–CD Inclusion Complexes for
Objective In Vitro Study
• To apply a material sparing and efficient Based on the method reported by Tang et al.
screening method to identify liquid and (2016)
semisolid formulations Objective
Equipment and Reagents • To prepare the solid water-soluble
• JNJ-25894934, JNJ-3026582 complexes of poorly water-soluble drug
• Gelucire® 44/14 and CDs for in vitro fungal testing
• Hydroxypropyl-β-cyclodextrin Equipment and Reagents
• Tween 20, Tween 80 • Posaconazole (POS)
• Volpo 10 • β-cyclodextrin (β-CD)
• Capmul® MCM, PG8 • 2,6-di-O-methyl-β-cyclodextrin
• Captex 200 (DM-β-CD)
• Maisine 35–1 • Acetone
• Myvacet® 9–45 • Methanol
• Oleic acid Method
• Capric acid • The solid inclusion complexes of
• Vitamin E TPGS POS-β-CD and POS-DM-β-CD were
• TECAN Genesis Workstation (96 Well prepared, sieved through a 60-mesh sieve,
Model) and stored in sealed glass containers.
• The samples were characterized using vari- complexes formed in comparison with
ous solid-state analytical techniques includ- those of the parent β-CD.
ing FT-IR, PXRD, DSC, 1H-NMR,
and SEM. Method Capsule 4
• Molecular modeling was performed using
Cross-linking of Soft Gelatin and Hard Gelatin
the software.
Capsules
• Phase solubility of the complexes was
Based on the method reported by Meyer et al.
studied.
(2000)
• Solubility in water of the complexes was
Objective
carried out using saturated solutions
• To utilize in vitro analysis to predict bio-
approach.
equivalent and bioinequivalent capsules.
• In vitro dissolution was demonstrated, and
the concentrations of the drug were
• Acetaminophen
evaluated using UV spectrophotometer at
• Lactose
263 nm.
• Polyethylene Glycol 600, 1000
• In vitro antifungal susceptibility tests were
• Hard Gelatin Capsules, Size 1
performed in different types of common
• Type B Gelatin, 150 bloom limed-bone
pathogenic fungi.
gelatin
Results
• Glycerin
• Phase solubility study provided both
• Sorbitol
solubilizing ability and the apparent stabil-
Method
ity constant of the complexes.
• Hard gelatin capsules were stressed by fill-
• The DM-β-CD complexes showed greater
ing with lactose containing 20 ppm and
stability constants when compared with
120 ppm of formaldehyde while storing
those of the β-CD.
for 6 days at room temperature and 1 day
• The characterization methods confirmed
at 40 C/75% RH. Capsules were then emp-
the formation of drug–CD complexes.
tied and manually filled with
• Molecular modeling illustrated the most
acetaminophen.
likely configuration prepared, which is the
• Soft gelatin capsules were prepared
1:2 drug–CD complex.
containing 0, 20, and 80 ppm, with a stor-
• Solubility of the β-CD and DM-β-CD
age period of over 30 weeks at 25 C/60%
complexes increased about 10 and
RH and 40 C/75% RH.
90 folds, respectively form the pure POS.
• In vitro dissolution was conducted using
• Owing to the hydrophilic exterior surface of
USP Apparatus II with 900 ml of simulated
the CDs, the in vitro dissolution rate signif-
gastric fluid containing pepsin at 50 rpm.
icantly increased from the intact drug.
• Two separate 24-subject, three-way cross-
• The drug–CD inclusion complexes pre-
over, bioequivalence studies using three
served the antifungal activity of the drug.
different lots of hard gelatin capsules and
• These two CD complexes decreased the
three different lots of soft gelatin capsules,
viability of the fungi in a dose-dependent
all having experienced different levels of
manner, and the DM-β-CD complex is
stress.
superior to the β-CD complex.
Results
• In summary, in this study the substitution of
• Hard gelatin capsules exposed to increased
methyl groups in DM-β-CD resulted in an
levels of formaldehyde failed to meet USP
expansion of the hydrophobic region which
dissolution testing requirements in SGF and
increased the greater binding with drug and
water.
led to the better performance of the
• Soft gelatin capsules containing 20 ppm • In vivo pharmacokinetic study in fasted

formaldehyde met dissolution nonnaïve male beagle dogs was conducted,
requirements; however, higher levels failed and the relative bioavailability was
to comply with USP specifications after calculated.
55-day storage at 40 C/75% RH. Results
• Oral bioavailability of stressed capsules • Binary solid dispersions were successfully
showed similar AUC when compared to prepared by KinetiSol for Lots 1–5.
nonstressed product; however, a statisti- • The dissolution study showed that Lot
cally significant increase in Tmax was 5 which consisted of drug and HPBCD
observed for stressed product due to the had the best performance with a dramati-
cross-linked-induced delayed release. cally higher relative AUDCTotal of
1173.4%, compared to neat API.
Method Capsule 5 • However,obvious precipitation of abiraterone
was observed in FaSSIF media. Hence, ter-
Improving Bioavailability of Abiraterone by
nary systems containing drug, HPBCD, and
Forming Drug–Cyclodextrin Complex
polymers were investigated to solve this issue.
Through KinetiSol®
• Drug was found to be amorphous in Lots
Based on the method reported by Gala et al.
6–8 by using XRPD.
(2020)
• Improved supersaturation was observed in
Objective
ternary formulations.
To improve the solubility and bioavailability of
• Results from in vivo pharmacokinetic study
abiraterone by combining hydroxypro-
showed Lots 5 and 6 can significantly
pyl-β-cyclodextrin during KinetiSol®
enhance drug bioavailability by 12.4-fold
Processing
and 13.8-fold, respectively.
• Abiraterone
Method Capsule 6
• HPMC
• Polyvinyl pyrrolidone Development of Remdesivir-Captisol®-Based
• Polyvinyl acetate phthalate Dry Powder formulation by Thin Film
• Kleptose® HPB Freezing (TFF).
• Sodium carboxymethyl cellulose
• Microcrystalline cellulose Based on the method reported by Sahakijpijarn
Method et al. (2020).
• The physical blends containing abiraterone Objective
and cyclodextrin were processed by • To apply TFF technology to develop
KinetiSol at different grinding speeds remdesivir-CD dry powder formulation for
between 10,000 and 20,000 for 60 s. The inhalation administered by a commercially
processing temperature was monitored by available dry powder inhaler.
an infrared probe, and the mass was ejected Equipment and Reagents

when temperature reached 160 C. • Remdesivir
• The quenched mass was further milled by a • Captisol®
IKA tube mill and was then passed through • Lactose monohydrate
a #60 mesh screen ( 250 μm). • Leucine
• Physicochemical characterization studies • Polysorbate 20
were performed by XRPD, mDSC, • High resistance Monodose RS00 dry pow-
and HPLC. der inhaler
• A nonsink pH shift dissolution study was Method
conducted to analyze the dissolution profile • Remdesivir and different excipients were
of abiraterone. first fully dissolved into different solvent
systems such as acetonitrile/water (50/50 v/ Armstrong NA, James KC, Pugh WKL. Drug migration
v) and 1,4-dioxane/water (50/50 v/v). into soft gelatin capsule shells and its effect on in-vitro
availability. J Pharm Pharmacol. 1984;36:361–5.
• A precooling process at 2–8 C at a refrig- Atwood JL. Comprehensive supramolecular chemistry
erator was conducted before TFF. II. Elsevier; 2017.
• The solutions of different formulations Bai Y, Wang J, Bashari M, Hu X, Feng T, Xu X, Jin Z,
were passed through a syringe needle and Tian Y. A thermogravimetric analysis (TGA) method
developed for estimating the stoichiometric ratio of
dropped to a cryogenically cooled stainless- solid-state α-cyclodextrin-based inclusion complexes.
steel drum. The solutions were frozen and Thermochim Acta. 2012;541:62–9.
collected for the following drying process. Ban MM, Chakote VR, Dhembre GN, Rajguru JR, Joshi
• Frozen samples were primarily dried at DA. In-situ gel for nasal drug delivery. Int J Dev Res.
2018;8(2):18763–9.
-40 C for 20 h, and then 25 C for another Banchero M, Ronchetti S, Manna L. Characterization of
20 h. ketoprofen/methyl-β-cyclodextrin complexes prepared
• Characterization studies were performed by using supercritical carbon dioxide. J Chem. 2013:1–8.
HPLC, XRPD, mDSC, SEM, and NMR. Barzegar-Jalali M, Valizadeh H, Shadbad MR, Adibkia K,
Mohammadi G, Farahani A, Arash Z, Nokhodchi
• In vitro dissolution study and in vivo phar- A. Cogrinding as an approach to enhance dissolution
macokinetic study were also conducted. rate of a poorly water-soluble drug (gliclazide). Powder
Results Technol. 2010;197(3):150–8.
• All formulations exhibited a different Betageri GV, Makarla KR. Enhancement of dissolution of
glyburide by solid dispersion and lyophilization
degree of porous matrix structure observed techniques. Int J Pharm. 1995;126(1–2):155–60.
by SME. Specifically, the formulation Bettinetti GP, Sorrenti M, Rossi S, Ferrari F, Mura P,
containing Captisol® and leucine showed Faucci MT. Assessment of solid-state interactions of
relatively smaller nanostructure aggregates naproxen with amorphous cyclodextrin derivatives by
DSC. J Pharm Biomed Anal. 2002;30:1173–9.
than mannitol and lactose formulations. Bindra WTD, Stella VJ. Degradation of O6-benzylguanine
• No drug characteristic peaks were found on in aqueous polyethylene glycol 400 (PEG 400)
XRD patterns for all samples, which solutions: concern with formaldehyde in PEG 400.
indicates remdesivir was in an amorphous Pharm Res. 1994;11(7):1060–4.
Bley H, Fussnegger B, Bodmeier R. Characterization and
state. stability of solid dispersions based on PEG/polymer
• Aerodynamic studies showed that high- blends. Int J Pharm. 2010;390:165–73.
drug loadings would yield relatively high Borba PA, Pinotti M, Andrade GR, da Costa NB,
FPF<5μm and lower mass median aerody- Olchanheski LR, Fernandes D, de Campos CE, Stulzer
HK. The effect of mechanical grinding on the forma-
namic diameter (MMAD) value. tion, crystalline changes and dissolution behaviour of
the inclusion complex of telmisartan and
β-cyclodextrins. Carbohydr Polym. 2015;133:373–83.
Bouchal F, Skiba M, Chaffai N, Hallouard F, Fatmi S,
References Lahiani-Skiba M. Fast dissolving cyclodextrin com-
plex of piroxicam in solid dispersion Part I: Influence
Al-Marzouqi A, Jobe B, Dowaidar A, Maestrelli F, Mura of β-CD and HPβ-CD on the dissolution rate of
P. Evaluation of supercritical fluid technology as pre- piroxicam. Int J Pharm. 2015;478(2):625–32.
parative technique of benzocaine–cyclodextrin Brodin A, Frank S, Ye C. Method of preparing solid
complexes–comparison with conventional methods. J dispersions. US Patent Publication 2003/
Pharm Biomed Anal. 2007;43:566–74. 0104065; 2003.
Al-Marzouqi AH, Elwy HM, Shehadi I, Adem Cade D, Madit N. Liquid filling in hard gelatin capsules-
A. Physicochemical properties of antifungal drug– preliminary steps. Bull Tech Gattefosse. 1996;88:15–9.
cyclodextrin complexes prepared by supercritical car- Cade D, Cole ET, Mayer JPH, Witter F. Liquid filled and
bon dioxide and by conventional techniques. J Pharm sealed hard gelatin capsules. Drug Dev Ind Pharm.
Biomed Anal. 2009;49:227–33. 1986;12(11–13):2289–300.
Anguiano-Igea S, Otero-Espinar FJ, Vila-Jato JL, Blanco- Çelik SE, Özyürek M, Güçlü K, Apak R. Antioxidant
Mendez J. The properties of solid dispersions of clofi- capacity of quercetin and its glycosides in the presence
brate in polyethylene glycol. Pharm Acta Helv. of β-cyclodextrins: influence of glycosylation on inclu-
1995;80:57–66. sion complexation. J Incl Phenom Macro. 2015;83
(3–4):309–19.
Chatham SM. The use of bases in SSM formulations. STP acetate succinate based spray dried dispersions: an
Pharma. 1987;3(7):575–82. overview. Mol Pharm. 2008;5(6):1003–19.
Chiou WL, Riegelman S. Pharmaceutical application of Frömming KH, Szejtli J. Cyclodextrins in pharmacy, vol.
solid dispersion systems. J Pharm Sci. 1971;60(9): 5. Springer; 1993.
1281–301. Gala U, Miller D, Williams RO. Improved dissolution and
Christoforides E, Mentzafos D, Bethanis K. Structural pharmacokinetics of abiraterone through kinetisol®
studies of the inclusion complexes of the (+)- and (2)- enabled amorphous solid dispersions. Pharmaceutics.
borneol enantiomers in α– and β–cyclodextrin. J Incl 2020;12(4):357.
Phenom Macrocycl Chem. 2015;81:193–203. Garnero C, Longhi M. Study of ascorbic acid interaction
Cole ET, Cade D, Benameur H. Challenges and with hydroxypropyl-β-cyclodextrin and
opportunities in the encapsulation of liquid and semi- triethanolamine, separately and in combination. J
solid formulations into capsules for oral administra- Pharm Biomed Anal. 2007;45:536–45.
tion. Adv Drug Del Rev. 2008;60:747–56. Garnero C, Aiassa V, Longhi M. Sulfamethoxazole:
Collnot EM, Baldes C, Wempe MF, Kappl R, hydroxypropyl-β-cyclodextrin complex: preparation
Huttermann J, Hyatt JA, Edgar KJ, Schaefer UF, and characterization. J Pharm Biomed Anal. 2012;63:
Lehr CM. Mechanism of inhibition of P-glycoprotein 74–9.
mediated efflux by vitamin E TPGS: influence on Ginés-Dorado JM, Arias-Blanco MJ, Rabasco-Álvarez
ATPase activity and membrane fluidity. Mol Pharm. AM, González-Rodríguez ML, Cózar-Bernal MJ,
2007;4(3):465–74. Sánchez-Soto PJ. Application of hot stage microscopy
Connons KA. Measurement of cyclodextrins complex sta- to the thermal study of two binary systems of pharma-
bility constants, Comprehensive supermolecular chem- ceutical interest: triamterene–polyethyleneglycol 6000
istry. Oxford: Elsevier; 1996. p. 205–41. and triamterene–β-cyclodextrin. In: Méndez-Vilas A,
Corciova A, Ciobanu C, Poiata A, Mircea C, Nicolescu A, editor. Microscopy: advances in scientific research and
Drobota M, Varganici CD, Pinteala T, Marangoci education, vol. 2. Formatex Research Center; 2014.
N. Antibacterial and antioxidant properties of hesperi- p. 1015–9.
din:cyclodextrin complexes obtained by different Grein J, Ohmagari N, Shin D, Diaz G, Asperges E,
techniques. J Incl Phenom Macrocycl Chem. Castagna A, et al. Compassionate use of remdesivir
2015;81:71–84. for patients with severe Covid-19. N Engl J Med.
Craig DQM. Polyethylene glycols and drug release. Drug 2020;382(24):2327–36.
Dev Ind Pharm. 1990;16(17):2501–26. Gullapalli RP. Soft gelatin capsules. J Pharm Sci. 2010;99
Cyclodextrin News (2013) 27(2). (10):4107–48.
D’souza AA, Shegokar R. Polyethylene glycol (PEG): a Gupta V, Ain S, Babita K, Ain Q, & Dahiya J. Solubility
versatile polymer for pharmaceutical applications. enhancement of the poorly water-soluble antiulcer
Expert Opin Drug Deliv. 2016;13(9):1257–75. drug famotidine by inclusion complexation. IJPSN.
Davis ME, Brewster ME. Cyclodextrin-based 2013;6(1):1983–89.
pharmaceutics: past, present and future. Nat Rev Ha ES, Lee SK, Choi DH, Jeong SH, Hwang SJ, Kim
Drug Discov. 2004;3(12):1023–35. MS. Application of diethylene glycol monoethyl ether
Del Valle EM. Cyclodextrins and their uses: a review. in solubilization of poorly water-soluble drugs. J
Process Biochem. 2004;39(9):1033–46. Pharm Investig. 2020;50(3):231–50.
Desai KGH, Park HJ. Solubility studies on valdecoxib in Hamburger R, Azaz E, Donbrow M. Autoxidation of
the presence of carriers, cosolvents and surfactants. polyoxyethylenic non-ionic surfactants and of polyeth-
Drug Dev Res. 2004;62:41–8. ylene glycols. Pharm Acta Helv. 1975;50(1/2):10–7.
Dodziuk H, Bielejewska A, Schiff W, Ueda Hancock BC, Parks M. What is the true solubility advan-
H. Complexation of [2.2]para- cyclophane with β- tage for amorphous pharmaceuticals? Pharm Res.
and γ-cyclodextrins studied by HPLC and NMR. 2000;17(4):397–404.
Centr Eur J Chem. 2011;9:1056–61. Heydari A, Iranmanesh M, Doostan F, Sheibani
Emas M, Nyqvist H. Method of studying aging and stabi- H. Preparation of Inclusion Complex Between Nifedi-
lization of spray-congealed solid dispersions with car- pine and Ethylenediamine-β-Cyclodextrin as
nauba wax 1. microcalorimetric investigation. Int J Nanocarrier Agent. Pharm Chem J. 2015;49(9):
Pharm. 2000;197:117–27. 605–12.
European Medicines Agency. Summary on Hohne H, Lahr W, Schmersahl HU. Nifedipine-containing
Compassionate Use: Remdesivir Gilead. Amsterdam: form of administration and method for its production.
European Medicines Agency; 2020. US Patent. 1990;4:894–235.
Ferreira MJG, García A, Leonardi D, Salomon CJ, Lamas Holm P, Buur A, Elma MO, Mollgarrd B, Holm JE,
MC, Nunes TG. 13 C and 15 N solid-state NMR Schultz K. Controlled agglomeration. US Patent.
studies on albendazole and cyclodextrin albendazole 2007;7:217–341.
complexes. Carbohyd Polym. 2015;123:130–5. Holm R, Olesen NE, Alexandersen SD, Dahlgaard BN,
Friesen DT, Shanker R, Crew M, Smithy DT, Curatolo Westh P, Mu H. Thermodynamic investigation of
WJ, Nightingale JAS. Hydroxypropyl methylcellulose the interaction between cyclodextrins and
preservatives—Application and verification in a Loftsson T, Duchêne D. Cyclodextrins and their pharma-

mathematical model to determine the needed preser- ceutical applications. Int J Pharm. 2007;329(1):1–11.
vative surplus in aqueous cyclodextrin formulations. Loftsson T, Brewster ME, Másson M. Role of
Eur J Pharm Sci. 2015; cyclodextrins in improving oral drug delivery. Am J
Iglesias E. Inclusion complexation of novocaine by beta- Drug Deliv. 2004;2(4):261–75.
cyclodextrin in aqueous solutions. J Org Chem. Luppi B, Bigucci F, Corace G, Delucca A, Cerchiara T,
2006;71:4383–92. Sorrenti M, et al. Albumin nanoparticles carrying
Iglesias-Garcia I, Brandariz I, Iglesias E. Fluorescence cyclodextrins for nasal delivery of the anti-Alzheimer
study of tetracaine-cyclodextrin inclusion complexes. drug tacrine. Eur J Pharm Sci. 2011;44(4):559–65.
Supramol Chem. 2010;22:228–36. Mackaplow MB, Zarraga IE, Morris JF. Rotary spray
Jambhekar SS, Breen P. Cyclodextrins in pharmaceutical congealing of a suspension: effect of disk speed and
formulations II: solubilization, binding constant, and dispersed particles properties. J Microencapsul.
complexation efficiency. Drug Discov Today. 2016;21 2006;23(7):793–809.
(2):363–8. Mansky P, Dai WG, Li S, Pollock-Dove C, Daehne K,
Jimerson RF. Soft gelatin capsule update. Drug Dev Ind Dong L, Elchenbaum G. Screening method to identify
Pharm. 1986;12(8&9):1133–44. preclinical liquid and semi-solid formulations for low
Jouyban A. Prediction of drug solubility in water- solubility compounds: miniaturization and automation
propylene glycol mixtures using Jouyban-Acree of solvent casting and dissolution testing. J Pharm Sci.
model. Die Pharmazie-Int J Pharm Sci. 2007;62(5): 2007;96(6):1548–63.
365–7. Martin A, Wu PL, Adjei A, Mehdizadeh M, James KC,
Jouyban A, Acree WE Jr. Mathematical derivation of the Metzler C. Extended hildebrand solubility approach:
Jouyban-Acree model to represent solute solubility testosterone and testosterone propionate in binary
data in mixed solvents at various temperatures. J Mol solvents. J Pharm Sci. 1982;71(12):1334–134.
Liq. 2018;256:541–7. McGinity JW, Maincent P, Steinfink H. Crystallinity and
Juppo A (2004) Novel modified release formulation. US dissolution rate of tolbutamide solid dispersions
Patent Application 20,040,067,256. prepared by the melt method. J Pharm Sci. 1984;10:
Khaled KA, Yousif A, Asiri B, El-sayed YM. In vitro 1441–4.
evaluation of hydrochlorothiazide liquisolid tablets in Meier MM, Luiz MTB, Szpoganicz B, Soldi V. Thermal
beagle dogs. Int J Pharm. 2001;222:1–6. analysis behavior of β- and γ-cyclodextrin inclusion
Killeen MJ. The process of spray drying and spray complexes with capric and caprylic acid. Thermochim
congealing. Pharm Eng. 1993;13(4):56–64. Acta. 2001;375:153–60.
Kuentz M, Rothlisberger D. Determination of the optimal Mennini N, Bragagni M, Maestrelli F, Mura
amount of water in liquid filled masses for hard gelatin P. Physicochemical characterization in solution and in
capsules by means of texture analysis and experimental the solid state of clonazepam complexes with native
design. Int J Pharm. 2002;236:145–52. and chemically-modified cyclodextrins. J Pharm
Kulthe VV, Chaudhari PD. Effectiveness of Spray Biomed Anal. 2014;89:142–9.
Congealing to Obtain Physically Stabilized Amor- Meyer MC, Ab S, Mhatre RM, Hussain A, Shah VP, Bottom
phous Dispersions of a Poorly Soluble CB, Cole ET, Lesko LL, Mallinowski H, Williams
Thermosensitive API. AAPS PharmSciTech. 2014;15 RL. The effect of gelatin cross-linking on the bioequiva-
(6):1370–7. lence of hard and soft gelatin acetaminphoen capsules.
Kurkov SV, Loftsson T. Cyclodextrins. Int J Pharm. Pharm Res. 2000;17(8):962–6.
2013;453(1):167–80. Mohan PRK, Sreelakshmi G, Muraleedharan CV, Joseph
Law D, Wang W, Schmitt EA, Qiu YH, Krill SL, Fort R. Water soluble complexes of curcumin with
JJ. Properties of rapidly dissolving eutectic mixtures of cyclodextrins: characterization by FT-Raman spectros-
poly(ethylene glycol) and fenofibrate: the eutectic copy. Vib Spectrosc. 2012;62:77–84.
microstructre. J Pharm Sci. 2003;92(3):505–15. Moneghini M, Kikic I, Voinovich D, Perissutti B,
Lo Meo P, D’Anna F, Riela S, Gruttadauria M, Noto Filipovic-Grcic J. Processing of carbamazepine-PEG
R. Binding properties of heptakis-(2,6-di-O- 4000 solid dispersions with supercritical carbon diox-
methyl)-β-cyclodextrin and mono-(3,6-anhydro)-β- ide: preparation, characterization and in vitro dissolu-
cyclodextrin: a polarimetric study. J Incl Phenom tion. Int J Pharm. 2001;222:129–38.
Macrocycl Chem. 2011;71:121–7. Moore WE. The use of an approximate dielectric constant
Lobenberg R, Amidon GL. Modern bioavailability, bio- to blend solvent systems. J Pharm Sci. 1958;47(12):
equivalence and biopharmaceutics classification sys- 855–7.
tem new scientific approaches to international Mura P, Faucci MT, Manderioli A, Bramanti G, Parrini
regulatory standards. Eur J Pharm Biopharm. P. Thermal behavior and dissolution properties of
2000;50:3–12. naproxen from binary and tertiary solid dispersions.
Loftsson T, Brewster ME. Pharmaceutical applications of Drug Dev Ind Pharm. 1999;25(3):257–64.
cyclodextrins: basic science and product development.
J Pharm Pharmacol. 2010;62(11):1607–21.
Nema S, Washkuhn RJ, Brendel RJ. Excipients and their Sharma SC, Shaw JJ, Yang RK. Novel drug delivery
use in injectable products. PDA J Pharm Sci Technol. system. US Patent. 1989;4,797,288
1997;51(4):166–71. Sherje AP, Londhe V. Development and evaluation of
Ogawa N, Higashi K, Nagase H, Endo T, Moribe K, pH-responsive cyclodextrin-based in situ gel of
Loftsson T, Yamamoto K, Ueda H. Effects of paliperidone for intranasal delivery. AAPS
cogrinding with β-cyclodextrin on the solid state fenta- PharmSciTech. 2018;19(1):384–94.
nyl. J Pharm Sci. 2010;99:5019–29. Sheth A, Jarowski CI. Use of powdered solutions to
Osborne DW. Diethylene glycol monoethyl ether: an improve the dissolution rate of polythiazide tablets
emerging solvent in topical dermatology products. J Drug Dev. Ind Pharm. 1990;16(5):769–77.
Cosmet Dermatol. 2011;10(4):324–9. Song W, Yu X, Wang S, Blasier R, Markel DC, Mao G,
Pápay ZE, Sebestyén Z, Ludányi K, Kállai N, Balogh E, Shi T, Ren W. Cyclodextrin–erythromycin complexes
Kósa A, Somavarapu S, Böddi B, Antal I. Comparative as a drug delivery device for orthopedic application. Int
evaluation of the effect of cyclodextrins and pH on J Nanomedicine. 2011;6:3173–86.
aqueous solubility of apigenin. J Pharm Biomed. Spireas S. Liquisolid systems and method of preparing
2016;117:210–6. same. US Patent. 2002;6,423,339
Passerini N, Perissutti B, Albertini B, Voinovich D, Spireas SS, Jarowski CI, Rohera BD. Powdered solution
Moneghini M, Rodriguez L. Controlled release of technology: principles and mechanisms. Pharm Res.
verapamil hydrochloride from waxy microparticles 1992;9(10):1351–8.
prepared by spray congealing. J Control Release. Stain D, Bindar DS. Stabilization of hard gelatin capsule
2003;88:263–75. shells filled with polyethylene glycol matrices. Pharm
Pouton CW. Formulation of self-emulsifying drug deliv- Dev Technol. 2007;12:71–7.
ery systems. Adv Drug Del Rev. 1997;25:47–58. Suzuki ÉY, Amaro MI, de Almeida GS, Cabral LM, Healy
Qiu N, Cheng X, Wang G, Wang W, Wen J, Zhang Y, AM, de Sousa VP. Development of a new formulation
Song H, Ma L, Wei Y, Peng A, Chen L. Inclusion of roflumilast for pulmonary drug delivery to treat
complex of barbigerone with hydroxypro- inflammatory lung conditions. Int J Pharm. 2018;550
pyl-β-cyclodextrin:preparation and in vitro evaluation. (1–2):89–99.
Carbohydr Polym. 2014;101:623–30. Szejtli J. Introduction and general overview of cyclodex-
Ragab MT, El-Rahman MKA, Ramadan NK, El-Ragehy trin chemistry. Chem Rev. 1998;98(5):1743–54.
NA, El-Zeany BA. Novel potentiometric application Szejtli J. Past, present and futute of cyclodextrin research.
for the determination of pantoprazole sodium and Pure Appl Chem. 2004;76(10):1825–45.
itopride hydrochloride in their pure and combined dos- Szente L, Szejtli J. Highly soluble cyclodextrin
age form. Talanta. 2015;138:28–35. derivatives: chemistry, properties, and trends in devel-
Rajewski RA, Stella VJ. Pharmaceutical applications of opment. Adv Drug Deliv Rev. 1999;36(1):17–28.
cyclodextrins. 2. In vivo drug delivery. J Pharm Sci. Tang P, Wang L, Ma X, Xu K, Xiong X, Liao X, Li
1996;85(11):1142–69. H. Characterization and In Vitro Evaluation of the
Rizzo C, D'Anna F, Marullo S, Vitale P, Noto Complexes of Posaconazole with β-and 2, 6-di-O-
R. Two-Component Hydrogels Formed by methyl-β-cyclodextrin. AAPS PharmSciTech. 2016:
Cyclodextrins and Dicationic Imidazolium Salts. Eur 1–11.
J Org Chem. 2014;5:1013–24. Taupitz T, Dressman JB, Buchanan CM, Klein
Sahakijpijarn S, Moon C, Koleng JJ, Christensen DJ, S. Cyclodextrin-water soluble polymer ternary
Williams RO III. Development of remdesivir as a dry complexes enhance the solubility and dissolution
powder for inhalation by thin film freezing. behaviour of poorly soluble drugs. Case example:
Pharmaceutics. 2020;12(11):1002. itraconazole. Eur J Pharm Biopharm. 2013;83(3):
Salústio PJ, Feio G, Figueirinhas JL, Pinto JF, Cabral 378–87.
Marques HM. The influence of the preparation Tiwari G, Tiwari R, Rai AK. Cyclodextrins in delivery
methods on the inclusion of model drugs in a systems: applications. J Pharm Bioallied Sci. 2010;2
β-cyclodextrin cavity. Eur J Pharm Biopharm. (2):72.
2009;71:377–86. Trollope L, Cruickshank DL, Noonan T, Bourne SA,
Saokham P, Muankaew C, Jansook P, Loftsson Sorrenti M, Catenacci L, Caira MR. Inclusion of trans-
T. Solubility of cyclodextrins and drug/cyclodextrin resveratrol in methylated cyclodextrins: synthesis and
complexes. Molecules. 2018;23(5):1161. solid-state structures. J Org Chem. 2014;10:3136–51.
Scott MW. Pharmaceutical tablet excipients of solid U. S. Food and Drug Administration. Inactive ingredient
particles of a binary solid solution of mannitol with a search for approved drug products (2021).
sugar. US Patent. 1967;3:341,415. Unga J, Tajarobi F, Norder O, Frenning G, Larsson
Scriba GKE. Chiral Separations in Capillary Electropho- A. Relating solubility data of parabens in liquid
resis. Reference Module in Chemistry, Molecular PEG400 to the behavior of PEG400-parabens solid
Sciences and Chemical Engineering; 2015. dispersions. Eur J Pharm Biopharm. 2009;73:260–8.
Serajuddin ATM, Sheen PC, Augustine MA. Water migra- Ungaro F, Giovino C, Catanzano O, Miro A, Mele A,
tion from soft gelatin capsule shell to fill material and its Quaglia F, La Rotonda MI. Use of cyclodextrins as
effect on drug solubility. J Pharm Sci. 1986;75:82–4. solubilizing agents for simvastatin: effect of
hydroxypropyl-β-cyclodextrin on lactone/hydroxyacid Yang P, Li Y, Li W, Zhang H, Gao J, Sun J, et al.

aqueous equilibrium. Int J Pharm. 2011;404:49–56. Preparation and evaluation of carfentanil nasal spray
Urbanetz NA, Lippold BC. Solid dispersions of employing cyclodextrin inclusion technology. Drug
nimodipine and polyethylene glycol 2000: dissolution Dev Ind Pharm. 2018;44(6):953–60.
properties and physico-chemical characterization. Eur Yano H, Kleinebudde P. Improvement of dissolution
J Pharm Biopharm. 2005;59:107–18. behavior for poorly water-soluble drug by application
Venuti V, Cannavà C, Cristiano MC, Fresta M, of cyclodextrin in extrusion process: comparison
Majolino D, Paolino D, Stancanelli R, Tommasini S, between melt extrusion and wet extrusion. AAPS
Ventura CA. A characterization study of resveratrol/ PharmSciTech. 2010;11(2):885–93.
sulfobutyl ether-β-cyclodextrin inclusion complex and Yousaf AM, Kim DW, Cho KH, Kim JO, Yong CS, Choi
in vitro anticancer activity. Colloids Surf B HG. Effect of the preparation method on crystallinity,
Biointerfaces. 2014;115:22–8. particle size, aqueous solubility and dissolution of dif-
Villiers, A. (1891). Sur la fermentation de la fécule par ferent samples of the poorly water-soluble fenofibrate
l’action du ferment butyrique. Compt Rend Acad Sci, with HPβCD. J Incl Phenom Macro. 2015;81(3–4):
112, 536-538 347–56.
Washington N, Washington C, Wilson C. Physiological Yousaf AM, Malik UR, Shahzad Y, Mahmood T, Hussain
pharmaceutics: barriers to drug absorption. CRC T. Silymarin-laden PVP-PEG polymeric composite for
Press; 2000. enhanced aqueous solubility and dissolution rate: prep-
Werner B (1988) Soft gelatin capsules and method for aration and in vitro characterization. J Pharm Anal.
their production. US Patent 4,744,988. 2019;9(1):34–9.
Wu CY, Benet LZ. Predicting drug disposition via appli- Zhang JQ, Li K, Cong YW, Pu SP, Zhu HY, Xie XG,
cation of BCS: transport/absorption/elimination inter- Jin Y, Lin J. Preparation, characterisation and bioactiv-
play and development of biopharmaceutics drug ity evaluation of the inclusion complex formed
disposition classification system. Pharm Res. between picoplatin and β-cyclodextrin. Carbohydr
2005;22:11–23. Res. 2014;396:54–61.
Yajima T, Umeko N, Itai S. Optimum spray congealing Zhao L, Yalkowsky SH. Stabilization of eptifibatide by
conditions for masking the bitter taste of cosolvents. Int J Pharm. 2001;218:43–56.
clarithromycin in wax matrix. Chem Pharm Bull. Zhou Y, Zhao Y, Niu B, Luo Q, Zhang Y, Quan G, et al.
1999;47(2):220–5. Cyclodextrin-based metal-organic frameworks for pul-
Yalkowsky SH, Flynn GL, Amidon GL. Solubility of monary delivery of curcumin with improved solubility
nonelectrolytes in polar solvents. J Pharm Sci. and fine aerodynamic performance. Int J Pharm.
1972;61(6):983–4. 2020;588:119777.
Yang T, Hussain A, Paulson J, Abbruscato TJ, Ahsan Zoppi A, Delrivo A, Aiassa V, Longhi MR. Binding of
F. Cyclodextrins in nasal delivery of low-molecular- sulfamethazine to β-cyclodextrin and methyl-
weight heparins: in vivo and in vitro studies. Pharm β-cyclodextrin. AAPS PharmSciTech. 2013;14:727–
Res. 2004;21(7):1127–36. 35.
Injectable Formulations of Poorly
Water-Soluble Drugs 6
Hannah L. O’Mary and Zhengrong Cui
Abstract conventional and investigational, to overcome

Each year, an increasing number of new challenges of poor aqueous solubility. The
molecular entities in the drug development current authors would like to thank and
pipeline are characterized as having poor acknowledge the previous authors’ significant
water solubility. While solubility limitations contribution from the first and second editions.
are critically important in all formulations, This current third edition chapter is a revision
failure to properly understand and address sol- and update of the original authors’ work from
ubility challenges in parenteral formulations the first and second editions.
can result in significant toxicities and adverse
events. As such, it is imperative that scientists Keywords
in the parenteral development space under- Parental formulation · Injectable drug
stand the vast array of solubilization delivery · Solubility enhancement ·
techniques available for injectable Emulsions · Liposomes · Nanosuspensions ·
formulations. Relatively simple approaches Nanoparticles · Micelles · Hydrogels · In-situ
such as pH manipulation, salt formation, use forming implants
of a cosolvent system, addition of a surfactant,
or complexation may be used to provide suffi-
cient aqueous solubility and favorable drug 6.1 Introduction
product properties. More complex approaches
may be warranted for tissue targeting, alter- Injectable formulations encompass a wide range
ation of pharmacokinetic profiles, and of uses and routes of administration. Although
sustained drug release. This chapter provides sometimes less preferred due to patient compli-
an overview (though certainly not exhaustive) ance and strict production requirements,
of applicable formulation approaches, both injectable formulations are critically important
in a variety of indications. By direct administra-
tion to systemic circulation or target tissue,
H. L. O’Mary (*)
Sterile and Specialty Drug Products, Merck & Co., Inc., injectable formulations reduce inter/intra subject
Rahway, NJ, USA variability, allow for rapid onset of action, and
e-mail: [email protected] enable greater control of pharmacokinetics.
Z. Cui (*) Increasingly, using self-administrable devices,
Division of Molecular Pharmaceutics and Drug Delivery, injectable formulations represent an important
The University of Texas at Austin, Austin, TX, USA category of medications in the treatment of
218 H. L. O’Mary and Z. Cui
Table 6.1 Newly approved injectable products of poorly water-soluble drugs (2020)
Route of Solubilization
Product administration Indication technique
BARHEMSYS® Intravenous Prevention of postoperative nausea/vomiting Salt formation
(amisulpride) infusion
injection
BYFAVO™ Intravenous Induction and maintenance of procedural sedation Salt formation
(remimazolam) for injection
injection
IMCIVREE Subcutaneous Chronic weight management due to proopiomelanocortin Salt formation
(setmelanotide) injection (POMC), proprotein convertase subtilisin/kexin type
injection 1 (PCSK1), or leptin receptor (LEPR) deficiency
VEKLURY® Intravenous Treatment of COVID-19 requiring hospitalization Cyclodextrin
(remdesivir) for infusion complexation
injection
ZEPZELCA™ Intravenous Treatment of metastatic small cell lung cancer (SCLC) pH adjustment
(lurbinectedin) for infusion with disease progression on or after platinum-based
injection chemotherapy.
chronic diseases. In 2019, 7 of the top 10 best- design” approach, patient-centric quality
selling commercial drug products were injectable characteristics can be translated into quality
formulations, and the top seller (Humira®) is attributes of the drug product, known as a target
formulated as a prefilled pen for at-home admin- product profile (TPP). Details in a comprehensive
istration by patients (Top selling drugs by TPP include intended clinical use, route of admin-
revenue, 2020). istration, dosage form, dosage strength, drug
In addition to the many commercial product release rate, and product sterility, stabil-
indications, injectable formulations are often ity, or purity. Careful consideration of the TPP
used during drug product development in preclin- allows for the design and implementation of a
ical and clinical studies to assess toxicity and consistent manufacturing process. In addition to
pharmacokinetic parameters. Reports suggest the TPP, regulatory implications and properties of
that up to 90% of drug candidates in the pharma- the therapeutic moiety will also determine the
ceutical pipeline fall under BCS Class II or Class acceptable characteristics of the final injectable
IV, indicating poor solubility. Examples of newly product.
approved, poorly water-soluble products are
listed in Table 6.1 (Bahremsys 2020, Gilbert
2020, Byfavo 2020, Imcivree 2020, Veklury 6.3 Routes of Administration
2020, Larson 2020, Zepzelca 2020). As this per-
centage has risen, so have the methods and 6.3.1 Intravenous
approaches for formulating these compounds.
This chapter provides guidance for improving Intravenous (IV) administration involves injec-
the solubility of poorly water-soluble drugs and tion of the drug product directly into a patient’s
for developing a stable product for injectable vein. This can be performed as a bolus or “push”
delivery. dose where the drug is administered rapidly, as a
slow infusion where the drug is given over a few
minutes, or as a continuous infusion where
6.2 Target Product Profile administration can last several hours to days.
Bolus dosing allows for rapid serum concentra-
Prior to developing an injectable formulation, it is tion increases and is useful when an immediate
critical to define the objectives for the final drug therapeutic effect is warranted. Because it can be
product. Using a pharmaceutical “quality by maintained for prolonged periods, IV infusion
6 Injectable Formulations of Poorly Water-Soluble Drugs 219
ensures more accurate serum concentrations and depot systems for sustained release (Bari 2010,
can be used to administer larger volumes to the Bhalla 2007).
patient. Drugs with short half-lives may also be Approved in 2019, REDITREX (methotrex-
administered through IV infusion if continued ate) injection for SC use is supplied as a prefilled
therapeutic effect is necessary (Aschenbrenner syringe in a needle safety device for treatment of
and Venable 2009). severe rheumatoid arthritis and polyarticular juve-
IV injection of a drug product is typically nile idiopathic arthritis.
performed using a sterile needle and syringe or
using a sterile catheter as part of an IV infusion
set. Drug products that might be irritating through 6.3.3 Intramuscular
other routes (such as intramuscular or subcutane-
ous) are often given through IV administration. Intramuscular (IM) administration involves the
Drug products that require precise control over injection of a bolus dose of drug product directly
serum concentrations utilize IV administration as into the muscular tissue, most commonly the glu-
it avoids the need for absorption across a mucosal teal, vastus lateralis, or deltoid muscles. As with
membrane to reach blood circulation making drug SC injections, the rate of absorption into the blood
immediately bioavailable. Because IV adminis- stream is dependent on the properties of the formu-
tration is often reserved for critically ill patients lation and therapeutic moiety. However, the rate of
receiving multiple medications, compatibility absorption for IM injections is typically quicker
with various diluents and IV-line materials is and more predictable than that of SC injections,
essential and will be discussed further in Sect. as more blood flow is available to the muscle
6.4.7. than to the fatty tissue of the subcutaneous
layer (Bhattacharjee and Thoma 2010). IM
injections are limited by injection volume as well,
6.3.2 Subcutaneous which is typically 4 mL or less, but may consist of
suspensions or sustained-release products. This
Subcutaneous (SC) administration involves route is typically used for vaccines but may also
injecting the drug product below the dermis and be used to deliver other therapeutics including
epidermis into the fatty, subcutaneous layer in the corticosteroids and antibiotics. Poorly soluble
skin. Typical sites of injection include the upper forms of antipsychotic drugs (e.g. aripiprazole
arm, upper leg, or abdomen, and the injection is and paliperidone) have been formulated into
usually performed with a small, thin needle. The extended-release products for intramuscular
drug can diffuse into the vasculature within the administration, designed to increase patient com-
subcutaneous layer, though the time for this dif- pliance. Both INVEGA TRINZA®, paliperidone
fusion is dependent on both the formulation and palmitate, and ARISTADA™, aripiprazole
properties of the active pharmaceutical lauroxil, utilize fatty acids to sustain release of
ingredient. drug into the bloodstream, achieving durations of
SC administration is limited by volume (less 3 months and 6 weeks, respectively (Carter 2012).
than 2 mL) but is often used for products that
cannot be given orally due to instability or poor
absorption. An example of this would be insulin, 6.3.4 Intrathecal
which is typically administered through this
route. Suspensions, which are typically not viable Intrathecal (IT) injection is a route of intraspinal
for IV administration, may also be given through administration that delivers drug directly into the
the SC route. Oil-based formulations may be cerebrospinal fluid (CSF). Injection involves
given through SC administration to form drug direct placement of the needle into the
subarachnoid space of the spinal column. Drugs 6.3.5 Intra-Articular

delivered through this route are not available to
systemic circulation and are maintained within Injectable formulations may also be administered
the CSF, allowing for drug to bypass the blood– directly at the site of action, as is the case with
brain barrier (Dodou 2012). intra-articular (IA) injections. IA injections
The CSF is a sterile liquid without immune involve the administration of drug into the joint
defense capabilities; therefore, sterility of space using a sterile needle and syringe. The
formulations intended for intrathecal administra- needle size and dose volume are dependent upon
tion is exceedingly important. Additionally, due the joint that is being injected (Cardone and Tallia
to the potential for inflammatory-mediated 2003). For example, injections of 3 mL with an
reactions and neurotoxicity, intrathecal 18- to 22-gauge needle are tolerated in the knee,
preparations must be preservative-free, reiterating while injection volumes of only 0.25 to 0.5 mL
the need for aseptic processing. with a 25-gauge needle are acceptable in the first
Dosing volumes of 5 mL or less are preferred metacarpal joint of the hand. This route is fre-
for intrathecal administration to reduce the risk quently used to administer products for the relief
for increases in intracranial pressure. Specially of pain and inflammation and may be beneficial to
designed intrathecal catheter sets may be used in deliver drugs that are poorly distributed to the
patients requiring frequent and repeated joint following IV administration or that have
injections. Because the CSF is isotonic, products significant systemic toxicities (Gerwin et al.
should be prepared as close to physiological 2006, Lavelle et al. 2007, Neustadt 2006).
tonicity as possible. While intrathecal Potential formulations for IA injection may
preparations may contain buffers, inclusion of include solutions, suspensions, liposomal
additional excipients should be limited. Because preparations, hydrogels, or micro/nanoparticles.
the CSF is susceptible to introduced foreign mat- Solutions may be prone to rapid clearance from
ter, the limits of contamination are restricted fur- the joint and require repeated administrations,
ther for intrathecal preparations than for other increasing the risk for infection. Preferably, IA
injectable formulations. Therefore, it is essential injectables should provide sustained therapeutic
that excipients be limited to reduce potential effect and be free from any potential immuno-
exposure and possible toxicity. Table 6.2 lists genic components that might elicit an unwanted
examples of poorly soluble drugs formulated immune response (Larsen et al. 2008). Table 6.3
into commercial products for intrathecal injec- lists selected IA products that are currently
tion (Clorotekal 2017, Lioresal Intrathecal 2013).
Table 6.2 Commercial products for intrathecal injection

Product Formulation Inactive ingredients Dose
CLOROTEKAL® (chloroprocaine Solution Hydrochloric acid 1 N (for pH adjustment), sodium 5 mLa
hydrochloride) injection chloride, water for injection
LIORESAL® INTRATHECAL Solution 0.5 mg or 2 mg and sodium chloride 9 mg in water 0.5 to
(baclofen) injection for injection 2 mL
a
Prescribing information recommends 5 mL but suggests a weight-based calculation to confirm dose
Table 6.3 Commercial products for intra-articular injection

Product Formulation Dose and frequency Status
CELESTONE® SOLUSPAN injection (betamethasone Suspension 0.5 to 2 mL (0.08 to Marketed
acetate, betamethasone sodium phosphate) 0.33 mg) per day
ZILRETTA injection (triamcinolone acetonide) Microspheres 0.06 to 2.5 mL (2.5 to Marketed
100 mg) per day
Table 6.4 A comparison of common injectable routes of administration

Intravenous Subcutaneous Intramuscular Intrathecal Intra-articular
Maximum 10 mL (bolus) 2 mL 4 mL 10 mL Site dependent (e.g. 3 mL
volume No limit for for knee, 0.5 mL for first
infusion (max metacarpal joint)
rate 3 mL/min).
Formulations Solutions. Solutions. Solutions. Solutions. Solutions.
Colloidal Suspensions. Suspensions. Suspensions. Suspensions.
dispersions Oil-based Oil-based Oil-based formulations.
(<1 μm). formulations. formulations.
Time for Instantaneous, Typically Less than No absorption, Local action only.
absorption no absorption within 20 minutes direct delivery
phase. 20 minutes. (usually faster to brain.
than SC).
Injection site Peripheral veins Upper arm. Dorsogluteal Subarachnoid Directly into joint space.
(s) (smaller Upper leg. muscle. space in spinal
volumes). Abdomen. Deltoid muscle. column.
Central veins Ventrogluteal
(larger muscle.
volumes). Vastus lateralis
muscle.
marketed (Celestone Soluspan 2015, Zilretta range dependent on the volume, rate of adminis-
2017). tration, and buffer capacity of the product. With
regard to volume, pH restrictions are tighter on
large volume administrations (defined by the USP
as IV volumes exceeding 100 mL) as compared to
6.3.6 Other Injectable Routes
slowly injected small volume preparations. While
the range of 2 to 12 is typically accepted, products
Injectable formulations are not limited to the pre-
should ideally be formulated as close to the phys-
viously mentioned routes, which are summarized
iological pH as possible, with pH values greater
in Table 6.4. The FDA also lists several other
than 3 and less than 10.5 preferred.
routes of administration, many of which would
Subcutaneously-administered products are
require the use of an injectable product. Though
restricted even further to a range of 2.7 to 9, due
less commonly used, these routes may be required
to the decreased capacity for in vivo dilution (Shi
for specialized delivery of a parenteral product.
et al. 2009). Careful considerations must be taken
Careful investigation of the challenges and
to ensure that products formulated away from
limitations of each of these routes must be thor-
physiological pH do not precipitate upon dilution
oughly explored before development of a drug
in circulation following administration. Addition-
product.
ally, pH must be considered regarding overall
product stability, including solubility of the active
pharmaceutical ingredient and the functionality of
6.4 Formulation Considerations other excipients, such as preservatives.
for Injectable Products
6.4.1 pH
6.4.2 Buffer Systems
One of the initial parameters that need to be
considered when formulating injectable products A buffered system is used to maintain the
is pH. Formulations designed for IV or IM admin- intended pH of a formulation throughout the
istration may be developed at a pH range from shelf-life of the product. Preformulation solubility
2 to 12; however, there are constraints to this and stability assessments are conducted to
determine the pH range that provides optimal Table 6.5 A list of tonicity-modifying agents from
solubility and chemical stability from degradation USP43-NF38 2S
products. Thorough reviews of preformulation sol- Electrolytes Mono- or disaccharides Others
ubility and stability assessments have been provided Sodium Dextrose Glycerin
by several authors (see Halbert 2009; Tong and chloride
Potassium Mannitol
Wen 2008) and will not be discussed here.
chloride
Injectable formulations commonly make use Corn syrup
of buffer systems containing a weak acid or Corn syrup
weak base with a corresponding salt. Once the solids
optimal pH range is determined, appropriate
buffers for this range can be tested for compati-
bility with the drug product. If possible, multiple potential complications by utilizing specific
buffers should be tested to ensure minimal degra- means of preparation or by altering the route of
dation of the drug product in the buffer or changes administration of the product. Drug products may
in the solubility of the drug. In addition to probe mixed with sufficient isotonic diluent (which
moting the shelf-life of the drug product, buffer will be discussed further in Sect. 6.4.7) prior to
concentrations should be selected that minimize dosing, or a tonicity modifier may be added into
alterations in blood pH following injection and the formulation. A list of common tonicity-
decrease the potential for injection site pain or modifying agents is provided in Table 6.5 (from
irritation. USP43-NF38 2S 2021). Small volumes of hypo-
tonic or hypertonic solutions may be administered
without tonicity modification either by using a
low injection rate or by injecting into large veins
6.4.3 Tonicity and Biological
where they are able to undergo rapid dilution in
Implications
the circulation (Troy 2006).
Methods to characterize and quantify the
Tonicity is another critical factor that must be
tonicity of a solution include the freezing-point
considered when developing an injectable formu-
depression method, sodium chloride equivalent
lation. “Isotonic” solutions are defined as having
method, Van’t Hoff equation method, and the
an equal osmotic pressure with respect to a par-
White–Vincent method. Of these, freezing-point
ticular membrane, and therefore no net movement
depression and vapor-pressure osmometer are
of fluid from one side to another. In the case of
most used to quickly determine the osmolality
intravenous drug products, the membrane of
of solutions and compare it to that of plasma
interest is that of the blood cells and cardiovascu-
(275 to 299 mOsm/kg) (Troy 2006). Unlike tonic-
lar tissue. The consequences of altered tonicity on
ity, osmolality takes into consideration both per-
red blood cells are swelling (in the case of hypo-
meable and impermeable solute molecules. As
tonic solutions) or crenation (in the case of hyper-
such, tonicity is always equal to or less than
tonic solutions), both of which can result in cell
osmolality and can be defined as the osmolality
destruction and toxicity. To prevent cell damage
minus the concentration of freely permeable
and maintain fluid homeostasis within the blood
solutes (See USP-NF<785> Osmolality and
cells, an injectable product should be developed
Osmolarity, 2020 for reference). Knowing this,
as close as possible to the tonicity of physiologi-
it is possible to estimate the potential fluid
cal fluids.
interactions with plasma of an injectable product.
Often, however, it is not feasible to develop an
Solutions with a high osmolality or that contain
isotonic formulation due to product-specific
excipients that alter freezing but do not contribute
factors, such as high drug concentrations, low
to tonicity (e.g. propylene glycol) may be
injection volumes, or stability considerations. In
iso-osmotic but not isotonic. In this instance, it
these instances, it is possible to overcome
may be more appropriate to evaluate the tonicity the membrane upon cell death. Also known as
of such a solution through observation of erythro- lipopolysaccharide (LPS), these endotoxins cause
cyte changes, such as red blood cell volume, in a host of inflammatory responses at certain levels
the presence of the formulation. of exposure (Tran and Whitfield 2009).
A more predictive and definitive test for As more components are added to a formula-
evaluating the compatibility of a formulation tion, including active ingredients, excipients, and
with blood is the in vitro hemolysis test, which solvents, the risk for endotoxin exposure is
provides visual observations on the changes red increased for a given drug product. Each compo-
blood cells may undergo after mixing with the nent of a formulation should be scrutinized for
injectable solution. In vivo tests are also endotoxin levels to reduce patient exposure. For
performed to determine potential adverse events nonintrathecal injectable products, the limit is
during administration, including precipitation 5 endotoxin units (EU) per kg, equivalent to
upon injection, phlebitis (inflammation of the 500 pg of LPS from Escherichia coli, divided
vasculature), pain, or irritation at the injection by the maximum human dose given per hour.
site. Regulatory guidelines should be consulted Methods for determining endotoxin levels are
for the specific route of administration to deter- detailed in USP<85> (see USP<85> Bacterial
mine the appropriate dosing volume, injection Endotoxins Test, 2020).
rate, and levels of ingredients that limit
injection-related adverse events.
6.4.5 Particulate Matter
6.4.4 Sterility and Endotoxin Particulate matter, according to USP < 788>,
Requirements consists of “mobile undissolved particles, other
than gas bubbles, unintentionally present in solu-
For all injectable products, sterility must be tion.” Visible and subvisible particulate matter
demonstrated in accordance with microbial limit must be determined in an injectable product, and
assay requirements of USP < 71> (see USP<71> the two methods for doing so are detailed in
Sterility Tests, 2020). As seen in Table 6.6, steril- Table 6.7. These methods are reviewed more
ization methods may or may not be terminal. extensively in USP < 788> (see USP<788> Par-
From a cost and regulatory standpoint, terminal ticulate Matter in Injections, 2020). For drug
methods are preferred; however, this option is not products where viscosity or sample clarity limits
available to many drug products due to stability the ability to be tested by one, or both, of these
concerns. For these products, sterile filtration is methods, the solution may be diluted with
often used. particulate-free diluent in order to conduct the
In addition to microbial stability, limits of test. Parenteral products must also be inspected
endotoxins are also placed on injectable products. for visual particulates, which is outlined in
Endotoxins are a part of the outer cell wall in USP < 790> (see USP<790> Visible Particulates
Gram-negative bacteria that are released from in Injections, 2020).
Table 6.6 A list of methods of sterilization

Terminal methods Nonterminal methods
Steam (autoclaving) Filtration
Dry heat Aseptic processing
Radiation (gamma or electron-beam)
Gas (ethylene oxide)
Table 6.7 Methods for determining particulate matter from USP<788>

Light obscuration particle count test (Method 1) Microscopic particle count test (Method 2)
Principle Light blockage determines the size and number of Particles are counted and size determined by use
particles. of an ocular micrometer-equipped microscope.
Methods Samples of 5 mL or more are drawn from the Samples are filtered, and the membrane filter is
preparation. allowed to dry in a petri dish.
Particles equal to or greater than 10 or 25 μm are Filter is examined under the microscope.
counted.
Mean number of particles calculated. Particles equal to or greater than 10 or 25 μm are
counted.
Mean number of particles is calculated.
Evaluation Solutions greater than 100 mL must not exceed Solutions greater than 100 mL must not exceed
25 per mL of particles >10 μm and 3 per mL of 12 per mL of particles >10 μm and 2 per mL of
particles >25 μm. particles >25 μm.
Solutions less than 100 mL must not exceed 6000 Solutions less than 100 mL must not exceed 3000
per container of particles >10 μm and 600 per per container of particles >10 μm and 300 per
container of particles >25 μm. container of particles >25 μm.
Limitations May not be suitable for emulsions, colloids, or
liposomal preparations.
Table 6.8 Common preservatives used in parenteral formulations

Maximum potency per
Preservative unit dose Example
Benzalkonium 0.02% w/v CELESTONE® SOLUSPAN injection (betamethasone acetate,
chloride betamethasone sodium phosphate)
Benzyl alcohol 2.02% w/v SOLU-MEDROL® for injection (methylprednisolone sodium succinate)
Chlorobutanol 0.57% w/v DDAVP® injection (desmopressin acetate)
Metacresol 0.3% w/v INTRON® A for injection (interferon alfa-2b, recombinant)
Methylparaben 5% w/v Hydralazine hydrochloride for injection, USP
Phenol 0.55% w/v ZANTAC® for injection (ranitidine hydrochloride)
Propylparaben 0.2% w/v Heparin sodium injection, USP
Thimerosal 0.3 mg Fluvirin® (influenza virus vaccine)
6.4.6 Preservatives high cosolvent concentration or the inclusion of

a bacteriostatic drug, many formulations will
If an injectable product is intended for multidose require the use of preservatives (De Spiegeleer
applications, it must meet regulatory standards et al. 2006). Table 6.8 contains a list of
regarding antimicrobial effectiveness testing preservatives and their typical concentrations
(AET). As detailed in USP <51> (see USP<51> used in injectable formulations (Celestone
Antimicrobial Effectiveness Testing, 2020), AET Soluspan 2015, Solu-Medrol 2011, DDAVP
requirements involve inoculating the drug prod- 2007, Intron 2015, Hydralazine hydrochloride
uct with a known concentration of specific injection 2013, Zantac 2015, Fluvirin 2015).
microorganisms (i.e. bacteria, yeasts, and molds) When choosing a preservative, it is imperative
and evaluating for resistance to microbial prolif- that factors which can affect the performance of the
eration. In order to pass these requirements, drug preservative are evaluated. Both the pH of the drug
products must provide a specific log reduction in product and other excipients included in the formu-
bacteria and no growth of yeast or molds at lation can decrease the effectiveness of
defined time points after inoculation. While preservatives (Meyer et al. 2007). Furthermore, the
some formulations may be able to accomplish patient population and route of administration
this through self-preserving properties such as should be evaluated regarding the inclusion of
Table 6.9 Common diluents for parenteral preparations

Diluent Tonicity
Sterile water for injection Hypotonic
0.9% sodium chloride for injection Isotonic
0.45% sodium chloride for injection Hypotonic
5% dextrose for injection Isotonic
Lactated Ringer’s for injection Isotonic
5% dextrose in lactated Ringer’s for injection Hypertonic
5% dextrose in 0.9% sodium chloride for injection Hypertonic
5% dextrose in 0.45% sodium chloride for injection Hypertonic
preservatives in a formulation. For example, syringes, IV infusion bags, IV tubing lines, and
injections intended for newborns, as well as other materials used as part of administration sets.
injections intended for epidural or intrathecal injec- These materials must be assessed, as well, for
tion, should be preservative-free (Soni et al. 2005). compatibility with the drug product. Examples
of incompatibilities include adsorption of the
drug, alteration in pH of the drug product, and
leaching of materials into the medication.
6.4.7 Device and Diluent
Compatibility
As previously mentioned, diluents may be used 6.4.8 Packaging and Manufacturing

with many injectable products prior to adminis- Considerations
tration. Therefore, it is crucial that the compati-
bility between a given diluent and diluent volume While careful consideration is taken into ensuring
is carefully assessed during product development. drug product stability during development and
The selection of a diluent is product-specific and formulation, one important component that is
dependent on formulation factors including drug often overlooked is the container closure of the
concentration, dosing regimen, pH, and tonicity. drug product, which includes the primary and
Table 6.9 lists common diluents used with paren- secondary packaging that holds the product for
teral preparations (Trissels 2002). the duration of its shelf-life. These studies should
To test diluent and drug product compatibility, be conducted early in the development process to
the drug is diluted with diluent to a final dosing identify any potential stability issues that may
concentration and evaluated for physical and arise between the drug product and packaging
chemical stability with respect to time, tempera- and to ensure that stability requirements are
ture, and storage conditions to ensure a sufficient maintained throughout the shelf-life of the prod-
stability of the drug product during the intended uct. A thorough review by Akers et al. (Akers
use period. In many cases, the stability of the drug 2012) details some of the options available for
product after dilution is limited; therefore, the product packaging and their benefits in specific
drug must be delivered within the intended use formulations.
period. Additionally, drugs delivered via the IV Product packaging must not only be compati-
route are often given alongside other IV ble with drug stability, it must also adhere to
medications and fluids. For this reason, safety particulate and sterility requirements. Container
and compatibility will need to be established for closure integrity tests are performed to validate
product labeling purposes. that no leakage occurs from the package and that
During preparation and administration, sterility is not compromised. Seal integrity testing
injectable formulations will also encounter many is also important for drug products prone to oxi-
different materials, including filters, needles, dation and that require inert headspace for
stability. Light sensitivity must also be considered manufactured, from both a time and cost perspec-
when evaluating product packaging. Amber vials tive, as well as efficiently utilized by the end
or opaque outer cartons may be used to protect the consumer.
product, depending on the level of light sensitiv-
ity. Ultimately, the appropriate container closure
selection is mediated by the properties of the drug 6.5.1 pH Adjustment
formulation, including product type
(i.e. lyophilized powder, solution, or powder for For compounds that ionize, adjusting the pH of
reconstitution), compatibility with packaging the solvent is one of the first techniques that can
components, method of sterilization, and route be explored to improve solubility. Knowledge of
of administration. Regulatory restrictions on spe- the pKa of a drug can be used to formulate a
cific materials must also be taken into consider- product at a pH that provides maximum solubility
ation. (Troy 2006). of the compound in solvent. The solubility of a
During manufacture, the drug product may weak acid, for example, may be increased
also be exposed to a number of materials. Com- ten-fold by adjusting the pH to 1 pH unit above
patibility studies must be conducted here, as well, the acidic pKa. Likewise, the solubility of a weak
to demonstrate that the manufacturing process base may be increased by adjusting the pH to
does not alter the stability and release 1 pH unit below the basic pKa. As mentioned
specifications of the drug product. Incompat- previously, there are acceptable pH ranges for
ibilities may exist, for example, between some each route of administration, and as such, the
liquid formulations and stainless-steel adjustments made to improve solubility will be
compounding vessels, resulting in color change defined by these ranges. Hydrochloric acid and
or degradant formation. Increased residence time sodium hydroxide are the preferred acid and base
in tubing during filling or filtration processes may used to titrate the pH of a solution.
result in a loss of potency due to sorption. Satura- Controlling the final pH of a solution may
tion of lines and filters and flushing with sufficient prove difficult, in which case buffers can be
volumes of drug product prior to filling can be added to the formulation to ensure that an accu-
utilized to ensure the proper potency of the final rate pH-solubility profile is produced. To assess
product. solubility and stability of a compound, buffer
solutions are prepared at a specific pH and con-
stant ionic strength. These solutions are then
6.5 Vehicle Selection mixed in appropriate concentrations to prepare
and Solubilization solutions between pH 3 and 9. Buffers that are
effective in this pH range are required, such as
Characterization of the solubility and physical/ citrate (effective from pH 2 to 6) and phosphate
chemical stability of a drug can be used to select (effective from pH 6 to 8).
the appropriate vehicle to deliver the injectable An example of an injectable product that
product. Simple formulations of an aqueous vehi- utilizes pH adjustment to improve solubility is
cle are often not enough to provide acceptable Cardene® I.V. premixed injection (Cardene
solubility for poorly water-soluble compounds. 2010). This formulation is a solution of nifedi-
Instead, these compounds may require additional pine, a BCS Class III drug, in either dextrose or
excipients to promote solubility or special sodium chloride. Nifedipine is a weak acid (pKa
processing methods to maintain stability. A bal- 3.93) with poor water solubility (approximately
ance must be achieved between improving the 10 μg/mL). The pH of Cardene® I.V. premixed
solubility and stability of a compound and injection is adjusted to a range of 3.7 to 4.7 in
maintaining the simplicity of the formulation. In order to provide sufficient solubility of the drug
addition to meeting regulatory requirements, an for administration at 20 mg per 200 mL (0.1 mg/
injectable product must also be efficiently mL).
Table 6.10 Available counter ions for salt formation

Cations Anions
Aluminum Magnesium Acetate Lactobionate
Arginine Histidine Aspartate Malate
Benzathine Lithium Benzenesulfonate Maleate
Calcium Meglumine Benzoate Mandelate
Chloroprocaine Potassium Besylate Mesylate
Choline Procaine Bicarbonate Methylsulfate
Diethanolamine Sodium Bitartrate Mucate
Ethanolamine Triethylamine Bromide Napsylate
Ethylenedamine Zinc Camsylate Nitrate
Lysine Carbonate Octanoate
Chloride Oleate
Citrate Pamoate
Decanoate Pantothenate
Edetate Phosphate
Esylate Polygalacturonate
Fumarate Propionate
Gluceptate Salicylate
Gluconate Stearate
Glutamate Acetate
Glycolate Succinate
Hexanoate Sulfate
Hydroxynaphthoate Tartrate
Iodide Teoclate
Isethionate Tosylate
Lactate
Table adapted from Gupta et al. (2018)
6.5.2 Salt Formation most soluble. Counter ions (as shown in

Table 6.10) may be used to prepare various salt
The addition of a buffer to a solution may influ- forms of the compound that may then be screened
ence the solubility or stability of a compound. In for the desired solubility and stability profile.
some instances, the solubility may decrease in the Characterization of the solid-state properties of
presence of a buffer, indicating formation of a these salt forms, including crystallinity, polymor-
less-soluble salt, polymorph, or the common-ion phism, and stability, must be conducted prior to
effect. However, in some instances, the presence further development (Gupta 2018).
of a buffer may improve solubility, specifically Salt formation is a common technique
through the formation of a more soluble salt. employed to improve drug solubility in many
Therefore, salt formation may also be utilized as commercialized, injectable products. Doxycy-
a technique to improve solubility (Serajuddin cline hyclate salt, originally marketed under the
2007). brand name of Vibramycin® and currently sup-
While many compounds are supplied as the plied as a generic product, improves the solubility
more soluble salt, it is also possible to form of the weak base doxycycline for intravenous use.
these in situ in the presence of counter ions and Doxycycline has poor water solubility, whereas
at a given pH. Though HCl and NaOH may be doxycycline hyclate is soluble in water, allowing
used as pH modifiers, these salts formed with the for the formation of a product for
active ingredient are typically not stable nor the reconstitution (Vibramycin 2014).
6.5.3 Cosolvents allows for intravenous or intramuscular injections

of 2 or 4 mg of the drug in a volume as small as
In cases where the aforementioned techniques are 1 mL (Ativan 1980).
not sufficient to achieve an acceptable solubility,
the use of a cosolvent may be implemented. A
cosolvent is a water-miscible organic solvent 6.5.4 Surfactants
used in an aqueous solution, or in combination
with other cosolvents, to increase solubility. This The addition of surfactants is another technique
is done by altering the polarity of the solution, by which solubility may be increased or stability
typically decreasing it such that less polar improved in a formulation. Because of their
compounds may be solubilized in the vehi- amphiphilic structure, surfactants contain both
cle (Vemuri 2010). Propylene glycol and ethanol hydrophilic and hydrophobic groups that allow
are two of the most used cosolvents; however, for reduction in the interfacial tension between
there are many others used in commercially avail- solid particles and liquid, thus improving solubil-
able products. ity. Lipophilic, nonpolar, and nonionizable
As with pH adjustment and buffer systems, the compounds are typically the best candidates for
addition of a cosolvent changes the properties of a improving the solubility with a surfactant-
solution, which might also alter the stability of the containing formulation. Nonionic surfactants are
compound. The appropriate concentration of a more widely used in the pharmaceutical industry
cosolvent should be selected to optimize both as compared with anionic, cationic, or amphoteric
solubility and stability, while not compromising surfactants because of their effectiveness and
the compatibility of the injection with blood and lower toxicity.
the intended route of administration. Table 6.11 Due to the amphiphilic nature of surfactants,
lists some of the cosolvents used in injectable the concentration of a surfactant is important in
formulations and their acceptable levels and predicting its behavior in solution. A value
potential toxicity limits (Mottu 2015). In order known as the “critical micelle concentration”
to reduce pain or injection site discomfort, the (CMC) defines the concentration at which
administration of cosolvents typically involves surfactants begin to form micellar structures.
decreasing the injection rate or, more often, dilu- Listed in Table 6.12 are the CMC values of
tion prior to administration. Cosolvents may be selected surfactants. Below the CMC, as the con-
used along with pH adjustment, buffers, salt for- centration of surfactants increases, as does the
mation, or a combination of these methods to reduction in surface tension. At and above the
achieve the desired solubility and stability of a CMC, surfactants organize into micelles,
product (Strickley 2004). structures often with a hydrophobic core and an
As stated before, formulations may utilize one outer hydrophilic layer. Because of this hydro-
or more cosolvents to enhance solubility. phobic core, micelles are capable of sequestering
ATIVAN® injection contains polyethylene glycol lipophilic drugs and therefore increasing their
400 in propylene glycol to solubilize the water- solubility. They may also protect drugs from deg-
insoluble active ingredient, lorazepam. This radation, including hydrolysis and
Table 6.11 Injectable cosolvents and concentrations in approved injectable products

Cosolvent Range of maximum potency per unit dose (w/v) LD50 in vivo
Propylene glycol 0.5 to 82.04% 26 g/kg (dog)
Polyethylene glycol 300 44.22 to 65% 8 g/kg (rat)
Polyethylene glycol 400 18 to 75.58% 7.4 g/kg (rat)
Ethanol 0.01 to 1.5% 4 mL/kg (mouse)—75%
2.5 mL/kg (mouse)—95%
Dimethyl sulfoxide 5 to 10% IV, 104 mg for SC implant 5.2 to 8.1 mg/kg (rat)
Table 6.12 CMC values of surfactants in aqueous solutions

Surfactant CMC at 25 C
Polysorbate 20 0.006 g/100 mL
Polysorbate 80 0.001 g/100 mL
Solutol HS-15 0.005 to 0.02 g/100 mL
Poloxamer 188 0.0014 g/100 mL (37 C)
Kolliphor EL 0.02 g/100 mL
Table 6.13 Selected cyclodextrins for parenteral use

Cyclodextrin Aqueous solubility Example
Hydroxypropyl-β-cyclodextrin 2300 g/L at 25 C VIBATIV® (televancin) for injection
Hydroxypropyl-α-cyclodextrin 143 g/L at 20 C Edex® (aloprostadil) for injection
Sulfobutylether- β-cyclodextrin >1000 g/L at 25 C NEXTERONE® (amiodarone) premixed injection
oxidation (Croy and Kwan 2006). While the for- polysorbate 80 to solubilize 20 mg of docetaxel
mation of micelles may be beneficial in some through micelle formation. The 40 mg/mL of
formulations, this mechanism of encapsulation docetaxel is further diluted with a water/
may be undesirable in others. For example, dehydrated alcohol cosolvent system to achieve
injectable formulations that are intended to be a final concentration of 10 mg/mL before further
multidose and contain a preservative may dilution for infusion (Taxotere 2010).
become susceptible to contamination through
the formation of micelles, which can entrap
preservatives in the hydrophobic core making 6.5.5 Cyclodextrins
them less available in the aqueous phase to pre-
vent microbial growth. For some poorly water-soluble compounds, com-
Changes in factors that affect the ionic and plexation may be a method to increase their solu-
nonionic properties of a solution, such as temper- bility. Cyclodextrins can be used to form water-
ature, pH, electrolytes, or solvents, may also alter soluble, reversible, noncovalent inclusion
the CMC (Kamat and DeLuca 2010). Therefore, complexes. Cyclodextrins are cyclic oligosac-
it is important that the CMC for a formulation be charides with a hydrophilic exterior and a
well-characterized in order to understand the lipophilic core, which can be used to complex
physical stability of the solution. Techniques, poorly water-soluble compounds. The ability of
such as dynamic light scattering, may be used to a cyclodextrin to solubilize a drug is dependent
determine the CMC, as well as plots of surfactant on several factors, including the size and structure
concentration versus surface tension. of the drug molecule, the stoichiometry of the
Furthermore, careful consideration must be complex, the complexation equilibrium constant,
made when choosing to utilize a surfactant in a and the overall free energy of the system (Challa
formulation. Reports of hypersensitivities, nonal- et al. 2005). Cyclodextrins are typically only
lergic anaphylaxis, rash, and infusion-related slightly soluble in water, which limits their use,
adverse events have been reported, particularly particularly in formulations that require a high
with the use of polysorbate 80 (Schwartzberg level of solubility. Furthermore, this reduced sol-
and Navari 2018). ubility in water poses a toxicity concern for
An example of a marketed product that utilizes injectable formulations due to the potential for
a surfactant is the TAXOTERE® injection. precipitation after administration (Shi et al.
Docetaxel, the active ingredient, is practically 2009). Table 6.13 summarizes the most com-
insoluble in water. Each milliliter of monly used cyclodextrins and their aqueous
TAXOTERE® injection contains 0.54 g of solubilities (Vibativ 2009, Edex 2011, Nexterone
2011). Of these, hydroxypropyl-β-cyclodextrin formulation that utilizes sesame oil to deliver a

and sulfobutylether-β-cyclodextrin are preferred depot of the water-insoluble testosterone deriva-
for use in injectable formulations due to their high tive into deep muscle tissue by IM injec-
aqueous solubilities, efficient complexation tion (Invega Sustenna 2015, Aristada 2015,
capacity, and reduced toxicity. DEPO-Testosterone 2011).
Although complexes between drug molecules
and cyclodextrins are typically 1:1, it is possible
for higher order inclusions to occur where 6.6 Dispersed Systems
complexed molecules are in a rapid dynamic
equilibrium with the free drug molecules in solu- 6.6.1 Emulsions
tion. Therefore, it is important to evaluate how
free drug concentration in solution changes as a Emulsions are abundantly used in pharmaceutical
function of cyclodextrin concentra- formulations for oral, topical, and parenteral
tion (Thompson 1997, Loftsson et al. 2005). formulations. An emulsion is a heterogeneous
Additionally, complexed drug molecules may dispersed system of two immiscible liquids, and
have altered physicochemical properties or classically it is composed of two phases, namely
biological activity as compared with free drug the dispersed phase and the external (continuous)
molecules, which must be characterized prior to phase. The dispersed phase is emulsified into the
further development of the drug product. continuous phase through energy input to form
For example, PREVYMIS™ (letermovir) small-sized droplets that are stabilized by an
injection for IV use, a product used to treat cyto- emulsifier (i.e. emulsifying agent). Emulsions
megalovirus (CMV) infection, utilizes cyclodex- can be oil-in-water (o/w) or water-in-oil (w/o),
trin complexation to promote solubility. The although oil-in-oil (o/o) and multiple emulsions
formulation contains 150 mg/mL of (e.g. w/o/w or w/o/o) can also be found in litera-
hydroxypropyl betadex, also known as ture (Wu et al. 2009, Asano et al. 2015, Goncalez
(2-hydroxypropyl)-β-cyclodextrin, to complex et al. 2015).
and solubilize 20 mg of letermovir, which is Injectable emulsions have been on the market
sparingly soluble in aqueous for over 50 years. One of the earliest products,
buffers (Prevymis 2017). Intralipid®, is a sterile oil-in-water emulsion par-
enteral nutrition product for critically ill patients
comprised of soybean oil (20 or 30%), egg yolk
6.5.6 Nonaqueous and Oily Vehicles phospholipids, glycerin, and water for injection.
Sodium hydroxide is added to adjust the pH to
The use of nonaqueous or oily vehicles may also 8.0. Intralipid® can also be used to dilute other
be employed for poorly water-soluble drugs for emulsion formulations (e.g. Diazemuls®). Other
IM injection. These formulations are typically examples of marketed nutritional emulsions
long acting and require less frequent administra- include SMOFLIPID, which contains 6% soybean
tion to patients. Many antipsychotic medications oil, 6% medium chain triglyceride (MCT), 5%
have been formulated using this method olive oil, and 3% fish oil and OMEGAVEN®,
(e.g. INVEGA SUSTENNA®, ARISTADA®). used in pediatric patients, which contains 10 g
DEPO-Testosterone® injection also utilizes a of fish oil per 100 mL, comprising
nonaqueous oily vehicle to achieve sustained eicosapentaenoic acid (EPA), docosahexaenoic
release. This formulation contains cottonseed oil acid (DHA), and α-tocopherol (Intralipid 1996,
to solubilize water-insoluble testosterone Diazemuls 1998, Smoflipid 2012, Omegaven
cypionate, an ester of testosterone. This product 2010).
is designed for intramuscular use only and is Parenteral emulsions also have application
administered into the gluteal muscle once every outside of intravenous nutrition. Clevidipine
two to four weeks. Nandrolone decanoate injec- butyrate, available as CLEVIPREX®, is a water-
tion (e.g. Deca-Durabolin®) is another insoluble, fourth-generation calcium channel
Table 6.14 Emulsion preparation techniques

Method Final size range Suitable for nanoemulsions? (size in nm scale)
High-pressure homogenization 1 nm to 5 μm Yes
Microfluidization 10 nm to 1 μm Yes
Ultrasonication 20 nm to 1 μm Yes
Phase inversion 10 nm to 200 nm Yes
Spontaneous emulsification 10 nm to 200 nm Yes
Solvent evaporation 10 nm to 200 nm Yes
Hydrogel evaporation 10 nm to 200 nm Yes
Mechanical stirring 1 μm to 15 μm No
Colloidal mills 10 μm to 50 μm No
Membrane emulsification 200 nm to 100 μm No
blocker used to treat hypertension. CLEVIPRE commonly implemented for long-acting

X® is formulated as an oil-in-water emulsion with formulations, particularly for products where
soybean oil (200 mg/mL), oleic acid (0.3 mg/mL) fluctuations in plasma concentration are not
(as the oil phase), glycerol (22.5 mg/mL), and egg well-tolerated or where patient compliance may
yolk phospholipids (12 mg/mL) (as stabilizers). be an issue. Coarse suspensions are not suitable
Clevidipine is solubilized in the oil phase at a fi for IV injection, but may be injected IM or SC to
nal concentration of 0.5 mg/mL. DIPRIVAN®, p provide a controlled release depot or to provide a
ropofol, is another common emulsion product, dose above the solubility limits of the active
indicated for the induction of general anesthesia ingredient in biocompatible vehicles. Colloidal
and sedation. Like CLEVIPREX®, it is also an oil- suspensions (e.g. nanoparticles) may be suitable
in-water emulsion comprised of soybean oil (100 for IV injection, as well, and will be discussed
mg/mL), glycerol (22.5 mg/mL), egg lecithin (12 later. Several aspects must be considered in the
mg/mL), and disodium edetate (0.005%), with development of a parenteral suspension. The
sodium hydroxide to adjust pH (Cleviprex 2011). choice of excipients, particle size distribution,
There are several methods to manufacture physical stability (i.e. caking, Ostwald ripening,
emulsions, which can be found in Table 6.14 (Kale etc.), and sterilization complicate the develop-
and Deore 2017, Nikam et al. 2018). As ment of injectable suspensions. Particle size and
indicated, some methods can produce oil droplets viscosity can affect syringeability and
in the nanoscale range, which may provide addi- injectability, and must remain unchanged
tional stability to the final drug product. Globule throughout the shelf-life of the product.
size distribution of the final emulsion must also be Resuspendability (i.e. the ability to resuspend
controlled, specifically the mean droplet size and after a few shakes) also must be maintained dur-
large-diameter tail. Methods to characterize lipid ing storage. Postproduction sterilization methods
droplet size are outlined in USP <729> and are should not affect the stability of the suspension;
summarized in Table 6.15. (see USP<729> Glob- otherwise, processing of the suspension under
ule Size Distribution in Lipid Injectable aseptic conditions may be required. A list of
Emulsions, 2020). marketed poorly soluble coarse suspensions is
provided in Table 6.16. (Sublocade 2017,
Scenesse 2019, Perseris 2018, Abilify Maintena
6.6.2 Coarse Suspensions 2016, Zyprexa Relprevv 2009, Cabenuva 2021,
Triptodur 2017).
An injectable suspension is a dispersed system Biodegradable polymeric microspheres repre-
that comprises an insoluble solid material sent another type of coarse dispersion that can
suspended in a liquid vehicle (usually water for provide sustained and controlled drug release.
injection or oily vehicles). Suspensions are Polymeric biodegradable microspheres are
Table 6.15 USP<729> Methods for lipid droplet size determination

Method 1 Method 2
Technique Dynamic light scattering (DLS). Light obscuration (LO).
Laser diffraction (LD). Light extinction (LE).
Standard Concentrated suspension, containing National Concentrated suspension, containing National
Institute of Standards and Technology (NIST)- Institute of Standards and Technology (NIST)-
traceable polystyrene latex standard particles or traceable polystyrene latex standard particles or
other suitable nanospheres diluted in water. other suitable nanospheres diluted in water.
Tested at three different standard sizes – 100, 250, Tested at 5 μm and 10 μm in triplicate).
400 nm (in triplicate).
Sample prep Test sample diluted in water to an appropriate Test sample diluted in water to an appropriate
concentration for detection method. concentration for detection method.
Testing should be done in triplicate or, as Testing should be done in triplicate or, as
appropriate, based on method validation data. appropriate, based on method validation data.
Test method Set at ambient temperature. Threshold of detection ¼ 1.8 μm.
Set scattering angle to 90 . Upper limit ¼ 50 μm.
Vary the concentration and/or data collection
times such that there is at least a factor of two in
the difference of the total number of globules that
measure >5 μm between at least two sample runs.
Data Chi-square (χ2) goodness-of-fit parameter must The volume-weighted, large-diameter fat globule
interpretation remain acceptably low (per instrument limits of the dispersed phase, expressed as the
specifications). percentage of fat residing in globules larger than
The intensity-weighted mean droplet diameter 5 μm (PFAT5) for a given lipid injectable
(MDD) for lipid injectable emulsions must be less emulsion, must not exceed 0.05%.
than 500 nm or 0.5 μm, irrespective of the
concentration of the dispersed lipid phase.
Table 6.16 Marketed long-acting injectable formulations of poorly soluble drugs

Route of
Product Formulation technique Duration administration
SUBLOCADE™ (buprenorphine ATRIGEL® delivery system (50:50 poly 1 month Subcutaneous
extended-release) injection (DL-lactide-co-glycolide) polymer and a
biocompatible solvent, N-methyl-2-
pyrrolidone (NMP))
SCENESSE® (afamelanotide) implant Poly (DL-lactide-co-glycolide) implant 2 months Subcutaneous
PERSERIS™ (risperidone) for extended- Poly (DL-lactide-co-glycolide) polymer 1 month Subcutaneous
release injectable suspension and Nmethyl-2-pyrrolidone
Provided as a two-syringe mixing system
ABILIFY MAINTENA® (aripiprazole) for Carboxymethyl cellulose sodium 1 month Intramuscular
extended-release injectable suspension Provided as a two-syringe mixing system
ZYPREXA RELPREVV (olanzapine) for Carboxymethylcellulose sodium, 1 month Intramuscular
extended release injectable suspension polysorbate 80
CABENUVA (cabotegravir extended- Supplied as two separate vials for 1 month Intramuscular
release injectable suspension; rilpivirine simultaneous injection at separate sites
extended-release injectable suspension) Cabotegravir—Polyethylene glycol (PEG)
3350, polysorbate 20
Rilpivirine—Poloxamer 338
TRIPTODUR (triptorelin) for extended- Microgranule formulation (poly-d,l- 24 weeks Intramuscular
release injectable suspension lactide-co-glycolide)
monolithic microparticles where the drug is dis- structure that characterizes microcapsules. Drug
persed homogeneously within the bulk of the release is usually a result of combined diffusion
particle, without having a distinct core-shell and erosion of the polymeric matrix. This is a
Table 6.17 Microsphere preparation techniques

Method References
Solvent evaporation McGinity and O’Donnell (1997)
Spontaneous emulsification/solvent diffusion Horisawa et al. (2002)
Salting-out/emulsification diffusion Chaisri et al. (2009)
Supercritical fluid technology Duarte et al. (2006)
Electrospray technology Malik et al. (2016)
Table 6.18 Poorly soluble compounds marketed in liposomal formulations

Available products Formulation
AMBISOME® (amphotericin B) Hydrogenated soy phosphatidylcholine, cholesterol,
liposome for injection Distearoylphosphatidylglycerol, alpha tocopherol, sucrose, disodium succinate
hexahydrate
MEPACT® (mifamurtide)a 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC)
1,2-Dioleoyl-sn-glycero-3-phospho-L-serine monosodium salt (OOPS)
VISUDYNE® (verteporfin for Lactose, egg phosphatidylglycerol, dimyristoyl phosphatidylcholine, ascorbyl
injection) palmitate and butylated hydroxytoluene
a
Approved by EMA only, not FDA
major difference between polymeric (LUVs), large multilamellar vesicles (MLVs), or

microspheres and micronized drug dispersions, multivesicular vesicles (MVVs). Liposomes, as
where the drug particles form a depot at the injec- well as other nanocarriers, may exhibit
tion site, and drug release is mainly dependent on completely different pharmacokinetic and
dissolution in the physiological fluid. Drug disso- biodistribution profiles as compared to free drug
lution may also impact drug release from poly- following IV injection, especially when
meric microspheres in instances where the drug is formulated with surface modification, such as
suspended (rather than dissolved) in the polymer polyethylene glycol (PEGylated or stealth
solution during preparation (e.g. solid-in-oil-in- liposomes). PEGylation helps liposomes, and
water (S/O/W) emulsion-based method). other injectable carrier systems, avoid being pre-
Methods of microsphere preparation are listed in maturely taken up by cells of the mononuclear
Table 6.17. phagocyte system (MPS) and achieve
prolonged circulation time. In addition,
liposomes and other nanocarriers may be able
6.6.3 Liposomes and Other Colloidal to extravasate through the leaky ill-formed
Systems tumor vasculature and thus accumulate at the
tumor site in a phenomenon called enhanced
Liposomes permeation and retention (EPR) (Maeda et al.
Liposomes are carrier systems that are suitable for 2000, Fang et al. 2011, Lammers et al. 2012).
delivery of both lipophilic and hydrophilic Due to their unique structure, liposomes can
molecules. These vesicles are assembled as an incorporate water-soluble drugs inside the
aqueous core enclosed by amphiphilic lipids, aqueous core, or water-insoluble drugs within
which are usually bilayer-forming phospholipids. the bilayer. Although the majority of commer-
Commonly used phospholipids include phospha- cially marketed liposomal products to date are
tidylcholine, phosphatidylethanolamine, comprised of hydrophilic drugs, liposomes may
phosphatidylserine, phosphatidic acid, phosphati- still be utilized in the formulation of poorly
dylinositol, and phosphatidylglycerol cardiolipin. soluble compounds, such as amphotericin B
Depending on the number of lamellae and size, and verteporfin, as detailed in
liposomes are categorized as small unilamellar Table 6.18 (Bulbake et al. 2017, Ambisome
vesicles (SUVs), large unilamellar vesicles 2012, Mepact 2009, Visudyne 2013). Although
not true liposomes, two other formulations of with weakly acidic or basic (ionizable) terminal
amphotericin B, ABELCET® and groups. As described in Sect. 6.5.5, the poorly
®
AMPHOTEC , also utilize phospholipids to soluble drugs can be complexed with
generate formulations of ribbon-like bilayered cyclodextrins, which could then be actively
membranes and colloidal dispersions of disc- loaded into the liposomes (Aloisio 2017). Inter-
like particles, respectively (Płaczek et al. 2019). estingly, this method successfully loaded both
Drug loading into liposomes can either be hydrophobic and hydrophilic drugs with high
passive or active. The passive loading method is efficiency.
only used for water-soluble drugs, and the load- A variety of manufacturing techniques exist
ing efficiency is usually low. In contrast, active for the preparation of liposomes, including thin-
(remote) loading can result in significantly higher film hydration, reversed-phase evaporation, sol-
loading efficiencies. The method relies on creat- vent injection, and detergent dialysis. While these
ing a transmembrane pH gradient in which the methods are often successful, they can pose
drug of interest (or a part of it) is present outside challenges related to particle size, sterility, and
the liposomal membrane in nonionized form. safety. These methods may result in larger-than-
This encourages the transport of drug across the desired liposomes or formulations with broad
membrane to the inside of the liposomal core, in particle size distributions. Because many of
which it becomes ionized (e.g. an amine group these traditional methods require the use of
becomes protonated) due to the pH difference, organic solvents, concerns may arise regarding
and will consequently be trapped inside, as an residual solvent content and subsequent safety
ionized species can no longer pass across the upon administration. Likewise, many lipids used
lipid bilayer. However, as is the case with many to formulate liposomes are sensitive to high
poorly soluble drugs, there are no ionizable temperatures, which may make sterilization a
groups which could take advantage of this active challenge. Lastly, these methods may not provide
loading method (Eloy et al. 2014). An alternative adequate reproducibility. As such, many newer
method, developed by Sur et al. (2014), methods have been developed to overcomes
incorporates chemically-modified cyclodextrins these challenges and are described in Table 6.19.
Table 6.19 Newer techniques for liposome preparation

Method Advantages References
Supercritical antisolvent method No organic solvent required Lesoin et al. (2011)
Supercritical phase-evaporation Improved entrapment efficiency Otake et al. (2001) and Otake et al.
method No organic solvent required (2006)
Dual asymmetric centrifugation Produces small particle size Massing et al. (2008)
High entrapment efficiency of water-
soluble drugs
No organic solvent required
Membrane contractor technology High entrapment efficiency of Laouini et al. (2011) and Jaafar-Maalej
lipophilic drugs et al. (2011)
Produces small particle size
Easily scaled-up
Crossflow filtration detergent Production of homogenous, stable Peschka et al. (1998)
depletion method liposomes
Shorter production time
Production of sterile products
Freeze-drying double emulsion Efficient encapsulation Van Winden et al. (1997) and Wang
method Produces small particles with good et al. (2006)
stability
Reproducible
Production of sterile products
Table adapted from Huang et al. (2014)
Nanosuspensions loading can be increased such that the volume of

To achieve high drug loading for compounds injection is greatly reduced. Additionally, poten-
with poor solubility in water, lipids, and/or tially toxic agents, such as surfactants, may be
organic solvents, nanosuspensions may be con- avoided and replaced with lower levels of more
sidered. By formulating an injectable as a tolerate agents, including phospholipids or albu-
nanosuspension, drug loading per unit volume min. Due to reduced particle size, the
can be substantially increased, reducing the over- syringeability of nanosuspensions is often
all volume of injection. Nanosuspensions are improved over coarse dispersions of equivalent
stabilized dispersions of submicron drug drug load. However, as with coarse suspensions,
particles, which can be produced by both challenges such as foreign matter contamination
top-down and bottom-up preparation methods and sterilization also apply to nanosuspension
(Table 6.20 and Fig. 6.1). formulations (Anton et al. 2016). Table 6.21
There are several advantages to formulating a lists currently marketed nanosuspensions of
poorly soluble drug as a nanosuspension. Since poorly soluble drugs (Abraxane 2015, Aristada
the drug is not required to be in solution, drug 2015, Aristada Initio 2020, Invega Sustenna
Table 6.20 Types of preparation methods for nanosuspensions

Method Approach Examples
Milling Top-down Colloid milling, jet-milling, Nanomill®, NanoCrystal®
High-pressure homogenization Top-down Microfluidization, Dissocubes®, Nanopure®
Precipitation Bottom-up Hydrosol, Nanomorph®, NanoCrySP
Combination techniques Bottom-up Nanopure XP, Nanoedge™,SmartCrystal®
Adapted from Sheokand et al. (2018)
Fig. 6.1 Preparation Methods for Nanosuspension (from Patel and Agrawal 2011)
Table 6.21 Marketed nanosuspensions of poorly soluble drugs

Route of
Product Formulation administration
ABRAXANE® for injectable suspension Albumin–paclitaxel conjugate Intravenous
ARISTADA™ (aripiprazole lauroxil) extended-release Nanocrystal (Alkermes Intramuscular
injectable suspension NanoCrystal® technology)
ARISTADA INITIO™ (aripiprazole lauroxil) extended-release Nanocrystal (Alkermes Intramuscular
INVEGA SUSTENNA® (paliperidone palmitate) extended- Nanocrystal (Alkermes Intramuscular
release injectable suspension NanoCrystal® technology)
INVEGA TRINZA® (paliperidone palmitate) extended-release Nanocrystal (Alkermes Intramuscular
RYANODEX® (dantrolene sodium) for injectable suspension Nanocrystal Intravenous
2015, Invega Trinza 2015, Ryanodex 2014, Choi 2012 (Shetab Boushehri et al. 2020). While sev-
and Han 2018, Shetab Boushehri et al. 2020, eral products have been approved in recent years
Ventola 2017). that are solid lipid nanoparticle formulations, par-
ticularly for the delivery of RNA-based therapies,
Nanoparticles including the recent mRNA-based COVID-19
Nanoparticles may also provide an alternative vaccines from Moderna and Pfizer-BioNTech,
formulation approach for poorly soluble drugs, none exist to date of poorly water soluble drugs;
specifically in cases where alteration of the phar- however, clinical trials evaluating the use of lipid
macokinetics may be desired, or tissue targeting nanoparticles for delivery of poorly soluble drugs
is an advantage. Generally, nanoparticles are for oncology and infectious disease are ongoing
defined as dispersed, solid, supramolecular (e.g. NCT04148833, NCT04616872,
structures smaller than 500 nm, typically com- NCT04610567). Many methods exist to prepare
posed of polymers and/or lipids. Nanoparticles nanoparticles, which are outlined in
may be either developed as a matrix particles or Table 6.22 (Crucho and Barros 2017, Duarte
reservoir particles, depending on the method of et al. 2006, Mishra et al. 2018). However,
preparation. Because of the EPR effect, challenges remain regarding efficiency of drug
nanoparticles can be utilized for passive targeting loading and rapid premature release (known as
of tumor tissue. burst release), both of which may contribute to
Though nanoparticles comprise a broad and potential toxicity concerns (Anselmo and
diverse array of formulation approaches, poly- Mitragotri 2019).
meric nanoparticles and solid lipid nanoparticles
are most common (Teixeira et al. 2017). For Micelles
example, ELIGARD® (leuprolide acetate) sus- Micelles are nanoaggregates composed of self-
pension for subcutaneous injection is a marketed assembled amphiphilic molecules that exist in an
polymeric nanoparticle product that utilizes the aqueous solution above their CMC. As mentioned
ATRIGEL® Delivery System, which is a poly- in Sect. 6.5.4, they can efficiently solubilize
meric (nongelatin containing) delivery system poorly soluble drugs by entrapping them within
consisting of a biodegradable poly (DL-lactide- their hydrophobic core, while their hydrophilic
co-glycolide) (PLGH or PLG) polymer formula- tails are directed outward. Polymers may be
tion dissolved in a biocompatible solvent, added, in the formulation, along with surfactants,
N-methyl-2-pyrrolidone (NMP) (Eligard 2015). to provide additional stability of the micelles (see
Another example, OPAXIO® from CTI Table 6.23 for preparation methods). As a deliv-
BioPharm, contains polyglutamic acid- ery system, micelles can extend circulation time
conjugated (poliglumex) paclitaxel and was and provide passive targeting into tumor tissue.
granted Orphan Drug Status by the FDA in GENEXOL®, for example, is a polymeric
Table 6.22 Methods of preparation for nanoparticles

Method Application
High shear homogenization Solid lipid nanoparticles
Ultrasonication/high-speed homogenization Solid lipid nanoparticles
Hot homogenization Solid lipid nanoparticles
Cold homogenization Solid lipid nanoparticles
Microemulsion Solid lipid nanoparticles
Supercritical fluid Solid lipid nanoparticles, polymeric nanoparticles
Solvent emulsification/evaporation Solid lipid nanoparticles, polymeric nanoparticles
Double emulsion Solid lipid nanoparticles, polymeric nanoparticles
Spray drying Solid lipid nanoparticles, polymeric nanoparticles
Emulsification-solvent diffusion Polymeric nanoparticles
Emulsification reverse salting-out Polymeric nanoparticles
Nanoprecipitation Polymeric nanoparticles
Dialysis Polymeric nanoparticles
Table 6.23 Drug-loaded methods for preparation of polymeric micelles

Method Application
Chemical Water-insoluble drugs may be covalently linked to the hydrophobic core of the copolymer.
conjugation
Dialysis Water-insoluble drug and polymer are dissolved in solvent, then dialyzed against a solvent
selective for the hydrophilic portion of the copolymer.
Oil-in-water Water-insoluble drug in volatile solvent is added to an aqueous solution of copolymer; micelles
emulsion form as solvent evaporates.
micellar formulation of paclitaxel comprised of a formulations may be suitably injectable and

PEG and poly (lactic acid) copolymer syringeable (McKenzie et al. 2015). Figure 6.2
(PEG-PLA) indicated for treatment of various highlights some of the mechanisms by which
cancers (Genexol 2015). gelation occurs in hydrogels.
Though promising, drug delivery using
hydrogels is not without challenges, especially
regarding poorly soluble drugs. Because of their
6.6.4 Hydrogels and In-Situ Implants
hydrophilic nature, efficient encapsulation,
homogeneity within the hydrogel matrix, and
Hydrogels are reversible, three-dimensional
sustained release are not always attainable using
networks of cross-linked polymers that swell in
conventional hydrogel polymers and preparation
the presence of water or biological fluid. Because
methods. However, approaches to overcoming
of their ability to absorb water, they simulate
these challenges can be found in literature. In a
living tissue and have favorable biocompatibility
study published by Ju et al. (Ju et al. 2013),
and biofunctionality. Though commonly
N-octyl-O-sulfate chitosan (NOSC) micelles
implemented in tissue engineering approaches as
were used to enhance the solubility of paclitaxel.
a biomaterial scaffold, hydrogels are also applica-
These micelles were then dispersed in
ble in drug delivery, particularly for injection.
carboxymethyl chitosan (CMCS)-conjugated
Hydrogels are insoluble in water upon gelation
poloxamer P407 gels and mixed with glutaralde-
(i.e. cross-linking) and as such may be utilized as
hyde to form cross-links with amine groups on
a sustained-release vehicle. Furthermore, because
CMCS, resulting in a gel depot upon intratumoral
increases in viscosity of hydrogels only occur
injection. Similarly, another study by Gat et al.
upon gelation (either through chemically, physi-
used PLGA-PEG-PLGA micelles to incorporate
cally, or environmentally induced stimuli),
Fig. 6.2 Mechanisms of gelation in hydrogels (from Lee 2018)
docetaxel inside a hydrogel matrix. The In another example, published by Benhabbour

docetaxel-loaded PLGA-PEG-PLGA micelles et al. (2019), two antiretroviral drugs, MK-2048
also formed a gel depot at the injection site. and dolutegravir, were evaluated as potential
In-situ forming implants (ISFIs) may also be long-acting INFIs for HIV treatment.
an attractive alternative for long-acting injectable Formulations were prepared by dissolving
drug delivery. ISFIs utilize a water-miscible sol- PLGA (50:50 LA/GA, 27 kDa) in N-methyl-2-
vent to dissolve both drug and biodegradable pyrrolidone (NMP) at varying weight ratios. The
polymer. Upon subcutaneous or intramuscular resulting formulations demonstrated both physi-
injection, the solvent diffuses into the cal and chemical stability upon testing, indicating
surrounding aqueous environment and, by its no interaction of drug with polymer. Release of
water-miscible nature, allows penetration of drug from the ISFIs was found to be related to the
water into the polymer solution causing phase weight ratio of PLGA to NMP, with higher
inversion and the formation of solid implant of amounts of NMP resulting in greater burst release
biodegradable polymer and drug. Because of this, and quicker overall release kinetics. The ISFIs
ISFIs are inherently compatible with poorly solu- also maintained zero order release, with 100%
ble drugs and provide advantages over traditional release achieved by day 208 for MK-2048 at by
implant manufacturing and insertion 1:2 (PLGA to NMP) weight ratio formulation and
methods (Calo and Khutoryanskiy 2015). 100% release achieved by day 120 for DTG at a
BEPO®, a technology from MedinCell, is 1:8 weight ratio, while the 1:2 weight ratio DTG
composed of diblock and triblock polyethylene formulation had 34% release at day 120.
glycol-polyester copolymers which form a matrix
depot via solvent exchange upon injection. In a
publication by Roberge et al. (2020), in-situ
6.7 Strategies for Improving
forming depots prepared with this technology
Stability
were able to provide up to 50 days and 100 days
of sustained release for bupivacaine and ivermec-
Once the drug is solubilized to the required con-
tin, respectively.
centration by any of the aforementioned
techniques, the physicochemical stability of the
Table 6.24 Hydrogel-based commercially marketed products

Route of
Product Formulation administration
SUPPRELIN® LA (histrelin MedLaunch™ polymer reservoir (hydrophilic cartridge composed Subcutaneous
acetate) of 2-hydroxyethyl methacrylate, 2-hydroxypropyl methacrylate,
trimethylolpropane trimethacrylate, benzoin methyl ether,
Perkadox-16, and Triton X-100)
DEXTENZA® 4-arm polyethylene glycol (PEG) N-hydroxysuccinimidyl glutarate Intracanalicular
(dexamethasone ophthalmic (20 K), trilysine acetate, N-hydroxysuccinimide-fluorescein, sodium
insert) phosphate dibasic, sodium phosphate monobasic, water for injection
formulation must be assessed. If stability is inad- oxidation-catalyzing impurities may be

equate, several techniques may be used to introduced by other excipients, making the addi-
improve it, including the addition of stabilizing tion of antioxidant(s) or chelating agent(s) benefi-
excipients and adjustment in processing cial (Chen 2015). A list of antioxidants used in
techniques. The following sections will discuss parenteral formulations and their typical
how to use these techniques to improve the concentrations is provided in Table 6.24 (Suprelin
stability of compounds prone to hydrolysis 2011, Dextenza 2019) Antioxidants and chelators
and oxidation, two of the major chemical prevent oxidative degradation through different
degradation pathways for small-molecule mechanisms. A true antioxidant, such as sodium
drugs (Vemuri 2010). bisulfite, reacts with free radicals to terminate the
oxidative chain reaction; a reducing agent
(e.g. citric acid) is preferentially oxidized and
6.7.1 Addition of Stabilizing reduces the level of oxygen or oxidants in the
Excipients formulation; a chelating agent, such as ethylene-
diaminetetraacetic acid (EDTA), complexes with
Although the inclusion of a buffer system in a trace metals in the drug product, thereby blocking
formulation may be beneficial in maintaining the the oxidative degradation pathway. In some
appropriate pH of the formulation, the concentra- cases, the addition of metal chelator may not be
tion and type of buffer must be carefully screened efficient to eliminate trace metals. For example,
and selected such that the buffer does not nega- the foreign metal contaminant may be distributed
tively impact the solubility or stability of the between the chelator and the drug, depending on
active ingredient. The formulation should be the respective binding constant of each (Hovorka
assessed to ensure that precipitation of the poorly and Schoneich 2001). In other cases, the metal–
soluble drug does not occur in the buffer. Stability chelator complex intermediate may not be redox-
must also be evaluated to confirm that the buffer inert. Similarly, selecting an antioxidant should
does not reduce the stability by acting as a cata- consider the rate constant of the reaction of the
lyst for degradation reactions. suspected oxidizing species with the drug sub-
For drugs that are likely to undergo oxidative stance and with the antioxidant. Sometimes, a
degradation that is catalyzed by metal ions, combinational approach is utilized to increase
hydrogen, or hydroxyl ions, their stability can be the overall antioxidant effectiveness. As with
enhanced by displacing oxygen from the vial with any excipient, it is important to assess the regu-
an inert gas (e.g. nitrogen) and/or by adding latory acceptability of the antioxidant/chelating
antioxidants or chelating agents to the formula- agent. Also, depending on the properties of the
tion (Waterman et al. 2002, Chen 2015). While drug molecule, the antioxidant may either
oxygen displacement alone can provide some improve or worsen the solubility of the drug in
deterrence to oxidation, trace amounts of the formulation (Table 6.25).
Table 6.25 Commonly used antioxidants in injectables

Antioxidant Typical concentration (%, w/v)
Ascorbic acid 0.01
Butylated hydroxyanisole (BHA) 0.02
Butylated hydroxytoluene (BHT) 0.02
Cysteine 0.5
Monothioglycerol 0.5
Sodium bisulfite 0.15
Sodium metabisulfite 0.2
Glutathione 0.1
6.7.2 Adjustment in Processing frozen formulation through sublimation, resulting

Techniques in a dry powder (or cake) that can be easily
reconstituted for administration. During the pro-
An improvement in the stability of a formulation cess of lyophilization, the drug solution typically
may also be achieved using specific processes transitions into a porous, amorphous cake.
that reduce or eliminate potential causes of degra- Because the lyophilized powder/cake lacks a
dation or instability. As discussed previously, rigid crystalline structure, it has the highest ther-
oxidation may be an issue in the stability of modynamic solubility and is more easily
many injectables. In these cases, the removal of dissolved, which can improve the reconsti-
oxygen from the primary container closure can tutability of the product.
increase the stability of the drug product. Sparg- As with other processing methods, lyophiliza-
ing (a process during which a chemically inert gas tion must be carried out under aseptic conditions.
is bubbled through the solution prior to packag- In this process, a solution of the formulation and
ing) and blanketing the vials with nitrogen are possibly a bulking agent (an excipient used to
two techniques that may be used to remove oxy- provide an elegant, noncollapsed cake) are filled
gen from the formulation and achieve the desired into vials, which are loaded into a lyophilizer
shelf-life of the product. containing temperature-controlled shelves. The
Hydrolytic degradation, another common solution is frozen, and the pressure is decreased
mechanism of drug-product instability, may also to facilitate the primary drying phase, where
be addressed through processes that remove water water can sublimate. The temperature may then
from the formulation. This can be accomplished be increased for secondary drying, which
by removing water from the formulation until removes residual water from the cake. Lyophili-
administration, or by forming a solid dispersion zation requires formulation-specific optimization
of insoluble drug to isolate water. Filling of dry and may be costly as compared with other
powder into vials allows for water to be added to methods of processing; however, it has been
the product only at the time of reconstitution just extensively characterized and utilized at the
prior to use; however, the powder must be sterile industrial level.
and aseptically packaged, which may prove diffi- Depending on the route of parenteral adminis-
cult. Additionally, powder flow properties and tration, some drugs susceptible to hydrolysis in
particle size can contribute to difficulties in filling solution may be formulated into a suspension. By
and constituting the product. Sterile filtered using the least soluble form of the active com-
solutions may be used to crystallize or spray dry pound, the drug may be isolated from the
the product to reduce contamination and maintain surrounding vehicle to improve stability. These
sterility. suspensions may be formulated as ready-to-use
Lyophilization (or freeze drying) is a more products where water has already been added, or
robust method of removing water from drug prod- as powders-for-reconstitution with a sterile dilu-
uct. In this process, water is removed from the ent. As discussed earlier in the chapter,
suspensions may be prone to physical • UV-VIS Spectrophotometer.

instabilities, and these issues must be addressed • Centrifuge.
as well; however, the chemical stability of the • Magnetic stirrer.
active ingredient may be significantly improved
using this method. Method
• Prepare standard drug solutions in a range
from 10 to 100 μM:
6.8 Injectable Product – For gliclazide, glyburide, glipizide, and
Development Workflow glimepiride, solubilize drug in
0.1 M NaOH.
As detailed by the information presented in this – For pioglitazone and rosiglitazone, solu-
chapter, there are many considerations that must bilize drug in 0.1 M HCl.
be made in the development of an injectable • Using an ultraviolet absorption spectrome-
product containing a poorly water-soluble com- try, determine the extinction coefficient of
pound. Properties of the drug and end-product drug in the relevant solvent at wavelength
objectives must be determined and will dictate of maximum absorption:
the approaches taken in the formulation develop- – Gliclazide and glyburide—226 nm,
ment process. Figure 6.3 is an example flowchart glipizide—275 nm, glimepiride—
that depicts the variables that must be considered 228 nm, pioglitazone—269 nm,
and evaluated to develop a safe and successful rosiglitazone—317 nm.
parenteral product. • Add excess drug to relevant solvent in con-
ical flask. Seal and stir on magnetic stirrer
Method Capsule 1 for 24 hours at 25 C.
• Centrifuge sample, remove supernatant,
Micellar Solubilization of a Poorly Soluble and filter through 0.45-μm filter.
Antidiabetic Drugs • Dilute supernatant in appropriate solvent
Based on a method reported by Seedher and and read using UV-VIS spectrophotometer.
Kanojia (2008). • Calculate micellar concentration by C micel-
Objective
lar ¼ C surf CMC.
• To examine the effect of micellar solubili- Results
zation using both nonionic and ionic • Solubility for all drugs is increased in the
surfactants on poorly soluble antidiabetic presence of 50 mM surfactant.
drugs—gliclazide, glyburide, glipizide, • Gliclazide, glyburide, and glimepiride have
glimepiride, pioglitazone, and increased solubility in nonionic surfactant
rosiglitazone. (Tween 80), whereas pioglitazone,
Equipment and Reagents rosiglitazone, and glipizide show increased
• Antidiabetic compounds (listed above). solubility in ionic surfactants
• Tween 80. (SDS, CTAB).
• Sodium dodecyl sulfate (SDS).
• Cetrimonium bromide (CTAB). Method Capsule 2
• Glass conical flasks (5 mL minimum
volume). Solubilization of Albendazole by
• Double distilled water. Cyclodextrins in Aqueous and Acetic Acid
• 0.1 M NaOH Solutions
• 0.1 M HCl Based on a method reported by Moriwaki et al.
• 0.45-μm filter (2008)
Fig. 6.3 Injectable product development flowchart

Objective Objective
• To evaluate complexation of albendazole • To prepare microspheres of norquetiapine
(ABZ) and β-cyclodextrin (BCD) at differ- freebase for long-acting injection.
ent molar ratios. Equipment and Reagents
Equipment and Reagents • Norquetiapine (NQ) freebase (prepared
• ABZ. according to a method described in
• BCD. reference).
• Phenolphthalein (PHE). • Poly(d,l-lactic-co-glycolic acid) (PLGA)
• Acetic acid. 502, 502H, 503, and 503H.
• 0.45-μm filter • Polyvinyl alcohol 500 (PVA 500).
• UV-VIS Spectrophotometer. • Methanol.
• Centrifuge. • Dichloromethane.
• Magnetic stirrer. • Distilled water.
Method • 0.45-μm filter
• Dissolve ABZ standards 5, 7.5, 10, 15, and • High-speed dispersion homogenizer
20 mg per 3 mL of concentrated acetic acid (S18N-19G).
and add water to a final volume of 50 mL. • Laboratory lyophilizer.
Measure absorbance values at 295 nm • Magnetic stir plate.
against solvent blank. • Centrifuge.
• Prepare ABZ and BCD solutions at the Method
following molar ratios—1:1, 1:2, 1:4, 1:6, • Dissolve 300 mg each of the PLGA
1:8, and 1:10. polymers and 60 mg of NQ freebase in
• Agitate test samples at 175 rpm for 3 days 15 mL of a 9:1 (v/v) dichloromethane and
and 37 C. methanol solution.
• After agitation, allow suspensions to cool to • Inject the PLGA-NQ solution into 300 mL
room temperature and filter the supernatant of a 0.2% (v/v) PVA 500 aqueous solution
through a 0.45-μm filter. using a high-speed dispersion homogenizer
• Dilute filtered supernatant with water and a at 1 mL/s. The homogenizer should be set at
solution of PHE. 2800 rpm. Homogenize for 1 minute to
• Characterize samples by UV-VIS spectros- form an oil-in-water emulsion.
copy and determine BCD concentration • Transfer the emulsion to a magnetic stir
based on absorbance of PHE at 550 nm plate and stir for 3 hours at 24–27 C for
(absorption should be minimal for dilution and evaporation of the solvent.
complexed PHE/BCD at this wavelength • Centrifuge resulting suspension at
compared to free PHE). 1500 rpm for 3 minutes and rinse three
Results times with distilled water.
• Spectrophotometric analyses of the • Filter suspension through a 0.45-μm filter
preparations of ABZ:BCD complex and dilute with 5 mL distilled water.
indicated the formation of a complex, and • Lyophilize for two days at 75 C.
the previous solubilization of ABZ with Results
acetic acid allowed a more concentrated • Use of the NQ freebase increases encapsu-
complex solution to be prepared. lation efficiency dramatically from practi-
cally zero for the hydrochloride salt to
Method Capsule 3 30.33 0.51% (502), 46.19 0.33%
(502H), 32.69 1.28% (503), and
Preparation of Poly (Lactide-co-Glycolide)
48.30 3.04% (503H).
Microspheres Containing Norquetiapine
Based on a method reported by Park et al.
(2018)
• All formulations showed sustained release nitrogen flushing the liquid surface at
of NQ (following an initial burst release) 30 C.
for 20 days. • Centrifuge at 600 g for 5 minutes to remove
free breviscapine from the MVLs and
Method Capsule 4 resuspend in a buffered saline solution.
Results
Preparation of a Multivesicular Liposome For-
• Breviscapine MVLs produced by this
mulation for Sustained IM Delivery
method significantly prolong the release of
Based on the method reported by Zhong et al.
drug both in vitro (5–6 days) and in vivo
(2005)
(IM injection in rats lasted 4–5 days) com-
Objective
pared to other the liposome preparation
• To design a multivesicular liposome
techniques investigated.
(MVL) sustained delivery formulation of
breviscapine for IM injection.
Method Capsule 5
• Breviscapine (bioactive ingredient). Preparation of Supramolecular Hydrogels
• Phosphatidylcholine. Based on Poly (Ethylene Glycol)-Poly (Lac-
• Phosphatidylglycerol. tic Acid) Block Copolymer Micelles and
• Triolein or tricaprylin. α-Cyclodextrin
• Cholesterol. Based on the method reported by Poudel et al.
• Sucrose and glucose. (2018)
• Solutions and buffers: 50 mM arginine- Objective
containing buffer (pH 7), and 40 mM L- • To prepare hydrogels of hydrophobic drugs
Lysine. for long-acting injection.
• Chloroform–diethyl ether (1:1, v/v). Equipment and Reagents
• T 18 basic Ultra-Turrax mixer. • Poly (ethylene glycol)-block-poly (lactic
Method (Use the Double-Emulsion Process to acid) (PEG-b-PLA) copolymer (see Xiao
Produce Breviscapine MVLs) et al. 2017).
• Prepare a lipid mixture in 1 mL • α-cyclodextrin (α-CD),
chloroform–diethyl ether with 40 mg phos- • Tetrahydrofuran (THF).
phatidylcholine, 8 mg phosphatidyl- • Deionized water.
glycerol, 40 mg cholesterol, and triolein or • Magnetic stir plate.
tricaprylin at a 5.75:1 molar ratio of phos- • Dialysis bags.
phatidylcholine to the total triglyceride • Vortex mixer.
content. • Ultrasonicator.
• Prepare an aqueous solution containing Method
40 mg/mL of breviscapine, 4% (w/v) • Prepare micelles using a cosolvent method.
sucrose in 50 mM arginine buffer (pH 7). Dissolve 100 mg of PEG-b-PLA copolymer
• Emulsify the lipid and aqueous solutions to in 2 mL of THF.
make a W/O emulsion at 10,000 rpm for • Add deionized water slowly (1 mL/h) while
8 minutes with the mixer. vigorously stirring to a final water content
• Prepare the second aqueous solution with of 50% wt.
40 mM L-Lysine and 3.4% (w/v) glucose • Quench aggregates by adding micellar sus-
and emulsify with the W/O emulsion to pension to 6 mL of deionized water.
form a W/O/W emulsion. Transfer the • Dialyze for 3 days against deionized water
emulsion to an Erlenmeyer flask and to remove any residual solvent.
remove the chloroform–diethyl ether by
• Once dialyzed, add α-CD to 10 mg/mL of • YMC-pack ODS-AM-302 (150 mm x

PEG-b-PLA suspension at concentrations 4.6 mm) HPLC column.
of 7%, 8%, 9%, 10%, and 11% (w/v). • Mobile phase of 0.2% acetic acid/acetoni-
• Homogenize using a vortex mixer for trile in a ratio of 65:35 (flow rate 1 mL/min,
1 minute at 150 rpm followed by UV detection wavelength 240 nm, injection
ultrasonication for 10 minutes. volume 15 μL).
Results Method
• Resultant PEG-b-PLA micelles have spher- • Prepare surfactant solutions by dissolving
ical morphology with a diameter ranging an adequate amount of surfactant (Tween
from 40 to 180 nm. 80 or SDS) in water or cosolvent solution.
• The initial gelation time for the hydrogel for • Add an excess amount of PHT to a glass
7% (w/v) α-CD was 92.6 minutes, but, as vial and introduce 4 mL of surfactant-in-
the concentration of the α-CD was water or surfactant-in-cosolvent solution.
increased to 8%, 9%, 10%, and 11% • Rotate vials at 50 rpm at 25 C for 24 hours.
(w/v), the gelation time decreased to 63.3, • Filter samples using a 0.45-μm syringe
38, 13, and 6.3 minutes, respectively. filter.
• Viscosity of the hydrogels can be modified • Analyze PHT content of samples using
by altering the amount of α-CD, with lower HPLC to determine solubility.
amounts producing lower viscosities. Results
• Hydrogels can self-heal following shear • The addition of up to 20% (w/v) cosolvent
stress (12 minutes for 11% w/v increases the solubility of PHT in a
preparation). log-linear fashion, with the exception of
• α-CD/micellar hydrogels can effectively glycerol.
load a hydrophobic drug and provide • The solubilization capacity for DMA, etha-
sustained release over 168 hours. nol, and PEG is 11.9 (w/w), 7.2, and 7.2,
respectively.
Method Capsule 6 • The pH values for all cosolvent solutions
are between 3.0 and 6.5, well below the
Solubilization of a Poorly Soluble Drug Using
pKa value of PHT (8.3); thus,
Cosolvents
pH-dependent solubility is not a
Based on a method reported by Kawakami
contributing factor.
et al. (2006).
Objective
Method Capsule 7
• To examine the solubility of a poorly water-
soluble drug, phenytoin (PHT), in the pres- Solubilization of a Novel Poorly Soluble Drug
ence of the cosolvents dimethylacetamide Using Hydroxypropyl-β-Cyclodextrin
(DMA), ethanol, polyethylene glycol Based on a method reported by Kim et al.
400 (PEG400), and glycerol. (2004).
Equipment and Reagents Objective
• PHT. • To improve the solubility of a poorly water-
• DMA. soluble novel compound, CKD-732,
• Ethanol. through the addition of hydroxypro-
• PEG400. pyl-β-cyclodextrin (HP-β-CD) at various
• Glycerol. concentrations.
• Glass vials. Equipment and Reagents
• Mechanical rotator. • CKD-732.
• 0.45-μm syringe filter • HP-β-CD.
• Glass vials.
• Orbital shaker. Equipment and Reagents

• Centrifuge. • Nimodipine (compound for the treatment of
• 0.2-μm syringe filter subarachnoid hemorrhage-related
• Kromasil® C18 (250 mm x 4.6 mm, 5 μm) vasospasm).
HPLC column. • Sodium deoxycholate.
• Mobile phase of acetonitrile/20 mM ammo- • Poloxamer 188.
nium acetate buffer (pH 4.2) in a ratio of 45: • Mannitol.
55 (flow rate 1.2 mL/min, UV detection • Polysorbate 80.
wavelength 306 nm). • MC One (fluid jet mill).
Method • Niro-Soavi NS1001L (high-pressure
• Prepare solutions of HP-β-CD at various homogenizer).
concentrations (from 0 to 0.2 M) in distilled • Laboratory freeze drier.
water or pH-adjusted buffers. Method
• Add an excess amount of CKD-732 to the • Perform all the following steps under
HP-β-CD solutions in glass vials. reduced lighting conditions to protect the
• Shake vials at 200 rpm for at least 72 hours light-sensitive drug.
to achieve equilibrium. • Jet mill the coarse powder of nimodipine to
• Centrifuge samples and filter using a 0.2-μ produce a microparticulate powder.
m syringe filter. • Disperse with a magnetic stirrer 0.5% (w/v)
• Analyze CKD-732 content of samples the milled powder in an aqueous solution
using HPLC to determine solubility. composed of 0.6% (w/v) poloxamer
Results 188, 0.4% (w/v) sodium cholate, and 4.0%
• The aqueous solubility of CKD-732 (w/v) mannitol.
increases linearly as a function of • Perform premilling using the high-pressure
HP-β-CD concentration, suggesting the for- homogenizer (maintain sample temperature
mation of a 1:1 complex. at 25–30 C) by starting with the settings at
• Solubility of CKD-732 increased as a func- 200 bar with two cycles then increasing to
tion of HP-β-CD concentration at pH values 500 bar with five cycles.
of 6.7 and 9.7. However, the solubility • Follow the premilling step with 15–-
values at pH 6.7 at different HP-β-CD 20 cycles at 1500 bar to produce the
concentrations are approximately four-fold nanosuspension.
higher than their counterparts at pH 9.7. • Lyophilize the nanosuspension by drying
for 15 hours at 15 C (below
Method Capsule 8 200 mTorr), with a secondary step of
3 hours at 5 C, and a final step of
Preparation of an IV Nanosuspension Formu-
2 hours at 20 C.
lation for Reduced Irritation and Phlebitis
• Sterilize by gamma irradiation for 6 hours
Upon Injection
with an absorbed dose of 12 kGy.
Based on the method reported by Xiong et al.
Results
(2007).
• Injection toleration studies performed in
Objective
rabbits demonstrate that this formulation
• To improve upon a current clinical IV for-
reduces the occurrence of phlebitis and
mulation for nimodipine, which contains a
minimizes local irritation when compared
high concentration of ethanol that causes
with the clinical ethanol-based product.
injection site pain, irritation, and phlebitis.
Method Capsule 9 • For animal injection, the lyophilized

microspheres were resuspended in a vehicle
Preparation of Poly (Lactide-co-Glycolide)
composed of sterile normal saline,
Microspheres with Microencapsulated
NaCMC, and polysorbate 80, then injected
Prednisolone for Sustained Release Follow-
SC in mice with an adjuvant-induced rheu-
ing IM or SC Injection
matoid arthritis.
Based on the method reported by Khaled et al.
Results
(2010)
• The microsphere formulation significantly
Objective
reduces the inflammation in the feet of the
• To provide a sustained release of the anti-
mice where Freund’s complete adjuvant is
inflammatory steroid prednisolone over
injected, compared to both the control mice
3–4 weeks following the IM or SC injection
(no drug treatment) and the prednisolone
for chronic inflammation conditions.
single-dose-treated mice.
• Prednisolone.
• Poly (lactide-co-glycolide) (PLG 50:50,
Resomer RG 503H). References
• Polyvinyl alcohol (PVA, partially
hydrolyzed 88%). Abilify Maintena. Product label. Tokyo: Otsuka Pharma-
ceutical; 2016. https://www.accessdata.fda.gov/
• Polysorbate 80. drugsatfda_docs/label/2016/202971s008lbl.pdf.
• Carboxymethylcellulose Sodium salt Accessed 27 Feb 2021
(NaCMC). Abraxane. Product label. Summit, NJ: Celgene corpora-
• Sterile normal saline. tion; 2015. http://www.abraxane.com/downloads/
Abraxane_PrescribingInformation.pdf. Accessed
• Bath sonicator. 27 Feb 2021
• Benchtop homogenizer. Akers M. Parenteral preparations. In: Felton L, editor.
• Magnetic stirrer. By. Remington - essentials of pharmaceutics. 1st ed.
• Centrifuge. London: Pharmaceutical Press; 2012. p. 495–532.
Aloisio C, Antimisiaris SG, Longhi MR. Liposomes
• Laboratory freeze drier. containing cyclodextrins or meglumine to solubilize
Method and improve the bioavailability of poorly soluble
• Prepare a suspension (solid-in-oil, s/o) of drugs. J Mol Liq. 2017;229:106–13.
prednisolone in dichloromethane (DCM, Ambisome. Product label. San Dimas, CA: Gilead
Sciences, Inc.; 2012. https://www.ambisome.com.
1.5 mL) using a bath sonicator. PLG Accessed 27 Feb 2021
(12.5% w/v) is completely dissolved in Anselmo AC, Mitragotri S. Nanoparticles in the clinic: an
dichloromethane prior to prednisolone addi- update. Bioeng Transl Med. 2019;4(3):e10143.
tion. Drug-to-polymer ratio is 1:4 (w/w). Anton N, Hallouard F, Attia MF, Vandamme TF. Nano-
emulsions for drug delivery and biomedical
• Prepare a solid-in-oil-in-water emulsion imaging. In: Prokop A, Weissig V, eds. Intracellular
(s/o/w) of the oil phase with an aqueous delivery III. Springer; 2016:273–300.
phase composed of 1% (w/v) PVA solution Aristada. Product label. Waltham: Alkermes; 2015. http://
in water (50 mL) by adding the oil phase aristada.com/hcp. Accessed 27 Feb 2021
Aristada Initio Kit. Product label. Waltham: Alkermes.
dropwise into the aqueous phase during 2020. https://www.aristadahcp.com/downloadables/
stirring at 2000 rpm using the benchtop ARISTADA-INITIO-PI.pdf Accessed 27 Feb 2021.
homogenizer. Asano J, So S, Lodge TP. Location and influence of added
• The formed emulsion is left to stir at room block copolymers on the droplet size in oil-in-oil
emulsions. Langmuir. 2015;31(27):7488–95.
temperature to evaporate the solvent for a Aschenbrenner DS, Venable SJ. Drugs affecting cardiac
few hours using a magnetic stirrer. rhythm. In: Aschenbrenner DS, Venable SJ, editors.
Microspheres are collected by centrifuga- Drug therapy in nursing. 3rd ed. Philadelphia: Wolters
tion, washed three times with water, and Kluwer Health, Lippincott Williams & Wilkins; 2009.
then lyophilized overnight.
Ativan. Product label. Deerfield: Baxter Healthcare; 1980. pharmacokinetics. J Pharm Investig. 2018;48(1):
https://www.accessdata.fda.gov/drugsatfda_docs/ 43–60.
label/2006/018140s028lbl.pdf. Accessed 27 Feb 2021 Clorotekal. Product label. Bethlehem: B. Braun Medical;
Bari H. A prolonged release parenteral drug delivery 2017. https://www.accessdata.fda.gov/drugsatfda_
system-an overview. Int J Pharm Sci Rev Res. 2010;3 docs/label/2017/208791s000lbl.pdf. Accessed
(1):1–11. 27 Feb 2021
Benhabbour SR, Kovarova M, Jones C, et al. Ultra-long- Croy SR, Kwon GS. Polymeric micelles for drug delivery.
acting tunable biodegradable and removable controlled Curr Pharm Des. 2006;12(36):4669–84.
release implants for drug delivery. Nat Commun. Crucho CIC, Barros MT. Polymeric nanoparticles: a study
2019;10(1):4324. on the preparation variables and characterization
Bhalla S. Parenteral delivery. In: Desai A, Lee M, editors. methods. Mater Sci Eng C. 2017;80:771–84.
Gibaldi’s drug delivery systems in pharmaceutical DDAVP injection. Product label. Bridgewter: Sanofi-
care. Bethesda: American Society of Health-Systems Aventis; 2007. http://products.sanofi.us/ddavp_iv/
Pharmacists; 2007. ddavp_iv.pdf. Accessed 27 Feb 2021
Bhattacharjee H, Thoma A. Parenteral drug administra- De Spiegeleer B, Wattyn E, Slegers G, et al. The impor-
tion: routes of administration and devices. In: tance of the cosolvent propylene glycol on the antimi-
Nema S, Ludwig JD, editors. Pharmaceutical dosage crobial preservative efficacy of a pharmaceutical
forms parenteral medications. 3rd ed. New York: formulation by DOE-ruggedness testing. Pharm Dev
Informa Healthcare; 2010. Technol. 2006;11(3):275–84.
Bulbake U, Doppalapudi S, Kommineni N, Khan Dextenza. Product label. Bedford: Ocular Therapeutix;
W. Liposomal formulations in clinical use: an updated 2019. https://www.accessdata.fda.gov/drugsatfda_
review. Pharmaceutics. 2017;9(4):12. docs/label/2019/208742s001lbl.pdf. Accessed
Byfavo. Product label. Indianapolis: Acacia Pharam; 2020. 27 Feb 2021
https://www.accessdata.fda.gov/drugsatfda_docs/ Diazemuls. Product monograph. In: Gillis MC, editor.
label/2020/212295s000lbl.pdf. Accessed 27 Feb 2021 CPS compendium of pharmaceuticals and specialties.
Cabenuva. Product label. Research Triangle Park: Viiv; 33rd ed. Ottawa: Canadian Pharmacists
2021. https://www.accessdata.fda.gov/drugsatfda_ Association; 1998.
docs/label/2021/212888s000lbl.pdf. Accessed Dodou K. Intrathecal route of drug delivery can save lives
27 Feb 2021 or improve quality of life. Pharm J. 2012;289:501–2.
Caló E, Khutoryanskiy VV. Biomedical applications of Duarte ARC, Costa MS, Simplício AL, Cardoso MM,
hydrogels: a review of patents and commercial Duarte CMM. Preparation of controlled release
products. Eur Polym J. 2015;65:252–67. microspheres using supercritical fluid technology for
Cardene. Product label. Deerfield: Baxter Healthcare; delivery of anti-inflammatory drugs. Int J Pharm.
2010. http://www.accessdata.fda.gov/drugsatfda_ 2006;308(1–2):168–74.
docs/label/2011/019734s017lbl.pdf. Accessed Edex. Product label. Lake Forest: Actient
27 Feb 2021 Pharmaceuticals; 2011. http://www.edex.com/assets/
Cardone D, Tallia A. Diagnostic and therapeutic injection pdf/prescribing_information.pdf. Accessed
of the hip and knee. Am Fam Physician. 2003;67(10): 27 Feb 2021
2147–52. Eligard. Product label. Fort Collins: Tolmar
Carter N. Extended-release intramuscular Paliperidone Pharmaceuticals Inc.; 2015. http://dailymed.nlm.nih.
palmitate a review of its use in the treatment of schizo- gov/dailymed/drugInfo.cfm?setid¼b78d1919-9dee-
phrenia. Drugs. 2012;72(8):1137–60. 44fa-90f9-e0a26d32481d.. Accessed 27 Feb 2021
Celestone Soluspan. Product label. Whitehouse Station: Eloy JO, de Souza MC, Petrilli R, Barcellos JPA, Lee RJ,
Merck; 2015. http://www.accessdata.fda.gov/ Marchetti JM. Colloids Surf B Biointerfaces.
drugsatfda_docs/label/2015/014602s059lbl.pdf. 2014;123:345–63.
Accessed 27 Feb 2021 Fang J, Nakamura H, Maeda H. The EPR effect: unique
Chaisri W, Hennink WE, Okonogi S. Preparation and features of tumor blood vessels for drug delivery,
characterization of cephalexin loaded PLGA factors involved, and limitations and augmentation of
microspheres. Curr Drug Deliv. 2009;6(1):69–75. the effect. Adv Drug Deliv Review. 2011;63:136–51.
Challa R, Ahuja A, Ali J, et al. Cyclodexrins in drug Genexol PM. 2015. https://www.samyangbiopharm.com/
delivery: an updated review. AAPS PharmSciTech. eng/ProductIntroduce/injection01. Accessed
2005;6(2):329–57. 27 Feb 2021.
Chen J, Sun P, Zhou X, Zhang Y, Huang C-H. Cu(II) Gerwin N, Hops C, Lucke A. Intraarticular drug delivery
catalyzed transformation of Benzylpenicillin revisited: in osteoarthritis. Adv Drug Deliv Rev. 2006;58(2):
the overlooked oxidation. Environ Science Technol. 226–42.
2015;49:4218–25. Gilbert J. US Patent No. 10525033B2. 2020.
Choi YH, Han H-K. Nanomedicines: current status and Washington, DC: U.S. Patent and Trademark Office.
future perspectives in aspect of drug delivery and Goncalez ML, Marcussi DG, Calixto GM, Correa MA,
Chorilli M. Structural characterization and in vitro
antioxidant activity of kojic dipalmitate loaded w/o/ European Journal of Pharmaceutical Sciences, 2006;
w multiple emulsionsintended for skin disorders. 2015. 28(1–2):7–14, ISSN 0928-0987, https://doi.org/10.
Gupta D, Bhatia D, Dave V, Sutariya V, Varghese 1016/j.ejps.2005.11.012.
GS. Salts of therapeutic agents: chemical, physico- Khaled KA, Sarhan HA, Ibrahim MA, Ali AH, Naguib
chemical, and biological considerations. Molecules. YW. Prednisolone-loaded PLGA microspheres. In
2018;23(7):1719. vitro characterization and in vivo application in
Halbert GH. Preformulation. In: Florence AT, Shipman J, adjuvant-induced arthritis in mice. AAPS
editors. Modern phramaceutics: basic principles and PharmSciTech. 2010;11(2):859–69.
systems, vol. 1. New York: Informa Healthcare; 2009. Jae-Hyun Kim, Su-Kyung Lee, Min-Hyo Ki, Won-Kyu
Horisawa E, Kubota K, Tuboi I, et al. Size-dependency of Choi, Soon-Kil Ahn, Hee-Jong Shin, Chung Il Hong,
DL-lactide/glycolide copolymer particulates for intra- Development of parenteral formulation for a novel
articular delivery system on phagocytosis in rat angiogenesis inhibitor, CKD-732 through complexa-
synovium. Pharm Res. 2002;19(2):132–9. tion with hydroxypropyl-β-cyclodextrin, International
Hovorka S, Schoneich C. Oxidative degradation of Journal of Pharmaceutics, 2004;272(2):79–89, ISSN
pharmaceuticals: theory, mechanisms and inhibition. 0378-5173, https://doi.org/10.1016/j.ijpharm.2003.11.
J Pharm Sci. 2001;90:253–69. 034.
Huang Z, Li X, Zhang T, et al. Progress involving new Lammers T, Kiessling F, Hennink W, Storm G. Drug
techniques for liposome preparation. Asian J targeting to tumors: principles, pitfalls and (pre-) clini-
Pharmaceut Sci. 2014;9(4):176–82. cal progress. J Control Release. 2012;161:175–87.
Hydralazine hydrochloride injection. Product label. Laouini A, Jaafar-Maalej C, Sfar S, Charcosset C, Fessi
Shirley: American Reagent, Inc. 2013. http://www. H. Liposome preparation using a hollow fiber mem-
accessdata.fda.gov/drugsatfda_docs/label/2013/ brane contactor--application to spironolactone encap-
040136s005lbl.pdf. Accessed 27 Feb 2021. sulation. Int J Pharm. 2011;415(1–2):53–61.
Imcivree. Product label. Boston: Rhythm Pharmaceuticals; Larsen C, Østergaard J, Larsen S, Jensen H, Jacobsen S,
2020. https://www.accessdata.fda.gov/drugsatfda_ Lindegaard C, et al. Intra-articular depot formulation
docs/label/2020/213793s000lbl.pdf. Accessed principles: role in the management of postoperative
27 Feb 2021 pain and arthritic disorders. J Pharm Sci. 2008;97
Intralipid. Product label. Deerfield: Baxter Healthcare; (11):4622–54.
1996. https://www.accessdata.fda.gov/drugsatfda_ Larson N. US Patent No. 10,675,296. 2020.
docs/label/2007/017643s072,018449s039lbl.pdf. Washington, DC: U.S. Patent and Trademark Office.
Accessed 27 Feb 2021 Lavelle W, Lavelle E, Lavelle L. Intra-articular injections.
Intron A. Product label. Whitehouse Station: Merck; 2015. Anesthesiol Clin. 2007;25(4):853–62.
https://www.merck.com/product/usa/pi_circulars/i/ Lee JH. Injectable hydrogels delivering therapeutic agents
intron_a/intron_a_pi.pdf. Accessed 27 Feb 2021 for disease treatment and tissue engineering. Biomater
Invega Sustenna. Product label. Titusville: Janssen Res. 2018;22(1):27.
Pharmaceuticals, Inc; 2015. http://www.janssencns. Lesoin L, Crampon C, Boutin O, Badens E. Preparation of
com/invegasustenna/. Accessed 27 Feb 2021 liposomes using the supercritical anti-solvent (Sas)
Invega Trinza. Product label. Titusville: Janssen process and comparison with a conventional method.
Pharmaceuticals Inc; 2015. https://www.accessdata. J Supercrit Fluids. 2011;57(2):162–74.
fda.gov/drugsatfda_docs/label/2017/207946s003lbl. Lioresal Intrathecal. Product label. Minneapolis:
pdf. Accessed 27 Feb 2021. Medtronic, Inc. 2013. http://www.accessdata.fda.gov/
Jaafar-Maalej C, Charcosset C, Fessi H. A new method for drugsatfda_docs/label/2013/020075s026lbl.pdf.
liposome preparation using a membrane contactor. J Accessed 27 Feb 2021.
Liposome Res. 2011;21(3):213–20. Loftsson T, Jarho P, Másson M, et al. Cyclodextrins in
Ju C, Sun J, Zi P, Jin X, Zhang C. Thermosensitive drug delivery. Expert Opin Drug Deliv. 2005;2(2):
micelles–hydrogel hybrid system based on poloxamer 335–51.
407 for localized delivery of paclitaxel. J Pharm Sci. Maeda H, Wu J, Sawa T, Matsumura Y, Hori K. Tumor
2013;102(8):2707–17. vascular permeability and the EPR effect in macromo-
Kale S, Deore S. Emulsion micro emulsion and Nano lecular therapeutics: a review. J Control Release.
emulsion: a review. Systematic Reviews in Pharmacy. 2000;65(1–2):271–84.
2017;8(1):39–47. Malik SA, Ng WH, Bowen J, et al. Electrospray synthesis
Kamat M, DeLuca PP. Formulation development of small and properties of hierarchically structured PLGA TIPS
and large volume injections. In: Nema S, Ludwig JD, microspheres for use as controlled release
editors. Pharmaceutical dosage forms parenteral technologies. J Colloid Interface Sci. 2016;467:220–9.
medications. 3rd ed. New York: Informa Massing U, Cicko S, Ziroli V. Dual asymmetric centrifu-
Healthcare; 2010. gation (Dac) – a new technique for liposome prepara-
Kawakami K, Oda N, Miyoshi K, Funaki T, Ida Y, tion. J Control Release. 2008;125(1):16–24.
Solubilization behavior of a poorly soluble drug
under combined use of surfactants and cosolvents,
McGinity J, O’Donnell P. Preparation of microspheres by with flexible drug delivery kinetics modulation. J Con-
the solvent evaporation technique. Adv Drug Deliv trol Release. 2020;319:416–27.
Rev. 1997;28(1):25–42. Ryanodex. Product label. Woodcliff Lake: Eagle
McKenzie M, Betts D, Suh A, Bui K, Kim L, Cho Pharmaceuticals, Inc. (2014. http://www.ryanodex.
H. Hydrogel-based drug delivery systems for poorly com/wp-content/uploads/2014/07/ryanodex-prescrib
water-soluble drugs. Molecules. 2015;20(11): ing-information.pdf. Accessed 27 Feb 2021.
20397–408. Scenesse. Product label. West Menlo Park: Clinuvel Inc.
Mepact. Product label. Linz: Takeda Austria GmbH; 2009. 2019. https://www.accessdata.fda.gov/drugsatfda_
https://www.ema.europa.eu/en/documents/product- docs/label/2019/210797s000lbl.pdf. Accessed
information/mepact-epar-product-information_en.pdf. 27 Feb 2021.
Accessed 27 Feb 2021 Schwartzberg LS, Navari RM. Safety of polysorbate 80 in
Meyer B, Ni A, Hu B, Shi L. Antimicrobial preservative the oncology setting. Adv Ther. 2018;35(6):754–67.
use in parenteral products: past and present. J Pharm Serajuddin A. Salt formation to improve drug solubility.
Sci. 2007;96(12):3155–67. Adv Drug Del Rev. 2007;59:603–16.
Mishra V, Bansal K, Verma A, et al. Solid lipid Sheokand S, Reddy V, Bansal A. Pharmaceutical
nanoparticles: emerging colloidal nano drug delivery nanocrystals: from fundamentals to advances. Pharma
systems. Pharmaceutics. 2018;10(4):191. Times. 2018;50(2):20–5.
Mottu F. Organic solvents for pharmaceutical parenterals Shetab Boushehri M, Dietrich D, Lamprecht
and embolic liquids: a review of toxicity data. PDA J A. Nanotechnology as a platform for the development
Pharm Sci Technol. 2015;54:456–69. of injectable parenteral formulations: a comprehensive
Neustadt D. Intra-articular injections for osteoarthritis of review of the know-hows and state of the art.
the knee. Cleve Clin J Med. 2006;73(10):897–8. Pharmaceutics. 2020;12(6):510.
Nikam TH, Patil MP, Patil SS, Vadnere GP, Lodhi Shi Y, Porter W, Merdan T, Li LC. Recent advances in
S. Nanoemulsion: a brief review on development and intravenous delivery of poorly water-soluble
application in parenteral drug delivery. Adv Pharm compounds. Expert Opin Drug Deliv. 2009;6(12):
J. 2018;3(2):43–54. 1261–82.
Otake K, Imura T, Sakai H, Abe M. Development of a new Smoflipid Product monograph. 2012. Fresenius Kabi
preparation method of liposomes using supercritical Canada. http://fresenius-kabi.ca/en/wp-content/
carbon dioxide. Langmuir. 2001;17(13):3898–901. uploads/sites/2/2013/06/SMOFLipid-PM.pdf.
Otake K, Shimomura T, Goto T, et al. Preparation of Accessed 27 Feb 2021.
liposomes using an improved supercritical reverse Solu-Medrol. Product label. New York: Pfizer; 2011.
phase evaporation method. Langmuir. 2006;22(6): http://www.accessdata.fda.gov/drugsatfda_docs/label/
2543–50. 2011/011856s103s104lbl.pdf. Accessed 27 Feb 2021
Patel VR, Agrawal YK. Nanosuspension: an approach to Soni MG, Carabin IG, Burdock GA. Safety assessment of
enhance solubility of drugs. J Adv Pharm Technol Res. esters of p-hydroxybenzoic acid (parabens). Food
2011;2(2):81–7. Chem Tox. 2005;43:985–1015.
Perseris. Product label. North Chesterfield: Indivior; 2018. Strickley RG. Solubilizing excipients in oral and injectable
https://www.accessdata.fda.gov/drugsatfda_docs/ formulations. Pharm Res. 2004;21(2):201–30.
label/2018/210655s000lbl.pdf. Accessed 27 Feb 2021 Sublocade. Product label. North Chesterfield: Indivior;
Peschka R, Dennehy C, Szoka FC. A simple in vitro model 2017. https://www.accessdata.fda.gov/drugsatfda_
to study the release kinetics of liposome encapsulated docs/label/2017/209819s000lbl.pdf. Accessed
material. J Control Release. 1998;56(1–3):41–51. 27 Feb 2021
Nexterone. Product label. Deerfield: Baxter; 2011. https:// Suprelin LA. Product label. Chadds Ford: Endo
www.accessdata.fda.gov/drugsatfda_docs/label/2011/ Pharmaceuticals; 2011. https://www.accessdata.fda.
022325s002lbl.pdf. Accessed 27 Feb 2021 gov/drugsatfda_docs/label/2011/022058s006lbl.pdf.
Płaczek M, Wątróbska-Świetlikowska D, Stefanowicz- Accessed 27 Feb 2021
Hajduk J, Drechsler M, Ochocka JR, Sznitowska Sur S, Fries AC, Kinzler KW, Zhou S, Vogelstein
M. Comparison of the in vitro cytotoxicity among B. Remote loading of preencapsulated drugs into
phospholipid-based parenteral drug delivery systems: stealth liposomes. PNAS. 2014;111:2283–8.
emulsions, liposomes and aqueous lecithin dispersions Taxotere. Product label. Bridgewter: Sanofi-Aventis;
(Wlds). Eur J Pharm Sci. 2019;127:92–101. 2010. http://www.accessdata.fda.gov/drugsatfda_
Prevymis. Product label. Whitehouse Station: Merck docs/label/2010/020449s059lbl.pdf. Accessed
Sharpe & Dohme; 2017. https://www.accessdata.fda. 27 Feb 2021
gov/drugsatfda_docs/label/2006/018140s028lbl.pdf. Teixeira MC, Carbone C, Souto EB. Beyond liposomes:
Accessed 27 Feb 2021 recent advances on lipid based nanostructures for
Roberge C, Cros J-M, Serindoux J, et al. BEPO®: poorly soluble/poorly permeable drug delivery. Prog
Bioresorbable diblock mPEG-PDLLA and triblock Lipid Res. 2017;68:1–11.
PDLLA-PEG-PDLLA based in situ forming depots
Thompson DO. Cyclodextrins: enabling excipients - their Ventola CL. Progress in nanomedicine: approved and
present and future use in pharmaceuticals. Crit Rev investigational nanodrugs. P T. 2017;42(12):742–55.
Ther Drug Carrier Syst. 1997;14(1):1–104. Vibativ. Product label. Deerfield: Astellas Pharma; 2009.
Tong W, Wen H. Preformulation aspects of water insolu- http://www.accessdata.fda.gov/drugsatfda_docs/label/
ble compounds. In: Liu R, editor. Water-insoluble drug 2009/022110s000lbl.pdf. Accessed 27 Feb 2021
formulation. 2nd ed. Boca Raton: CRC Press; 2008. Visudyne. Prescribing information. Valeant
Top selling drugs by revenue. Pharmaceutical Technol- Pharmaceuticals. 2013. http://www.bausch.com/
ogy. 2020. https://www.pharmaceutical-technology. Portals/109/-/m/BL/United%20States/Files/Package%
com/features/top-selling-prescription-drugs/ 20Inserts/Pharma/visudyne-package-insert.pdf.
Tran AX, Whitfield C. Lipopolysaccharides Accessed on 27 Feb 2021.
(endotoxins). In: Schaechter M, editor. Encyclodepia Wang T, Deng Y, Geng Y, Gao Z, Zou J, Wang
of microbiology. 3rd ed. London: Academic; 2009. Z. Preparation of submicron unilamellar liposomes by
p. 513–28. freeze-drying double emulsions. Biochimica et
Triptodur. Product label. Atlanta: Arbor Pharmaceuticals; Biophysica Acta (BBA) Biomembranes. 2006;1758
2017. https://www.accessdata.fda.gov/drugsatfda_ (2):222–31.
docs/label/2017/208956s000lbl.pdf. Accessed Waterman K, Adami R, Alsante K, Hong J, Landis M,
27 Feb 2021 Lombardo F, Roberts C. Stabilization of
Trissel LA, editor. Handbook of injectable drugs. 12th pharmaceuticals to oxidative degradation. Pharmaceu-
ed. Bethesda: American Society of Health-System tical Dev Technol. 2002;7:1–32.
Pharmacists; 2002. Wu J, Wu L, Xu X, Xu X, Yin X, Chen Y, Hu
Troy DB, editor. Remington: the science and practice of Y. Microspheres made by w/o/o emulsion method
pharmacy. 21st ed. Philadelphia: Lippincott Williams with reduced initial burst for long-term delivery of
& Wilkins; 2006. endostar, a novel recombinant human endostatin. J
United States Pharmacopeia and National Formulary Pharm Sci. 2009;98(6):2051–8.
(USP43-NF38 2S). https://online.uspnf.com/uspnf/?_ Xiao L, Huang L, Moingeon F, Gauthier M, G. YangpHre-
ga¼2.19661552.952334861.1554828162-258162058. sponsive poly(ethylene glycol)-block-polylactide
1554828162. Accessed 27 Feb 2021. micelles for tumor-targeted drug delivery Biomacro-
USP<51> Antimicrobial Effectiveness Testing. The molecules. 2017;18(9):2711–2722.
United States Pharmacopeia 43rd revision. Xiong R, Lu W, Li J, Wang P, Xu R, Chen T. Preparation
Rockville. 2020. and characterization of intravenously injectable
USP<71> Sterility Tests. The United States Pharmacopeia nimodipine nanosuspension. Int J Pharm. 2008;28
43rd revision. Rockville. 2020. (350):338–43. https://doi.org/10.1016/j.ijpharm.2007.
USP<729> Globule Size Distribution in Lipid Injectable 08.036
Emulsions. The United States Pharmacopeia 43rd revi- Zantac. Product label. Mississauga: GlaxoSmithKline, Inc.
sion. Rockville. 2020. 2015. http://www.gsk.ca/english/docs-pdf/product-
USP<785> Osmolality and Osmolarity. The United States monographs/Zantac.pdf. Accessed 27 Feb 2021.
Pharmacopeia 43rd revision. Rockville. 2020. Zepzelca. Product label. Palo Alto: Jazz Pharmaceuticals;
USP<788> Particulate matter in injections. The United 2020. https://www.accessdata.fda.gov/drugsatfda_
States Pharmacopeia 43rd revision. Rockville. 2020. docs/label/2020/213702s000lbl.pdf. Accessed
USP<790> Visible Particulates in Injections. The United 27 Feb 2021
States Pharmacopeia 43rd revision. Rockville. 2020. Zilretta. Product label. Burlingame: Flexion Therapeutics;
USP<85> Bacterial Endotoxins Test. The United States 2017. https://www.accessdata.fda.gov/drugsatfda_docs/
Pharmacopeia 43rd revision. Rockville. 2020. label/2017/208845s000lbl.pdf. Accessed 27 Feb 2021
Vibramycin. Product label. New York: Pfizer; Zyprexa Relprevv. Product label. Indianapolis: Eli Lilly
2014. https://www.accessdata.fda.gov/drugsatfda_ and Company; 2009. https://www.accessdata.fda.gov/
docs/label/2013/050442s016lbl.pdf. Accessed 27 Feb drugsatfda_docs/label/2009/022173lbl.pdf. Accessed
2021. 27 Feb 2021
van Winden EC, Zhang W, Crommelin DJ. Effect of
freezing rate on the stability of liposomes during
freeze-drying and rehydration. Pharm Res. 1997;14
(9):1151–60. Capsule References
Veklury. Product label. Foster City: Gilead Sciences;
2020. https://www.accessdata.fda.gov/drugsatfda_
docs/label/2020/214787Orig1s000lbl.pdf. Accessed Moriwaki C, Costa GL, Ferracini CN, et al. Enhancement
27 Feb 2021 of solubility of albendazole by complexation with
Vemuri N. Preformulation. In: Nema S, Ludwig JD, editors. β-cyclodextrin. Braz J Chem Eng. 2008;25(2):255–67.
Pharmaceutical dosage forms parenteral medications. Park C-W, Lee H-J, Oh D-W, Kang J-H, Han C-S, Kim
3rd ed. New York: Informa Healthcare; 2010. D-W. Preparation and in vitro/in vivo evaluation of
PLGA microspheres containing norquetiapine for
long-acting injection. Drug Des Devel Ther. 2018;12: Seedher N, Kanojia M. Micellar solubilization of some
711–9. poorly soluble antidiabetic drugs: a technical note.
Poudel AJ, He F, Huang L, Xiao L, Yang AAPS PharmSciTech. 2008;9(2):431–6.
G. Supramolecular hydrogels based on poly (ethyl- Zhong H, Deng Y, Wang X, et al. Multivesicular liposome
ene glycol)-poly (lactic acid) block copolymer formulation for the sustained delivery of breviscapine.
micelles and α-cyclodextrin for potential injectable Int J Pharm. 2005;301:15–24.
drug delivery system. Carbohydr Polym. 2018;194:
69–79.
Lipid-Based Formulations
7
Daniel A. Davis Jr., Han-Hsuan Peng, and Robert O. Williams III
Abstract provided on the solidification of LBFs and

The understanding in design, classification, the benefits and challenges associated with
and characterization of lipid-based the techniques involved. The current authors
formulations (LBFs) has evolved greatly over would like to thank and acknowledge the pre-
the last two decades. LBFs include simple vious authors’ significant contribution from
lipid solutions, self-emulsifying and self- the first and second editions. This current
microemulsifying drug delivery systems third edition chapter is a revision and update
(SEDDS and SMEDDS), and surfactant- of the original authors’ work form the first and
cosolvent solutions, designated as Type I to second editions.
IV systems depending on their composition
and properties. Notably, this chapter details Keywords
several mechanistic studies that have helped Lipid-based formulations (LBFs) · Self-
to elucidate the pathways involved for Emulsifying Drug Delivery Systems
increased absorption of poorly soluble drugs, (SEDDS) · Self-Microemulsifying Drug
which have aided in the design and Delivery Systems (SMEDDS) ·
standardization of LBF in vitro characteriza- Supersaturation · Digestion
tion methods. Much of this work has evolved
through the recent Lipid Formulations Classi-
fication System Consortium, an academic- 7.1 Introduction
industrial group of experts and stakeholders
that have worked to advance the state of the Oral delivery of lipid-based formulations (LBFs)
art. The role of sustained supersaturation, trig- has been utilized for decades as a means to
gered by LBF dilution and digestion, in improve absorption of poorly water-soluble,
enhancing drug absorption is discussed, lipophilic drugs. For dissolution rate-limited
along with formulation variables intended for drugs, LBFs present the drug in a presolubilized
this purpose. Finally, a short review is form to the gastrointestinal tract for absorption.
For solubility-limited drugs, several mechanisms
have been proposed for enhanced absorption.
These include increased drug permeability via
D. A. Davis Jr. (*) · H.-H. Peng · R. O. Williams III transporters (Constantinides and Wasan 2007;
Present Address: The University of Texas at Austin Goole et al. 2010), through presentation after
Austin, TX, USA collisional transfer of bile-salt-mixed micelles
254 D. A. Davis Jr. et al.
a Solubilization effect b Solubilization and

supersaturation effect
Higher drug solubility in
GI fluids plus the LBF
Mass of solubilized drug in the
Mass of solubilized drug in the

sustained supersaturation
Solubility decreases during LBF

LBF dispersion digestion supersaturation
effect
GI tract
GI tract
LBF solubilization LBF solubilization
effect effect
Crystalline drug Drug solubility in Crystalline drug

tablet/capsule GI fluids tablet/capsule
time time
Fig. 7.1 Schematic representation of the solubilization drug. In addition, the changing nature of colloidal spe-
and supersaturation effects afforded by lipid-based cies formed as LBFs disperse and become integrated
formulations (LBFs). Lipids and surfactants (and to a into lipid digestion and absorption pathways, and the
lesser extent cosolvents) boost drug solubilization progressive decrease in drug solubility in these colloids
capacity in the GI tract following the dispersion and that typically occur as these processes continue is such
digestion of LBFs. This is described in (a) as the “solu- that LBFs have the capacity to generate supersaturation,
bilization effect,” and as shown, this solubilization described above in (b) as the “supersaturation effect.”
effect affords higher stable drug concentrations than (Reproduced with permission from Williams et al.
traditional solid dosage forms containing crystalline 2013b)
(Gao and Morozowich 2007; Aburahma 2016), exceeds the available solvent capacity, the drug
decreased first-pass metabolism (Patel and will either be in a supersaturated state driving
Brocks 2009; Trevaskis et al. 2006), and, for enhanced absorption, or will rapidly precipitate
highly lipophilic drugs (e.g., log P > 6), an resulting in limited absorption, as shown in
increase in lymphatic transport (O'Driscoll 2002; Fig. 7.1b (Williams et al. 2013b). The sustained
Trevaskis et al. 2008; Vishwakarma et al. 2019). supersaturation pathway provides agreement
Recently, however, understanding has evolved between the demonstrated success of several
implicating dilution and/or digestion-induced LBFs to enhance absorption, and the recently
supersaturation as a predominant mechanism of described solubility–permeability interplay that
enhanced absorption of poorly water-soluble clearly demonstrated a solubilization effect
drugs in many examples of LBFs (Gao and alone by the formation of lipid mixed micelles
Morozowich 2006; Yeap et al. 2013), potentially would result in lower absorption due to reduced
including notable marketed LBFs such as free drug concentration, as demonstrated in
Neoral® and Agenerase® (Strickley 2004; cosolvent, surfactant, and cyclodextrin systems
Williams et al. 2013b). (Dahan et al. 2010; Miller et al. 2012; Miller
Upon ingestion of an LBF, solubilization of et al. 2011). Promoting sustained supersaturation
the drug is altered over time due to dilution/dis- versus drug precipitation in LBFs has been
persion in aqueous gastrointestinal fluid and bile- attributed to the properties of the lipid
salt secretions (Devraj et al. 2013), as well as components used (e.g., solubility, digestibility),
digestion of lipid formulation components by the maximum degree of supersaturation based on
gastric and pancreatic lipases (Anby et al. 2012; the administered dose, and the inclusion of poly-
Bakala-N'Goma et al. 2015), as shown in meric precipitation inhibitors (PPIs) within the
Fig. 7.1a. As the mass of drug in the GI tract LBF.
7 Lipid-Based Formulations 255
Given the performance of a given LBF 7.2 Classification and Components

depends on the formulation composition and of Lipid-Based Formulations
lipid digestion, developing discriminating and
biorelevant in vitro methods to guide formula- LBFs commonly contain pharmaceutical
tion selection is a complex but necessary ingredients, including natural oils, fatty acids,
endeavor. This chapter will highlight the prog- partial glycerides, surfactants, and cosolvents
ress that has been made in recent years to (Dahan and Hoffman 2008; Trevaskis et al.
develop and standardize in vitro methods in 2008). LBFs are classified into four categories
order to adequately assess the diverse types of according to their composition, as listed in
formulations that are classified as LBFs (see Table 7.1; additionally, the table describes differ-
Sect. 7.2), and how these methods have helped ent characteristics, advantages, and disadvantages
to elucidate the mechanisms of enhanced for each category (Pouton 2000, 2006; Vithani
absorption via sustained supersaturation. et al. 2019b). In the simplest lipid-based
Comparisons to in vivo data will be made formulations (Type I), the drug is dissolved in
where available. Finally, a review of recent glycerides. Type I formulations are nondispersive
efforts to develop solid-LBFs by absorbing liq- and require digestion. Digestive products of
uid components on solid substrates or through glycerides mix with bile salts in the GI tract to
the use of lipids that are solid at room tempera- create mixed micelles where the drug remains
ture will be provided. solubilized. The micelles function as drug
Table 7.1 Lipid-based formulation classification system and associated characteristics, advantages, and disadvantages
of various types of LBFs
Type I Type II Type IIIA Type IIIB Type IV
Materials Oils (100%) Oils Oils (40–80%), Oils (<20%), water- Water-soluble and
without (40–80%) water-soluble and/or soluble and/or water- (0–20%)/or water-
surfactants and water- water-insoluble insoluble surfactants insoluble
insoluble surfactants (20–50%) and surfactants
surfactants (20–40%), and hydrophilic (30–80%) and
(20–60%) hydrophilic cosolvents hydrophilic
cosolvents (0–40%) (20–50%) cosolvents
(0–50%)
Characteristics Nondispersing, SEDDS SEDDS/SMEDDS SMEDDS with Oil-free
requires without with water-soluble water-soluble formulation
digestion water- substances substances and low
soluble oil content
substances
Advantages Simple Less likely Forms a clear or Forms a clear and High solvent
solution, to lose almost clear solution transparent solution capacity for many
excellent solvent upon dispersion, upon dispersion, drugs, forms
capsule capacity drug absorption may drug absorption may micellar solution
compatibility upon occur without occur without upon dispersion
dispersion digestion digestion
Disadvantages Poor solvent Turbid Likely to lose solvent Likely to lose Likely to lose
capacity unless O/W capacity upon solvent capacity solvent capacity
drug is highly emulsion dispersion, less upon dispersion upon dispersion,
lipophilic (particle easily digested may not be
size digestible
0.25–2 μm)
Adapted with permission from Vithani et al. (2019b)
reservoirs from which drug molecules may dif- temperature. On the basis of fatty acids’ carbon
fuse into the aqueous phase and be absorbed. backbone length, triglycerides can be classified as
Drug loading in Type I formulations is limited short-chain (<6 carbon), medium-chain (6–12
by the drug’s lipophilicity due to the low solvent carbon), and long-chain (>12 carbon)
capacity of glycerides. In Type II formulations, triglycerides (Gibson 2007). Refined natural oils
water-insoluble surfactants are incorporated to are commonly used for the delivery of poorly
improve the emulsification of glycerides. Type soluble drugs. Refined natural oils are unsatu-
II formulations are considered self-emulsifying rated, long-chain triglycerides consisting of tri-
drug delivery systems (SEDDS), forming turbid, glyceride of fatty acids, small amount of free
coarse oil-in-water emulsions upon dispersion in fatty acids, and plant sterols. Most commonly
the GI tract in the size range of few microns. Type used refined natural oils are coconut oil, corn
III formulations are also SEDDS or self- oil, soybean oil, and sesame oil. These oils differ
microemulsifying drug delivery systems in the chain length of fatty acids, and the number
(SMEDDS) and are further divided into Type and position of unsaturated bonds. At the initial
IIIA or Type IIIB depending on the proportions stage of production for refined natural oils, crude
of lipid components to hydrophilic surfactants natural oils are obtained from the plant seeds by
and/or cosolvents. Type III formulations can rap- expression in a hydraulic press or by solvent
idly lose solvent capacity when dispersed due to extraction. The refining process includes several
the dissolution of the hydrophilic components. critical steps to ensure product quality, such as the
Additionally, the smaller droplet sizes formed removal of impurities such as free fatty acids and
allow for absorption without digestion but may phospholipids; decolorizing with activated car-
provide a greater surface area for lipid digestion. bon; and deodorization with steam. Medium-
Finally, Type IV formulations contain no oils and chain triglycerides are liquid at room temperature
instead are comprised of a mixture of hydrophilic and are chemically synthesized via
surfactants and cosolvents. These formulations re-esterification of glycerol with medium-chain
may be digestible depending on the type of sur- fatty acids (caprylic and capric fatty acids).
factant used. Type IV formulations typically have Medium-chain fatty acids are produced via the
a higher solvent capacity to solubilize the hydrolysis of fixed oils and the distillation pro-
intended dose but will rapidly lose solvent capac- cesses. Coconut oil is the primary source for
ity upon dispersion. Descriptions and examples of medium-chain fatty acid. Given their more polar
the excipients used in Type I, II, III, and IV nature, medium-chain triglycerides have better
formulations are provided in the following solvent capacity as compared to long-chain
subsections. triglycerides.
Mixed glycerides are manufactured through
transesterification or direct esterification. In the
7.2.1 Triglycerides and Mixed transesterification process, triglycerides react
Glycerides with glycerol under heating with alkaline
catalysts. Direct esterification takes places
The simplest LBFs, Type I, consist of drug between the selected fatty acids and glycerol.
dissolved in lipids such as triglycerides or mixed Mixed glycerides are mixtures of fatty acids,
mono- and diglycerides. Triglycerides are lipids glycerol, mono-, di-, and triglycerides. Due to
consisting of one glycerin and three fatty acid the nature of manufacturing processes,
molecules. On the basis of the C–C backbone compendial specifications for mixed triglycerides
composition, triglycerides can be classified as allow for variations in the composition. Formula-
saturated (no C¼C bond in C–C backbone) and tion scientists must pay close attention to the
unsaturated (C¼C bond in C-C backbone) compositional difference between the materials
triglycerides. Saturated triglycerides are also from different manufacturers and between differ-
called “fat,” which are solids at room ent batches from the same manufacturer.
7.2.2 Surfactants emulsions by localizing at the oil–water interface

(Gursoy and Benita 2004). Additionally, at low
In more complex LBFs (Type II–IV), hydrophilic concentrations, surfactants are present as
and lipophilic surfactants are utilized to improve monomers and adsorb at interfacial surfaces,
drug solubilization and stabilize emulsion reducing interfacial energy, which in turn reduces
droplets in LBFs. Due to their amphiphilic nature, the surface tension (Fig. 7.2c, d. The concentra-
surfactants can increase the solvent capacity of an tion and type of surfactant used can be adjusted to
LBF in order to dissolve the required dose of a form droplets of varying sizes.
hydrophobic drug, and this interaction is shown Surfactants can be classified by their molecular
in Fig. 7.2a. The polar heads and hydrophobic weight and their hydrophilic-lipophilic balance
tails (Fig. 7.2c, d) associate, making the polar (HLB), as shown in Fig. 7.3 (Bnyan et al. 2018).
hydrophilic heads face the aqueous media, while The low-molecular-weight surfactants are
the lipophilic tails repel and face inward categorized as cationic (i.e. positively charged),
(Fig. 7.2). After the dispersion of LBFs, anionic (i.e. negatively charged), zwitterionic,
surfactants can also stabilize oil and water also referred to as amphoteric, and nonionic
a b
Hydrophilic drug
Lipophilic drug
Amphipathic drug
Lipid molecule
Bilayer membrane
Aqueous core (Lipid, surfactant)
surfactant molecule
c d
e
Hydrophilic
(polar) head
Hydrophobic f
(non-polar) tail
Fig. 7.2 (a) The dispersion of drug molecules within the surfactant monomer with 2 hydrocarbon tails, (e) micelle
lipid-based vesicles, (b) comparison between conventional (surfactant assembly) in aqueous medium, (f) micelle in
liposome on right half and transfersome (elastic liposome) nonaqueous medium. (Adapted with permission from
left side, (c) surfactant monomer with one tail, (d) Bnyan et al. 2018)
Fig. 7.3 Diagram of surfactant classification based on molecular weight and hydrophilic–lipophilic balance. (Adapted
with permission from Bnyan et al. 2018)
surfactants. In general, nonionic surfactants are ethanol, polyethylene glycol, glycerin, and pro-
preferred in LBFs due to increased safety at high pylene glycol, which were described in greater
concentrations as compared to ionic surfactants. detail in a previous chapter.
A list of nonionic surfactants and their
corresponding HLB values is listed in Table 7.2.
For HLB classification, the surfactants that are 7.3 Characterization of Lipid-Based
more lipophilic (HLB values of 3–6) tend to Formulations
form water-in-oil (w/o) emulsions, which exhibit
a greater solubility increase in nonaqueous media. Even with the development of the Lipid Formula-
Furthermore, the surfactants that are more hydro- tion Classification System (LFCS) to describe
philic and water soluble (HLB values of 8–18) LBFs of varying composition, the ability to col-
tend to form oil-in-water (o/w) emulsions, which lectively compare different types of formulations
exhibit a greater solubility increase in aqueous has been hampered by a lack of standardized
media. Nevertheless, an HLB value between in vitro methods to characterize representative
7 and 9 is regarded as a wetting agent, exhibiting formulations (Pouton 2006). Established
both properties. Surfactants commonly used in compendial dissolution methods are inadequate
LBFs have an HLB value of 10 or higher due to for discriminating the performance of LBFs, and
the formation of o/w emulsions or even with the adoption of in vitro dispersion-
microemulsions. digestion tests, the experimental parameters
(e.g., test pH, fluid volume, concentrations of
bile salt, calcium, and buffer agents) often varied
7.2.3 Hydrophilic Cosolvents greatly among laboratories. In an effort to better
understand the effect of experimental variables
In Type IIIB and IV formulations, hydrophilic and standardize in vitro test methods for LBFs,
cosolvents are often added to aid in solubilization the Lipid Classification System Consortium
of drugs that are hydrophobic, but not necessarily (LFCS Consortium) was formed in 2009 to define
lipophilic (Pouton 2006). Common cosolvents and communicate standardized testing methods
acceptable for oral dosage forms are limited to (Williams et al. 2012b).
Table 7.2 Pharmaceutical nonionic surfactants

Nonproprietary name HLB Value
Glyceryl Monooleate 3.3
Poloxamer 181, 182, 331 1–7
Poloxamer 124 12–18
Poloxamer 407 18–23
Poloxamer 188, 237, 338 >24
Polyoxyethylene alkyl ether (Brij L4) 9.7
Polyoxyethylene castor oil derivative 35 12–14
Polysorbate 20 16.7
Polysorbate 21 13.3
Polysorbate 40 15.6
Polysorbate 60 14.9
Polysorbate 61 9.6
Polysorbate 65 10.5
Polysorbate 80 15.0
Polysorbate 81 10.0
Polysorbate 85 11.0
Polysorbate 120 14.9
Propylene glycol Monolaurate 4.5–5
Sorbitan monostearate 4.7
Sorbitan monolaurate 8.6
Sorbitan monooleate 4.3
Sorbitan monopalmitate 6.7
Sorbitan monostearate 4.7
Sorbitan sesquioleate 3.7
Sorbitan trioleate 1.8
Sorbitan tristearate 2.1
Tricaprylin 7.0
Vitamin E polyethylene glycol succinate 13.2
To date, much of the work of the LFCS Con- but also provides standard methods to evaluate
sortium has focused on understanding and saturation solubility within the formulation, upon
establishing discriminating methods for the con- dispersing in aqueous media, and within digested
duct of in vitro dispersion and digestion tests. media to understand the dynamics of solvent
Eight exemplary LBFs were defined to conduct capacity and its impact on drug solubilization or
the studies as shown in Fig. 7.4, which spanned precipitation.
Type I through Type IV LBFs with Type I, II, and
IIIA formulations divided into examples
containing either long-chain (LC) or medium- 7.3.1 Solubility
chain (MC) triglycerides. Formulations were
developed based on their capacity to self- One of the first steps in developing an LBF is to
emulsify, digestibility, drug solvency, and parti- generate a formulation that contains enough sol-
cle size on dispersion. The appearance, particle vent capacity to dissolve the required dose of the
size, and polydispersity of the selected drug substance. In order to determine saturation
formulations are shown in Table 7.3. While solubility, standard shake-flask methods have
refinements to methods will continue over time, been used with an excess of drug substance
the work to date not only establishes baseline added to a model LBF and saturation determined
parameters for dispersions and digestion testing, after consecutive samples vary by less than 5% in
Type I Type II Type IIIA Type IIIB Type IV
35% 35% 25%

100% 65% 65% 50% 50% 50%
25%
LC & MC LC & MC LC & MC MC

Lipids:
Long-chain = Corn oil and Maisine™ 35-1 (in a 1:1 ratio)
Medium-chain = Captex® 300 and Capmul® MCM EP (in a 1:1 ratio)
Lipophilic surfactant: Tween® 85
Hydrophilic surfactant: Cremophor® EL
Cosolvent: Transcutol HP
Fig. 7.4 Exemplary lipid-based formulations utilized by or medium-chain (MC) triglycerides within their composi-
the Lipid Formulation Classification System Consortium. tion, whereas Type IIIB only consists of a composition
Type I, II, and IIIA formulations are represented by that contains medium-chain triglycerides. (Adapted with
versions that contain either long-chain (LC) triglycerides permission from Williams et al. 2013a)
Table 7.3 Physical Properties of the LFCS Consortium Exemplary LBFs After Dispersion in Digestion Media
Formulation Type Appearance on Dispersiona Particle Sizeb (nm) Polydispersityb
I-MC Opaque emulsion n/ac –
II-MC Opaque emulsion 1 in 36: n/ac 0.068 0.05
1 in 250: 190.2 2.7
IIIA-MC Ultrafine dispersion 1 in 36: 29.1 0.6 0.062 0.01
1 in 250: 36.2 0.9 0.088 0.02
IIIB-MC Ultrafine dispersion 1 in 36: 21.4 0.5 0.111 0.01
1 in 250: 14.0 2.7 0.138 0.10
I-LC Coarse emulsion n/ac –
II-LC Coarse–opaque emulsion n/ac –
IIIA-LC Ultrafine dispersion 1 in 36: 61.8 0.4 0.197 0.01
1 in 250: 58.4 1.3 0.149 0.01
IV Transparent solution 1 in 36: 13.1 0.4 0.033 0.01
1 in 250: 9.9 0.5 0.111 0.06
Adapted with permission from Williams et al. (2012b)
a
Appearance following dispersion of 1 g LBF in 36 mL digestion medium (37 C)
b
Mean particle size of the dispersed LBF following a 1 in 36 and 1 in 250 dilution in the digestion medium (37 C).
Values are expressed as means (n ¼ 3) 1 SD
c
Colloids formed on dispersion were too large to be measured accurately (i.e. polydispersity >0.5)
LC long chain, MC medium chain
concentration (Williams et al. 2012b). Depending (>50 mg/g) only in the more polar Type IIIB
on the properties of the drug substance and the and Type IV formulations as compared to
desired dose, the type of LBF employed may be fenofibrate, which has relatively high solubility
limited based on the relative lipophilicity or across all types of LBFs and low variability in
hydrophilicity of the formulation. In Fig. 7.5, for solubility between formulations in which MC
example, danazol has relatively high solubility lipids are used (Williams et al. 2012b; Williams
Danazol
80
Equilibrium solubility (mg/g)

65.5
H
O
60
H 48.7
N H H
40
O 27.2 28.3
24.8
19.0
molecular weight: 337.5 20 16.9
melting temperature: 224.2°C 8.4
log P: 4.5
solubility in water: 18 μg/ml
0
C
LC
C
LC
III C
C
IV
I-L
I-M
M
II-
A-
II-
A-
B-
III
III
200 189.1
Fenofibrate
Equilibrium solubility (mg/g) 149.6 152.1
O 143.1
150 135.6
O 106.8
Cl O 98.1 99.9
100
O
molecular weight: 360.8

50
melting temperature: 81.1°C
log P: 5.24
solubility in water: 0.3 μg/ml
0
C
LC
LC
III C
C
IV
I-L
I-M
M
II-
A-
II-
A-
B-
III
III
100
Tolfenamic acid
77.0
HO O 80
H
N Cl
54.8
60
40 32.6
molecular weight: 261.7 26.0 25.9
melting temperature: 212.1°C 20.6 17.3
log P: 5.7 20
solubility in water: <0.1 μg/ml 8.8
pKa (37°C): 4.1

0
C
LC
LC
III C
C
IV
I-L
I-M
M
II-
A-
II-
A-
B-
III
III
Fig. 7.5 Equilibrium solubility values for danazol, fenofibrate, and tolfenamic acid in the eight LBFs investigated.
(Adapted with permission from Williams et al. 2012b; 2013a)
et al. 2013a). Furthermore, the different salt and example, Tay et al. investigated the change in
ionic forms of a drug exhibit different solubilities solubility of different salt and ionic forms of
depending on the LBF classification. For lumefantrine (Fig. 7.6). The free base exhibited
Fig. 7.6 Equilibrium solubility of lumefantrine IV formulation, and the data shown reflect the measured
compounds in lipid-based formulations. Data are concentration after was miscible in the Type IV formula-
expressed in lumefantrine free base equivalents. Data are tion, and the data shown reflect the measured concentra-
mean SD, n ¼ 3. * Lumefantrine docusate expressed in tion after mixing the docusate and formulation in a 1:1
lumefantrine free base equivalents. Data are mean SD, w/w ratio. Mixing the docusate and formulation in a 1:1
n ¼ 3. * Lumefantrine docusate was miscible in the Type w/w ratio. (Adapted with permission from Tay et al. 2019)
higher solubility in the more lipophilic digestion and potentially the precipitation behav-
formulations (i.e. Types I and II), while the salts ior of the drug (Williams et al. 2012a).
and ionic liquids exhibited higher solubilities in When LBFs are dispersed into aqueous media,
the more hydrophilic formulations (i.e. Types III a drop in solvent capacity can occur as the con-
and IV). Additionally, all formulations achieved centration of water increases. In order to quantify
higher solubilities when medium-chained lipids the maximum effect of water concentration on
were used compared to long-chain lipids (Tay solvent capacity, a phase diagram including
et al. 2019). water within the formulation can be constructed
The importance of determining saturation sol- and solubility measurements can be taken with
ubility within the LBF, rather than individual progressively increasing concentrations of water.
components, is highlighted in Fig. 7.7. In this This is illustrated in Fig. 7.8 with four
example, while some general trends in solubility formulations of varying medium-chain mono-
are observed for danazol in individual glyceride and polysorbate 80 ratios. As the con-
components (e.g., higher solubility in polar centration of water increases, a phase transition
excipients), it was not possible to calculate or occurs between a continuous oil phase to a
predict the solubility of danazol in the LBFs bicontinuous phase of oil and water, and the ther-
based on either solubility of individual modynamic impact at each of these water
components or the mass ratio of excipients within concentrations on the solubility of a lipophilic
the formulation (Williams et al. 2012b). Deter- drug can be measured. In this case, the solubility
mining the saturation solubility for a given LBF of the drug dropped by half with only 10% water
serves as a guide to the drug loading, or percent addition (Pouton 2006). Such an exercise serves
saturation, that will be necessary to dissolve the to quantify the thermodynamic impact of water
intended dose. The drug loading or degree of on drug equilibrium solubility, but this may only
saturation in the LBF can have a significant serve as a worse case depending on the kinetics of
impact on the solvent capacity after dispersion/ the dispersion process and drug precipitation.
Fig. 7.7 Equilibrium 80 71.8

solubility of danazol in lipid
excipients used in the

compositions of the eight
LFCS Consortium LBFs. 60
from Williams et al. 2012b)
40
27.9
23.0
19.1
20
9.3
6.3
2.8
0
85
l
-1
P
oi
30
rE
H
C
35
n
ol
or
ho
ee
x
™
te
ut
ul
C
op
Tw
ne
ap
sc
m
m
ap
an
si
C
re
ai
Tr
M
Fig. 7.8 Equilibrium solubility of a lipophilic drug in medium-chain monoglyceride (MCMG) and polysorbate 80 (P80)
formulations with increasing water content. (Adapted with permission from Pouton 2006)
Additionally, this exercise fails to take into conducted in a dispersed aqueous phase (APDISP)
account the effect of digestion products and the and a digested aqueous phase (APDIGEST)
impact of increasing bile-salt concentrations on (Williams et al. 2013a). For these studies, drug-
drug solubility. free LBFs should undergo standardized disper-
In order to quantify the impact of not only sion and digestion testing under the same
dilution, but the effects of colloid stability due conditions as active LBFs (standardized digestion
to digestion and the presence of bile salts on parameters defined by the LFCS Consortium will
solvent capacity, solubility studies can be be discussed in the next subsection). For
solubility determination in APDISP, media digestion are monitored for a period of 30 to

samples can be taken after the drug-free LBF 60 min (Sek et al. 2002). The effect of gastric
has been dispersed for a determined time period lipases on lipid digestion (Carriere et al. 1993;
and added to a container containing an excess of Lengsfeld et al. 2006; Fernandez et al. 2007) is
crystalline drug that is incubated at 37 C. For currently neglected assuming that the pH is too
solubility determination in APDIGEST, samples of low in fasted-state conditions to support lipase
digestion medium can be taken after a activity (Bakala-N'Goma et al. 2015). Addition-
predetermined time period, and digestion can be ally, a commercial source for gastric lipase does
halted with the addition of a digestion inhibitor not exist, practically limiting its use. In order to
(1.0 M 4-bromophenylboronic acid in methanol). standardize digestion test parameters representing
The APDIGEST media can then be added to a the fasted state in the small intestine, the impact of
container containing excess crystalline drug and several experimental parameters including pH,
incubated at 37 C. Samples at select time bile-salt concentration, separation methods, drug
intervals can then be taken, separated by centrifu- type, drug loading, and pancreatin concentration
gation, and analyzed for equilibration of drug on LBF performance has been evaluated by the
content. For certain LBFs, a drop in solubility LFCS Consortium.
may occur over time if the colloidal phase is The developed digestion test method utilizes a
unstable and solubility values at earlier time pH-stat apparatus. This allows for maintenance of
points prior to the solubility loss should be the media pH over the course of the experiment as
utilized (Williams et al. 2012a). Knowledge of well as the ability to quantify the extent of lipid
drug solubility in APDISP and APDIGEST is useful digestion by pancreatic lipases. As lipids are
to understand the impact of loss of solvent digested during the experiment, free fatty acids
capacity and propensity for drug precipitation are liberated, and the pH of the media is con-
given the amount of drug present in an active stantly monitored and maintained by titration
LBF and the volume of digestion medium, with a concentrated sodium hydroxide buffer to
which will be discussed further in the following the reaction vessel. The amount of sodium
section. hydroxide added during the experiment will
reflect the amount of ionized free fatty acids
liberated during digestion. Since not all of the
7.3.2 Dispersion and Digestion fatty acids may be ionized at the test pH (e.g.,
Testing pH ¼ 6.5), additional amount of sodium hydrox-
ide to adjust the media pH to 9 at the end of the
The primary focus of the LFCS Consortium has experiment, with subtraction of the amount of
been to develop standardized in vitro digestion titrant needed in the absence of lipids, allows for
tests that can discriminate between LBFs. Diges- calculation of the unionized fatty acids liberated
tion test methods have been developed mimicking during digestion testing. The summation of the
the fasted-state behavior in the small intestines ionized and unionized fatty acids and knowledge
that includes a 10-min dispersion (dilution) step of the theoretical maximum amount of fatty acids
in buffered media containing bile salts and in the LBF provides an estimation of the extent of
phospholipids, followed by the addition of pan- digestion during the experiment according to
creatic lipases where drug solubilization and lipid Eq. 1 (Williams et al. 2012b).
ionized fatty acid þ unionized fatty acid

Extent of Digestion ¼ ð7:1Þ
theoretical maximum fatty acids in LBF
Fig. 7.9 Apparent titration Extent of

A
of fatty acids (FAs) released 3.0 Digestion
and estimated extent of
digestion during in vitro
Titrated fatty acid (mmol)

digestion of (a) medium- I-MC 90%
chain lipid-based
formulations (LBFs) and 2.0 IIIA-MC 90-105%
(b) long-chain LBFs. II-MC 82-100%
Digestion was initiated at
t ¼ 0 min on addition of
pancreatin, and pH was
maintained constant at 1.0
pH 6.5. Titrated FA was IIIB-MC 71-117%
detected by pH titration,
and values are expressed as
means (n ¼ 3) 1 SD and
have been corrected for the 0.0
level of FA released in 0 10 20 30
background digestion tests
(no lipid formulation). Time (min)
(Adapted with permission B
from Williams et al. 2012b) 0.4
IIIA-LC 67-95%
Titrated fatty acid (mmol)
0.2
II-LC 27-37%
I-LC 25%
IV 15%
0.0
0 10 20 30
Time (min)
The titrated fatty acid concentration and the can have a significant impact on drug solvent
extent of digestion estimated by Eq. 1 for the capacity.
eight exemplary formulations evaluated by the With the ability to estimate the extent of diges-
LFCS Consortium are shown in Fig. 7.9 tion in situ, experimental parameters can be
(Williams et al. 2012b). It can be observed that varied and their impact on digestion and LBF
formulations containing medium-chain performance can be assessed. Initial baseline
triglycerides are digested to a much greater extent parameters for an in vitro digestion test were
compared to formulations containing long-chain determined based on previous experiments and
triglycerides. Additionally, LBFs containing low biorelevance to conditions in the small intestine.
levels of lipids (e.g., Type IV) result in minimal Cross-site reproducibility was demonstrated for
digestion reflecting the small concentration of the eight-model LBFs, and the centrifugation
digestible surfactants in these formulations. method was initially refined (Williams et al.
Finally, it is evident that a great extent of diges- 2012b). Baseline conditions were set at a pH of
tion occurs in the first few minutes after the 6.5 using a 2 mM Tris-maleate buffer with
lipases are present, reflecting a rapid process that 150 mM NaCl, 1.4 mM CaCl2-2H2O, 3 mM
Table 7.4 Proposed LFCS consortium fasted-state digestion testing Parameters

Digestion condition
Fasted
Test preparation
LBF mass (g) 1.083
pH of digestion medium 6.5
Buffer concentration: (Tris-maleate with NaCl) (mM) 2 mM/150 mM
CaCl2 concentration 1.4 mM
Bile salt: Phospholipid concentration (mM) 3:0.75
Digestion medium volume (mL) 39
Dispersion phase
Duration 10 min
Samples 3 1 mL
Digestion phase
Pancreatin (~1000 TBU/mL) 4 mL added after 10-min dispersion
Duration 30–60 min
Samples 4 1 mL for benchtop centrifugation
2 4 mL for ultracentrifugation
sodium taurodeoxycholate (NaTDC), and altered solvent capacity. Given an increase in

0.75 mM phosphatidylcholine (PC) based on calcium concentration may overestimate drug
reported in vivo conditions. Drug separation precipitation, physiological levels of 1.4 mM are
based on ultracentrifugation and standard centri- recommended. Based on the studies to date, the
fugation was compared, and it was recommended proposed standardized parameters for fasted state
that formulations that produce an oil-rich phase in vitro digestion testing of LBFs are listed in
utilize ultracentrifugation. Further studies Table 7.4.
evaluated the impact of bile-salt concentration In addition to LBF formulation composition
(e.g., 0, 3, 5, and 10 mM NaTDC), pancreatin and digestion testing parameters, three model
concentration (e.g., 150, 300, 600, and 900 USP drugs have been evaluated using the proposed
U/mL), and calcium concentration (0, 1.4, 5, and digestion method including danazol (neutral),
10 mM) on digestion (Sassene et al. 2014; fenofibrate (neutral), and tolfenamic acid (weak
Williams et al. 2012a). Results indicated that the acid) (Williams et al. 2013a). The type of drug
bile-salt concentration had only moderate effects used and drug loading had little impact on digest-
above a concentration of 3 mM, thus this concen- ibility, but drug loading did have a significant
tration was recommended. Increasing pancreatin effect on drug precipitation. As drug loading is
concentration influenced the extent of digestion increased, the extent of drug precipitation increased
for both medium-chain and long-chain LBFs, but across all formulation types. However, a general
had little effect on drug distribution; thus, physi- trend was realized across all model drugs and LBF
ological quantities were recommended. Calcium types when the maximum drug concentration in the
had a smaller effect on digestion, but generally digested aqueous phase (APMAX) was normalized
influenced drug distribution for Type IIIA-LC, to the thermodynamic solubility within APDIGEST.
I-MC, II-MC, and IIIA-MC LBFs due to the This value was termed the maximum supersatura-
formation of free fatty acid-calcium soaps that tion ratio, SRM, as shown in eq. 2.
Maximum Drug Concentration in APDIGEST ðAPMAX Þ

SRM ¼ ð7:2Þ
Drug Solubility in APDIGEST
Fig. 7.10 (continued)
in such cases. Similar trends have been observed

in LBFs containing other types of surfactants in
which an SRM limit of ~2.6 to 2.7 was observed
in 6 of the 7 LBFs evaluated (Devraj et al. 2014).
This realization of sustained drug supersaturation
at moderate supersaturation levels has contributed
to the evolution that digestion-triggered supersat-
uration is a predominant mechanism of enhanced
absorption by LBFs (Anby et al. 2012) and
highlights the importance of developing discrimi-
nating in vitro methods to monitor drug supersat-
uration in biorelevant conditions.
Alskär et al. demonstrated that acidic and neu-
tral drugs exhibit a higher solubility in the disper-
Fig. 7.10 Three-dimensional surface plots to illustrate sion media compared to the digestion media;
the relationship between SRM, APMAX, and drug solubility
in the APDIGEST for (a) danazol, (b) fenofibrate, and (c)
whereas weakly basic drugs exhibit higher solu-
tolfenamic acid. SRM is the maximum degree of supersat- bility in the digestion media compared to the
uration that is attained following 30–60-min in vitro diges- dispersion media. The improved solubility for
tion of a lipid-based formulation. An SRM threshold of ~3 basic drugs is associated with favorable electro-
was observed for all drugs. (Adapted with permission from
Williams et al. 2012b, 2013a)
static interactions between the charged weakly
basic drug and the charged free fatty acid. The
authors concluded that for the weakly basic
Applying the SRM equation, it was observed that compounds studied, lipolysis-triggered precipita-
drug precipitation occurred when this value tion was unlikely due to the high equilibrium
exceeded ~2.5 to 3 as shown in Fig. 7.10 for solubility of the weakly basic drugs in the diges-
danazol, fenofibrate, and to tolfenamic acid. tion media (Alskar et al. 2018). However, similar
This suggests that after dispersion/digestion of to the findings of Williams et al., when the weakly
an LBF, a supersaturated state of the drug may acidic and neutral compounds underwent disper-
be sustained when the degree of supersaturation sion, rapid supersaturation occurred, and in all
does not exceed 3x. Once this level is exceeded, cases following the digestion phase drug precipi-
the driving force for drug precipitation becomes tation occurred (e.g., between 50 and 70% recrys-
too great, and drug crystallization was confirmed tallization, Fig. 7.11).
Fig. 7.11 Lipolysis profiles of drug-loaded type IIIB-MC (black dotted line). Even though drug precipitation
for danazol, griseofulvin, felodipine, indomethacin, and occurred, all drugs remained above the solubility value in
niclosamide. Drug concentration in the aqueous phase the digestion medium, and hence stayed supersaturated
(blue dots) and the pellet phase, i.e., precipitated drug throughout the lipolysis. (Adapted with permission from
(orange squares) during lipolysis. Thermodynamic drug Alskar et al. 2018)
solubility at 10 min of dispersion and 60 min of digestion
To further develop discriminating in vitro elucidate the performance of different types of

digestion tests, “stressed” digestion testing LBFs. In addition to the proposed stressed diges-
conditions have been proposed by the LFCS Con- tion conditions, an A-D performance classifica-
sortium along with a new performance classification system was proposed based on a passing or
tion system (Williams et al. 2014a). The stressed failed performance in 1) dispersion testing, 2)
digestion test is intended to follow standard fasted-state digestion testing, and 3) stressed
fasted-state digestion testing in order to further digestion testing as shown in Fig. 7.12. An LBF
progress lead LBF candidates. The stressed diges- that passed all three in vitro tests would receive an
tion conditions can be utilized to alter SRM in A grade, whereas one that failed dispersion test-
APDIGEST by altering the digestion or solvent ing would receive a D grade.
capacity properties. Some example stressed In order to develop LBFs that can sustain
digestion conditions are shown in Table 7.5 and supersaturation at SRM values greater than ~3,
compared to the standard fasted conditions. For polymeric precipitation inhibitors (PPIs) have
example, decreasing the amount of LBF or been investigated as an additional formulation
increasing the concentration of bile salts can component. A variety of polymers have been
impact the solvent capacity in APDIGEST by shown to be effective precipitation inhibitors in
greater aqueous dilution or formation of bile-salt a wide range of studies, particularly in amorphous
micelles. A lower LBF mass was shown to be a solid dispersions (Konno et al. 2008; Ueda et al.
stressor particularly for medium-chain-rich LBFs 2013; Warren et al. 2010). Gao and his colleagues
due to increased dilution. Adjustments of media were the first to investigate PPIs’ effect to main-
pH can additionally be utilized for ionizable tain supersaturation and enhance absorption of
drugs that have pH-dependent solubility to further poorly soluble drugs from supersaturable
Table 7.5 Proposed LFCS consortium fasted-state and stressed digestion testing parameters
Digestion Condition (Fasted + Stressed)
Fasted #LBF "BS #pH
Test preparation
LBF mass (g) 1.083 0.173 1.083 1.083
pH of digestion medium 6.5 6.5 6.5 4.5
BS:PL concentration (mM) 3:0.75 3:0.75 10:2.5 3:0.75
Digestion medium volume (mL) 39
Dispersion phase
Duration 10 min
Samples 3 1 mL
Digestion phase
Pancreatin 4 mL added after 10-min dispersion
Duration 30–60 min
Samples 4 1 mL for benchtop centrifugation
2 4 mL for ultracentrifugation
Adapted with permission from Williams et al. (2014a)
Fig. 7.12 A proposed new LBF performance classification based on in vitro evaluations. (Adapted with permission
from Williams et al. 2014a)
SEDDS (S-SEDDS) formulation approaches in digestion tests, which further confirmed an

(Gao et al. 2004; Gao et al. 2003). The mecha- increase in SRM values up to ~8 in some cases,
nism of such polymeric PPIs in maintaining as shown in Fig. 7.14 (Anby et al. 2012). How-
highly supersaturated state was also investigated ever, it was found that in certain cases, the pre-
and hypothesized by Gao et al. (Gao et al. 2009). cipitation inhibition properties were
As shown in Fig. 7.13, the inclusion of 5% overestimated in vitro compared to in vivo analy-
HPMC resulted in a significant increase in pacli- sis. Still, the inclusion of PPIs to LBFs is an
taxel supersaturation in a simple dispersion test as additional important formulation variable in
compared to lipid formulations without PPIs, and order to achieve supersaturation and enhance the
this resulted in enhanced absorption in vivo in oral absorption of poorly water-soluble drugs.
rats. The impact of PPIs has also been evaluated
Fig. 7.13 a) Apparent in vitro paclitaxel concentration– rats after oral administration of 10-mg/kg paclitaxel in rats.
time profiles in simulated gastric fluid from the SEDDS (Adapted with permission from Gao et al. 2003; Williams
formulations (S-SEDDS) with and without HPMC and b) et al. 2013b)
Mean plasma concentration–time profiles of paclitaxel in
A B C
100 100 100
% Drug solubilised
80 80 80
60 60 60
40 40 40
20 20 20
0 0 0
0 2 4 6 8 0 2 4 6 8 0 2 4 6 8
SMDISP SMDIGEST SMDISP or DIGEST
Fig. 7.14 Percentage drug solubilized versus SM at 40% presence (open) and absence (closed) of PPI. Arrow
(orange) and 80% (black) on (a) dispersion and (b) diges- illustrates the effect of PPI on percentage of solubilized
tion. Panel C shows percent drug solubilized versus drug at constant SMDISP&DIGEST. (Reproduced with per-
SMDISP&DIGEST at 40% (orange) and 80% (black) in the mission from Anby et al. 2012)
Although HPMC is regarded as one of the 7.3.3 In Vivo Characterization

most effective PPIs, HPMC is not soluble in
LBFs and forms a two-phase system; therefore, The discriminating power of in vitro digestion test
Suys et al. studied the impact of various PPIs that methods for LBFs with direct relevance to in vivo
can be dissolved within LBFs while still performance has been an active research subject.
preventing precipitation for the drug fenofibrate Dahan and Hoffman investigated the distribution
(Figure 7.15a). The authors revealed that there and solubilization pattern of griseofulvin across
was no significant difference in the PPIs’ ability different phases of an in vitro digestion model
to maintain supersaturation throughout dispersion resulting from LCT, MCT, and SCT-based lipid
and digestion regardless of its miscibility with the systems compared to in vivo evaluations of these
LBF (Figure 7.15b), indicating all PPI selections formulations (Dahan and Hoffman 2007). The
should be extensively screened for each composi- in vitro lipolysis results showed significant
tion (Suys et al. 2018). differences between these formulations. MCT
Fig. 7.15 In vitro evaluation of the impact of the pres- polymer (also shown as the black (control) bar). Data
ence of PPIs predispersed in different media on fenofibrate above the line represent formulations where polymer addi-
solubilization and supersaturation after dispersion tion leads to lower drug precipitation during formulation
(15 min) and digestion (60 min) in a Type IV formulation dispersion and digestion (and therefore greater drug solu-
at 85% saturated solubility [mean SD (n 3)]. Panel A: bilization). The green panel highlights the area where the
It shows the % drug solubilized over the period of formu- % solubilized fenofibrate exceeds that of the formulation
lation dispersion and digestion, calculated by dividing the without polymer (control). Panel B: Mean supersaturation
AUC of the concentration versus time profile obtained in ratio (S) obtained over the period of dispersion and diges-
the in vitro test by the AUC of the APMAX over the same tion of the formulations; data grouped for polymers with
time period (representing 100% solubilized). The horizon- different solubility profiles in the LBF. (Adapted with
tal black dotted line represents the extent of drug solubili- permission from Suys et al. 2018)
zation obtained from the formulation in the absence of
lipid led to high amounts of solubilized drug in digestion testing, the Type IIIA formulations
the aqueous phase (>95%), while the LCT lipid exhibited significantly less precipitation com-
resulted in a smaller amount, and the SCT proved pared to the Type IIIB/IV formulation. When
the lowest (Figure 7.16a). All three lipid dosed orally in pigs, each of the formulations
formulations proved to be advantageous over the performed similarly, demonstrating that the
control, an aqueous suspension. A similar trend digestion test was over discriminating the LBFs,
was confirmed by the in vivo PK study. As shown as shown in Fig. 7.17. A similar overdiscri-
in Figure 7.16b, the exposure of griseofulvin was minating power was observed using the digestion
directly related to the nature of the lipid compo- testing protocol proposed by the LFCS Consor-
nent, showing the rank order MCT > LCT > SCT tium in LBFs containing PPIs (Anby et al. 2012).
> water. An excellent correlation (Figure 7.16c) Here, it was found that the digestion test failed to
between the percentage of griseofulvin dose that predict the in vivo absorption in dogs, with no
was solubilized in the aqueous phase of the observable difference between LBFs with and
in vitro lipolysis medium and the in vivo AUC without PPIs at high drug loadings, despite a
values upon oral administration of these four difference observed in digestion testing.
formulations was observed. There are a number of reasons that the in vitro
For other researchers, establishing an in vitro– digestion tests may overdiscriminate the in vivo
in vivo correlation between digestion tests and PK performance of LBFs. One primary reason that
studies has proved elusive (Griffin et al. 2014). has been discussed is the lack of an absorptive
When comparing in vitro tests of varying com- sink in the digestion test, which may drive super-
plexity (i.e. dilution, dispersion, and digestion), saturation ratios to higher levels than what would
Griffin and colleagues found that solubility was be observed in vivo. Some researchers have had
similar among three LBFs (Types IIIA-LC, IIIA- success correlating dispersion methods that con-
MC, and IIIB/IV) containing fenofibrate in the tain a sink compartment to in vivo data for
dilution and dispersion tests, whereas during supersaturating lipid systems. For example, Shi
a b
100 2.0
Aqueous phase
Sediment phase LCT
80 Lipid phase MCT
Griseofulvin conc. (μg/ml)

% of griseofulvin dose
1.5 SCT
H2O
60
1.0
40
0.5
20
0 0.0
MCT LCT SCT H2O 0 5 10 15 20 25
c Time (hr)
Griseofulvin in vivo AUC values (μg-hr/ml)
20
MCT
16
LCT
12
8
SCT
4 H2O
0
0 20 40 60 80 100
% of Griseofulvin dose in the in vitro aqueous phase
Fig. 7.16 (a) Distribution of griseofulvin in the aqueous IVIVC between solubilized griseofulvin in the aqueous
phase, sediment, and lipid phase, (b) Plasma phase upon in vitro lipolysis and in vivo AUC. (Adapted
concentration–time profiles of griseofulvin, and (c) with permission from Dahan and Hoffman 2007)
and coworkers evaluated a novel in vitro biphasic Other notable in vivo/ex vivo studies with
dissolution test that involved both aqueous and LBFs have helped to elucidate the precise
organic phases. The drug is released from LBFs mechanisms of enhanced absorption. Given that
in the biorelevant aqueous media under nonsink enhanced solubilization in lipid micelles by LBFs
condition and partitions into the organic phase may have a negative impact on absorption by
(e.g., octanol) under sink condition. They decreasing the free drug concentration, two alter-
correlated the drug concentration in the octanol native absorption mechanisms were evaluated by
phase to in vivo human data with promising Yeap and colleagues (Yeap et al. 2013): (1) direct
results, as shown in Fig. 7.18 (Shi et al. 2010). transfer of drug from the colloidal lipid phase to
Adding an organic phase as an “absorptive com- the absorptive membrane by collisional transfer,
partment” to the current in vitro digestion test and (2) enhanced flux through the absorptive
may be a future option, though this adds signifi- membrane due to a transient, supersaturated
cant complexity to the in vitro test, and this com- state of the drug. A single-pass rat jejunum perfu-
plexity needs to be weighed against the sion model was used to investigate the two
probability of false positives in lipid formulation mechanisms utilizing the poorly soluble drug
development. cinnarizine. In this study, collisional transport
Fig. 7.17 Comparison of in vitro dispersion tests to tests were more predictive to in vivo bioavailability data in
in vitro digestion tests and their predictability to in vivo pigs compared to the more complex digestion tests.
data. In this case with fenofibrate, the simpler dispersion (Adapted with permission from Griffin et al. 2014)
was shown to not play a significant role in absorp- processes (e.g., soft/hard shell capsule filling
tion; however, bile dilution-induced drug super- and banding), potential for volume reduction,
saturation was shown to have a significant more precise dosing, and improved patient com-
positive effect on absorption. For colloidal pliance. A number of processing techniques are
systems, supersaturation was maintained for a available for solidification of LBFs as shown in
long enough period to promote enhanced absorp- Fig. 7.20, which has been the topic of several
tion, but vesicular systems promoted rapid pre- recent reviews (Dening et al. 2015; Jannin et al.
cipitation and failed to enhance oral absorption. 2008; Tan et al. 2013). Among the available
The bile-mediated solubilization and supersatura- techniques, physical adsorption on to materials
tion pathways are illustrated in Fig. 7.19. such as fumed silica and magnesium aluminum
silicate has been the most extensively studied.
Many researchers have reported favorable mate-
rial properties (e.g., flow, compaction) at high
7.4 Solid Lipid-Based
lipid loadings in adsorptions systems (Agarwal
Formulations—A Current
et al. 2009; Hansen et al. 2004). Alternatively,
Trend in Lipid-Based
other researchers have evaluated the used of syn-
Formulation Design
thetic poloxamers as emulsifying and solidifying
agents with lipid-based components via melt
The solidification of LBFs continues to be an
mixing methods (Shah and Serajuddin 2012;
active area of pharmaceutical research. Solidifica-
Tran et al. 2013). These types of formulations
tion can offer several advantages over liquid-
have performed favorably in vivo, with
based formulations including enhanced stability,
poloxamer potentially having a dual role as a
elimination of liquid-based manufacturing
Media USP II with dual paddle
Drug Organic layer

product
Pump
Aqueous layer
Returned
Aqueous Layer Media
Organic Layer
0.16 0.5
CEB in aq. phase (mg/mL)
CEB in octanol (mg/mL))

S-SEDDS
S-SEDDS 0.4 Solution
0.12 Solution Celebrex
Celebrex 0.3
0.08
0.2
0.04
0.1
0.00
0.0
0 30 60 90 120 0 30 60 90 120
Time (min) Time (min)
170
160
% Relative AUC (in vivo)
150
in vivo AUC
140 S-SEDDS
S-SEDDS
130
120
110 Capsule y = 71.92x + 86.07

100
Capsule Solution Suspension R2 = 0.97
80 90
0.0 1.0 2.0 3.0 4.0 0.0 0.2 0.4 0.6 0.8 1.0
In vitro AUC in vitro AUC
Fig. 7.18 A biphasic dissolution test comparing a human PK AUC, as compared to the right panel where a
supersaturable SEDDS formulation to a solution and good correlation is observed for the organic phase AUC
marketed formulation. In the left panel, poor correlation and human PK AUC. (Adapted with permission from Shi
is observed in the aqueous phase AUC relative to the et al. 2010)
precipitation inhibitor. While solidification of solid formulations were compared both in vitro
LBFs has shown some success and offers some and in vivo (Van Speybroeck et al. 2012). As
notable advantages in terms of stability and shown in Fig. 7.21, when the formulations were
manufacturability, these advantages need to be compared in dispersion and digestion tests, an
weighed against their performance in vivo com- ~35% decrease in the dose was solubilized in
pared to their liquid counterparts. the solid formulations as compared to the liquid
Recently, direct comparisons between liquid counterparts. Furthermore, when these
and adsorption-based solid LBFs on in vitro and formulations were administered to rats, the bio-
in vivo performance have been performed. In one availability of the solid formulations was approx-
study, two self-emulsifying LBFs (one containing imately half of the liquid formulations.
MC lipids and one containing LC lipids) were Investigations of the poor performance from the
adsorbed onto Neusilin US2, and the liquid and solid formulations were attributed to incomplete
Fig. 7.19 The dual role of bile during lipid digestion and supersaturation and enhances drug absorption by increas-
dispersion. (i) Bile-mediated solubilization of lipid diges- ing drug thermodynamic activity in colloidal phases. D
tion products. (ii) The continuing interaction of secreted represents the free concentration of drug available for
bile with existing lipid colloidal phases in the lumen absorption. Dss is used to signify the increase in free
results in progressively less lipid-rich phases with lowered concentration resulting from bile-mediated supersaturation
solubilization capacity. Thus, ongoing bile-mediated dilu- that drives increases in drug absorption. Reproduced with
tion of lipid colloidal phases promotes drug permission from Yeap et al. (2013)
Fig. 7.20 Overview of

solidification techniques
commonly used for
transforming liquid and
semisolid lipid-based
formulations into solid
dosage forms. (Adapted
with permission from Tan
et al. 2013)
Fig. 7.21 Percentage of the danazol dose recovered in the initiate digestion. C) Plasma concentration versus time
aqueous phase during in vitro evaluation, consisting of a profiles for danazol after oral administration to overnight
30-min dispersion followed by a 30-min digestion phase. fasted rats. Treatments consisted of MC-SEDDSliquid,
(A) Medium-chain (MC) lipid SEDDS formulations. MC-SEDDSNeusilin, LC-SEDDSliquid, or LC-SEDDSNeusilin.
(B) Long-chain (LC) lipid SEDDS formulations. Values All animals received a nominal dose of 3 mg. Average
are expressed as means (n ¼ 3) SD. The vertical dotted values (n ¼ 4) SD are depicted. (Adapted with permis-
line indicates the point at which pancreatin was added to sion from Van Speybroeck et al. 2012)
desorption of the lipid ingredients from the solid hydrophilic surfactants performed the best in
substrate. Further studies investigating a broader terms of degree of desorption. High hydrophilic
range of model drugs and LBF compositions surfactant containing LBFs were also the least
revealed that incomplete desorption was ubiqui- affected by short-term storage, whereas other
tous among all of the formulations (Williams LBFs were found to show a significant decrease
et al. 2014b). An example is shown in Fig. 7.22 in degree of desorption after 1 month of storage.
for LBFs containing the model drug danazol. Such a change in formulation/drug release stabil-
Some degree in incomplete desorption is ity represents a significant challenge for solidified
observed across all LBFs which varied in triglyc- LBFs.
eride chain length (SC, MC, and LC), surfactant Syringe-based extrusion 3-dimensional print-
loading (15, 30, and 60%), and surfactant type ing is a technique has been adapted to overcome
(Tween 85, Tween 80, Tween 20, Cremophor EL, limitations associated with the aforementioned
and Cremophor RH40), though it was found that solidification techniques. In their study, Vithani
formulations containing high (>30%) amounts of et al. demonstrated the ability to print different
A 100
Tween 85 Tween 80 Tween 20 Cremophor EL Cremophor RH40
Solubilized doseafter 4 h
75
dispersion (%)
50
25
0
15% 30% 60% 15% 30% 60% 15% 30% 60%
Surfactant in long-chain Surfactant in medium-chain Surfactant in short-chain
lipid LBFs (%) lipid LBFs (%) lipid LBFs (%)
B Surfactant in long-chain Surfactant in medium-chain Surfactant in short-chain

lipid LBFs (%) lipid LBFs (%) lipid LBFs (%)
Difference in danazol desorption
20
following 1-month storage (%)
15% 30% 60% 15% 30% 60% 15% 30% 60%

0
–20
–40
–60
–80
Tween 85 Tween 80 Tween 20 Cremophor EL Cremophor RH40
Fig. 7.22 (a) Plot summarizing the effect of increasing chain lipids: triacetin (n ¼ 1). (b) The extent of danazol
the proportion of different nonionic surfactants in LBFs desorption from the formulations following 1-month stor-
consisting of long-, medium-, and short-chain lipids on age time relative to respective initial performance. Nega-
desorption behavior from Neusilin. Bars represent the tive values indicate a decrease in the danazol desorption
extent of danazol desorption (expressed as a percentage following storage. Due to an analytical error, the result for
of the total dose) after 4-h dispersion in SIF*. Long-chain the LC lipid formulation containing 15% Cremophor EL is
lipids: soybean oil–MaisineTM 35–1 (1:1), medium-chain missing from this dataset (n ¼ 1). (Adapted with permis-
lipids: CaptexR_ 355 CapmulR_ MCM EP(2:1), short- sion from Williams et al. 2014b)
S-SMEDDS formulations, in different that can solubilize lipophilic molecules

geometries, without the use of a solvent or a (Fig. 7.23). Various solid lipids include
solid-phase carrier (Vithani et al. 2019a). This monoglycerides, diglycerides, triglycerides, fatty
approach circumvents the large amount of solid- acids, steroids, and waxes (Kaur et al. 2012).
phase carrier required for S-SMEDDS formula- SLNs have significant advantages, for instance,
tion and offers the ability to control drug release they are nontoxic, easy to produce, and are highly
based on the printed geometries. stable, which helps control the release of both
lipophilic and hydrophilic drugs.
To demonstrate the applicability of SLNs for
the delivery of small water-insoluble molecules,
7.4.1 Solid Lipid Nanoparticle (SLN)
Hassan et al. loaded acyclovir, a BCS class III
drug, in SLNs. Following SLNs’ optimization,
Solid lipid nanoparticles (SLNs) are regarded as
administration of the SLNs in rats (Fig. 7.24)
an alternative carrier system to traditional lipid-
showed a 423% increase in bioavailability com-
based formulations as they combine the
pared to commercial Zovirax® suspension, which
advantages of traditional colloidal carriers while
exhibits between 10% and 20% bioavailability in
avoiding some of their main drawbacks
humans after ingestion. Acyclovir, a BCS Class
(Shirodkar et al. 2019). SLNs are typically spher-
III drug, is relevant in the scope of this book as
ical particles (size range 10–1000 nm) with a
this study teaches to improve its poor
solid lipid core matrix (stabilized by surfactants)
Fig. 7.23 Schematic

representation of structure
of solid lipid nanoparticles.
from Vishwakarma et al.
2019)
Fig. 7.24 Plasma

concentration versus time
profile after oral
administration of acyclovir-
loaded SLNs or commercial
suspensions in rats. The
data represent mean SD
(n ¼ 6). (Adapted with
permission from Hassan
et al. 2020)
permeability, and these findings can be applied to

7.5 Conclusion
improve BCS Class IV drugs’ permeability.
Furthermore, the surface charge and size of the
As evident in this chapter, the formulation and
micelles of SLNs formed after lipolysis impact
characterization of LBFs continue to be an active
drug absorption. Ban et al. demonstrated the abil-
area of research in pharmaceutical sciences.
ity to modify both the surface charge and size to
While significant progress has been made to clas-
control the bioavailability levels of curcumin, a
sify and standardize LBF characterization
poorly water-soluble, small molecule with anti-
methods, continued efforts are warranted to
inflammatory properties (Ban et al. 2020).
develop methods that are predictive of in vivo
Selecting long-PEGylated SLNs enabled rapid
bioperformance. Formulation techniques to solid-
permeation of the epithelium, attributed to the
ify LBFs are notable in that they make LBFs a
neutral charge, and achieved greater than a
competitive alternative to other solid enabling
12-fold increase in bioavailability compared to
techniques for poorly water-soluble drugs (e.g.,
the curcumin solution administered in rats. The
amorphous dispersions), but certain limitations
effects of different SLNs’ formulation modifica-
including incomplete desorption and physical
tion on absorption are shown in Fig. 7.25.
Fig. 7.25 Schematic of the absorption of curcumin and permeation through the gut membrane with a mucus
loaded in an emulsion and SLN. Arrow thickness indicates layer. (Adapted with permission from Ban et al. 2020)
the intensity of the absorption procedure, such as lipolysis,
stability need to be addressed. Continued efforts Equipment and Reagents

to understand the mechanisms involved in • Danazol.
sustained supersaturation and increased absorp- • Corn oil, Maisine 35–1, Captex
tion in LBFs will enable greater adoption and 300, Capmul MCM, Tween 85, Cremophor
success of these systems in academic and indus- EL, and Transcutol HP.
trial settings to address the unique needs of poorly • Digestion buffer (pH 6.5) containing 2 mM
soluble compounds. Tris-maleate, 1.4 mM CaCl22H2O,
150 mM NaCl, 3.0 mM sodium taurodeox-
Method Capsule 1 ycholate (NaTDC), and 0.75 mM
phosphatidylcholine (PC).
Drug Solubility Determination in Model LBFs
• Pancreatin.
and in Digested Aqueous Media
• pH-stat apparatus, centrifuge, HPLC,
Based on the method reported by Williams
Method
et al. (2012a) and Williams et al. (2012b)
• For solubility determination in LBFs, stan-
Objective
dard shake-flask methods have been used
• To determine the equilibrium drug solubil-
with an excess of drug substance added to a
ity of danazol in Type I, II, IIIA, IIIB, and
model LBF at 37 C and saturation deter-
IV LBFs and the impact of dilution and
mined after consecutive samples vary by
digestion of lipid formulation components
less than 5% in concentration.
on equilibrium solubility.
• For solubility determination in digested • Danazol, fenofibrate, and tolfenamic acid.

media APDIGEST, drug-free LBFs should • Corn oil, Maisine 35–1, Captex
undergo standardized dispersion and diges- 300, Capmul MCM, Tween 85, Cremophor
tion testing under the same conditions as EL, and Transcutol HP.
active LBFs. Samples of digestion medium • Digestion buffer (pH 6.5) containing 2 mM
can be taken after a predetermined time Tris-maleate, 1.4 mM CaCl22H2O,
period, and digestion can be halted with 150 mM NaCl, 3.0 mM sodium taurodeox-
the addition of a digestion inhibitor (1.0 M ycholate (NaTDC), and 0.75 mM
4-bromophenylboronic acid in methanol). phosphatidylcholine (PC).
The APDIGEST media can then be added to • Pancreatin.
a container containing excess crystalline • pH-stat apparatus, centrifuge, HPLC,
drug and incubated at 37 C. Samples at • Method
select time intervals can then be taken, • Eight model LBFs (1.083 g) were evaluated
separated by centrifugation, and analyzed with three model drugs (danazol,
for equilibration of drug content. fenofibrate, and tolfenamic acid) by dis-
Results persing for 10 min in 39 mL of digestion
• The equilibrium solubility of danazol in the buffer, followed by the addition of 4 mL of
eight LBFs is showed a trend toward pancreatin solution and a 30–60-min diges-
increasing solubility with increasing polar- tion period. Samples were taken at 5- and
ity of the formulation, ranging from a solu- 10-min time points during the dispersion
bility of 8.4 mg/g in the Type I-LC phase and at 5, 15, 30, and 60 min after
formulation to the highest solubility of digestion was initiated for drug content
65.5 mg/g in the Type IV formulation. analysis. Extent of digestion was monitored
• Danazol solubility in excipients alone was through back titration calculations to
explored and found to be nonpredictive of maintain pH.
the results in the multicomponent LBF Results
mixtures. Calculated solubilities in the • Type I-MC formulations were approxi-
model LBF formulations on the basis of mately 90% digested under the current
individual values or weight-normalized experimental conditions. Formulation
ratios were all lower than values measured digestion was 82–100%, 90–105%, and
in each formulation. 71–117% for Type II-MC, IIIA-MC, and
• For solubility determinations in IIIB-MC formulations, respectively. Type
APDIGEST, danazol solubility decreased IV formulation, which contains Cremophor
to less than 200 μg/mL for all of the EL as the only source of FA, revealed
LBFs, illustrating the potential drop in sol- approximately 15% digestion.
vent capacity upon dilution and digestion of • Estimations of the extent of LC formulation
LBF components. digestion were 25% for the Type I-LC
systems, and between 27–37% and
Method Capsule 2 67–95% for the Type II-LC and IIIA-LC
formulations, respectively.
Lipid Formulation Classification System Con-
• The type of drug incorporated or the drug
sortium Fasted-State Digestion Test
load had little effect on the digestion
properties, but drug load did impact precip-
et al. (2012a) and Williams et al. (2013a).
itation based on a maximum calculated
Objective
supersaturation ratio, SRM.
• To standardize experimental parameters for
• Drug crystallization was only evident in
in vitro dispersions and digestion testing of
instances when the calculated maximum
Type I, II, IIIA, IIIB, and IV LBFs.
supersaturation ratio (SRM) was >3,
• Equipment and Reagents
which was consistent across all LBF.
Method Capsule 3 Results

• Decreased LBF mass was a stressor to
Stressed Digestion Test and LBF Performance
medium-chain glyceride-rich LBFs, but
Classification
not more hydrophilic surfactant-rich LBFs,
whereas decreasing pH stressed tolfenamic
et al. (2014a).
acid LBFs, but not fenofibrate LBFs.
Objective
• An A–D performance classification system
• To develop “stressed” digestion method
was proposed based on a passing or failed
parameters to further discriminate between
performance in 1) dispersion testing, 2)
LBFs and guide lead candidate optimiza-
fasted-state digestion testing, and 3)
tion through a performance classification
stressed digestion testing. An LBF that
system.
passed all three in vitro tests would receive
an A grade, whereas one that failed disper-
• Fenofibrate and tolfenamic acid.
sion testing would receive a D grade.
• Corn oil, Maisine 35–1, Captex
300, Capmul MCM, Tween 85, Cremophor
Method Capsule 4
EL, and Transcutol HP.
• Fasted digestion buffer (pH 6.5) containing Characterization of S-SEDDS Formulation of
2 mM Tris-maleate, 1.4 mM CaCl22H2O, Celecoxib by In Vitro Biphasic Test and its
150 mM NaCl, 0.3 mM sodium taurodeox- Relevance to In Vivo Absorption in Humans
ycholate (NaTDC), and 0.75 mM Based on the method reported by Shi
phosphatidylcholine (PC). et al. (2010).
• Decreased pH digestion buffer (pH 4.5) Objective
containing Tris-maleate, CaCl22H2O, • To characterize drug release profiles of
150 mM NaCl, 0.3 mM sodium taurodeox- three formulations of celecoxib (CEB)
ycholate (NaTDC), and 0.75 mM including an S-SEDDS formulation by an
phosphatidylcholine (PC). in vitro biphasic test and establish the rele-
• Increased Bile-Salt Digestion Buffer vance of the in vitro profiles to in vivo
(pH 6.5) containing 2 mM Tris-maleate, absorption in human subjects.
1.4 mM CaCl22H2O, 150 mM NaCl, • Equipment and Reagents
10.0 mM sodium taurodeoxycholate • Celecoxib, marketed product (Celebrex®).
(NaTDC), and 2.5 mM • Ethanol, PEG400, PVP-12 PF, HPMC
phosphatidylcholine (PC). (E5), oleic acid, Tween 80, octanol.
• Pancreatin. • Gelatin capsule (size 1).
• pH-stat apparatus, centrifuge, HPLC, • USP II dissolution apparatus with a vessel
Method containing two phases: aqueous phase and
• Model LBFs containing fenofibrate or octanol phase. The aqueous phase was
tolfenamic acid were evaluated in a stan- 250 mL 80 mM phosphate buffer, and the
dard dispersion test, a standard fasted-state octanol was 200 mL. A dual paddle was
digestion test, and a “stressed” digestion used at 75 rpm.
test. Stressed digestion tests included a • USP IV dissolution apparatus (a flow cell)
decreased pH version for ionizable, was connected to the USP II vessel through
pH-dependent compounds such as a closed loop with the aqueous dissolution
tolfenamic acid; an increased bile-salt con- medium circulated at 30 mL/min.
centration version that contained 10 mM of Method
NaTDC; and a diluted version which • CEB S-SEDDS formulation was prepared
consisted of dispensing 0.173 g of LBF in by mixing the drug in the vehicle
39 mL of fasted digestion buffer. containing PEG 400, oleic acid and Tween
80 under gentle heating (~50 C) and as a trigger for supersaturation: evaluation of the
vortexing to yield a clear solution. impact of supersaturation stabilization on the in vitro
and in vivo performance of self-emulsifying drug
• Pharmacokinetic study conducted in human delivery systems. Mol Pharm. 2012;9(7):2063–79.
volunteers at a dose of 200 mg as reported https://doi.org/10.1021/mp300164u.
previously in the reference therein. Bakala-N'Goma JC, Williams HD, Sassene PJ, Kleberg K,
Results Calderone M, Jannin V, Igonin A, Partheil A, March-
aud D, Jule E, Vertommen J, Maio M, Blundell R,
• The biphasic test results reveal that the Benameur H, Mullertz A, Pouton CW, Porter CJ,
S-SEDDS formulation of CEB rapidly Carriere F. Toward the establishment of standardized
yields and sustains a supersaturated state in vitro tests for lipid-based formulations. 5. Lipolysis
as compared to the marketed capsule and of representative formulations by gastric lipase. Pharm
Res. 2015;32(4):1279–87. https://doi.org/10.1007/
suspension formulation. The high mean s11095-014-1532-y.
Cmax and AUC observed in human Ban C, Jo M, Park YH, Kim JH, Han JY, Lee KW, Kweon
subjects from the S-SEDDS formulation DH, Choi YJ. Enhancing the oral bioavailability of
are attributed to the supersaturated state of curcumin using solid lipid nanoparticles. Food Chem.
2020;302 https://doi.org/10.1016/j.foodchem.2019.
CEB. 125328.
• In vitro CEB concentration–time profiles Bnyan R, Khan I, Ehtezazi T, Saleem I, Gordon S,
observed in the aqueous phase (with and O'Neill F, Roberts M. Surfactant effects on lipid-
without the octanol phase) are found to based vesicles properties. J Pharm Sci. 2018;107:
1237–46. https://doi.org/10.1016/j.xphs.2018.01.005.
bear no relationship to the in vivo AUC in Carriere F, Barrowman JA, Verger R, Laugier R. Secretion
human subjects. and contribution to lipolysis of gastric and pancreatic
• Good in vitro–in vivo relationship (IVIVR) lipases during a test meal in humans. Gastroenterology.
is obtained between the CEB concentration 1993;105(3):876–88.
Constantinides PP, Wasan KM. Lipid formulation
in the octanol phase at t ¼ 2 h of the three strategies for enhancing intestinal transport and absorp-
formulations and their mean in vivo AUC tion of P-glycoprotein (P-gp) substrate drugs: in vitro/
and Cmax values observed in human in vivo case studies. J Pharm Sci. 2007;96(2):235–48.
subjects. https://doi.org/10.1002/jps.20780.
Dahan A, Hoffman A. The effect of different lipid based
• The biphasic test method appears an effec- formulations on the oral absorption of lipophilic drugs:
tive tool for screening and optimizing the ability of in vitro lipolysis and consecutive ex vivo
formulations of BCS II drugs. intestinal permeability data to predict in vivo bioavail-
ability in rats. Eur J Pharm Biopharm. 2007;67(1):
96–105. https://doi.org/10.1016/j.ejpb.2007.01.017.
Dahan A, Hoffman A. Rationalizing the selection of oral
lipid based drug delivery systems by an in vitro
References dynamic lipolysis model for improved oral bioavail-
ability of poorly water soluble drugs. J Control
Aburahma MH. Bile salts-containing vesicles: promising Release. 2008;129(1):1–10. https://doi.org/10.1016/j.
pharmaceutical carriers for oral delivery of poorly jconrel.2008.03.021.
water-soluble drugs and peptide/protein-based thera- Dahan A, Miller JM, Hoffman A, Amidon GE, Amidon
peutics or vaccines. Drug Deliv. 2016;1–21. https:// GL. The solubility-permeability interplay in using
doi.org/10.3109/10717544.2014.976892 cyclodextrins as pharmaceutical solubilizers: mecha-
Agarwal V, Siddiqui A, Ali H, Nazzal S. Dissolution and nistic modeling and application to progesterone. J
powder flow characterization of solid self-emulsified Pharm Sci. 2010;99(6):2739–49. https://doi.org/10.
drug delivery system (SEDDS). Int J Pharm. 2009;366 1002/jps.22033.
(1–2):44–52. https://doi.org/10.1016/j.ijpharm.2008. Dening TJ, Rao S, Thomas N, Prestidge CA. Novel
08.046. nanostructured solid materials for modulating Oral
Alskar LC, Keemink J, Johannesson L, Porter CJH, drug delivery from solid-state lipid-based drug deliv-
Bergström CAS. Impact of drug physicochemical ery systems. AAPS J. 2015; https://doi.org/10.1208/
properties on lipolysis-triggered drug supersaturation s12248-015-9824-7.
and precipitation from lipid-based formulations. Mol Devraj R, Williams HD, Warren DB, Mohsin K, Porter CJ,
Pharm. 2018;15:4733–44. https://doi.org/10.1021/acs. Pouton CW. In vitro assessment of drug-free and
molpharmaceut.8b00699. fenofibrate-containing lipid formulations using disper-
Anby MU, Williams HD, McIntosh M, Benameur H, sion and digestion testing gives detailed insights into
Edwards GA, Pouton CW, Porter CJ. Lipid digestion
the likely fate of formulations in the intestine. Eur J Biomedecine & pharmacotherapie. 2004;58(3):
Pharm Sci. 2013;49(4):748–60. https://doi.org/10. 173–82. https://doi.org/10.1016/j.biopha.2004.02.001.
1016/j.ejps.2013.04.036. Hansen T, Holm P, Schultz K. Process characteristics and
Devraj R, Williams HD, Warren DB, Porter CJ, Pouton compaction of spray-dried emulsions containing a drug
CW. Choice of nonionic surfactant used to formulate dissolved in lipid. Int J Pharm. 2004;287(1–2):55–66.
type IIIA self-emulsifying drug delivery systems and https://doi.org/10.1016/j.ijpharm.2004.08.014.
the physicochemical properties of the drug have a Hassan H, Bello RO, Adam SK, Alias E, Affandi M,
pronounced influence on the degree of drug supersatu- Shamsuddin AF, Basir R. Acyclovir-loaded solid
ration that develops during in vitro digestion. J Pharm lipid nanoparticles: optimization, characterization and
Sci. 2014;103(4):1050–63. https://doi.org/10.1002/jps. evaluation of its pharmacokinetic profile. Nano.
23856. 2020;10(9):1785. https://doi.org/10.3390/
Fernandez S, Jannin V, Rodier JD, Ritter N, Mahler B, nano10091785.
Carriere F. Comparative study on digestive lipase Jannin V, Musakhanian J, Marchaud D. Approaches for
activities on the self emulsifying excipient Labrasol, the development of solid and semi-solid lipid-based
medium chain glycerides and PEG esters. Biochim formulations. Adv Drug Deliv Rev. 2008;60(6):
Biophys Acta. 2007;1771(5):633–40. https://doi.org/ 734–46. https://doi.org/10.1016/j.addr.2007.09.006.
10.1016/j.bbalip.2007.02.009. Kaur J, Singh G, Saini S, Rana A. Innivative growth in
Gao P, Morozowich W. Development of supersaturatable developing new methods for formulating solid lipid
self-emulsifying drug delivery system formulations for nanoparticles and microparticles. J Drug Deliv Therap.
improving the oral absorption of poorly soluble drugs. 2012;2(5) https://doi.org/10.22270/jddt.v2i5.295.
Expert Opin Drug Deliv. 2006;3(1):97–110. https:// Konno H, Handa T, Alonzo DE, Taylor LS. Effect of
doi.org/10.1517/17425247.3.1.97. polymer type on the dissolution profile of amorphous
Gao P, Morozowich W. Case studies: rational develop- solid dispersions containing felodipine. Eur J Pharm
ment of self-emulsifying formulations for improving Biopharm. 2008;70(2):493–9. https://doi.org/10.1016/
the oral bioavailability of poorly soluble, lipophilic j.ejpb.2008.05.023.
drugs. Drugs Pharm Sci. 2007;170:273. Lengsfeld H, Beaumier-Gallon G, Chahinian H, De
Gao P, Rush BD, Pfund WP, Huang T, Bauer JM, Caro A, Verger R, Laugier R, Carrière F (2006) Physi-
Morozowich W, Kuo MS, Hageman ology of gastrointestinal lipolysis and therapeutical use
MJ. Development of a supersaturable SEDDS of lipases and digestive lipase inhibitors. Lipases and
(S-SEDDS) formulation of paclitaxel with improved Phospholipases in Drug Development: From Biochem-
oral bioavailability. J Pharm Sci. 2003;92(12): istry to Molecular Pharmacology.
2386–98. https://doi.org/10.1002/jps.10511. Miller JM, Beig A, Krieg BJ, Carr RA, Borchardt TB,
Gao P, Guyton ME, Huang T, Bauer JM, Stefanski KJ, Lu Amidon GE, Amidon GL, Dahan A. The solubility-
Q. Enhanced oral bioavailability of a poorly water permeability interplay: mechanistic modeling and pre-
soluble drug PNU-91325 by supersaturatable dictive application of the impact of micellar solubiliza-
formulations. Drug Dev Ind Pharm. 2004;30(2): tion on intestinal permeation. Mol Pharm. 2011;8(5):
221–9. https://doi.org/10.1081/DDC-120028718. 1848–56. https://doi.org/10.1021/mp200181v.
Gao P, Akrami A, Alvarez F, Hu J, Li L, Ma C, Surapaneni Miller JM, Beig A, Carr RA, Webster GK, Dahan A. The
S. Charac terization and optimization of AMG solubility-permeability interplay when using
517 supersaturatable self-emulsifying drug delivery cosolvents for solubilization: revising the way we use
system (S-SEDDS) for improved oral absorption. J solubility-enabling formulations. Mol Pharm. 2012;9
Pharm Sci. 2009;98(2):516–28. https://doi.org/10. (3):581–90. https://doi.org/10.1021/mp200460u.
1002/jps.21451. O'Driscoll CM. Lipid-based formulations for intestinal
Gibson L. Lipid-based excipients for oral drug delivery. lymphatic delivery. Eur J Pharm Sci. 2002;15(5):
Drugs Pharm Sci. 2007;170:33. 405–15.
Goole J, Lindley DJ, Roth W, Carl SM, Amighi K, Patel JP, Brocks DR. The effect of oral lipids and
Kauffmann JM, Knipp GT. The effects of excipients circulating lipoproteins on the metabolism of drugs.
on transporter mediated absorption. Int J Pharm. Expert Opin Drug Metab Toxicol. 2009;5(11):
2010;393(1–2):17–31. https://doi.org/10.1016/j. 1385–98. https://doi.org/10.1517/
ijpharm.2010.04.019. 17425250903176439.
Griffin BT, Kuentz M, Vertzoni M, Kostewicz ES, Fei Y, Pouton CW. Lipid formulations for oral administration of
Faisal W, Stillhart C, O'Driscoll CM, Reppas C, drugs: non-emulsifying, self-emulsifying and 'self-
Dressman JB. Comparison of in vitro tests at various microemulsifying' drug delivery systems. Eur J
levels of complexity for the prediction of in vivo per- Pharm Sci. 2000;11(Suppl 2):S93–8.
formance of lipid-based formulations: case studies with Pouton CW. Formulation of poorly water-soluble drugs
fenofibrate. Eur J Pharm Biopharm. 2014;86(3): for oral administration: physicochemical and physio-
427–37. https://doi.org/10.1016/j.ejpb.2013.10.016. logical issues and the lipid formulation classification
Gursoy RN, Benita S. Self-emulsifying drug delivery system. Eur J Pharm Sci. 2006;29(3–4):278–87.
systems (SEDDS) for improved oral delivery of https://doi.org/10.1016/j.ejps.2006.04.016.
lipophilic drugs. Biomed Pharmacother ¼
Sassene P, Kleberg K, Williams HD, Bakala-N'Goma JC, and lymphatic drug transport in the rat. Drug Metab
Carriere F, Calderone M, Jannin V, Igonin A, Dispos. 2006;34(5):729–33. https://doi.org/10.1124/
Partheil A, Marchaud D, Jule E, Vertommen J, dmd.105.008102.
Maio M, Blundell R, Benameur H, Porter CJ, Pouton Trevaskis NL, Charman WN, Porter CJ. Lipid-based
CW, Mullertz A. Toward the establishment of delivery systems and intestinal lymphatic drug trans-
standardized in vitro tests for lipid-based formulations, port: a mechanistic update. Adv Drug Deliv Rev.
part 6: effects of varying pancreatin and calcium levels. 2008;60(6):702–16. https://doi.org/10.1016/j.addr.
AAPS J. 2014;16(6):1344–57. https://doi.org/10.1208/ 2007.09.007.
s12248-014-9672-x. Ueda K, Higashi K, Yamamoto K, Moribe K. Inhibitory
Sek L, Porter CJ, Kaukonen AM, Charman effect of hydroxypropyl methylcellulose acetate succi-
WN. Evaluation of the in-vitro digestion profiles of nate on drug recrystallization from a supersaturated
long and medium chain glycerides and the phase solution assessed using nuclear magnetic resonance
behaviour of their lipolytic products. J Pharm measurements. Mol Pharm. 2013;10(10):3801–11.
Pharmacol. 2002;54(1):29–41. https://doi.org/10.1021/mp400278j.
Shah AV, Serajuddin AT. Development of solid self- Van Speybroeck M, Williams HD, Nguyen TH, Anby
emulsifying drug delivery system (SEDDS) I: use of MU, Porter CJ, Augustijns P. Incomplete desorption
poloxamer 188 as both solidifying and emulsifying of liquid excipients reduces the in vitro and in vivo
agent for lipids. Pharm Res. 2012;29(10):2817–32. performance of self-emulsifying drug delivery systems
https://doi.org/10.1007/s11095-012-0704-x. solidified by adsorption onto an inorganic mesoporous
Shi Y, Gao P, Gong Y, Ping H. Application of a biphasic carrier. Mol Pharm. 2012;9(9):2750–60. https://doi.
test for characterization of in vitro drug release of org/10.1021/mp300298z.
immediate release formulations of celecoxib and its Vishwakarma N, Jain A, Sharma R, Mody N, Vyas S,
relevance to in vivo absorption. Mol Pharm. 2010;7 Vyas SP. Lipid-based Nanocarriers for lymphatic
(5):1458–65. https://doi.org/10.1021/mp100114a. transportation. AAPS PharmSciTech. 2019;20
Shirodkar RK, Kumar L, Mutalik S, Lewis S. Solid lipid (83) https://doi.org/10.1208/s12249-019-123-3.
nanoparticles and nanostructured lipid carriers: Vithani K, Goyanes A, Jannin V, Basit A, Gaisford S,
emerging lipid based drug delivery systems. Pharm Boyd BJ. A proof of concept for 3D printing of solid
Chem J. 2019;53:440–53. https://doi.org/10.1007/ lipid-based formulations of poorly water-soluble drugs
s11094-019-02017-9. to control formulation dispersion kinetics. Pharm Res.
Strickley RG. Solubilizing excipients in oral and injectable 2019a;36:102. https://doi.org/10.1007/s11095-019-
formulations. Pharm Res. 2004;21(2):201–30. 2639-y.
Suys EJA, Chalmers DK, Pouton CW, Porter CJH. Poly- Vithani K, Jannin V, Pouton CW, Boyd BJ. Colloidal
meric precipitation inhibitors promote Fenofibrate aspects of dispersion and digestion of self-dispersing
supersaturation and enhance drug absorption from a lipid-based formulations for poorly water-soluble
type IV lipid- based formulation. Mol Pharm. drugs. Adv Drug Deliv Rev. 2019b;142:16–34.
2018;15(6):2355–71. https://doi.org/10.1021/acs. https://doi.org/10.1016/j.addr.2019.01.008.
molpharmaceut.8b00206. Warren DB, Benameur H, Porter CJ, Pouton CW. Using
Tan A, Rao S, Prestidge CA. Transforming lipid-based polymeric precipitation inhibitors to improve the
oral drug delivery systems into solid dosage forms: absorption of poorly water-soluble drugs: a mechanis-
an overview of solid carriers, physicochemical tic basis for utility. J Drug Target. 2010;18(10):
properties, and biopharmaceutical performance. 704–31.
Pharm Res. 2013;30(12):2993–3017. https://doi.org/ Williams HD, Anby MU, Sassene P, Kleberg K, Bakala-
10.1007/s11095-013-1107-3. N'Goma JC, Calderone M, Jannin V, Igonin A,
Tay E, Nguyen TH, Ford L, Williams HD, Benameur H, Partheil A, Marchaud D, Jule E, Vertommen J,
Scammells PJ, Porter CJH. Ionic liquid forms of the Maio M, Blundell R, Benameur H, Carriere F,
antimalarial Lumefantrine in combination with LFCS Mullertz A, Pouton CW, Porter CJ. Toward the estab-
type IIIB lipid-based formulations preferentially lishment of standardized in vitro tests for lipid-based
increase lipid solubility, in vitro solubilization behav- formulations. 2. The effect of bile salt concentration
ior and in vivo exposure. Pharmaceutics. 2019;12(1): and drug loading on the performance of type I, II, IIIA,
17. https://doi.org/10.3390/pharmaceutics12010017. IIIB, and IV formulations during in vitro digestion.
Tran PH, Tran TT, Piao ZZ, Vo TV, Park JB, Lim J, Oh Mol Pharm. 2012a;9(11):3286–300. https://doi.org/
KT, Rhee YS, Lee BJ. Physical properties and in vivo 10.1021/mp300331z.
bioavailability in human volunteers of isradipine using Williams HD, Sassene P, Kleberg K, Bakala-N'Goma JC,
controlled release matrix tablet containing self- Calderone M, Jannin V, Igonin A, Partheil A, March-
emulsifying solid dispersion. Int J Pharm. 2013;450 aud D, Jule E, Vertommen J, Maio M, Blundell R,
(1–2):79–86. https://doi.org/10.1016/j.ijpharm.2013. Benameur H, Carriere F, Mullertz A, Porter CJ, Pouton
04.022. CW. Toward the establishment of standardized in vitro
Trevaskis NL, Porter CJ, Charman WN. An examination tests for lipid-based formulations, part 1: method
of the interplay between enterocyte-based metabolism parameterization and comparison of in vitro digestion
profiles across a range of representative formulations. J Williams HD, Sassene P, Kleberg K, Calderone M,
Pharm Sci. 2012b;101(9):3360–80. https://doi.org/10. Igonin A, Jule E, Vertommen J, Blundell R,
1002/jps.23205. Benameur H, Mullertz A, Porter CJ, Pouton CW,
Williams HD, Sassene P, Kleberg K, Calderone M, Communicated on Behalf of the LC. Toward the estab-
Igonin A, Jule E, Vertommen J, Blundell R, lishment of standardized in vitro tests for lipid-based
Benameur H, Mullertz A, Pouton CW, Porter CJ, Con- formulations, part 4: proposing a new lipid formulation
sortium L. Toward the establishment of standardized performance classification system. J Pharm Sci.
in vitro tests for lipid-based formulations, part 3: 2014a;103(8):2441–55. https://doi.org/10.1002/jps.
understanding supersaturation versus precipitation 24067.
potential during the in vitro digestion of type I, II, Williams HD, Van Speybroeck M, Augustijns P, Porter
IIIA, IIIB and IV lipid-based formulations. Pharm CJ. Lipid-based formulations solidified via adsorption
Res. 2013a;30(12):3059–76. https://doi.org/10.1007/ onto the mesoporous carrier Neusilin(R) US2: effect of
s11095-013-1038-z. drug type and formulation composition on in vitro
Williams HD, Trevaskis NL, Yeap YY, Anby MU, Pouton pharmaceutical performance. J Pharm Sci. 2014b;103
CW, Porter CJ. Lipid-based formulations and drug (6):1734–46. https://doi.org/10.1002/jps.23970.
supersaturation: harnessing the unique benefits of the Yeap YY, Trevaskis NL, Quach T, Tso P, Charman WN,
lipid digestion/absorption pathway. Pharm Res. Porter CJ. Intestinal bile secretion promotes drug
2013b;30(12):2976–92. https://doi.org/10.1007/ absorption from lipid colloidal phases via induction
s11095-013-1126-0. of supersaturation. Mol Pharm. 2013;10(5):1874–89.
https://doi.org/10.1021/mp3006566.
Structured Development Approach
for Amorphous Systems 8
Susanne Page, Reto Maurer, Nicole Wyttenbach,
and Felix Ditzinger
Abstract the contribution of the authors of this chapter

A structured development approach is from the previous editions. This current third
presented to guide the development of stable edition chapter is a revision and update of the
and commercially viable polymer-based amor- second edition.
phous formulations. The proposed approach
should not only enable the delivery of poorly Keywords
soluble drugs but also help to reduce the API In silico evaluation · High-throughput
needs, reduce in vivo screening, minimize screening assays · In vitro dissolution
risks for late-stage development, and ensure methods · Amorphous solid dispersions
consistent quality. During initial assessment, (ASDs) · Miscibility · Solubility parameter
a guided evaluation of the physicochemical calculations · Molecular modeling · Molecular
properties of the API helps to assess the degree dynamic simulation · Flory-Huggins theory ·
of difficulty for the development. A range of Miniaturized systems · Supersaturation
tests including in silico evaluation, high- screening · X-ray powder diffraction (XRPD) ·
throughput screening assays, and miniaturized Microscopy
screening tools provide a road map for
selecting the appropriate polymer, drug load-
ing, and suitable manufacturing process. A 8.1 Introduction
dedicated section provides a review of the
characterization tools to assess and quantify Over the last decades, amorphous solid
the crystallinity, understanding the phase dispersions (ASDs) have emerged as a method
behavior of amorphous solid dispersions, and of choice for improving the dissolution behavior
designing the in vitro dissolution methods. and bioavailability of poorly water-soluble active
Finally, a reference chart is provided that pharmaceutical ingredients (Chiou and
summarizes the key concepts proposed as Riegelmann 1970; Hancock and Parks 2000;
part of the structured development approach Mishra et al. 2015; Jermain et al. 2018; Edueng
that can serve as a blueprint for the develop- et al. 2017; Vinarov et al. 2021). As amorphous
ment of amorphous formulations. The current compounds are thermodynamically unstable, they
authors would like to thank and acknowledge may crystallize over pharmaceutically relevant
timescales and thus negate any solubility advan-
S. Page (*) · R. Maurer · N. Wyttenbach · F. Ditzinger tage. Therefore, amorphous compounds are often
F. Hoffmann-La Roche AG, Basle, Switzerland stabilized by combining the active ingredient with
288 S. Page et al.
a carrier polymer to form an amorphous, molecu- chemical properties and that therefore no univer-
lar-level solid dispersion, as described in several sal formulation can exist (Liu et al. 2015). Conse-
comprehensive reviews (Leuner and Dressman quently, his/her primary focus in the early stage of
2000; Serajuddin 1999; Van den Mooter et al. development is the selection of a suitable carrier
2001; Janssens et al. 2010; Mishra et al. 2015; system and drug load, followed by the
He and Ho 2015). The properties of the resultant manufacturing of some prototype formulations
solid dispersions are influenced by the physico- for pharmacokinetic testing in animals, which
chemical properties of both the active pharmaceu- might be replaced by in silico and more advanced
tical ingredient and the carrier polymer. Based on in vitro assessment in the future. Based on the
the physical state and the composition of the results of the in vivo behavior of the formulation,
carrier, ASDs have been classified into different an intensive analytical characterization and stabil-
categories: first-generation ASDs contain crystal- ity testing of the final formulation composition
line carriers (e.g., urea, mannitol), while second- can be defined. For the next step of the develop-
generation ASDs consist of amorphous carriers ment (e.g., process development and up-scaling),
(mostly polymers). For some time, surface active the quality by design (QbD) approach was suc-
agents or self-emulsifiers have been used as cessfully used by a number of researchers so far
carriers or additives to improve the dissolution (Agrawal et al. 2015; Sanghvi et al. 2015;
rate and to further reduce the risk of API precipi- Grymonpré et al. 2017b). As amorphous solid
tation and recrystallization (third-generation dispersions are often used for formulating BCS
ASDs). In fourth-generation ASDs, water insolu- II or IV compounds (BCS ¼ biopharmaceutical
ble or swellable polymers were introduced to classification system), the critical quality
enhance the solubility as well as to ensure the attributes (CQAs) for the product in development
drug release in a controlled manner (controlled- are dissolution, bioavailability, and solid-state
release solid dispersions) (Mishra et al. 2015; Vo stability (Siew 2014). Once an acceptable target
et al. 2013; Ditzinger et al. 2019a). Recently, new is identified for each CQA, design of experiments
carrier systems such as mesoporous silica (see can be performed in order to identify the relation-
Chap. 13 “Emerging Technologies”) or amino ship between the critical process parameters and
acids (Lobmann and Grohganz 2013; Jensen the CQAs that allows setting a design space
et al. 2015; Lenz et al. 2015) were used, or func- acceptable for consistent product quality. The
tional additives such as plasticizers, solubilizers, focus of this chapter is to describe a structured
wetting agents, superdisintegrants, antioxidants, approach for the development of amorphous
and pH modifiers were added. Over the past formulations that should help bolster confidence
years, the characterization methods as well as and provide a formulation with the best chance
the manufacturing technologies further advanced. for success
The expanded theoretical and practical knowl-
edge of amorphous systems with respect to ther-
modynamic as well as kinetic stability and the 8.2 Ideal Amorphous Formulation:
availability of various modern instrumental Structured Development
techniques to qualitatively and quantitatively
characterize the amorphous system has led to An ideal amorphous formulation should provide
more amorphous drug products being introduced the maximum physical stability during processing
into the market place (Table 8.1) (Wyttenbach and storage and maintain supersaturation while
and Kuentz 2017; Tan et al. 2020). the drug is being dissolved and absorbed in the
This increase might be a result of a more gastrointestinal tract. Among various factors, the
structured development approach in comparison selection of polymer and drug loading are two
to the empirical approach used in the past. The key aspects in the development of an ideal amor-
formulation scientist is aware of the fact that each phous formulation. The inhibitory effects of
drug candidate has its own unique physical and polymers against crystallization in the solid state
8
Table 8.1 Examples of marketed polymeric amorphous solid dispersions
Drug Drug Tmb or Drug
FDA MWa Tdecc Tgd Processing
approval Trade name Manufacturer Drug name (g mol-1) ( C) ( C) Carrier technology
1975 Gris-PEG® Pedinol Pharm Griseofulvin 353 218f 89f PEG 6000 Melt blending
1981 Isoptin SR AbbVie Verapamil hydrochloride 491 148i 61i HPC/HPMC Melt extrusion
1985 Cesamet® Lilly/Valeant Nabilone 373 160e n.a. PVP Melt extrusion
1989w Nivadil® Fujisawa Nilvadipine 385 168m 49m HPMC Melt extrusion
1992 Sporanox® Janssen Itraconazole 706 168f 58f HPMC Bead coating
1994 Prograf® Fujisawa Tacrolimus 804 128e 76o HPMC Solvent
process
1997 Rezulin® Pfizer Troglitazone 442 125/175B 63o PVP Melt extrusion
(Parke-Davis)
2000 Afeditab® CR Elan/Watson Nifedipine 346 173o 47o PVP/poloxamer Melt/absorb on
carrier
2003 Crestor® AstraZeneca Rosuvastatin calcium 1001 n.a. 121o HPMC Spray drying
2006 Nimotop® Bayer Nimodipine 418 125o 14o PEG Spray drying/
fluid bed
2007 Fenoglide® LifeCycle Pharma Fenofibrate 361 81o 19o PEG 6000 Melt
poloxamer 188 granulation
Structured Development Approach for Amorphous Systems
2007 Kaletra® AbbVie Ritonavir/lopinavir 721 126f 49f PVP VA 64 Melt extrusion
629 125j 78k
2008 Intelence® Janssen Etravirine 435 264h 100h HPMC Spray drying
2009 Samsca® Otsuka Tolvaptan 449 221o 122o HPMC Granulation
2010 Certican®/ Novartis Everolimus 958 115e 50p HPMC Spray drying
Zortress®
2010 Norvir® AbbVie Ritonavir 721 126f 49f PVP VA 64 Melt extrusion
2010 Onmel® GSK/Stiefel Itraconazole 706 168f 58f HPMC Melt extrusion
2011 INCIVEK™/ Vertex/Janssen Telaprevir 680 246g 105g HPMCAS Spray drying
Incivo®
2011 Zelboraf® Roche Vemurafenib 490 272n 105n HPMCAS Coprecipitation
2012 Kalydeco® Vertex Ivacaftor 392 292l 175l HPMCAS Spray drying
2012 Stivarga® Bayer Regorafenib 483 208A 86A PVP K25 –
2013 Noxafil® Merck/Schering Posaconazole 701 168x 59x HPMCAS Melt extrusion
2014 Belsomra® Merck Suvorexant 451 125o 75o PVP VA 64 Melt extrusion
2014 Harvoni Gilead Ledipasvir/sofosbuvir PVP VA 64 Spray drying
289
(continued)
Drug Drug Tmb or Drug
290
FDA MWa Tdecc Tgd Processing

approval Trade name Manufacturer Drug name (g mol-1) ( C) ( C) Carrier technology
889 168–172q 160q
529 120–125r 59o
2014 Lynparza® AstraZeneca Olaparib 434 210 o 95 o PVP VA 64 Melt extrusion
2014 Viekira XR™ AbbVie Dasabuvir sodium monohydrate/ombitasvir/ 534 n.a. n.a. PVP VA 64 Melt extrusion
paritaprevir/ritonavir 894 n.a. n.a.
766 n.a. n.a.
721 126e 49e
2016 Epclusa® Gilead Velpatasvir/sofosbuvir 883 109/177s 155o PVP VA 64 Spray drying
529 120–125r 59o
2016 Orkambi® Vertex Lumacaftor/ivacaftor 452 204o 84o HPMCAS Spray drying
392 292l 175l
2016 Venclexta™ AbbVie Venetoclax 868 138z 116o PVP VA Melt extrusion
64, polysorbate 80
2016 Zepatier® Merck Elbasvir/grazoprevir 882 n.a.y 159o PVP VA Spray drying
767 184o 150o 64, TPGS,
HPMC
2017 Mavyret™ AbbVie Glecaprevir/pibrentasvir 839 n.a.u 146o PVP VA 64, TPGS Melt extrusion
1113 n.a.u 160o
2017 Vosevi® Gilead Voxilaprevir/velpatasvir/sofosbuvir 869 278o 145o PVP VA 64 Spray drying
883 109/177s 155o
529 120–125r 59o
2018 Braftovi® Array BioPharma/ Encorafenib 540 182o 76o PVP VA Melt extrusion
Pierre Fabre 64, poloxamer 188
2018 Erleada® Janssen Apalutamide 477 194o 100o HPMCAS Spray drying
2018 Symdeko® Vertex Tezacaftor/ivacaftor 452 172–178t 76o HPMCAS Spray drying
392 292l 175l
2019 Trikafta™/ Vertex Elexacaftor/tezacaftor/ivacaftor 598 ~215 Cv n.a. HPMCAS Spray drying
Kaftrio® 452 172–178t 76o
392 292l 175l
a
Molecular weight; bMelting point; cDecomposition temperature; dGlass transition temperature; eHuang and Dai (2014), fBaird et al. (2010), gKwong et al. (2011), hWeuts et al.
(2011), iAdrjanowicz et al. (2010), jPatel et al. (2015), kLemmer and Liebenberg (2013), lHadida et al. (2014), mMiyazaki et al. (2007), nShah et al. (2013), oMeasured according to
Wyttenbach et al. (2016), pMeasured by modulated DSC, qScott et al. (2013), rMansuri et al. (2018), sSolvated form, different isostructural channel solvates are reported by Lapina
et al. (2015), tKeshavarz-Shokri et al. (2011), uDifferent crystalline solvate forms are reported by Chen et al. (2015) and by Califano et al. (2016), vAbela et al. (2018),
w
Nilvadipine is approved in Japan and in many European countries but is not approved in the USA, xMudie et al. (2020), yDifferent crystalline solvate forms and elbasvir
pseudopolymorphs are reported by Mcnevin et al. (2015); zEMEA. Assessment Report Venclyxto (2016), AChen et al. (2019), BSuzuki et al. (2002)
S. Page et al.
8 Structured Development Approach for Amorphous Systems 291
Table 8.2 Overview of the structured development approach

Stage Bioavailability Stability Manufacturability
Initial assessment Prediction of the Glass-forming ability Preselection of
solubility advantage of Glass stability/fragility manufacturing
the amorphous form technologies
Polymer screening – Part I – API-polymer miscibility Further restriction of
(theoretical) Specific interactions polymers based on the
Hygroscopicity and preselected manufacturing
water activity technology
Polymer screening – Part II Supersaturation Solid-state screening
(miniaturized assays) screening
Prototype Analytical In vitro dissolution In-depth analytical Selection of the
manufacturing characterization and/or in vivo PK characterization and manufacturing technology
studies solid-state stability and selection of suitable
(stress) testing process parameters
Downstream Investigate the release Investigate the physical Selection of the
processing and supersaturation from and chemical stability of downstream process based
considerations the drug product the drug product on the properties of the DP
(intermediate) (intermediate) intermediate
Selection of excipients
have been attributed to various mechanisms • Initial assessment of physicochemical drug

including anti-plasticization by the polymers properties to evaluate the suitability of the
(Van den Mooter et al. 2001; Oksanen and drug substance for amorphous solid dispersion
Zografi 1990), interactions between the API and development
polymers in solid dispersions (Aso et al. 2002; • Definition of the formulation composition,
Taylor and Zografi 1997; Miyazaki et al. 2004), a e.g., by polymer-excipient screening
reduction in local molecular mobility due to cou- • Manufacturing of prototype formulations
pling between the polymer and API motions (Aso • Characterization of the amorphous solid
and Yoshioka 2006), and an increase in the acti- dispersion
vation energy for nucleation (Marsac et al. 2008). • Downstream processing of the amorphous
On the other hand, maintaining supersaturation formulation
during the dissolution process has been attributed
to the inhibition of API crystallization from the We believe that using a structured approach to
supersaturated solution by the polymer (Gupta amorphous formulation development consisting
et al. 2004; Tanno et al. 2004; Price et al. 2019) of an evaluation of drug substance properties,
and increased equilibrium solubility of the API the selection of a suitable polymer and concentra-
due to complexation with the polymer (Usui et al. tion, and the use of a proper process will enable
1997; Acartürk et al. 1992; Loftsson et al. 1996). the development of amorphous formulations with
Therefore, the ideal formulation should provide optimum solid-state stability and dissolution per-
solid-state stability during downstream formance. An overview of the different stages of
processing and storage as well as maintenance the structured approach is shown in Table 8.2.
of supersaturation during dissolution (generally
about 2–4 h) not only initially but also throughout
the product’s shelf life.
The following steps are described in the chap-
8.3 Initial Assessment
ter as part of the structured development approach
to support and strengthen amorphous The aim of the initial assessment is to assess the
suitability of the drug substance for amorphous
formulations:
solid dispersion development in terms of the
292 S. Page et al.
potential solubility advantage of the amorphous to 75 C with a cooling rate of 20 C/min, and
over the crystalline form and the stability of the finally reheated at 10 C/min to temperatures
amorphous form. Furthermore, it should provide above the melting point. Although it is known
the formulator with some criteria to preselect a that the ability of a compound to form a glass is
manufacturing technology. strongly dependent on the cooling rate applied
and that a minimum cooling rate is required,
they fixed the cooling rate to 20 C/min for the
8.3.1 Initial Assessment in Terms screening method. Compounds, which
of Bioavailability recrystallized upon cooling, are classified as
class I compounds (weak glass formers), whereas
Over the years many researchers tried to predict molecules which showed a crystallization upon
the solubility advantage of the amorphous form of reheating are put into class II. Class III is
a drug substance in comparison to the solubility comprised of compounds which did not show
of the crystalline form in order to estimate the recrystallization at all (strong glass formers).
potential effect on the bioavailability. In a recent Although not explicitly discussed, class I
publication, Paus et al. compared the temperature- compounds may represent a higher risk for the
dependent solubility advantage of an amorphous development of stable amorphous formulations,
drug substance versus its crystalline form while class III compounds are the ideal candidates
predicted using the perturbed-chain statistical for successful ASD formation. This DSC-based
associating fluid theory (PC-SAFT) with the classification system provides an early readout
results obtained from the Gibbs energy difference from a simple experiment to assess ASD feasibil-
(GED) method and experimental data. Overall, ity (with exception of thermally labile
the results obtained from this research group compounds). Overall, Baird et al. investigated
indicated that the predictions from PC-SAFT 51 structurally diverse organic molecules and
were more accurate than the results obtained tried to relate the GFA and GS with physicochem-
from the GED method for the five investigated ical properties of the compounds. Compounds,
compounds (Paus et al. 2015). which are weak glass formers, are usually low
molecular weight compounds with rigid structure,
whereas strong glass formers tend to have a
8.3.2 Initial Assessment in Terms higher molecular weight and a more complex
of Stability structure (Baird et al. 2010). A follow-up investi-
gation by the same research group investigated
Other key factors for successful ASD develop- the crystallization tendency of the same set of
ment are the glass-forming ability (GFA) as well compounds following rapid solvent evaporation.
as glass stability (GS) of the compound. These Overall, the comparison showed identical
properties can either be predicted based on classifications in 68% of all cases. Furthermore,
measured and calculated parameters or they can they could confirm that high molecular weight
be experimentally evaluated. The ease of compounds with a large number of rotatable
vitrification of a liquid on cooling is described bonds are strong glass formers (Van Eerdenburgh
as glass-forming ability, whereas the resistance of et al. 2010).
a material to crystallization is known as glass The relationship between molecular weight
stability in the literature (Baird et al. 2010). and glass-forming ability was confirmed by
Baird et al. developed a classification system other research groups as well. Mahlin and
based on a fast DSC screening method to assess Bergstrom (2013) showed that the glass-forming
the glass-forming ability of organic molecules. ability can be easily predicted from the molecular
The sample is first heated at 10 C/min to approx. weight of the compound. Molecules with > 300 g/
10 C above the melting temperature of the com- mol are expected to be strong glass formers that
pound, held at this temperature for 3 min, cooled can be easily transformed into the amorphous
state. The glass stability was originally related to also of critical importance for the feasibility of
the glass transition temperature, but newer developing a stable amorphous system.
investigations by Alhalaweh et al. (2015) showed
that the physical stability of the amorphous drug
is related to π-π interactions and aromaticity. 8.4 Polymer Screening: Part
I (Theoretical)
8.3.3 Initial Assessment in Terms Due to the thermodynamically unstable nature of

of Manufacturability the amorphous form, many drug substances are
stabilized by forming a polymeric amorphous
The physicochemical properties of the com- solid dispersion (Baghel et al. 2016).
pound, usually evaluated during preformulation, Preformulation data of the drug substance such
can be used to preselect the manufacturing tech- as type and number of hydrogen-bond donor/
nology. The assessment is usually based on: acceptor groups, ionic groups, partition coeffi-
cient, hygroscopicity, and ratio of Tm/Tg provide
• Melting point and heat of fusion of the crystal- some key characteristics for developing a poly-
line form and the glass transition temperature meric amorphous solid dispersion. Friesen et al.
of the amorphous form (if it can be (2008) systematically showed the ability to form
determined) amorphous systems for a large variety of drug
• Thermal stability of the crystalline and amor- substances based on the Tm/Tg ratio and the
phous form octanol-water partition coefficient (log P). Fig-
• Solubility in organic solvents with low and ure 8.1 provides an overview on a couple of
high boiling point drug substances relating the drug load of the
amorphous solid dispersion to the physical stabil-
For solvent-based processes, e.g., spray drying ity of the amorphous solid dispersions as well as
or microprecipitation, the compounds need a rea- their dissolution performance.
sonable solubility in organic solvents. For spray In order to preselect the most promising poly-
drying organic solvents with low boiling point are mer(s) for amorphous drug stabilization, the
used (volatile solvents, e.g., ethanol, acetone), miscibility of the drug substance and polymer as
whereas for the microprecipitation process, well as the potential interactions between the drug
solvents with high boiling point are used (nonvol- and polymer can theoretically be assessed by the
atile solvents, e.g., dimethylacetamide, dimethyl- chemical structure of the compound and the
formamide). Hot-melt extrusion is usually polymers. The phase behavior of polymer-
difficult for compounds with high melting point stabilized amorphous formulations depends on
(> 200 C) or compounds which are thermally the specific interactions between the polymer
unstable. and the drug. A key to the selection of polymer
In summary, the knowledge of the basic drug and optimal drug loading is to maximize the
substance properties is of utmost importance for interactions between the drug and the polymer.
assessing the suitability of the compound for a As shown in subsequent sections, specific
successful ASD development. Additional drug interactions between the drug and polymer are
substance properties might be assessed in case the strongest and, if favorable, have the best
other manufacturing technologies are intended probability of achieving the desired stability.
to be used. As all the necessary data for this The list of polymers can sometimes be further
evaluation are usually determined in the early narrowed by excluding manufacturing
phases of drug development (preformulation), technologies which cannot be pursued due to
there is no additional effort expected. In addition, specific properties of the drug substance. The
the clinical dose of the drug substance is of course data used during this assessment either are based
294 S. Page et al.
Fig. 8.1 Tm/Tg vs log P diagram adapted from Friesen ASD. Group 3 (squares): Drug propensity to crystallize
et al. (2008). Group 1 compounds (circles): High Tg may limit drug loading in ASD (10–35% w/w). Group
allows high drug loadings in ASD ( 50% w/w). Group 4 (triangles): Drug loading can be limited by dissolution
2 (diamonds): Typically 35–50% w/w drug loading in rate (10–35% w/w)
on the structure or are usually measured during polydispersity, and monomer levels, are also
early phases of drug development essential and should be evaluated during develop-
(preformulation). The key steps considered dur- ment mainly as part of critical material attributes
ing initial assessment include: during the quality by design phase (QbD) to
ensure consistent processing and performance.
• Assessing the miscibility of drug and polymer
Some of the key factors related to polymer selec-
(e.g., solubility parameters, glass transition
tion in the design of stable amorphous
temperature, phase diagrams)
formulations are discussed in the following
• Assessing the potential of specific interactions
sections.
between the compound and polymer
• Further restriction of polymers based on the
preselected manufacturing technology
8.4.1 Assessing the Miscibility
of Drug and Polymer
A list of commonly used pharmaceutical
polymers is compiled in Table 8.3, along with
The drug solubility determines the upper limit of
some relevant properties. As one can imagine,
the drug concentration in the drug-polymer mix-
the list of pharmaceutically acceptable polymers
ture, in which the drug exists in a molecularly
for amorphous formulations is somewhat limited.
dispersed state and no phase separation or crys-
Although new polymers are being added continu-
tallization will occur during storage. Any drug
ously to the list, a mechanistic understanding of
amount higher than that solubility will exist in
polymer properties needed to stabilize the amor-
the metastable state and is prone to revert to a
phous system remains somewhat elusive. Other
low-energy crystalline state under normal stresses
properties, such as composition of polymer,
of temperature, pressure, and humidity.
8
Table 8.3 Relevant properties of common pharmaceutical polymers
Hygroscopicity
δ pH (moisture at 75%RH/
Polymer Tg (or Tm) ( C) MW (g/mol) (MPa0.5) solubility RT) Comments
Cellulose based
Hypromellose 2910 170–180 10,000–50,000 23.8 1–10 ~10% Used in Sporanox™
Hydroxypropylcellulose 100–150 80,000 31.5 1–10 12% (at 84% RH) Thermo-reversible gel
EFa
Hydroxyethylcellulose LFa 95,000 31.0
Hydroxyethylcellulose 1,150,000
HFa
Hypromellose acetate 111–115 55,000–93,000 40.5 >5.5 7–8% Can stabilize due to hydrophobicity and possibility of
succinate (HPMCAS), LFb, forming colloidal structures in aqueous solutions
c
HPMCAS, MF b,c 111–115 55,000–93,000 31.2 >6.0 6–7%

HPMCAS, HFb,c 111–115 55,000–93,000 – >6.5 5–6%
Cellulose acetate 160–170 (192) N/A 27 >6.0 7–8%
phthalateb
Cellulose acetate butyrated, 130 (155–165) 30,000 28.7 Negligible N/A
e
Structured Development Approach for Amorphous Systems
Cellulose acetateb 170–190 30,000–60,000 25.8–26.2 N/A N/A

(230–300)
Hypromellose phthalateb,f 133–137 (150) 20,000–200,000 28 >5.0 7–8%
Ethyl celluloseb 129–133 – – Insoluble 3% Controlled release
PEG or PEG copolymer (semicrystalline)
PEG 6000 (55–63) 6000 24.0 1–10.0 0.90% Primarily provides crystalline solid dispersions
PEG 35000 S (64–66) 35,000 24.0 N/A Virtually
nonhygroscopic
Poloxamer 188 (52–57) 7680–9510 23.7 1–10 <0.5%
Poloxamer 407 (52–57) 9840–14,600 – 1–10 <0.5%
Solutol HS 15 (Solidification 344.53 – 1–10 N/A
at 25–30)
Vinylpyrrolidone based
PVP K30b 175 30,000–50,000 27.7 1–10 40% Used in several commercial melt extruded formulations
PVP K90b 180 100,000 27.7 1–10 40% (Rezulin™, Kaletra™, etc.)
Copovidone PVP VA 64b 106 45,000–70,000 25.6 1–10 10% at 50% RH
295
(continued)
296
Hygroscopicity
δ pH (moisture at 75%RH/
Polymer Tg (or Tm) ( C) MW (g/mol) (MPa0.5) solubility RT) Comments
Polyvinyl alcoholb (228, 180–190) 20,000–200,000 – 1–10 –
Crospovidoneb >1,000,000 – Insoluble Max. 60%
Soluplus 60 64,000 22.1 1–10 – Especially suited for melt extrusion
Methacrylate based g
Eudragit E100 48 147,000 19.3 <5.0 N/A Enteric, polymer. Suitable for variety of processing
Eudragit L100-55 110 278,000 23.4 >5.5
Eudragit L100 >150 123,000 23.5 >6.0
Eudragit S100 >150 123,000 7.0
Eudragit RL 70 31,000 Insoluble Controlled release
Eudragit RS 65 30,000 18.9 Insoluble
a
Ashland Inc. Product literature
b
Rowe et al. (2010)
c
Harke Group Pharmaceutical Polymers
d
Eastman Chemical Company
e
Acros Organics
f
Shin-Etsu Chemical Co. Ltd
g
Evonik Pharma Polymers Literature
S. Page et al.
Furthermore, these systems may not provide con- calculate the solubility parameter as shown by
sistent dissolution (or in vivo performance) due to Gupta et al. (2011). The authors see a clear advan-
the chaotic nature of the reversion process. tage of these MD simulations as the solubility
Increasing the drug loading further, e.g., above parameter can be calculated as a function of tem-
the miscibility, will lead to spontaneous phase perature, and additional functional groups as well
separation and further down the line to crystalli- as secondary interactions that are not covered in
zation of the compound. the group contribution methods can also be
Although significant efforts have been made to calculated.
understand and determine drug solubility (crys-
talline drug) and miscibility (amorphous) in
Glass Transition Temperature
polymers, it still remains a challenge to estimate
Amorphous solids are frequently characterized by
these values due to low diffusivity and high
the glass transition temperature (Tg) that
molecular weight even at high temperatures or
corresponds to the temperature at which an amor-
the relaxation of the amorphous system at lower
phous material undergoes a transition from a
temperature (Qian et al. 2007; Marsac et al. 2008;
“glassy state” to a “rubbery state.” Unlike a melt-
Huang et al. 2008). Commonly used methods for
ing endotherm, this transition is a second-order-
estimating the solubility of drugs in polymers
like transition and is associated with continuous
include solubility parameter calculations, molec-
changes (as opposed to abrupt changes) in ther-
ular modeling, thermodynamic modeling (Prudic
modynamic properties, such as heat capacity, vis-
et al. 2014), molecular dynamic simulation
cosity, entropy, and volume. Due to the nature of
(Gupta et al. 2011), and Flory-Huggins interac-
the transition, its measurement is sensitive to
tion parameter (using thermal analysis and
many factors including sample history, rate of
solubility).
cooling, the presence of impurities, and water
content of the sample (Newman and Zografi
Solubility Parameters
2020).
Calculated solubility parameters have been used
A molecular dispersion of an amorphous drug
for decades to predict the miscibility of drugs
with low Tg in a polymer with high Tg will lead to
with polymers (Hancock et al. 1997). Greenhalgh
an ASD with a Tg intermediate of the Tg of the
et al. observed miscibility between ibuprofen and
two components. As a general guiding principle,
several excipients when the drug substance and
amorphous systems with high Tgs are preferred to
the polymer had a difference in the total solubility
improve stability as they can exist in a glassy state
parameter of less than 7 MPa1/2 and immiscibility
at room temperature, which has substantially high
when the difference in the solubility parameter
viscosity (>1013 P) limiting the configurational
was above 10 MPa1/2 (Greenhalgh et al. 1999).
changes and rendering the system immobile. A
There exists a significant body of evidence
low-Tg system can be considered only if there are
suggesting that the ranges suggested by
significant interactions between the materials
Greenhalgh et al. (1999) could provide good
(Chokshi et al. 2008).
guidance. Based on this, an improved group con-
Based on free volume theory, the Tg of a
tribution parameter set for hot-melt extrusion
mixture as a function of polymer concentration
application was established by Just et al. (2013).
is generally expressed as a weighted average of
Furthermore, the absolute difference between sol-
the Tgs of the pure components and can be calcu-
ubility parameters of a drug and polymer should
lated using the Gordon-Taylor equation (Gordon
not be considered as an exclusion criterion
and Taylor 1952):
because other aspects, such as formation of ionic
interactions, may help overcome the solubility w1 T g1 þ kw2 T g2
Tg ¼
limitations. w1 þ kw2
With increasing computational power, molec-
ular dynamic (MD) simulations can be used to
298 S. Page et al.
In this equation, Tg is the Tg of the blend, Tg1 molecular weight analogue of the polymer
and Tg2 are the Tgs of the pure components, w1 (Knopp et al. 2015). A comparison of all these
and w2 are the weight fractions of each compo- methods showed that the magnitude of the
nent in the blend, and k is a constant calculated predicted solubilities from the solubility in the
using true density (ρ) and the difference between liquid analogue correlated well with the results
expansion coefficients of the melt and the glass from the recrystallization and melting point
(Δα) (k ¼ ρ1Δα1/ρ2Δα2). For early assessments, depression method (Knopp et al. 2015). In a
k is generally considered a constant. A simplified follow-up investigation, Knopp et al. (2016) criti-
version of this equation for an ideal system, when cally reviewed the use of the melting point
k ¼ 1, is known as the Fox equation. depression method for calculating the interaction
As an initial assessment and rough rule of parameter χ. The authors performed a statistical
thumb, polymers with high Tgs are preferred analysis of the method proposed by Lin and
especially those that can provide amorphous Huang and showed that the predicted miscibility
solid dispersions with a single composite Tg of curve could not be trusted with statistical confi-
75 C or higher (i.e., 50 C above the storage dence. They rather propose that the DSC
temperature). measurements, which are used to make
miscibility predictions, should be examined by
Prediction of Phase Diagrams deriving an objective function. This would result
In recent years several attempts are made in order in an unbiased, minimum variance property of the
to construct temperature-composition phase least square estimator. Nevertheless, they also
diagrams to support the rational selection of state that additional arguments are needed to
drug load and polymer type during the formula- prove the underlying physical assumptions, such
tion design phase. The temperature-composition as the temperature dependence of χ in order to
phase diagram usually contains the solubility fully believe in the predictions (Knopp et al.
curve, the miscibility curve, and the glass transi- 2016). Another alternative in order to determine
tion curve. Tian et al. have predicted the phase the Flory-Huggins interaction parameter χ is the
diagram for a solid dispersion containing use of molecular dynamic simulations or alterna-
cinnarizine (CIN) and Soluplus®. They calcu- tive approaches such as the extended Flory-
lated the solubility curve using the solid-liquid Huggins theory as implemented in the Material
equilibrium (SLE) equation and the miscibility Studio Blends module from Accelrys Inc. The
curve, which corresponds to the spinodal curve, latter was used by Pajula et al. to predict the
based on the Flory-Huggins theory (Tian et al. phase stability of small-molecule binary mixtures
2015). The Flory-Huggins interaction parameter (Pajula et al. 2010).
χ was calculated using the solubility parameter Another approach to generate phase diagrams
determined by the group contribution method. is the use of thermodynamic models such as
Finally, the glass transition curve was estimated PC-SAFT. It has been used, for example, for the
using the Fox equation. There was a good agree- prediction of the phase behavior of APIs and
ment between the theoretically calculated and polymers in the dry state (Prudic et al. 2014) as
experimentally determined solubility of CIN in well as in presence of humidity (Lehmkemper
Soluplus® as well as the miscibility level (Tian et al. 2017). A theoretical example of such a
et al. 2015). phase diagram is depicted in Fig. 8.2.
Alternatively, thermal analysis methods can be
used for the determination of the Flory-Huggins
interaction parameter χ, including the recrystalli- 8.4.2 Specific Interactions
zation method, the dissolution end point method,
and the melting point depression method, or the It is well recognized that interactions between a
estimation of the solubility of a drug in a polymer drug and polymer have a significant effect on the
from the solubility of the drug in a liquid low stability of a high-energy amorphous system. The
Fig. 8.2 Theoretical phase

diagram of a polymer-drug
mixture with the combined
glass transition temperature
(dashed line)
Table 8.4 Typical bond energy and relative strength of different intermolecular forces
Type of interactions Bond energy (kJ/mol) Approximate relative strength
Ionic interactions 850–1700 1000
Hydrogen bonding 50–170 100
Dipole-dipole interactions 2–8 10
Van der Waals interactions 1 1
Adapted from Yang and Han (2008)
interactions between a drug and a polymer can Nevertheless, the presence of ionizable groups
result from several types of intermolecular can provide secondary structures that are steri-
interactions, e.g., hydrophobic interactions (due cally stabilized in addition to reducing mobility.
to dispersions forces), hydrogen bonding, or elec- Yoo et al. (2009) investigated the miscibility of
trostatic (polar or induced-dipole) interactions. As polymers and highly crystalline additives. They
shown in Table 8.4, electrostatic interactions, showed that the likelihood of obtaining a miscible
being strong forces, can provide stability to amor- system was the highest in the case where an acid-
phous solid dispersions. base ionic interaction is involved in the formation
An understanding of the interactions and their of the amorphous state. In the absence of ionic
effect on the solubility of a drug in a polymer and interactions, systems with similar solubility
the resultant phase diagram determine the space parameters and partition coefficients showed
within which high-energy systems can provide miscibility. Similar interactions are expected
maximal benefit. A theoretical basis for the calcu- between drugs and polymers. Additionally,
lation of thermodynamic solubility and kinetic Forster et al. (2001a, b) also showed that ionic
miscibility has been discussed by several interactions and solubility parameters play a role
researchers (Zhao et al. 2011; Janssens et al. in the formation of the amorphous state. The
2010; Paudel et al. 2010). development of modern ASDs broadened the
As shown in Table 8.3, polymers with ioniz- interaction screen to targeted interactions between
able groups present opportunities for the forma- polymer, drug, and additives that stabilize the
tion of hydrogen bonding and/or ionic amorphous form in the solid state and the super-
interactions. It has been generally recognized saturated state upon aqueous dispersion (Kasten
that proton transfer and exact stoichiometry may et al. 2019; Newman et al. 2018; Ditzinger et al.
not be a requirement for drug-polymer blends due 2019b).
to large differences in the molecular weight.
300 S. Page et al.
8.4.3 Hygroscopicity and Water interfere with amorphous systems in many differ-
Activity ent ways. Therefore, polymers with low water
activity are expected to provide the best stability
Hygroscopicity of a polymer plays an important for amorphous systems (Nair et al. 2020).
role in determining the physical stability of an
amorphous product, especially during storage.
Adsorption of water can act in many ways to 8.4.4 Further Restriction of Polymers
destabilize amorphous systems, such as by weak- Based on the Manufacturing
ening the interactions between the drug and poly- Technology
mer, lowering the solubility or miscibility of the
drug in the polymer, and lowering the glass tran- Once the formulator has preselected those
sition temperature of the polymer (plasticization) polymers that provide miscibility and/or specific
(Newman and Zografi 2020). Although the effect interactions with the drug substance, he/she can
of water on the interactions between the drug and narrow down the list further by considering the
polymer is not easy to assess, several authors preselected manufacturing technology (see Sect.
have evaluated the effect of water on drug solu- 8.3.3.). For solvent-based processes, the polymer
bility and glass transition temperature. Rumondor needs to have a reasonable solubility in organic
et al. (2009b) showed that a small amount of solvents (e.g., for spray drying in organic solvents
water can significantly lower the solubility of with low boiling point and for the micropreci-
felodipine in PVP. Similarly, in another study, it pitation process solvents with high boiling
was shown by differential scanning calorimetry point). For hot-melt extrusion, the polymer
(DSC), atomic force microscopy (AFM), and should have a reasonable low melt viscosity in
transmission electron microscopy (TEM) the temperature range needed for processing the
measurements that water can irreversibly disrupt drug substance, and it needs to be thermally sta-
the favorable interactions between a drug and a ble. In summary, the knowledge of the basic
polymer, thus resulting in phase separation that polymer properties is very helpful for assessing
eventually leads to crystallization (Marsac et al. the suitability of the polymer for a certain
2008). A modified Flory-Huggins equation con- manufacturing technology. Additional properties
sidering a water-drug-polymer ternary system might be considered for newer technologies such
was developed by Rumondor et al. (2009b) to as electrospinning or KinetiSol®.
estimate the effect of water on the interaction
parameters. It was shown that ingress of water
can weaken the interactions between hydrophobic 8.4.5 Summary of Initial Assessment
drug and hydrophilic polymer, thus resulting in
drug-rich phases that induce crystallization. In a As described in this section, it is very important to
study where different polymers were compared, know the physicochemical properties of the drug
the authors concluded that use of a hydrophobic substance with respect to its thermal behavior,
polymer, such as hypromellose acetate succinate hydrogen-bond acceptor and donor groups,
(HPMCAS), can be beneficial over more hydro- molecular weight, hydrophobicity, melting
philic polymers, such as povidone or copovidone, point, glass transition temperature, solubility
to achieve better stability. A similar finding has parameter, hygroscopicity, logP, and Tm/Tg
been shown by Friesen et al. (2008) where the use ratio. The basic preformulation data can help
of HPMCAS was shown to provide better super- assess the degree of difficulty a molecule may
saturation during dissolution that was related to present in being converted to and maintained in
the formation of aggregated structures by an amorphous form (Friesen et al. 2008). To best
HPMCAS (Friesen et al. 2008). The aforemen- match the properties of a drug substance, it is also
tioned discussion clearly shows that water can important to know the physicochemical
properties of polymers such as their thermal
behavior, hydrogen-bond acceptor and donor software and database tools, allowing dozens or
groups, molecular weight, hydrophobicity, melt- hundreds of screening experiments in a short
ing point/glass transition temperature, solubility period of time. As discussed before, the develop-
parameter, and hygroscopicity. A comprehensive ment of suitable amorphous systems requires
summary of polymer properties is provided in miniaturization of the preparation methods and
Table 8.3. The initial selection of polymers can characterization tools including the analytical
be based on solubility parameters, functional methods to evaluate the API supersaturation
groups, and thermal behavior including melting upon dissolution, as well as to assess the solid
point and Tg. Thermal stability and solubility in state and physical stability of the systems.
organic solvents are critical for the selection of The ASD processing technologies can gener-
the processing method (further discussed in Sect. ally be divided into two main classes, solvent-
8.6). The initial screening will help to narrow based and fusion-based methods. Solvent-based
down the list of polymers, e.g., considering simi- technologies include spray drying, fluid bed
lar solubility parameters or the possibility of ionic layering, coprecipitation (e.g., MBP ¼ micropre-
interactions between a weakly basic drug and cipitated bulk powder), etc. Fusion-based
enteric polymers. The next step in the process of technologies include melt extrusion, melt granu-
structured development is the conduction of lation, KinetiSol® technology, etc. The selection
small-scale experiments to further narrow down of an appropriate processing technology strongly
the choice of polymer and drug loading that has depends on the API properties. Understanding the
the potential to provide the best stability and physicochemical properties of API, therefore,
supersaturation. will streamline the amorphous formulation
screening design. Most of ASD screening
approaches described in the literature are focused
8.5 Miniaturized Methods for ASD either on the assessment of the supersaturation
Screening potential of the API-polymer combination in
solution or on the evaluation of amorphous drug
After a careful evaluation of drug and polymer stabilization in the solid state. So far, only a few
properties and in silico assessment, the next phase papers have considered the use of a combination
in the development of amorphous dispersions is of different miniaturized assays to identify
the evaluation of different compositions by polymers with appropriate dual function, that is,
miniaturized or small-scale experiments to nar- (i) generation and maintenance of supersaturation
row down the choice of polymer and to get an and (ii) stabilization of the amorphous drug in the
assessment of drug loading. Nowadays, formulation matrix. Examples for such combined
miniaturized experimental screening systems are approaches include the 96-well plate vacuum dry
being utilized for the evaluation of suitable system for ASD screening described by Chiang
polymers and additives (or mixtures thereof) for et al. (2012), the SPADS approach (Wyttenbach
amorphous formulations. These systems use typi- et al. 2013), or the MiCoS method (Hu et al.
cally less than 10 mg of compound per test sam- 2013). An overview of various miniaturized
ple and work in the 96-well format. The use of ASD screening methods is shown in Table 8.5.
these miniaturized systems has the potential to At an early development stage, when API sup-
facilitate amorphous formulation development ply is limited, a step-by-step ASD screening strat-
by saving valuable time and resources (manpower egy is recommended: (1) Drug candidates
and compound). Ideally, the miniaturized assays suitable for spray drying (SDP candidates) with
are partly or fully automated by an assembly of good solubility in volatile solvents (API solubility
different modules, for instance, robotic systems, in volatile solvents > 20 mg/ml) are first screened
high-throughput analytical systems, and specific with solvent casting methods for selecting
302 S. Page et al.
Table 8.5 ASD miniaturized screening methods

ASD processing
96-well technology
Supersaturation Solid-state (parallel) test assessment
Method References screening screening reported SDP HME MBP
Solvent shift Guzmán et al. (2007) Y N Y Y Y Y
Warren et al. (2010)
Solvent casting Singh et al. (2007) Y Y Y Y Y/N N
Chiang et al. (2012)
Wyttenbach et al. (2013)
Coprecipitation Hu et al. (2013) Y Y Y N N Y
Melt fusion Forster et al. (2001a, b) N Y N Y/N Y N
DSC Zhao et al. (2011)
Hot stage Sarode et al. (2013)
microscopy Agrawal et al. (2016)
Freeze drying Moes et al. (2011) Y Y N Y/N Y/N N
Spin coating Lee and Lee (2003) N Y N Y Y/N N
Konno and Taylor
(2006), Konno et al.
(2008)
Y yes
N no
Y/N, possibly; under certain conditions
appropriate polymers and drug loads. (2) HME 8.5.1 Supersaturation Screening
candidates with low melting point (MP < 200 C)
can be evaluated by small-scale fusion-based The concept of generating and maintaining super-
methods, e.g., drug-polymer mixtures treated by saturation has been described as the “spring and
heat-cool-heat cycling in DSC experiments. If parachute approach” by Guzmán et al. (2007). In
HME candidates have good solubility in volatile the case of a molecular dispersion (solid solu-
solvents, solvent casting methods can also be tion), release of drug molecules is dictated by
used. (3) Coprecipitation (MBP) candidates dissolution of the hydrophilic carrier (spring func-
(with poor solubility in volatile organic solvents tion) and leads to a supersaturated state of the
and high melting point) with good solubility in drug in solution. Two different miniaturized
nonvolatile water-miscible solvents (API solubil- methods have been applied to assess the supersat-
ity in nonvolatile solvents ideally > 50 mg/ml) are uration potential of excipients with poorly water-
screened by miniaturized coprecipitation screen- soluble drugs: cosolvent quenching (solvent-shift
ing (e.g., MiCoS) for ASD feasibility. In addition, method) and amorphous film dissolution (solvent
the generally very simple and fast solvent-shift casting method). Examples for both methods
method can be used as a supportive method for all described in the literature are shown in Table 8.6.
ASD processing technologies (e.g., SDP, HME, The cosolvent quench (solvent-shift) method
MBP) for a first assessment of the supersaturation is a common method used for initial polymer
potential of different compositions. If drug screening. In this method, drug is dissolved in a
candidates are amendable for different ASD water-miscible solvent with high drug solubility.
processing technologies and if API supply is not A small aliquot of the solution is then dispersed in
the limiting factor, a parallel screening approach the aqueous phase to create a supersaturated sys-
with different screening methods (solvent casting, tem. In order to determine the extent of drug
melt fusion, coprecipitation, etc.) can also be precipitation, the concentration of the dissolved
adopted. Some of the miniaturized ASD screen- drug within the aqueous phase or the mass of drug
ing methods are described in more detail in the precipitated can be assayed or measured indi-
following subsections. rectly by turbidity measurement. Alternatively,
Table 8.6 Miniaturized methods for the assessment of the supersaturation potential
Method of
References Screening method analysis Compound(s) tested
Cosolvent quench (solvent-shift) method
Guzmán et al. 96-well microplate format; drug dissolved in Nephelometry Celecoxib
(2007) sodium hydroxide solution
Vandecruys et al. 10 mL format; dimethylformamide and UV 25 different (not specified) drug
(2007) dimethylacetamide used as solvent spectroscopy candidates
Janssens et al. 10 mL format; dimethylformamide used as UV Itraconazole
(2008) solvent spectroscopy
Curatolo et al. 10 mL syringe/filter method; HPLC Different drugs and drug
(2009) dimethylacetamide used as solvent candidates
De Maesschalk 96-well microplate format; dimethylacetamide Nephelometry J&J 1
et al. (2010) used as solvent
Warren et al. 96-well microplate format; propylene glycol Nephelometry Danazol
(2010) used as solvent
Yamashita et al. 96-well microplate format; drug dissolved in HPLC Model compound X
(2010) simulated gastric fluid (SGF)
Amorphous film dissolution (solvent casting) method
Singh et al. 96-well microplate format HPLC Indomethacin, haloperidol,
(2007) progesterone
Barillaro et al. 4 mL format HPLC Phenytoin
(2008)
Shanbhag et al. 96-well microplate format UV JNJ-25894934
(2008) spectroscopy
Swinney et al. 96-well microplate format UV Different (not specified) drug
(2009) spectroscopy candidates
De Maesschalk 96-well microplate format UV J&J 1
et al. (2010) spectroscopy
Chiang et al. Amorphous films in 96-well microplate format HPLC Acetaminophen,
(2012) indomethacin, celebrex,
griseofulvin, compound A
Wyttenbach et al. Amorphous films in 96-well microplate format UPLC CETP(2) inhibitor
(2013)a
a
SPADS approach
the film dissolution method involves parallel dis- amorphous system include amorphous sample
solution screening of solid dispersions with dif- preparation by solvent casting, coprecipitation,
ferent compositions and drug loads prepared by melt fusion, freeze drying, or spin coating
solvent casting. Amorphous drug films are (Table 8.5) and analysis for recrystallization or
prepared from mixtures of drug and excipient(s) phase separation by different analytical
dissolved in a volatile organic solvent. The techniques, e.g., by polarized light microscopy,
organic solvent is then evaporated resulting in a DSC, XRPD, FT-IR spectroscopy, Raman spec-
thin film of the formulation. The dissolution troscopy, or atomic force microscopy (AFM).
medium is added and the drug concentration is Most commonly, amorphous films are prepared
determined as a function of time. by evaporation of organic drug-excipient
solutions on different carrier systems, such as
glass slides, cover slips, 96-well microplates, or
8.5.2 Solid-State Screening aluminum pans. The stability of the amorphous
systems is assessed by reanalysis of the samples
Miniaturized methods used to experimentally after storage at accelerated conditions (tempera-
evaluate the homogeneity and stability of an ture and humidity). Different miniaturized
304 S. Page et al.
Table 8.7 Miniaturized methods for the evaluation of amorphous drug stabilization
Method of
References Screening method analysis Compound(s) tested
Methods to determine drug-polymer miscibility/solubility/stability
Forster et al. Quench cooled melts in aluminum DSC and hot Indomethacin, lacidipine
(2001a, b) pans and drug-excipient blends on stage microscopy
glass slides
Lee and Lee Spin-coated films on silicon wafer Light microscopy Sulfisoxazole, griseofulvin, ketoprofen,
(2003) chips flurbiprofen
Konno and Spin-coated films on glass cover Polarized light Felodipine
Taylor (2006), slips and ZnS disks microscopy,
Konno et al. FT-IR
(2008) spectroscopy
Singh et al. Films in 96-well microplates XRPD Indomethacin, haloperidol,
(2007) progesterone
Swinney et al. Films in 96-well microplates and in Birefringent Different (not specified) drug
(2009) aluminum pans imaging, XRPD, candidates
DSC
Van Spin-coated films on glass cover Polarized light Benzamide, phenacetin, flurbiprofen,
Eerdenbrugh and slips microscopy flufenamic acid, chlorpropamide,
Taylor (2010) chlorzoxazone, bifonazole, lidocaine
Lauer et al. Film fracture surfaces on glass Raman NK1 receptor antagonist, CETP
(2011) slides microscopy, inhibitor
AFM
Chiang et al. Amorphous films in 96-well XRPD Acetaminophen, indomethacin,
(2012) microplate format celebrex, griseofulvin, compound A
Hu et al. (2013) Coprecipitate in 1 ml glass vials in XRPD, Raman Felodipine, glybenclamide, nifedipine
96-well microplate format and
96-well filter plates
Wyttenbach Films in aluminum pans (96-well FT-IR CETP(2) inhibitor
et al. (2013)a microplate format) and film microscopy,
fracture surfaces on glass slides AFM
a
SPADS approach
methods to evaluate amorphous drug stability are assessed on the second heating scan after the
presented in Table 8.7. system has initially been heated beyond the melt-
With respect to the amorphous drug stability, ing point of the drug. If the drug is thermally
drug-polymer miscibility (homogeneity) is essen- labile, amorphous drug and polymer can be
tial because immiscibility can result in the forma- casted in a DSC pan followed by evaporation of
tion of drug-rich domains that may be prone to the solvent. A general rule of thumb is that a
recrystallization. Other important factors strongly single Tg suggests miscibility, observation of the
affecting the stability of amorphous systems two Tgs corresponding to individual components
include solid-state solubility and drug-polymer suggests complete immiscibility, and two Tgs in
molecular interactions (Konno and Taylor 2006; between the two individual Tgs suggest partial
Konno et al. 2008; Wyttenbach et al. 2013). miscibility.
Qualitatively, the miscibility of drug and poly- This DSC technique is simple and rapid in
mer can be assessed by simple DSC obtaining essential information on miscibility
measurements, in which mixtures of drug and but is very crude in that quantitative miscibility
polymer are mixed and treated by heat-cool-heat information is lacking and the miscibility is given
cycling in a DSC pan to observe the change of only near the glass transition temperature. In
glass transition temperatures (Sarode et al. 2013). addition, if the drug and polymer have similar
Generally, miscibility of drug and polymer are Tgs or the change in heat capacity around the Tg
is small, then this method does not provide much promising prototype amorphous compositions
insight into the miscibility of the drug and the and drug loads that will be subjected to small-
polymer. scale solid dispersion preparation for further
in vitro and in vivo assessment (dissolution and
solid-state characterization, animal PK studies).
8.5.3 The SPADS (Screening These scaled-down processes include mini-
of Polymers for Amorphous extrusion for hot-melt extrusion, mini-fluid bed
Drug Stabilization) Approach: systems for coated bead systems, or mini-spray
Example for a Combined ASD dryers if spray drying will be an option. Based on
Screening Approach our experience, this step is very important since
various amorphous solid dispersions often have
The SPADS approach has been developed at different stability and biopharmaceutical
F. Hoffmann-La Roche Ltd. for solid dispersion properties based on how they were generated,
screening by Wyttenbach et al. (2013). It even when containing the same composition and
combines the assessment of supersaturation drug load.
potential, the evaluation of drug-polymer
miscibility, and the stability of amorphous
systems. The aim is to rapidly identify prototype 8.6 Selection of Most Suitable
amorphous compositions suitable for preclinical Technology
studies and early-stage clinical trials.
The SPADS approach consists of three differ- After having identified suitable polymers and
ent miniaturized assays: (1) SPADS dissolution, drug loads, selecting the right processing technol-
(2) SPADS imaging, and (3) SPADS interaction ogy is essential. The selection of the most appro-
assay. It is a two-step approach; in a first step, priate process for the manufacture of a solid
drug dissolution and supersaturation are dispersion depends on the physicochemical
evaluated on amorphous drug-polymer films properties of the API and the polymer (solubility,
prepared in 96-well plates by solvent-based film thermal stability, melting point, and Tg). The
casting. The drug-excipient combinations that manufacturing process directly impacts the com-
emerge from this initial screening as positive are plexity of further downstream processing (e.g.,
then subjected to a more thorough study of their particle size and bulk density) and properties of
solid-state properties in a second step. Optical the finished drug product (e.g., stability and dis-
microscopy and AFM are applied to analyze the solution). Additional criteria that should be taken
molecular homogeneity and stability of promising into consideration are the availability of the
amorphous API-polymer combinations (SPADS equipment, the robustness of the manufacturing
imaging assay), while Fourier transform infrared process, the impact on the cost of goods (energy
(FTIR) microspectroscopy is used to study consumption and equipment footprint) and the
molecular interactions (SPADS interaction intellectual property considerations. These
assay). properties should be assessed to allow the formu-
lator to rank order different technologies. In a
second step, the ranking should be confirmed by
8.5.4 Summary of the Small-Scale manufacturing trials on small-scale equipment.
Experiments: Selection A physically stable solid dispersion of an
of Polymer and Drug Load amorphous API in a polymer can be achieved if
the drug and the polymer are intimately mixed at
All analytical results obtained from initial assess- the molecular level, i.e., a solid solution.
ment, miniaturized or small-scale experiments, Molecular-level mixing is achieved either by dis-
are collected, assessed, and compared. The col- solution of both components in a solvent (solvent-
lected data are used to preselect the most based technologies) followed by solvent removal
306 S. Page et al.
or by directly mixing both components as liquids for spray drying can have an impact on the physi-
as is accomplished by melting (fusion) methods cal stability as well as the in vitro dissolution
(Chiou and Riegelman 1971). Currently the most kinetics of the resulting spray-dried solid
relevant technologies for the manufacture of solid dispersions. Further insights into the effect of
dispersions in the pharmaceutical industry are the solvent on long-term physical stability were
(1) spray drying, (2) melt extrusion, and evaluated using the thermodynamic model
(3) coprecipitation. The following section PC-SAFT (Luebbert et al. 2018; Dohrn et al.
discusses these technologies including a brief 2020a, b, 2021), as well as light scattering and
overview, schematic of the physical processes, viscometric techniques (Defrese et al. 2020). The
and a summary table with the associated results clearly showed that solvent-induced amor-
advantages and disadvantages (Table 8.8). Sev- phous phase separation occurs regardless of the
eral other technologies are described in the litera- API when the polymer-solvent system undergoes
ture to produce amorphous solid dispersions such phase separation (Luebbert et al. 2018) and that
as spray granulation, fluid bed layering, spray although API and polymer were initially
congealing, co-grinding, spin coating, electro- completely dissolved in a solvent mixture,
static spinning, microwave technology, liquid-liquid phase separation can occur during
KinetiSol® dispersing, ultrarapid freezing, and drying due to different solvent volatilities
supercritical fluid processing (He and Ho 2015), (Dohrn et al. 2020b) leading to inhomogeneous
but with the exception of fluid bed layering they final ASDs. In a follow-up study (Dohrn et al.
have not yet been applied for marketed products. 2020a), the phase behavior of pharmaceutically
relevant polymer-solvent mixtures was
investigated, and a number of appropriate
8.6.1 Overview of Key Manufacturing solvents for PVP and PVP VA 64 were identified.
Technologies Unfortunately, all of the tested solvents including
acetone and water were immiscible with
Spray Drying HPMCAS. Using the PC-SAFT model, solvent
Spray drying is well established as an industrial mixtures of certain acetone/water ratios were
process for transforming solutions, emulsions, or predicted to completely dissolve HPMCAS
suspensions into a dry powdered form and can be (Dohrn et al. 2020a). On top of this, a suitable
performed over a wide range of scales (from solvent or solvent mixture must meet the follow-
milligrams/grams to tons of drugs) (Friesen et al. ing criteria (Table 8.9).
2008). In principle, the process consists of 4 unit Besides the solvent selection, also process-
operations: spray solution preparation, solution related criteria have to be considered. In order to
atomization/spray drying, collection of the replace the error-prone visual inspection and
spray-dried powder, and removal of the residual offline high-performance liquid chromatography
solvent/secondary drying. A thorough discussion (HPLC), Lee et al. (2019) developed an online
of spray-drying technology and its application to NIR method, which is capable of defining the end
the formulation of poorly water-soluble drugs is point of the spray-drying solution preparation and
provided in Chap. 10 or can be found in review accurately predicting the API concentration. Once
articles such as Ziaee et al. (2017). the feed solution is manufactured, it is pumped to
As spray drying is a solvent-based an atomizer to produce droplets inside the drying
manufacturing process, the selection of a suitable chamber. As the droplet size distribution has a
organic solvent or solvent mixture is one of the direct influence on the size and porosity of the
key parameters. In recent years the impact of the resulting particles, a lot of effort is put into the
solvent or solvent system was investigated in development of predictive models (Poozesh et al.
more detail. Al-Obaidi (2009), Hugo et al. 2018) and characterization of the rheological
(2013), and Paudel and Van Den Mooter (2012) properties of the spray solutions to predict the
described that the solvent or solvent mixture used droplet size distribution (Porfirio et al. 2021).
Table 8.8 Pros and cons of different processing technologies

Process Pros Cons
Spray drying Rapid removal of solvent and fast Use of organic solvents (environmental
solidification safety)
Equipment available from lab to full-scale Difficulty to identify a common volatile
commercial production solvent for API and polymer
Relatively low-temperature processing Residual solvent content potentially
feasible for highly volatile solvents (reducing requires secondary drying process (usually
thermal stress and degradation of the API) batch process)
Continuous processing High manufacturing cost
Generally results in very fine particles with
low bulk density and poor flow properties
(increased complexity of downstreaming
process)
Powder properties depend on
manufacturing scale
Melt extrusion Non-solvent processing (eliminate the need High energy mainly related to shear forces
for solution preparation and removal steps) and temperature (high thermal stress in case
Customizable process (screw/die design, of high melting APIs)
temperature profile, and solvent addition) High melt viscosity causing torque
Effect of humidity and oxygen can be almost limitations. Use of processing aids
completely eliminated (plasticizers) required
Robust process control, application of PAT Limited selection of pharma grade polymers
tool and easy scale-up High density and low porosity of the
Continuous process thermoplastic extrudates reduce the
Small equipment footprint compactibility of the material
Coprecipitation Suitable for compounds that cannot be Requires polymers with differentiated
processed by spray drying (due to low solubility in solvent and antisolvent
solubility in volatile organic solvents) or melt Limited selection of pharma grade polymers
extrusion (due to high melting point with Amorphous intermediate is exposed to
thermal degradation) antisolvent, often water, for substantial
Amenable for continuous processing amount of time
Weak bases (and acid drugs) exhibit
significant solubility in acidic (and basic)
solvents
May require multiple washings to remove
solvents
Table 8.9 Selection criteria for solvents or solvent mixtures for spray drying
Selection criteria Description
High drug and polymer A high drug loading of the spray solution (feed solution) is a prerequisite in order to
solubility reduce the amount of solvent needed, thereby reducing the overall processing time
and the environmental burden. For a commercially viable process, solubility of
approx. 100 mg/mL and above is preferred; however, the solubility of 50 mg/mL is
considered essential
Low viscosity of the feed To improve atomization and facilitate solvent evaporation. A highly viscous feed may
not allow for sufficient atomization, leading to insufficient evaporation of the solvent
after drying that may impact the stability of the solid solution (plasticizing effect of
residual solvents)
Low toxicity of the solvent(s) Preferable ICH Q3C Class 2 and 3 solvents for personal and environmental safety and
to meet specifications in case of high residual solvent content
Low boiling point of the To ease evaporation of the solvent: drying can be performed at lower temperature,
solvent(s) lower thermal stress, and less thermal degradation of the drug
Solvent mixtures – preferable Solvents that can form azeotropes are preferred because non-azeotropic solvent
azeotropes mixtures will have different evaporation rates during drying that may lead to either
phase separation or crystallization of the API during evaporation
Chemical stability of the API Sufficient chemical stability of the API in the feed solution is required
in solvent(s)
308 S. Page et al.
Various approaches have been used in the past in screw design is self-wiping, where it minimizes
order to scale up spray-drying processes. Further, the stationary zone and prevents localized
more in-depth information are provided, for overheating of material in the extruder. The
instance, in Gil et al. (2010), Thybo et al. screw is typically modular and divided into
(2008), Dobry et al. (2009), and Kalb three sections along the lengths of the barrel:
et al. (2013). feeding, melting or compression, and metering.
Spray-dried material is often very fine and In the feeding section, the material is transferred
contains a high amount of residual solvents. from the hopper to the barrel. In the compression
This may have a negative effect on the down- zone, the polymer usually starts to melt due to the
stream process (flow and density of particles) thermal energy that is generated by shear forces
and the stability of the spray-dried solid imposed by the rotating screw and from conduc-
dispersions (recrystallization of the amorphous tion from the barrel via electrical heating bands.
drug). In addition the residual solvent content The temperature of the melting zone is normally
can easily exceed ICH limits. Solvent reduction set 10–60 C above the melting point of semi-
would require a secondary drying step which very crystalline polymers or the glass transition tem-
often transforms the continuous spray-drying pro- perature of amorphous polymers to ensure
cess into a batch process. In order to improve the consistent flow. The metering zone is designed
properties of the spray-dried material, more to reduce pulsating flow of the molten polymer
advanced spray-dryer layouts with an additional and provide uniform delivery of the material
external fluid bed or an internal fluid bed are through the die. The screw configuration can be
available to improve the drying efficiency modified by changing the screw elements to opti-
(avoiding secondary drying) and to agglomerate mize the shear, residence time distribution, and/or
the fines (Masters 1991). Recently, Shepard et al. product characteristics (Crowley et al. 2007).
(2020) introduced solvent-assisted secondary HME requires pharmaceutical grade polymers
drying as a new more effective process for sec- that can be processed at relatively low
ondary drying. Unfortunately, the effects of meth- temperatures due to the thermal sensitivity of
anol and water, which were used to accelerate the drugs and polymers. For most polymers,
residual solvent removal from the spray-dried temperatures between 120 and 180 C have been
material, on the long-term physical stability used. For a complete discussion of HME technol-
have not been investigated. ogy, the reader is referred to Chap. 9.
The viability of melt extrusion depends on the
Hot-Melt Extrusion ability to form a one-phase solid solution. API
Hot-melt extrusion (HME) is frequently used in and polymers are subject to elevated
the pharmaceutical industry for developing solid temperatures, high pressure, and intensive mixing
dispersion formulations. HME is the process of during the HME process (Patterson et al. 2008).
pumping raw materials with a rotating screw(s) Depending on the process conditions, the crystal-
under elevated temperatures through a die to form line drug either melts or becomes solubilized in
a product of uniform shape. The intense mixing the molten polymer. The latter allows the manu-
and agitation imposed by the rotating screws facture of a solid dispersion at temperatures
result in a uniform dispersion. Generally, the below a drug’s melting point. The recrystalliza-
extruder consists of one (single-screw) or two tion of the drug during the cooling of the
(twin-screw) rotating screws in a stationary bar- extrudate is decelerated due to reduced solute
rel. Twin-screw extruders can be operated in migration and the reduction of nucleation kinetics
corotating or counter-rotating modes. Corotating by the viscous polymer medium. Nevertheless the
twin-screw extruders are widely used since they cooling rate has to be carefully evaluated as it
can be operated at high screw speeds, yielding may impact the quality (physical stability) of the
high output, good mixing, and good conveying obtained extrudates.
characteristics. A fully intermeshing type of
Table 8.10 Key considerations for HME

Key considerations Description
Melting point/Tg/processing Minimum temperatures are required for extrusion. In order to reduce melt viscosity
temperature and facilitating material transfer, the processing temperature should be set
approximately 10–20 C above the melting point of a semicrystalline polymer, or the
Tg of an amorphous polymer (Chokshi et al. 2005). If the temperature is too high,
thermal stresses during melt extrusion may cause degradation of API and polymer
Melt viscosity of the polymer Polymers with low melt viscosities and high thermal conductivity yield a more
efficient melting process (Crowley et al. 2007). If the melt viscosity of the polymer is
too high, it may limit miscibility of the API and the polymer (Forster et al. 2001a, b)
High melt viscosity results in high torque and pressure which may exceed the
capabilities of the equipment. In that case the plasticizing effect of the API has to be
evaluated or use of processing aids (plasticizers) to be considered
Miscibility In order to form a one-phase mixture, the two molten components have to be miscible.
The changes in melting point or Tg as a function of polymer concentration provide the
phase diagram to establish the boundary of solid-state miscibility and help to select the
processing temperature (Chokshi et al. 2005)
Solubility The solubility of the crystalline drug in the polymer is critical for the stability of the
extrudate. Crystallization of a miscible API-polymer system can occur if the solubility
limit has been exceeded (Marsac et al. 2006; Breitenbach 2002). Therefore, the drug
load of the formulation needs to be adapted accordingly
Downstream process Depending on subsequent purpose of the HME strands, the downstreaming process
can have increased complexity. Direct shaping (calendaring) would require a very
steady/non-pulsatile melt flow
Powder blends for capsule filling and tablet compression require pelletization/milling
step of the extrudate strands. Downsizing can be the limiting process step as extrudates
very often show unfavorable milling properties. It can be overcome by increasing
brittleness of HME strands, for example, by injecting supercritical CO2 or introducing
brittle excipients like mannitol or usage of cooled milling equipment (Read et al.
2010). It also has to be considered that the properties of the final drug product can
depend on the particle size of the milled intermediate (Jijun et al. 2010; Feng et al.
2012 and Deng et al. 2013)
The selection of optimal melt extrusion feasible HME process conditions and investigate
conditions depends on the chemical stability of the plasticizing effect of the API is to use a torque
the drug and the polymer and the physical rheometer for preliminary experiments. It allows
properties of the polymer. For example, the running temperature and screw speed ramps in
throughput can be very limited if fine and volu- order to investigate the thermic and mechanical
minous polymers need to be fed into the extruder. impact on the given system. This allows the for-
The key considerations to establish the most mulator to identify process temperatures and the
appropriate process parameters for HME are required mechanical energy input (screw design).
shown in Table 8.10. Current research in the field focuses on the
Tg analysis by DSC as a function of the poly- application of HME to incorporate APIs in a
mer concentration provides a baseline for setting mesoporous silica formulation (Genina et al.
the extrusion temperature and helps to assess the 2018; Williams III et al. 2020) or in
miscibility of the system. In addition, rheological co-amorphous systems (see Chap. 13).
properties (zero rate viscosity and activation In 2004 the Food and Drug Administration
energy) of the material as a function of shear (FDA) encouraged the pharmaceutical industry
rate and temperature are key considerations in to introduce a quality by design (QbD) approach
establishing the HME process. In case of a new and usage of process analytical tools (PAT) in
product (API-polymer combination) with order to improve process understanding and
unknown properties, one approach to identify ensure the quality of final drug product.
310 S. Page et al.
Furthermore, a recent review by Butreddy and the drug and the polymer(s) simultaneously to
colleagues provided guidance and highlighted form the MBP. Under appropriate processing
the necessity of QbD-based formulation and pro- conditions, the drug is uniformly embedded in
cess design for the development of ASDs an amorphous form in the polymer. The MBP
(Butreddy et al. 2021). HME as a continuous process is particularly suitable for compounds
process is an obvious choice to introduce PAT that may have low solubility in commonly used
tools (Patil et al. 2016; Martin 2016). Over the solvents, such as acetone or ethanol, but have
years, in-line NIR and in-line Raman spectros- sufficient solubility in solvents, such as
copy were implemented in order to understand dimethylacetamide (DMA), dimethylsulfoxide
drug-polymer interactions, to continuously quan- (DMSO), dimethylformamide, or N-methylpyr-
tify the API amount and to monitor the solid-state rolidone (NMP). Based on the fact that usually
property of the API in solid dispersions (Saerens ionic polymers are used in this process, the
et al. 2011; Repka and Langley 2013). Due to the solvent-controlled precipitation is carried out
noninvasive nature of PAT tools, they can be used under either acidic or basic conditions. The resid-
to investigate the impact of potential critical pro- ual solvent is removed by a series of washing
cess parameters on critical quality attributes of the steps followed by filtration and drying. Due to
final product. In addition the real-time monitoring the nature of the process, it is used only when
of an HME process allows to immediately adjust other means of making amorphous dispersion are
the process parameters if process deviations are not feasible. The key considerations for the MBP
observed (Islam et al. 2015). process are listed in Table 8.11. A more detailed
discussion of MBP technology and its application
Coprecipitation to the formulation of poorly water-soluble drugs
The term coprecipitation has been used in the is provided in Chap. 5.
literature to describe amorphous solid dispersions More recently, Mann et al. (2018) used a
produced when the drug and the polymer are coprecipitation process in order to manufacture
precipitated together by changing the solubility an amorphous solid dispersion of compound A, a
conditions, either by addition of an antisolvent or low-solubility weak acid, in combination with
by evaporating the solvent (Simonelli et al. 1969). copovidone. Precipitation was induced using
Further attempts have been made to coprecipitate methyl-t-butyl-ether as nonaqueous antisolvent.
a solution of drug and polymer in ethanol by The obtained coprecipitated material was
addition into aqueous solution; however, the characterized, downstream processed to tablets,
characteristic crystalline and thermal peaks were and compared to a spray-dried ASD of the same
still present, indicating incomplete conversion to compound. The authors finally concluded that
the amorphous form (Kislalioglu et al. 1991). coprecipitation is a viable alternative to spray
Based on the principle of solvent-controlled drying for some APIs. Additional research done
precipitation, a technology referred to as by the same group (Schenck et al. 2019) clearly
microprecipitated bulk powder (MBP) has been showed that the material properties of the
developed that provides complete conversion of a coprecipitated material such as wettability and
drug to the amorphous form, for a large variety of dissolution behavior can be improved by adding
compounds (Albano et al. 2002; Shah et al. water-soluble excipients prior to evaporative iso-
2008). The MBP technology is particularly suit- lation of the coprecipitated material. In addition to
able for highly insoluble compounds for which these publications, Hou et al. (2019)
the utilization of spray-drying or hot-melt extru- demonstrated in their research work that acoustic
sion technologies is not feasible. With this tech- mixing is a scalable new method to make ASDs
nology, a solution of drug and stabilizing polymer through coprecipitation.
is introduced into an antisolvent that precipitates
Table 8.11 Key considerations for MBP process

Key considerations Description
Selection of a suitable solvent In addition to the requirements for amorphous form stability, the MBP process
and antisolvent requires careful evaluation of API and polymer solubility in solvents and
antisolvents. Since ionic polymers are primarily used for this process, the pH of the
precipitating medium is also critical. Furthermore, the residual solvent in the
product can also be critical
Solvent-antisolvent ratio To enable rapid quenching of the amorphous form, the solvent-antisolvent ratio
needs to be optimized. Generally, a ratio of 1:5-1:10 is required to achieve rapid
precipitation rate
Processing conditions The mode of solution addition to antisolvent, feed rate, hydrodynamic conditions,
and precipitation temperature merit careful evaluation
Isolation of the precipitated The material can be converted to powder form by several means ranging from spray
material drying and lyophilization to filtration and conventional drying (fluid bed/forced air
oven). The drying of aqueous material needs careful evaluation of the temperature/
time profile as the wet amorphous material in the aqueous phase could be more
susceptible to reversion
copovidone, was published by Iyer et al. (2013).

8.7 Downstream Processing
Their results clearly showed that the ASD
and Final Product Properties
manufacturing technologies altered the mechani-
cal properties of the polymers in a way that it
Once the solid dispersion is manufactured, it
could impact the downstream processing of the
needs to be further downstream processed to
material by reducing the deformation and in case
derive the final drug product. The selection of
of HME by increasing the density of the material.
the downstream manufacturing process is based
Similar observations were done by Davis et al.
on the properties of the solid dispersion (e.g.,
(2018) investigating a ternary ASD of
particle size, bulk density, mechanical properties)
itraconazole manufactured by spray drying and
as well as the properties of the finished drug
hot-melt extrusion, respectively. While the
product derived from the quality target product
spray-dried powder exhibited poor flow
profile (QTPP).
properties, due to the small particle size, it could
While some early research articles describe
be compressed into stronger compacts which
formulation compositions and manufacturing
released the drug faster. A number of additional
aspects for ASDs of particular compounds,
studies investigated the impact of formulation and
recently more studies are performed in order to
process parameters on the mechanical properties
investigate the downstream processing of new
and/or tabletting behavior of milled extrudates.
type of ASD systems, such as mesoporous silica
Figure 8.3 summarized the main findings of
(Hanada et al. 2020), for ASDs derived by new
these investigations, with parameters marked in
technologies, such as electrospinning (Démuth
bold having a clear impact and those marked in
et al. 2017), as well as to systematically investi-
italic showing no or only minor impact on the
gate the downstream processing of ASDs derived
mechanical properties/tabletting behavior.
by the established technologies mentioned in
This clearly shows the importance of
Sect. 8.6. The latter focused in particular on hav-
investigating the bulk powder properties of
ing an improved understanding on factors
ASDs prior to downstream processing, despite
impacting the mechanical properties as well as
the fact that certain properties such as the brittle-
the disintegration/dissolution behavior.
ness of copovidone-based ASDs are purely
One of the first publications, which
related to the polymer itself (Flügel et al. 2021).
investigated the mechanical properties of spray-
While ASDs manufactured by HME can be
dried and hot-melt extruded HPMCAS and
312 S. Page et al.
Fig. 8.3 Impact of hot-melt extrusion formulation and process parameter on mechanical properties and/or tabletting
behavior (Grymonpré et al. 2016, 2017a, b; Patel et al. 2017; Davis et al. 2018; Flügel et al. 2020, 2021)
directly downstream processed to tablets or which have an impact on the critical quality
capsules, an additional densification step is attributes (CQAs) of the final drug product. Typi-
required for drug product intermediates with cal CQAs of solid dispersion-based drug products
small particles and/or low density such as spray- are the amount of residual solvents, impurities,
dried powders or coprecipitates. This is normally crystallinity (initial and over-storage time; physi-
done using dry granulation processes such as cal stability), and the release profile of the drug.
roller compaction. Another alternative of One important aspect, which was investigated in
generating particles with high drug load after this context in recent years, is impact of compres-
coprecipitation and improved properties for sion on the physical stability of the final drug
downstream processing was recently described product (Singh et al. 2016; Berziņš et al. 2021).
by Schenck et al. (2019). For the characterization of the solid dispersion
The high amount of polymer present in the and the finished product, several analytical
ASD is not only challenging in terms of technologies are available, as described in
manufacturing tablets with reasonable mechani- Sect. 8.8.
cal strengths but also in terms of disintegration.
Water ingress into the tablet during the dissolu-
tion process often leads to polymer gelling and 8.8 Characterization
slow disintegration. This effect can be overcome of the Amorphous Systems
by addition of small amounts of inorganic salts
into the formulation as shown by Takano et al. An amorphous material is generally defined as a
(2020) and Xi et al. (2020). solid material that lacks long-range symmetry
In addition to the formulation composition, operators (translational, orientational, and confor-
which should be optimized with respect to bio- mational operators), which are characteristics of
availability and stability, it is of utmost impor- crystalline material. An amorphous system is con-
tance that the formulator investigates the impact sidered to be a disordered system with random
of potential critical process parameters on the molecular configuration. Furthermore, amor-
performance of the final drug product and finally phous formulations may possess local and short-
defines the critical process parameters (CPPs) range order crystallites, residual crystallinity, and
different molecular density regions. Conse- 8.8.1 Detection of Crystallinity

quently, gaining insight into amorphous in Amorphous System
systems to address the formulation challenge
is also an analytical challenge and frequently Many attempts have been made to determine the
limits development of a stable amorphous degree of crystallinity or amorphicity of amor-
formulation. phous formulations. Without doubt, this is the
For successful amorphous formulation devel- most critical testing in the evaluation of amor-
opment, it is imperative to have suitable analyti- phous formulations. Inappropriate selection of
cal techniques to characterize the materials polymer, inadequate drug loading (beyond
produced and to assess success at each step. The miscibility), and poor process control during
complexity of analytical testing may vary pharmaceutical manufacturing will result in
depending on the developmental stage, from sim- incomplete conversion of a crystalline drug to
ple qualitative testing at an initial stage to com- an amorphous state. Any trace level of crystalline
prehensive and validated quantitative testing at drug in amorphous formulations may serve as
the clinical manufacturing stage. Nevertheless, seeds for recrystallization during in vitro and
for successful amorphous formulation develop- in vivo dissolution or during storage, which may
ment, it is critical to understand how the amor- jeopardize the entire development program. The
phous system is being formed, to determine following techniques are frequently employed to
whether or not the crystalline drug is completely detect the degree of crystallinity.
converted to an amorphous state, to understand
how the drug and the polymer molecules are X-Ray Powder Diffraction
arranged in an amorphous solid dispersion, and X-ray powder diffraction (XRPD) is the most
to determine the uniformity of the formulation widely used and perhaps the most definitive tech-
produced. nique used in detection and quantification of crys-
In this section, commonly used analytical tallinity. Absence of sharp Bragg’s peaks
techniques will be briefly discussed for their corresponding to the crystalline drug suggests
merits and limitations in two groups distinguished a formulation is in an amorphous state. It is
by the information acquired. First the set of important, however, to note that XRPD detects
techniques used to exclude crystallinity or deter- the presence of molecular order; therefore, the
mine the level of residual crystallinity in an amor- disorder (amorphous state) is only implied by
phous system is described. The second set of the absence of the order (long-range symmetry
techniques includes the tests that are used to order). The limit of detection of the method
study the properties of amorphous formulations, ranges from 0.2 to 5% in an amorphous or
i.e., their molecular arrangement and their excipient mixture and dispersion and is depen-
behaviors. Understanding the molecular arrange- dent on the instrument used, the sampling pro-
ment of drug and polymer molecules in amor- tocol, and its characteristic (Palermo et al.
phous formulations is critical to minimize risk of 2012).
recrystallization upon storage and to understand
the in vitro/in vivo performance. Lastly, trends in IR and Raman Spectroscopy
amorphous formulation stability predictions are Vibrational spectroscopy, such as IR and Raman
briefly discussed, together with experimental spectroscopy, can be used to detect the variations
techniques that can be employed. Table 8.12 in vibrational energy between amorphous and
highlights the differences in physiochemical crystalline states. In general, sharp vibrational
properties between crystalline and amorphous peaks indicate crystallinity, whereas a broad
materials that can be further explored. hump suggests amorphicity, as a result of disorder
314 S. Page et al.
Table 8.12 Comparison of physicochemical properties between crystalline and amorphous drug
Attributes Crystalline drug Amorphous drug
Thermal Exhibits well-defined thermal behavior such as Exhibits no clear melting phenomenon, but
behaviors melting point and heat of fusion usually exhibits glass transition temperature
Birefringence Except cubic, crystalline material is anisotropic Amorphous material is isotropic and exhibits no
and exhibits birefringence birefringence under cross polarization
X-ray Crystalline material including liquid crystal and Lacking periodicity, does not reflect X-ray beam
diffraction plastic crystal reflects X-ray radiation exhibiting and exhibits characteristic amorphous diffused
characteristic diffraction pattern halo
Energy level Comparatively low in energy state, exhibits lower Comparatively higher in energy state, exhibits
solubility, slower dissolution, and more stable higher solubility, faster dissolution, and less
stable
Mechanical Lower specific molecular volume, leading to Randomness and irregular packing causes higher
properties denser and harder material molecular volume and less dense material
Spectroscopy Intermolecular interaction to adjacent molecules Varying configurational states and intermolecular
is well defined, resulting in well characteristic interaction to adjacent molecules results in broad
spectrum and diffused spectrum
in molecular arrangements. IR and Raman spec- characterization, but exposes limitations presum-
troscopy can be used to determine the amount of ably due to the requirement of experience in the
crystallinity of an amorphous formulation when a interpretation and quantification of data. Never-
calibration model is used. Advanced Raman spec- theless, many researchers have used the cross-
troscopy with improved signal intensity and ana- polarized microscope to study the kinetics of
lytical sensitivity became a versatile tool to detect crystallization and to predict the stability of amor-
and quantify amorphous drug content (Wabuyele phous systems (Taylor and Zografi 1997; Yu
et al. 2017). 2001).
Differential Scanning Calorimetry Other Techniques

DSC is probably the most versatile and widely Water vapor sorption can be used to discriminate
used technique in the characterization of amor- between amorphous and crystalline materials if
phous formulations including quantification of hygroscopicity is different and in the absence of
crystallinity. From thermal events, melting and interferences. Isothermal microcalorimetry was
(re)crystallization energy and changes in heat one of the earliest techniques used to study
capacity at glass transition can be measured, amorphous systems with remarkable sensitivity.
which can be used to quantify the crystallinity This is based on the principle that enthalpy
or the amorphicity. Nevertheless, with DSC changes at constant temperature and relative
small domains of crystalline material (< 30 nm) humidity are associated with (re)crystallization
cannot be detected (Meng et al. 2015), and small of amorphous material. However, this technique
portions of crystal traces may be dissolved in the is extremely sensitive to many operational
liquified polymer matrix during the first heating conditions, which makes it difficult for routine
cycle. DSC can also be used to assess the molec- use. Besides NIR and Raman, other spectro-
ular arrangement of the ASD (see also Sect. scopic techniques like terahertz-pulsed spectros-
8.8.2.1.). copy (Sibik et al. 2015) or solid-state nuclear
magnetic resonance (Paudel et al. 2014) are
Microscopic Technique used in order to investigate amorphous systems
Cross-polarized microscopy is one of the most or to quantify crystallinity.
powerful techniques in amorphous formulation
8.8.2 Determination of Molecular studied the differences between intermolecular

Arrangement in Amorphous interactions in amorphous and crystalline phases
Systems of celecoxib, valdecoxib, rofecoxib, and
etoricoxib using FTIR (Kaushal et al. 2008).
As discussed earlier in this chapter, an ideal Konno et al. studied interactions between amor-
formulation is one that provides the maximum phous felodipine and PVP, HPMCAS, and
physical stability over the intended period and HPMC using FTIR (Konno and Taylor 2006). In
maintains supersaturation while the drug is another case study, it was shown that the strength
being absorbed. Physically, the maximum sta- of the drug-polymer H-bonding varied with the
bility and optimal performance are achieved composition and process employed (Paudel et al.
when drug molecules are molecularly dispersed 2012).
in a polymer matrix with appropriate intermo- NIR chemical imaging combines NIR spec-
lecular interactions between the drug and the troscopy with digital imaging. Due to the syn-
polymer(s). In recent years, various techniques ergistic capabilities, this method can be used to
have been used in order to get more insight into assess the physical distribution of components
the phase behavior of amorphous solid in solids. Ma et al. (2013) used the method in
dispersions. order to determine the drug distribution in an
ASD by comparing it to that in a homogeneous
physical mixture with the same composition
and to facilitate the selection of the formulation
A first insight into the molecular arrangement of
composition and manufacturing technology for
the ASD can be obtained by determining the glass
an ASD.
transition temperature of the system using DSC.
Binary drug-polymer systems showing a single Tg
are generally expected to be miscible, whereas Atomic Force Microscopy
systems with two Tgs or the appearance of a Atomic force microscopy (AFM) allows to get
melting endotherm of the drug would generally a more in-depth view by assessing the phase
indicate immiscibility. behavior on a nanometer scale. It is a scanning
probe technique which has a high spatial reso-
lution. Phase separation of the ASD leads to
(N)IR and Raman Spectroscopy
differences in the local material properties that
Vibrational spectroscopy methods such as IR or
can be assessed by using AFM in tapping
Raman spectroscopy are utilized to investigate
mode. Lauer et al. (2013) used AFM in order
the molecular arrangement in the ASD. In com-
to compare the homogeneity of hot-melt
parison to the conventional Raman techniques,
extrudates obtained from a small-scale extruder
micro-Raman has a much higher detection limit
with data obtained from screening experiments.
and provides localized compositional information
Lamm et al. (2016) investigated the phase
that are distinct from the bulk of the material;
behavior of ASD containing copovidone and
however the method has only low spatial resolu-
vitamin TPGS in dependence of various process
tion (Meng et al. 2015). More recently,
parameters used to prepare the extrudates. Fur-
low-frequency Raman spectroscopy was
thermore, they analyzed the nature of the
described in order to monitor the crystallization
domains present in some of the samples by
kinetics of amorphous systems (Bērziņš et al.
generating force-displacement curves. This
2019).
technology is also known as force spectroscopy
IR spectroscopy has been extensively used to
(Lamm et al. 2016). Other researchers coupled
study the molecular arrangement of drug and
AFM with local or nanothermal analysis
polymer and their interaction. Kaushal et al.
316 S. Page et al.
(Qi et al. 2011; Purohit and Taylor 2015), 8.8.3 Dissolution Method
photothermal Fourier transform infrared for Amorphous Formulations
microspectroscopy (PT-FTIR: Qi et al. 2011),
or nanoscale infrared spectroscopy (Purohit and The dissolution profile is considered to be a good
Taylor 2015) to get further insight into the predictor for the in vivo performance; however it
different phases formed. is difficult to establish an in vitro-in vivo correla-
tion due to the complex nature of supersaturation
generation and maintenance in combination with
Solid-State NMR
the driving force triggering absorption of ASDs
Solid-state NMR (ssNMR) has been increasingly
(Baghel et al. 2016). In general it is assumed that
used to study miscibility and molecular interac-
the drug release mechanism of an ASD depends
tion between amorphous drug and polymer and to
on the type of polymer used. The dissolution
investigate the dynamics and phase compositions
profile of an ASD containing hydrophilic,
of amorphous dispersions (Paudel et al. 2014). A
medium-soluble polymers, which dissolve in
single relaxation time for the drug and polymer
the medium, is characterized by an initial surge
indicates complete miscibility (Meng et al. 2015).
of supersaturation followed by a decrease in
Mistry et al. (2015) used FTIR in combination
drug concentration due to precipitation (Park
with ssNMR to probe the interactions between
2015). The degree of precipitation depends on
ketoconazole and three different polymers. In
the type of polymer used. ASDs containing
another study by Song et al. (2016), ssNMR was
medium-insoluble polymers lack the initial fast
used to confirm the strong intermolecular acid-
increase in concentration and supersaturation
base interactions present between drug and
generation, and their overall release profile is
polymer.
sustained for an extended period of time (Park
2015). For these types of ASDs, the rate of
Other Techniques supersaturation generation has a huge impact
In order to assess the molecular arrangement/ on the kinetic solubility profile as discovered
miscibility of the ASD, XRPD is often com- by Sun and Lee (2013).
bined with computational methods such as pair Based on the abovementioned considerations,
distribution functions (PDFs), pure curve reso- the development of meaningful dissolution
lution method (PCRM), and alternate least methods for kinetically unstable systems of
square (ALS) (Meng et al. 2015). Bates et al. poorly soluble drugs is challenging. Neverthe-
(2006) have used XRPD in combination with less, the requirements in terms of dissolution
the pairwise distribution function to investigate might change during the whole development
the local structure of amorphous formulations. ranging from predicting the in vivo performance
Mistry et al. (2015) used variable temperature to control batch to batch consistency. In the first
XRPD in order to determine the crystallization phase, the physiologically relevant conditions,
onset temperature of various ketoconazole such as composition of the medium, pH of the
solid dispersions. Furthermore, Song et al. medium, dose to GI fluid volume ratio, and
(2016) used X-ray photoelectron spectroscopy exposure time, should be considered (He and
(XPS) to determine the binding energy of Ho 2015). For these studies the use of non-sink
an API and polymers. Besides NIR, solid- conditions, the measurement of the pH at the end
state NMR, and Raman, other spectroscopic of the dissolution test, and the characterization of
techniques like terahertz-pulsed spectroscopy the precipitate should be considered (He and Ho
(Sibik et al. 2015) are described in order to 2015). The use of a diffusion study might also
investigate the dynamics of amorphous be useful, in order to relate the drug release with
systems. the absorption process. Later in the
development, the focus of the method has maximum amount of drug dissolved (Cmax) at the
changed toward quality control (QC), and it is time to achieve the maximum concentration
vital to demonstrate adequate discriminating (Tmax) could be used for kinetic solubility deter-
power for the QC method. mination. Furthermore, the dissolution profile is
A perfect sink condition will not help to ascer- characterized by the duration in which supersatu-
tain the ability of a system to maintain supersaturation is maintained and by the area under the
ration. However, an overly non-sink condition curve (AUC).
may result in overdiscrimination and potential Based on the kinetic solubility, dissolution
elimination of a viable formulation. There are conditions can be varied with regard to dose,
several publications related to dissolution of volume of dissolution medium, and surfactant
amorphous formulations (Law et al. 2004; Chan concentration of dissolution medium for screen-
and Kazarian 2004; Doherty and York 1987; ing amorphous formulations. A dissolution
Janssens et al. 2010). The effect of polymer type medium representing 100% saturation is the
(Konno et al. 2008; Saboo and Moseson 2020) most efficient way of screening formulations as
and drug load (Saboo et al. 2019) on the dissolu- it avoids unnecessary stress during dissolution
tion rate of amorphous systems has been testing. An example for such a dissolution test is
described as well in current literature. shown in Fig. 8.5 for a research compound. Addi-
Unlike for crystalline APIs, equilibrium solu- tionally, a formulation maintaining supersatura-
bility cannot be determined for amorphous tion for at least 2–4 h (physiologically relevant)
materials because, at equilibrium, drug solubility would represent a viable formulation. On the
using the crystalline or the amorphous form is contrary, a formulation maintaining supersatura-
similar, by definition. Therefore, the method has tion for less than 60 min would need careful
to determine the kinetic solubility in physiologi- evaluation.
cally relevant conditions. Additional research was done by Lynne
The kinetic solubility of amorphous material S. Taylor's group in order to better understand
can be determined in bio-relevant or aqueous the mechanism of precipitation by investigating
medium with or without surfactant. If we accept the solution phase behavior during dissolution.
the fact that precipitation follows supersaturation, They investigated, for instance, the dissolution
then solubility of amorphous material at a 30–60- behavior of danazol from ASDs containing dif-
min time window or at the peak of the upward ferent polymers and found out that upon dissolu-
slope could be used for determination of kinetic tion prior to crystallization of the drug, liquid-
solubility. This is illustrated in Fig. 8.4, where the liquid phase separation occurred. The phase
Fig. 8.4 Schematic

representation for kinetic
solubility determination for
amorphous materials
318 S. Page et al.
Fig. 8.5 Effect of

surfactant concentration on 100
the dissolution behavior of
amorphous solid dispersion
and supersaturation 80
% Dissolved
60
40
100 % Saturation
20 200 % Saturation
300 % Saturation
400 % Saturation
0
0 40 80 120 160 200 240 280 320 360
Time [min]
behavior of the solutions present during dissolu- addition, the effect of water on molecular mobil-
tion was monitored using UV and fluorescence ity and stability must be accounted for. Water can
spectroscopy (Jackson et al. 2016). In another exist in an amorphous state with a Tg of about
research article, the release profile of celecoxib 138 C (135 K). As such, water can reduce
ASDs with high drug loading was optimized in glass transition temperatures of amorphous
terms of rapid release and crystallization inhibi- systems substantially. The effect on Tg can be
tion by using a combination of two different more pronounced for partially amorphous
polymers. Besides the drug release, which was systems. In such cases, most of the water will be
determined “at sink” conditions, the authors also located in the disordered region, and the Tg in that
determined the nucleation induction time in a region will be considerably lower (Marsac et al.
precipitation assay using light scattering in order 2010; Rumondor et al. 2009a, b). Temperature
to optimize the crystallization inhibition (Xie and naturally has a direct effect on molecular mobil-
Taylor 2016). ity. Molecular mobility is generally considered
one of the key factors in determining amorphous
system stability. The molecular mobility of a sys-
tem can be quantified using dielectric spectros-
8.8.4 Stability Prediction
copy. This technology was employed by Mistry
et al. in order to compare ASD of ketoconazole
The goal for amorphous formulation stability is to
with three different polymers. They could show
maintain long-term physical stability with respect
that the molecular mobility was only moderately
to solid-state properties and maintain supersatura-
reduced in case of PVP and poly(2-hydroxyethyl
tion during the time course of the dissolution
methacrylate), whereas a strong reduction was
process mimicking the in vivo dissolution win-
observed for poly(acrylic acid) (Mistry et al.
dow, which is generally about 2–4 h. The key to
2015).
understanding stability is to understand that
In spite of intensive research to predict the
molecules in an amorphous system can have sig-
stability of amorphous systems through molecular
nificant molecular motion both above and below
mobility concepts, the prediction of physical sta-
the glass transition temperature. Molecular
bility is rather unpredictable. Often, physical
motion in the form of translational and rotational
instability occurs in a nonlinear fashion after
diffusion can generally be described in terms of
varying induction periods. Although the stability
molecular relaxation time or annealing time. In
of amorphous systems is not predictable, one can and mechanical properties of polymers will help
use accelerated stability conditions of severe in the determination of the suitability of
humidity and temperature to rank order the polymer(s) for a given drug for an amorphous
formulations. A commonly used accelerated sta- solid dispersion. Besides theoretical
bility condition for formulation screening is considerations in terms of solubility and
40 C/100% RH or 40 C/75%RH open. In this miscibility, miniaturization techniques, such as
case, amorphous solid dispersions are subjected the SPADS approach, help in the selection of
to stress conditions and are evaluated by PXRD, specific polymers and drug loads for maximum
FTIR, and DSC analysis. In addition, the sus- solid-state stability and supersaturation.
pension stability that is performed to enable the In a second phase, the thorough evaluation of
use of amorphous dispersion for toxicology stud- physicochemical properties of drugs and
ies is also helpful to assess physical stability. For polymers with respect to thermal behavior, sol-
toxicology purposes, the preferred vehicle is the vent, and aqueous solubility and stability guides
one having the lowest solubility for amorphous the selection of appropriate processing
solid dispersion and should maintain amorphous techniques for amorphous solid dispersions.
form stability for at least 4 h at room During the development of the amorphous for-
temperature. mulation, the impact of various process
parameters needs to be investigated in order to
ensure consistent quality in terms of dissolution
performance and stability. Different instrumental
8.9 Overall Summary
techniques are presented to characterize and
quantify the crystallinity and for understanding
In this chapter, we have proposed a structured
the phase behavior of amorphous solid
development approach for amorphous
dispersions. The use of more than one technique
formulations based on sound physicochemical
is preferred to establish confidence. Amorphous
principles of the drug and the polymer with the
formulation stability is difficult to predict; how-
goal of maximizing success rates and reducing
ever, accelerated storage conditions such as
risks. The proposed approach consists of a com-
40 C/100% RH and suspension stability test
prehensive evaluation of the drug substance and
provide a useful insight into the stability of
polymer properties and understanding the basic
these amorphous systems. Similarly, approaches
principles to help design amorphous dispersions
are presented to help develop appropriate disso-
that provide consistent in vitro and in vivo per-
lution methods based on kinetic solubility and
formance. The assessment of basic physico-
consideration of the supersaturation level to
chemical properties of drug substance including
avoid false negatives. Finally, a flowchart
its thermal behavior (Tg, Tm, stability, and
shown in Fig. 8.6 is proposed as a blueprint for
recrystallization), hydrogen-bond acceptor and
structured development of amorphous
donor groups, molecular weight, hydrophobicity,
formulations.
solubility parameters, solubility (aqueous and
organic solvents), and hygroscopicity forms the
Acknowledgments The authors of the third edition of the
basis of determining the degree of difficulty a book chapter would like to thank their former co-workers
molecule may present in converting to and Navnit Shah, Harpreet Sandhu, Duk Soon Choi, and Oskar
maintaining an amorphous form. Likewise, a Kalb for the great work done together for the first edition.
thorough understanding of the physicochemical
320 S. Page et al.
API Polymer (s)
Physicochemical properties:
Solubility (solvents)
Solubility (aq. and solvents) Tg, MW, Viscosity
Tm, Tg, MW, Log P etc. Thermal stability
Chemical stability Availability of pharm. Grade
Thermal stability Safety
Suitability of API for ASD
Prediction of the solubility advantage of

the amorphous form
Glass forming ability
Glass stability
Screening
API – polymer miscibility

Specific interactions
Hygroscopy and water activity
Theoretical considerations
Screening - Manufacturability
Pre-selection of manufacturing technologies based on physicochemical properties
Miniaturized screening (experimental)
Supersaturation screening
Solid state screening
Interaction screening
Prototype development
Small scale/lab scale development - Selection of the manufacturing technology

In vitro dissolution and in vivo PK studies Analytical characterization of ASD
Selection of process parameters
Solid state stability (stress) testing Detection of crystallinity (XRPD, Raman, IR, DSC,
cross-polarized microscopy)
Determination of molecular arrangement (IR,
Raman, AFM, ssNMR)
Dissolution kinetics
Drug product development
Selection of excipients
Investigate release and supersaturation of drug product
Selection of the downstream process
QbD (Identification of cQA’s and cPP’s, PAT)
Investigate physical and chemical stability of the drug product
Packaging of Drug product
Stability assessment
Fig. 8.6 Structured development approach for amorphous formulation development

References Berziņš K, Fraser-Miller SJ, et al. Low-frequency Raman

spectroscopic study on compression-induced destabili-
zation in melt-quenched amorphous celecoxib. Mol
Abela AR, Alcacio T et al. Modulator of cystic fibrosis
Pharm. 2019;16:3678–86.
transmembrane conductance regulator, pharmaceutical
Berziņš K, Fraser-Miller SJ, et al. Combined effect of the
compositions, methods of treatment, and process for
preparation method and compression on the physical
making the modulator. International Patent Publication
stability and dissolution behavior of melt-quenched
2018/107100Al. 2018.
amorphous celecoxib. Mol Pharm. 2021;18(3):1408–18.
Acartürk F, Kislal o, et al. The effect of some natural
Breitenbach J. Melt extrusion: from process to drug deliv-
polymers on the solubility and dissolution
ery technology. Eur J Pharm Biopharm. 2002;54(2):
characteristics of nifedipine. Int J Pharm. 1992;85
107–17.
(1–3):1–6.
Butreddy A, Bandari S, et al. Quality-by-design in hot melt
Adrjanowicz K, Kaminski K, et al. Dielectric relaxation
extrusion based amorphous solid dispersions: An
studies and dissolution behavior of amorphous verapa-
industrial perspective on product development. Eur J
mil hydrochloride. J Pharm Sci. 2010;99(2):828–39.
Pharm Sci. 2021;158:105655.
Agrawal A, Dudhedia M, et al. Development of tablet
Califano JC, Fickes MG et al. (2016) Solid forms of anti-
formulation of amorphous solid dispersions prepared
viral compounds US Patent Publication 2016/
by hot melt extrusion using quality by design
0090373A1.
approach. AAPS Pharmscitech. 2015;17:214–32.
Chan KLA, Kazarian SG. FTIR spectroscopic imaging of
Agrawal AM, Dudhedia MS, et al. Hot melt extrusion:
dissolution of a solid dispersion of nifedipine in poly
development of an amorphous solid dispersion for an
(ethylene glycol). Mol Pharm. 2004;1(4):331–5.
insoluble drug from mini-scale to clinical scale. AAPS
Chen S, Gates BD, et al. Crystal forms. International
PharmSciTech. 2016;17(1):133–47.
Patent Publication 2015/188045Al. 2015.
Albano AA, Phuapradit W, et al. Stable complexes of
Chen Y, Chen H, et al. A single hydrogen to fluorine
poorly soluble compounds in ionic polymers. US Pat-
substitution reverses the trend of surface composition
ent Office, United States of America, F. Hoffmann-La
enrichment of Sorafenib amorphous solid dispersion
Roche Ltd, 7. 2002.
upon moisture exposure. Pharm Res. 2019;36(7):105.
Alhalaweh A, Alzghoul A, et al. Physical stability of drugs
Chiang P-C, Ran Y, et al. Evaluation of drug load and
after storage above and below the glass transition tem-
polymer by using a 96-well plate vacuum dry system
perature: relationship to glass-forming ability. I J
for amorphous solid dispersion drug delivery. AAPS
Pharm. 2015;495(1):312–7.
PharmSciTech. 2012;13(2):713–22.
Al-Obaidi H, Brocchini S, et al. Anomalous properties of
Chiou WL, Riegelman S. Pharmaceutical applications of
spray dried solid dispersions. J Pharm Sci. 2009;98
solid dispersion systems. J Pharm Sci. 1971;60(9):
(12):4724–37.
1281–302.
Aso Y, Yoshioka S. Molecular mobility of nifedipine–
Chiou WL, Riegelmann S. Oral absorption of griseofulvin
PVP and phenobarbital–PVP solid dispersions as
in dogs: increased absorption via solid dispersion in
measured by 13 C-NMR spin-lattice relaxation time.
polyethylene glycol 6000. J Pharm Sci. 1970;59:937–42.
J Pharm Sci. 2006;95(2):318–25.
Chokshi RJ, Sandhu HK, et al. Characterization of
Aso Y, Yoshioka S, et al. Effect of water on the molecular
physico-mechanical properties of indomethacin and
mobility of sucrose and poly(vinylpyrrolidone) in a
polymers to assess their suitability for hot-melt extru-
colyophilized formulation as measured by 13 C-NMR
sion process as a means to manufacture solid disper-
relaxation time. Chem Pharm Bull. 2002;50(6):822–6.
sion/solution. J Pharm Sci. 2005;94(11):2463–74.
Baghel S, Cathcart H, et al. Polymeric amorphous solid
Chokshi RJ, Shah NH, et al. Stabilization of low glass
dispersions: a review of Amorphization, crystalliza-
transition temperature indomethacin formulations:
tion, stabilization, solid-state characterization, and
impact of polymer-type and its concentration. J
aqueous solubilization of biopharmaceutical classifica-
Pharm Sci. 2008;97(6):2286–98.
tion system class II drugs. J Pharm Sci. 2016; https://
Crowley MM, Zhang F, et al. Pharmaceutical applications
doi.org/10.1016/j.xphs.2015.10.008.
of hot-melt extrusion: part I. Drug Dev Ind Pharm.
Baird JA, Van Eerdenburgh B, et al. A classification
2007;33(9):909–26.
system to assess the crystallization tendency of organic
Curatolo W, Nightingale J, et al. Utility of hydroxypropyl-
molecules from undercooled melts. J Pharm Sci.
methylcellulose acetate succinate (HPMC-AS) for ini-
2010;99(9):3787–806.
tiation and maintenance of drug supersaturation in the
Barillaro VR, Pescarmona PP, et al. High-throughput
GI milieu. Pharm Res. 2009;26(6):1419–31.
study of phenytoin solid dispersions: formulation
Davis MT, Potter CB, et al. Downstream processing of a
using an automated solvent casting method, dissolution
ternary amorphous solid dispersion: the impacts of
testing, and scaling-up. J Comb Chem. 2008;10(5):
spray drying and hot melt extrusion on powder flow,
637–43.
compression and dissolution. Int J Pharmaceut.
Bates S, Zografi G, et al. Analysis of amorphous and
2018;544:242–53.
nanocrystalline solids from their X-ray diffraction
De Maesschalk R, Stokbroekx S, et al. Development of a
patterns. Pharm Res. 2006;23(10):2333–49. Epub
96-well plate dissolution method for screening solid
22 Sept 2006
dispersions: COMPARISON to classical USP methods
322 S. Page et al.
and its use in predicting oral bioavailability in animals. Forster A, Hempenstall J, et al. Selection of excipients for
AAPS annual meeting and exposition, Ernest melt extrusion with two poorly water-soluble drugs by
N. Morial Convention Center, New Orleans. 2010. solubility parameter calculation and thermal analysis.
Defrese MK, Farmer MA, et al. Approaches to understand- Int J Pharm. 2001b;226(1–2):147–61.
ing the solution-state organization of spray-dried dis- Friesen DT, Shanker R, et al. Hydroxypropyl methylcellu-
persion feed solutions and its translation to the solid lose acetate succinate-based spray-dried dispersions:
state. Mol Pharm. 2020;17(12):4548–63. an overview. Mol Pharm. 2008;5(6):1003–19.
Démuth B, Farkas A, et al. Development and tableting of Genina N, Hadi B, et al. Hot melt extrusion as solvent-free
directly compressible powder from electrospun technique for a continuous manufacturing of drug-
nanofibrous amorphous solid dispersion. Adv Powder loaded mesoporous silica. J Pharm Sci. 2018;107:
Technol. 2017;28:1554–63. 149–55.
Deng W, Majumdar S, et al. Stabilization of fenofibrate in Gil M, Vicente J, Gaspar F. Scale-up methodology for
low molecular weight hydroxypropylcellulose matri- pharmaceutical spray drying. Chem Today. 2010;28
ces produced by hot-melt extrusion. Drug Dev Ind (4):18–22.
Pharm. 2013;39(2):290–8. Gordon M, Taylor JS. Ideal copolymers and the second-
Ditzinger F, Price DJ, et al. Lipophilicity and order transitions of synthetic rubbers. I. Noncrystalline
hydrophobicity considerations in bio-enabling oral copolymers. J Appl Chem. 1952;2:493–500.
formulations approaches – a PEARRL review. J Greenhalgh DJ, Williams AC, et al. Solubility parameters
Pharm Pharmacol. 2019a;71:464–82. as predictors of miscibility in solid dispersions. J
Ditzinger F, Scherer U, et al. Modified polymer matrix in Pharm Sci. 1999;88(11):1182–90.
pharmaceutical hot melt extrusion by Molecular Grymonpré W, Jaeghere WD, et al. The impact of hot-melt
interactions with a carboxylic Coformer. Mol Pharm. extrusion on the tableting behaviour of polyvinyl alco-
2019b;16(1):141–50. hol. Int J Pharm. 2016;498:254–62.
Dobry DE, Settell DM, et al. A model-based methodology Grymonpré W, Verstraete G, et al. In-line monitoring of
for spray-drying process development. J Pharm Innov. compaction properties on a rotary tablet press during
2009;4(3):133–42. tablet manufacturing of hot-melt extruded amorphous
Doherty C, York P. Mechanisms of dissolution of solid dispersions. Int J Pharm. 2017a;517:348–58.
frusemide PVP solid dispersions. Int J Pharm. Grymonpré W, Bostijn N, et al. Downstream processing
1987;34(3):197–205. from hot-melt extrusion towards tablets: a quality by
Dohrn S, Reimer P, et al. Thermodynamic modeling of design approach. Int J Pharm. 2017b;531:235–45.
solvent-impact on phase separation in amorphous solid Gupta P, Kakumanu VK, et al. Stability and solubility of
dispersions during drying. Mol Pharm. 2020a;17(7): celecoxib-pvp amorphous dispersions: a molecular
2721–33. perspective. Pharm Res. 2004;21:1762–9.
Dohrn S, Luebbert C, et al. Phase behavior of Gupta J, Nunes C, et al. Prediction of solubility parameters
pharmaceutically relevant polymer/solvent mixtures. and miscibility of pharmaceutical compounds by
Int J Pharm. 2020b;577:119065. molecular dynamics simulations. J Phys Chem
Dohrn S, Luebbert C, et al. Solvent influence on the phase B. 2011;115(9):2014–23.
behavior and glass transition of Amorphous solid Guzmán HR, Tawa M, et al. Combined use of crystalline
dispersions. Eur J Pharm Biopharm. 2021;158:132–42. salt forms and precipitation inhibitors to improve oral
Edueng K, Mahlin D, et al. The need for restructuring the absorption of celecoxib from solid oral formulations. J
disordered science of amorphous drug formulations. Pharm Sci. 2007;96(10):2686–702.
Pharm Res. 2017;34:1754–72. Hadida S, Van Goor F, et al. Case history: Kalydeco®
EMEA.. Assessment report Venclyxto, Procedure (VX-770, Ivacaftor), a CFTR potentiator for the treat-
No. EMEA/H/C/004106/0000. 2016. https://www. ment of patients with cystic fibrosis and the G551D-
ema.europa.eu/en/documents/assessment-report/ CFTR mutation. Annu Rep Med Chem. 2014;49:383–
venclyxto-epar-public-assessment-report_en.pdf. 98.
Accessed 15 Mar 2021. Hanada M, Jermain SV, et al. Ternary amorphous solid
Feng J, Xu L, et al. Evaluation of polymer carriers with dispersions containing a high-viscosity polymer and
regard to the bioavailability enhancement of bifendate mesoporous silica enhance dissolution performance.
solid dispersions prepared by hot-melt extrusion. Drug Mol Pharm. 2020;18:198–213.
Dev Ind Pharm. 2012;38(6):735–43. Hancock BC, Parks M. What is the true solubility advan-
Flügel K, Hennig R, et al. Impact of structural relaxation tage for amorphous pharmaceuticals? Pharm Res.
on mechanical properties of amorphous polymers. Eur 2000;17(4):397–404.
J Pharm Biopharm. 2020;154:214–21. Hancock BC, York P, et al. The use of solubility
Flügel K, Schmidt K, et al. Impact of incorporated drugs parameters in pharmaceutical dosage form design. Int
on material properties of amorphous solid dispersions. J Pharm. 1997;148:1–21.
Eur J Pharm Biopharm. 2021;159:88–98. He Y, Ho C. Amorphous solid dispersions: utilization and
Forster A, Hempenstall J, et al. The potential of small- challenges in drug discovery and development. J
scale fusion experiments and the Gordon-Taylor equa- Pharm Sci. 2015;104(10):3237–58.
tion to predict the suitability of drug/polymer blends Hou HH, Rajesh A, et al. Impact of method of preparation
for melt extrusion. Drug Dev Ind Pharm. 2001a;27(6): of amorphous solid dispersions on mechanical
549–60.
properties: comparison of coprecipitation and spray Kaushal AM, Chakraborti AK, et al. FTIR studies on
drying. J Pharm Sci. 2019;108 differential intermolecular association in crystalline
Hu Q, Choi DS, et al. Highly efficient miniaturized and amorphous states of structurally related
coprecipitation screening (MiCoS) for amorphous non-steroidal anti-inflammatory drugs. Mol Pharm.
solid dispersion formulation development. Int J 2008;5(6):937–45.
Pharm. 2013;450:53–62. Keshavarz-Shokri A, Zhang, et al. Solid forms of (r)-1
Huang Y, Dai W-G. Fundamental aspects of solid disper- (2,2-difluorobenzo[d][1,3]dioxol-5-yl) -n-
sion technology for poorly soluble drugs. Acta Pharm (1-(2,3-dihyderoxypropyl)-6-fluoro-2-(1-hydroxy-2-
Sinica B. 2014;4(1):18–25. methylpropan-2-yl)-1h-indol-5-yl)cyclopropane-
Huang J, Wigent RJ, et al. Drug-polymer interaction and carboxamide. International Patent Publication
its significance on the physical stability of nifedipine 2011119984A1. 2011.
amorphous dispersion in microparticles of an ammonio Kislalioglu MS, Khan MA, et al. Physical characterization
methacrylate copolymer and ethylcellulose binary and dissolution properties of ibuprofen: Eudragit
blend. J Pharm Sci. 2008;97(1):251–62. coprecipitates. J Pharm Sci. 1991;80(8):799–804.
Hugo M, Kunath K, et al. Selection of excipient, solvent Knopp MM, Tajber L, et al. Comparative study of differ-
and packaging to optimize the performance of spray- ent methods for the prediction of drug–polymer solu-
dried formulations: case example fenofibrate. Drug bility. Mol Pharm. 2015;12:3408–19.
Dev Ind Pharm. 2013;39(2):402–12. Knopp MM, Olesen NE, et al. Statistical analysis of a
Islam M, Scoutaris N, et al. Implementation of transmis- method to predict drug–polymer miscibility. J Pharm
sion NIR as a PAT tool for monitoring drug transfor- Sci. 2016;105:362–7.
mation during HME processing. Eur J Pharm Konno H, Taylor LS. Influence of different polymers on
Biopharm. 2015;96:106–16. the crystallization tendency of molecularly dispersed
Iyer R, Hedge S, et al. The impact of hot melt extrusion amorphous felodipine. J Pharm Sci. 2006;95(12):
and spray drying on mechanical properties and 2692–705.
tableting indices of materials used in pharmaceutical Konno H, Handa T, et al. Effect of polymer type on the
development. J Pharm Sci. 2013;102(10):3604–13. dissolution profile of amorphous solid dispersions
Jackson MJ, Kestur US, et al. Dissolution of Danazol containing felodipine. Eur J Pharm Biopharm.
amorphous solid dispersions: supersaturation and 2008;70(2):493–9.
phase behavior as a function of drug loading and poly- Kwong AD, Kauffman RS, et al. Discovery and develop-
mer type. Mol Pharm. 2016;13(1):223–31. ment of telaprevir: an NS3-4A protease inhibitor for
Janssens S, Nagels S, et al. Formulation and characterization treating genotype 1 chronic hepatitis C virus. Nat
of ternary solid dispersions made up of Itraconazole and Biotechnol. 2011;29(11):993–1003.
two excipients, TPGS 1000 and PVPVA 64, that were Lamm MS, DiNunzio J, et al. Assessing mixing quality of
selected based on supersaturation screening study. Eur J a copovidone-TPGS hot melt extrusion process with
Pharm Biopharm. 2008;69:158–66. atomic force microscopy and differential scanning cal-
Janssens S, De Zeure A, et al. Influence of preparation orimetry. AAPS PharmSciTech. 2016;17(1):89–98.
methods on solid state supersaturation of amorphous Lapina OV, Shi B, et al. Solid forms of an antiviral
solid dispersions: a case study with itraconazole and compound. International Patent Publication 2015/
Eudragit E100. Pharm Res. 2010;27(5):775–85. 191431Al. 2015.
Jensen KT, Blaabjerg LI, et al. Preparation and characteri- Lauer M, Grassmann O, et al. Atomic force microscopy-
zation of spray-dried co-amorphous drug–amino acid based screening of drug-excipient miscibility and stabil-
salts. J Pharm Pharmacol. 2015;68(5):615–24. ity of solid dispersions. Pharm Res. 2011;28(3):572–84.
Jermain SV, Brough C, et al. Amorphous solid dispersions Lauer M, Siam M, et al. Rapid assessment of homogeneity
and nanocrystal technologies for poorly water-soluble and stability of amorphous solid dispersions by atomic
drug delivery – an update. Int J Pharm. 2018;535(1–2): force microscopy – from bench to batch. Pharm Res.
379–92. 2013;30(8):2010–22.
Jijun F, Lili Z, et al. Stable nimodipine tablets with high Law D, Schmitt EA, et al. Ritonavir-PEG 8000 amorphous
bioavailability containing NM-SD prepared by hot-melt solid dispersions: in vitro and in vivo evaluations. J
extrusion. Powder Technol. 2010;204:214–21. Pharm Sci. 2004;93(3):563–70.
Just S, Sievert F, et al. Improved group contribution Lee T, Lee J. Drug-carrier screening on a chip. Pharm
parameter set for the application of solubility Tech N Am. 2003;27(1):40–8.
parameters to melt extrusion. Eur J Pharm Biopharm. Lee YC, Zhou G, et al. Application of online near infrared
2013;85(3):1191–9. for process understanding of spray-drying solution
Kalb O, Page S, et al. Scale-up of solid dispersions. In: preparation. J Pharm Sci. 2019;108(3):1203–10.
Encyclopedia of pharmaceutical science and technol- Lehmkemper K, Kyeremateng SO, et al. Impact of poly-
ogy. 4th ed; 2013. https://doi.org/10.1081/E- mer type and relative humidity on the long-term physi-
EPT4-120050349. cal stability of amorphous solid dispersions. Mol
Kasten G, Lobmann K, et al. Co-former selection for Pharm. 2017;14(12):4374–86.
co-amorphous drug-amino acid formulations. Int J Lemmer HJR, Liebenberg W. Preparation and evaluation
Pharm. 2019;557:366–73. of metastable solid-state forms of lopinavir. Pharmazie.
2013;68:327–32.
324 S. Page et al.
Lenz E, Jensen KT, et al. Solid-state properties and disso- crystallization inhibition in ketoconazole solid
lution behaviour of tablets containing co-amorphous dispersions. Mol Pharm. 2015;12(9):3339–50.
indomethacin–arginine. Eur J Pharm Biopharm. Miyazaki T, Yoshioka S, et al. Ability of polyvinylpyr-
2015;96:44–52. rolidone and polyacrylic acid to inhibit the crystalliza-
Leuner C, Dressman J. Improving drug solubility for oral tion of amorphous acetaminophen. J Pharm Sci.
delivery using solid dispersions. Eur J Pharm 2004;93:2710–7.
Biopharm. 2000;50(1):47–60. Miyazaki T, Yoshioka S, et al. Crystallization rate of amor-
Liu H, Taylor LS, et al. The role of polymers in oral phous nifedipine analogues unrelated to the glass transi-
bioavailability enhancement; a review. Polymer. tion temperature. Int J Pharm. 2007;336:191–5.
2015;77:399–415. Moes J, Koolen S, et al. Pharmaceutical development and
Lobmann K, Grohganz H. Amino acids as co-amorphous preliminary clinical testing of an oral solid dispersion
stabilizers for poorly water soluble drugs – Part 1: formulation of docetaxel (ModraDoc001). J Pharm Sci.
preparation, stability and dissolution enhancement. 2011;420:244–50.
Eur J Pharm Biopharm. 2013;85:873–81. Mudie DM, Stewart AM, et al. Novel high-drug-loaded
Loftsson T, Fririksdóttir H, et al. The effect of water- amorphous dispersion tablets of posaconazole; in vivo
soluble polymers on aqueous solubility of drugs. Int J and in vitro assessment. Mol Pharm. 2020;17(12):
Pharm. 1996;127(2):293–6. 4463–72.
Luebbert C, Real D, et al. Choosing appropriate solvents for Nair AR, Lakshman YD, et al. Overview of extensively
ASD preparation. Mol Pharm. 2018;15(11):5397–409. employed polymeric carriers in solid dispersion tech-
Ma HM, Choi DS, et al. Evaluation on the drug–polymer nology. AAPS PharmSciTech. 2020;21:309.
mixing status in amorphous solid dispersions at the Newman A, Zografi G. Commentary: considerations in the
early stage formulation and process development. J measurement of glass transition temperatures of phar-
Pharm Innov. 2013;8:163–74. maceutical amorphous solids. AAPS PharmSciTech.
Mahlin D, Bergstrom CAS. Early drug development 2020;21:26.
predictions of glass-forming ability and physical sta- Newman A, Reutzel-Edens SM, et al. Coamorphous active
bility of drugs. Eur J Pharm Sci. 2013;49(2):323–32. pharmaceutical ingredient–small molecule mixtures:
Mann AKP, Schenck L, et al. Producing amorphous solid considerations in the choice of Coformers for enhanc-
dispersions via co-precipitation and spray drying: ing dissolution and oral bioavailability. J Pharm Sci.
impact to physicochemical and biopharmaceutical 2018;107(1):5–17.
properties. J Pharm Sci. 2018;107:183–91. Oksanen CA, Zografi G. The relationship between the
Mansuri R, Patel D, et al. In depth investigation of quanti- glass transition temperature and water vapor absorp-
tative analytical and bioanalytical techniques of tion by poly(vinylpyrrolidone). Pharm Res. 1990;7(9):
hepatitic drug sofosbuvir in different matrices: a 654–7.
review. J Anal Pharm Res. 2018;7(2):206–20. Pajula K, Taskinen M, et al. Predicting the formation and
Marsac PJ, Konno H, et al. A comparison of the physical stability of amorphous small molecule binary mixtures
stability of amorphous felodipine and nifedipine from computationally determined FloryHuggins
systems. Pharm Res. 2006;23:2306–16. interaction parameter and phase diagram. Mol Pharm.
Marsac PJ, Konno H, et al. Recrystallization of nifedipine 2010;7(3):795–804.
and felodipine from amorphous molecular-level solid Palermo RN, Anderson CA, et al. Review: use of thermal,
dispersions containing poly(vinylpyrrolidone) and diffraction, and vibrational analytical methods to deter-
sorbed water. Pharm Res. 2008;25(3):647–56. mine mechanism of solid dispersion stability. J Pharm
Marsac PJ, Rumondor ACF, et al. Effect of temperature Innov. 2012;7:2–12.
and moisture on the miscibility of amorphous Park K. Drug release mechanism from amorphous solid
dispersions of felodipine and poly(vinyl pyrrolidone). dispersions. J Control Release. 2015;211:171.
J Pharm Sci. 2010;99(1):169–85. Patel KP, Pathak CJ, et al. Nanostructured lipid carrier – a
Martin C. Twin screw extruders as continuous mixers for novel dosage form to improve the oral bioavailability of
thermal processing: a technical and historical perspec- Lopinavir. Eur J Biomed Pharm Sci. 2015;2(2):295–311.
tive. AAPS PharmSciTech. 2016;17(1):3–19. Patel S, Kou X, et al. Mechanical properties and tableting
Masters K. Spray drying handbook. Burnt Mill: Longman behavior of amorphous solid dispersions. J Pharm Sci.
Scientific & Technical, Wiley; 1991. 2017;106:217–23.
Mcnevin M, Liu Y, et al. Pseudopolymorphs of an hcv Patil H, Tiwari R, et al. Hot-melt extrusion: from theory to
ns5a inhibitor and uses thereof. International Patent application in pharmaceutical formulation. AAPS
Publication 2015/065817A1. 2015. PharmSciTech. 2016;17(1):20–42.
Meng F, Dave V, et al. Qualitative and quantitative Patterson JE, James MB, et al. Melt extrusion and spray
methods to determine miscibility in amorphous drug- drying of carbamazepine and dipyridamole with
polymer systems. Eur J Pharm Sci. 2015;77:106–11. polyvinylpyrrolidone/vinyl acetate copolymers. Drug
Mishra DK, Dhote V, et al. Amorphous solid dispersion Dev Ind Pharm. 2008;34:95–106.
technique for improved drug delivery: basics to clinical Paudel A, Van den Mooter G. Influence of solvent compo-
applications. Drug Deliv Transl Res. 2015;5:552–65. sition on the miscibility and physical stability of
Mistry P, Mohapatra S, et al. Role of the strength of drug– naproxen/PVP K 25 solid dispersions prepared by
polymer interactions on the molecular mobility and cosolvent spray-drying. Pharm Res. 2012;29:251–70.
Paudel A, Van Humbeeck J, et al. Theoretical and experi- Saboo S, Moseson DE. Patterns of drug release as a
mental investigation on the solid solubility and function of drug loading from amorphous solid
miscibility of naproxen in poly(vinylpyrrolidone). dispersions: a comparison of five different polymers.
Mol Pharm. 2010;7(4):1133–48. Eur J Pharm Sci. 2020;155:105514.
Paudel A, Nies EG, et al. Relating hydrogen-bonding Saboo S, Mugheirbi NA, et al. Congruent release of drug
interactions with the phase behavior of naproxen/PVP and polymer: a “sweet spot” in the dissolution of
K 25 solid dispersions: evaluation of solution-cast and amorphous solid dispersions. J Control Release.
quench-cooled films. Mol Pharm. 2012;9(11):3301–17. 2019;298:68–82.
Paudel A, Geppi M, et al. Structural and dynamic Saerens L, Dierickx, et al. Raman spectroscopy for the
properties of amorphous solid dispersions: the role of in-line polymer-drug quantification and solid state char-
solid-state nuclear magnetic resonance spectroscopy acterization during pharmaceutical hot-melt extrusion
and relaxometry. J Pharm Sci. 2014;103:2635–62. process. Eur J Pharm and Biopharm. 2011;77:158–63.
Paus R, Ji Y, et al. Predicting the solubility advantage of Sanghvi T, Katstra J, et al. Formulation development of
amorphous pharmaceuticals: a novel thermodynamic amorphous dispersions, Pharmaceutical amorphous
approach. Mol Pharm. 2015;12(8):2823–33. solid dispersions. 1st ed. Hoboken: Wiley; 2015.
Poozesh S, Setiawan N, et al. Assessment of predictive Sarode AL, Sandhu H, et al. Hot melt extrusion (HME) for
models for characterizing the atomization process in a amorphous solid dispersions: predictive tools for
spray dryer’s bi-fluid nozzle. Chem Eng Sci. 2018;180: processing and impact of drug–polymer interactions
42–51. on supersaturation. Eur J Pharm Sci. 2013;48:371–84.
Porfirio T, Galindo-Rosales FJ, et al. Rheological charac- Schenck L, Mann AKP, et al. Building a better particle:
terization of polymeric solutions used in spray drying Leveraging physicochemical understanding of amor-
process. Eur J Pharm Sci. 2021;158:105650. phous solid dispersions and a hierarchical particle
Price DJ, Ditzinger F, et al. Approaches to increase mecha- approach for improved delivery at high drug loadings.
nistic understanding and aid in the selection of precipita- Int J Pharm. 2019;559:147–55.
tion inhibitors for supersaturating formulations – a Scott RW, Wang F, et al. Solid forms of an antiviral com-
PEARRL review. J Pharm Pharmacol. 2019;71:483– pound. US Patent Publication 2013/0324496A1. 2013.
509. Serajuddin ATM. Solid dispersion of poorly water-soluble
Prudic A, Ji Y, et al. Thermodynamic phase behavior of drugs: early promises, subsequent problems, and recent
api/polymer solid dispersions. Mol Pharmaceut. breakthroughs. J Pharm Sci. 1999;88:1058–66.
2014;11:2294–304. Shah N, Sandhu H, et al. Solid complexes with ionic
Purohit HS, Taylor LS. Miscibility of Itraconazole- polymers: pharmaceutical technology is pleased to rec-
Hydroxypropyl methylcellulose blends: insights with ognize the winners of its innovations in pharma science
high resolution analytical methodologies. Mol Pharm. awards. Pharm Technol. 2008;32(12):2.
2015;12(12):4542–53. Shah N, Iyer RM, et al. Improved human bioavailability of
Qi S, Belton P, et al. Compositional analysis of low Vemurafenib, a practically insoluble drug, using an
quantities of phase separation in hot-melt-extruded amorphous polymer-stabilized solid dispersion
solid dispersions: a combined atomic force micros- prepared by a solvent-controlled coprecipitation pro-
copy, photothermal Fourier-transform infrared cess. J Pharm Sci. 2013;102(3):967–81.
microspectroscopy, and localized thermal analysis Shanbhag A, Rabel S, et al. Method for screening of solid
approach. Pharm Res. 2011;28(9):2311–26. dispersion formulations of low-solubility compounds-
Qian F, Tao J, et al. Mechanistic investigation of Miniaturization and automation of solvent casting and
Pluronic® based nano-crystalline drug-polymer solid dissolution testing. Int J Pharm. 2008;351(1–2):209–18.
dispersions. Pharm Res. 2007;24(8):1551–60. Shepard KB, Dower AM, et al. Solvent-assisted secondary
Read MD, Coppens KA, et al. Hot melt extrusion technol- drying of spray-dried polymers. Pharmaceut Res.
ogy for the manufacture of poorly soluble drugs with 2020;37:156.
controlled release dissolution profiles. Sibik J, Lobmann K, et al. Predicting crystallization of
ANTEC:1203-1206. 2010. amorphous drugs with terahertz spectroscopy. Mol
Repka M, Langley N. Melt extrusion; materials, Technol- Pharm. 2015;12(8):3062–8.
ogy and drug product design. New York: Siew A. Solving Poor Solubility with Amorphous Solid
Springer; 2013. Dispersions: Weighing the pros and cons of hot-melt
Rowe R, Shesky P, et al. Handbook of pharmaceutical extrusion and spray drying. Pharm Tech. 2014;38(1):
excipients. 4th ed. Washington: APhA 30–5.
Publications; 2010. Simonelli AP, Mehta SC, et al. Dissolution rates of high
Rumondor A, Stanford L, et al. Effects of polymer type energy polyvinylpyrrolidone (PVP)-sulfathiazole
and storage relative humidity on the kinetics of coprecipitates. J Pharm Sci. 1969;58(5):538–49.
felodipine crystallization from amorphous solid Singh H, Atef E, et al. High throughput screening of solid
dispersions. Pharm Res. 2009a;26(12):2599–606. dispersion using solvent evaporation technique aaps
Rumondor ACF, Marsac PJ, et al. Phase behavior of poly annual meeting and exposition. San Diego: San
(vinylpyrrolidone) containing amorphous solid Diego Convention Center; 2007.
dispersions in the presence of moisture. Mol Pharm. Singh A, Bharati A, et al. Effect of compression on the
2009b;6(5):1492–505. molecular arrangement of Itraconazole–Soluplus solid
326 S. Page et al.
dispersions: induction of liquid crystals or exacerba- opinion of the UNGAP network. Adv Drug Deliv
tion of phase separation? Mol Pharmaceut. 2016;13(6): Rev. 2021;171:289–331.
1879–93. Vo CL, Chulhun Park C, et al. Current trends and future
Song Y, Zemlyanov D, et al. Acid–base interactions of perspectives of solid dispersions containing poorly
polystyrene sulfonic acid in amorphous solid water-soluble drugs. Eur J Pharm Biopharm. 2013;85
dispersions using a combined UV/FTIR/XPS/ssNMR (3):799–813.
study. Mol Pharm. 2016;13(2):483–92. Wabuyele BW, Sotthivirat S, et al. Dispersive Raman
Sun DD, Lee PI. Evolution of supersaturation of amor- spectroscopy for quantifying amorphous drug content
phous pharmaceuticals: the effect of rate of supersatu- in intact tablets. J Pharm Sci. 2017;106:579–88.
ration generation. Mol Pharm. 2013;10(11):4330–46. Warren DB, Benameur H, et al. Using polymeric precipi-
Suzuki N, Kasahara K, et al. Physical property of tation inhibitors to improve the absorption of poorly
troglitazone, an equal mixture of four stereoisomers. water-soluble drugs: a mechanistic basis for utility. J
Int J Pharm. 2002;248(1–2):71–80. Drug Target. 2010;18:704–31.
Swinney K, Herman J, et al. Configuration of an Weuts I, Van Dycke F, et al. Physicochemical properties
automated screening tool to facilitate solid dispersion of the amorphous drug, cast films, and spray dried
development AAPS annual meeting and exposition. powders to predict formulation probability of success
Los Angeles: Los Angeles Convention Center; 2009. for solid dispersions: Etravirine. J Pharm Sci. 2011;100
Takano R, Maurer R, et al. Formulating amorphous solid (1):260–74.
dispersions: impact of inorganic salts on drug release Williams III RO, Hanada M, et al. Drug compositions
from tablets containing itraconazole-HPMC extrudate. containing porous carriers made by thermal or fusion-
Mol Pharm. 2020;17(8):2768–78. based processes. US Patent Publication 2020/
Tan DK, Davis DA, et al. Innovations in thermal 0405643A1. 2020.
processing: hot-melt extrusion and KinetiSol® dispers- Wyttenbach N, Kuentz M. Glass-forming ability of
ing. AAPS PharmSciTech. 2020;21:312. compounds in marketed amorphous drug products.
Tanno F, Nishiyama Y, et al. Evaluation of hypromellose Eur J Pharm Biopharm. 2017;112:204–8.
acetate succinate (HPMC-AS) as a carrier in solid Wyttenbach N, Janas C, et al. Miniaturized screening of
dispersions. Drug Dev Ind Pharm. 2004;30(1):9–17. polymers for amorphous drug stabilization (SPADS):
Taylor LS, Zografi G. Spectroscopic characterization of rapid assessment of solid dispersion systems. Eur J
interactions between PVP and indomethacin in amor- Pharm Biopharm. 2013;84:583–98.
phous molecular dispersions. Pharm Res. 1997;14: Wyttenbach N, Kirchmeyer W, et al. Theoretical
1691–8. considerations of the prigoginedefay ratio with
Thybo P, Hovgaard L, et al. Scaling up the spray drying regard to the glass-forming ability of drugs from
process from pilot to production scale using an undercooled melts. Mol Pharm. 2016;13:241–50.
atomized droplet size criterion. Pharm Res. 2008;25 Xi H, Ren J, et al. The effect of inorganic salt on disintegra-
(7):1610–20. tion of tablets with high loading of amorphous solid
Tian B, Wang X, et al. Theoretical prediction of a phase dispersion containing copovidone. Pharm Res. 2020;37
diagram for solid dispersions. Pharm Res. 2015;32: (4):70.
840–51. Xie T, Taylor LS. Improved release of celecoxib from high
Usui F, Maeda K, et al. Inhibitory effects of water-soluble drug loading amorphous solid dispersions formulated
polymers on precipitation of RS-8359. Int J Pharm. with polyacrylic acid and cellulose derivatives. Mol
1997;154(1):59–66. Pharm. 2016;13(3):873–84.
Van den Mooter G, Wuyts M, et al. Physical stabilisation Yamashita T, Kokubo T, et al. Antiprecipitant screening
of amorphous ketoconazole in solid dispersions with system for basic model compounds using bio-relevant
polyvinylpyrrolidone K25. Eur J Pharm Sci. 2001;12 media. J Assoc Lab Automat. 2010;15(4):306–12.
(3):261–9. Yang Z, Han CD. Rheology of miscible polymer blends
Van Eerdenbrugh B, Taylor LS. Small scale screening to with hydrogen bonding. Macromolecules. 2008;41(6):
determine the ability of different polymers to inhibit 2104–18.
drug crystallization upon rapid solvent evaporation. Yoo SU, Krill SL, et al. Miscibility/stability considerations
Mol Pharm. 2010;7(4):1328–37. in binary solid dispersion systems composed of func-
Van Eerdenburgh B, Baird JA, et al. Crystallization ten- tional excipients towards the design of multi-
dency of active pharmaceutical ingredients following component amorphous systems. J Pharm Sci. 2009;98
rapid solvent evaporation—classification and compari- (12):4711–23.
son with crystallization tendency from undercooled Yu L. Amorphous pharmaceutical solids: preparation,
melts. J Pharm Sci. 2010;99(9):3826–38. characterization and stabilization. Adv Drug Deliv
Vandecruys R, Peeters J, et al. Use of screening method to Rev. 2001;48:27–42.
determine excipients which optimize the extend and Zhao Y, Inbar P, et al. Prediction of the thermal phase
stability of supersaturated drug solutions and applica- diagram of amorphous solid dispersions by Flory-
tion of this system to solid formulation design. Int J Huggins theory. J Pharm Sci. 2011;100(8):3196–207.
Pharm. 2007;342:168–75. Ziaee A, Albadarin AB, et al. Spray drying ternary amor-
Vinarov Z, Abrahamsson B, et al. Current challenges and phous solid dispersions of ibuprofen – An investigation
future perspectives in oral absorption research: an into critical formulation and processing parameters.
Eur J Pharm Biopharm. 2017;120:43–51.
Melt Extrusion
9
Stephen A. Thompson, Daniel A. Davis Jr., James C. DiNunzio,
Charlie Martin, Robert O. Williams III, and Feng Zhang
Abstract well-characterized nature of the process

Techniques to overcome poor aqueous solubil- provides for ease of scale-up and process opti-
ity of active pharmaceutical ingredients (APIs) mization while also affording benefits of con-
continue to gain interest, with a reported 30% tinuous manufacturing and adaptability to
of marketed compounds classified as BCS II process analytical technology in an ever-
and 10% classified as BCS IV. Approximately changing regulatory and fiscal environment
70% of new chemical entities under develop- where manufacturing efficiencies must be
ment may be classified as BCS II and another maximized to reduce cost and improve product
20% as BCS IV (Siew Pharm Technol 39:20– quality. This chapter details the basic engi-
27, 2015). Driven by this need to enable neering principles of the melt extrusion pro-
therapies of poorly soluble compounds cess and provides a fundamental
through the generation of amorphous solid understanding of formulation development of
dispersions, pharmaceutical scientists have melt-extruded solid dispersions for bioavail-
adapted a number of technologies from other ability enhancement. Several recent case stud-
industries to provide reliable and robust drug ies are also described to highlight the
product manufacturing. Melt extrusion is an technology’s applicability to developmental
example of such a technology. Originally and marketed products within the industry.
developed in the plastic industry in the
1960s, it has been applied to pharmaceutical Keywords
systems over the last three decades to generate Melt extrusion · Specific mechanical energy
some of the most cutting-edge delivery (SME) · Amorphous solid dispersions
systems seen in the industry to date. The (ASDs) · Solubility limitations · Quality by
design (QbD) · Screws · Melt residence time ·
Convective heat transfer
S. A. Thompson · D. A. Davis Jr. · R. O. Williams III ·
F. Zhang (*)
Division of Pharmaceutics, College of Pharmacy,
The University of Texas at Austin, Austin, TX, USA 9.1 Introduction
J. C. DiNunzio Extrusion was developed in the 1960s and was
Pharmaceutical Sciences & Clinical Supplies, Merck & primarily used by the rubber, food, and plastic
Co., Inc., Kenilworth, NJ, USA
industries. It was adapted to the pharmaceutical
C. Martin industry in the 1980s with numerous patents and
Leistritz Extrusion, Somerville, NJ, USA
328 S. A. Thompson et al.
research papers published since then in addition Based on decades of research and recent applica-
to the approval of commercial products. Since tion of the technology to the production of phar-
then, the advantages of extrusion for preparing maceutical products, one major facet of design
pharmaceuticals have become clear. Extrusion has become evident: formulation and process go
provides both continuous processing and ease of “hand in hand” to yield desired critical product
adaptability to in-line analytical technology; attributes. Melt viscosity and solubility of a for-
twin-screw melt extrusion has been extensively mulation directly influence processability factors
utilized in several pharmaceutical technologies, such as motor load, pressure, and residence time
including solid dispersion production, advanced distribution (Repka et al. 1999; Schilling et al.
device manufacturing, 3D printing of dosage 2007). Similarly, screw design, throughput, and
forms, continuous melt granulation, abuse deter- die geometry may impact solubilization and deg-
rence formulations, and the development of radation rates of the formulation due to microen-
controlled-release products for oral delivery vironmental changes within the control volume
(Agrawal et al. 2015; Jedinger et al. 2015; Keen (DiNunzio et al. 2010b; DiNunzio 2010). During
et al. 2015; Khaled et al. 2014; Maniruzzaman early development using material-sparing
et al. 2015a, b). Available in multiple variants approaches, it is critical to understand the inter-
based on equipment geometry and processing play of formulation and process variables to allow
temperature, twin-screw melt extrusion has for rational design. Identifying primary factors
emerged as a viable technology in the pharma- governing critical product attributes and the abil-
ceutical industry to produce solid dispersions ity to address issues associated with their control
(Breitenbach and Mägerlin 2003; Martin 2016; are crucial to establishing a basic formulation and
Patil et al. 2015; Crowley et al. 2008). While process to support developmental activities. In a
each of the applications has had a significant later development, full optimization using a qual-
impact on pharmaceutical development, this ity by design (QbD) methodology requires an
chapter is dedicated to a discussion of amorphous understanding of the complex formulation-
solid dispersions to address solubility limitations. process interaction to establish a robust and
Covered within this chapter are equipment selec- highly efficient design space to support routine
tion, a detailed discussion of the fundamental market production. Given that early development
theory of melt extrusion, preformulation and later-stage optimization are both driven by
procedures, process considerations, critical the interplay between formulation and process,
aspects of formulation design for producing sta- this section describes melt extrusion’s fundamen-
ble amorphous solid dispersions, and scale-up tal engineering aspects.
strategies. Major case studies covering the most Pharmaceutical melt extruders are classified
recent commercial applications of melt extrusion using several fundamental characteristics of the
for improving bioavailability through solubility equipment, including the number of screws and
enhancement of poorly water-soluble compounds the direction of rotation of the screws (corotating
are also presented. or counter-rotating). Single-screw extruders can
be used to provide more consistent pumping (Kim
and Kwon 1996a, b; Luker 2003). Example
9.2 Equipment Design products prepared by single-screw extrusion
and Engineering Principles include oral filaments, catheter tubes, and multi-
layer films. Twin-screw extruders are frequently
Twin-screw melt extrusion was developed during used in compounding applications such as amor-
the 1930s and has been extensively studied for a phous solid dispersions due to their higher mixing
number of polymer processing operations efficiencies (Thiele 2003; McGinity et al. 2007).
(Mollan 2003; Thiele 2003). Today, the theory For twin-screw extruders, the direction of rotation
of the technology has been well described and the and type of intermesh play a major role in the
processing equipment extensively developed. equipment efficiency and purpose. Extruder
9 Melt Extrusion 329
screws relative to one another may either be 4. Screw elements

operated in a corotating or counter-rotating con- 5. Dies
figuration, with corotating units having both
The basic equipment parts, their variations,
extruder screws rotating in the same direction.
and their role in extrusion have been summarized
This geometry is also associated with opposing
in Table 9.1. For more detailed information, the
surface velocities in the intermesh, which results
reader is directed to Brown et al. (2014), Crowley
in a self-wiping effect. In this system, the rota-
et al. (2007), Martin (2016), and Patil
tional motion of one screw wipes material from
et al. (2015).
the other to further convey material down the
Each of the major components of the extruder
length of the barrel. Counter-rotating
needs to be considered when designing a robust
intermeshing designs can be utilized during the
process. The critical zones that make up the barrel
formulation of heat-sensitive materials as this
from Table 9.1 are shown in Fig. 9.1. Each sec-
type of design provides advantages for these
tion serves specific functionality, allowing for
compounds, which may require a tight residence
multiple-unit operations to be achieved when
time distribution. For the purposes of this chapter,
matched with the desired screw design.
equipment discussion will focus primarily on the
Within twin-screw extruders, pools result
intermeshing corotating twin-screw extruder.
within screw flights and barrel walls the rotational
While frequently generalized as a single-unit
movement of the screw (Fig. 9.2). The material
operation, melt extrusion is better understood as a
follows the perimeter path along the cross section
combination of multiple operations, each
of the barrel in the screw channel. A peak shear
contributing to the operational performance. Fun-
occurs within the overflight and intermesh
damentally, melt extrusion is divided into seven
regions, while channel regions exhibit a shear
basic steps (Schenck et al. 2011):
minimum (Thiele 2003). The intermesh region
1. Metering of raw materials (independent of and lobal pool (where extension mixing occurs)
screws rpms) are also comparatively high shear regions. This
2. Solids conveying unique mixing pattern generally results in greater
3. Melting efficiencies than single-screw extruders and
4. Mixing allows for effective compounding on the twin-
5. Devolatilization screw extruder platform.
6. Pumping Based on the different screw elements featured
7. Shaping in Table 9.1 and varying the geometries from
Fig. 9.3, the screw design can be optimized for
any process. Generally, feed regions contain
9.2.1 Basic Equipment Description elements with the longest pitch to maximize con-
veyance from the zone, while transitions to pres-
Modern twin-screw extruders offer a modular sured zones will be supplemented with a decrease
design that allows scientists to incorporate basic in pitch. Shear-inducing elements are to facilitate
unit operations in desired sequence to achieve the melting and mixing (distributive and dispersive).
desired product attributes. This modularity of the Due to most pharmaceutical agents’ sensitivity,
process gives rise to a need to address the general the minimum number of mixing elements is used
description of the equipment. A twin-screw to enhance distributive and dispersive mixing and
extruder consists of several basic components, to avoid degradation.
each of which participates in the cumulative man- During twin-screw extrusion, there are two
ufacture of the drug product. These parts are: main types of mixing: distributive mixing and
dispersive mixing. Distributive mixing involves
1. Gearbox and motor
melt division and recombination, while dispersive
2. Barrel with integral heating and cooling
mixing involves shear and elongational mixing. A
3. Shafts
Table 9.1 Twin-screw extruder equipment parts, variations, and functions

Parts and
Component Hardware feature variations Features/notes
1. Gearbox and Size of extruder Screw diameter 12 to 18 mm for development
motor up to 100 mm + for commercial
Rotation Corotating Intermeshing
Counter-rotating Intermeshing
2. Barrels L:D ratio Length-to-screw 25:1 to 60:1
diameter ratio
Configuration Clamshell
Sequential blocks
Zone types (Fig. 9.1) Solid feed Located upstream to introduce
materials
Liquid feed Intro. liq. and semi-solids via inj.
nozzle downstream
Eliminate premixing
Side feed Intro. solids downstream
Reduces heat exposure to
sensitive materials
Venting Remove volatile materials
(moisture, residual solvents)
With or without vacuum
Closed segment Site for mixing, kneading,
pumping
3. Screw Shaft-element interface (torque load Key Least torque capacity
elements and transferred to shaft from screws) Spline More torque capacity
shafts Asymmetric Most torque capacity
spline
4. Screw Types Conveying Move material forward or
elements backward
Mixing Distributive, planar
Kneading Distributive and dispersive
mixing
Limited by drug-substance
sensitivity
Offset paddles (30 , 60 , 90 )
Forward, neutral, reverse
5. Dies Common geometries Rod Pelletized and milled
Film Easily cooled on chill roller
Fig. 9.1 Types of block

sections from a sequential
block barrel design, moving
from left to right, solid feed,
liquid feed, side feed
with vent

critical geometric
descriptors of twin-screw
extruder elements.
(Courtesy of Leistritz)
Fig. 9.3 Geometric

descriptors for conveying
elements
graphic illustration of two types of mixing is can be performed under vacuum and may occur at
presented in Fig. 9.4. Screw design can be made a series of positions along the barrel. Devolati-
shear intensive or passive. Mixing elements may lization efficiency can be improved with longer
be dispersive, distributive, or a balance of each/ residence time under the vent, higher surface
both. Screw elements (e.g., wide-lobe kneading renewal, more nucleation, growth and rupture of
element) that accentuate extensional mixing and bubbles, and higher vacuum levels (Jerman
planar shear effects are dispersive in nature, as 2006).
compared to elements that facilitate melt division/ The die can also be a factor in maintaining the
recombination (e.g., combing mixer), which are stability of the material. Many pharmaceutical
more distributive. active ingredients and polymers are heat sensi-
Another common operation during twin-screw tive. Depending on the pressure, formulation vis-
melt extrusion is devolatilization, which is a pro- cosity, screw design, and screws rpms utilized,
cess to remove various amounts of gases from the Todd et al. correlated the pressure buildup to a
process melt including residual solvent, water, localized temperature increase where the temper-
and other volatile contaminants. Vent zones facil- ature increases by up to 1 C for every 30 psi
itate the removal of volatiles. The venting process generated (Todd 1995). For common
acceleration of the surface extrudate layer, manip-

ulation of material linear velocity, die landing
length, die surface coating/polishing, and temper-
ature can all be used to reduce shark skinning in
production (Perdikoulias and Dobbie 2003;
Kulikov and Hornung 2001). These surface
defects of extrudates are of less a concern for
melt extrusion of amorphous solid dispersions
for bioavailability enhancement since those
extrudates are generally pelletized and milled
into granules.
Feeders
During manufacturing runs, a continuous stream
of material is metered to the extruder when
operating at steady state. This is achieved by
Fig. 9.4 Graphic illustration of distributive and disper- gravimetric or volumetric feeders to meter solid,
sive mixing
liquid, or gaseous materials into the process sec-
tion. A gravimetric feeder is simply a volumetric
pharmaceutical applications where pressures are
feeder situated on a load cell that is used to
generally 300–600 psi at the 18-mm scale, this
modulate the feed mechanism to maintain a con-
can result in local temperature increases of
stant mass flow rate to the extruder. Selection of
10–20 C due to viscous dissipation during tran-
feed screw geometry can play a significant role in
sit. When processing heat-sensitive products,
process uniformity, reducing the feed stream’s
changes to the die geometry can minimize pres-
avalanching and pulsing behavior (Schenck
sure buildup and reduce impurity formation dur-
et al. 2011; Doetsch 2003). Examination of
ing production. Pressure-generating devices, such
avalanching behavior shows that the period of
as a gear pump, can be mated to the twin-screw
pulsing is a function of the flight design. For
extruder to help manage pressures and therefore
example, a transition from a single-flight to a
also melt temperature.
dual-flight system operated at 60 rpm reduces
For the extrusion of products with precision
the pulse period from 1 to 0.3 s. Further monitor-
dimension such as sheets and filament, other more
ing of material feed during production with gravi-
general issues with the quality of extruded
metric feeders is achieved using a feedback
products are die swell, shark skinning, and melt
control loop that monitors real-time weight loss
fracture (Perdikoulias and Dobbie 2003). Die
and adjusts screw speed to maintain a consistent
swell is caused by the entropic restriction placed
feed rate. Low bulk density and poorly flowing
on material through the die that causes polymer
materials can present unique feeding challenges
molecules to orient parallel to the direction of
even within the gravimetric feeder. In addition,
flow (Perdikoulias and Dobbie 2003; Wang and
some polymers commonly used in pharmaceuti-
Drda 1997). Upon exiting, the polymer chains
cal extrusion contain high moisture content,
reorient to the natural random coiled structure
which can affect the consistency of feeding and
due to elastic recovery, resulting in swelling of
extruding. These include polyvinylpyrrolidone
the material. Shark skinning is the appearance of
(PVP or povidone) and PVP vinyl acetate
surface roughness of the extrudate material on
(PVPVA, copovidone, or VA64). For polymers
exit from the die (Kulikov and Hornung 2001).
such as these, it is recommended to dry overnight
The magnitude of these defects is directly related
at a temperature sufficient to remove excess mois-
to both material properties and die geometry.
ture. One paper determined that 40 C was insuf-
Originating from factors related to the
ficient to affect PVP moisture content before
extrusion (Alshahrani et al. 2015), while others points along the barrel to prevent off-gassing
have used 70 C for drying VA64 with success during manufacture.
(Ma et al. 2019). If not dried, material can cake in
the feed area over long extrusion periods and
cause flowability issues. Self-wiping auger 9.2.2 Melt Extrusion Processing
designs and non-intermeshing undercut designs
can improve material throughputs, while driven As a continuous process, melt extrusion can be
agitators can minimize powder bridging during viewed to have two specific operational modes:
manufacture. dynamic (start-up and shutdown) and steady
In the case of liquid feed streams, control can state. In the dynamic mode, a mass balance will
be achieved using volumetric or gravimetric not be achieved. This stage occurs during start-up
systems depending on the feed rates and material or major system perturbations and represents only
properties. These systems are generally closed a very small sequence of the overall run. The
systems that feed directly into the process section majority of the process will be conducted at
using liquid injection ports. Systems can be steady state where a mass balance is achieved,
designed to perform at ambient or elevated and no accumulation occurs. While transient
temperatures, which allow the addition of low- processing is difficult to model, steady-state oper-
melting-point solids as liquids into the system. ation of extrusion processes has been well
Similar to gravimetric solids feeders, liquid described. In addition to the engineering
feeders must be able to provide a consistent and principles, the process may also be described
reproducible feed stream at steady state that can from the perspective of the formulation based on
be ensured through proper equipment and process the drug-substance melting temperature in rela-
design. tion to the processing conditions. This section
Selection of the type of pump is dependent on provides fundamental engineering and process
the viscosity of the liquid. Nozzle design and concepts as related to pharmaceutical production
positioning relative to the screw design ensure using melt extrusion.
that liquid is fed into the extruder at sufficient
pressure to prevent clogging (Schenck et al. Steady-State Processing and Production
2011). Accurate feeding can be achieved with Feedback
gear-type pumping systems or pulse dampening During extrusion operations, several fundamental
piston pump. Smart control systems that can engineering principles describe the behavior of
assess changes to processing and feed conditions the system. As mentioned previously, most
can also help compensate for drift that occurs twin-screw melt extrusion applications are
over extended runs, thereby allowing the system conducted in a “starve-fed” manner where mate-
to maintain steady state. Generally speaking, liq- rial added into the system is controlled by the
uid pumps do not drift and the scales are used for feeder (Thiele 2003). In such cases, the feed rate
monitoring purpose. Gas-assisted and supercriti- is independent of the extruder screw speed, and
cal fluid injections have been applied to pharma- the rate of material conveyance into the flighted
ceutical processes (Verreck et al. 2006a, b, 2007; sections in the extruder is greater than the rate at
Lyons et al. 2007; Terife et al. 2012). In the case which material is being provided to the system.
of amorphous solid dispersions, gas-assisted Establishing a control volume around the process
injection can increase the surface area and section of the extruder, as illustrated in Fig. 9.5,
increase the drug’s release. In general, the feed- allows for the establishment of a mass balance
stock will be a gas/supercritical liquid and will that accounts for all possible addition and
use metering and injection systems where critical removal streams from the process. When
pressures help to add material into the process. operating at steady state, a mass balance is
Applications will also be supported by screw achieved where the mass into the system is
design, where melt seals will be formed at various equal to that leaving the system.
factors impacting the RTD are feed rate and

screw profile.
Beyond understanding the overall residence
time within the system, melt residence time can
play a critical role in product performance. Melt
residence time is defined as the time during which
material exists within the molten state. As mate-
rial enters the system, it displays a specific set of
Fig. 9.5 Steady-state control volume for hot-melt-extru-
solid-state characteristics and has a defined tem-
sion operations
perature, which is generally that of the ambient
environment. During conveyance from the feed
In the majority of applications, 100% fill of the
zone to downstream regions within the extruder,
extruder will not be achieved across the length of
mechanical energy is imparted into the material
the process section (Todd 1995). Most screw
along with conductive heat transfer from the bar-
flighted elements will be starved, with only
rel wall. Most of the total energy originates from
100% fill achieved in the higher-pressure regions
the screws and this energy drives a temperature
leading to kneading/mixing elements and the die.
increase of the material. At some point in a twin-
High levels of fill will also be achieved in the
screw extruder, generally the first set of kneaders
kneading/mixing sections themselves because of
along the length of the extruder, the material
the limited conveyance provided by these types of
transitions from solid to melt, resulting in peak
elements. As such, one can also define a fill
observed shear and a well-delineated change of
volume within the extruder system and correlate
flow characteristics. This behavior is generally
residence time as a function of mass flow rate and
due to one or more of the components melting
volume. It is important to note that residence time
or achieving a temperature greater than the glass
is not a discrete value but a distribution that can
transition temperature. At this point, the material
be described in average or cumulative terms for
may be considered a melt, and the melt residence
any fraction of material. A hypothetical residence
time begins. At this point, the polymer starts to
time distribution and cumulative plot are shown
exhibit properties characteristic of a solvent as a
in Fig. 9.6. For pharmaceutical compounding
result of lower viscosity and greater molecular
applications, the residence time and
mobility. As the molten dissolution process
corresponding distribution determine the ability
begins, the drug substance will convert from crys-
to achieve the desired target product attributes
talline to amorphous, defining the time zero point
without excessive degradation of the materials.
for the dissolution process and signaling the point
In some cases, these attributes will be defined by
at which materials start to become more suscepti-
the mean residence time ðt Þ , whereas other
ble to degradation due to the absence of crystal-
applications, such as heat-sensitive material
line morphology, which can inhibit certain
processing, may best be defined by the time for
decomposition processes. Exposure to the high-
99% of the material to leave the extruder (t99%).
energy environment also increases the rates at
For early development work, one can manipulate
which these reactions occur. It is not surprising
residence time and distribution by changing feed
to see that the melt residence time plays an impor-
rate and screw speed. Changes to screw design
tant role in the production of high-melting-point
and formulation can also impact residence time
compounds and heat-sensitive compounds. In
distribution and often prove to be useful
these cases, optimization of barrel temperature
parameters to control during development
can be used, albeit to a limited extent, to control
(Schenck et al. 2011; Todd 1995; Ganjyal and
the melt region’s size along the length of the
Hanna 2002). A summary of these process
screw. Reducing barrel temperatures in concert
parameters and the corresponding impact is
with screw design and screw speed can increase
provided in Table 9.2. The most significant
heat transfer rates from the system and help to
Fig. 9.6 Hypothetical

residence time distribution
profile plotted as discrete
value and cumulative
curves
Table 9.2 General impact of process parameters on the residence time of melt-extruded products
Effect on residence time
Variable, RTD
Parameter Magnitude effect Description
Process Proportional #, # Reducing will result in shorter
length mean residence time
Mass flow Major (greatest ", # Increasing will result in shorter
rate effect) mean residence time and tighter RTD
Screw Major (less than #, # Reduced number of mixing/kneading
design rate) sections and reverse elements will contribute
to shorter mean residence time and tighter RTD
Screw speed Intermediate ", # Increasing will generally mean shorter residence
by a little and also will widen residence time distribution
Temperature Minor ", # Increasing will generally lower viscosity and reduce residence
time.
Some formulations may thin excessively limiting flow
Formulation Minor Dependent Formulation determines viscosity and which can impact
residence time and distribution
lower localized temperatures within the process polymer and drug resisting the rotation of the
(Todd 1995). At a critical threshold, barrel tem- screw. Viewing the extruder barrel section as a
perature reduction may become too great and stationary boundary and the rotating screw as a
result in an inability to process or render target moving boundary, the shear rate of the molten
product properties. polymer can be defined as a function of geometry
During the extrusion process, the majority of and screw speed, as shown in (9.1). Further
energy is imparted due to the mechanical input understanding of the shear stress applied to the
provided by the screw, which manifests itself as system is obtained as the product of shear rate and
shear and viscous dissipation of frictional energy viscosity, described in (9.2). Given that the shear
resulting in temperature increases. Viscous dissi- rate and shear stress are functions of the clearance
pation is the irreversible conversion of mechani- geometry, a profile will be obtained across the
cal to thermal energy that results from the length of the process section as a function of
elemental design. Detailed knowledge of peak dominance of heating via viscous dissipation ver-
shear and cumulative shear experienced during sus the loss of thermal energy from the system via
production is necessary for successful process thermal conduction (Marschik et al. 2018). In
optimization and scale-up. Altering feed rate pro- other words, it predicts if extruder parameters
portionally to screw speed can also allow for will cause a temperature increase of the extruded
throughput increases without major changes to material over the set barrel temperature. In gen-
target product properties so long as maximum eral, increasing the rate of mechanical energy
peak shear does not achieve a critical threshold input per unit mass into the system by increasing
to drive degradation of the formulation. screw speed or lowering throughput rate can
cause significant increases in the temperature of
π ∙ d0 ∙ n
Shear Rate ¼ ð9:1Þ the system. Because temperature affects drug sol-
ðd 0 d i Þ
ubility, conversion from crystalline drug to amor-
phous, and drug degradation, it is important to
consider this phenomenon, particularly in larger
Shear Stress : γ ¼ τ ∙ η ð9:2Þ extruders. Its relevance and use in melt extrusion
has been recently reviewed by Thompson and
In the molten region of the process, the
Williams (2021).
materials behave as viscous liquids, and as a
After extrusion, the material is discharged
result of motion and flow restrictions, pressure
from the die and cooled using a number of differ-
fields build up across the length of the process
ent technologies. In most cases, dispersions are
section (Lim and White 1994). Pressure increases
cooled by forced air convection to room tempera-
occur as a function of screw design, with mixing
ture, which can be an effective technique when
and kneading elements that provide minimal con-
working with laboratory-scale processes where
veyance and having higher fill yielding greater
strand dimensions and linear velocities are rela-
increases. These pressure differentials provide a
tively small. The general process may be
specific function, allowing for separation and
expressed by (9.3), noting that h represents the
sealing of different unit operations along the
convective heat transfer coefficient. When
length of the screw. The magnitude of pressure
throughputs reach levels where convective
observed is also a function of the formulation.
cooling is no longer effective due to equipment
High-melt-viscosity systems will generate larger
geometries or product-specific cooling
pressures at equivalent mass flow rates compared
requirements, chill roller technology may be a
to lower viscosity formulations. An example of
viable solution. Within the chill roller, two or
this would be a system based on hypromellose
more rolls maintained at definable temperatures
acetate succinate where flow rates of 1 kg/
press the discharged extrudate to a film of desired
h through a 2.5-mm die would result in a pressure
thickness. This process serves to increase surface
of ~600 psi, whereas a similar system using
area while also providing direct contact for con-
copovidone would only yield ~200 psi when
ductive heat transfer between solids. This is anal-
processed at 170 C. Pressure buildup also results
ogous to the convective situation; however, h is
in a localized temperature increase due to viscous
replaced by the conductance of the roller material
dissipation, which can impact heat-sensitive
(U). Generally, conductance values of roller
materials’ product attributes.
materials are significantly greater than the
Another consideration during processing is the
coefficients provided by convective systems,
potential for local temperature increases in the
which utilize airflow from fans. Further improv-
barrel due to viscous heating. While not com-
ing heat transfer rates for chill roller systems is the
monly used in pharmaceutical processing, the
substantially greater surface area and reduced film
Nahme number, which is sometimes referred to
thickness compared to the cylindrical extrudate.
as the Nahme-Griffith number or Brinkman num-
Additionally, greater control of temperature
ber, can be used to describe the relative
differentials for chill roller systems allows for
superior regulation of cooling rates, which can While its importance in extrusion for other
present unique advantages for post-process treat- fields has long been recognized, recognition of
ment where roll designs integrate internal flow the influence of SME in melt extrusion is cur-
paths for liquid temperature control which help rently growing. It has been found that SME is a
determine the heat transfer capabilities of the good indication that there is the successful
system. processing of ASDs without degradation
(Ma et al. 2019; Haser et al. 2017; Evans et al.
dQ
ConvectiveHeat TransferEquation : 2019) as well as affecting the dissolution profiles
dt
of ASDs in some cases (Hanada et al. 2018). In
¼ hAðT T o Þ: ð9:3Þ general, there is a necessary combination of
energy from mechanical and thermal sources to
Numerous engineering factors determine the
convert extrudate from the crystalline physical
ability to process during melt extrusion, with
blend into an ASD. Recent uses of SME in the
process parameter modifications directly
formulation of poorly water-soluble drugs have
impacting the behavior of the system and formu-
been reviewed recently and include optimizing
lation. The development of a robust design space
ASD formulations as well as scaling up between
becomes imperative, and an understanding of
extruders (Thompson and Williams 2021).
processing ranges should be established at an
Equations (9.5) and (9.6) show some of the
early stage to justify the use of melt extrusion as
different parameters that affect a system’s SME
the production technology.
(Haser et al. 2017). Equation (9.5) describes the
total power applied by the machine, while
Importance of Specific Mechanical Energy Eq. (9.6) describes the power per unit processed
in Melt Extrusion (throughput rate). Several of the parameters are
Of further importance to the process’ design is the extruder-specific constants, such as the motor
energy input imparted by the melting, mixing, power rating (KW) and the maximum rpm of
and pumping mechanisms. Termed specific the machine. Screw speed (rpm running) and
mechanical energy (SME) input, this is defined feed rate are independent variables (i.e., they do
as the energy input provided by the motor of the not rely on nor are directly affected by other
system to the extrusion process and is shown variables), while %Torque is a function of the
mathematically in (9.4). SME plays an important screw speed and viscosity of the system as well
role in optimization and scale-up; many formula- as the maximum motor rating of the engine
tion properties are a function of a minimum spe- (Thompson and Williams 2021). Because SME
cific energy input to achieve the desired results. is dependent on each of these variables, it acts as a
SME can be calculated for a system by the com- composite that captures a more complete picture
bination of two Eqs. (9.5) and (9.6). of the mechanical energy input from the extruder
into the extrudate than each of the other variables
€
E
Specific Energy Input ¼ ð9:4Þ alone (Table 9.3). From these equations, it is
m_ apparent that SME increases with increasing
KWðappliedÞ ¼ KWðmotor ratingÞ screw speed and torque and decreases with
%Torque increasing feed rate. Less obviously, increasing
the screw speed does not always lead to a linear
rpm running
ð9:5Þ increase in SME because most polymers used in
rpm maximum
HME undergo shear thinning (i.e., decreased
Specific Mechanical Energy ðSMEÞ complex viscosity when exposed to high shear).
With the lowered viscosity of the system at higher
KWðappliedÞ
¼ , ð9:6Þ screw speeds, each individual rotation requires
Feed rate
less energy than at lower screw speeds. When
optimizing SME for formulation or other
Table 9.3 General impact of process parameters on the specific mechanical energy of melt-extruded products
Effect on SME
Effect of
Parameter Magnitude parameter Description
Screw speed Less than " Increasing screw speed generally increases SME despite a possible
linear decrease in torque from shear thinning and can cause lower screw fill
Torque Linear " Torque is dependent on viscosity, percentage of fill, and screw speed
Feed rate Linear # Inversely proportional in the SME equation. It also affects the degree of
fill, which can increase SME
Viscosity Varies " Dependent variable on temperature, screw speed, polymer choice, and
MW. Can screen for appropriate viscosity using rheology
Temperature Greater # Viscosity of polymers generally decreases exponentially with
than linear temperature
Screw design Varies " Increased kneading zone number or intensity generally increases SME,
intensity but it can increase shear thinning or temperature
Formulation Varies Varies Excipient and polymer choice as well as API loading can affect viscosity
and the Tm or Tg of the system
Adapted with permission Thompson and Williams (2021)
purposes such as scaling up, altering screw speed Under this definition, systems can include
and feed rate is generally sufficient to achieve formulations ranging from discrete amorphous
desired values. A developing body of work is drug distributed in a carrier phase to systems
being done to model SME of systems using where varying drug concentration gradients in
Ludovic® software based on the rheological the carrier manifest themselves as multiple glass
properties of mixtures to formulate ASDs transition temperatures. A specific subgroup of
(Bochmann et al. 2018). amorphous solid dispersions is amorphous solid
solution, where the drug substance is molecularly
and homogeneously dispersed within the carrier
Processing Regimes for Pharmaceutical
phase. In such systems, a single glass transition
Melt Extrusion
temperature is observed. However, one should
Pharmaceutical extrusion is conducted to com-
note that the definition provided may be subject
pound a drug substance into a larger carrier
to the limitations of the quantitative methods used
matrix to enhance product performance. In gen-
to assess the distribution. For example, DSC
eral, dispersions may be classified into three basic
methodologies have detection limitations of
types of material, depending on the distribution
approximately 30 nm (Newman et al. 2008).
and physical state of the active ingredient (Brown
This allows for heterogeneity in what is perceived
et al. 2014; Brough and Williams 2013; Van den
as an amorphous solid solution even when
Mooter 2012). Represented schematically in
distributions of less than 30 nm are present. Ana-
Fig. 9.7, these states are crystalline solid disper-
lytical technology development in the area of
sion, amorphous solid dispersion, and amorphous
dispersion characterization has received substan-
solid solution. Crystalline solid dispersions pres-
tial attention recently, and several other
ent multiple material phases, including a discrete
technologies have been applied to characteriza-
crystalline phase and a uniquely identifiable car-
tion. Techniques such as pair distribution analysis
rier phase. Using differential scanning calorime-
of X-ray diffraction patterns and spectroscopic
try (DSC), the polymer carrier phase will be
assessment continue to improve drug distribution
identified by a glass transition temperature,
resolution within the formulation (Newman et al.
while the melting endotherm of the crystalline
2008; Tumuluri et al. 2008). Based on the
drug substance will also be observed. Amorphous
limitations of detection of current analytical
solid dispersions are systems where one or more
technologies, the boundary between a solution
amorphous drug-containing phases are identified.
Crystalline Solid Dispersion Amorphous Solid Dispersion Amorphous Solid Solution
Fig. 9.7 Schematic diagram of the different states of a solid dispersion
with a single Tg and a dispersion with multiple Tg energy input to be dispersed and form a homoge-
can become blurry for nanoscale distributions that nous material. The dissolution of the drug can be
cannot be differentiated from solution. Because of described in terms of the Noyes-Whitney equa-
this, for the purposes of this chapter, the term tion given as (9.7) where D is the diffusion coef-
amorphous solid dispersion will include amor- ficient, A is the surface area of the drug substance,
phous solid solutions. For further discussion on Co is the saturation solubility of the drug sub-
detailed analytical methodologies for stance in the polymer, C is the concentration of
characterizing amorphous formulations, the drug substance in the bulk polymer melt, and h is
reader is referred to section “Solid-state charac- the diffusion boundary thickness. Raising
terization of melt-extruded amorphous temperatures above the glass transition tempera-
dispersions”. ture results in greater diffusivity, as described in
Most pharmaceutical systems currently pro- the Stokes-Einstein equation presented in (9.8)
duced using melt extrusion for bioavailability- where kB is Boltzmann’s constant, T is tempera-
enhancement applications are intended to create ture, η is viscosity, and r is the radius of the drug-
an amorphous solid dispersion. This system substance particle. Temperature increase also
provides the free energy benefits of an amorphous raises the equilibrium solubility of a drug in the
form to increase dissolution rates and solubility carrier (Co). This behavior results in greater dis-
while also drawing on the molecularly disperse solution rates (dM/dt) which can be further
nature of the system to maximize specific surface increased by providing higher shear levels to the
area at a molecular level. During the system, driving a decrease in boundary layer
manufacturing of solid dispersions, the material thickness (h).
may be processed in one of two distinct regimes Noyes Whitney Equation Describing Disso-
(Brown et al. 2014). The first system involves the lution Behavior of Solute in Solvent:
dissolution of the solid drug substance into the
liquid-like polymer. The second system involves
dM DA
the mixture of two miscible liquids of differing ¼ ðCo Cðt ÞÞ, ð9:7Þ
dt h
viscosities. The two systems can be further
divided based on the melting point of the drug Stokes Einstein equation for diffusivity :
substance, the extent of melting-point depression
kB T
observed in the presence of polymer, and the melt D¼ : ð9:8Þ
6πηr
viscosities of the drug substance and polymer
(Brown et al. 2014). Based on the Noyes-Whitney equation,
According to Brown et al., in the first regime, increasing the surface area of the drug substance
drug substance is paired with a highly viscous or and reducing the boundary thickness are critical
inviscid (less viscous) polymer. Drug substances to producing a homogenous system. To
in this class have a high melting point and require
accomplish this, this regime requires high sufficiently stable amorphous form cannot be
processing temperatures and longer residence generated. These systems may also be prepared
times. In order to prevent degradation and to provide controlled-release functionality to a
improve processing conditions, pretreatment dosage form. The top-down crystalline
steps can be done. Micronizing the drug sub- dispersions, particularly useful for controlled-
stance can help reduce necessary processing release application, are also generated in the sol-
conditions if it is thermally labile, and ubilization regime; however, the goal is to find
pre-blending drug with the polymer can also materials in which the drug has extremely limited
improve processing performance by maximizing solubility (Co ~ 0). Processing conditions will
initial drug-polymer contact. also be modified to minimize temperature to
The second regime is best described by liquid/ lower equilibrium solubility and diffusivity
liquid mixing due to the miscibility of the drug while also providing only sufficient shear to
and polymer as observed by the melting-point reduce particle size to the desired target product
depression of the drug substance in the presence property while not imparting excess energy to
of polymer or due to a low-melting-point drug substantially raise localized temperatures or sig-
substance. In the case of the high-melting-point nificantly reduce boundary layer thickness. By
drug substances in this regime, the drug melts careful identification of nonsolvent materials and
below its melting point due to the presence of optimizing processing conditions, it is possible to
polymer, which creates two liquids (polymer create crystalline solid dispersions using melt
and drug) of varying viscosities in the extruder. extrusion.
The melting-point depression allows more mod-
erate processing conditions in comparison to
systems in the first regime. The challenge with 9.2.3 Distributive and Dispersive
this regime is the extent of mixing required to Mixing in HME
achieve a well-dispersed system. Distributive
and dispersive mixing are both required to break The shear force and SME from the extruder mix
up the droplets of drug substance and fully dis- the extruded material in two ways distributive and
solve the drug substance to form a homogenous dispersive mixing. As described above, these two
system and are discussed more fully in the next processes increase the dissolution rate of a drug
section. This challenge is reduced with a less substance in the carrier polymer. This increase in
viscous polymer. dissolution is described by both the Noyes-
In this regime, the other case is the lower- Whitney (9.7) and the Stokes-Einstein (9.8)
melting-point drug substances that require less equations. With formulation of poorly water-
thermal energy but still require mixing. These soluble drugs with hydrophilic polymers, the
systems, both with viscous and inviscid mixing during HME is analogous to the mixing
polymers, result in liquid/liquid mixing. They of any hydrophobic and hydrophilic compounds.
are the least complex cases to process. Lower Emin and Schuchmann use computer
processing temperatures due to the miscibility modeling to describe the dispersive and distribu-
and plasticizing effect of the drug substance are tive mixing of a similar system of triglycerides in
required, and viscosity reduction of the polymer starch during HME processing (2013). In HME,
is not essential. dispersive mixing is the forced disintegration of
Manufacturing of a crystalline solid dispersion hydrophobic aggregates by high shear, which
is performed to reduce particle size for increased causes their extension into long filaments that
dissolution rate where the free energy benefit of are less physically stable and subsequently break
an amorphous form is not required. Such a case into smaller droplets. The hydrophilic substance’s
may arise for compounds that have sufficiently tendency to disintegrate depends on the relative
low lattice energy that dissolution is not hindered viscosities of each compound and the extent of
by the crystalline structure or in cases where a shear stress. In HME, the location of greatest
Fig. 9.8 Depiction of API distribution by kneading zones in HME. Kneading and mixing elements rapidly distribute
API, shown as blue dots, within the bulk media in extruder. From left to right, each figure represents 1 s passing at
500 RPM. (Reproduced with permission)
shear is generally in the overflight and can ideally achieve amorphicity while avoiding
intermeshing regions. While occurring largely in degradation by excess shear forces.
the same location of the barrel as dispersive
mixing, distributive mixing is a distinct phenom-
enon that describes the intermixing of compounds 9.3 Formulation Design
with minimal reduction of particle size as for Melt-Extruded Dispersions
depicted in Fig. 9.8. In HME, both the
intermixing of the physical blend distributive Melt-extruded solid dispersions can contain a
mixing and the reduction of particle size by dis- number of different materials, each serving a
persive mixing increase the rate and extent of specific and necessary role in the formulation.
dissolution in ASDs. Consisting of drug substance, stabilizing poly-
This effect was demonstrated in an experiment mer, plasticizer or surfactant, melt solubilizer,
examining the effect of the presence or absence of and glidant, determination of component need
kneading elements when using HME to process a and level is driven by early preformulation
mixture of nifedipine and hydroxypropyl methyl- assessment and tuned during the formulation
cellulose phthalate (HPMCP) (Nakamichi et al. optimization period based on manufacturability,
2002). When processed without a kneading ele- bioavailability, and stability of the dispersion
ment, the extrudate and physical mixture had system. Utilizing a rational design concept, a
similar dissolution profiles, and it was observed rapid prototyping approach for the development
that the extrudate neither melted nor became of melt-extruded solid dispersions, outlined in
amorphous. When processed under identical Fig. 9.9, can be used to identify lead
parameters except for the addition of a kneading compositions providing desired target properties.
element, the extrudate was found to be amor- Within this path, property mapping and molecu-
phous and had a significant improvement in dis- lar modeling concepts are utilized to justify for-
solution compared to the physical mixture. These mulation selection leading to prototype
observations highlight the importance of the dis- development. Prototype formulations can then
tributive and dispersive mixing, which primarily be rapidly screened using small-scale extrusion
occurs in the extruder’s kneading zones. equipment and evaluated for manufacturability,
Optimizing the screw design during formulation bioavailability enhancement, and stability. This
NO
Re–evaluate formulations
and processing
conditions to correct
Accelerated
Stability
Identify Acceptable
Prototype
Candidate Dispersion
Manufacturing YES
Molecule Characteristics
Continue to BA
Analytical evaluation & dosage
Accelerated Stability development
Characterization
Tg Open & Closed Condition:
Tm SEM
Solubility Parameter Assay/Impurites
Thermal Stability DSC
Modeling to establish loading XRD
and prototypes
Produce prototypes to Assay
assess manufacturability Impurities
Tg
XRD
Non-Sink Dissolution
Fig. 9.9 Pathway for prototype solid dispersion development
section provides a basic understanding of formu- enhancement. Once identified as a specific need
lation design for melt-extruded solid dispersions for the compound, a range of more preferred
while also describing the current performance solubility-enhancement options are available,
assessment strategies to support early formula- including salt formation, polymorph selection,
tion development. particle-size reduction, and pH modification.
Each of these options has been utilized more
extensively than amorphous dispersions to
improve oral bioavailability and provide greater
9.3.1 Preformulation Assessment
physical stability and production within conven-
to Support Hot-Melt Extrusion
tional pharmaceutical unit operations. Further
information on these processes is provided in
Solid dispersion formulation development begins
various chapters of this text. In cases where
with a detailed preformulation evaluation
these options do not provide sufficient solubility
designed to identify the need for an amorphous
improvements, amorphous dispersions become a
formulation and the viability of melt extrusion as
viable option.
a production platform. Assessment of molecular
Once the need has been identified, the next
properties, particularly around aqueous solubility
step for preformulation characterization is to
and solid-state properties, allows one to identify
assess the viability of producing a stable solid
the need for an amorphous form, with solid dis-
dispersion. Evaluation of solubility parameters,
persion formulation intervention required for
melting point, glass transition temperature, crys-
molecules exhibiting low aqueous solubility.
tallization tendency, and lipophilicity can be used
Generally, if the solubility is insufficient to
to assess viability quickly. Using these values, it
allow the target dose to dissolve in less than
is possible to project the likelihood of success for
250 ml, the compound may require solubility-
an amorphous formulation. Development of basic
enhancement technologies to achieve the desired
property maps, such as shown in Fig. 9.10, can
exposure (Amidon et al. 1995). In general, many
predict the likelihood of success for the
developmental compounds exhibit solubilities
formulations by quickly identifying the maxi-
severalfold below that, clearly delineating the
mum attainable drug loading and comparing this
need for an amorphous formulation. Compounds
to the target dose estimated from the amorphous
exhibiting slow intrinsic dissolution rates or site-
form (Friesen et al. 2008).
specific absorption may also require solubility
Fig. 9.10 Property 1.6

mapping of solid
dispersions to determine
maximum attainable 1.5
theoretical drug loading Tg Limited, Loading < 35%
1.4
Tm/Tg
1.3
Moderate Loading 35% - 50% Dissolution Limited Loading, < 35%
1.2
1.1
High Loading > 50%

1
1 2 3 4 5 6 7 8 9 10 11 12
log p
One quick assessment of the ability of the drug required processing temperatures (DiNunzio
substance to form a stable glass dispersion is by et al. 2008). Further utilization of the Gordon-
the Tm/Tg ratio (Debenedetti and Stillinger 2001; Taylor equation can be conducted to identify
Angell 1995, 2002). Values greater than 1.3 indi- glass transition temperature at varying plasticizer
cate that the material is a weak glass and will have levels to aid in processing and provide stability
a propensity to recrystallize on storage, while assessment of theoretical formulations based on
values less than 1.3 indicate a strong glass. the rule of 50. Originally developed by Zografi
Recently, other assessments have been proposed and co-workers using the Williams-Landel-Ferry
to classify the glass-forming ability of drugs. The equation and experimentally demonstrated using
crystallization tendency from an undercooled indomethacin, this model describes a solid disper-
melt is measured, and molecules are divided into sion in terms of molecular mobility as a function
three separate classes based upon crystallization of glass transition temperature (Yoshioka et al.
observed during heating/cooling/heating by DSC. 1995; Hancock et al. 1998). Theoretically, molec-
Class I molecules recrystallize upon cooling and ular mobility increases as a function of tempera-
can be divided into two groups, those that can be ture, with significant mobility changes observed
melt quenched to avoid crystallization upon when the temperature is increased above the glass
cooling and those that cannot. Class II and III transition point. Zografi’s team identified a dis-
molecules are less likely to crystallize. Class II persion Tg 50 C greater than the storage
molecules recrystallize upon heating, while Class conditions to sufficiently inhibit mobility so as
III molecules do not exhibit any crystallization to provide an acceptable product shelf life.
(Baird et al. 2010). This system can quickly While this rule has shown validity in many
assess the glass-forming ability and foreshadows cases, there are also numerous exceptions,
how crucial choosing a stabilizing polymer for showing incompatibility when Tg – Tstorage > 50 C
the system will be. and acceptable stability when Tg – Tstorage < 50 C
Polymers can be evaluated for their ability to (Vyazovkin and Dranca 2007). Due to the
stabilize the amorphous form of the drug. Extrap- inconsistencies, molecular mobility remains an
olation of proposed solid dispersion glass transi- active area of research (Dantuluri et al. 2011;
tion temperature using the Gordon-Taylor Bhardwaj and Suryanarayanan 2012). In general,
equation can be used to identify appropriate cases of incompatibility in systems having a Tg –
materials and drug loadings capable of facilitating Tstorage > 50 C are observed with systems having
stable product formation while also indicating a high drug loading, a compound prone to rapid
recrystallization, or a system with only partial development of stable or metastable solid

miscibility. For cases where stability is observed dispersions.
at differentials below the rule of 50, it is most Despite its widespread use, there are
often due to a specific intermolecular interaction limitations to the use of the Flory-Huggins
that results in an elevated activation energy for model. It ignores important factors that can affect
recrystallization. Specific examples of such miscibility, such as hydrogen bonding between
interactions include amine groups interacting drug and polymer. In addition, the interaction
with carboxylic acid groups of enteric polymers parameter (χ) is an average of all interactions
and hydrogen-bond donor groups interacting with and so does not give information about any spe-
the carbonyl acceptor group of the vinylpyr- cific interaction occurring (Anderson 2018). A
rolidone polymers. novel alternative method to DSC for determining
Solubility parameters provide a measure of miscibility of drug in polymer at various
cohesive energy density for a material and can temperatures and drug loads has been used
be used to assess the interaction potential between recently to develop high drug-load (w/w) ASDs.
different formulation additives (Greenhalgh et al. This method uses small-angle and wide-angle
1999). The contributions of dispersive, polar, and X-ray scattering with a temperature-controlled
hydrogen-bonding components of molecular vacuum chamber to achieve solubility
structure are shown in Eq. 9.10. Differential measurements (Tian et al. 2020). Finally, there
values between components of less than is a method to determine the chemical potential
7.0 MPa1/2 have been correlated with miscibility, while remaining at room temperature; using solu-
while values greater than 10.0 MPa1/2 indicated tion calorimetry to determine the heat of mixing,
immiscibility. Applying this simple technique the solubility of a drug in the polymer can be
allows for rapid screening of candidate dispersion calculated (Marsac et al. 2012).
polymers related to the drug-substance properties.
Further thermodynamic evaluation of miscibility
and dispersion stability can also be assessed using Solubility parameter equation : δ2T
adaptations of the Flory-Huggins model to gener- ¼ δ2d þ δ2p þ δ2h : ð9:10Þ
ate phase diagrams of the solid dispersion at vary-
ing temperatures and compositions (Marsac et al. Beyond formulation selection to achieve a sta-
2006; Zhao et al. 2011). Small-scale melting- ble and orally bioavailable formulation, the pro-
point depression trials can also be used to accu- cessability of a melt-extruded system is highly
rately determine API solubility in the molten dependent on the formulation characteristics.
polymer as a function of temperature, allowing Melt viscosity plays an instrumental role in deter-
for identification of required processing mining the motor load and pressure during manu-
temperatures and maximum theoretically attain- facture (Schilling et al. 2007). Ongoing research
able drug loading prior to prototype in this area can help develop accurate extrudable
manufacturing. In these studies, a physical mix- temperature ranges for each polymer. In one
ture of drug with small fractions of polymer is example, the viscoelastic properties of poly
prepared and analyzed by DSC to identify (vinylpyrrolidone), cellulosic, and polymetha-
melting-point depression. By plotting melting- crylates/polymethacrylic acid-based polymers
point depression, it is possible to determine the were evaluated, and the intersection point
Flory-Huggins interaction parameter (χ) and between the storage modulus and loss modulus
develop bimodal and spinodal curves delineating (elastic and viscous term, respectively) of the
the two-phase region from the homogeneous polymers was plotted against temperature, and
single-phase region. From such phase diagrams, the point of intersection, tan δ ¼1, noted (Gupta
as presented in Fig. 9.11, one can identify the et al. 2014; Meena et al. 2014). This point
required processing temperatures and estimate represents the transition of the material to a
the appropriate quench rates required for the more liquid-like state, indicating a processable
Fig. 9.11 Phase diagram 100

of solid dispersion Spinodal point
80
60
Temperature (°C)
40
20
–20
–40
0 0.2 0.4 0.6 0.8 1
Φdrug
temperature. This method has the advantage of preformulation by reducing the need for multiple
providing a more accurate temperature range than extrusion runs to optimize processing parameters.
glass transition alone. In the case of Soluplus®, The system’s viscosity is also critical for pro-
which has a minimum extrusion tempera- cesses operated in the solubilization regime where
ture > 10 C above its glass transition, the method lower melt viscosities increase diffusivity and
found tan δ ¼1 at 83 C or 13 C above its glass dissolution rate of the solid drug particles in the
transition. This method could be extended to pre- molten polymer. By incorporating plasticizers
dict appropriate ranges for drug-polymer systems and surfactants into the system, processing
(Gupta et al. 2015a, b). temperatures may be lowered to improve the
An emerging method for determining the system’s thermal stability and aid in the manufac-
miscibility of drug in polymer during ture of solid dispersions with high-melting-point
preformulation is to use rheology to determine compounds. This is because the complex viscos-
the “critical temperature,” or the temperature at ity of systems containing even small amounts of
which a specific system (w/w) of a drug and plasticizers (i.e., ~5%) can be reduced by an order
polymer becomes molecularly dispersed (Yang of magnitude compared to those without (Evans
et al. 2016). To determine the critical temperature, et al. 2019). Counterintuitively, because the suc-
Yang et al. performed a rheological temperature cessful formulation of ASDs by HME relies on
sweep of 50:50 nifedipine/VA64 that determined the transfer of a sufficient quantity of energy to
that crystalline drug was dissolving in VA64 melt the mixture and convert the drug amorphous,
below the Tm of nifedipine (173 C) at around it has been seen that the presence of plasticizers
160 C based on the plateauing of the damping such as water (Haser et al. 2017) or Tween
factor at that temperature. They followed up with 80 (Evans et al. 2019) can reduce the imparted
HME with temperatures chosen above and below SME and cause incomplete conversion to amor-
this critical temperature. They observed that the phous drug. This is in part due to the reduced
extrudate processed below 160 C exhibited a viscosity causing less viscous dissipation from
small endotherm by DSC, indicating that it was the lower required torque at equivalent extruder
not completely amorphous. In contrast, all parameters. In these cases, other variables were
samples processed at or above the critical process altered to overcome these observed phenomena.
were determined to be amorphous (Yang et al. The beneficial aspects of processing using
2016). This method for assessing the miscibility plasticizers are also counteracted by a limitation
of drug in polymer at certain temperatures has the to physical stability. Although incorporating
potential to save significant resources during these additives has been shown to compromise
Table 9.4 Troup classification system characterizing dispersion production risk, adapted from Shah, Sandhu et al.
Melting temp Melting-point Polymer
Class of API depression system Complexity Phase attributes
I High Negligible Viscous Mixing degradation Solid/viscous
liquid
II High Negligible Inviscid Mixing degradation Solid/inviscid
liquid
III High Significant Viscous Mixing Liquid/liquid
IV High Significant Inviscid Mixing Liquid/liquid
V Low N/A Viscous Mixing for extreme Liquid/liquid
viscosity ratios
VI Low N/A Inviscid Mixing for extreme Liquid/liquid
viscosity ratios
the physical stability of susceptible amorphous extrusion are in development and have recently
forms, resulting in recrystallization on storage, begun to reach the market. Affinisol™ HPMC
several commercial products contain a surfactant HME developed by the Dow Chemical Company
in their formulation (Rosenberg et al. 2012, 2013; was recently made available as an alternative to
Liepold et al. 2011). traditional hypromellose (HPMC) with more
The two processing regimes presented in sec- desirable properties. Cellulosic polymers are typ-
tion “Processing regimes for pharmaceutical melt ically challenging to extrude due to their high
extrusion”. have been further classified. The melt viscosity and degradation at temperatures
Troup classification system (TCS) is presented where melt viscosity is extrudable. Affinisol™
in Table 9.4. It divides the risk of dispersion has a lower glass transition temperature range,
production into six classes based on phase 117–128 C, than traditional HPMCs which
attributes determined from melting temperature, have a range of 160–210 C. This is due to the
extent of melting-point depression, polymer sys- different substitution architecture of methoxyl
tem rheology, and complexity (Shah et al. 2014). and hydroxypropyl groups. As a result, it also
Polymer systems are classified as viscous or not has a significantly lower melt viscosity.
viscous. Class I is the most difficult, with Class With these properties, melt extrusion can be
VI being the least complex. Many examples in the performed at a wider range of temperatures and
literature exist covering the extrusion of Classes without plasticizers while maintaining the ability
IV–VI systems (Verreck et al. 2003; Keen et al. to stabilize an API. The success of Affinisol™ has
2014; DiNunzio et al. 2010a; Chokshi et al. been recently published comparing it to conven-
2005); however, fewer have been described for tional extrusion polymers (Huang et al. 2015;
the more challenging Classes I–III (Hughey et al. Gupta et al. 2015a, b). Ashland®, Inc., has also
2010). been investigating another cellulosic extrusion
Several polymeric materials are currently grade polymer, HPMCAS. Due to the hydropho-
utilized to produce melt-extruded solid bic interactions between drug and polymer in
dispersions and include nonionic and ionic solution, HPMCAS is an excellent crystallization
polymers (LaFountaine et al. 2015). In most inhibitor; however current commercial grades
cases, these materials were designed for other have a glass transition of 120 C and require

pharmaceutical technologies and have been extrusion temperature above 170 C
®
applied to melt extrusion, including enteric (LaFountaine et al. 2015). Ashland ’s HPMCAS
polymers, which are commonly used for coatings, under development has a glass transition closer to
and binders, which are commonly used for 100 C (Tanno et al. 2004). By being designed
granulations and compression. While these explicitly for melt extrusion, these materials pro-
materials have shown sufficient applicability, vide benefits for processing and bioavailability
new materials specifically designed for melt enhancement. A summary of materials commonly
Table 9.5 Commonly used polymers for pharmaceutical melt extrusion

Polymer Tg ( C) Grades Notes
Hypromellose 170–180 Methocel® E5 Non-thermoplastic
API must plasticize
Excellent nucleation inhibition
Difficult to mill
Modified hypromellose 117–128 Affinisol® Designed for melt extrusion
15cp, 100cp, and Broad processing window
4M API plasticization not required
Excellent nucleation inhibition
Vinylpyrrolidone 168 Povidone® K30 API must plasticize
Plasdone® K29/32 Potential for H-bonding
Hygroscopic
Residual peroxides
Easily milled
Vinylpyrrolidone-vinyl 106 Kollidon® VA 64 Easily processed by melt extrusion
acetate Plasdone® S630 No API plasticization required
copolymer More hydrophobic than vinylpyrrolidone
Processed around 130 C
Polyethylene glycol, vinyl 70 Soluplus® Designed for melt-extruded dispersions
acetate, Easily process by melt extrusion
vinyl caprolactam graft Low Tg can limit stability
copolymer Not of compendial status
Stable up to 180 C
Polymethacrylates 130 Eudragit® Not easily extruded without plasticizer
L100–55 Degradation onset is 155 C
Eudragit® L100 Ionic polymer soluble above pH 5.5
Hypromellose acetate 120–135 AQOAT® – L Easily extruded without plasticizer
succinate AQOAT® – M Process temperatures >140 C
AQOAT® – H Ionic polymer soluble above pH 5.5 depending on
AquaSolve™ – L grade
AquaSolve™ – M Excellent concentration-enhancing polymer
AquaSolve™ – H Stable to 190 C depending on processing
conditions
Amino methacrylate 56 Eudragit® E PO Process temperature 100 C
copolymer Degradation onset is >200 C
Low Tg can limit stability
used for melt extrusion to support bioavailability the optimum dispersion while also conserving
enhancement is provided in Table 9.5. valuable drug substance.
Utilizing basic information gathered from
drug-substance characterization, it is possible to
identify potential solid dispersion materials 9.3.2 Assessing Formulation
appropriate for development. Understanding Performance During Early
polymer properties coupled with thermodynamic Development
modeling of dispersion properties as a function of
temperature and drug loading will provide a basis After identifying appropriate materials and a tar-
for selecting early prototype formulations. In this get prototype formulation, small-scale
manner, one can significantly reduce the time dispersions will need to be manufactured to
required in subsequent development to arrive at assess three critical performance areas:
manufacturability, bioavailability, and stability. mixers, which have blades that rotate rapidly
While generalizations of solid dispersion perfor- inside of a container to mimic HME with the
mance with respect to bioavailability and stability advantage of achieving greater control over resi-
can be inferred from the behavior of systems dence time (Liu et al. 2010).
prepared using other methodologies, some
variations of physical properties may exist. Cur-
Manufacturability
rently, there is no small-scale method capable of
After careful selection of a polymer carrier and/or
providing the manufacturing performance assess-
other additives, melt extrusion is performed and
ment seen during melt extrusion, although new
manufacturability assessed. Instrumentation used
research in this area continues (Lauer et al. 2013).
to control the melt extrusion process provides
With the existing methods, formulation
significant insight into the performance of the
compositions may be erroneously disregarded
formulation. Key measurements of torque, tem-
because of the way they are prepared during
perature, and pressure provide information on
screening. For instance, the use of ovens or
material flow behavior within the process section.
thermogravimetric analysis (TGA) to simulate
Torque values are an indicator of the amount of
extrusion leads to exaggerated thermal exposure
energy that is being imparted into the materials
compared to extrusion residence times. Extending
being processed. Formulations exhibiting
heating time can lead to polymer and/or drug
increased levels may have high melt viscosities
degradation (DiNunzio et al. 2010a). Screening
and require the addition of a plasticizer. In addi-
methods, like cyclical DSC, can also underesti-
tion to plasticizers, dissolved supercritical fluids
mate the miscibility of a system such as with
like CO2 have been used to reduce the tempera-
HPMC that cannot sufficiently mix upon heating.
ture and/or stress needed to form a homogenous
Attempts have been made to improve thermal
melt (Ashour et al. 2015; Sauceau et al. 2011;
methods by particle-size reduction and systemati-
Chiou et al. 1985). Injection of supercritical sCO2
cally varying heating rates (Sun et al. 2009, 2010;
into the extrusion process results in a
Mahieu et al. 2013). Rheometers have been used
plasticization effect for polymers with high
in place of DSC to induce shear and assess the
sCO2 solubility. The sCO2 can be removed fol-
system under stress conditions (Yang et al. 2011).
lowing extrusion so the resultant glass transition
While able to simulate stress, they are unable to
temperature is not depressed and molecular
provide distributive mixing experienced in
mobility is not increased. This provides additional
extruders. Due to the restrictions and false
manufacturing advantages since the resulting
predictions, accurate development requires the
extrudate is typically porous, further facilitating
use of small-scale extruders to process materials.
post-processing steps such as milling and improv-
The miniaturization of the process has seen sig-
ing compression (Sauceau et al. 2011).
nificant development to support pharmaceutical
In addition to high torque, elevated melt pres-
development. Currently, systems as small as
sure and temperature may also provide informa-
3 mm can provide accurate manufacturing perfor-
tion about formulation flow while also identifying
mance assessments, allowing for batch sizes of
if die bore cross-sectional area is insufficient for
less than 5 g of material. This provides an accu-
the flow rates used in manufacturing. Representa-
rate and representative platform for evaluating
tive information about manufacturing perfor-
formulation performance while conserving a lim-
mance can also begin to be characterized, which
ited early-stage drug supply. Additionally, minia-
will serve as a basis for later-stage scale-up eval-
turization and advancement of analytical
uation. Determination of specific energy input
methodologies allow for small batch sizes to pro-
and screw design required to achieve target prod-
vide useful information about solid dispersion
uct properties establishes the beginnings of an
stability and bioavailability (Zecevic and Wagner
operational design space. Similarly, the early
2013). Another method for examining
stage’s temperature profiles are translated to
preformulation performance is by use of intensive
larger geometrically similar equipment as devel- optical imaging of chiral crystals (SONICC), a
opment progresses. type of nonlinear optical microscopy. The tech-
nique works for chiral molecules, which is inher-
Solid-State Characterization ent in most pharmaceutical compounds. The
of Melt-Extruded Amorphous Dispersions selectivity for crystalline drug substance has
Assessment of solid dispersion amorphous nature allowed measurements with detection on the
is conducted using at-line evaluations and tradi- order of parts per billion (Wanapun et al. 2010,
tional analytical characterization methods. Dur- 2011; Kestur et al. 2012). Coupling SONICC
ing production, amorphous systems frequently with Raman spectroscopy or two-photon fluores-
form transparent glasses, similar to that shown cence may provide quicker identification and bet-
in Fig. 9.12, which allows for easy identification ter discriminating power (Toth et al. 2012;
of residual crystallinity and provides a first-line Schmitt et al. 2015).
assessment of product attributes. Polarized light Aside from residual crystallinity, amorphous-
microscopy is an excellent qualitative technique amorphous phase separation also needs to be
that can support at-line determination of the assessed. Amorphous phase separation can be
properties. Further confirmation of amorphous indicative of an unstable system. Detecting
nature is achieved by using powder X-ray diffrac- amorphous-amorphous phase separation can be
tion, DSC, and Fourier transform infrared spec- challenging due to inherently less sharp analytical
troscopy, where characteristic markers associated signatures of amorphous forms. Commonly,
with the crystalline form can be detected. Using phase distribution within the dispersion is
DSC, a crystalline dispersion will contain both a achieved with modulated differential scanning
glass transition temperature associated with the calorimetry (mDSC), where a single glass transi-
polymer phase and the melting endotherm of the tion temperature denotes homogeneity. This tech-
drug substance. This type of dispersion will also nique has two significant limitations. The first
be visually opaque at-line, allowing for clear limitation presents itself when the phase
identification of amorphous and crystalline separating materials have glass transitions that
forms. Similarly, the immiscibility of formulation are too close together and the method is unable
additives can result in opacity of the system that to resolve them. The second limitation arises from
will also manifest itself in multiple detectable falsely characterizing a dispersion as homoge-
phases by mDSC. In extreme cases of immisci- nous if the measurement itself homogenizes the
bility, visual observations of phase separation sample.
may occur at-line where phases of different vis- Within the last decade, many publications
cosity are observed to exit the die region. have been published identifying amorphous
While these techniques have general detection phase separation. Techniques being explored are
limits of ~2–5% crystalline material, method mathematically transformed X-ray data, vibra-
development and optimization can be used to tional spectroscopy, confocal Raman spectros-
refine these and other technologies to detect resid- copy, ssNMR, and AFM (Newman et al. 2008;
ual crystallinity levels to less than 0.1% w/w Rumondor et al. 2009). In one example, two
(Newman et al. 2008). Solid-state nuclear mag- samples both characterized by a single glass tran-
netic resonance (ssNMR) can be used to achieve sition and extruded at varied conditions showed
detection limits very low when the drug contains stability differences (Qian et al. 2010a). Phase
an atom of high natural abundance that is present separation of the original extrudate was detected
only in the drug (Paudel et al. 2014). The reader is by Raman spectroscopy. In the case of
referred to an in-depth review of the advances in transformed X-ray data, an amorphous X-ray pat-
ssNMR for pharmaceutical applications recently tern is coupled with pair distribution functions.
published by Li et al. (2020). New techniques are Newman et al. could detect phase separation of
being investigated, including second-harmonic trehalose-dextran and poly(vinylpyrrolidone) that
generation (SHG) or second-order nonlinear was not detected by DSC. In addition to its use in
Fig. 9.12 Image of an

amorphous extrudate
detecting trace crystalline materials, ssNMR has to reach moisture equilibrium with the environ-
recently been employed to detect amorphous ment. Since moisture is a potent plasticizer,
phase separation (Pham et al. 2010). Atomic allowing the system to reach moisture equilib-
force microscopy (AFM) provides high spatial rium with the environment presents a worst-case
resolution and an ability to exam the molecular scenario for storage. Further increases in temper-
surface of extrudates in a nondestructive manner ature and humidity can provide greater stress on
(Lauer et al. 2013). Pulsed-force AFM coupled the material but may not be representative of the
with FTIR demonstrated the ability to detect final storage conditions.
phase separation in felodipine/Eudragit® E PO Samples can then be evaluated for chemical
systems (Qi et al. 2011). Recently, AFM was and physical stability. Chemical stability of the
used to detect phase separation in copovidone/ dispersion can be performed using conventional
vitamin E TPGS solid dispersions prepared by analytical techniques such as high-performance
melt extrusion (Lamm et al. 2015). The authors liquid chromatography, with the specific method
concluded that lower screw speed resulted in highly dependent on the properties of the drug
phase separation, and 600 RPM was required to substance. Physical stability is assessed by a com-
form one phase. bination of qualitative and quantitative methods
aimed at identifying recrystallization. Commonly
Stability used techniques include those mentioned above
Further evaluation of the extrudate to predict perfor initial quality evaluation along with qualita-
formance is necessary. Stability evaluation of tive methods such as polarized light microscopy
melt-extruded solid dispersions is conducted sim- (PLM) and scanning electron microscopy (SEM)
ilarly to other types of solid dispersions by plac- to visually identify the presence of crystalline
ing materials under aggressive storage conditions materials (Bruce et al. 2007). Owing to increased
(elevated temperature and humidity) to induce entropic freedom at the surface, molecular mobil-
stress on the system. They are then evaluated ity at the interfacial regions is greater than that in
using a battery of qualitative and quantitative the bulk, facilitating recrystallization on the exte-
techniques. From a regulatory perspective, the rior of solid dispersion particles (Zhu et al. 2008;
most aggressive condition the material will see Qian et al. 2010b). Exposure to elevated moisture
during shelf life is storage at 40 C and 75% and temperature helps to plasticize the interface
relative humidity. For rapid stability assessment, and further increase molecular mobility, allowing
it is recommended to store dispersions at these SEM to be used for identification of recrystalliza-
conditions in an open manner to allow the system tion on storage.
Similarly, PLM can be used to identify process, polyvinyl acetate phthalate (PVAP)
crystallized material in the solid dispersion creates an acidic moiety, phthalate acid, that
particles. Both methods generally tend to be reacts with the primary hydroxyl groups present
more sensitive than quantitative techniques; how- on the API, dipyridamole (DPD), promoting deg-
ever, they cannot be used to identify the extent of radation of DPD. Process optimization only
recrystallization. For this, methods mentioned in improved the purity profile from 0.8% to 7.5%,
section “Solid-state characterization of melt- deeming a PVAP-DPD ASD by HME not feasi-
extruded amorphous dispersions” are utilized. ble (Monschke et al. 2020). In contrast, DPD
Storage placement of samples under different ASDs with HPMC E50 and copovidone have
moisture controls allows for decoupling of mois- been successfully manufactured within accept-
ture and temperature effects responsible for able degradation limits (Vora et al. 2015; Zecevic
recrystallization while also providing the ability et al. 2018). During formulation, a fundamental
to extrapolate performance. If moisture absorp- understanding of the APIs degradation pathway
tion is identified as a failure mechanism can facilitate polymer selection and expedite
associated with dispersion stability, it may be product development.
possible to limit recrystallization with moisture-
resistant dosage form development or product Bioavailability
packaging. Temperature-associated failures Amorphous solid dispersions prepared by melt
become more difficult to address; however, in extrusion are undertaken to enable drug delivery
rare cases, they could be mitigated through strin- through improved solubility and bioavailability.
gent and restrictive storage requirements for the During development, in vitro characterization of
drug product, such as maintaining the material performance is necessary to identify lead
under refrigeration. formulations, but conventional techniques are
often insufficient to describe an amorphous
Drug-Polymer Incompatibilities form’s dissolution behavior. Resulting from the
Successful ASD formation by HME is a balanc- amorphous nature, formulations exposed to an
ing act that includes many variables. The preven- aqueous environment will rapidly dissolve and
tion of degradation typically requires the selection drive a supersaturated concentration (Brouwers
of a suitable polymer that solubilizes an API at et al. 2009; Alonzo et al. 2010). Nucleation and
temperatures below its melting point, as high- growth will begin to occur, causing precipitation,
melting-point drugs tend to degrade at which drives the system to thermodynamic equi-
temperatures close to their melting point or at librium. As a result of this behavior, a number of
the onset of melting. Additionally, SME different colloidal species will be present in the
components (e.g., screw RPM, feed rate) and solution, many of which will exist below 200 nm,
screw configuration impart significant mechanical thus limiting the ability of conventional analytical
energy that can contribute to the degradation and technologies to identify species in the solution.
must be considered to prevent degradation, espe- Within the current dissolution theory of amor-
cially for shear-sensitive API. Furthermore, poly- phous formulations, free-drug concentration is
mer selection cannot be chosen solely based on its believed to be the primary contributor to oral
capacity to solubilize the API; functional groups bioavailability enhancement. Free drug is defined
present can negatively interact with the API and as individual drug molecules dissolved in solution
catalyze degradation. For example, selecting an without significant complexation to other multi-
incompatible polymer that promotes degradation, molecular aggregates. Other species present in the
regardless of its solubilizing capacity, creates for- solution include drug-polymer nanostructures,
mulation obstacles that may not be overcome, nano-aggregates, free polymer, dissolving solid
regardless of changes in the drug loading and dispersion, bile salt micelles, and larger
process parameters. Mocschke et al. illustrate precipitated aggregates. Each of these species
this point; they found that during the extrusion contributes to a pseudo-equilibrium with the free
drug in solution, during which time free drug is three compounds (Chen et al. 2016). More
absorbed while the system approaches thermody- recently, the interactions between surfactants
namic equilibrium. Formulating solid dispersions and their abilities to maintain supersaturation
requires the identification of materials capable of have been examined more extensively (Zhang
maximizing the free-drug concentration while et al. 2019; Deshpande et al. 2018; Chen et al.
also extending the duration of time at which ele- 2015).
vated concentrations are maintained. If supersatu- Once the solubility issue has been overcome,
ration occurs too rapidly, the particles may permeability is the next factor to address. While a
nucleate and recrystallize. Surfactants have been formulation can perform well during in vitro dis-
used in many of the commercial formulations solution, it may be limited by its permeability.
(Rosenberg et al. 2012, 2013; Liepold et al. Systems other than amorphous solid dispersions
2011, 2014). Surfactants can influence dissolu- such as cyclodextrin complexes or cosolvent
tion rate and ultimately bioavailability (Brown systems have shown to suffer from low perme-
et al. 2014). Materials capable of providing such ability due to the decreased fraction of free drug
an effect are termed concentration-enhancing in the gastrointestinal tract (Dahan et al. 2010;
excipients, the majority of which are ionic amphi- Miller et al. 2012a). Amorphous solid dispersions
philic macromolecules that are ionized at intesti- have an advantage over these systems (Miller
nal pH (Friesen et al. 2008). From a functional et al. 2012b). Miller et al. showed that amorphous
perspective, the large molecular weight provides solid dispersions enable increased apparent solu-
steric hindrance to recrystallization, while the bility without sacrificing intestinal membrane
ionized state prevents aggregation and growth of permeability.
particles, leading to prolonged durations of super- Due to the small-size domain of the different
saturation (Friesen et al. 2008; Brouwers et al. species in solution and the varying degrees they
2009; DiNunzio et al. 2010c). The material’s contribute to bioavailability, analytical
amphiphilic nature allows hydrophilic groups to technologies used for characterization must be
interact with water, while hydrophobic capable of separating free drug from the other
interactions maintain stabilization of the free species. Current free-drug characterization
drug in solution (DiNunzio et al. 2010c). technologies rely on passive diffusion through
One concern with utilizing surfactants is the an isolating membrane to assess concentration
potential for them to interact competitively with levels. Dialysis methods and adaptations of the
the carrier polymer and reduce the drug’s bio- parallel artificial membrane permeability assay
availability. A telling example of this was (PAMPA) have been successfully implemented.
shown when the in vitro dissolution of ASD These techniques utilize membrane cutoffs to
formulations containing the poorly water-soluble ensure that higher-molecular-weight material is
drug, posaconazole, and the anionic polymer, non-permeable into the receptor component,
HPMCAS, with or without sodium lauryl sulfate while low-molecular-weight drug molecules can
(SLS), a surfactant, was compared to their in vivo diffuse through the system. More accurate than
bioavailabilities (Chen et al. 2016). The formula- many of the centrifuge and filtration counterparts,
tion containing SLS showed increased AUC and permeability through the membrane is used to
Cmax during dissolution compared to the one determine the donor compartment’s free-drug
without. Because with poorly water-soluble concentration. While diffusion techniques pro-
drugs the rate-limiting step is dissolution, it was vide the most accurate assessment, frequently, it
hypothesized that bioavailability should be may be acceptable to use alternate techniques as
improved. Despite this, when the formulations surrogates by measuring total drug in solution,
were dosed in beagle dogs, the formulation with- which also includes many of the colloidal species.
out SLS achieved ~3 greater bioavailability Compared to conventional sink dissolution test-
than that with SLS, indicating the complicated ing, these methods can be more reliable predictors
chemical interactions occurring between the of in vivo performance.
Another option for screening bioavailability 9.4.1 Process Optimization

enhancement of melt-extruded dispersions is the
use of preclinical animal models, although phys- The initiation of process optimization is largely
iological differences compared to man may limit scale-independent and driven by the properties of
predictability. These issues can become even the system being developed. For systems showing
more challenging for BCS II/IV compounds, process sensitivity, such as heat-sensitive
where volumes, pH, transit times, and mem- materials, there will generally be a need for opti-
brane permeability may not accurately predict mization at an earlier stage than that required for a
formulation performance in humans. While a more robust product.
detailed discussion of animal models is outside Early-stage process optimization will be
this chapter’s scope, it is vital to highlight the conducted to address issues with drug product
potential issues of predicting the oral bioavail- attributes or improve production efficiencies.
ability of poorly soluble formulations with pre- During this time, key-controlled process-indepen-
clinical models. One example of this variability dent variables of barrel temperature, feed rate,
by species is membrane permeability. In relation screw speed, and screw design will be modulated
to human models, primate models tend to pro- to assess the impact to target product attributes.
vide more similar permeabilities than canines. Since these variables represent underlying depen-
As a result, oral bioavailability may be dent properties of the system which are key to the
overpredicted in dogs due to higher permeabil- functionality of the process, such as melt viscos-
ity. This behavior may become even more ity, fill, and shear rate, significant interactions
exaggerated in amorphous formulations of BCS may be observed requiring a knowledge-based
II and IV compounds. approach to development and the use of a statisti-
cal design of experiments to better determine the
operational space which is represented
schematically in Fig. 9.13. Factorial designs capa-
9.4 Process Optimization
ble of being expanded to response surface studies
and Scale-Up
can provide a sound methodology for identifying
primary control variables as well as optimizing
The development of melt-extruded formulations
their responses around the minimization of
requires the design of compositions that facilitate
impurities and maximization of feed rate. During
processing, provide appropriate stability, and
this stage, drug product attributes related to drug
enable therapy by yielding greater bioavailability.
and carrier materials will be evaluated. Changes
Formulation additives must provide functionality
to production properties have been demonstrated
to processing and solubility enhancement. For-
to have a substantial effect on excipient properties
mulation development will begin with a sound
as well. One also notes that residence time is
preformulation screening and subsequent proto-
coupled with feed rate, which will drive a general
type production to identify final compositions
reduction of chemical impurities as throughput
through a combination of in vitro and in vivo
increases. For amorphous dispersions produced
screening. Further optimization ultimately leads
in the solubilization regime, the reduction of resi-
to the selection of a final formulation that will
dence time may be brought to a critical point
support commercial manufacturing. This section
where residual crystalline material exceeds the
describes the current scale-up and process optimi-
allowable limits of the process, or the homogene-
zation strategies for melt extrusion processes
ity of the dispersion is not achieved. For systems
within the framework of quality by design
developed in the miscibility regime, they will
(QBD).
Fig. 9.13 Representative

diagram of design space for
hot-melt extrusion
Temperature (°C)
Residual
Crystallinity
Degradation
s) She
e( ar (γ
Tim )
primarily suffer from distribution limitations. The polymer may be dried during extrusion,
This illustrates the critical balance that must be adding devolatilization steps and subsequent bar-
achieved during optimization to achieve design rel zones or length to the process.
product rates and product properties. When pro- For a given extruder, the process can be scaled
duction rates cannot be maximized to the desired up by increasing feed rate and screw speed pro-
throughputs while yielding the required product portionally to maintain specific energy input.
attributes, scale-up will be triggered to support However, care must be taken to ensure peak
continued development. shear rates at higher screw speeds do not
adversely impact product attributes.
Similar to scaling for other unit operations,
9.4.2 Scale-Up of Extrusion geometric similarity between extruders should
Operations be maintained when scaling up to a larger unit.
It is recommended to adhere to the same design of
The need and rate of scale-up will largely be the extruder. To achieve geometric similarity,
driven by drug product demand to support clinical length-to-diameter ratio, screw diameter ratio,
trials, noting that extrusion’s continuous nature and overall screw design should be kept as con-
does not limit production by batch size but rather stant as possible.
by throughput limitation. While the authors have The classic scale-up strategies focus on the
provided several other references for detailed process-limiting factors, including volumetric,
melt extrusion optimization and scale-up power, and shear rate-based scale-up. Volumetric
procedures (Steiner 2003; Schenck 2010; Todd scale-up aims to maintain the same mean resi-
1995; Dreiblatt 2003), this section summarizes dence time, while power scale-up aims to main-
the key aspects of the development for pharma- tain the same specific mechanical energy, and
ceutical applications. shear rate scale-up seeks the same peak shear. It
Scale-up is typically encountered during phase is also important to understand the energy balance
III and product launch. Other than the equipment of a given process. With small-scale extruders,
itself, changes to how the material is fed and the the energy that is imparted into the compounded
number of unit operations during extrusion can be material is significantly lower than the calculated
altered. For instance, powder ingredients may be SME because large portion of SME is “lost”
fed separately without blending or after blending. through thermal conduction via the barrel. There
Melt Extrusion
Formulation Equipment Process

-Solubilization Capacity -Feeder Type -Feeder Rate
-Melting Point -Screw Configuration -Screw Speed
-Viscosity -Process Length -Barrel Temperature
-Moisture Content -Die -Vacuum
-Density & Physical State -Cooling Train -Cooling Rate
Fig. 9.14 Diagram of melt-extrusion-process properties
are a limited number of publications that address achieved by implementing a more rigorous QbD
scaling up of pharmaceutical melt extrusion pro- approach, implementing process analytical tech-
cess (Haser et al. 2018; Chen et al. 2016). nology (PAT).
Difficulties can arise for geometric scaling
when extruder geometries are dissimilar. In
these situations, one method for scaling-up is to 9.4.3 QbD and PAT for Melt Extrusion
model processing parameters on the larger Processes
extruder to equal the SME used in the smaller
one. Because SME depends on screw speed, Early development approaches provide useful
feed rate, torque, and extruder max power and fundamental strategies for addressing process
RPM, any specific SME can be achieved in vari- limitations; however, the complex interaction of
ous ways by altering some of these parameters dependent properties that govern the extrusion
while holding others constant (Haser et al. 2018). process requires rigorous approaches to fully
For more detailed information and additional optimize the manufacturing process to maximize
scale-up strategies, the reader is directed to the production value. Examination of the process
Brown et al. (2014). In support of early develop- shows that a number of factors contribute to the
ment, basic process optimization around key con- final product attributes, each of which requires
trollable independent process variables can be in-depth consideration and unique approaches to
used to deliver target product properties of the assess their impact. Represented schematically in
solid dispersion. Balancing throughput and Fig. 9.14, these properties can be classified into
energy input considerations can be used to pro- three general types (formulation, equipment, and
vide a knowledge-based approach to develop- process), which contribute to the behavior of
ment, while a DOE-based approach can yield key-dependent variables within the system. For
greater information about the interaction of pro- successful optimization, the ability to decouple
cess parameters. Furthermore, scale changes can and describe key process-dependent variables
be accommodated using a series of relatively from the process-independent controls to accu-
simple principles, provided geometric similarity rately describe the system’s behavior is neces-
is maintained. While these rules will help to work sary. The complexity of the task and the amount
within the design space, they do not ensure a fully of data generated during such an undertaking also
optimized production process, which can only be make the application of PAT a necessity to
support the optimization. The application of a design space, changes in parameters, if they are
in-line PAT technology allows for accurate real- within the defined space, do not require additional
time monitoring of chemical compositions, which regulatory approval as the product is not consid-
can be provided by multiple independent feed ered to have been changed. In HME, relevant
streams. Real-time oscillations associated with processing parameters within a design space
system perturbations can be accurately monitored could include extruder type, screw speed, screw
and modeled to ensure that the system provides design, and formulation composition, among
sufficient dampening to account for natural pro- others. For a more comprehensive review of
cess variations. Step change modifications to feed QBD and PAT applications to melt extrusion
rates of different components can be used to iden- process development, the reader is referred to
tify process response factors, while pulsed com- Schenck et al. (2011).
ponent inputs can determine residence time
distributions. Computational and mathematical
modeling of the system based on reaction kinetics 9.5 HME for Nonbinary Systems
models of CSTRs and PFRs can then be used to
interpolate performance within the studied design As experience with HME for pharmaceutical
space ranges. Continued PAT application in rou- purposes has grown, more complicated drug
tine production also assists in the identification delivery systems are being studied to overcome
and isolation of out-of-specification material, shortcomings of binary (i.e., drug substance and
which results from unanticipated system one carrier polymer with other excipients) ASD
perturbations. In the future, this may also lead to systems that include low drug loading and signif-
the real-time release of material from in-line icant hygroscopicity of common polymers. These
measurements during production. include multidrug co-amorphous systems which
In-line infrared and Raman spectroscopy have may or may not contain carrier polymers (Arnfast
been commonly used to analyze extrudate in real et al. 2017); co-amorphous systems containing
time (Islam et al. 2015); the use of UV-vis is a drug, a small molecule such as an amino acid,
promising technique that aims to reduce setup and carrier polymer (Lenz et al. 2017); co-crystal
time as well as expense during processing systems (2017); and ternary systems containing
(Schlindwein et al. 2018). Schlindwein et al. drug, carrier polymer, and a non-soluble
demonstrated the use of UV-vis to assist with mesoporous carrier (Hanada et al. 2018). Each
formulating an ASD of piroxicam and VA64 by system type holds significant promise for future
HME (2018). They utilized a sequential DOE developments within the HME field.
based on various HME parameters, including
screw speed, feed rate, and drug concentration,
and confirmed the amorphicity and lack of degra- 9.5.1 Co-amorphous
dation by UV-vis. Based on their findings,
optimized parameters were chosen, and reproduc- Recent work has shown that it is possible to
ibility was shown by repetition by multiple formulate stable amorphous solid dispersions
operators on different occasions. This example with minimal or no carrier polymer when
shows the utility of incorporating PAT into the formulated as co-amorphous systems. Arnfast
processing of ASDs by HME. et al. recently examined the feasibility of
Defining a design space by way of DOE has manufacturing by HME a co-amorphous system
important regulatory advantages as well, where a containing equal weight indomethacin, a nonste-
design space is “the multidimensional combina- roidal anti-inflammatory drug, as well as cimeti-
tion of input variables (e.g., material attributes) dine, a H2 receptor antagonist, with and without
and process parameters that have been 5% polyethylene oxide (PEO) (2017). Both drug
demonstrated to provide assurance of quality” substances are small molecules that had previ-
(ICH 2009). When manufacturing is done within ously been made co-amorphous by a solvent
evaporation method. The processability by HME indomethacin, arginine, and copovidone showed
was determined by rheological temperature significantly greater dissolution under nonsink
sweep examining the effect of temperature and conditions than neat indomethacin and an ASD
composition on complex viscosity. It was found containing only indomethacin and arginine.
that 5% w/w PEO acted as a plasticizer and While there are still challenges with using amino
increased the ASD’s hygroscopicity compared acids in HME, novel formulation techniques can
to the ASD without it. Surprisingly, despite the allow for successful extrusion of difficult to pro-
plasticization by the PEO, the ASD containing cess materials. An in-depth discussion of
PEO showed better physical stability compared co-amorphous systems can be found in
to that of the neat drug. This was demonstrated by Chap. 13, emerging technologies to increase the
DSC, where one Tg was observed in the bioavailability of poorly water-soluble drugs.
indomethacin-cimetidine-PEO ASD after storage
at 11% RH and 25 C after 1 week while the
indomethacin-cimetidine ASD had two Tgs. The 9.5.2 Pharmaceutical Co-crystals
presence of multiple Tgs suggests that phase sep-
aration has occurred. The authors hypothesize An in-depth discussion of the physiochemical
that PEO acts as a cosolvent that stabilizes the properties, formulation methods, and solubility
ASD by reducing the extent of saturation of both benefits of pharmaceutical co-crystals is found
drugs (Arnfast et al. 2017). These findings sug- in Chap. 3. Briefly, pharmaceutical co-crystals
gest that certain co-amorphous systems may contain two or more ionic compounds in stoichio-
benefit from adding small amounts of carrier metric ratios held together by non-covalent
polymer to aid in stabilization. interactions, without a proton transfer occurring.
The creation of co-amorphous systems The resulting solid-state co-crystal is a neutral,
containing an amino acid has been growing in crystalline material (Ross et al. 2016; Berry and
popularity. A major limitation of these systems Steed 2017). In general, APIs in the crystalline
has been that amino acids tend to degrade near state are more stable than the amorphous state.
their melting temperature of ~200 C, which Additionally, pharmaceutical co-crystal forma-
precludes the use of HME with many drugs. tion offers an alternative formulation strategy to
Lenz et al. overcame this issue and formulated improve the solubility and bioavailability of
several indomethacin-arginine ASDs with and poorly water-soluble drugs. Therefore, when
without polymer by HME through a novel liquid ASDs encounter stability challenges, co-crystal
injection of a heated aqueous arginine solution formation is a viable alternative.
into the extruder (2017). Initially, indomethacin, Adapting HME to form co-crystals overcomes
arginine, and polymer were blended and extruded many of the limitations of solvent evaporation
with various extruder parameters, but no and grinding, the two most common methods to
conditions provided sufficient energy to convert prepare co-crystals. In regard to solvent evapora-
the arginine amorphous. Because of the high tion, HME does not require similar solubilities in
water content during extrusion with the arginine a common solvent or slow drying steps to elimi-
solution, the extrudate was dried under a vacuum nate any residual organic solvents. Additionally,
before milling. This formulation’s success is solid-state techniques like grinding are limited
thought to be dependent on the basicity of argi- when attempting to scale up processing produc-
nine, as experiments with the neutral amino acid tion (Tan et al. 2020). In the first publication to
tryptophan did not cause amorphicity. There was successfully demonstrate HMEs applicability for
concern that because indomethacin degrades by co-crystal formation, Medina et al. attribute the
base-catalyzed hydrolysis, the presence of basic success to the twin-screw design providing
arginine would cause chemical stability issues, homogenous and efficient mixing in the closely
but the measurement of degradation products packed barrel (Medina et al. 2010). Dhumal et al.
was minimal. The ASDs consisting of reported the impact of HME processing
conditions on co-crystal formation, notably Soluplus®, 20:30:50 ibuprofen/Soluplus®/

finding processing at temperatures above the mesoporous silica mixtures were extrudable
eutectic point facilitates successful co-crystal for- Genina et al. (2018). The use of mesoporous
mation (Dhumal et al. 2010). carriers holds promise for overcoming issues of
Limitations of co-crystal systems arise from ASDs which can include poor dissolution or
their ability to form multidirectional hydrogen physical stability.
bonds, particularly at higher humidity, promoting Due to the general requirement of binary
hydrate formation in the presence of water ASDs to contain a large percentage (w/w) of
molecules. Liu et al. invented the matrix- carrier polymer to produce supersaturation and
associated co-crystallization (MAC) approach to prevent precipitation, it can be challenging to
circumvent this stability issue (2012). In this formulate a high drug-load dosage form such as
approach, a polymer matrix functions to catalyze tablets or capsules. This is particularly relevant
the interaction between co-former and API, when dosage forms require large doses of drug
improve processability and tableting properties, substance such that multiple, large tablets or
and promote stability (Liu et al. 2012; Boksa et al. capsules may be necessary to achieve a therapeu-
2014; Karimi-Jafari et al. 2019; Ross et al. 2019). tic effect. One of the challenges is that dosage
In contrast to ASD formation, the MAC approach forms containing significant polymer can form a
preferably utilizes inert, nonmiscible polymers, gel upon contact with water that protects the drug
typically at low polymer concentrations (e.g., from water, thereby reducing dissolution (Hanada
5% w/w) (Li et al. 2016). Selecting a polymer et al. 2020). In some situation, including a
that is not miscible with the API and co-former is disintegrant (e.g., croscarmellose sodium) in the
crucial due to the polymer’s potential to compete dosage form can reduce this tendency. Another
for the non-covalent interactions and prevent approach is to formulate the ASD in a ternary
complete co-crystal formation. This emerging mixture of drug, polymer, and a mesoporous
field warrants further investigation; a detailed non-polymer carrier such as mesoporous silica.
review of all relevant co-crystal works utilizing Hanada et al. recently formulated a high drug-
HME was recently published by Tan et al. load ASD tablet containing itraconazole, a poorly
water-soluble antifungal drug, Affinisol 4M
(AF4M), a high-viscosity hypromellose, and
9.5.3 Ternary Mixtures mesoporous silica (XDP) using HME Hanada
of Non-polymer Carriers et al. (2020). Ternary mixtures in which all three
were coextruded were compared to binary
Mesoporous Silica extrudate containing either itraconazole and
Researchers are increasingly using mesoporous XDP or itraconazole and AF4M, which then
carriers such as mesoporous silica in HME to was physically blended with the other excipient.
improve the release of poorly water-soluble The extrudates containing XDP had significantly
drugs. Until recently, most formulations of lower ASD granule size, which the authors
ASDs with mesoporous carriers used solvent hypothesized was due to the integration of ASD
methods, but its inclusion in HME and hot melt granules into the XDP, which reduced the
is becoming more common (Hanada et al. 2018; polymer’s tendency to gel. The authors note that
Lizoňová et al. 2018; Maniruzzaman et al. this size difference could not be rectified by
2015a, b). One of the main challenges with increased milling due to the hardness of the
incorporating mesoporous carriers into extrusion itraconazole-AF4M extrudate. Dissolution of the
processes is the high viscosity of the undissolved tablets in basic, nonsink conditions showed a ~ 4x
carriers and subsequent buildup in the extruder. increase in Cmax of the ternary ASD over binary
Genina et al. discovered that 50:50 ibuprofen/ itraconazole-AF4M extrudate with added XDP,
mesoporous silica mixtures could not be extruded while the binary mixture of itraconazole-XDP
but that with the addition of the polymer with added AF4M showed almost no
improvement over crystalline solubility. In addi- AbbVie

tion, in a pH transition dissolution experiment, AbbVie applied the Meltrex® technology to their
the ternary mixture showed slightly improved human immunodeficiency virus (HIV) drugs
dissolution over the two binary mixtures. The Norvir® and Kaletra®. HIV affects nearly 0.6%
discrepancy in performances can be partially of the world’s population. Many of these
explained by the significantly higher solubility therapies are classified as BCS IV compounds
of itraconazole in acidic media. This study due to extensive metabolism and poor aqueous
shows that incorporating novel excipients which solubility (Rosenberg et al. 2012). Intervention
alter the properties of ASDs can assist in addressing both of these limitations is necessary
overcoming problems common in more tradi- when addressing oral bioavailability of BCS IV
tional, binary systems. The use of mesoporous compounds.
materials is discussed at greater length in AbbVie’s single HIV compound, ritonavir,
Chap. 13, emerging technologies to increase the exhibits low aqueous solubility and a tendency
bioavailability of poorly water-soluble drugs. to crystallize. Two polymorphic forms were
identified; form II was the higher-energy form
and provided a solubility benefit but also
presented a potential for conversion to the
9.6 Case Studies
lower-energy form over time. Norvir® was origi-
nally developed as a liquid-filled capsule formu-
The majority of compounds in many drug discov-
lation containing ritonavir form II. The product
ery pipelines have limited solubility that require
received approval and was marketed for a short
formulation intervention to improve delivery.
period of time before form conversion of the
Over the last decade, many new products have
unstable drug substance was observed (Bauer
been developed utilizing solid dispersion technol-
et al. 2001). Due to the conversion, the product’s
ogy to improve bioavailability. This section
bioavailability was reduced, which required a
details several of the most recent developmental
product recall and ultimate reformulation.
and marketed products using hot-melt extrusion
Meltrex® technology was used to achieve a stable
to enable therapeutic performance.
amorphous solid dispersion in the reformulation.
It consists of ritonavir, Kollidon® VA 64, and a
surfactant (Rosenberg et al. 2012). The release
9.6.1 Oral Formulations mechanism is extended type, which presents the
of Amorphous Solid drug as a nanoparticle dispersion capable of
Dispersions maintaining supersaturation and increasing oral
bioavailability (Kanzer et al. 2010; Toth et al.
The last decade has seen the approval of multiple 2012; United States Pharmacopeia and National
commercial products made by extrusion. AbbVie Formulary (USP 38-NF 33) 2015). Through this
and Merck & Co. have both had approvals and technology, issues of form conversion have been
continue to develop poorly soluble compounds eliminated while also providing patients with a
with extrusion. AbbVie acquired Meltrex® extru- more robust dosage form that does not require
sion technology from Soliqs in 2001 and adapted refrigeration to maintain stability.
it to commercialize multiple amorphous solid
dispersions (Thayer 2010). Merck & Co. has Similar to Norvir®, first-generation Kaletra®
also utilized melt extrusion to address the poorly was formulated as a soft-gelatin capsule, which
water-soluble compounds in its pipeline, includ- required frequent multiple-unit administration
ing approval of an insomnia treatment and multi- and refrigerated storage temperature to maintain
ple compounds in development, most notably the stability. Kaletra®, a combination product
next-generation cholesteryl ester transfer protein containing lopinavir and ritonavir, addresses the
inhibitor (Martin 2016).
Process Section Configuration Novel Mixing Element Design

Power Liquid Gas
Connecting Zone Mixing Zone Pumping Zone
Low Temperature Moderate Temperature High Temperature

20°C 80°C 100°C
Fig. 9.15 Novel screw designs for production of Kaletra® solid dispersions
metabolic conversion of the target moiety with gelatin capsules to provide a finished dosage
the inclusion of a secondary component to inhibit form. Although able to provide improved oral
or saturate conversion pathways in vivo. This bioavailability, the product still exhibited variable
strategy, commonly referred to as a “booster,” absorption and substantial food effect, requiring
allowed for improved lopinavir levels with the patients to eat a meal before taking the capsules.
incorporation of ritonavir as a booster Instead, in Onmel®, itraconazole is extruded with
(Breitenbach 2006). Despite the clinical HPMC (40% itraconazole, 60% HPMC) to pro-
advantages, ritonavir exhibited heat sensitivity, duce a stable amorphous solid dispersion that
which challenged dispersion production by melt eliminated the food effect (Baert et al. 2003).
extrusion. Applying some unique screw elements Plasma levels in fasted healthy volunteers were
and designs, shown in Fig. 9.15, to minimize measured after administering Sporanox® capsules
localized temperature buildup and residence and the melt-extruded formulation, Onmel®. The
time, it was possible to reduce impurities (Kessler AUC of the extruded formulation was 2.3 times
et al. 2009). The final dosage form consists of 4% greater than the Sporanox® capsules (Baert et al.
ritonavir, 17% lopinavir, 78% Kollidon® VA 2003).
64, 7% Span 20 as a surfactant, and 1% colloidal In 2014, another therapy in the antiviral cate-
silica (Rosenberg et al. 2013). Through careful gory was approved by the FDA, made by
consideration of the solid-state properties, engi- AbbVie. The race to find a cure for hepatitis C
neering aspects of melt processing, and bioavail- virus (HCV) began in the 1990s and has recently
ability considerations, it was possible to design a been a much-debated topic following the
solid oral dosage form capable of providing the approval of multiple HCV drugs with >94%
benefits of soft-gelatin capsule systems with cure rate (PhRMA Foundation 2014). Like HIV,
improved stability aspects and size properties to complete and efficient removal of the virus
enhance patient compliance. Kaletra® represents requires multiple inhibitors. AbbVie has devel-
a successful melt-extruded dosage form to oped an oral tablet that contains all three drugs,
address solubility limitations. ombitasvir, paritaprevir, and ritonavir (Miller
AbbVie’s Meltrex® technology was also used et al. 2015). These drugs are all practically insol-
in the development of Onmel®, a second- uble in water, limiting their absorption and bio-
generation itraconazole dosage form. Marketed availability (Liepold et al. 2011, 2014; Rosenberg
as Sporanox®, the initial dispersion is prepared et al. 2012). Early in development, ombitasvir and
by an organic solvent drug layering process of paritaprevir were manufactured using spray dry-
itraconazole and hypromellose (HPMC) onto ing (European Medicines Agency (EMA) 2014).
sugar spheres (Gilis et al. 1997). The material is However, a solvent-free process was preferred, so
then surface coated with a protective layer of melt extrusion was employed. An in vivo study
polyethylene glycol, dried, and metered into was performed to compare the relative
bioavailabilities of ombitasvir and paritaprevir points and analyzed by HPLC. Vitamin E TPGS
prepared by both processing techniques, and a at 7% was capable of fully dispersing the API at
higher Cmax and AUC were observed from the 10% loading, while the same level of Tween
extruded formulation for both drugs (European 80 was capable of fully dispersing 12% API. A
Medicines Agency (EMA) 2014). Each active is temperature of 150 C was required to produce a
extruded individually to form three separate solid clear extrudate at 7% surfactant level. The lead
dispersions that are combined into a single dosage formulation (12% API, 80% Kollidon® VA
form (Liepold et al. 2011, 2014). This 64, 7% Tween 80, and 1% Aerosil) was tested
manufacturing technique allows for the in a canine model against a reference lipid formu-
challenges of each active to be addressed inde- lation. Plasma concentrations obtained from the
pendently. Additionally, melt extrusion solid dispersion surpassed the lipid formulation
eliminated the need for secondary drying and by 61% (AUC0-1 of 205.7 and 127.6 μgh/mL,
other post-processing steps required for solvent- respectively). In a Phase 1 study, a twofold higher
based techniques. The final formulation of each than predicted AUC of 58 μgh/mL was observed,
extrudate contains Kollidon® VA 64 and a sur- proving the ability of melt extrusion to enhance
factant, in addition to the amorphous form of the the bioavailability of venetoclax. Venetoclax is an
active (Liepold et al. 2011, 2014; Miller et al. excellent example of melt extrusion being used in
2015; Rosenberg et al. 2012). a new therapeutic area highlighting its versatility
Venclexta (venetoclax) is another marketed and continued need (Birtalan et al. 2012).
product by AbbVie formulated by extrusion
(LaFountaine et al. 2015; Birtalan et al. 2012). Merck
An oncology compound, venetoclax, was granted Merck is another large pharmaceutical company
accelerated approval in 2018 and regular approval that uses melt extrusion to improve bioavailabil-
in 2020 for the treatment of acute myeloid leuke- ity and commercialize poorly water-soluble
mia (Humphreys 2016; Center for Drug Evalua- compounds. Their first product of this type
tion and Research 2020). A major hallmark of approved was Noxafil®. Noxafil® consists of
cancer is the evasion of apoptosis (Hanahan and posaconazole, a weakly basic triazole antifungal
Weinberg 2000). One of the primary ways that that exhibits pH-dependent solubility. Without
cancer cells evade apoptosis is by upregulation of intervention, the drug dissolves in acid conditions
anti-apoptotic proteins of the Bcl-2 family. The but rapidly precipitates in the upper small intes-
small molecule in development has shown to tine. Using a conventional suspension formula-
successfully suppress Bcl-2 proteins, preventing tion (first-generation Noxafil®), the formulation
tumors from evading apoptosis. The aqueous sol- exhibits highly variable bioavailability with a
ubility and bioavailability of the small molecule is substantial food effect. In addition to the solubil-
very low; therefore, a solubility-enhancing tech- ity issues, posaconazole also exhibits significant
nology was necessary to improve oral bioavail- degradation at temperatures above 160 C, creat-
ability and achieve a therapeutically effective ing a challenge for extrusion (Fang et al. 2009).
dose (Birtalan et al. 2012). To overcome these However, a formulation was designed in which
challenges, venetoclax is extruded with the molten polymer (HPMCAS) solubilizes
Kollidon® VA 64 and surfactant (Birtalan et al. posaconazole below its melting point. This eutec-
2012). Variables considered during development tic behavior allows for the production of the
were surfactant level, drug load, extrusion tem- material at reduced temperatures to improve the
perature, and appearance of the extrudate. drug’s stability. The resulting formulation
Dispersibilities of the compound from exhibits enteric protection of the formulation,
formulations prepared with various surfactants thus preventing recrystallization in the small
and levels were screened by dispensing 25 mg intestine. Results showed limited dissolution
equivalent of granules in 75 mL of 0.1 N HCl at occurred in the acidic environment. Following
37 C. Samples were taken at predetermined time pH change, substantial dissolution and
Fig. 9.16 Posaconazole 14000 12,400

oral bioavailability 11,300
12000
10000 8,750
AUC (ng*hr/ml)
8000
6000
3,400
4000
2000
0
Fed Fasted Fed Fasted
Oral Suspension Melt Extruded Tablet
supersaturation occur, which provide more effi- differentiation in the tablets despite all but Batch
cient targeting of the small intestine. The melt- E being fully dissolved by 60 min in line with
extruded formulation was compared to the sus- their immediate-release properties. Harder tablets
pension at a 100-mg dose in a crossover study in had a slower dissolution. The tablets’ disintegra-
healthy volunteers examining fed-state and tion took from 5 to 30 min, increasing with
fasted-state plasma profiles. The results presented increasing tablet hardness, contributing to the
in Fig. 9.16 show substantial oral bioavailability differences in dissolution.
improvements with reduced food effect. The plasma concentration profiles over time
for the four test batches are presented in
Following the approval of Noxafil®, Fig. 9.18. The AUC01 was approximately the
Belsomra®, a treatment for patients suffering same for all batches; however, the observed geo-
from insomnia, was approved. Belsomra® was metric mean Cmax for the higher hardness batches
commercialized based on extrusion of the active (D and E) were ~ 13–14% less than the Phase III
suvorexant, a BCS II drug due to its poor water formulation (C) with the lowest hardness batch
solubility. To achieve supersaturation and (A) having the highest Cmax.
increase suvorexant bioavailability, Merck In addition, anacetrapib, a cholesteryl ester
scientists used a pH-independent polymer, transfer protein (CETP) inhibitor with poor aque-
Kollidon® VA 64, to reduce a delayed Tmax ous solubility, was being developed using extru-
(food effect) seen with pH-dependent polymers sion (Krishna et al. 2011). The competitive
(Harmon and Variankaval 2015). In addition to market in this therapeutic area has been reduced
food effects, during development, it became with Pfizer backing out after adverse events
apparent that tablet compression had a significant including heart failure and increased rates of
impact on suvorexant tablets’ disintegration time, angina were observed during a long-term ILLU
thus affecting dissolution and potentially absorp- MINATE trial. Hoffman-La Roche also
tion (Kesisoglou et al. 2015). During a Phase I terminated their CETP program when Phase III
study, tablets prepared at four different hardness trials failed to show clinically meaningful effi-
levels (0.8, 22.1, 32.0, and 38.1 kPa) were cacy. CETPs are challenging to deliver because
evaluated, and a correlation between dissolution of their lipophilicity and chemical instability of
and disintegration with pharmacokinetics was the amorphous form in the solid state. Merck
obtained (Multiple Level C IVIVC). Figure 9.17 showed the success of anacetrapib in a preclinical
shows the average dissolution curves of the tested study using rhesus monkeys. Developmental
batches used in the study. The data shows clear formulations were prepared using spray drying
Fig. 9.17 Average

dissolution curves for 100
batches of suvorexant
Suvorexant dissolved (%)

tablets made with varied
80
hardness (n ¼ 12) from
Kesisoglou et al. (2015)
60
40
Batch A: 1.8 kPa
Batch C: (reference) 22.1 kPa
20
Batch D: 32.0 kPa
Batch E: 38.1 kPa
0
0 10 20 30 40 50 60
Time (min)
Fig. 9.18 Average

Mean suvorexant plasma concentration (μM)
1.6 Batch A: 14.8 kPa

dissolution curves for
Batch C: (reference): 22.1 kPa
batches of suvorexant 1.4 Batch D: 32.0 kPa
tablets made at different Batch E: 38.1 kPa
hardness values (n ¼ 12) 1.2
from Kesisoglou et al.
(2015) 1.0
0.8
0.6
0.4
0.2
0.0
0 12 24 36 48 60 72
Time (h)
and melt extrusion (Geers et al. 2010). Melt- 2017 after insufficient benefit was shown to jus-
extruded formulations consisted of Kollidon® tify further development. While ultimately not
VA 64 and surfactants at levels up to 15%. Dos- submitted for approval, anacetrapib shows melt
age forms were evaluated in comparison to crys- extrusion’s continuing relevance to enhance drug
talline drug-substance formulations and spray- product performance (LaMattina 2015).
dried dispersion formulations made with
HPMCAS. Summary data from the preclinical
evaluations are provided in Table 9.6 (Geers 9.6.2 Non-oral Sustained-Release
et al. 2010). Melt-extruded solid dispersions Formulations
using Kollidon® VA 64 and vitamin E TPGS
provided superior bioavailability and reduced While most marketed HME products are oral
variability compared to the spray-dried dispersion dosage forms, Merck has developed several
product while increasing oral bioavailability by sustained-release hormonal contraceptive
almost threefold when compared to the crystalline products using HME. These include NuvaRing®
formulation. Merck discontinued the project in (etonogestrel, ethinyl estradiol), an intravaginal
Table 9.6 Preclinical oral bioavailability of anacetrapib solid dispersions

Metric Crystalline capsule SDD capsule HME tablet
Drug loading in dosage form 5.0% 8.8% 12.0%
Stabilizing polymer None HPMCAS-LF Copovidone
Surfactant 5% SDS None 6% vitamin E TPGS
Cmax (mM) 0.04 0.03 0.12 0.08 0.13 001
Tmax (h) 18.7 9.2 6.7 2.3 6.7 1.2
AUC0–24 (mM∙h) 0.67 0.032 1.99 1.10 2.29 0.21
insert that is replaced monthly, and Nexplanon® of this, the hormones in the core layer permeate
(etonogestrel), an intramuscular implant that lasts more rapidly into the skin layer than they can
for 3 years. Implanon®, a similar product, was permeate out of the skin layer into the
replaced by Nexplanon®, which has an improved surrounding tissue. In non-supersaturated
injector device. Non-oral contraceptive birth con- systems, the permeation from the inner layer to
trol has gained popularity in the past few decades the outer layer might be expected to decrease over
to improve patient compliance by reducing dos- time as the drug concentration in the reservoir
age intervals and allowing for lower systemic decreases. Because the etonogestrel in NuvaRing
exposure of hormones with similar birth control is supersaturated, this drop in permeation does not
efficacy (Algorta et al. 2017). While oral dosage occur during the intended lifetime of the medical
forms of ASDs are generally developed to device. In this way, drug release is controlled
increase drug bioavailability or reduce the food primarily by the outer layer’s permeability,
effect, these are not the most difficult problems to allowing for pseudo-zero-order release, as
overcome for hormonal contraception. For these depicted in Fig. 9.19 (Van Laarhoven et al. 2002).
two products, HME was used to prepare While supersaturation of etonogestrel is neces-
formulations that release hormones in pseudo- sary to achieve the controlled release of drug over
zero-order release or an almost continuous release the intended period of use, supersaturated
despite a decrease in drug concentration over solutions are inherently unstable and tend to pre-
time. Because Nexplanon is not an ASD, this cipitate out of solution over time. This tendency
section will focus on NuvaRing as an example. can be seen in Fig. 9.18, where there is a “burst”
NuvaRing achieves pseudo-zero-order release effect that causes an initial large release of drug
by a unique co-extrusion process; it is a dual- which drops off to the equilibrium rate. In this
layer, coaxial fiber shaped into a ring consisting system, the burst effect is also due to migration of
of ethylene-vinyl acetate (EVA) copolymer, etonogestrel from the core to the skin during
where the inner “core” layer is loaded with 0.7% storage. Ideally, the extent of supersaturation
w/w etonogestrel and 0.1% w/w ethinyl estradiol, should be high enough to ensure constant drug
while the outer “skin” layer is drug-free upon release while not being so high that it causes
manufacturing (Groenewegen 1999). The core physical instability over the product’s shelf life.
layer etonogestrel concentration is approximately The physical stability of two different
2 the solubility at 25 C of 0.35% w/w and formulations containing etonogestrel, one
serves as a drug reservoir. Notably, the vinyl containing 0.57% w/w, which is slightly lower
acetate content of the EVA inner layer is signifi- than the level in NuvaRing, and one containing
cantly higher at 28% compared to the outer layer, 5% w/w in EVA, was examined (Groenewegen
which has only 9% vinyl acetate. The increased 1999). The 5% w/w drug loading represented an
concentration of vinyl acetate has been shown to approximate 14 supersaturation at room tem-
increase the permeability of EVA for poorly perature, and after 6 months of storage at room
water-soluble drugs. This is because vinyl acetate temperature, it was found to have 450 mcg of
is amorphous and disrupts the semicrystalline etonogestrel crystallized on the surface as com-
ethylene backbone (Tallury et al. 2007). Because pared to the 0.57% w/w formulations, which had
Fig. 9.19 In vitro release 250

of various w/w etonogestrel 0.25 % etonogestrel
0.3 % etonogestrel
extrudates from EVA 0.57 % etonogestrel
Release etonogestrel (μg/day)

coaxial fibers at 37 C 200 0.69 % etonogestrel
immediately. The saturation 0.75 % etonogestrel
solubility of etonogestrel at
25 C is 0.35% w/w and all 150
fibers with a greater
concentration are
supersaturated. Note the 100
initial large drug release
rate for the supersaturated
fibers. (Adapted from Van 50
Laarhoven et al. 2002)
0
0 5 10 15 20 25
Time (days)
less than 10 mcg crystallized. In addition, visible 40% w/w drug loading by use of hot-melt extru-
crystals were seen after 1 day of storage at room sion (Breitenbach and Wolff 2015). The drug was
temperature when another 5% w/w etonogestrel shown to be stable during HME processing up to
extrudate was examined (Van Laarhoven et al. 160 C despite a tendency to undergo oxidative
2002). Understanding the limitations of supersat- degradation. Not only did HME processing allow
urated solutions’ physical stability, an optimized for a greater drug concentration, but HME
supersaturation factor was chosen for the processing also allowed for decreased cost and
marketed product. From this data, the FDA pack- continuous manufacturing of product compared
age insert for NuvaRing indicates that while it to solvent methods.
must be stored refrigerated at between 2 and Likely because of the significantly higher drug
8 C, it may be stored at room temperature for achieved using HME, the approved product was
up to 4 months after dispensing to the user. In this recalled in 2008 in the USA by the manufacturer
way, HME was used to create a controlled-release Schwarz Pharma because of the presence of
dosage form that achieves nearly zero-order rotigotine “snowflake” crystals (Sharma et al.
release over 21 days, offering an alternative solu- 2018). The crystals were not present during
tion for users’ hormonal contraception. manufacturing, indicating that storage conditions
Melt extrusion has also been utilized to formu- could be responsible for drug precipitation. It was
late extended-release transdermal patches. First determined that the precipitated rotigotine was a
approved by the FDA in 2007, Neupro® new polymorph with increased thermodynamic
(rotigotine) is a once-daily transdermal system stability and was not due to any contamination.
containing a non-ergoline dopamine agonist for Any drug precipitation in a transdermal system
the treatment of Parkinson’s disease and restless reduces the driving force for permeation and drug
leg syndrome. Previous methods for preparing absorption into the skin and could reduce drug
rotigotine transdermal systems have only bioavailability to sub-therapeutic levels. To com-
achieved 10–15% w/w drug loading using solvent bat this precipitation, cold-chain distribution was
methods, and there was a need for a greater drug implemented in Europe to allow patients to con-
loading (Muller and Peck 2005). The formulation tinue receiving this medication, and the shelf life
approved in 2007 was a silicone-based matrix was decreased from 24 to 6 months (Chaudhuri
transdermal system that contained rotigotine, sili- 2008). Neupro® was eventually reintroduced to
cone, as well as polyvinylpyrrolidone as a solu- the market in 2012 with a new formulation that
bility enhancer, and it achieved a maximum of did not have stability issues. This case
demonstrates the importance of understanding the Equipment and Reagents

long-term physical stability of drug systems, par-
ticularly those with high drug loads. • Acetaminophen
• Polyethylene oxide (MW ~ 100,000)
• Oscillatory rheometer (Rheometric
Mechanical Spectrometer RMS-800)
9.7 Summary
• Capillary rheometer
• Differential scanning calorimetry (DSC
Over the last quarter-century, a tremendous tran-
Q-100)
sition has been observed in the pharmaceutical
Method
industry resulting from several factors: techno-
logical advancements yielding more compounds
• Measure dynamic viscosity of a series of
with lower solubility, a need for innovative and
acetaminophen drug loadings (0%, 10%,
integrated drug products to provide greater intel-
20%, 30%, 40%, and 50%) at fixed
lectual property protection, and a global drive to
temperatures (80, 100, 120, and 140 C).
increase the usage of cost-effective
• Measure steady viscosity of a series of acet-
manufacturing platforms. Fortunately, melt extru-
aminophen drug loadings (0%, 10%, 20%,
sion has emerged as an innovative solution within
30%, 40%, 5%, and 0%) at fixed
the industry and is supplemented by over a cen-
temperatures (80, 100, 120, and 140 C).
tury of technological achievements in other fields.
Drug loadings of 40% and 50% only
While perceived by many to be a recent develop-
evaluated at 120 and 140 C.
ment for pharmaceuticals, the first marketed melt-
• Measure thermal behavior of formulations
extruded drug product using melt extrusion
using a heat-cool cycle from 0 to 100 C at
reached the market in 1981, with subsequent
10 C/min.
products developed periodically over the next
Results
30 years. As summarized in Table 9.7, the num-
ber of publicly disclosed programs has increased
• Viscosity curves as a function of drug con-
substantially as a result of the different industry
centration showed a “V”-shaped inflection
pressures, and undoubtedly many more programs
that indicated the solubility of drug in mol-
are currently undisclosed within development.
ten polymer.
One also notes that these programs span a range
• Increasing stress rates showed a reduction
of applications, which includes the use of melt
in method sensitivity which indicated that
extrusion to replace conventional technologies to
testing for melt solubility should be
advanced drug delivery systems. Regulatory
conducted at low shear.
initiatives and cost constraints will also utilize
• High shear conditions were more represen-
the continuous nature of the process to help
tative of extrusion conditions and should be
shape drug delivery in the twenty-first century.
studied to evaluate formulation perfor-
mance under extrusion conditions.
Method Capsule 1
• Small-scale viscosity measurements can be
Characterizing Drug Solubility Using Quanti- an effective method for evaluating drug
tative Preformulation Techniques solubility in the molten polymer to con-
serve drug substance in early development.
Based on the method reported by Suwardie
et al. (2011). Method Capsule 2
Objective
Dissolution Behavior of Solid Dispersions
• To determine the melt solubility of acet-
Based on the method reported by Alonzo
aminophen in polyethylene oxide using
et al. (2010).
rheological properties of the materials.
Table 9.7 Currently marketed and developed drug products using hot-melt extrusion
Product Indication HME purpose Company Approval stage
Lacrisert Dry eye Shaped system Merck Marketed
(HPC rod) syndrome
Zoladex Prostate cancer Shaped system AstraZeneca Marketed
(goserelin acetate implant)
Implanon Contraceptive Shaped system Merck Withdrawn (replaced
(etonogestrel) with Nexplanon)
Nexplanon Contraceptive Shaped system Merck Marketed
(etonogestrel)
NuvaRing® Contraceptive Shaped system Merck Marketed
(etonogestrel, ethinyl
estradiol)
Ozurdex® Macular edema Shaped system Allergan Marketed
(dexamethasone)
Palladone™ Pain Controlled release Purdue pharma Withdrawn (dose
(hydromorphone HCl) dumping)
Nucynta Pain Controlled release Depomed Inc. Marketed
(tapentadol)
Opana ER Pain Controlled release Endo Marketed
(oxymorphone HCl) pharmaceuticals
Eucreas® Diabetes Melt granulation Novartis Marketed
(vildagliptin, metformin HCl)
Zithromax® Antibiotic Melt congeal Pfizer Marketed
(azithromycin) (taste masking)
Gris-PEG Antifungal Crystalline Pedinol Marketed
(griseofulvin) dispersion Pharmacal
Rezulin Diabetes Amorphous Wyeth Withdrawn (API
(troglitazone) dispersion toxicity)
Norvir® Antiviral (HIV) Amorphous AbbVie Marketed
(ritonavir) dispersion
Kaletra® Antiviral (HIV) Amorphous AbbVie Marketed
(ritonavir/lopinavir) dispersion
Viekira Pak Antiviral Amorphous AbbVie Marketed
(ombitasvir, paritaprevir, (HCV) dispersion
ritonavir, dasabuvir)
Mavyret Antiviral Amorphous AbbVie Marketed
(glecaprevir, pibrentasvir) (HCV) dispersion
Belsomra Insomnia Amorphous Merck Marketed
(suvorexant) treatment dispersion
Noxafil® Antifungal Amorphous Merck Marketed
(posaconazole) dispersion
Onmel® Antifungal Amorphous Merz North Marketed
(itraconazole) dispersion America, Inc.
Anacetrapib Cardiovascular Amorphous Merck Withdrawn before
disease dispersion approval
Venclexta Oncology Amorphous AbbVie Marketed
(venetoclax) dispersion
Braftovi Oncology Amorphous Pfizer Marketed
(encorafenib) dispersion
Objective Method Capsule 3

Evaluation of Processability of Modified
• To study the supersaturated dissolution
HPMC, Affinisol™ for Extrusion
behavior of felodipine and indomethacin
solid dispersions prepared by thermal
Based on the method reported by Gupta et al.
processing.
(2015a, b).
Objective
• Felodipine
• To assess the extrudable window for
• Indomethacin
modified HPMC (Affinisol™) as compared
• Hypromellose
to traditional HPMC and copovidone.
• Hypromellose acetate succinate
• Polyvinylpyrrolidone K29/32
• Polarized light microscopy
• Affinisol™ HPMC HME (15 LV, 100LV,
• Raman spectroscopy
and 4 M)
• UV spectroscopy
• Methocel™ K100LV
Method
• Copovidone (Kollidon® VA 64)
• Differential scanning calorimeter
• Solid dispersions were prepared by melt
• Thermal gravimetric analyzer
mixing at temperatures 10 C above the
• Discovery hybrid rheometer-2 (DHR-2)
drug-substance melting temperature and
Method
quench cooling. Material was then cryo-
genically milled to less than 300 μm.
• Perform thermal gravimetric analysis to
• Nonsink dissolution was performed pION
identify polymer degradation window. Per-
m-Diss Profiler at pH 2 for indomethacin.
form modulated differential scanning calo-
• Polarized light microscopy, powder X-ray
rimetry to identify polymer Tg.
diffraction, and Raman spectroscopy were
• Evaluate viscoelastic properties of
used to investigate recrystallization behav-
polymers with rheometer at various
ior of solid dispersions in the aqueous
temperatures and shear rates.
environment.
Results
Results
• No polymers experienced significant degra-
• Amorphous formulations were prepared by
dation until 220 C.
melt mixing and quenching.
• The Tg of the Affinisol™ polymer (103 C)
• Analysis of aqueous slurries showed rapid
with comparable MW to Methocel E5 was
crystallization of amorphous drug sub-
approximately 70 C lower than the Tg of
stance which was inhibited by the presence
Methocel E5 (170 C).
of polymeric additives to the formulation.
• Kollidon® VA 64 had higher viscosity
• Nonsink dissolution illustrated the ability to
below 150 C than Affinisol™ polymers,
supersaturate the aqueous environment.
but lower viscosity >150 C.
• Polarized light microscopy showed how
• Affinisol™ polymers showed broader
crystal growth occurred in supersaturated
range of processing temperatures for extru-
solutions and could be inhibited by the
sion due to significant drop in viscosity
presence of polymeric additives.
with shear.
Method Capsule 4 • Dissolution of both extrudates was compa-

rable after tableting
Use of Pressurized CO2 as Plasticizer and
• All extrudates remained amorphous
Foaming Agent for Preparation of Amor-
throughout stability study
phous Solid Dispersion by Melt Extrusion
Method Capsule 5
Based on the method reported by Ashour
et al. (2015). Melt-Extruded Solid Dispersions for
Objective Improved Oral Bioavailability
• To use pressurized CO2 to reduce extrusion Based on the method reported by Miller et al.
temperature and improve post-processing Miller et al. (2008a).
properties of extrudate while maintaining Objective
performance.
Equipment and Reagents • To investigate the use of high-molecular-
weight stabilizer to improve oral bioavail-
• Ketoprofen ability of itraconazole.
• Hydroxypropyl cellulose (HPC) Equipment and Reagents
• Differential scanning calorimetry
• Thermo-Prism 16-mm hot-melt extruder • Itraconazole
• Comminuting mill • Eudragit® L100–55
• Manual tablet press • Triethyl citrate
Method • Carbomer 974P
• Haake Minilab II
• Extrude ketoprofen/HPC blends with and Method
without pressurized CO2
• Confirm ketoprofen/HPC extrudates are • Formulations containing 33% itraconazole,
amorphous by DSC Eudragit® L100–55 plasticized with 20%w/
• Mill granules and compress to tablets w triethyl citrate, and varying levels of
• Evaluate friability and hardness of tablets carbomer 974P were prepared by hot-melt
prepared with both extrudates extrusion at 130 C.
• Perform dissolution using USP type II • Extrudate milled and screened to a particle
apparatus with 1000 mL 0.05 M phosphate size of less than 250 μm.
buffer pH 7.4, and analyze with HPLC • Nonsink dissolution testing was performed
• Assess crystallinity of extrudates with and using a two-stage acid-phosphate buffer
without CO2 injection by DSC after pH 6.8 using USP apparatus II.
3-month storage at 25 C/60%RH • Bioavailability enhancement of
Results formulations was evaluated using male
Sprague-Dawley rats at a dose of 30 mg/kg.
• Extrudates without CO2 injection were Results
denser and characteristically difficult to
mill, while extrudates with CO2 injections • Melt-extruded material was successfully
were porous and readily milled prepared using the Haake Minilab II, yield-
• Injection of CO2 reduced extrusion ing amorphous formulations that exhibited
processing temperature by ~30 C multiple Tg’s associated with the solid dis-
• Foamy extrudates had better compressibil- persion and the carbomer phase.
ity and binding properties • Nonsink dissolution showed that carbomer
reduced the rate of dissolution and also
extended the duration of supersaturation, Baird JA, Van Eerdenbrugh B, Taylor LS. A classification
with the most effective level being system to assess the crystallization tendency of organic
molecules from undercooled melts. J Pharm Sci.
observed at 20% w/w. 2010;99(9):3787–806.
• Oral bioavailability was improved with the Bauer J, et al. Ritonavir: an extraordinary example of
20% carbomer formulation and also conformation polymorphism. Pharm Res. 2001;18(6):
exhibited a significant reduction of 859–66.
Berry DJ, Steed JW. Pharmaceutical cocrystals, salts and
variability. multicomponent systems; intermolecular interactions
and property based design. Adv Drug Deliv Rev.
2017;117:3–24.
Bhardwaj SP, Suryanarayanan R. Molecular mobility as an
References effective predictor of the physical stability of amor-
phous trehalose. Mol Pharm. 2012;9(11):3209–17.
Agrawal AM, Dudhedia MS, Zimny E. Hot melt extru- Birtalan E, Hoelig P, Lindley DJ, et al. Melt-extruded solid
sion: development of an amorphous solid dispersion dispersions containing an apoptosis-inducing agent.
for an insoluble drug from mini-scale to clinical scale. US Patent 2012/0108590; 2012.
AAPS PharmSciTech. 2015;17(1):133–47. Bochmann ES, Gryczke A, Wagner KG. Validation of
Algorta J, Diaz M, de Benito R, Lefebvre M, Sicard E, model-based melt viscosity in hot-melt extrusion
Furtado M, Regidor PA, Ronchi C. Pharmacokinetic numerical simulation. Pharmaceutics. 2018;10(3):
bioequivalence, safety and acceptability of Ornibel®, a 9–12.
new polymer composition contraceptive vaginal ring Boksa K, Otte A, Pinal R. Matrix-assisted cocrystallization
(etonogestrel/ethinylestradiol 11.00/3.474 mg) com- (MAC) simultaneous production and formulation of
pared with Nuvaring® (etonogestrel/ethinylestradiol pharmaceutical cocrystals by hot-melt extrusion. J
11.7/2.7 mg). Eur J Contracept Reprod Health Care. Pharm Sci. 2014;103(9):2904–10.
2017 Nov 2;22(6):429–38. Breitenbach J. Melt extrusion can bring new benefits to
Almeida A, et al. Ethylene vinyl acetate as matrix for oral HIV therapy. Am J Drug Deliv. 2006 Jun;4(2):61–4.
sustained release dosage forms produced via hot-melt Breitenbach J, Mägerlin M. Melt-extruded solid
extrusion. Eur J Pharm Biopharm. 2011;77(2): dispersions. In: Pharmaceutical extrusion technology.
297–305. New York: Informa Healthcare; 2003.
Alonzo DE, et al. Understanding the behavior of amor- Breitenbach A, Wolff HM, inventors; UCB Pharma
phous pharmaceutical systems during dissolution. GmbH, assignee. Hot melt TTS for administering
Pharm Res. 2010;27(4):608–18. rotigotine. United States patent US 9,186,335; 2015
Alshahrani SM, Morott JT, Alshetaili AS, Tiwari RV, Nov 17.
Majumdar S, Repka MA. Influence of degassing on Brough C, Williams R. Amorphous solid dispersions and
hot-melt extrusion process. Eur J Pharm Sci. 2015 Dec nano-crystal technologies for poorly water-soluble
1;80:43–52. drug delivery. Int J Pharm. 2013;453(1):157–66.
Amidon GL, et al. A theoretical basis for a Brouwers J, Brewster ME, Augustijns P. Supersaturating
biopharmaceutic drug classification: the correlation of drug delivery systems: the answer to solubility-limited
in vitro drug product dissolution and in vivo bioavail- oral bioavailability? J Pharm Sci. 2009;98(8):2549–72.
ability. Pharm Res. 1995;12(3):413–20. Brown C, DiNunzio J, Eglesia M, et al. HME for solid
Anderson BD. Predicting solubility/miscibility in amor- dispersions: scale-up and late-stage development in:
phous dispersions: it is time to move beyond regular amorphous solid dispersions: theory and practice.
solution theories. J Pharm Sci. 2018;107(1):24–33. New York: Springer; 2014. p. 231–60.
Angell CA. Formation of glass from liquids and Bruce C, et al. Crystal growth formation in melt
biopolymers. Science. 1995;267(5206):1924–35. extrudates. Int J Pharm. 2007;341(1–2):162–72.
Angell CA. Liquid fragility and the glass transition in Center for Drug Evaluation and Research. FDA grants
water and aqueous solutions. Chem Rev. 2002;102 regular approval to venetoclax in combination for
(8):2627–50. untreated [Internet]. U.S. Food and Drug Administra-
Arnfast L, Kamruzzaman M, Löbmann K, Aho J, tion. FDA; 2020 [cited 2021 Jan 15]. Available from:
Baldursdottir S, Rades T, Rantanen J. Melt extrusion https://www.fda.gov/drugs/drug-approvals-and-
of high-dose co-amorphous drug-drug combinations. databases/fda-grants-regular-approval-venetoclax-com
Pharm Res. 2017 Dec 1;34(12):2689–97. bination-untreated-acute-myeloid-leukemia
Ashour A, et al. Influence of pressurized carbon dioxide on Chaudhuri KR. Crystallisation within transdermal
ketoprofen-incorporated hot-melt extruded low molec- rotigotine patch: is there cause for concern? Expert
ular weight hydroxypropylcellulose. Drug Dev Ind Opin Drug Deliv. 2008 Nov 1;5(11):1169–71.
Pharm. 2015;42(1):1–8. Chen J, Ormes JD, Higgins JD, Taylor LS. Impact of
Baert LEC, Verreck G, Thoné D. Antifungal compositions surfactants on the crystallization of aqueous
with improved bioavailability. US Patent suspensions of celecoxib amorphous solid dispersion
6,509,038; 2003.
spray dried particles. Mol Pharm. 2015 Feb 2;12(2): melt extrusion and Kinetisol® dispersing. Eur J
533–41. Pharm Biopharm. 2010b;74(2):340–51.
Chen B, Zhu L, Zhang F, Qiu Y. Chapter 31: Process DiNunzio JC, et al. Production of advanced solid
development and scale-up: twin-screw extrusion. In: dispersions for enhanced bioavailability of
Qiu Y, Chen Y, Zhang G, Yu L, Mantri RV, editors. itraconazole using KinetiSol dispersing. Drug Dev
Developing solid oral dosage forms. 2nd Ind Pharm. 2010c;36(9):1064–78.
ed. Amsterdam: Elsevier; 2016. p. 48. 1176p. (Book Doetsch W. Material handling and feeder technology. In:
chapter). Ghebre-Sellassie I, Martin C, editors. Pharmaceutical
Chiou JS, Barlow JW, Paul DR. Plasticization of glassy extrusion technology. New York: Marcel Dekker;
polymers by CO2. J Appl Polym Sci. 1985;30(6): 2003. p. 111–34.
2633–42. Dover HW. Apparatus for insulating or covering strands
Chokshi RJ, Sandhu HK, Iyer RM, et al. Characterization and forming same into cables. US Patent
of physico-mechanical properties of indomethacin and 699,459; 1902.
polymers to assess their suitability for hot-melt extru- Dreiblatt A. Process design. In: Ghebre-Sellassie I,
sion process as a means to manufacture solid disper- Martin C, editors. Pharmaceutical extrusion technol-
sion/solution. J Pharm Sci. 2005;94(11):2463–74. ogy. New York: Marcel Dekker; 2003. p. 149–70.
Crowley MM, et al. Pharmaceutical applications of Emin MA, Schuchmann HP. Analysis of the dispersive
hot-melt extrusion: part I. Drug Dev Ind Pharm. mixing efficiency in a twin-screw extrusion processing
2007;33(9):909–26. of starch based matrix. J Food Eng [Internet]. 2013;115
Crowley M, Zhang F, Koleng J, et al. Hydrophobic abuse (1):132–143. Available from: https://doi.org/10.1016/
deterrent delivery system for hydromorphone. US Pat- j.jfoodeng.2012.10.008.
ent 20080020032; 2008. European Medicines Agency (EMA). Ozurdex EPAR
Dahan A, Miller JM, Hoffman A, Amidon GE, Amidon summary for the public. London: EMA; 2014. Docu-
GL. The solubility–permeability interplay in using ment Reference: EMA/477896/214, p 32
cyclodextrins as pharmaceutical solubilizers: mecha- Evans RC, Bochmann ES, Kyeremateng SO, Gryczke A,
nistic modeling and application to progesterone. J Wagner KG. Holistic QbD approach for hot-melt
Pharm Sci. 2010;99(6):2739–49. extrusion process design space evaluation: linking
Dantuluri AK, Amin A, Puri V, et al. Role of α-relaxation materials science, experimentation and process
on crystallization of amorphous celecoxib above T g modeling. Eur J Pharm Biopharm. 2019 Aug 1;141:
probed by dielectric spectroscopy. Mol Pharm. 2011;8 149–60.
(3):814–22. Fang LY, et al. High density compositions containing
Davis DA, Miller DA, Su Y, Williams RO. Thermally posaconazole and formulations comprising the same.
conductive excipient expands KinetiSol® processing US Patent Application CA2720849; 2009.
capabilities. AAPS PharmSciTech. 2020;21(8):1–15. Friesen DT, et al. Hydroxypropyl methylcellulose acetate
Debenedetti PG, Stillinger FH. Supercooled liquids and succinate-based spray-dried dispersions: an overview.
the glass transition. Nature. 2001;410(259–267):259. Mol Pharm. 2008;5(6):1003–19.
Deshpande TM, Shi H, Pietryka J, Hoag SW, Medek Ganjyal G, Hanna M. A review of residence time distribu-
A. Investigation of polymer/surfactant interactions tion in food extruders and study on the potential of
and their impact on itraconazole solubility and precipi- neural networks in RTD modeling. J Food Sci.
tation kinetics for developing spray-dried amorphous 2002;67(6):1996–2002.
solid dispersions. Mol Pharm. 2018 Jan 18;15(3): Geers S, et al. Polymer formulations of CETP inhibitors.
962–74. US Patent 8030359; 2010.
Dhumal RS, Kelly AL, York P, Coates PD, Paradkar Genina N, Hadi B, Löbmann K. Hot melt extrusion as
A. Cocrystallization and simultaneous agglomeration solvent-free technique for a continuous manufacturing
using hot melt extrusion. Pharm Res. 2010;27(12): of drug-loaded mesoporous silica. J Pharm Sci. 2018
2725–33. Jan 1;107(1):149–55.
DiNunzio JC. Formulation design of melt extruded Gilis PM, De Condé VFV, Vandecruys RPG. Beads hav-
systems for oral bioavailability enhancement. ing a core coated with an antifungal and a polymer. US
presented in Leistritz pharmaceutical extrusion semi- Patent 5633015; 1997.
nar, Clinton, NJ; 2010 Greenhalgh DJ, et al. Solubility parameters as predictors of
DiNunzio JC, et al. Amorphous compositions using con- miscibility in solid dispersions. J Pharm Sci. 1999;88:
centration enhancing polymers for improved bioavail- 1182–90.
ability of itraconazole. Mol Pharm. 2008;5(6):968–80. Groenewegen RJ, inventor; Akzo Nobel NV, assignee.
DiNunzio JC, Martin C, Zhang F. Melt extrusion: shaping Drug delivery system for two or more active
drug delivery in the 21st century. Pharmaceutical tech- substances. United States patent US 5,989,581; 1999
nology, pp s30–7, November supplemental; 2010a. Nov 23.
DiNunzio JC, et al. Fusion production of solid dispersions Gupta SS, Meena A, Parikh T, et al. Investigation of
containing a heat-sensitive active ingredient by hot thermal and viscoelastic properties of polymers rele-
vant to hot melt extrusion – I: polyvinylpyrrolidone
and related polymers. J Excipients Food Chem. 2014;5 conference-harmonisation-technical-requirements-reg

(1):32–45. istration-pharmaceuticals-human-use_en-11.pdf.
Gupta SS, Parikh T, Meena AK, et al. Effect of carbamaz- Islam MT, Scoutaris N, Maniruzzaman M, Moradiya HG,
epine on viscoelastic properties and hot melt extrud- Halsey SA, Bradley MSA, Chowdhry B, Snowden MJ,
ability of Soluplus(®). Int J Pharm. 2015a;478(1): Douroumis D. Implementation of transmission NIR as
232–9. https://doi.org/10.1016/j.ijpharm.2014.11.025. a PAT tool for monitoring drug transformation during
Gupta SS, Solanki N, Serajuddin AT. Investigation of HME processing. Eur J Pharm Biopharm. 2015;96:
thermal and viscoelastic properties of polymers rele- 106–16.
vant to hot melt extrusion, IV: Affinisol™ HPMC Jedinger N, Schrank S, Fischer JM, Breinhälter K,
HME polymers. AAPS PharmSciTech. 2015b;17(1): Khinast J, Roblegg E. Development of an abuse- and
1–10. alcohol-resistant formulation based on hot-melt extru-
Hanada M, Jermain SV, Lu X, Su Y, Williams sion and film coating. AAPS PharmSciTech. 2015;17
RO. Predicting physical stability of ternary amorphous (1):68–77.
solid dispersions using specific mechanical energy in a Jerman R, editor. Devolatilization of polymers via twin
hot melt extrusion process. Int J Pharm. 2018;548(1): screw extrusion. Proceeding of the Leistritz Twin
571–85. screw Extrusion Workshop; 2006 Bridgewater,
Hanada M, Jermain SV, Thompson SA, Furuta H, NJ, USA.
Fukuda M, Williams RO III. Ternary Amorphous Kanzer J, Hupfeld S, Vasskog T, et al. In situ formation of
Solid Dispersions Containing a High-Viscosity Poly- nanoparticles upon dispersion of melt extrudate
mer and Mesoporous Silica Enhance Dissolution Per- formulations in aqueous medium assessed by asym-
formance. Mol Pharmaceutics. 2020 Dec 8;18(1): metrical flow field-flow fractionation. J Pharm Biomed
198–213. Anal. 2010;53(3):359–65.
Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. Karimi-Jafari M, Ziaee A, Iqbal J, O’Reilly E, Croker D,
2000;100(1):57–70. Walker G. Impact of polymeric excipient on cocrystal
Hancock BC, Christensen K, Shamblin SL. Estimating the formation via hot-melt extrusion and subsequent
critical molecular mobility temperature (TK) of amor- downstream processing. Int J Pharm. 2019;566
phous pharmaceuticals. Pharm Res. 1998;15(11): (January):745–55.
1649–51. Keen JM, Martin C, Machado A, et al. Investigation of
Hanna MA, Gennadios A, Mandigo RW. Restructuring of process temperature and screw speed on properties of a
pork meat in a twin-screw extruder. J Food Process pharmaceutical solid dispersion using corotating and
Preserv. 1996;20(5):391–402. counter-rotating twin-screw extruders. J Pharm
Harmon PA, Variankaval N. Solid dosage formulations of Pharmacol. 2014;66(2):204–17.
an orexin receptor antagonists. World Patent Keen JM, Foley CJ, Hughey JR, Bennett RC, Jannin V,
WO2013181174 A3; 2015. Rosiaux Y, Marchaud D, McGinity JW. Continuous
Haser A, Huang S, Listro T, White D, Zhang F. An twin screw melt granulation of glyceryl behenate:
approach for chemical stability during melt extrusion development of controlled release tramadol hydrochlo-
of a drug substance with a high melting point. Int J ride tablets for improved safety. Int J Pharm. 2015;487
Pharm. 2017 May 30;524(1–2):55–64. (1):72–80.
Haser A, Haight B, Berghaus A, Machado A, Martin C, Kesisoglou F, Hermans A, Neu C, et al. Development of
Zhang F. Scale-up and in-line monitoring during con- in vitro–in vivo correlation for amorphous solid disper-
tinuous melt extrusion of an amorphous solid dispersion immediate-release suvorexant tablets and applica-
sion. AAPS PharmSciTech. 2018 Oct 1;19(7): tion to clinically relevant dissolution specifications and
2818–27. in-process controls. J Pharm Sci. 2015;104(9):
Huang S, O’Donnell KP, Keen JM, et al. A new extrudable 2913–22.
form of hypromellose: AFFINISOL™ HPMC HME. Kessler T, et al. Process for producing a solid dispersion of
AAPS PharmSciTech. 2015;17(1):1–14. an active ingredient. US Patent 20090302493; 2009.
Hughey JR, DiNunzio JC, Bennett RC, et al. Dissolution Kestur US, Wanapun D, Toth SJ, et al. Nonlinear optical
enhancement of a drug exhibiting thermal and acidic imaging for sensitive detection of crystals in bulk
decomposition characteristics by fusion processing: a amorphous powders. J Pharm Sci. 2012;101(11):
comparative study of hot melt extrusion and 4201–13.
KinetiSol® dispersing. AAPS PharmSciTech. Khaled SA, Burley JC, Alexander MR, Roberts
2010;11(2):760–74. CJ. Desktop 3D printing of controlled release pharma-
Humphreys A. Annual report: top 10 pipelines (2016). ceutical bilayer tablets. Int J Pharm. 2014;461(1):
Retrieved 02/23/16, from http://www.pharmalive. 105–11.
com/annual-report-top-10-pipelines/. Kim SJ, Kwon TH. Enhancement of mixing performance
ICH-Q8(R2). ICH guideline Q8 (R2) on pharmaceutical of single-screw extrusion processes via chaotic flows:
development EMA/CHMP/ICH/167068/2004. 2009;8 part I basic concepts and experimental study. Adv
(June). Available from: https://www.ema.europa.eu/ Polym Technol. 1996a;15(1):41–54.
en/documents/scientific-guideline/international-
Kim SJ, Kwon TH. Enhancement of mixing performance loading and drug release from mesoporous silica
of single-screw extrusion processes via chaotic flows: carriers. Eur J Pharm Biopharm. 2018 Sep 1;130:
part II numerical study. Adv Polym Technol. 1996b;15 327–35.
(1):55–69. Luker K. Single-screw extrusion and screw design. In:
Krishna R, Bergman AJ, Green M, et al. Model-based Ghebre-Sellassie I, Martin C, editors. Pharmaceutical
development of anacetrapib, a novel cholesteryl ester extrusion technology. New York: Marcel Dekker;
transfer protein inhibitor. AAPS J. 2011;13(2):179–90. 2003. p. 54–83.
Kulikov OL, Hornung K. A simple geometrical solution to Lyons JG, et al. Preparation of monolithic matrices for oral
the surface fracturing problem in extrusion processes. J drug delivery using a supercritical fluid assisted hot
Nonnewton Fluid Mech. 2001;98(2–3):107–15. melt extrusion process. Int J Pharm. 2007;329(1–2):
LaFountaine JS, McGinity JW, Williams RO. Challenges 62–71.
and strategies in thermal processing of amorphous Ma X, Huang S, Lowinger MB, et al. Influence of mechan-
solid dispersions: a review. AAPS PharmSciTech. ical and thermal energy on nifedipine amorphous solid
2015;17(1):43–55. dispersions prepared by hot melt extrusion: preparation
Lakshman JP, et al. Application of melt granulation tech- and physical stability. Int J Pharm [Internet]. 2019 Apr
nology to enhance tabletting properties of poorly com- 20;561:324–34. Available from: https://doi.org/10.
pactible high-dose drugs. J Pharm Sci. 2010;100(4): 1016/j.ijpharm.2019.03.014.
1553–65. Mahieu AL, Willart J-FO, Dudognon E, et al. A new
LaMattina, J. Forget HDL, Merck is banking on LDL for protocol to determine the solubility of drugs into poly-
its heart drug, anacetrapib. Forbes; 2015, Feb. mer matrixes. Mol Pharm. 2013;10(2):560–6.
Lamm MS, DiNunzio J, Khawaja NN, Crocker LS, Pecora Maniruzzaman M, Nair A, Scoutaris N, Bradley MS,
A. Assessing mixing quality of a copovidone-TPGS Snowden MJ, Douroumis D. One-step continuous
hot melt extrusion process with atomic force micros- extrusion process for the manufacturing of solid
copy and differential scanning calorimetry. AAPS dispersions. Int J Pharm. 2015a Dec 30;496(1):42–51.
PharmSciTech. 2015;17(1):89–98. Maniruzzaman M, Islam MT, Halsey S, Amin D,
Lauer ME, Siam M, Tardio J, et al. Rapid assessment of Douroumis D. Novel controlled release polymer-lipid
homogeneity and stability of amorphous solid formulations processed by hot melt extrusion. AAPS
dispersions by atomic force microscopy—from bench PharmSciTech. 2015b;17(1):191–9.
to batch. Pharm Res. 2013;30(8):2010–22. Marsac PJ, Shamblin SL, Taylor LS. Theoretical and prac-
Lenz E, Löbmann K, Rades T, Knop K, Kleinebudde tical approaches for prediction of drug-polymer
P. Hot melt extrusion and spray drying of miscibility and solubility. Pharm Res. 2006;23(10):
co-amorphous indomethacin-arginine with polymers. 2417–26.
J Pharm Sci. 2017 Jan 1;106(1):302–12. Marsac P, Taylor L, Xi H, et al. A novel method for
Li S, Yu T, Tian Y, McCoy CP, Jones DS, Andrews accessing the enthalpy of mixing active pharmaceutical
GP. Mechanochemical synthesis of pharmaceutical ingredients with polymers. Presented at American
cocrystal suspensions via hot melt extrusion: feasibility Association for Pharmaceutical Scientists National
studies and physicochemical characterization. Mol Meeting, Chicago, IL, USA; 2012.
Pharm. 2016;13(9):3054–68. Marschik C, Roland W, Miethlinger J. A network-theory-
Li M, Xu W, Su Y. Solid-state NMR spectroscopy in based comparative study of melt-conveying models in
pharmaceutical sciences. TrAC Trends Anal Chem. single-screw extrusion: A. Isothermal flow. Polymers
2020 Dec;15:116152. (Basel). 2018;10(8):929.
Liepold B, Rosenblatt K, Hoelig P, et al. Solid Martin C. Twin screw extruders as continuous mixers for
compositions. World Patent 2011156578 A1; 2011. thermal processing: a technical and historical perspec-
Liepold B, Jung T, Hölig P, et al. Solid compositions. US tive. AAPS PharmSciTech. 2016;17(1):3–19.
Patent 8686026 B2; 2014. McGinity JW, et al. Hot-melt extrusion technology. In:
Lim S, White JL. Flow mechanisms, material distributions Encyclopedia of pharmaceutical technology. Hoboken:
and phase morphology development in a modular Informa Healthcare; 2007. p. 2004–20.
intermeshing counter-rotating twin screw extruder of Medina C, Daurio D, Nagapudi K, Alvarez-Nunez
Leistritz design. Int Polym Process. 1994;9:33–45. F. Manufacture of pharmaceutical co-crystals using
Liu H, Wang P, Zhang X, Shen F, Gogos CG. Effects of twin screw extrusion: a solvent-less and scalable pro-
extrusion process parameters on the dissolution behav- cess. J Pharm Sci. 2010;99(4):1693–6.
ior of indomethacin in Eudragit® E PO solid Meena A, Parikh T, Gupta SS, et al. Investigation of
dispersions. Int J Pharm. 2010;383(1–2):161–9. thermal and viscoelastic properties of polymers rele-
Liu X, Lu M, Guo Z, Huang L, Feng X, Wu C. Improving vant to hot melt extrusion, IV: Affinisol™ HPMC
the chemical stability of amorphous solid dispersion HME polymers. AAPS PharmSciTech. 2014;17(1):
with cocrystal technique by hot melt extrusion. Pharm 148–57.
Res. 2012;29(3):806–17. Miller DA, et al. Targeted intestinal delivery of supersatu-
Lizoňová D, Mužík J, Šoltys M, Beránek J, Kazarian SG, rated itraconazole for improved oral absorption. Pharm
Štěpánek F. Molecular-level insight into hot-melt Res. 2008a;25(6):1450–9.
Miller DA, et al. Enhanced in vivo absorption of Pham TN, Watson SA, Edwards AJ, et al. Analysis of
itraconazole via stabilization of supersaturation follow- amorphous solid dispersions using 2D solid-state
ing acidic-to-neutral pH transition. Drug Dev Ind NMR and 1H T1 relaxation measurements. Mol
Pharm. 2008b;34(8):890–902. Pharm. 2010;7(5):1667–91.
Miller DA, et al. Hot-melt extrusion for improved dissolu- PhRMA Foundation. Twenty-five years of progress
tion properties of poorly water-soluble molecules. In against hepatitis C: setbacks and stepping stones.
U.S.P.a.T. Organization, editors. Basel: F. Hoffmann- Washington, DC: PhRMA; 2014.
La Roche, Inc.; 2011 Pipe CJ, Majmudar TS, McKinley GH. High shear rate
Miller JM, Beig A, Carr RA, et al. The solubility–perme- viscometry. Rheol Acta. 2008;47(5–6):621–42.
ability interplay when using cosolvents for solubiliza- Qi S, Belton P, Nollenberger K, et al. Compositional
tion: revising the way we use solubility-enabling analysis of low quantities of phase separation in hot-
formulations. Mol Pharm. 2012a;9(3):581–90. melt-extruded solid dispersions: a combined atomic
Miller JM, Beig A, Carr RA, et al. A win–win solution in force microscopy, photothermal Fourier-transform
oral delivery of lipophilic drugs: supersaturation via infrared microspectroscopy, and localised thermal
amorphous solid dispersions increases apparent solu- analysis approach. Pharm Res. 2011;28(9):2311–26.
bility without sacrifice of intestinal membrane perme- Qian F, Huang J, Zhu Q, et al. Is a distinctive single Tg a
ability. Mol Pharm. 2012b;9(7):2009–16. reliable indicator for the homogeneity of amorphous
Miller JM, Morris, JB, Sever, NE, et al. Solid antiviral solid dispersion? Int J Pharm. 2010a;395(1):232–5.
dosage forms. US Patent 20150258093; 2015. Qian F, Huang J, Hussain MA. Drug-polymer solubility
Mollan M. Historical overview. In: Ghebre-Sellassie I, and miscibility: stability considerations and practical
Martin C, editors. Pharmaceutical extrusion technol- challenges in amorphous solid dispersion develop-
ogy. New York: Marcel Dekker; 2003. p. 1–18. ment. J Pharm Sci. 2010b;99(7):2941–7.
Monschke M, Kayser K, Wagner KG. Processing of poly- Repka MA, et al. Influence of plasticizers and drugs on the
vinyl acetate phthalate in hot-melt extrusion—prepara- physical-mechanical properties of hydroxypropyl-
tion of amorphous solid dispersions. Pharmaceutics. cellulose films prepared by hot melt extrusion. Drug
2020;12:337. Dev Ind Pharm. 1999;25(5):625–33.
Muller W, Peck JV, inventors; LTS Lohmann Therapie Repka MA, et al. Pharmaceutical applications of hot-melt
Systeme GmbH and Co KG, Aderis Pharmaceuticals extrusion: part II. Drug Dev Ind Pharm. 2007;33(10):
Inc, assignee. Transdermal therapeutic system which 1043–57.
contains a d2 agonist and which is provided for treating Repka MA, et al. Applications of hot-melt extrusion for
Parkinsonism, and a method for the production thereof. drug delivery. Expert Opin Drug Deliv. 2008;5(12):
United States patent US 6,884,434; 2005 Apr 26. Col- 1357–76.
umn 4, Lines 6–16. Rosenberg J, Reinhold U, Liepold B, et al. Solid pharma-
Nakamichi K, Nakano T, Yasuura H, Izumi S, Kawashima ceutical dosage form. US Patent 8268349 B2; 2012.
Y. The role of the kneading paddle and the effects of Rosenberg J, Reinhold U, Liepold B, et al. Solid pharma-
screw revolution speed and water content on the prep- ceutical dosage form. US Patent 8399015; 2013.
aration of solid dispersions using a twin-screw Ross SA, Lamprou DA, Douroumis D. Engineering and
extruder. Int J Pharm. 2002;241(2):203–11. manufacturing of pharmaceutical co-crystals: a review
Newman A, Engers D, Bates S, et al. Characterization of of solvent-free manufacturing technologies. Chem
amorphous API: polymer mixtures using X-ray powder Commun. 2016;52(57):8772–86.
diffraction. J Pharm Sci. 2008;97(11):4840–56. Ross SA, Ward A, Basford P, McAllister M, Douroumis
Patil H, Tiwari RV, Repka MA. Hot-melt extrusion: from D. Coprocessing of pharmaceutical cocrystals for high
theory to application in pharmaceutical formulation. quality and enhanced physicochemical stability. Cryst
AAPS PharmSciTech. 2015;17(1):20–42. Growth Des. 2019;19(2):876–88.
Paudel A, Geppi M, Van den Mooter G. Structural and Rumondor AC, Marsac PJ, Stanford LA, et al. Phase
dynamic properties of amorphous solid dispersions: the behavior of poly (vinylpyrrolidone) containing amor-
role of solid-state nuclear magnetic resonance spectros- phous solid dispersions in the presence of moisture.
copy and relaxometry. J Pharm Sci. 2014;103(9): Mol Pharm. 2009;6(5):1492–505.
2635–62. https://doi.org/10.1002/jps.23966. Sauceau M, Fages J, Common A, et al. New challenges in
Peng T, She Y, Zhu C, Shi Y, Huang Y, Niu B, et al. polymer foaming: a review of extrusion processes
Influence of polymers on the physical and chemical assisted by supercritical carbon dioxide. Prog Polym
stability of spray-dried amorphous solid dispersion: Sci. 2011;36(6):749–66.
dipyridamole degradation induced by enteric Schenck L. Devolatilization using hot melt extrusion.
polymers. AAPS PharmSciTech. 2018;19(6):2620–8. Presented in Leistritz pharmaceutical extrusion semi-
Perdikoulias J, Dobbie T. Die design. In: Ghebre- nar, Clinton, NJ, USA; 2010.
Sellassie I, Martin C, editors. Pharmaceutical extrusion Schenck L, et al. Achieving a hot melt extrusion design
technology. New York: Marcel Dekker; 2003. space for the production of solid solutions. In: Chemi-
p. 99–110. cal engineering in the pharmaceutical industry: R&D to
manufacturing. New York: Wiley; 2011.
Schilling SU, et al. Citric acid as a solid-state plasticizer Thayer AM. Finding solutions. Chem Eng News. 2010;88:
for Eudragit RS PO. J Pharm Pharmacol. 2007;59(11): 13–8.
1493–500. Thiele W. Twin-screw extrusion and screw design. In:
Schlindwein W, Bezerra M, Almeida J, Berghaus A, Pharmaceutical extrusion technology. New York: Mar-
Owen M, Muirhead G. In-line UV-Vis spectroscopy cel Dekker; 2003. p. 69–98.
as a fast-working process analytical technology (PAT) Tho I, Liepold B, Rosenberg J, Maegerlein M, Brandl M,
during early phase product development using hot melt Fricker G. Formation of nano/micro-dispersions with
extrusion (HME). Pharmaceutics. 2018 Dec;10(4):166. improved dissolution properties upon dispersion of
Schmitt PD, Trasi NS, Taylor LS, et al. Finding the needle ritonavir melt extrudate in aqueous media. Eur J
in the haystack: characterization of trace crystallinity in Pharm Sci. 2010;40(1):25–32.
a commercial formulation of paclitaxel protein-bound Thompson SA, Williams III RO. Specific mechanical
particles by Raman spectroscopy enabled by second energy – an essential parameter in the processing of
harmonic generation microscopy. Mol Pharm. 2015;12 amorphous solid dispersions. Adv Drug Deliv Rev;
(7):2378–83. 2021 Mar 27; https://doi.org/10.1016/j.addr.2021.
Shah N, Sandhu H, et al. Amorphous solid dispersions: 03.006.
theory and practice. New York: Springer; 2014. Tian Y, Jacobs E, Jones DS, McCoy CP, Wu H, Andrews
Sharma PK, Panda A, Pradhan A, Zhang J, Thakkar R, GP. The design and development of high drug loading
Whang CH, Repka MA, Murthy SN. Solid-state stabil- amorphous solid dispersion for hot-melt extrusion plat-
ity issues of drugs in transdermal patch formulations. form. Int J Pharm. 2020;586:119545.
AAPS PharmSciTech. 2018 Jan 1;19(1):27–35. Todd DB. Introduction to compounding. In: Plastics
Siew A. Solving poor solubility to unlock a drug’s poten- compounding: equipment and processing. New York:
tial. Pharm Technol. 2015;39:20–7. Hanser Publishers; 1995. p. 1–12.
Steiner R. Extruder design. In: Pharmaceutical extrusion Toth SJ, Madden JT, Taylor LS, et al. Selective imaging of
technology. New York: Marcel Dekker; 2003. active pharmaceutical ingredients in powdered blends
p. 20–38. with common excipients utilizing two-photon excited
Sun Y, Xi H, Ediger M, et al. Diffusion-controlled and ultraviolet-fluorescence and ultraviolet-second order
“diffusionless” crystal growth near the glass transition nonlinear optical imaging of chiral crystals. Anal
temperature: relation between liquid dynamics and Chem. 2012;84(14):5869–75.
growth kinetics of seven ROY polymorphs. J Chem Tumuluri VS, et al. Off-line and on-line measurements of
Phys. 2009;131(7):074506. drug-loaded hot-melt extruded films using Raman
Sun Y, Tao J, Zhang GG, et al. Solubilities of crystalline spectroscopy. Int J Pharm. 2008;357(1–2):77–84.
drugs in polymers: an improved analytical method and United States Pharmacopeia and National Formulary. USP
comparison of solubilities of indomethacin and nifedi- 38-NF 33 [Internet]. 2015 [revision 24 Apr 2015; cited
pine in PVP, PVP/VA, and PVAc. J Pharm Sci. 15 Jan 2020]. Available from: https://www.uspnf.com/
2010;99(9):4023–31. official-text
Suwardie H, et al. Rheological study of the mixture of Van den Mooter G. The use of amorphous solid
acetaminophen and polyethylene oxide for hot-melt dispersions: a formulation strategy to overcome poor
extrusion application. Eur J Pharm Biopharm. solubility and dissolution rate. Drug Disc Today
2011;78(3):506–12. Technol. 2012;9(2):e79–85.
Tadmor Z, Gogos CG. Principles of polymer processing. Van Laarhoven JAH, Kruft MAB, Vromans H. In vitro
Hoboken: Wiley; 2013. release properties of etonogestrel and ethinyl estradiol
Tallury P, Alimohammadi N, Kalachandra S. Poly (ethyl- from a contraceptive vaginal ring. Int J Pharm.
ene-co-vinyl acetate) copolymer matrix for delivery of 2002;23:163–73.
chlorhexidine and acyclovir drugs for use in the oral Verreck G, Six K, Van den Mooter G, et al. Characteriza-
environment: effect of drug combination, copolymer tion of solid dispersions of itraconazole and
composition and coating on the drug release rate. Dent hydroxypropylmethylcellulose prepared by melt extru-
Mater. 2007 Apr 1;23(4):404–9. sion—part I. Int J Pharm. 2003;251(1):165–74.
Tan DK, Davis DA, Miller DA, et al. Innovations in Verreck G, et al. The effect of pressurized carbon dioxide
thermal processing: hot-melt extrusion and KinetiSol® as a plasticizer and foaming agent on the hot melt
dispersing. AAPS PharmSciTech. 2020;21:312. extrusion process and extrudate properties of pharma-
https://doi.org/10.1208/s12249-02001854-2. ceutical polymers. J Supercrit Fluids. 2006a;38(3):
Tanno F, Nishiyama Y, Kokubo H, Obara S. Evaluation of 383–91.
hypromellose acetate succinate (HPMCAS) as a carrier Verreck G, et al. Hot stage extrusion of p-amino salicylic
in solid dispersions. Drug Dev Ind Pharm. 2004;30(1): acid with EC using CO2 as a temporary plasticizer. Int
9–17. J Pharm. 2006b;327(1–2):45–50.
Terife G, Wang P, Faridi N, Gogos CG. Hot melt mixing Verreck G, et al. The effect of supercritical CO2 as a
and foaming of Soluplus® and indomethacin. Polym reversible plasticizer and foaming agent on the hot
Eng Sci. 2012;52(8):1629–39. stage extrusion of itraconazole with EC 20 cps. J
Supercrit Fluids. 2007;40(1):153–62.
Vora C, Patadia R, Mittal K, Mashru R. Preparation and Yang F, Su Y, Zhang J, DiNunzio J, Leone A, Huang C,
characterization of dipyridamole solid dispersions for Brown CD. Rheology guided rational selection of
stabilization of supersaturation: effect of precipitation processing temperature to prepare copovidone–nifedi-
inhibitors type and molecular weight. Pharm Dev pine amorphous solid dispersions via hot melt extru-
Technol. 2015;21(7):847–55. sion (HME). Mol Pharm. 2016 Oct 3;13(10):
Vyazovkin S, Dranca I. Effect of physical aging on nucle- 3494–505.
ation of amorphous indomethacin. J Phys Chem Yoshioka M, Hancock BC, Zografi G. Inhibition of indo-
B. 2007;111(25):7283–7. methacin crystallization in poly(vinylpyrrolidone)
Wanapun D, Kestur US, Kissick DJ, et al. Selective detec- coprecipitates. J Pharm Sci. 1995;84(8):983–6.
tion and quantitation of organic molecule crystalliza- Zecevic DE, Wagner KG. Rational development of solid
tion by second harmonic generation microscopy. Anal dispersions via hot-melt extrusion using screening,
Chem. 2010;82(13):5425–32. material characterization, and numeric simulation
Wanapun D, Kestur US, Taylor LS, et al. Quantification of tools. J Pharm Sci. 2013;102(7):2297–310.
trace crystallinity in an organic powder by nonlinear Zecevic DE, Evans RC, Paulsen K, Wagner KG. From
optical imaging: investigating the effects of mechanical benchtop to pilot scale–experimental study and compu-
grinding on crystallinity loss. In: Abstracts of papers of tational assessment of a hot-melt extrusion scale-up of
the American Chemical Society meeting, Washington, a solid dispersion of dipyridamole and copovidone. Int
DC, USA; 2011. J Pharm. 2018;537(1–2):132–9.
Wang S-Q, Drda P. Molecular instabilities in capillary Zhang W, Hate SS, Russell DJ, Hou HH, Nagapudi
flow of polymer melts: interfacial stick-slip transition, K. Impact of surfactant and surfactant-polymer inter-
wall slip and extrudate distortion. Macromol Chem action on desupersaturation of clotrimazole. J Pharm
Phys. 1997;198(3):673–701. Sci. 2019 Oct 1;108(10):3262–71.
Yang M, Wang P, Suwardie H, et al. Determination of Zhao Y, et al. Prediction of the thermal phase diagram of
acetaminophen’s solubility in poly(ethylene oxide) by amorphous solid dispersions by Flory-Huggins theory.
rheological, thermal and microscopic methods. Int J J Pharm Sci. 2011;100:3196–207.
Pharm. 2011;403(1–2):83–9. Zhu L, Wong L, Yu L. Surface-enhanced crystallization of
amorphous nifedipine. Mol Pharm. 2008;5(6):921–6.
Spray-Drying Technology
10
Dave A. Miller, Daniel Ellenberger, Tiago Porfirio,
and Marco Gil
Abstract spray-drying process and its uses as a formula-

This chapter provides an in-depth review of tion technology toward the enhancement of
spray-drying technology and its application to drug delivery with poorly water-soluble
the formulation of poorly water-soluble drugs. compounds. The current authors would like
In the early part of the chapter, the to thank and acknowledge the previous
fundamentals of the process are discussed, authors’ significant contribution from the first
including process theory, process components, and second edition. This current third edition
equipment options, equipment by scale, vari- chapter is a revision and update of the original
ous feeds, and typical solvent systems. In the authors’ work from the first and second
latter part of the chapter, the application of editions.
spray drying to the formulation of poorly
water-soluble drugs is discussed. Particular Keywords
emphasis is given to spray drying for amor- Spray congealing · Microencapsulation ·
phous solid dispersion systems. The path Atomization · Nozzles · Cyclones · Filter bags ·
toward developing an amorphous spray-dried Alcohols · Ketones · Tetrahydrofuran ·
dispersion and conversion to a final dosage Tablets · Wall materials
form is covered in detail. Additionally, several
academic and industrial examples are
presented, illustrating the benefits of the pro- 10.1 Background
cess as a formulation technology and its com-
mercial viability. Finally, the application of At its essence, spray drying is a continuous means
spray drying to inhalation and emerging of extracting dry solids from a fluid by evapora-
applications, i.e., spray congealing and micro- tion of the carrier liquid. By this process, one can
encapsulation, are reviewed. This chapter start with a solution, suspension, slurry, emulsion,
provides comprehensive coverage of the low-viscosity paste, or the like and convert it into
a freely flowing powder in a single step (Celik
D. A. Miller (*) · D. Ellenberger and Wendel 2005). One of the earliest
DisperSol Technologies, LLC, Georgetown, TX, USA descriptions of a spray-drying process was
e-mail: [email protected]; daniel. published in 1872 in US Patent 125,406 entitled
[email protected] “Improvement in Drying and Concentrating Liq-
T. Porfirio · M. Gil uid Substances by Atomizing.” This patent
Hovione Farmaciência SA, Lisbon, Portugal describes a process whereby a fluid is converted
e-mail: tporfi[email protected]; [email protected]
378 D. A. Miller et al.
to a state of “minute division” with simultaneous technology as they apply to the formulation of
exposure to “currents of air or other gasses” for poorly water-soluble drugs.
the purpose of rapidly drying a solid substance
from a liquid carrier (Percy 1872). From this early
description, spray drying has since evolved into a
10.2 Process Overview
diverse technology with a long history of success-
ful commercial applications.
Spray drying consists of feeding a liquid stream
Spray drying has been utilized industrially for
that is continuously divided in fine droplets
well over a century and consequently has fully
(atomization) into a chamber (drying chamber).
matured and is well understood. Owing to its
In the drying chamber, the droplets come into
simplicity, efficiency, and robustness, the tech-
contact with a hot gas and, by an evaporative
nology has been employed in a variety of
cooling process, are converted into solid particles.
industries for the production of a vast array of
These particles are then separated from the wet
products. It is commonly used to process milk,
drying gas by a suitable separation system, most
eggs, ceramics, fertilizers, detergents, and numer-
commonly a cyclone or filter bag. To summarize,
ous other chemicals (Broadhead et al. 1992). In
the process is divided into four main steps:
the pharmaceutical industry, spray drying has
(1) spray formation, atomization of the feed solu-
been used for the production of bulk actives,
tion/suspension/emulsion; (2) droplet–gas con-
including small molecules, vitamins, peptides,
tact; (3) droplet drying and particle formation;
and proteins (Celik and Wendel 2005). It has
and (4) separation of solid particles from the wet
also been employed as a process technology to
drying gas (Masters 2002). A schematic diagram
impart unique functional attributes onto
of the spray-drying process is provided in
excipients, such as lactose, mannitol, and micro-
Fig. 10.1.
crystalline cellulose, to name a few. Additionally,
spray drying has been employed for the produc-
tion of novel drug-delivery systems with primary
applications in granulation and particle 10.2.1 Solution Preparation
engineering.
Spray drying has recently been the focus of The first step of spray-drying process is the solu-
increased interest as a technology applicable to tion preparation. Depending on the product appli-
improving bioavailabilities of poorly water- cation, the spray material could be in a suspension
soluble drugs. In this chapter, the utilization of (for instance, for microencapsulation purposes),
spray-drying technologies toward the enhance- in an emulsion, or in a solution in which the
ment of formulations containing poorly water- material is dissolved in a suitable solvent mixture.
soluble drugs is discussed in detail. Particular The feed solution composition impacts in the
emphasis is given to the use of spray drying for product attributes and processability and its defi-
the formulation of amorphous solid dispersions as nition should take some considerations such as:
this is perhaps the most common application of
• Solid content: is strongly influenced and lim-
the technology to this field. In addition, this chap-
ited by solubility of the components, mainly
ter aims to provide the reader with a basic under-
the drug, on the solvent mixture. Such feed
standing of the process fundamentals, the various
concentration influences the density and size
process components, and the equipment options
of the final product. High concentration of
at different scales. For a more detailed discussion
solids may produce larger and denser particles
regarding the principles of spray drying, the
comparing to the dilute solution. This is
inquiring reader is referred to the Spray Drying
justified by the difference of liquid to evapo-
Handbook by Masters (1985). Finally, this chap-
rate and by the diffusion of solutes within the
ter covers emerging applications of spray-drying
droplet (Vicente et al. 2013).
10 Spray-Drying Technology 379
Fig. 10.1 Schematic diagram of the spray-drying process
• Rheological properties: both concentration The organic solvents are normally used for the
and composition (solvent mixture and solutes) production of SDD due to the limited water solu-
influence the rheological properties of the bility of the drugs.
solution such as viscosity, surface tension,
and rheological behavior. The rheological
properties impact directly in the droplet forma-
10.2.2 Atomization
tion and drying kinetics, mainly if the fluid has
a non-Newtonian behavior.
The solution prepared is then pumped and fed to
• Stability: chemical profile should remain sta-
an atomizer placed inside the spray-drying cham-
ble in solution for the process duration. The
ber. Atomization, the generation of very fine
combination of solvents, drug, polymers, and
droplets from a liquid feed, is a key aspect of
other excipient could be a concern in terms of
the spray-drying process. It substantially
chemical stability of the species. This becomes
increases the liquid surface area to vastly improve
more relevant in manufacturing scales in
the efficiency of heat and mass transfer. For
which the batches can take from few hours to
example, 1 cubic meter of liquid atomized into
several days. Drug degradation, precipitation,
droplets of 100 μm in diameter yields a surface
and/or impurity growth can occur during such
area of 60,000 m2. By the generation of such high
period due to the reactivity of the components
surface area, the drying process proceeds rapidly,
in solutions. As such, the solution stability
completing in seconds or fraction of a second
should be assessed over time to support the
depending on equipment scale. Control of the
solvent selection and processability (Singh and
atomization process enables tuning of the droplet
Van den Mooter 2016).
size and consequently product particle size. Parti-
cle size of a spray-dried product is especially
relevant to downstream processing and final dos- the edge of the disk/wheel; therefore, the degree of
age form production. atomization depends primarily on the peripheral
There are several types of atomizers that can speed which is a function of rotational velocity and
be used to produce the feed spray inside the diameter. Higher speeds are typically used with
drying chamber. These nozzles are classified lower-diameter nozzles that are installed in
according to the type of energy used and include smaller-scale spray driers (Masters 2002).
rotary nozzles (centrifugal energy), two-fluid or Rotary nozzles can be used to atomize slurries,
pneumatic nozzles (kinetic energy), pressure suspensions, or solutions of high viscosity.
nozzles (pressure energy), and ultrasonic nozzles Besides the feed properties, the operating
(acoustic energy). There are several variables that influence droplet size are feed
configurations available for each type of nozzle flow, rotational velocity, wheel diameter, and
which are able to produce sprays with specific design. The general correlation that can be used
features. Also, some nozzles combine the use of to estimate mean droplet size is the following:
more than one type of energy to generate the
spray. The feed stream properties (viscosity, sur- D50 ðμmÞ ¼ KF x N y dz ðnhÞw 104 :
face tension, solids load, etc.) impact the degree
In this equation, K is a feed-related constant;
of atomization achievable for all types of
F is the feed flow (kg/h); N is the wheel speed
atomizers, but their sensitivity to each feed stream
(rpm); d is the wheel diameter; n is the number of
property with regard to droplet size depends on
vanes, bushings pins, or holes; and h is the height
the type of nozzle. In the pharmaceutical industry,
of a vane or pin, or half of the circumference of
the most common nozzles are pneumatic
circular bushing openings (m). The x, y, z, and
(two-fluid) and pressure nozzles due to their scal-
w parameters are system dependent and must be
ability and reduced tendency to generate wall
determined experimentally.
deposits as compared to rotary nozzles.
Pneumatic Two-Fluid Nozzles

Rotary Nozzles
Pneumatic two-fluid nozzles use a gas stream to
For rotary nozzles, atomization is achieved by
atomize the feed which explains their designation.
centrifugal energy transmitted to the liquid stream
The mixing of the liquid stream and the atomizing
by a disk or wheel rotating at high speed (from
gas can be performed in the nozzle’s tip as
3000 to 50,000 rpm). The droplets are generated at
depicted in Fig. 10.2a (external mixing two-fluid
Fig. 10.2 Pneumatic a b

two-fluid nozzles: (a) Liquid Liquid
external mixing and (b)
internal mixing. Image (b)
was reproduced with Gas Gas
permission from Delavan
Spray Technologies and
Goodrich Corporation,
respectively
Gas
Mixing
chamber
nozzle) or in a chamber inside the nozzle as Lefebvre and Wang 1987; Vig and Morgen
depicted in Fig. 10.2b (internal mixing two-fluid 2017).
nozzle). Although the external mixing two-fluid These nozzles require the use of high-pressure
nozzles are the most common in small-scale and pumps as pressures of up to 450 bar can be
pilot-scale spray driers, the internal mixing required. They may be applied to all types of
nozzles are more efficient and preferred in feeds, but special attention must be paid when
larger-scale spray driers, especially when small spraying suspensions. The size of the suspended
particle sizes (less than 10 microns) are required. material should be controlled to avoid nozzle
In this type of nozzle, the gas mass flow-to-liquid clogging, nozzle erosion, and/or pump failure. In
mass flow ratio (often called the atomization larger-scale spray driers, these nozzles are good
ratio) is used to characterize the degree of for production of medium to large particles
atomization. (30–200 μ). They also produce more uniform
Correlations have been developed to estimate powders than pneumatic two-fluid nozzles and,
droplet size produced with two-fluid nozzles therefore, are preferred for the production of
based on liquid sheet formation and destabiliza- powders for oral dosage forms (Fig. 10.3).
tion through an accelerative mechanism. In the The modeling of droplet size for pressure
case of two-fluid nozzles, the Sauter mean diam- nozzles is more developed than for two-fluid
eter can be estimated by the following correlation: nozzles. This is attributable to the application of
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pressure nozzles in combustion engines and the
u
2π u σ:λKH optimization of those engines over the years.
SMD ¼ u
max
qffiffiffiffi 2 ,
5 t ρ There are several semiempirical correlations that
10:ρg : U G : 1 þ ρg U L can be easily used to estimate droplet size, such as
l
those proposed by Radcliffe, Jasuja, and Lefebvre

where σ is the surface tension, λKH max is the maxi- (Jasuja 1979; Lefebvre 1987; Radcliffe 1960).
mum Kelvin–Helmholtz wavelength, ρg is the Droplet size can be tuned by adjusting operating
atomization gas density, Ug is the atomization pressure, but feed flow is not independent of the
gas velocity at nozzle tip, ρL is the liquid density, pressure. With increasing atomization pressure,
and UL is the liquid velocity. the droplet size decreases and feed flow increases.
Therefore, the selection of the nozzle size must
Pressure Nozzles take into account not only pressure but also feed
With pressure nozzles the atomization is achieved flow requirements/drying capacity (Fig. 10.4).
by transforming pressure into kinetic energy. Dif-
ferent types of pressure nozzles are available in
Fig. 10.3 Pressure nozzle.
which pressure swirl is the most usually applied Reproduced with
to spray-drying process. In spray drying, pressure permission from Delavan
nozzles are designed to produce hollow swirling Spray Technologies and
sprays; hence, they are also known as pressure- Goodrich Corporation,
respectively
swirl nozzles. The pressure-swirl nozzle is
designed to have a swirl chamber immediately
before the discharge orifice. Inside the chamber
a swirl motion is produced through the passage of
the liquid in slots or holes. The swirl motion
pushes the liquid to the chamber walls creating
an air core from the discharge orifice. The liquid
fluid is ejected through the orifice in a sheet in a
hollow cone shape which spreads radially due to
the centrifugal forces (Gaspar et al. 2014;
Fig. 10.4 Droplet size and 160 90

feed flow variation with
140 80
atomization pressure
120 70
Droplet D50 (µm)
Feed flow (kg/h)

60
100
50
80
40
60
30
40 20
20 10
0 0
0 20 40 60 80 100 120 140 160 180 200 220
Atomization presssure (bar)
Ultrasonic Nozzles the capability to produce droplets of single

Ultrasonic nozzles use acoustic energy to pro- micron diameters. However, the most interesting
mote vibration of the nozzle tip where the forma- feature of this nozzle is the possibility to feed two
tion of droplets takes place. The droplet size can liquid streams. This enables the production of
be tuned by the ultrasound frequency and the solid dispersions by spray drying where the active
nozzle tip design. Their application is primarily ingredient and the polymer are not dissolved in
for laboratorial-scale spray driers due to two main the same solvent (Mizoe et al. 2008).
factors: (1) low droplet velocity enabling drying
of larger droplets than two-fluid nozzles and Monodisperse Nozzles
(2) limited spray throughput. An important limi- All of the nozzles described above produce poly-
tation of these nozzles is that drying temperatures disperse sprays; however, small and uniform
are typically limited to less than 110 C. particles are required for certain applications,
such as high-efficacy drug microcapsules, colloi-
Three- and Four-Fluid Nozzles dal drug delivery, and dry formulation vehicles.
Other nozzles are described in the literature fea- For these cases, the application of monodisperse
turing specific designs to accomplish particular droplet generators (MDGs) further enhances the
tasks. The so-called three-fluid nozzles are utility of spray-drying technology (Wu et al.
described by Kirkpatrick et al. (1986). This noz- 2007). MDGs were originally developed based
zle has two different stages of atomization and on ink-jet printing technology (Le 1998). The
claims to enable higher solids load and the capa- technology has since improved and the nozzle
bility of handling highly viscous feeds. has found applications in other areas. Among
According to its design, a liquid is premixed those are the development of mechano-
with the atomization gas creating a primary hydrodynamic droplet generators associated
spray of coarse droplets, not suitable for spray with piezoelectric transducers that enable the pro-
drying. This primary spray is then deflected radi- duction of fine mono-droplets (Wu et al. 2007).
ally outward by collision with a deflector plate. A Other systems are based on laser-drilled orifices
second atomizing gas stream, concentric to the and are used in the production of microchips, e.g.,
deflector, promotes further breakup of the pri- Nanomi microsieve™ (spans of less than 1.0
mary spray, thus creating a fine spray suitable were achieved with PLGA particles). However,
for spray drying. these systems are still restricted to laboratorial
A four-fluid nozzle has also been described in applications mainly due to cost and low through-
the literature by Mizoe et al. (2008). This nozzle put. Throughput is limited with these systems
has two liquid and two gas passages and claims because the Ohnesorge and Reynolds numbers
must be within certain ranges to achieve the The ASDs production is normally executed
monodisperse droplet regimen. with pressure nozzle to obtain a narrow droplet
size distribution and consequently particle distri-
Flash Nozzles bution (Vig and Morgen 2017). The scalability of
The use of spray-drying solution with the atomization device, easy operation, and
temperatures higher than the solvent mixture boil- cleaning procedures are also advantages of such
ing point may be considered to increase the nozzle to support the selection in the production
solutes solubility. Under such approach, the solu- of SDDs.
tion stream is pressurized to maintain the liquid
phase as described in US 9,248,584 (Friesen et al.
2011). The atomization is supported by flash 10.2.3 Gas–Droplet Contact
nozzles which promotes solution cavitation prior and Particle Formation
to the orifice exit. The cavitation is induced by the
pressure drop in the nozzle. The pressure The produced droplets are dried in the chamber
decreases from the nozzle inlet to the outlet using a hot gas to evaporate the solvent and form
achieving lower values than solvent mixture particles. The final product attributes depend on
vapor pressure promoting the partial vaporization the drying rate and process that the droplet
of the system. At the nozzle outlet, the solution undergoes. There are two configurations for
droplet is formed leaving with the gas-phase sol- spray drying, depending on the direction of the
vent. Sweep gas and nonstick coatings are used to drying gas flow and the feed being atomized:
reduce solids buildups and system clogging. (1) cocurrent and (2) countercurrent, as depicted
in Fig. 10.5. In cocurrent configuration, the inlet
Atomizer Selection drying gas and atomized feed stream both origi-
Selection of the appropriate atomizer depends on nate from the top of the drying chamber. The
the following: droplets descend with the drying gas and exit
through the bottom of the drying chamber. In
– Feed properties (suspended particles, viscos- countercurrent mode, the atomized feed and the
ity, rheologic behavior, etc.) drying gas originate from opposite ends of the
– Feed flow capacity requirements. drying chamber. By this configuration, residence
– Particle-size distribution requirements. time in the drying chamber is increased. Of the
– Gas disperser design of the spray-drying unit two configurations, cocurrent is the most com-
(see section “Gas dispersers”). monly used in the pharmaceutical industry
because residence time is shorter and, therefore,
In pharmaceutical spray drying, two-fluid noz- particles are subjected to less thermal stress.
zle and pressure nozzle are the most commonly The drying history of a droplet can be divided
used due to the scalability and easy operation and into different phases, as shown in Fig. 10.6. The
cleaning (Gaspar et al. 2014). The use of atomization process is an important step for an
two-fluid nozzle is normally dedicated to high- efficient drying phenomenon due to the produc-
viscosity solutions, suspensions, or when fine tion of large surface area of the droplets, its dis-
particles are required, for instance, in inhalation tribution, pattern, and velocity. As soon as the
products. The most common atomizers used in droplet–gas contact occurs, the evaporation pro-
lab scale are the pneumatic two-fluid nozzles due cess starts, and the heat and mass transfer evolu-
to their simplicity and flexibility. Ultrasonic noz- tion determine the final product attributes. The
zle is restricted to lab scale; nevertheless, it is initial temperature of the droplet increases as
used to support scale-up activities and to isolate well as the drying rate. Then, the evaporation
shear-sensitive materials (biologics) (Fernandes occurs at a constant rate, called constant rate
et al. 2017; Vicente 2014). Table 10.1 compares period (stage 1), which is driven by heat transfer
various properties of different atomizers. to the droplet. At this point the droplet surface
Table 10.1 Comparison between types of atomizers (Lefebvre and McDonell 2017; Masters 1979)
T`wo-fluid nozzle Pressure nozzle Rotary nozzle Ultrasonic nozzle
Scheme Gas Liquid Gas Liquid Core head
Liquid Rotation driver Housing
Liquid Liquid
Slots
air core
Orifice
opening
Disk
External Internal
Energy Kinetic energy from Liquid kinetic energy Centrifugal energy Acoustic energy
high-velocity gas from pressure energy
stream
Droplet size Atomization gas flow Not possible without Rotation speed Frequency and power
adjustment rate changing feed flow rate
Droplet size 5–75 μm 30–400 μm 20–200 μm 30–120 μm
and Wide distribution Narrow distribution Narrow distribution Very narrow
distribution distribution
Sensitivity for No Yes No Possible
viscosity
Operability Easy to operate. Not suitable for Tendency to form Power required depends
Nevertheless, requires suspension, blockage wall deposits due to on liquid properties
a gas stream for can occur in presence of wide spray angle
operation particles
Scalability Both scalable. Internal Easy. Difficult to Easy Difficult
mixture is normally operate in small scales
used at large scale. such lab
Major limitations: Gas
stream consumption,
and pressure drop
Advantages Good atomization and Simple and cheap. Uniform Narrow distribution and
suitability for high Wide range of spray atomization and low droplet velocity.
viscous liquids angles good control of Low shear rate in the
atomization quality nozzle
Disadvantages Need of an external High pressure required. High spray angle Scale-up
source of gas stream spray angle varies with and cleaning issues
(with high pressure in pressure
case of internal
mixture)
temperature is at the wet bulb ztemperature and droplet surface forming a shell or skin. The stage
remains constant during this stage (Mezhericher 2 starts with the shell formation, and the drying
et al. 2007). The droplet diameter decreases due rate decreases due to the additional resistance to
to the solvent evaporation that migrates from the the mass transfer. The evaporation is then
core to the surface by diffusion. Moreover, the dominated by the solid layer that creates a barrier
solute present in surface migrates for the droplet for the evaporated solvent (Singh and Van den
core due to the concentration gradient between Mooter 2016). Therefore, the pressure inside the
core and surface (Kim et al. 2003). This stage droplet/particle increases. The drying rate
continues while the saturation conditions are becomes nil when the equilibrium of moisture
maintained at droplet surface. content is reached. The final product attributes
When the solute diffusion is not able to follow are strongly dependent on the crust properties
the droplet reduction, precipitation occurs at the (as strength and thickness) and the ability to
Fig. 10.5 Spray-drying

configurations: (a)
cocurrent and (b)
countercurrent
Fig. 10.6 Drying process

of droplet. Adapted from
Singh and Van den Mooter
(2016)
support the internal pressure buildup. According and a decrease can be just achieved if exposed to a
to that, the droplet/particle can inflate, shrink, or different atmospheric relative saturation
break. Polymeric system as the case of SDDs has (modifying the equilibrium) (Santos et al. 2015).
normally the ability to inflate or shrink due to the The final product attributes are influenced by
elastic properties. Nonetheless, this ability not the drying mechanism depending on the evapora-
only depends on the polymer attributes but also tion rate and, therefore, operating conditions.
on drug load of the formulation. Finally, in stage Table 10.2 summarizes the impact of the most
3 the formed particle heats up due to the sensible important parameters in the product properties
drying. No mass variation is expected at this stage and the variables definition method. Besides the
once the equilibrium was already achieved. process conditions, the dissolved solutes and
Therefore, the solvent content remains constant, concentrations contribute to the particle
Table 10.2 Influence of process variables in product properties

Parameter Defined by Product properties impact Remarks
Inlet Physical and Low Tin promotes a slow drying and Normally the process is controlled by
temperature, chemical stability consequently high density with Tout, being Tin a surrogate of this variable
Tin of the shriveled particles
formulation
Feed flow Target Increases droplet size, consequently Increases heat requirements
rate, Ffeed evaporation rate particle size
Increases outlet relative saturation,
therefore the level of residual solvents
Drying gas Scale of Limited effect on product attributes Influences the droplet/particle trajectories
flow rate, operation which may influence the chamber wall
Fdrying deposits
Outlet Particle attributes Low Tout promotes a slow drying and Depending on particle formation and
temperature, and physical consequently high density with materials properties, it influences the
Tout stability shriveled particles particle breakage, stringing, and bearding
Condenser Equipment Promotes the solvent removal, Dependent on temperature and flow rate
temperature, limitations, therefore decreases outlet relative of the refrigeration fluid
Tcond drying rate of saturation, and promotes product
product dryness
formation phenomenon by changing the fluid (Fig. 10.7) depending on two dimensionless num-
properties. Bercea et al. (2009) and Paudel and bers: Weber number, We, and an impact parame-
Van den Mooter (2012) demonstrated the influ- ter, B. Each number is specific for the droplet
ence of solvent and fluid properties on the evapo- involved on the process (size and properties).
ration rate drying kinetics and phase separation. The droplet can interact not only with other
In fact, a deviation in drying kinetics due to the droplets but also with sticky particles and with
presence of polymer is observed by changing dried particles. In this case, the collision outcome
solvent systems, using PVP in methanol, acetone, can be the agglomeration of the material
and dichloromethane and mixtures thereof. depending on velocity, force, and time of contact
(Finotello 2019). Nevertheless, the interaction
between droplet and sticky particles or dried
Coalescence
particles is not likely to occur due to the expan-
The spray produces a distribution of droplets
sion of the spray plume. A narrow droplet size
which will differ in the drying history and solvent
distribution may avoid the agglomeration once
concentration evolution and consequently final
the droplet will have similar drying history.
properties. After the formation, the droplets can
Finotello (2019) simulated the spray drying of
interact with each other by their proximity, trajec-
milk to predict the collision rate of droplets and
tory, and velocity. The droplet–droplet collision
particles (Finotello et al. 2019). The study
can promote coalescence, changing the output
concluded that the collisions of droplet/particles
excepted mainly in final product particle size.
and particles–particles (either sticky or dried) are
Nevertheless, the droplet coalescence process is
very rare comparing with droplet–droplet
dependent not only of the initial droplet sizes but
interactions. Depending on the fluid properties
also on the droplet’s velocity and properties. The
and concentration, the interaction besides
collision of two droplets can result in coales-
droplet–droplet represents up to 2% of the total
cence, reflexive separation, stretching separation,
collisions and 9% for the case of high-viscosity
or bouncing. The outcome could be assessed by
fluids (1200 cP).
collision maps which are proposed in the litera-
Besides the droplet/particles velocities,
ture. Jian et al. (1996) (Jiang et al. 1992), Qian
trajectories, and properties, other variables may
and Law (1997) (Qian and Law 1997), and Ko
promote the interaction of objects during the
and Ryou (2005) proposed an outcome map
Fig. 10.7 Droplet collision outcome map (left) and regimes description: bouncing, coalescence, reflexive separation,
and stretching separation (right). Adapted from Finotello et al. (2018, 2019) and Qian and Law (1997)
spray-drying process. Besides feed flow rate and countercurrent) and the atomization nozzle to be
atomization pressure (both influence the velocity installed. Since cocurrent spray driers equipped
and size of droplets), spray angle and drying gas with pneumatic or pressure nozzles are the most
flow rate may impact on the collision. Intuitively, commonly used in the pharmaceutical industry,
the influence of spray angle appears to promote only this case will be discussed.
the collision process: if narrow angle is used, the The dispersers in these cases can be divided
droplets–particles number per volume unit into two groups depending on the flow pattern.
increases, increasing the probability of collision. One design creates a rotation pattern around the
Regarding the drying gas flow rate, the increment atomization nozzle (suitable for rotary, pneu-
of gas velocity impacts on the droplet/particle matic, and pressure nozzles). This rotation
trajectory and velocity. The combination of both improves the solvent evaporation capacity but
variables (narrow angles and high gas flow rate) promotes recirculation of powder at the top of
may increase the collision process. the chamber potentially leading to product
deposits on the ceiling. Also, in the case of
low-density particles, the swirling air pattern sig-
Gas Dispersers
nificantly increases the residence time of powder
Gas dispersers are very important in the spray-
inside the chamber. This swirling pattern can also
drying unit because they define the air flow pat-
lead to product adhesion and buildup at the nozzle
tern inside the drying chamber, especially near
tip that, in severe cases, can obstruct the atomiza-
the nozzle (in the cocurrent flow spray driers).
tion nozzle.
They are located just before the entrance of the
The other type of disperser aims to provide
drying gas into the chamber and impact the dry-
streamlined flow parallel to the chamber’s vertical
ing efficiency, particle residence time, and drop-
axis. The chambers associated with this disperser
let/particle collisions. It is crucial to have a
type are usually taller than those coupled to rota-
uniform distribution of the drying gas flow inside
tional dispersers due to the flow pattern. The main
the chamber to achieve homogeneous drying of
advantage of this disperser type is reduced wall
the droplets. Otherwise, wall deposits can occur
deposit formation. In terms of design, the most
leading to lower yields.
common streamlined air dispersers are based on
Gas disperser design has been evolving over
perforated plates or guide vanes. The former can
the years, with recent advances realized by the
be very efficient in providing laminar vertical
application of computational fluid dynamics. The
flow of the drying gas, but application in the
most optimum design depends on the configura-
food and pharmaceutical industry is limited by
tion of the spray drier (cocurrent or
cleaning challenges. The latter is preferred for the available as a function of particle size (Masters
pharmaceutical industry due to its sanitary design 2002). These curves are plots of collection per-
and lower-pressure drop over the disperser centage versus particle diameter enabling the rel-
(Hansen and Ullum 2009). ative comparison between cyclones. These
The chamber dimensions depend on the atom- efficiencies cannot be directly correlated with
izer and air disperser design as described previ- spray-dried products due to differences in true
ously. However, another important feature is the density and omission of agglomeration effects
safety data of the product to be handled, e.g., often encountered with pharmaceutical spray-
explosivity and autoignition. In some extreme dried products. However, there are correlations
cases, the chamber has to be equipped with sup- that enable the correction of these plots that
pressant and/or pressure-relief systems. These achieve very precise estimations of efficiency.
systems are typically preferred, because the con- Particle agglomeration effects are one of the
struction of chambers with high-pressure shock reasons that cyclone efficiencies are often higher
ratings is very expensive. than would be expected from manufacturers’ effi-
ciency curves. A typical spray-drying system in
the pharmaceutical industry is equipped with a
10.2.4 Collection Systems single cyclone. However, several cyclones can
be utilized in parallel or in series to increase
After exiting the drying chamber, the powder collection efficiency. Very high-performance
must be collected by appropriate systems. The cyclones exist for particle sizes of about 5 microns
most appropriate collection system depends on that can yield collection efficiencies greater than
handling requirements and product 95%. These cyclones are particularly well suited
characteristics, e.g., particle size and density. for inhalable products.
There are three main types of collectors based Cyclone efficiency also depends on gas veloc-
on different separation techniques: centrifugal ity as this influences the pressure drop, which is
force (cyclones), filtration (filter bags), and elec- typically between 70 and 250 mm
trostatic precipitators. H2O. Therefore, for a given system, cyclone effi-
ciency may be improved by increasing drying gas
Cyclones flow rate or by adding an extra gas stream before
The most common collection systems used in the cyclone. However, this is limited by the
pharmaceutical spray drying are cyclones. In capacity of the ventilators to handle such
this system, the gas coming from the drying increases in gas flow.
chamber enters tangentially creating a downward The “cutoff” point of a cyclone is the particle
vortex that, through centrifugal forces, sweeps the size at which it can only capture 50% of the
material against the walls. In this way, the dried product particles. This metric is also a straightfor-
particles are collected in the bottom of the ward means of comparing performance. Major
cyclone. At the same time, the gas reverses direc- product losses may occur if there are leaks in the
tion and travels upward to the top of the cyclone bottom of the cyclone or if the base of the cyclone
forming an inner vortex (lower-pressure region). becomes obstructed by product. For the later, the
This upward current leaves the system as exhaust inner vortex carries the product to the exhaust
gas carrying with it entrained particles and sol- leading to potential for significant losses. Systems
vent vapors. Both capital investment and may be implemented to avoid obstruction of the
operating costs are low for cyclones as compared exit orifice such as hammering devices at the
to other collection systems. Moreover, recent discharge.
developments in design have further increased As in the drying chamber, some products tend
their efficiency (Salcedo and Pinho 2002). to stick to the walls of the cyclone. This aspect
Cyclone efficiency curves constructed with can be especially critical with thermoplastic
data from specific products are generally products and particularly in the case of solid
Table 10.3 Collection efficiencies for different collection systems

Particle size 10 μ 5μ 1μ
High-efficiency cyclones 95–100% 90–95% 10%
Filter bags 100% 100% 99
Electrostatic precipitators 100 99 86
dispersions when the drying temperature is rela- are required, Teflon filters are preferred; however,
tively close to the glass transition temperature. It they are also more expensive.
can also be an issue when relative solvent satura- Filter bags require greater capital investment
tion levels in the drying gas are high. and maintenance costs than cyclones (some
systems include 24 filters per bag house), but
the balance is favorable for high-cost products if
Filter Bags
they provide a significantly better product yield.
Filter bags are typically incorporated into spray-
Even if not the primary collection system, filter
drying systems downstream of the cyclone to
bags are widely used downstream of cyclones to
filter out entrained particles in the exhaust gas.
remove airborne particles from the exhaust of
This prevents powder from exiting into the envi-
laboratorial-, pilot-, and industrial-scale units.
ronment or contaminating auxiliary equipment
(ventilators, condensers, heaters, etc.). Filter
bags consist of a number of bags (depending on Electrostatic Precipitators
the drying gas flow and bag permeability) Electrostatic precipitators create an electrical field
installed inside a rigid container. To facilitate through which the spray-dried powder passes,
continuous operation, a dry clean-in-place system thereby charging the particles which subsequently
is usually installed. This consists of a pulsed flow adhere to oppositely charged plates. These
of a pressurized gas (same type and grade as systems are found in some laboratorial units to
drying gas) that dislodges the powder entrained collect very small particles (nanoparticles).
in the bags. Depending on the product being Despite their high recovery efficiencies, even for
spray dried and the material of construction of very small particles (particles in the nanometer
the bags, the efficiencies of these cleaning size range included), electrostatic precipitators are
systems vary. rarely used in industrial-scale pharmaceutical
In terms of powder collection efficiency, filter spray drying. This might be explained by the
bags can collect finer particles than cyclones high capital and operating costs and by the fact
irrespective of particle density. However, for that the previously described systems provide
some products, namely, those intended for inha- high collection efficiencies in a great majority of
lation or parenteral administration, its use as a cases (Table 10.3).
primary collection system may not be adequate.
In these cases, special types of bag materials have
to be tested and assessed for the release of fine 10.2.5 Closed-Loop Versus Open-Loop
particles. Systems
The selection of construction materials for the
filters depends on the chemical properties of the The most common spray-drying configuration
product being collected and the outlet drying found in industry is the open-loop system. In
temperature. Typically, filter bags are constructed this system, the drying gas is used only once
of polyester fibers which have a maximum and is then exhausted to the atmosphere after
operating temperature of approximately 130 C appropriate posttreatment. However, in the phar-
and are highly resistant to acids, bases, and micro- maceutical industry and especially with the pro-
bial growth. When higher operating temperatures duction of solid dispersions, the closed-loop
configuration (drying gas is reheated after solvent
removal and reintroduced in the drying chamber) 10.3 Equipment by Scale

is most typical. In these systems, flammable
organic solvents are often used; therefore, an 10.3.1 Laboratory-Scale Equipment
inert gas (nitrogen) must be used as the drying
medium. In the closed-loop layout, gas consump- Laboratorial-scale spray driers are useful for pro-
tion is minimal as compared to the open-loop ducing small quantities of prototype formulations
configuration and is, therefore, more cost effec- in early-stage development. They can process
tive in regular operation despite greater initial very small quantities of solution (as low as
investments. 0.25 ml) with relatively high yield. Since spray
The closed-loop configuration requires drying is a continuous process, all units provide
removal of solvent vapors from the drying gas flexibility with respect to batch size. Therefore,
before it can be reintroduced into the drying pro- these small units are also capable of spray dying
cess. Several systems can be used for this opera- much greater quantities of feed by running con-
tion, with the most common being condensers tinuously for extended durations. Another unique
with adjustable temperatures. In the condenser, feature of these systems is that the drying cham-
the drying gas is cooled to a target temperature ber and cyclones are typically constructed of
at which point the gas becomes saturated with the glass, thus enabling visualization of the drying
solvent. The condensed solvent is then removed process.
and the drying gas is reheated and reintroduced in In Table 10.4, a list of the most common
the drying chamber. In closed-loop systems, the laboratorial-scale equipment found in the market
amount of solvent being recirculated is not negli- is presented. The maximum drying gas flow rate
gible, especially in the case of organics. This must is approximately 30 m3/h which enables the dry-
be considered when generating the heat and mass ing of about 1 L/h of water (higher for organic
balances. solvents).
The saturation level of drying gas for a given The standard atomization system for these
feed flow and drying temperature is always higher units is the two-fluid nozzle, but in some models,
in a closed-loop configuration (can be as high as ultrasonic (Mini Spray drier B-290, 4 M8-TriX),
50% relative saturation at the outlet of the drying piezoelectric (Nano Spray Dryer B-90), or mono-
chamber). Thus, the residual solvent level in the disperse nozzles (4 M8-TriX) can also be found.
spray-dried solids is also higher in these units. A Another important feature of these units is the
plot illustrating the influence of solvent content in collection system. Cyclones are the most com-
the drying gas on residual solvent in the product is mon collection systems used. Some
provided in Fig. 10.8. manufacturers offer the choice of different
Fig. 10.8 Product residual 3.0

% Relative Saturation Drying
solvent as a function of
drying gas relative 2.5
saturation
2.0
Gas
1.5
1.0
0.5
0.0
2.0 4.0 6.0 8.0 10.0 12.0
% Product Residual Solvent
Table 10.4 Laboratory-scale spray dryers

Evaporation Possible
Drying gas capacity (liter particle-size Smallest Closed Collection
Designation Manufacturer flow (kg/h) H2O/h) diameter sample loop system
Mini spray Büchi 40 1.0 1–25 μm 30 ml Yes Cyclone
dryer B-290
Nano spray Büchi 12 0.2 0.3–5 μm 1 ml Yes Electrostatic
dryer B-90
SD micro™ Niro 30 1.0 – 100 ml No Cyclone
MicraSpray Anhydro 30 1.0 – 100 ml No Cyclone
Spray dryer Procept 36 1.0 1–100 μm 0.25 ml Yes Cyclone
4 M8-TriX
GS-310 Yamato 40 1.3 – – Yes Cyclone
cyclones providing a range of collection 10.3.2 Pilot-Scale Equipment

efficiencies. The Nano Spray Dryer B-90 is the
only unit capable of producing particles in the Pilot-scale spray driers are suitable for batch sizes
nanometer range. In order to provide acceptable from hundreds of grams up to about 20 kg. These
product yields for these fine particles, the system units are very similar to production-scale equip-
is equipped with an electrostatic particle collector, ment with regard to configuration and material of
enabling recovery yields of up to 90%. construction, i.e., they are composed of stainless
Process scale-up from laboratorial-scale to steel as opposed to lab-scale, glass units. These
pilot- and large-scale spray driers has been stud- units consume significant drying gas and there-
ied by several authors. The main challenge in fore should be operated in the closed-loop config-
scale-up is related to particle size. The small uration when using nitrogen (required for drying
dimensions of the drying chambers limit the resi- organic solvents).
dence time distribution. Therefore, the atomized Examples of pilot-scale spray driers available
droplets must be small in order to complete the in the market are shown in Table 10.5. The
drying process prior to exiting the chamber or suppliers offer many possible system
impacting the chamber walls. Typically, the dry- configurations (closed loop, open loop, mirror
ing gas flow pattern in the smaller units is not polishing, aseptic versions, clean in place, etc.)
laminar which causes the droplets/particles to as well as different atomization systems. Pressure
collide with the chamber walls. This further nozzles, as mentioned previously, produce a more
reduces the duration for drying of the atomized homogenous particle-size distribution and higher-
feed. The Nano Spray Dryer B-90 and 4 M8-TriX density powders. Therefore, those units that
units claim to provide laminar flow patterns enable the use of this atomization system are
which is a great advantage not only in terms of preferred for spray-dried products used in oral
yield but also with regarding to utilizing the full dosage forms.
chamber length to enable production of larger Evaporation capacity of these units depends on
product particles. However, only the 4 M8-TriX the drying gas flow and maximum inlet tempera-
spray drier, when coupled with an ultrasonic noz- ture. The Niro Mobile Minor has a higher evapo-
zle, claims to be capable of producing particles in ration capacity due to the maximum inlet
the same size range as large-scale units (up to temperature (350 C) when compared to the
100 μm). MicraSpray 75 (maximum temperature 200 C).
Table 10.5 Pilot-scale spray dryers

Evaporation
Drying gas capacity (liter H2O/ Maximum inlet Atomization Closed
Designation Manufacturer flow (kg/h) h) temperature ( C) systems loop
Mobile Niro 80 7 350 Rotary Yes
minor Two-fluid
Pressure
MicraSpray Anhydro 75 2 200 Two-fluid Yes
75
MicraSpray Anhydro 150 14 350 Rotary Yes
150 Two-fluid
Table 10.6 Spray-drying evaporation capacity for the most common solvents
Drying gas flow (kg/h) Water (kg/h) Ethanol (kg/h) Acetone (kg/h) Dichloromethane (kg/h)
360 18 45 75 118
630 35 75 130 200
1250 70 160 275 400
4000 220 575 900 1300
However, in the pharmaceutical industry, the use

10.4 Feeds
of such high temperature is not recommended in
most cases. For a given temperature, the evapora-
A wide range of feed types can be spray dried.
tion capacity depends on the drying gas flow,
They can be solutions, suspensions, emulsions, or
which is very similar for these two units.
mixtures thereof. In the case of solid dispersions,
a solution of active ingredient and excipient
(s) dissolved in a common solvent system is the
10.3.3 Production-Scale Equipment most typical feed. Some important parameters to
take into account for feed systems are:
Production-scale equipment have all the features
• Solids load – This is primarily governed by the
as the pilot-scale units; however, given their
solubility of the constituents in the solvent
larger size, they are more flexible in terms of
system as well as the viscosity of the resulting
manipulation of product properties, e.g., particle
solution.
size and density. Taking advantage of the larger
• Viscosity – Very high-viscosity solutions can
dimensions of the drying chamber, the use of
be spray dried, but the atomization system
pressure nozzles, instead of two-fluid or rotary
must be properly selected (see Atomization).
nozzles, is a standard at this scale.
• Solvent system – All volatile solvents are suit-
Spray-drier manufacturers have specialized
able for spray drying. The use of inert drying
their equipment to the pharmaceutical industry
gas and intrinsically safe equipment enables
creating, in some cases, specific pharma lines of
safe drying of organic solvents.
equipment. Such equipment is available in a wide
range of production scales from 360 to 4000 kg/
h of nominal drying gas capacity, enabling the The typical solids load in a spray-drying pro-
evaporation of more than 200 kg/h of water. In cess is between 5 and 50% (w/w). The higher the
Table 10.6, the evaporation capacity for various solids load, the more cost-effective will be the
solvents at varying production-scale drying gas process because spray-drying capacity is mainly
flow rates is presented. limited by solvent evaporation capacity. The
other potential benefit of using higher by the viscosity of the solution which can
concentrations in the feed is the possibility to adversely effect the atomization process.
produce larger particles that have better flow Water is the solvent of choice whenever possi-
properties. ble for a spray-drying process because it is an
Regarding viscosity, it is chiefly determined environmentally friendly solvent, does not
by the excipients (polymers, cyclodextrins, etc.) require the use of inertized systems, and upon
used in solid dispersions. Many of them exhibit evaporation requires only a filtration step before
non-Newtonian behavior, namely, shear thinning being exhausted to the atmosphere. Also, open-
and/or extensional thickening properties (Porfirio loop systems are suitable for drying of water-
et al. 2021). This impacts the atomization process based feeds even for large-scale production, e.g.,
and may introduce some challenges in this regard. milk powder. However, for obvious reasons it is
The most common solvents used are low- of very limited use for the production of solid
boiling-point solvents such as water, ethanol, dispersions by spray drying being only seldom
methanol, and acetone. Other solvents, such as employed as a cosolvent in binary or ternary
tetrahydrofuran and dichloromethane, may also solvent systems.
be used, but special handling or waste-treatment All International Conference on
requisites must be taken into consideration. In the Harmonization (ICH) class II and III solvents
next section, common solvents utilized in spray (see Sect. 10.5.7) are suitable for use in the
drying are discussed in detail. spray-drying process. The most volatile are pre-
ferred due to the higher throughputs that can be
achieved and because they enable the use of lower
drying temperatures. The latter aspect is impor-
10.5 Solvents
tant when processing low glass transition
products as it results in higher process yields,
A frequent challenge in developing amorphous
less potential for product sticking to equipment
solid dispersions by spray drying is to find a
walls, and less potential for partial drug
suitable solvent system that dissolves, in an
crystallization.
appreciable amount, both drug and polymer. In
The most common solvents used in the spray
fact, many poorly water-soluble drugs also have
drying of solid dispersions include alcohols (e.g.,
very limited solubility in organic solvents. There-
methanol, ethanol, and isopropanol), ketones
fore, binary and ternary solvent mixtures are often
(e.g., acetone and methyl ethyl ketone),
required to achieve solution concentrations com-
dichloromethane, tetrahydrofuran, and ethyl ace-
patible with an industrial spray-drying process. It
tate. At the end of the spray-drying process, the
is important to emphasize that spray-drying
residual solvent content of the powder is almost
capacity is measured in terms of solvent evapora-
always above the admissible limits set by the
tion and independent of solids throughput. Thus,
ICH. Therefore, a secondary drying step is often
the higher the solids concentration, the more eco-
required to decrease the solvent content. How-
nomical will be the process since higher feed
ever, this must be assessed in an early stage of
flows can be used resulting in greater solids
development, because residual solvent removal
dried per unit time.
can be very difficult to achieve and, in some
In the case of solid dispersions, the solution
cases, is impossible to remove to the required
concentrations range from 5 and 50% (w/w), with
safety limits. In these instances, a change of sol-
a substantial majority of processes being effec-
vent would be needed. The various solvents and
tively run between 10 and 20% (w/w). The upper
their respective enthalpies of vaporization and
limit of concentration is dictated not only by the
boiling points are tabulated in Table 10.7.
components’ solubilities but also, in some cases,
Table 10.7 Vaporization enthalpies and boiling points for most common solvents
Solvent ΔH vap (kJ/mol) Boiling point ( C)
Water 40.7 100.0
Dichloromethane 28.0 39.8
Acetone 29.1 56.2
Methyl ethyl ketone 31.2 79.64
Methanol 35.3 65.0
Ethanol 38.7 78.5
Tetrahydrofuran 29.6 67.0
Ethyl acetate 31.9 77.1
10.5.1 Alcohols 10.5.3 Dichloromethane
Alcohols are polar protic solvents widely used in Dichloromethane (DCM) is a very complemen-
spray drying. Methanol and ethanol are the most tary solvent to alcohols and ketones due to its lack
used alcohols for this purpose due to their low of oxygen atoms. It provides good solubilization
boiling points and good solubilization properties. properties for certain drugs and polymers, and it
Relatively low drying temperatures (outlet enables the use of very low drying temperatures
temperatures of 30–40 C) can be used to evapo- (outlet temperatures around 20 C may be used)
rate these solvents in the spray-drying process. due to its low boiling point (39.8 C). Also,
Also, given that their freezing point is very low according to its low boiling point and evaporation
(<–98 C), the condenser temperature in the enthalpy, spray drying of DCM is highly efficient,
closed-cycle spray-drying units can be very low thus making it a favorable solvent for processing
(<0 C) and limited only by the capabilities of the large quantities. Its evaporation capacity is seven
cooling system. times greater than that of water for a given spray-
drying unit, i.e., a unit that can dry 70 kg/h of
water is able to dry approximately 400 kg/h of
DCM. However, the requirement for very low
10.5.2 Ketones
residual levels in pharmaceutical products (see
Table 10.8) and associated environmental hazards
Ketones are polar aprotic solvents that are also
may limit its use.
widely used for spray drying, namely, acetone
and methyl ethyl ketone. Given the good solubil-
ity of HPMCAS in acetone, this solvent is widely
used for the production of solid dispersions by 10.5.4 Tetrahydrofuran
spray drying. For instance, torcetrapib referred to
in section “Torcetrapib” was spray dried from a As insolubility becomes a greater issue in drug
solution of HPMCAS in acetone. Relatively low development, tetrahydrofuran (THF) is becoming
drying temperatures may be used (outlet a solvent of choice for spray drying. In terms of
temperatures of 30 C are common for this sol- volatility, it is very similar to acetone (vaporiza-
vent) which enables the production of relatively tion enthalpy of 29.6 kJ/mol and 19.1 kJ/mol,
dense powders. In terms of spray drying, methyl respectively), although it has a higher boiling
ethyl ketone has similar properties to acetone but point (67.0 C vs. 56.2 C). Nevertheless, it read-
has a higher boiling point and can provide higher ily forms explosive peroxides; thus, appropriate
solubilities with some drugs. Similarly to safety measures must be in place to process this
alcohols, very low temperatures in the drying solvent at an industrial scale. Antioxidants
condenser are acceptable with ketones. (stabilizers) are often added to THF to prevent
Table 10.8 Allowable limits of class 1 solvents in pharmaceutical products.

Solvent Allowable concentration (ppm) Issue
Benzene 2 Carcinogen
Carbon tetrachloride 4 Toxic/environmental
1,2-Dichloroethane 5 Toxic
1,1-Dichloroethane 8 Toxic
1,1,1-Trichloroethane 1500 Environmental
Adapted from ICH (2011)
peroxides formation in order to further reduce the they usually compensate in process robustness
explosion hazard. These antioxidant stabilizers, and throughput.
e.g., BHT, have high boiling points and thus
become concentrated in the final spray-dried
product. Removal is sometimes impossible; 10.5.6 Aqueous Systems
hence, their toxicity must be balanced with the
product daily dose. In cases where poorly soluble drugs are weak
Due to its miscibility with water, tetrahydrofu- acids or bases, aqueous solvents can occasionally
ran is often used in binary THF/water solvent be used for spray drying. Good candidates for this
systems. This system is easily spray dried, but type of aqueous drying are drugs having reason-
care should be taken in optimizing drying able solubility in acidic media (weak bases) or
temperatures as many times it produces alkaline buffer systems (weak acids). In these
low-density, broken particles. In some cases, cases, the drug and excipients are co-dissolved
THF can be difficult to remove from the product in the acidic or buffered system and spray dried
to levels permissible by ICH guidelines. There- from solution per normal operation. Common
fore, it is recommended to evaluate a secondary aqueous solvents for this application are dilute
drying step at an early stage of development. hydrochloric acid and mildly alkaline buffer
systems.
The key benefit to using aqueous solvent
10.5.5 Mixed Organic Solvent Systems systems, aside from the obvious toxicity, environ-
mental, and cost benefits associated with avoiding
It is not uncommon to find that a mixture of organic solvents, is that air can be used as the
organic solvents must be used to achieve solubil- drying gas. This aspect further improves the cost
ity of the drug and polymer at the desired solids efficiency of the spray-drying operation, as men-
concentration. In principle, all combinations are tioned previously. The potential disadvantages of
suitable for spray drying as long as the solvents using acid or alkaline buffer for spray drying are
are miscible (specifically required for amorphous instability of the drug and/or excipients in acidic
solid dispersions). Binary and ternary diagrams or basic solutions and detrimental effects on the
can be generated with common process software formulation (chemical stability or release profile)
to predict miscibility of complex systems. How- resulting from acid or buffer salt content in the
ever, for the sake of process robustness, formulation after drying. Despite these
processing close to those limits should be disadvantages, aqueous-based solvent systems
avoided, and variations in temperature should be can be used for spray drying of poorly water-
accounted for in the design of the system. Binary soluble drugs. Being that these types of weakly
solvent systems, such as acetone/DCM, ethanol/ basic or acidic drugs have solubility within the
water, THF/water, and methanol/acetone, have physiological pH range, the primary benefit of
been widely reported, to mention a few. Usually, spray-dried dispersions of these molecules is
these systems are more complex to optimize, but more related to reducing pharmacokinetic
variability rather than improving exposure.
10.5.7 ICH Guidelines on Residual Although these solvents are considered less harm-
Organic Solvents ful than class 1 solvents, exposures should be
limited to avoid toxic effects. Consequently,
In Impurities: Guideline for Residual Solvents, class 2 solvents are limited in concentration in
Q3C(R5), the ICH provides recommendations pharmaceutical products. The allowable limits of
for acceptable amounts of residual solvents in various class 2 solvents in pharmaceutical
pharmaceuticals for patient safety (ICH 2011). products are listed in Table 10.9.
The guideline recommends the use of less toxic
solvents and sets the standards for levels of resid-
Class 3 Solvents
ual solvents that are considered toxicologically
Class 3 solvents are described by the ICH
acceptable. The guideline defines residual
guidance as: “Solvents with low toxic potential
solvents in pharmaceutical products as: “organic
to man; no health-based exposure limit is needed”
volatile chemicals that are used or produced in the
(ICH 2011). This class contains no solvents that
manufacture of drug substances or excipients, or
are known health hazards to humans at typical
in the preparation of drug products. . .that are not
levels in pharmaceutical products. However, data
completely removed by practical manufacturing
concerning toxicity from long-term exposure are
techniques” (ICH 2011). The guidance states that
lacking for many of these solvents. The permitted
because residual solvents provide no therapeutic
daily exposure of class 3 solvents is 50 mg or
effect, they should be removed from the product
more. These solvents should be used preferen-
to meet the requirements of all quality-based
tially over class 1 and 2 solvents whenever possi-
specifications or practices. Further, drug products
ble. The acceptable concentration of a class
should contain no more residual solvent than can
3 solvent in a drug product should be dictated
be supported by safety data. This guideline
by GMP and/or product quality standards
classifies organic solvents into three classes
(Table 10.10).
which will be discussed below. The reader should
consider that the following sections are merely
brief summaries of aspects of the ICH guideline.
The actual guideline should be consulted when 10.6 Residual Solvent Content
evaluating a drug product intended for and Secondary Drying
consumption.
Although spray drying is a highly efficient pro-
cess for separating solids from a solvent system, it
Class 1 Solvents
is typical for some amount of residual solvent to
Class 1 solvents are “known human carcinogens,
remain in the dried product. The residual solvent
strongly suspected human carcinogens, and envi-
amount can vary in the range of a few parts per
ronmental hazards” (ICH 2011). They should be
billion up to several percent of the dry powder
avoided in the production of drug products unless
weight depending on the solvent(s), solid(s), and
there is strong justification based on a risk to
processing conditions. In addition to the potential
benefit assessment. The concentrations of class
toxicity issues discussed previously, residual
1 solvents in drug products should be limited
solvents can also adversely impact product qual-
according to Table 10.8.
ity. In this section, quality concerns posed by
excessive residual solvent are discussed along
Class 2 Solvents with methods for secondary drying of spray-
Class 2 solvents are described as: “non-genotoxic dried materials to reduce or eliminate residual
animal carcinogens or possible causative agents solvent.
of other irreversible toxicity such as neurotoxicity
or teratogenicity. . . Solvents suspected of other
significant but reversible toxicities” (ICH 2011).
Table 10.9 Allowable limits of class 2 solvents in pharmaceutical products

Solvent Permitted daily exposure (mg/day) Concentration limit (PPM)
Acetonitrile 4.1 410
Chlorobenzene 3.6 360
Chloroform 0.6 60
Cumene 0.7 70
Cyclohexane 38.8 3880
1,2-Dichloroethane 18.7 1870
Dichloromethane 6.0 600
1,2-Dimethoxyethane 1.0 100
N,N-Dimethylacetamide 10.9 1090
N,N-Dimethylformamide 8.8 880
1,4-Dioxane 3.8 380
2-Ethoxyethanol 1.6 160
Ethylene glycol 6.2 620
Formamide 2.2 220
Hexane 2.9 290
Methanol 30.0 3000
2-Methoxyethanol 0.5 50
Methyl butyl ketone 0.5 50
Methylcyclohexane 11.8 1180
N-Methylpyrrolidone 5.3 530
Nitromethane 0.5 50
Pyridine 2.0 200
Sulfolane 1.6 160
Tetrahydrofuran 7.2 720
Tetralin 1.0 100
Toluene 8.9 890
1,1,2-Trichloroethene 0.8 80
Xylene 21.7 2170
Table 10.10 Class 3 solvents; daily permitted exposure >50 mg

Acetic acid Heptane
Acetone Isobutyl acetate
Anisole Isopropyl acetate
1-butanol Methyl acetate
2-butanol 3-Methyl-1-butanol
Butyl acetate Methyl ethyl ketone
Tert-Butylmethyl ether Methyl isobutyl ketone
Dimethyl sulfoxide 2-Methyl-1-propanol
Ethanol Pentane
Ethyl acetate 1-Pentanol
Ethyl ether 1-propanol
Ethyl formate 2-propanol
Formic acid Propyl acetate
10.6.1 Influence of Residual Solvents specifications for maximum residual solvent

on Product Quality content.
There are several different methods for mea-
The allowable product concentration limits for suring residual solvent content in pharmaceutical
class 1 and 2 solvents are determined according products. A detailed discussion of these methods
to safety and are sufficiently low not to pose any is beyond the scope of this chapter. For a compre-
product quality issues. For Class 3 solvents, the hensive review of these methods, the reader is
allowable product concentrations are not trivial referred to the review article by Witschi and
and can have a significant influence on product Doelker (1997).
quality. Residual solvents can present product
quality issues such as color changes, odors, chem-
ical instability, and physical instability (Witschi 10.6.2 Secondary Drying
and Doelker 1997). The issue of physical insta-
bility is particularly critical for amorphous spray- Residual solvent levels must be controlled to
dried systems, i.e., amorphous spray-dried disper- ensure safety and product quality. Secondary dry-
sion (ASDD) of a drug in an excipient (usually a ing of spray-dried product is often required to
polymer) carrier. reduce residual solvent levels to meet product
It has been demonstrated in the scientific liter- safety and quality specifications. The various
ature that organic solvents tend to be efficient equipment choices for secondary drying and
polymer plasticizers (Wicks 1986). As such, their applicability at different scales will be
they can significantly reduce the Tg of an amor- discussed in the following section.
phous drug–polymer composite at low
concentrations. The correlation between compos- Tray Drying
ite Tg and the physical stability of drugs in an Perhaps the simplest method for removing resid-
amorphous solid dispersion has been well ual solvent from spray-dried powder is tray dry-
established in the pharmaceutical literature ing. In practice, tray drying is accomplished by
(Matsumoto and Zografi 1999; Yoshioka et al. spreading the bulk powder on trays and drying in
1994). An ASDD-based product containing an oven at elevated temperatures for a time inter-
excessive residual solvent may therefore be at val sufficient to reduce the residual solvent to the
greater risk for recrystallization of the drug on specified concentration. During the process, loss
storage. Hence, it is understood that reducing on drying measurements are performed intermit-
residual solvent content in ASDDs is critical not tently to estimate the drying endpoint, and the
only for safety reasons but for product quality final residual solvent determination is performed
as well. by a suitable method, such as gas chromatogra-
In order to understand the influence of residual phy (GC).
solvent on product stability, an experimental Tray drying in a typical convection oven
design should be executed to evaluate the physi- employs both convection and conduction
cal stability of the ASDD with varying residual methods to dry the powder mass. In a typical
solvent content. The ASDD product should be convection oven, forced hot air is used to transfer
produced at conditions to generate variations in heat to the drying trays and also directly to the
residual solvent content. Powders with a range of powder bed. Heat is then exchanged by conduc-
solvent content below the limits set according to tion from the heated surfaces of the drying trays to
toxicity should be stored at accelerated conditions the powder mass. By these heat exchange pro-
and monitored periodically for the appearance of cesses, residual solvent contained in the spray-
drug precipitation/crystallization. This study will dried powder is vaporized and then transported
allow for the establishment of product quality from the oven to the exhaust. For this type of
system, both solid bottom and perforated trays
can be used. Paper or a suitable screen is often
placed over the bottom of a perforated tray to is governed by the inlet air temperature, flow rate,
support the powder bed. This allows air to flow solvent vapor content, and distribution in the bed
through the support and into the powder bed, (Frake et al. 1997). Heat transfer from the inlet
improving the efficiency of the convective drying gas to the particle over time causes residual sol-
component. vent bound to the particle surface to enter the
Vacuum ovens differ from convection ovens vapor state. Once in a gaseous state, residual
in that the heating method is accomplished strictly liquid becomes entrained in the drying gas stream
by conduction as there is no circulating air for and is carried out to the exhaust. Solvent from the
convection (Mujumdar 2007). Organic vapor interior of the particle can then migrate to the
evolved from the powder is transported from the surface by diffusion, and the surface-drying pro-
oven via the vacuum pump. Vacuum ovens offer cess is repeated. Drying by this mechanism
the advantage drying at lower temperatures due to continues until the solvent contained in the bulk
solvent boiling point reduction. The reduction of powder bed is reduced to meet the product
drying temperature is particularly advantageous specifications. Intermittent loss on drying or GC
for heat-sensitive actives. measurements should be used to monitor residual
Tray drying is a stagnant powder bed method; solvent levels and determine the drying endpoint.
thus, drying efficiency is inversely related to bed Fluidized bed drying is an efficient process for
depth (Carstensen and Zoglio 1982). Tray drying removing residual solvent from spray-dried pow-
is therefore most applicable to small batch sizes der. It is particularly useful for large masses of
where the powder can be spread thinly over the powder where tray drying is not practical.
tray surfaces. For a fixed oven size and number of Because particle size of spray-dried powders is
trays (fixed drying surface area), bed depth and typically fine, fluid bed drying can be problematic
drying time increase with increasing batch size with respect to powder loss in the exhaust and
(powder volume). For larger spray-dried batches, clogging of filters. This can usually be overcome
agitated bed methods such as those described by proper selection of filter bags and utilization of
below are recommended. blowback to dislodge powder from the filters.
Also, proper control of fluidization will limit par-
Fluid Bed Drying ticle entrainment in the drying gas that causes
One of the most commonly employed methods filter clogging. For cases where powder loss or
for removing residual solvent from spray-dried filter clogging is excessive, alternative drying
powder is fluid bed drying. By this method, a methods such as tumbling or agitated bed drying
powder bed is acted upon by an upward flowing are recommended.
gas stream with a flow rate sufficient to suspend
the particles without entrainment, thereby Rotary and Agitated Bed Driers
transforming the stagnant powder bed into a flu- Other equipment options for removing residual
idlike state (Parikh and Mogavero 2005). The solvent from spray-dried powder include agitated
mobility imparted on the particles by the air bed driers and rotary drum driers. Agitated bed
stream allows for greater surface contact between driers utilize mechanical agitation to stir the pow-
the solids and the drying gas, thus improving the der bed and improve drying efficiency, as
efficiency of heat and mass transfer. Conse- opposed to air fluidization. In a typical drier of
quently, drying time by this method is signifi- this type, an impeller distributes the powder bed
cantly shorter than by stagnant bed methods. within the drying vessel to provide even product
Solvent is removed from a particle by a contact with the heated surfaces of the chamber.
two-step cycle consisting of (1) surface evapora- Heat transfer in these systems is therefore
tion followed by (2) diffusion of solvent from the governed by conduction. Agitated bed driers are
interior to the surface of a particle (Mujumdar typically operated under vacuum to further
2007; Wildfong et al. 2002). Evaporation of sol- improve drying efficiency. These driers therefore
vent from a particle surface in a fluid bed system offer the benefits of vacuum oven drying with the
additional advantage of improved heat conduc- Therefore, the scale-up of spray drying has
tion. An EKATO vertical dryer is an example of been considered a resource-intensive task based
an agitated bed drier commonly used for second- on empirical activities dependent on feedstock
ary drying of spray-dried powders. properties, nonmeasured parameters, and phe-
Rotary driers employ the use of a rotating nomena that are mathematically complex (Gaspar
heated drying chamber to constantly tumble the et al. 2014). Nevertheless, model-based
powder bed for enhanced heat transfer. The drum approaches have been applied to the current pro-
is typically equipped with baffles to evenly dis- cess development methodologies founded on
tribute the powder along the walls of the chamber quality by design principles in which the process
and improve drying efficiency. These systems can understanding and modeling tools are used to
be operated under vacuum or purged with heated support scale-up.
drying gas. Several manufacturers offer rotary Dobry et al. (2009), Vicente (2014), Singh and
drier systems for various production scales. Van dun Mooter (2016), Al-Khattawi et al.
(2018), and Poozesh and Bilgili (2019), among
other authors, discussed and presented the major
challenges of scale-up in which the major
10.7 Spray-Drying Process
concerns are in atomization and drying kinetics.
Development
The importance of model-based approaches to
support is mentioned in all studies. In such
The scale-up is a constant need during drug devel-
works, the spray-drying scale-up methodologies
opment: from the lab scale to the final scale of the
are presented with general considerations, as
commercial product passing by intermediate
follows:
scales to support clinical supplies. The scale-up
aims to maintain the product attributes when the • Maintaining similar feeding system and fluid
process is transferred between different scales of properties as concentration and solvent(s).
production. Nevertheless, the differences between • Match the droplet size distribution.
scales are a challenge while scaling up spray- • Keep a similar drying process, i.e., evapora-
dried products. The differences in scales are sev- tion rate, HMT, Peclet number, and/or relative
eral: geometry, gas distribution and therefore saturation.
gas/spray interaction (and therefore droplet coa-
lescence), heat loss, and increase of drying capac- Nevertheless, it is referred to the difference
ity and throughput, among others. These features between the scale operations, especially from
may impact on the product quality in distinct lab scale to large production. The differences
perspectives: pointed are:
1. Chemical and physical behavior due to the • Atomization – the application of some
exposition at different temperature profiles atomizers does not support all scale of produc-
and evaporation profiles, which may result in tion. For instance, ultrasonic nozzle is not
distinct purity profiles, surface composition, applicable for large scale of production, and
and amorphous state. pressure nozzles are not easily implemented at
2. Particle characterization, viz., size and density low feed flow rate (i.e., lab scale).
impact by either atomization, gas/spray inter- • Chamber design – chamber length and geom-
action, or different drying kinetics. etry impact drying capabilities. Smaller scales
3. Powder flowability, which may be affected by are not able to dry larger droplets.
both the particle properties and morphology of • Gas exhaust – Large scale of production
the particles. operates normally in closed-loop system.
• Gas dispersion – gas distribution efficiency decrease the drying rate producing particles with
and interaction between gas and spray. distinguished attributes.
The mixing efficiency is increased by chang-
ing the operating conditions, mainly, gas and
liquid flow rates, spray angle, and process
10.7.1 Scale-Up Consideration
temperatures. CFD simulations are used to under-
for Atomization
stand and set the process conditions when the
decrease of mixing efficiency is severe. Besides
The change of production scale is related to the
CFD simulations, correlations and scale-
increase of process throughput and, consequently,
independent correlations are used based on
the increase of flow rates. Hence, the use of the
experimental data.
same nozzle and atomization conditions of the
In the scale-up case, this dimensionless param-
previous equipment cannot be applied – higher
eter can be simplified once the composition (and
feed flow rate requires higher energy to produce
consequently diffusion) and droplet size remain
the same droplet size. In the case of a two-fluid
constant. The drying mixing length (or spray
nozzle, a similar droplet size can be found by the
length) is determined experimentally, by CFD or
gas stream flow rate manipulation (Hede et al.
first principles models.
2008). For the pressure nozzle, the geometry
Other dimensionless number may be applied
and design (for instance, orifice diameter) must
to support scale-up and ensure chemical homoge-
be selected to target the droplet size of a smaller
neity and amorphous state of the product.
scale. The nozzle design and operating conditions
Vehring (2008) and other authors suggested
can be set using correlations (as explained above)
Peclet number as a descriptor of the particle for-
or by experimental measuring using light diffrac-
mation discussing its correlation to the final prod-
tion such as a phase Doppler anemometry (Singh
uct attributes such as density and morphology.
and Van den Mooter 2016).
Peclet number, which stands for the ratio of evap-
Besides the considerations to target the droplet
oration rate and diffusion of the solute in solution,
size, operational problems related to atomization
is referred to as an important parameter to
can occur in large scale such as bearding.
describe the drying kinetics and mechanism
Bearding consists in product buildup on the noz-
(Singh and Van den Mooter 2016; Vehring
zle discharge and around it (Gaspar et al. 2014;
2008). In cases with small Peclet number, the
Vig and Morgen 2017). This problem interferes in
evaporation rate is slower than the diffusion. Con-
the spray quality and can clog the nozzle. Engi-
sequently, the migration of the solute from the
neering solutions are available for this case
surface to the droplet core is promoted while the
(Vicente 2014).
drop reduces in size. The saturation stage is then
delayed producing denser particles. In the oppo-
site direction, high Peclet number represents a fast
10.7.2 Scale-Up Considerations evaporation rate comparing with the diffusion
for Droplet Drying rate. Therefore, the solute has no time to migrate
to the droplet center, and the saturation is reached
Concerning the effect of scale-up on the droplet quickly at the surface. Light and hollow particles
drying, an important aspect is the increase of are expected at these conditions due to the early
droplets in the spray plume due to the increase formation of surface layer.
of process throughput while maintaining the Quality attributes may be controlled using
droplet size (Vig and Morgen 2017). Conse- Peclet number (ratio of drying kinetics to mass
quently, the mixing efficiency between the drop- diffusivity), and the ratio of initial concentration
let and drying gas may be affected. A denser to solubility (ICS) is proposed to control surface
spray plume will increase the resistance to mass enrichment and physical particle properties (Hoe
and heat transfer at the droplets and, therefore, et al. 2014; Poozesh et al. 2017). Considering a
scale-up with the same ICS (i.e., using the same downstream processing efficiency by eliminating
liquid formulation), a higher Peclet number the need for a subsequent densification step,
implies to have a process controlled by the evap- thereby enabling direct tablet compression or cap-
oration rate. Considering the difference in diffu- sule filling (Bittorf et al. 2010).
sion rate of polymer and drug, it is expected to The FSD concept was patented in 1984 for the
have a skin formation promoted by polymer. This production of agglomerated milk powder with
leads to a polymer-enriched surface promoting better dissolution properties (Pisecky et al.
the increased Tg, as targeted in amorphous solid 1984). In FSD, the solution is atomized at the
dispersions (Poozesh et al. 2017). top of the drying chamber in cocurrent configura-
tion with the drying gas as in conventional spray
drying (see Fig. 10.10); however, the drying gas
exits through the top of the drying chamber.
10.8 Fluidized Spray Drying
Therefore, the flow patterns inside of the drying
chamber are totally different from conventional
The versatility of spray drying also encompasses
spray drying. This gas flow pattern promotes par-
its combination with other technologies.
ticle collisions and consequently agglomeration.
Fluidized spray drying (FSD) is the combination
After exiting the drying chamber through the top,
of fluid bed and spray drying in a single system.
the gas goes to the cyclone. Here, the particles are
By the incorporation of fluid bed drying into the
captured and reintroduced in the drying chamber.
spray-dried system, the product emerges from the
The gas stream coming from the cyclone may be
process with little to no residual solvent content,
introduced in a position concentric to the nozzle
i.e., there is no need for secondary drying. This
or tangentially to the drying chamber at a given
technology also enables the production of
height (Bittorf et al. 2010). This promotes more
agglomerated particles with particle sizes much
effective agglomeration because the particles
larger than in conventional spray drying. A pic-
coming from the cyclone collide with the
ture of an agglomerated particle produced by FSD
droplets/wetter particles nearer the top of the
is provided in Fig. 10.9. Additionally, FSD
chamber. In fact, all particle agglomeration takes
reduces the fines content in the spray-dried prod-
place within the drying chamber.
uct which improves powder flowability. Finally,
As particles agglomerate and become denser,
according to the particle agglomeration achieved
they will move down to the bottom of the drying
by FSD, spray-dried powders with much greater
chamber where the fluidized beds are installed. In
densities can be achieved. This improves
Fig. 10.9 Picture of an

agglomerated particle
produced by FSD.
Reproduced with
permission from Hovione
FarmaCiencia SA
Hot Drying Gas Feed Hot Drying Gas
Gas
Recycling
Wet Unit
Filter Gas
Bag
Drying
Chamber
Cyclone
Fines Return
FB2 FB1 FB3
Fluid Bed Drying Gas

Fluid Bed Drying Gas
Fluid Bed Drying Gas Spray Dried
Product
Fig. 10.10 Fluidized spray drying. Adapted from Bittorf et al. (2010)
Fig. 10.8, a schematic diagram is provided VX-950 solid dispersion obtained by FSD
depicting the fluidized spray-drying setup with (Bittorf et al. 2010).
three-fluid bed driers (FB1, FB2, and FB3). The main limitation of FSD is the absence of
The number of fluidizing beds may vary laboratorial-scale or pilot-scale units that are rep-
between equipment from one to three. Fluidizing resentative of the process at larger scales. This has
bed one (FB1) is directly connected to the drying prevented in many instances the development of
chamber and works as the main fluidizing cham- pharmaceutical products using this technology.
ber. It is the primary bed responsible to perform Recently, efforts have been made to design
particle selection by adjustment of the velocity of laboratorial-scale equipment that can replicate
the fluidizing gas. FB2 may be used to conduct a the FSD process at commercial scale (Assunção
secondary drying step to further decrease the 2011). The results were very promising, and the
residual solvent content, while FB3 may be used powder produced in this laboratorial scale unit
to cool down the product before final discharge. showed much better flowability than the typical
The passage between the fluidized beds is spray-dried powder obtained at the same scale.
achieved based on differential pressures across
the beds. Their temperatures are adjusted
independently. 10.9 Spray Drying for Amorphous
From the previous description, it is obvious Solid Dispersion Systems
that the residence time in the primary drying
process for FSD is much greater compared to The increasing frequency at which poorly water-
traditional spray drying. This enables the use of soluble molecules are entering drug development
even milder drying temperatures, which is bene- pipelines and the problem it poses to the pharma-
ficial for processing of low glass transition tem- ceutical industry have been stated previously in
perature products and for increasing product this book. The value of amorphous systems with
density. Bittorf and coworkers reported particles respect to overcoming intrinsic solubility
up to 259 μm and densities up to 0.32 g/ml for limitations and improving the bioavailability of
insoluble drugs is also covered in detail in this must be taken with post-processing (milling and
text (Chap. 3). More specifically, the value of sieving) to ensure that the product meets the
solid dispersion systems with respect to desired specifications.
stabilizing the amorphous form of a drug and The spray-drying process offers numerous
enhancing its bioavailability is a recurring theme advantages over batch solvent evaporation. First,
woven throughout the chapters. In this section, spray drying is continuous and as such is a more
the application of spray-drying technology for the commercially viable process. Second, atomiza-
production of amorphous solid dispersions tion of the liquid feed results in rapid evaporation
(ASDD) willbe discussed in detail. of the solvent from droplet surfaces and conse-
quently rapid droplet solidification. This is partic-
ularly critical for ASDDs with regard to
10.9.1 Solvent Evaporation Method preventing drug–carrier phase separation that
can lead to physical instability and poor product
There are principally two approaches for produc- performance (Friesen et al. 2008). Third, resi-
ing amorphous solid dispersion systems: (1) ther- dence time in the drying chamber is typically on
mal processing (see Chap. 9) and (2) solvent- the order of seconds; thus, exposure of API to
based processing (Chiou and Riegelman 1971). thermal stresses is minimal. Fourth, spray drying
Basic solvent processing for the production of is simple and easily tunable; hence, the process
amorphous solid dispersion systems involves dis- can be readily optimized to generate the desired
solution of the drug and the excipient carrier(s) in particle properties with limited batch-to-batch
a common solvent followed by rapid separation variability. This includes dry product with parti-
of the solids from the solvent to form, ideally, a cle properties amenable to direct tablet compres-
single-phase composite. This can be accom- sion and/or capsule filling, yet another advantage
plished in one of the two ways: (1) solvent evap- of spray drying. Finally, spray drying can be
oration or (2) induced precipitation by conducted in a wide spectrum of scales, from
introduction of an antisolvent (see Chap. 12). bench to commercial scale, without dramatic
From a process and scalability perspective, the changes to the process. Accordingly, spray drying
former method is the more straightforward and is easily transferred between scales to meet the
is thus more widely applied. increasing drug product needs of an evolving
Solvent evaporation for the production of solid development program. According to these
dispersion systems can be carried out as a batch attributes, spray drying and ASDD technology is
operation by drying a solution of drug plus applicable to drugs with widely varying
excipients with the use of a vacuum oven, lyoph- physiochemical properties and can be
ilizer, or a rotary evaporator. This is a commonly implemented at any stage in drug development.
used approach in early development stages when
API quantities are limited. One must be careful
when employing this technique as gradual solvent 10.9.2 Solvent Selection
evaporation can lead to drug–excipient phase sep-
aration (Curatolo et al. 2009). This can lead to The identification of a poorly water-soluble drug
false-negative results with respect to homogene- as a good candidate for spray drying begins dur-
ity, stability, and performance of the disperse ing the preformulation stages. Essentially, the
system. Also, exposure to thermal stress during determinant factor is the compound’s solubility
this process can be significant; thus, care must be and stability in organic solvents that are suitable
taken with this approach for thermally labile for spray drying. A solubility screening study
molecules. Finally, the dried solid material will should be conducted to determine the solvents
typically be in the form of a film or a solid foam, and cosolvent systems in which the drug has
and thus particle properties will be irregular and sufficient solubility to enable spray drying with
possibly heterogeneous from batch to batch. Care reasonable solids throughput. As a rule of thumb,
the drug should have solubility of at least 1–3% to cases in which amorphous systems can be suc-
consider spray drying a viable option for large- cessfully produced from liquid feeds in which the
scale manufacturing. Following selection of the drug is in solution and the polymer in suspension,
solvent system, stability studies of the drug in a complete solution is usually desired so as not to
solution should be conducted to identify instabil- risk the formation of a less stable two-phase com-
ity of the compound in the solvent and determine posite. A list of some suitable organic solvents for
the time scale of degradation in order to establish each commonly used polymer is provided for
a shelf life. Considering that in large-scale pro- reference in Table 10.10.
duction a solution may be sitting for several hours
or days before and during spray drying, stability Drug–Polymer Miscibility
of the drug in the solution is important. Selection of the appropriate polymer for an
ASDD should begin during the preformulation
stages. Once a drug candidate is identified as
10.9.3 Carrier Selection potentially requiring formulation as an amor-
and Optimization phous solid dispersion, efforts should begin to
identify polymers that are miscible with the com-
In an amorphous solid dispersion system, the pound. Drug–polymer miscibility is a critical con-
carrier principally serves the following functions: sideration during polymer selection, because in a
(1) imparts stability onto the amorphous drug in substantial majority of cases, drug–polymer
the solid state, (2) enhances solubility and/or miscibility is a requirement for physical stability
prolongs supersaturation in aqueous of the noncrystal-line drug.
environments, and (3) dictates the primary loca- As discussed in Chap. 2, calculation of solu-
tion of drug release in the GI tract. Owing to bility parameters can be useful for identifying
properties such as high molecular weights, high good polymer candidates for ASDD systems
Tg’s, water solubility, varying ionic character, with the drug of interest. Conversely, solubility
and, in some cases, surface activity, polymers parameters will aid in identifying polymers that
are almost exclusively utilized as the primary have limited miscibility with the drug that should
carriers for amorphous solid dispersion systems. be avoided. Although solubility parameters can
The most commonly used polymers for ASDD provide valuable insight into drug–polymer
applications are provided in Table 10.11. miscibility for a broad number of polymer
Nonpolymeric additives are also frequently candidates, these predictions can sometimes be
incorporated into carrier matrices, but typically misleading; therefore, empirical data is preferred.
serve secondary roles as functional adjutants, for In that way Duarte et al. (2015) proposed a
example, wetting agents, pore formers, moisture screening methodology based on model combin-
scavengers, densifying agents, anti-plasticizing ing thermodynamic and kinetics descriptors to
agents, etc. Selection of optimal polymers for fine-tune the carrier selection with a solvent cast-
amorphous solid dispersion systems is discussed ing protocol.
in detail elsewhere in the book (Chaps. 1 and 8). High-throughput, material-sparing screening
Therefore, discussion of carrier selection for methods for assessing drug–polymer miscibility,
ASDDs in this section is kept intentionally brief. like those described in Chap. 8 and elsewhere in
the pharmaceutical literature (Kwong et al. 2011;
Solubility in a Common Solvent Moser et al. 2008a, b; Shanbhag et al. 2008), are
For amorphous spray-dried dispersions, the poly- recommended for rapidly assessing a multitude of
mer must be soluble in a common solvent or carriers while consuming minimal API. Typically
solvent system with the drug. This is to ensure a 96-well robotic system is utilized to produce
mixing of the drug and polymer at a molecular and analyze solvent cast drug–excipient films.
level such that a single-phase system can be Following evaporation of solvent, the cast films
achieved on drying. Although there are some in each well are analyzed by high-throughput
Table 10.11 Polymers commonly used in solid dispersions produced by batch solvent evaporation and/or spray drying:
suitable organic solvents and corresponding references for use
Common
polymers Suitable organic solvents Selected references
Hypromellose Acetone, ethyl acetate, methanol/ethanol/ Tanno et al. (2004), Kennedy et al. (2008),
acetate succinate dichloromethane (1:1 v/v) Friesen et al. (2008), Curatolo et al. (2009)
(HPMCAS)
Methacrylic acid Acetone, ethanol, methanol, ethanol/ Tanno et al. (2004), De Jaeghere et al.
copolymers dichloromethane (1:1 v/v) (2000)
Hypromellose Ethanol/dichloromethane (1:1, 2:1 v/v), methyl Kohri et al. (1999), Yamashita et al.
(HPMC) acetate/methanol (1:1 v/v) (2003), Tanno et al. (2004), Kennedy et al.
(2008), Curatolo et al. (2009)
Povidone (PVP) Acetone, most alcohols dichloromethane, ethyl Yamashita et al. (2003), Paradkar et al.
acetate, methyl ethyl ketone, tetrahydrofuran (2004), Tanno et al. (2004), Ambike et al.
(2005), Curatolo et al. (2009), Paudel et al.
(2010)
Copovidone Acetone, most alcohols dichloromethane, ethyl Janssens et al. (2008a, b)
(PVPVA) acetate, methyl ethyl ketone, tetrahydrofuran
Polyethylene Acetone, most alcohols dichloromethane, ethyl Yamashita et al. (2003), Jung et al. (1999),
glycol (PEG) acetate, methyl ethyl ketone, tetrahydrofuran Law et al. (2004)
Poloxamers Acetone, most alcohols dichloromethane, ethyl Jung et al. (1999), Wong et al. (2006)
acetate, methyl ethyl ketone, tetrahydrofuran
Polyvinyl acetate Ethanol, methanol, acetone/ethanol (1:1 w/w), Curatolo et al. (2009)
phthalate (PVAP) acetone/methanol (1:1 w/w), methanol/
dichloromethane (1:1 w/w)
Cellulose acetate Acetone, methyl ethyl ketone, ethyl acetate Curatolo et al. (2009)
phthalate (CAP)
Amonio Acetone, methanol, ethanol, isopropyl alcohol, tert- Jung et al. (1999)
methacrylate butyl alcohol dichloromethane, ethyl acetate,
copolymer methyl ethyl ketone, tetrahydrofuran
Hypromellose Acetone (with heat); ethanol/dichloromethane (1:1, Kohri et al. (1999), Tanno et al. (2004),
phthalate 2:1), methanol/dichloromethane (1:1) Engers et al. (2010), Curatolo et al. (2009),
(HPMCP) Cui et al. (2006)
Hydroxypropyl Dichloromethane, ethanol, methanol, chloroform Curatolo et al. (2009), Cui et al. (2006)
cellulose (HPC)
PXRD, polarized light microscopy, and electron solubility enhancement, prolongation of supersat-
microscopy for evidence of drug crystals or phase uration, and finally supersaturated dissolution
separation (Moser et al. 2008a). A hybrid testing. With respect to solubility enhancement,
approach of first principles models and empirical screening studies can be conducted in which the
assessment may optimize the carrier selection crystalline drug is incubated under agitation with
expediting the screening of numerous polymers the excipient system in aqueous media for a suit-
with varying drug loads. Polymers which are able duration to establish equilibrium. The system
identified as miscible with the drug beyond the is then centrifuged or filtered and the supernatant
target drug loading range should then be selected or filtrate is analyzed for drug concentration.
for further evaluation on a larger scale. Excipients can then be rank ordered with respect
to drug solubility enhancement and selected
accordingly for further study. Typically,
Rapid in Vitro Screening of Carrier
polymers and surfactants are screened separately,
Excipients
and leads from both groups are combined to eval-
As discussed in Chap. 2, various in vitro tests can
uate potential synergies. By this method, one can
be conducted to assess performance of excipient
quickly determine lead polymers, surfactants, and
carriers for ASDD systems with respect to
combinations thereof for small-scale spray-drying concentrations identified, the influence of drug
studies. loading for the various carrier systems can be
In compliment to solubility enhancement, a evaluated. Ideally, at the completion of this pro-
supersaturation assay should be performed to cedure, lead polymers, surfactants, and
comparatively evaluate various polymers, combinations have been identified along with
surfactants, etc., with respect to prolongation of optimum drug loadings.
drug supersaturation. As described by Considering the results of the solubility
Vandecruys et al., this screening protocol enhancement screening, supersaturation assay,
involves pre-dissolving the excipient(s) in aque- and high-throughput dissolution screening
ous media and then introducing the drug as a together, the formulation scientist will have
solution in a water-miscible organic solvent such extensive information regarding carrier
as dimethylacetamide (Vandecruys et al. 2007). compositions that yield acceptable dissolution
The drug will initially be highly supersaturated performance. With this information, the formula-
but will precipitate with time at varying rates tion scientist can begin to design ASDD
depending on the drug–excipient interactions in formulations and produce powder on small-scale
solution. According to this rate of precipitation, a equipment, e.g., a Buchi Nano B-90 or Buchi
rank order for the evaluated excipients can be Mini B290 spray drier (see Table 10.3). At this
established. point, in vitro dissolution screening can be
A high-throughput screening method similar performed on the actual spray-dried powder
to that described above for drug–carrier formulations. This in vitro testing is critical
miscibility determination can also be employed because the above test methods are only
for rapid assessment of the dissolution perfor- approximations of the performance of the actual
mance for a vast array of compositions. ASDD system. The only means of truly
According to this method, dissolution testing is evaluating the performance of various carrier
conducted on the solvent cast films in each well of systems for ASDDs is to perform dissolution
a 96-well plate. Filtration and analysis of drug testing on the actual spray-dried powder. Any
concentration by HPLC can be automated to pro- unexpected results, whether positive or negative,
vide rapid results. For specific details regarding obtained from the above-described screening
this method, the reader is referred to Shanbhag methods can also be reevaluated with actual
et al. (2008) and Chap. 8 of this text. According to ASDD material to avoid unwarranted acceptance
the automated nature of the system and the num- or rejection.
ber of wells per plate, numerous dissolution
experiments can be conducted simultaneously. In Vitro Dissolution Testing of ASDDs
This provides opportunities for the formulation There are several options for in vitro dissolution
scientist to broadly screen numerous polymers, testing of spray-dried powder as described in
surfactants, other adjuvant excipients, and detail in Chap. 2. Selection of the appropriate
combinations thereof well beyond what is possi- dissolution test method should be made based
ble by manual means. on the quantity of material available, the
A typical high-throughput dissolution screen- properties of the drug, and the formulation
ing procedure employed for developing an ASDD components. Small-volume, material-sparing dis-
formulation will first involve screening a variety solution testing methods include the micro-
of polymers at a given drug loading. Following centrifuge (Curatolo et al. 2009) and syringe/filter
identification of lead polymers, another plate methods (Curatolo et al. 2009). Larger-volume
array will be designed to investigate synergies methods (500–1000 mL) typically involve the
between lead polymers and various surfactants. use of traditional USP Apparatus II systems (pad-
Next, concentrations of lead surfactants are varied dle method). Scaled-down versions of the paddle
to identify an optimum. Finally, with lead method (~100 mL) have also been used (Overhoff
polymers, surfactants, and surfactant et al. 2007).
Dissolution of ASDD systems should be Further description of this study is provided in

conducted at nonsink conditions in order to assess Method Capsule 10.2. At the conclusion of this
the extent and duration of supersaturation since series of in vitro screening studies, the formula-
these are the key metrics for predicting in vivo tion scientist should arrive at a short list of lead
performance (Miller et al. 2008a, b). Considering ASDD formulations for analytical characteriza-
that the supersaturated dissolution testing of tion and stability assessment.
amorphous solid dispersion systems will lead to
not only free drug in solution but also a variety of
particulate species (Friesen et al. 2008), it is 10.9.4 Analytical Characterization
imperative to also utilize an appropriate method of ASDD Formulations
for separating drug in solution from solid
particles. These can be accomplished by filtration Once spray-dried dispersions are produced, a bat-
or centrifugation of aliquots of the dissolution tery of analytical tests should be conducted to
media; however, these methods are not always thoroughly characterize the material. In
effective at eliminating particulates. Alternative Table 10.12, analytical tests most pertinent to
methods for measuring free drug include passive ASDD systems are listed along with a description
diffusion systems employing a semipermeable of purpose(s) and illustrating references. These
membrane and biphasic dissolution systems tests are critical to perform during formulation
containing an octanol layer into which the drug screening for comprehensive analysis toward the
partitions (Heigoldt et al. 2010; Shi et al. 2010). identification of lead candidates. It is also impor-
The appropriate dissolution media must be tant to perform these analyses each time a spray-
selected according to the properties of the drug dried batch is produced to ensure that it meets
and the formulation. If the drug has pH-dependent product specifications. Additionally, when
solubility, a pH shift method should be employed evaluating an ASDD product following storage
in order to capture pH influences on dissolution/ at accelerated conditions, it is important to assess
precipitation and evaluate formulation perfor- changes in the material by these methods. Specif-
mance accordingly. Also, if the polymer is ionic, ically, the use of polarized light microscopy
it should be soluble in at least one stage of the (PLM) and scanning electron microscopy (SEM)
dissolution test, i.e., one should not test an enteric is important with respect to identifying the emer-
polymer-based system only in acidic media. If the gence of drug crystals as these techniques are
drug and polymer are nonionic, a single-stage test straightforward and highly sensitive. For a more
should be sufficient. In both single-phase and thorough summary and detailed descriptions of
multiphase dissolution tests, selection of media these techniques, the reader is referred to Chap. 2.
should be based on iterative method development
to achieve the desired discrimination between
formulations and correlation to in vivo results. 10.9.5 Stability of Amorphous
Using small quantities of ASDD powder, Spray-Dried Dispersions
in vitro dissolution testing can be conducted as
described above to confirm (on a narrower group) Stability screening of ASDD formulations is as
the rank ordering of polymers, surfactants, and critical to selecting lead compositions as in vitro
multicomponent carriers as well as trends related performance. As with drug–carrier miscibility
to drug loading. An example of dissolution and in vitro testing, it is imperative to establish a
screening of ASDD powders to identify lead rank order of formulations with respect to stabil-
polymers was provided by Curatolo et al. ity such that at the end of screening, rank orders
(2009). As seen in Fig. 10.11, these authors of all criteria can be assessed together in the
established a rank order of various polymers by selection of lead ASDD formulations. To this
supersaturated dissolution testing of ASDD end, accelerated stability studies should be
formulations by the micro-centrifuge method. performed on binary ASDD systems (drug and
Fig. 10.11 Dissolution 200

performance of ASDDs at
Dissolved Drug Concentration (µg/mL)

10% drug loading of
“compound 5” with various
polymer carriers. 150 HPMCAS-MF
Reproduced from Curatolo
et al. (2009) with
permission from Springer
100
CAP
HPMC CAT
50
HPMCP
Amorphous Crystalline
0
0 20 40 60 80 100
Time (min)
Table 10.12 Analytical tests and their purpose for characterizing ASDDs
Analytical test Purpose (selected references provided)
Liquid chromatography (HPLC, Quantify drug and impurities content (Janssens et al. 2008b, Kennedy et al. 2008)
UPLC)
Powder X-ray diffraction Detect presence of crystalline drug to limit of detection (Curatolo et al. 2009,
(PXRD) Friesen et al. 2008, Kennedy et al. 2008)
Differential scanning calorimetry Detect presence of crystalline material to limit of detection, determine dispersed
(DSC) state of drug in carrier (Curatolo et al. 2009, Friesen et al. 2008, Kennedy et al.
2008)
Polarized light microscopy Detect presence of drug crystals (Law et al. 2001)
(PLM)
Scanning electron microscopy Assess particle size, shape, and morphology. Detect presence of drug crystals
(SEM) (Friesen et al. 2008, Kennedy et al. 2008)
Transmission electron Assess particle size and shape, evaluate dispersed state of drug in carrier (Friesen
microscopy (TEM) et al. 2008)
Thermogravitational analysis Quantify amount of volatile content (Kennedy et al. 2008)
(TGA)
Gas chromatography Quantify residual solvent content (Janssens et al. 2008b, Witschi and Doelker
1997)
Fourier transform infrared Determine drug–polymer interactions (Matsumoto and Zografi 1999)
spectroscopy (FTIR)
Laser light scattering, dynamic Determine particle-size distribution (Curatolo et al. 2009, Friesen et al. 2008,
light scattering Kennedy et al. 2008)
Helium pycnometry, mercury Quantification of surface area and apparent particle density (Moser et al. 2008a, b)
porosimetry
Dynamic vapor sorption Assess water uptake of the formulation (Friesen et al. 2008)
primary carrier) at varying drug loads between the 50 C, 60 C – in open and closed containers.
upper limit determined from miscibility studies Samples should be pulled according to a stability
and the lowest practical drug loading. Once pro- schedule and physical stability should be
duced, the binary mixtures should be stored at evaluated. PXRD is the most common method
accelerated conditions, e.g., 40 C/75% RH, for the detection of recrystallization; however,
Table 10.13 Example of an accelerated physical stability program for ASDD systems
Physical stability data for VX-950 spray-dried dispersion
A ¼ amorphous
C ¼ crystalline
Blank ¼ not tested
Formulation Condition Container Initial 1 Wk. 2 Wk. 1 Mo. 2 Mo.
Pure amorphous VX-950 40 C/75% RH Closed A A A A
60 C Closed A A A A
25 C/60% RH Closed A A A A
40 C/75% RH Open C
Solvent evaporation 40 C/75% RH Closed A A A
VX-950:PVP K30 (1:1), 1% SLS 60 C Closed A A A
25 C/60% RH Closed A A A
Spray dried 40 C/75% RH Closed A A A A
VX-950:PVP K30 (1:1), 1% SLS 60 C Closed A A A A
40 C/75% RH Open C
Solvent evaporation 40 C/75% RH Closed A A A A
VX-950:PVP K16 (1:1), 1% SLS 60 C Closed A A A A
40 C/75% RH Open A
Adapted from Cui et al. (2006)
the limit of detection of this method is typically performing (in vitro/in vivo) carrier system
around 5% (Friesen et al. 2008). Microscopy- tends to absorb moisture leading to destabiliza-
based methods, such as SEM and PLM, are there- tion of the amorphous API, certain packaging
fore recommended for fist-line detection of configurations can be utilized to protect the drug
recrystallization as it is possible to detect the product from moisture. In this example,
presence of drug crystals by these techniques accelerated stability testing serves as an early
well below the limit of detection of PXRD. An indication that packaging will be critical to prod-
example of such a stability program is provided in uct stability. Alternatively, if all other aspects of
Table 10.13 in which amorphous systems of several amorphous formulations are similar yet
VX-950 were stored at varying conditions and there are significant differences in physical stabil-
durations and then evaluated for the presence of ity, it would be wise to focus efforts on the most
crystallinity (Cui et al. 2006). It is observed that stable systems.
all samples were initially amorphous; however, In addition to physical stability, chro-
the pure amorphous drug and the ASDD formula- matographic analysis (HPLC) should be
tion with the PVP K30 carrier showed evidence of conducted on the accelerated stability samples to
crystallization after 1 month at 40 C/75% RH. detect changes in potencies and impurity profiles
Identification of physical instability of an with storage time. This will provide an early
amorphous system by an accelerated method is indication of chemical incompatibility of the
not necessarily a reason to eliminate a carrier drug (in a noncrystalline state) and polymer(s).
from further development, particularly if This is particularly important for amorphous
in vitro/in vivo performance suggests selection formulations, because excipient compatibility
of that polymer. This method is simply used to studies conducted during preformulation stages
rank order the physical stability of the amorphous typically utilize the crystalline drug in a binary
formulations to allow identification of the opti- blend with the excipient and therefore do not
mum polymer taking into account all attributes of adequately simulate the drug–excipient
the system. For example, if a certain high- interactions in an ASDD system. If the levels of
impurities are minor or if there is an opportunity 10.9.7 Preclinical Considerations

to qualify the impurities during toxicology stud-
ies, some chemical incompatibility may be man- Drug insolubility can be particularly challenging
ageable. However, if the formation of impurities in the preclinical phases of development. Specifi-
is substantial (significantly reducing product cally, achieving exposures in animal models that
potency) and there is no opportunity to qualify are sufficient to establish adequate safety margins
the impurities, then such chemical instability over the predicted efficacious dose can be difficult
would be an eliminating factor for the carrier. with poorly soluble compounds. Poor solubility
By accelerated stability programs, polymers limits not only exposure but dose linearity as
can be rank ordered with respect to physical and well. Dose linearity tends to be poor for insoluble
chemical stability of the amorphous drug. As key compounds as plateauing exposure is often seen
adjuvant excipients are identified by in vitro dis- during dose escalation and prior to achieving the
solution screening, SDD compositions containing desired exposure. Reducing the particle size of
these excipients should also be subjected to the crystalline API by micronizing, wet milling,
accelerated stability studies. Those systems with or nano-milling can improve exposures but is
unmanageable physical and chemical instability often not sufficient to achieve the target exposures
should be eliminated from consideration. for high-dose insoluble molecules due to inherent
solubility limitations of the crystalline form.
Solution formulations are another often-
employed approach that can yield improved
10.9.6 Pharmacokinetic Evaluation
exposures and dose linearity over the crystalline
drug; however, poor drug solubility in suitable
At the culmination of spray-dried dispersion
vehicles and vehicle dose restrictions often limit
development, an animal PK study should be
the amount of compound that can be administered
conducted to select the final ASDD formulation
with these formulations.
(s) from the leads. When developing a formula-
ASDD systems offer obvious advantages over
tion for toxicology studies, PK assessment should
crystalline suspension and solution formulation
be conducted in both animal species selected for
approaches. First, by converting the API to an
the toxicology studies. In this case, the best
amorphous form, the inherent solubility
performing formulation is directly selected. In
limitations associated with the crystalline API
the case of clinical and market formulation devel-
are eliminated. Second, if properly designed, the
opment, animal PK assessment of the lead ASDD
solid dispersion formulation can be easily
intermediates is done to reduce the leads to a
constituted in standard aqueous vehicles and can
number manageable for prototype dosage form
remain amorphous for several hours to enable a
development. This will include process develop-
sufficient time window for dosing. Therefore,
ment and scale-up which may result in further
amorphous solid dispersions offer the solubility
reduction of intermediate formulations based on
benefits of a solution formulation without the
manufacturability. Ultimately, the prototypes will
need for nonaqueous vehicles that can limit dos-
be evaluated again in animal PK studies before
ing volumes. Additionally, ASDDs are often-
the selection of clinical formulation(s). For entry
times found to provide improved exposure and
into humans, the best performer from this final
dose linearity over solution formulations. This is
animal study will proceed into clinical study. For
likely due to the intimate association between the
phase II/market formulations, this final animal PK
drug and the excipient carrier which delays pre-
study will be used to narrow the candidates to a
cipitation with respect to solution formulations
suitable number to test in a human bioavailability
which can be less efficient in keeping the drug
or bioequivalence study, with the final clinical/
in solution when mixed with GI fluids.
market formulation being selected based on these
results.
Fig. 10.12 Dose 600 Crystalline in 10% Tween 80

proportionality of Amorphous form in 0.5% MC/0.24% SLS
crystalline compound 3 in Solid dispersion
10% Tween ( filled 500
(50% compound 4/50% HPCPAS-HF)
triangle); amorphous form
in 0.5% Methocel/0.24%
400
SDS ( filled square); solid
AUC (µMxh)
dispersion at 50% drug
loading in HPMCAS-HF in 300
suspension ( filled circle)
from 10 mpk to 750 mpk
dosed at 5 ml/kg in Sprague 200
Dawley rats (n ¼ 4).
Reproduced from Kwong
et al. (2011) with 100
permission from Elsevier
0
0 100 200 300 400 500 600 700 800
Dose (mpk)
An example of improved exposure and dose 2008; Kohri et al. 1999; Miller et al. 2008a). The
linearity for an ASDD-based toxicology formula- additional benefit of these polymers for toxicol-
tion is described in an article by Kwong et al. ogy formulations is that they are insoluble in
(2011). In this example, the researchers devel- aqueous buffers of pH < 5. By properly adjusting
oped an ASDD formulation with an HPMCAS- the pH of the vehicle, the solid dispersion can
HF carrier for a poorly water-soluble molecule, remain amorphous for several hours to facilitate
“compound 3.” The powder was dosed as a sus- dosing.
pension from an aqueous vehicle containing 0.5% Suspension stability of an ASDD-based toxi-
methylcellulose and sodium lauryl sulfate (SLS). cology formulation as a function of vehicle pH
The dose proportionality comparison of this for- was demonstrated in articles by Moser and
mulation between suspensions of the pure crys- coworkers (Moser et al. 2008a, b). In this study,
talline and amorphous drug is shown in the authors describe the development of an
Fig. 10.12. From these results, it is observed that ASDD formulation with an HPMCAS-HF carrier
the ASDD formulation provided far superior in support of high-dose preclinical safety studies
exposure and dose linearity as compared to the of “compound A.” This formulation was found to
suspension formulations. yield a tenfold increase in exposure over a crys-
Selecting the appropriate carrier for an amor- talline formulation in dogs and thus enabled dose
phous spray-dried dispersion intended for toxicol- escalation to the desired exposures. As part of the
ogy studies is critical not only to achieve the toxicology formulation development process,
maximum drug exposure and dose linearity but these authors investigated the influence of vehicle
also to ensure that the ASDD formulation remains pH on the physical stability of the ASDD formu-
amorphous in the vehicle for a sufficient duration lation in suspension. The results of this study for
to facilitate dosing of multiple animals from the low concentration suspensions (1 mg/mL) in
same suspension lot. For this purpose, ionic deionized (DI) water and pH 4.0 acetate buffer
polymers tend to be the best carriers. There are are shown in Fig. 10.13. These results clearly
numerous articles in the literature which demon- demonstrate the influence of vehicle pH on the
strate the benefit of anionic polymers for intesti- physical stability of the ASDD formulation.
nal targeting of supersaturation to improve oral Recrystallization of compound A is evident only
absorption of insoluble compounds (Curatolo after 1 hour in DI water, whereas no appearance
et al. 2009; Friesen et al. 2008; Kennedy et al. of crystals is seen in acetate buffer for at least
Fig. 10.13 XRPD spectra of solids pulled from suspen- pH ¼ 4 acetate buffer (4, 2, 1 h), spray-dried powder,
sion (dried at R.T., vacuum) in both DI water and pH ¼ 4 crystalline API. Reproduced from Moser et al. (2008b)
buffer, spray-dried dispersion (control), and crystalline with permission from Russell Publishing, LLC
API. Order from top to bottom: DI water (4, 2, 1 h),
4 hours. This example thus illustrates the impor- For example, a suspension in which maximum
tance of utilizing a buffered vehicle of pH below viscosity is achieved at 20% (w/w) solids loading
the onset of dissolution of the carrier polymer to with an ASDD containing 10% active results in a
ensure physical stability of the ASDD system for maximum suspension concentration of only
a duration sufficient for study dosing. 20 mg API/mL, whereas a solid dispersion
As with any amorphous solid dispersion, containing 30% drug loading yields a maximum
optimizing drug loading of the formulation is of 60 mg API/mL suspension, thus enabling a
critical. It is particularly important for toxicology threefold higher dose per volume. Achieving ade-
formulations not only to achieve the required quate drug loading in the ASDD is essential to
drug exposures but also to ensure suitable drug achieving the target exposures for molecules
loading and amorphous stability in suspension. requiring highly elevated exposures for toxicity
The upper limit of solids loading of ASDD pow- assessment. However, drug loading must be bal-
der in aqueous vehicles is typically in the range of anced against suspension stability, as higher drug
15–30% (w/v). As solids loading is increased, the loading tends to result in faster recrystallization in
suspension becomes increasingly viscous up to a the vehicle because there is less excipient to pro-
point where dosing via gavage is prohibitively tect the amorphous drug. Experimentation is
difficult. Given this limitation, it is important to required to determine the maximum drug loading
maximize drug loading in the solid dispersion that can be maintained amorphous in suspension
system to achieve acceptable drug concentrations for a suitable dosing time interval.
in the suspension.
Spray drying as a formulation technology is production of ASDD formulations makes it an

also advantageous for toxicology formulations ideal process for the manufacture of formulations
with respect to the physical properties of the for preclinical safety studies at all development
powder product. Controlling particle size is stages.
important to ensuring that a formulation can be In summary, the advantages of spray-dried
administered via gavage needle of reasonable dispersions to preclinical development are mani-
gauge for rat dosing. Since it is straightforward fold. First, aptly formulated ASDDs yield a more
to produce spray-dried dispersions with a mean bioavailable form of the drug that is also physi-
particle diameter of less than 100 μm, spray dry- cally stable in suspension. Second, spray-dried
ing is an ideal process for producing particles that powder is ideal for constitution in aqueous
are easily delivered by standard animal dosing vehicles and forms a homogenous suspension
practices. It is also important that powders for which can be easily and accurately dosed by
constitution utilized for toxicology purposes wet gavage. Finally, the spray-drying process can be
and suspend well in the delivery vehicle. In this scaled up quickly to accommodate all stages of
regard, spray-dried powder is ideal in that it is toxicology assessment: from initial animal PK
typically high surface area and low bulk density. studies requiring only a few grams of drug prod-
Therefore, with appropriate carrier selection, wet- uct to multi-week GLP studies requiring tens of
tability and suspendability of spray-dried powder kilos.
is typically very good. Thus, spray drying
produces easily constitutable powders that yield
homogenous suspensions that are easily and accu-
10.9.8 Clinical Considerations
rately dosed to all animal species.
Spray-dried solid dispersions also offer
There are a few key benefits to utilizing an ASDD
processing advantages during preclinical phases
formulation strategy for a poorly water-soluble
of development. First, spray drying on very small
molecule for entry into and progression through
and very large scales can be accomplished rather
clinical development, aside from the obvious
easily. Therefore, it is straightforward to produce
enhancement of oral bioavailability. Paramount
the drug product as it progresses through the
among these benefits is that insoluble molecules
various phases of clinical candidate selection:
will typically require a solubility enhanced for-
from PK studies to dose range-finding studies
mulation for preclinical safety studies as well as
and to multi-week GLP toxicology studies.
for safety and efficacy assessment in humans.
Often the transition between these phases can be
Therefore, an optimized ASDD formulation
rather fast, and API may only become available
used in preclinical safety studies can often be
just prior to the study start. Hence, there is little
directly utilized in clinical formulation develop-
time and drug available for extensive process
ment. An obvious benefit to this situation is
development and scale-up. The concept of an
reduced development time. Another key benefit
ASDD can be proven on a very small scale,
relates to the safety assessment of the ASDD
using a Buchi Nano Spray Dryer B-90 system,
formulation. By using the same or similar
for example, and this powder can be used to
ASDD formulation in clinical studies as was
supply initial animal PK studies. At a small
used preclinical safety evaluation, any impurities
scale, optimizing the spray-drying process
unique to the ASDD formulation will have been
involves a simple design of experiments study to
previously qualified. Moreover, safety of the car-
identify acceptable process parameters. With the
rier formulation will also have been established.
basic process parameters established, stepwise
Since ASDD formulations often include excipient
scale-up of the spray-drying process to meet the
components for which there is no safety prece-
increasing needs of a progressing safety assess-
dent at the delivered amount, the latter benefit is
ment program is straightforward. Hence, the ease
an important one.
in transitioning between different scales for the
Fig. 10.14 Morphologies of spray-dried powder. Reproduced from Curatolo et al. (2009) and Friesen et al. (2008) with
permission from Springer and the American Chemical Society, respectively
ASDD formulations also offer advantages to 10.9.9 Final Dosage Form

clinical studies with respect to their applicability Development
to various final dosage forms, such as powder for
constitution, powder in a capsule, tablets, etc. After the formulation and processes have been
This flexibility provides formulation scientists developed, efforts are then focused on converting
various product options to choose from at each ASDD intermediates into final dosage forms. In
stage of clinical development while keeping the this section, downstream processing of spray-
critical component of the formulation constant. dried intermediates for the facilitation of final
This allows for quick reaction to the development dosage form manufacturing is discussed. Com-
program’s changing drug product needs. For mon issues encountered when post-processing
instance, to achieve rapid entry into clinical stud- spray-dried powders will be reviewed along
ies, a simple powder in a bottle concept can be with methods for overcoming these problems.
employed to minimize product delivery time to Then, final dosage form development is discussed
the clinic while providing dosing flexibility for with focus on the most common systems along
ascending-dose studies. As the program moves with strategies for utilization according to the
into small patient studies, a portable easy-to- stage of development.
dose formulation may be required, e.g., for take-
home studies. In this case, a powder in a capsule
Powder Densification
formulation can be rapidly developed from the
The particle morphology of spray-dried
ASDD formulation. Finally, as the program
dispersions, as illustrated by Fig. 10.14, tends to
moves into phase IIb, a close-to-market tablet
resemble a hollow or shriveled sphere structures.
formulation could be required. This tablet dosage
This particle morphology leads to high surface
form could also be developed in a relatively short
area which is advantageous for dissolution
period from the ASDD formulation. Keeping the
enhancement; however, it also results in low
critical component formulation constant through-
bulk density of the spray-dried powder which
out the various phases of development ensures
can be problematic with respect to dosage form
similar PK performance of the various delivery
production.
systems, thereby reducing risk of formulation-
The bulk density of spray-dried powder can
related issues in late-phase development. A
vary significantly, depending on the scale of
more detailed discussion of key considerations
manufacturing, formulation, process parameters,
for converting ASDD intermediates into these
and nozzle type. For typical amorphous solid
final dosage forms will be described in detail in
dispersions, bulk densities can range from
the following section.
<0.1 g/cm3 up to about 0.4 g/cm3. Bulk densities
Table 10.14 Example properties of a bulk spray-dried dispersion

Bulk properties (after secondary drying) Tray dried at 40 C
Bulk specific volume (g/cm3) 0.2
Tapped specific volume (g/cm3) 0.313
Hausner ratio 1.56
Mean particle diameter (μm) 80
D10, D50, D90* (μm) 25, 73, 143
Span (D90–D10)/D50 1.60
Residual acetone (before secondary drying) 3.0%
Adapted from Crew et al. (2007)
of greater than 0.4 g/cm3 are rarely achieved by could destabilize the amorphous material. Addi-
conventional spray drying alone. Typically, bulk tionally, many spray-dried powders have man-
densities of spray-dried powders fall in the range ageable flow properties, tend not to be tacky,
of 0.1–0.35 g/cm3. An example of the typical and bond well under pressure making them par-
properties of spray-dried dispersions is provided ticularly amenable to the roller compaction
in Table 10.14. It is typically the case that the process.
powder must be further densified in order to Roller compaction with ASDDs can be accom-
enable capsule filling or tablet compression. Den- plished without the inclusion of additional
sity is more critical with high-dose compounds excipients; however, to improve processing and
where a large amount of spray-dried powder will granulate performance, typically lubricant(s),
be loaded into the final dosage form. disintegrant(s), and filler(s) will be required. Fol-
The two primary options for downstream den- lowing roller compaction, the densified ribbons
sification of spray-dried powder are (1) dry gran- are milled to form the granulate product. Ideally,
ulation and (2) wet granulation. As discussed this granulate will have a bulk density of greater
previously, FSD is another option for improving than 0.3 g/cm3 to facilitate capsule filling or tablet
the density of spray-dried product and is particu- compression.
larly attractive because densification is achieved Dry granulation in itself is a complex science;
concurrently with the spray-drying process. How- a detailed discussion of the process is beyond the
ever, FSD has only just recently found application scope of this chapter. This section is included
in the pharmaceutical industry and has the restric- simply to make the reader aware of the options
tion of limited small-scale equipment availability. for densifying spray-dried powders to enable final
Hence, traditional spray drying is of primary dosage form production. The reader is refereed to
interest, and secondary powder densification is Kleinebudde (2004) and Miller (2010) for
typically a requirement. In this section, the use detailed discussions on roller compaction
of dry and wet granulation for the densification of technology.
ASDD powder is briefly reviewed. In contrast to dry granulation which can be
Dry granulation is a process of agglomerating applied to most ASDDs, wet granulation can
powders by the application of high pressure. The only be applied to amorphous systems which are
pharmaceutically relevant dry granulation pro- not sensitive to the granulation fluid. If the drug
cesses include slugging and roller compaction. and/or polymer has solubility in the granulation
Owing to greater throughput, more precise pro- liquid, the properties of the amorphous dispersion
cess control, and minimal need for lubricants, can be disrupted leading to recrystallization of the
roller compaction has emerged in recent years as drug. An example of a scenario in which wet
the preferred dry granulation process in the phar- granulation would be an option is one where the
maceutical industry (Kleinebudde 2004). Roller amorphous dispersion is stabilized with an ionic
compaction is ideal for densifying ASDDs polymer such as HPMCAS, HPMCP, methacrylic
because the process does not require liquid that acid copolymer, and amonio methacrylate
copolymer; the granulation fluid is of a pH at If wet granulation is selected, the formulation

which the polymer is not soluble. Even in this scientist should endeavor to thoroughly charac-
scenario, the drug must be highly stable in the terize the process at lab scales before utilizing for
dispersion for wet granulation to be viable. As a larger-scale drug product supply.
precursor experiment to performing wet granula-
tion, it is recommended to suspend the spray- Powder in a Bottle Formulation
dried material in the granulation fluid and monitor As discussed previously, ASDDs offer many
the morphology of the system periodically by a unique benefits as drug product for preclinical
suitable analytical technique to determine the safety studies. Without extensive (or any) addi-
onset of drug crystallization. If the amorphous tional downstream processing, ASDD
system remains unchanged for several hours in formulations can be used directly in animal toxi-
suspension, it is a good indication that the spray- cology studies. Pre-weighed powder filled into an
dried powder will remain unchanged by wet gran- appropriate container can simply be constituted
ulation. Nevertheless, extensive process develop- with the addition of the delivery vehicle, ade-
ment and evaluation is recommended at every quately mixed, and then dosed. In this context,
scale to ensure that wet granulation does not an ASDD powder in a bottle concept is ideal.
alter the properties of the ASDD. Similarly, this concept can be employed for
Wet granulation of ASDDs can be conducted early clinical studies as well. The ASDD powder
in a similar manner to traditional wet-granulation can be pre-weighed in individual containers
processes, either by low-shear, high-shear, or which can be combined with a specified volume
fluid-bed granulation techniques. As mentioned of a suitable liquid by shaking to create a homog-
previously, it is important that the drug and the enous suspension ready for administration. Alter-
carrier system be insoluble in the granulation natively, the ASSD powder can be supplied in
liquid to prevent alterations of ASDD properties. bulk to the dosing site to be weighed by the
To achieve the desired granulate properties, addi- attending pharmacist and pre-suspended in a suit-
tional excipients may be required, such as binder able liquid prior to administration. This approach
(s), disintegrant(s), and filler(s). As with any offers significant benefits with respect to reducing
wet-granulation process, the granulate will need time and resources spent on drug product devel-
to be dried to remove excess moisture. This can opment and manufacturing. Additionally, it
be achieved by traditional drying methods as allows for flexibility in dosing which can be ben-
described previously. The granulate yielded eficial in phase I ascending-dose studies.
from this process will ideally have a bulk density
in excess of 0.3 g/cm3 to facilitate capsule filling Powder in Capsule Formulations
or tablet compression. To reiterate, in most instances, particularly for
The amorphous nature of ASDDs coupled high-dose drugs, the ASDD must be densified
with the high surface area of the powders can by an appropriate granulation method to enable
lead to unusual responses to wet granulation. efficient filling of the required dose in a capsule of
Significantly more granulation fluid may be reasonable size. A standard size 1 capsule has a
required than with a traditional granulation to volume capacity of 0.5 ml. Assuming a density of
achieve the desired granulate. Also, uneven 0.2 g/mL for typical bulk spray-dried powder, a
binder distribution owing to inadequate agitation size 1 capsule can hold a maximum of 100 mg of
of the powder bed can lead to the rapid formation total powder. If a solid dispersion formulation
large agglomerates. Higher impeller speeds and contains only 30% drug, the maximum capsule
lower spray rates are therefore recommended at dose strength achievable is 30 mg. On the other
the start of granulation to facilitate homogenous hand, if by granulation the density of the ASDD
granule formation. Because of these challenges, powder is increased to 0.6 g/ml, then the theoreti-
the reader is encouraged to employ dry granula- cal maximum dose strength attainable in a size
tion methods for densification whenever possible. 1 capsule is a more reasonable 90 mg. It is then
understood that for high-dose applications and/or Fillers, disintegrants, flow aids, and lubricants
limitations on the number of capsules per dose will likely be required to ensure acceptable
administration, densification of ASDD powder is manufacturability, tablet hardness, and friability
required for capsule production. Additionally, and to achieve the desired performance with
increasing the density of the spray-dried powder regard to disintegration and dissolution.
will significantly improve flow, thereby enhanc- For products where the ASDD is the principal
ing the efficiency of the capsule-filling process. component of the final dosage form, e.g., greater
Another issue to consider regarding capsule than 50% w/w, carrier selection can dictate tablet
filling with ASDDs is plug formation on dissolu- performance with respect to disintegration and
tion. Plug formation in this context refers to the dissolution. Polymers which function as binders
phenomenon in which the powder content of a and/or gelling agents – such as hypromellose,
capsule does not disperse following dissolution of povidone, and copovidone – tend to produce
the capsule shell. Rather, the powder is held nondisintegrating tablets when external filler is
together presumably by hydrophobic interactions limited by dosage form size. For example, Kaletra
to form a loose solid mass or plug. These plugs and Norvir tablets are principally composed of
typically erode slowly in aqueous environments, amorphous solid dispersions (<20% drug load-
thereby significantly retarding the release of the ing) in copovidone (Berndl et al. 2008). With
drug. It has been explained that slow dissolution low drug loading in the solid dispersion and
of similar plugs stems from faster dissolution of drug doses of 250 mg and 100 mg for Kaletra
the hydrophilic carrier at the surface of the plug and Norvir, respectively, there is limited space in
leading to the formation of a hydrophobic drug- the formulations for external fillers to enhance
rich layer which acts as a barrier to drug release disintegration while maintaining a reasonable
(Serajuddin et al. 1988). dosage form size. As a result, both tablets are
To promote dispersion of capsule contents fol- eroding systems. For many drugs, steady drug
lowing dissolution of the shell, it is recommended release from the tablet surface over the course of
that the ASDD granulate be blended with GI transit is suitable. However, in some cases,
surfactants (Kennedy et al. 2008), wetting agents, site-specific absorption in the duodenum, for
soluble fillers, insoluble fillers, disintegrants, etc. example, immediate release, is required to maxi-
The addition of these external excipients mize oral absorption. In these cases, a
promotes wetting, acts as a stearic barrier to disintegrating tablet is usually required, and
hydrophobic interaction or gel formation, and therefore the previously mentioned polymers are
physically pushes adjacent particles away, thus not suitable as carriers for the solid dispersion. In
acting synergistically to facilitate dispersion of these cases, anionic polymers such as HPMCAS,
the powder from the capsule shell. The result is HPMCP, and methacrylic acid copolymers tend
an immediate release, as opposed to an undesired to be better carriers. In summary, carrier polymer
sustained-release capsule dosage form. selection for the ASDD can have a significant
influence on final dosage form performance. Con-
Tablet Formulations sideration should be given to this issue at the early
Similar to capsules, powder density is also a stages of amorphous solid dispersion
critical factor for tablet compression to improve development.
flow and ensure that the powder mass
corresponding to the required drug dose can eas-
ily fit inside the die cavity. Densification of the 10.9.10 Examples from
ASDD by the appropriate granulation method is the Pharmaceutical Literature
required to enable tablet compression in most
cases. As with most granulations, external In the previous sections, an overview of the
excipients are required to facilitate compression spray-drying process, the development of an
and to achieve the desired tablet properties. ASDD formulation, and final dosage form design
and production has been discussed. In the follow- amorphous form of G-F to significantly increase
ing sections, examples from both the pharmaceu- solubility.
tical literature and industry illustrating the use of Initially, solubility enhancements of G-F were
spray-drying technology for enhancement of dis- attempted with jet milling, cosolvents, and salt
solution properties, oral bioavailability, and ther- formation. These were unsuccessful as they
apeutic efficacy of poorly water-soluble drugs were unable to meet the needs of the efficacy
will be reviewed. and safety preclinical studies. Therefore, an amor-
phous solid dispersion by spray drying was cho-
sen as the method for improving drug solubility.
B-Raf Inhibitor (G-F) with HPMCAS-M
G-F has a melting point between 190 and 228 C,
In a study published by Cui et al., the authors
depending on the polymorph, with a Tg of
demonstrated the ability to generate a spray-dried
120 C. Due to this high Tg, the authors selected
amorphous dispersion of G-F, a B-Raf inhibitor
HPMCAS-MF as the polymer for the amorphous
(Cui et al. 2013). However, prior to developing
dispersion as well as for its resistance to water
the spray-dried composition, the feasibility of
absorption and capability to enhance bioavailabil-
using an amorphous dispersion was evaluated.
ity. A 25% drug load was selected for the disper-
Initially, the pH solubility profile for G-F was
sion. Acetone was chosen as the solvent with a
produced across a range of 1.54–12.05, with
total solids load of 5% (w/v). Pure amorphous
pKa values obtained at 0.68 (base), 7.46 (acid),
G-F, used above, was created by the same process
and 11.42 (acid). The intrinsic solubility was
but without HPMCAS-MF. The spray-dried
poor, several μg/ml or less, across the pH range
amorphous dispersion had a single Tg at 91 C.
of 2–7. The authors identified this as problematic
The Tg value greater than 90 C and the absence
as a dose of at least 100 mg/kg was necessary for
of a second Tg led the authors to predict good
preclinical evaluation of efficacy and safety. Tar-
physical stability for the amorphous dispersion.
get solubility enhancement of at least tenfold
Performance testing was carried out on crys-
would be necessary to enable these studies. A
talline, amorphous, and amorphous dispersion
prediction of amorphous solubility enhancement
G-F material. Nonsink dissolution was performed
could be made from the free energy difference
with pH 6.5 phosphate buffer as the media. The
between the amorphous and crystalline states.
amorphous G-F only demonstrated a twofold
This difference was calculated from determina-
increase over the crystalline form. While this
tion of the isobaric heat capacities of each state.
result was initially surprising, optical microscopy
Analyzing the isobaric heat capacities presented a
revealed recrystallization of G-F which likely led
challenge as the plot was discontinuous at the Tg
to the reduction in solubility. Similar performance
of G-F. To solve, the data was split into three
was observed when amorphous G-F was tested in
intervals: below, at, and above the glass transi-
media that had pre-dissolved HPMCAS-MF. By
tion. These heat capacity equations, along with
contrast, the amorphous dispersion containing
melting enthalpy/temperature, were plugged into
HPMCAS-MF reached a 20-fold increase over
integrated expressions from literature to obtain
the crystalline form. Additionally, the increase
the theoretical amorphous solubility increase.
was sustained for 0.5 hours before solubility
The general expression for this relationship is
began to decrease. The solubility was maintained
ΔGc ! a ¼ RT ln (Ca/Cc), where ΔGc ! a is the
above the crystalline level for over 2 hours. These
difference in free energy between the amorphous
results demonstrated a solubility/dissolution
and crystalline states, R is the gas constant, T is
benefit unique to the molecular interaction
the temperature of interest, and Ca/Cc are the
between G-F and HPMCAS-MF in the spray-
solubilities of the amorphous and crystalline
dried amorphous dispersion. Next, the crystalline
states. At 37 C, experimental results were 194-,
G-F and the spray-dried amorphous dispersion
246-, and 288-fold increases over crystalline state
were performance tested in a rat model. At a
indicating the thermodynamic potential for an
dose level of 100 mg/kg, the exposure increased
threefold. However, at a dose of 300 mg/kg, the diffraction and found to be amorphous with
exposure was 10- to 20-fold higher. The authors respect to sirolimus.
attributed the supra-proportional increase to satu- The four spray-dried compositions, an
ration of the clearance. Ultimately, there was HPMC–sirolimus physical mixture, and raw
good correlation between the in vitro and siroliums were analyzed following a procedure
in vivo performance. similar to the preformulation studies. As expected
from the preformulation study results, solid
Sirolimus with HPMC-TPGS dispersions with HPMC-TPGS and HPMC-
In a study published by Kim et al., the authors SE15 presented the highest supersaturation maxi-
generated and optimized formulations of mum and 24-hour concentration. The HPMC-
sirolimus solid dispersions by spray drying (Kim G44/14 comparator released to a similar maxi-
et al. 2013). Prior to spray drying, the authors mum value, but precipitated to levels similar to
utilized small-scale solvent casting to screen the physical mixture by the 8-hour time point. A
binary and ternary compositions of sirolimus follow-up study in rats returned similar results.
with five polymers (HPC, HPMC, PVP VA64, Statistical analysis (p < 0.05) determined that
PEG 8000, and PVP K30) and seven surfactants HPMC-TPGS ¼ HPMC-
(SLS, TPGS, G50/13, M52, P 407, SE15, and SE15 > HPMC ¼ HPMC-G44/14 > physical mix-
G44/14). For the binary studies, test formulations ture. The in vitro drug concentration data
were solvent cast at drug–excipient ratios of 50: correlated well to the in vivo rat data with R2
50, 20:80, and 10:90 and analyzed after incuba- values of 0.9860 and 0.8616 for the AUC and
tion times of 0.5, 1, 2, and 24 hours in water Cmax, respectively. These studies show the viabil-
media. Plots were generated of drug concentra- ity of spray drying as an option for improved
tion versus incubation time. The surfactants sirolimus supersaturation and bioavailability.
demonstrated a higher capacity for solubility
enhancement than the polymers. TPGS and SLS Telmisartan with Povidone K30 and a pH
had the highest effect on both initial solubility and Modifier
supersaturation maintenance. Also of note, SE15 In a study published by Marasini et al., the
had the slowest dissolution rate of the surfactants. authors developed pH-modulated solid
Next, ternary studies were performed by solvent dispersions of telmisartan (TMS) by spray drying
casting drug–polymer–surfactant ratios of 1:8:1 (Marasini et al. 2012). TMS is targeted for the
for all of the polymers/surfactants examined in reduction of blood pressure as an angiotensin II
the binary study. These compositions were receptor agonist. While it is practically insoluble
incubated in water media for 1 and 24 h followed in water, it has pH-dependent release as its free
by drug concentration analysis. In the ternary acid form is ionizable. To improve the solubility
compositions, TPGS, SE15, and P 407 were the of TMS, the authors targeted the adjustment of
most effective surfactants, and HPMC was the microenvironmental pH to improve dissolution
most effective polymer. Particularly, HPMC- performance. During preformulation work, the
TPGS and HPMC-SE15 were the most effective authors examined carriers and alkalizers at 1%
at both time points. level for their benefit to drug solubility. Among
The results from the preformulation screening the carriers, PVP K30 provided the maximum
were used to select three sirolimus formulations solubility benefit. The authors also noted its
for investigation as spray-dried compositions: advantages of a high melting point, low toxicity,
HPMC, HPMC-TPGS, and HPMC-SE15. tolerability, and precipitation inhibition. Among
HPMC-G44/14 was also chosen as a comparator. the alkalizers, three performed well: magnesium
The solvent for spray drying was a 50/50 (v/v) oxide, sodium hydroxide, and sodium carbonate.
ethanol/dichloromethane mixture with solids con- However, the authors discovered issues with
centration of 20 mg/mL. All four solid magnesium oxide (insoluble in water) and sodium
dispersions were analyzed by powder X-ray hydroxide (instability) that led them to select
sodium carbonate as the alkalizer moving superiority over the current commercial product
forward. approach.
A spray-dried solid dispersion was developed
using TMS, PVP K30, and sodium carbonate. To MK-0364 with HPMCAS and Copovidone
determine the optimum formulation, drug and In a study published by Sotthivirat et al., the
sodium carbonate concentrations were constant authors developed solid dispersions of MK-0364
at 2 and 3 g, respectively, while the K30 amount by spray drying and hot-melt extrusion
was varied. TMS has a melting point of approxi- (Sotthivirat et al. 2013). MK-0364 is targeted as
mately 268 C. A 95/5 water/ethanol mix was a monotherapy in the treatment of obesity. It is a
selected as the solvent with 2 g of drug per poorly water-soluble drug (<0.4 μg/mL) with no
100 mL of spray-dried media. It was noted that discovered stable salt form. Preformulation
the use of significant water in spray drying was evaluations of dispersions were carried out by
unconventional and offered the advantages of solvent casting at a 10% drug load with polymers
lower environmental impact, reduced toxicity Eudragit L100–55, HPMCP HP-55, and
risk, and greater safety with respect to the HMPCAS-LF, with the greatest solubility
manufacturing process. enhancement from HPMCAS. The effect of
After the development of spray-dried surfactants was also investigated in copovidone
formulations, the resulting materials were tested solvent cast dispersions where it was determined
for solubility performance. Equilibrium solubility that polysorbate 80, sorbitan monooleate, and
testing in water revealed that drug solubility in the vitamin E TPGS provided the greatest benefit to
dispersions was substantially improved compared supersaturation.
to TMS powder alone. Interestingly, the formula- Formulations were prepared by hot-melt extru-
tion with the greatest solubility enhancement was sion with copovidone as the polymeric carrier due
the one which contained the least amount PVP to its thermal stability, lower Tg, and challenges
K30. It was suggested that the pH modulation of with processing HPMCAS by hot-melt extrusion.
the system was sufficient to enhance the release PVP was also examined as a means to improve
over any benefit of the added polymer. Interest- physical stability in the copovidone formulations.
ingly, dissolution testing in water showed no dif- A Thermo Scientific twin-screw extruder was
ference between the formulations in release rate used with operating temperatures from 130 to
or magnitude. As the PVP K30 did not appear to 160 C. The spray-dried approach examined
provide significant solubility benefit, the formu- HPMCAS with HPMC samples as a comparator
lation containing the minimum amount was to determine the benefit of cellulosic polymers.
selected for subsequent analysis. Additional dis- Spray drying was performed with a Niro
solution testing of this formulation against TMS SD-Micro™ spray dryer. Feed solution contained
powder and the commercial product, Micardis®, 5% solids with acetone and methanol used as the
was executed in pH 1.2, 4.0, and 6.8 and distilled solvents for the HPMCAS and HPMC
water. In pH 6.8 and water, both the commercial formulations, respectively. Additionally, the
product and the pH-modulated dispersion surfactants polysorbate 80/sorbitan monooleate
performed similarly and far exceed the powder and vitamin E TPGS were examined in the
result. However, in pH 1.2 and 4.0, the new hot-melt extrusion and HPMC spray-dried
formulation surpassed the commercial product samples for their benefit to drug solubility. How-
release profile, while the TMS powder remained ever, surfactant quantity was reduced by half
the lowest release. This result was further verified compared to the solvent cast preformulation due
in a rat model with the new formulation having to processing challenges with hot-melt extrusion.
significantly higher Cmax (>6x improvement) and It was determined that this had no negative effects
AUC (>2x improvement). Thus, a pH-modulated on supersaturation.
spray-dried dispersion of TMS with polymeric All samples prepared by both methods were
carrier and alkalizer demonstrated the potential found to be amorphous with respect to XRPD.
Initial physical characterization of samples tested in animal models and 21 different drugs
generated by hot-melt extrusion found that formulated as SDDs and tested in humans
samples containing PVP were heterogeneous (Friesen et al. 2008). By investigating amorphous
and not suitable for further analysis. While the spray-dried dispersion systems with such a large
dissolution profiles were similar for both the poly- number of compounds, these researchers were
sorbate 80/sorbitan monooleate and vitamin E able to identify general trends for amorphous
TPGS melt-extruded samples, the polysorbate dispersion systems with regard to identifying
80/sorbitan monooleate dispersion displayed optimal polymer carriers, determining optimal
slightly greater bioavailability in rhesus monkeys drug loading with regard to performance and sta-
and was selected as the lead hot-melt extrusion bility as a function of key physiochemical
candidate. For the spray-dried formulations, all of properties, properties of amorphous dispersions
the HPMCAS and HPMC formulations had both in supersaturated aqueous media, and correlating
good physical characterization with rapid in vitro and in vivo performance. Comprehensive
sustained supersaturation by dissolution. Spray- research articles summarizing this work were
dried dispersions containing HPMCAS and published by Friesen et al. (2008) and Curatolo
HPMC/polysorbate 80/sorbitan monooleate were et al. (2009) presenting several example
evaluated in the rhesus monkey model with both molecules to illustrate key concepts.
options performing similarly to the hot-melt The article by Friesen et al. (2008) thoroughly
extruded sample. HPMCAS was chosen as the describes the attributes of HPMCAS that makes it
lead spray-dried candidate due to its lower hygro- a superior carrier for ASDD systems and provides
scopicity compared to HPMC. Both the lead various examples demonstrating its benefits
hot-melt extrusion copovidone formulation and (Friesen et al. 2008). The specific properties of
the lead HPMCAS spray-dried formulation were the polymer identified as being amenable to
evaluated in a 54-week stability study with crys- amorphous spray-dried dispersions are as
tallinity detected in the hot-melt extruded sample, follows:
while none was detected in the spray-dried candi-
1. High Tg in the unionized state reducing drug
date. Thus, the stability benefit of the HPMCAS
mobility and enhancing physical stability of
spray-dried formulation was noted over the
the system. Also, the hydrophobicity of the
copovidone hot-melt extruded formulation. In
polymer in the unionized state prevents
the case of MK-0364, spray drying demonstrated
absorption of ambient moisture, further
its ability to improve drug bioavailability while
improving physical stability.
having better processing and stability
2. Soluble in volatile organic solvents, e.g., ace-
characteristics over the hot-melt extruded disper-
tone and methanol, making it amenable to
sion of the drug.
spray drying.
3. The formation of small stable colloids in aque-
Various Compounds with HPMCAS
ous media above pH 5 as a result of polymer
Some of the most comprehensive research
ionization resulting in greater drug available
conducted to date on ASDD systems resulted
for absorption in the intestinal tract.
from an extensive collaboration between Pfizer
4. Greater interactions with insoluble molecules
Inc. and Bend Research Inc. During this collabo-
in aqueous media resulting from the amphi-
ration, in-depth investigations were conducted on
philic nature of the polymer that produces
all aspects of amorphous spray-dried dispersions,
greater apparent solubilities and maintenance
including formulation design, solid-state charac-
of supersaturation.
terization, understanding physical stability,
in vitro/in vivo performance assessment, and The article demonstrates the formation of
spray-drying process design. This work involved XRD amorphous spray-dried HPMCAS systems
a vast number of compounds with more than for various compounds and discusses the use of
100 different drugs formulated as SDDs and DSC to further characterize these systems as
being single phase (single Tg). Also, described is pharmacokinetic data demonstrating greater than
the use of calculations based on Flory–Huggins sixfold increase in exposure of “compound 3”
theory in combination with experimental data for versus the crystalline drug (Table 10.15). With
the construction of phase diagrams for amorphous this example, this article also demonstrates the
systems with HPMCAS. These diagrams aid in benefit of spray-dried amorphous dispersions,
the understanding of the dispersed state of the particularly those containing HPMCAS, to the
drug in the polymer and provide valuable insight formulation of poorly water-soluble drugs.
into the physical stability of the system.
Accelerated stability studies are also described
Simvastatin with Povidone
to aid in the prediction of long-term physical
In a study published by Ambike et al., the authors
stability. Examples of systems showing physical
demonstrated the use of spray drying to produce
stability at ambient conditions on the order of
amorphous compositions of simvastatin (SIM), a
years are presented.
cholesterol-lowering agent (Ambike et al. 2005).
Solution properties of amorphous spray-dried
These authors found that by co-spray-drying SIM
dispersions are discussed at length with emphasis
with colloidal silica (Aerosil 200) and polyvinyl-
on distinguishing free drug from particulates and
pyrrolidone (PVP K30), the amorphous form of
further delineating the particulate species. It is
the low Tg (~35 C) compound could be
contended that understanding solution behavior
maintained for at least 3 months when stored at
of ASDDs is critical to distinguishing between
40 C and 75% relative humidity. The stability of
formulations in vitro and predicting in vivo per-
amorphous SIM at accelerated storage conditions
formance. Vastly improved dissolution perfor-
was attributed to decreased molecular mobility of
mance with amorphous HPMCAS spray-dried
the amorphous drug resulting from the anti-
dispersions over crystalline or neat amorphous
plasticizing effect of the polymer (composition
API is exemplified with a number of molecules.
Tg ¼ 115 C) as well as hydrogen bonding
This includes enhancement of dissolution rate as
between SIM and PVP. The presence of colloidal
well as extent and duration of supersaturation.
silica also improved stability by acting as a mois-
Most importantly, significant enhancement of
ture scavenger to prevent disruption of drug–
exposure in humans with HPMCAS spray-dried
polymer hydrogen bonding. Spray drying is key
dispersions is demonstrated for two compounds,
to these drug–excipient interactions because the
as shown in Fig. 10.15.
process is crucial to the formation of a molecu-
Similar to the article by Friesen and
larly disperse system. In such a system the drug
coworkers, Curatolo et al. (2009) demonstrate
has greater opportunity to interact with the poly-
the benefits of HPMCAS ASDDs to the formula-
mer, thereby decreasing drug–drug interactions
tion of insoluble drugs, using examples of various
and, consequently, the likelihood for SIM
compounds to illustrate key aspects. In this arti-
recrystallization.
cle, significant discussion is done on screening of
With respect to performance enhancement, a
precipitation inhibitors, and data are presented
fivefold increase in saturation solubility of SIM as
demonstrating the superiority of HPMCAS in
well as a substantial increase in the rate and extent
this capacity. The superior performance of the
of SIM dissolution was achieved from an ASDD
polymer in this regard stems from the attributes
containing SIM, Aerosil 200, and PVP in a 1:2:2
mentioned earlier – specifically, the ionizable and
(w/w/w) ratio. An in vivo study in rats revealed
amphiphilic nature of the polymer. Also, shown is
that the more soluble spray-dried SIM formula-
an example demonstrating the effect of drug load-
tion performed substantially better than pure SIM
ing on dissolution performance of ASDDs.
with respect to the lowering of total cholesterol
Shown in Fig. 10.16, these results clearly demon-
and triglyceride levels while increasing HDL cho-
strate the reduction of drug in solution resulting
lesterol (Fig. 10.17). This example therefore
from increased drug loading in the formulation.
demonstrates the use of spray drying to improve
Finally, the authors present human
Fig. 10.15 Comparison of human in vivo exposure for HPMCAS-M SDD (dosed at 300 mg fasted). Reproduced
crystalline drug (or crystalline drug in gelatin capsule) from Friesen et al. (2008) with permission from the Amer-
and SDDs: (a) 25 wt. % compound 1/HPMCAS-M SDD ican Chemical Society
(dosed at 30 mg fasted) and (b) 25 wt. % compound 3/
Fig. 10.16 Effect of drug 2000

loading on dissolution
Dissolved Drug Concentration (µg/mL)
performance of ASDDs 25% Compound 3:HPMCAS-MF

made with compound 3 and
HPMCAS-MF. 1500
Reproduced from Curatolo
et al. (2009) with
permission from Springer 33% Compound 3:HPMCAS-MF
1000
50% Compound 3:HPMCAS-MF

66% Compound 3:HPMCAS-MF
500
Crystalline Compound 3
0
0 50 100 150 200 250 300 350 400
Time (min)
the solubility and consequently the efficacy of a polymer, Eudragit L100–55, were produced by
cholesterol lower agent. spray drying and an emulsion–diffusion method,
respectively. The spray-dried microparticles were
produced by spray drying from a methanol solu-
Protease Inhibitor with Eudragit L100–55
tion using a Buchi Model 190 spray drier. Further
In a study published by De Jaeghere et al., amor-
details regarding the method are provided in
phous formulation technologies were utilized to
Method Capsule 10.1.
improve the bioavailability of a poorly water-
Both the spray-dried and nanoparticle
soluble (0.12 μg/ml) HIV-1 protease inhibitor
formulations showed a substantial improvement
known as CGP 70726 (De Jaeghere et al. 2000).
in the oral absorption of CGP 70726 in dogs over
Amorphous microparticles and nanoparticles of
the pure drug which did not generate quantifiable
the API (~ 20% w/w) with the stabilizing
Table 10.15 Compound 3 pharmacokinetics in fasted humans after dosing a 25% compound 3/HPMCAS-MF spray-
dried dispersion (n ¼ 4)
Formulation Dose (mg) Cmax (μg/mL) Tmax (h) AUC0-24 (μg*h/mL)
SDD suspension 300 8.4 1.1 2.5 0.6 46 7.6
Crystalline drug in suspension 300 1.3 0.3 2.3 1.3 7.4 3.3
Adapted from Curatolo et al. (2009)
Fig. 10.17 (a) Percent changes in serum total cholesterol levels of experimental groups at different time intervals. (b)
Percent increase in serum HDL-cholesterol levels of experimental groups at different time intervals. Reproduced from
Ambike et al. (2005) with permission from Springer
blood plasma levels (Table 10.16). The authors production of amorphous formulations in most
attributed the improvement in bioavailability to cases.
rapid dissolution of the amorphous API from the
high-surface-area particles that was specifically Ternary Systems of Itraconazole,
targeted to the primary site of absorption by the Copovidone, and a Surfactant
use of a carrier polymer with pH-dependent solu- Spray-dried amorphous dispersions of a poorly
bility. Interestingly, despite having a substantially soluble antifungal drug, itraconazole (ITZ), with
greater mean particle diameter (9.7 μm versus copovidone were described in articles by Janssens
286 nm), the spray-dried microparticles produced et al. (2008a, b). Copovidone was determined to
greater mean Cmax and AUC values when orally be an effective carrier polymer for ITZ with
administered to dogs in both the fed and fasted regard to miscibility and enablement of molecular
states. Owing to statistical insignificance, the dispersions with high drug loading. As is com-
authors offered no explanation as to the differ- mon to most solid dispersion systems, it was
ence. This result indicates that microparticles profound that increasing drug loading reduced the
duced by spray drying can provide equivalent, if rate and extent of dissolution. To overcome this
not greater, improvements in the oral absorption issue, a surfactant, Inutec SP1, was incorporated
of poorly water-soluble drugs than nanoparticles into the formulation, and ternary solid dispersions
produced by seemingly more sophisticated parti- were produced by spray drying (Janssens et al.
cle engineering methodologies. When also con- 2008a). The addition of the surfactant was found
sidering the relative simplicity and proven to improve the dissolution properties of the high
scalability of the spray-drying process, it is drug loading solid dispersions. However, ITZ and
clearly a much more viable technology for the Inutec SP1 were determined to be sparingly
Table 10.16 Pharmacokinetic parameter of GCP 70726 incorporated in Eudragit L100-55 pH-sensitive particles after
oral administration in dogs. Mean S.E.M (n ¼ 4)
Formulation Cmax S.E.M. (μmol/L) Tmax (h) AUC0-8 h S.E.M. (μmol/L)
Fasted state
Nanoparticles 1.62 0.04 2 5.83 0.77a
Microparticles 1.59 0.32 2 7.83 1.55
Fed state
Nanoparticles 0.86 0.21 1 2.00 0.50*
Microparticles 0.88 0.33 2 4.40 1.38
Adapted from De Jaeghere et al. (2000)
a
Statistically different (student test, P < 0.01)
miscible, and thus a crystalline ITZ phase was investigated the application of ASDD technology
observed in compositions with high drug loading toward the end of developing a solid dosage form
and surfactant content. It was therefore necessary with adequate bioavailability. Specifically, the
to reduce drug loading and optimize the polymer- authors investigated HPMCAS-MF and HPMC-
to-surfactant ratio in the carrier to yield a single- E5 as carriers in the ASDD formulations. A list of
phase dispersion. Ultimately, the optimum bal- the spray-dried formulations with corresponding
ance of drug loading and surfactant content was properties is provided in Table 10.17.
established, and a high-drug-loading solid disper- Dissolution screening of the four solid disper-
sion with improved dissolution properties was sion formulations was conducted at supersatu-
achieved. rated conditions in pH 6.8 phosphate buffer
This paper by Janssens et al. illustrates some compared to the micronized AMG-517 free base
key issues regarding formulation of ASDDs: first, as a control. The results revealed a two- to five-
the critical aspect of selecting a polymer in which fold increase in drug concentration with the solid
the drug has adequate miscibility to achieve the dispersion formulations compared to the crystal-
desired drug loading; second, the negative impact line drug. Also, the HPMCAS-MF formulations
of increasing drug loading on the dissolution exhibited superior dissolution performance over
properties of the system; third, the benefit of the HPMC formulations, and the 15% drug load-
incorporating a surfactant into a high drug load ing compositions performed better than the 50%
system with regard to enhancing wettability of the loading counterparts. It was noticed during these
system, increasing drug solubility, and ultimately dissolution studies that wetting of the solid dis-
improving dissolution performance; and, finally, persion powder was poor, resulting in gel forma-
optimizing drug loading and balancing carrier tion and poor dispersion in the media. The authors
composition to achieve a single-phase, stable resolved this issue by blending the ASDD powder
amorphous dispersion. Each of these aspects with a surfactant, sodium dodecyl sulfate (SDS),
should be carefully considered when formulating at a concentration of 5% by weight. The result
an amorphous spray-dried dispersion with high was a 12-fold increase in dissolution over the
drug loading. powder not containing surfactant.
In vivo performance of the AMG-517:
HPMCAS-MF (15:85) ASDD blended with
AMG-517 with HPMCAS and HPMC
SDS dosed in a capsule was evaluated in
AMG-517 is a VR1 antagonist indicated for the
cynomolgus monkeys against an “Ora-Plus” crys-
treatment of acute and chronic pain (Kennedy
talline suspension form of the drug which pro-
et al. 2008). The free base form of the drug is
duced the greatest exposures in preclinical
poorly soluble in aqueous media, having an inher-
studies. The results of this PK study are provided
ent solubility of <7.0 μg/mL in the physiologi-
in Table 10.18. From this study, the exposure of
cally relevant pH range. To overcome the
AMG-517 from the SDD capsule was found to be
solubility limitations of AMG-517, the authors
Table 10.17 Characterization summary for ASDD formulations of AMG 517 prepared from HPMCAS-MF and HPMC
E5 by spray drying
Formulation Drug Yield LEa d50b % wt. loss by Tg ( C) by
ID Polymer (wt %) (%) (%) (μm) TGA MDSC
A, lot 1 HPMCAS- 15 76 34.2 2.39 106
MF
A, lot 2 15 86 98.8 35.3c 0.32c 101c
B 50 55 97.5 40.7 2.84 98
C HPMC E5 15 61 99.1 34.6 1.94 117
D, lot 1 50 12d 99.0 34.6 1.88 106
D, lot 2 50 60 28.7 1.44 107
Adapted from Kennedy et al. (2008)
a
Load efficiency (LE)
b
Particle-size diameter at 50% cumulative volume %
c
For “A, Lot 2,” results are provided for material after secondary drying in a vacuum oven
d
Inlet drying temperature was 45 C
Table 10.18 AMB 517 PK parameters and summary statistics following oral administration at 12.5 mg/animal to male
cynomolgus monkeys
Formulation Tmax (h) Cmax (ng/mL) AUC0-inf. (ng*h/mL) Frel (%)
AMG 517 Ora-Plus suspension 1.5 (1.0–2.0) 1020 (189) 40,800 (10,300) 100
AMG 517 ASD in capsule 2.0 (1.0–2.0) 1480 (309) 66,600 (13,200) 163
20.9
Adapted from Kennedy et al. (2008)
163% of that provided by the Ora-Plus suspen- the use of ASDD technology, with proper down-
sion formulation. Therefore, the SDD capsule stream processing, for improving the exposure of
oral dosage form was selected for use in clinical a poorly water-soluble molecule and enabling
studies. For additional information regarding the development of a highly bioavailable solid oral
methods used in the development of the dosage form for use in clinical studies.
HPMCAS-MF ASDD system, see Method Cap-
sule 10.3. Albendazole with HPMCP and HPMC
This example not only illustrates, once again, As demonstrated by previous examples, anionic
the superiority of HPMCAS as a carrier for SDD polymers can be particularly effective as carriers
formulations but also describes the common wet- in ASDD systems owing to their ability to target
ting problem with these systems. The hydropho- supersaturation to the neutral pH environment of
bic nature of HPMCAS and similar polymers in the intestinal tract and maintain elevated free drug
the unionized state often leads to gel formation in concentrations for the duration of intestinal tran-
acidic aqueous media. In capsules, this phenome- sit. An early example of this concept was
non can lead to plug formation in which the provided in a paper by Kohri et al. (1999) in
powder does not disperse from the capsule shell. which the authors demonstrated the use of a
As in this study, the result is a substantial reduc- dual-polymer carrier system consisting of
tion in the dissolution of the API from the ASDD HPMCP and HPMC for improved oral delivery
system. Externally blending the ASDD with of albendazole, a poorly water-soluble, weakly
surfactants such as SDS or docusate sodium, as basic drug.
was done in this study, is a highly effective solu- Amorphous solid dispersions of albendazole/
tion. Incorporating surfactants directly into the HPMCP/HPMC (2:1:1 w/w) were produced by
ASDD formulation is another strategy often the solvent evaporation method from an ethanol–
employed. Ultimately, this article demonstrates dichloromethane solvent system. Using a gastric
Fig. 10.18 Dissolution

behavior of albendazole
from solid dispersions in
media of pH 1.2–6.5;
( filled square) physical
mixture; ( filled circle) solid
dispersion with HPMC and
HPMCP; (open circle) solid
dispersion with HPMC;
(open square) solid
dispersion with HPMCP.
Adapted from Kohri et al.
(1999)
transfer dissolution method, these authors were results, it was concluded that crystalline
able to demonstrate that according to its weakly albendazole dissolved moderately well under nor-
basic nature, albendazole dissolves extensively in mal gastric conditions, but minimally when gas-
acid but then rapidly precipitates following the tric acidity was reduced. On the other hand, the
transition into neutral pH. Conversely, the amor- amorphous dispersion formulation was able to
phous dispersion with the HPMCP:HPMC carrier generate relatively high levels of dissolved
showed minimal dissolution in acidic media albendazole irrespective of the gastric conditions.
followed by rapid release in neutral media and Therefore, exposure was much lower for the
prolonged supersaturation. The results of this dis- physical mixture when gastric acidity was
solution study are shown in Fig. 10.18. reduced, whereas exposure was less affected by
The in vivo performance of the dual-polymer the change in gastric conditions for the solid
solid dispersion system was then evaluated in two dispersion formulation. Hence, bioavailability
groups of rabbits: (1) rabbits of normal gastric enhancement from the amorphous albendazole/
acidity (pH approximately 1) and (2) rabbits HPMCP/HPMC system was a result of the
with low gastric acidity (pH > 5). Crystalline carrier’s ability to generate and maintain super-
albendazole physically mixed with lactose was saturated concentrations of albendazole in neutral
used as a control for this study. The results are pH environments. These authors thus
provided in Table 10.19. In both groups, the solid demonstrated a delivery system for albendazole
dispersion yielded higher exposures versus the that is unaffected by fluctuations in the gastric
control; however, the difference was most pro- environment. This represents a significant
nounced in the group with low gastric acidity improvement to therapy with albendazole as the
where the AUC from the solid dispersion formu- variability inherent to oral delivery was substan-
lation was greater than three times that of the tially reduced.
physical mixture. Considering the dissolution
Table 10.19 Pharmacokinetic parameters after administration of albendazole (5 mg/kg) to rabbits in a crossover study
Parameter Normal gastric acidity Low gastric acidity
Physical mixture Solid dispersion Physical mixture Solid dispersion
Cmax (μg/mL) 2.1 0.2 2.4 0.4 0.5 0.2 1.4 0.2*
Tmax (h) 40 0.8 3.7 0.7 6.3 1.6 6.7 1.1
AUC0-24 h (μg*h/mL) 22.4 3.5 31.8 4.7* 6.4 2.6 24.8 5.8*
t½ (h) 2.8 0.4 5.8 1.7 4.8 1.6 4.6 0.8
MRT (h) 7.7 1.1 10.6 2.0* 12.0 3.2 13.3 2.3
Bioavailability (%) 67.6 17.9 94.9 12.9* 21.3 9.9 68.8 13.2*
Adapted from Kohri et al. (1999)
Table 10.20 Pharmacokinetic parameters of CsA after oral administration of CsA products to dogs
Parameters Cmax (μg/mL) Tmax (h) AUC (μg*h/mL)
CsA powder alone 0.46 0.15 2.13 0.48 2.81 1.16
Sandimmune® 0.72 0.27* 1.50 0.45 3.62 1.63
CsA microspheres 0.71 0.08 1.88 0.48 4.85 0.98*
Adapted from Lee et al. (2001)
*P < 0.05 by Duncan method of ANOVA when compared to CsA powder
a
Data are expressed as mean S.D. (n ¼ 4)
This example aptly demonstrates the utility of evaluation of in vivo performance against pure
HPMCP in amorphous solid dispersion systems CsA powder and Sandimmune in a single-dose
as a concentration enhancer in neutral pH oral administration study in dogs. The pharmaco-
environments. It also illustrates the corresponding kinetic data from this study are provided in
improvement to the oral absorption of poorly Table 10.20.
water-soluble drugs. This is an early example of These results revealed that both the spray-
perhaps the most widely used ASDD formulation dried dispersion and commercial CsA
approach for oral delivery of poorly soluble formulations produce substantial improvements
molecules today. Since HPMCP is thermally in oral absorption over the pure drug powder.
labile, spray drying is perhaps the best process Improved absorption from the spray-dried formu-
to apply to the manufacture of amorphous lation versus Sandimmune was also observed,
systems containing this polymer as exposure to although the differences were not statistically sig-
thermal stress is minimal. nificant. The similarity in oral absorption pro-
duced by the spray-dried formulation versus the
Cyclosporine with a Nonpolymer Carrier commercial formulation represents a significant
In a study conducted by Lee et al. (2001), a spray- potential benefit to CsA therapy as the spray-
dried solid dispersion formulation utilizing a dried formulation is in a preferred solid form
nonpolymeric carrier was developed for cyclo- and produced by a more straightforward process.
sporine A (CsA) and evaluated against the com- Owing to the inherent disadvantages of liquid
mercial product Sandimmune®, a microemulsion- versus solid formulations and the complicated
based soft gelatin capsule formulation. nature of the soft gelatin encapsulation process,
Microspheres of CsA with dextrin and sodium spray drying is a more viable means of improving
lauryl sulfate (SLS) were produced by spray dry- CsA therapy via formulation. Hence, this study
ing and optimized in terms of drug–excipient exemplifies the potential opportunities to simplify
ratio by dissolution testing. The optimum existing commercial products and their
CsA/SLS/dextrin ratio with respect to in vitro corresponding processes by applying spray-dried
drug release was found to be 1/3/1 (w/w/w). solid dispersion technology.
This formulation was selected for comparative
10.9.11 Examples from Industry an ASDD approach was employed to achieve

adequate bioavailability. US patent application
Currently, there is only one pharmaceutical prod- 14/168,264 describes an ASDD of LED and
uct on the market formulated as a spray-dried copovidone in a 1:1 (w/w) ratio produced by
amorphous dispersion. However, there are several spray drying (Chal et al. 2014). The patent speci-
examples of drug molecules formulated by this fication also indicates that spray drying of the
technology which have reached late-stage devel- LED/copovidone (1:1 w/w) composition is most
opment. It is important to note that the increase in efficient from solution in ethanol, generating high
the number of poorly water-soluble molecules in product yields (>88%) over a wide range of outlet
development pipelines is a relatively recent event, temperatures (30 – 90 C). It was also noted that
and thus amorphous drug-delivery systems, par- LED exhibited good chemical stability in ethanol.
ticularly in commercial use, are also in their Substantially lower process efficiency was
infancy. Also, considering the lengthy drug observed when spray drying this ASDD formula-
development process and the frequency at which tion from dichloromethane with reported yields as
programs are halted, it is not surprising that few low as 44% and with presumably poor chemical
spray-dried amorphous formulations have been stability of the drug in solution. The LED ASDD
commercialized. The absence of commercialized formulation was reported in the above referenced
products is certainly not an indication of the lim- patent application as providing “improved in vivo
ited commercial viability of spray drying or amor- and in vitro performance and manufacturability/
phous delivery systems. In this section, the scalability relative to the other formulation
commercial and late-stage clinical success of approaches.” The improvement of LED bioavail-
ASDD systems is highlighted by reviewing a ability via an ASDD approach was critical not
few key examples. only with respect to the efficacy of the drug but
also with regard to enabling a single-tablet fixed-
dose combination with the high-dose SOF,
Ledipasvir
providing excellent convenience to the patient
Ledipasvir (LED) is a hepatitis C virus (HCV)
with a single-tablet, once-daily dosing schedule.
NS5A inhibitor developed by Gilead. The drug
Hence, the LED example is one that illustrates
was FDA approved on October 10, 2014, as a
enablement of a breakthrough cure for a wide-
fixed-dose combination product with sofosbuvir
spread disease with optimum patient convenience
(SOF), an NS5B polymerase inhibitor, under the
via the application of an ASDD formulation
trade name Harvoni®. The product contains
approach.
90 mg LED and 400 mg SOF in a single-tablet
dosage form. Harvoni is indicated for the treat-
Velpatasvir
ment of chronic HCV genotypes 1, 4, 5, or 6 and
Velpatasvir (VEL) is an investigational
represents a breakthrough in the treatment of the
pan-genotypic NS5A inhibitor that is being devel-
disease as the first combination oral product
oped by Gilead as a fixed-dose combination with
approved to treat chronic HCV without
SOF for the treatment of chronic genotype 1–6
co-administration of interferon injections or riba-
HCV (Gilead 2016). The US FDA granted prior-
virin (Harvoni 2015).
ity review of the VEL/SOF NDA on January
Harvoni contains LED in the form of an
4, 2016, based on its designation as a “Break-
ASDD (Martin 2015). LED is a poorly water-
through Therapy.” VEL is incorporated into the
soluble, weakly basic molecule that is slightly
single-tablet fixed-dose combination product in
soluble (1.1 mg/mL) below pH 2.3 and practically
the form of an ASDD (Martin 2015). US patent
insoluble (0.1 mg/mL) in a pH range from 3.0 to
application 14/168,313 describes an amorphous
7.5 (Harvoni 2015). The pH solubility profile of
composition of VEL and copovidone in a 1:1
the molecule limits the available free drug for
(w/w) ratio produced by spray drying from an
absorption in the intestinal environment; hence,
ethanolic solution at 20% dissolved solid content
(Gorman et al. 2015). The application further by Merck & Co. indicated for the treatment of
describes the in vitro dissolution performance of HIV/AIDS. Development of this product is in
the ASDD formulation as being “the fastest and Phase 3 clinical trials at the time of this writing.
most complete dissolution of all of the As described in WIPO filing WO/2015/077273, a
formulations that were tested.” While no pharma- MK-1439:HPMCAS (1:4 w/w) amorphous solid
cokinetic comparisons between the ASDD formu- dispersion formulation was produced by spray
lation and a crystalline control were provided in drying from a solution of acetone/water
the application, it is assumed that the exceptional (Lowinger et al. 2015). The resulting drug
dissolution performance correlates to improved products demonstrated good physical stability
oral bioavailability, hence justifying the utiliza- with no recrystallization for up to a 30% drug
tion of an ASDD approach for VEL. Much like load. The “wet hold time” between initial spray
the LED example, the VEL ASDD demonstrates drying and secondary drying was evaluated for
the criticality of spray-drying technology to the impact on physical characteristics. From physical
advancement of HCV drug therapy in its characterization data, it was determined that a
enablement of efficacious oral delivery of NS5A difference of greater than 20 C between wet Tg
inhibitors. and storage conditions was critical for
maintaining amorphous stability of the product.
Everolimus Thus, it was important that the resulting spray-
Everolimus (EVE) is a macrolide immunosup- dried material had an initial wet Tg at least 20 C
pressant indicated for the prophylaxis of organ higher than the highest storage temperature. Bio-
rejection in adult patients receiving kidney or pharmaceutical evaluation of a tableted form of
liver transplants (Zortress 2015). EVE was devel- the spray-dried product was conducted in fasted
oped and commercialized by Novartis under the beagle with AUC24 and Cmax both approximately
trade name Zortress® and approved by the FDA doubled over encapsulated crystalline MK-1439
on February 15, 2013. US 6,004,973 is an FDA with the Tmax remaining the same at 4 hours.
Orange Book listed patent for Zortress that These results demonstrate the potential viability
describes a solid dispersion of EVE/HPMC for a spray-dried amorphous MK-1439 product
E3/lactose in an approximate weight ratio of 1: with improved bioavailability and sufficient phys-
9:1 (Guitard et al. 1999). Per Example 1 of the ical stability.
patent specification, the composition is formed by
dissolving the active with the carrier medium in Etravirine
an absolute ethanol/acetone cosolvent mixture in Etravirine (TMC125) is an HIV-1-specific,
a 1:1 ratio by weight with subsequent evapora- non-nucleoside reverse transcriptase inhibitor
tion. The resulting dried solids are further milled indicated in combination with other antiviral
and processed into tablets with dose strengths of agents for the treatment of HIV-1 infection.
0.25 mg, 0.5 mg, and 0.75 mg. The patent docu- Etravirine was granted accelerated approval by
ment provides results of a PK study in rats com- the US Food and Drug Administration in January
paring the above composition to three other 2008 followed by traditional approval in
formulations that demonstrate up to a tenfold December 2009. The product is marketed by
increase in oral exposure (AUC) by this formula- Tibotec Pharmaceuticals under the brand name
tion approach. These results thus signify the sub- Intelence® and is available in tablets of 100 and
stantial bioavailability enhancement of EVE 200 mg strengths. Additional approval was
achieved by the application of ASDD technology granted in March 2012 for treatment-experienced
which enabled commercialization of Zortress. patients and made available 25 mg tablets for
weight dosing.
Doravirine Etravirine is poorly water soluble and exhibits
Doravirine (MK-1439) is a non-nucleoside low to moderate permeability and hence is classi-
reverse transcriptase inhibitor under development fied as a BCS IV compound (Kakuda et al. 2008).
The dose selected based on phase IIb trials was the range of the 800 mg twice-daily dose with
800 mg twice daily delivered from a tablet formu- the phase II formulation. This represented a four-
lation produced by conventional granulation fold decrease in dose with corresponding
(Kakuda et al. 2008). Concurrent to phase II decrease in pill burden. Based on this study, the
studies, a formulation development program was spray-dried tablet formulation was selected for
conducted with the aim of reducing dose and phase III clinical trials. Eventually, the amor-
corresponding pill burden (Kakuda et al. 2008). phous spray-dried tablet formulation of etravirine
From this program, a spray-dried amorphous became the final marketed formulation (EMA
solid dispersion formulation was developed, the 2008). This example demonstrates the commer-
details of which are provided in patent application cial viability of amorphous spray-dried
WO2007141308 (Kiekens et al. 2007). This pat- dispersions and the potential enhancement of
ent application describes a spray-drying process therapies with poorly water-soluble drugs with
in which etravirine is spray dried with HPMC respect to reduction of dose and pill burden.
2910 5 mPa.s from solution in a cosolvent system
of dichloromethane and ethanol. Also, described
Telaprevir
is the inclusion of microcrystalline cellulose
Telaprevir, also known as VX-950, is a hepatitis
suspended in the feed solution, presumably for
C virus protease inhibitor being developed by
the purpose of increasing the bulk density of the
Vertex Pharmaceuticals in association with
powder.
Tibotec and Mitsubishi Tanabe Pharma. It was
The tablet developed from this solid dispersion
approved by the FDA in May 2011 under the
formulation was found to yield a ninefold
brand name Incivek® and is given as two
increase in exposure over the phase II formulation
375 mg tablets three times daily. Owing to insol-
in a single-dose pharmacokinetic study in healthy
ubility of the molecule in water, oral formulations
subjects (Scholler et al. 2005). In HIV-1-infected
of telaprevir containing a crystalline form of the
patients, tablets of 100 and 200 mg strength were
API do not provide sufficient exposure to achieve
evaluated against 800 mg of the phase II formu-
therapeutic efficacy. Absolute bioavailability fol-
lation with twice-daily administration (Kakuda
lowing oral administration of micronized crystal-
et al. 2008). The pharmacokinetic parameters
line telaprevir in rats is less than 0.5% (Cui et al.
generated from this study are shown in
2006).
Table 10.21. From these results, it was deter-
In order to improve the solubility
mined that 200 mg twice daily of the spray-
characteristics and oral bioavailability of
dried tablet formulation produced exposures in
telaprevir, an ASDD formulation strategy was
Table 10.21 Pharmacokinetic parameters of etravirine after a single (day 1) administration of phase III formulation at
doses of 100 and 200 mg (test) and phase II formulation at a dose of 800 mg (reference)
Phase II formulation 800 mg Phase III formulation Phase III formulation
Parameter twice daily (reference) 100 mg twice daily (test) 200 mg twice daily (test)
Number of patients, 32 33 27
n
Median tmax, 4 (2–8) 4 (2–6) 4 (3–8)
h (range)
Mean Cmax, ng/ml 70.6 (72.7) 54.9 (54.0) 125.9 (96.0)
(SD)
Mean AUC0-12h, 434 (437) 312 (331) 745 (660)
ng*h/ml (SD)
LSM
Cmax (90% CI) – 0.81 (0.65–1.00) 1.97 (1.59–2.45)
AUC0-12h (90% CI) – 0.72 (0.59–0.88) 1.91 (1.54–2.36)
Adapted from Scholler et al. (2005)
Fig. 10.19 Rat PK study

with various telaprevir
spray-dried amorphous
dispersion formulations.
Reproduced from Cui et al.
(2006)
evaluated. US patent application 2006/0089385 discussion of the FSD processes used in this
describes compositions in which telaprevir is dis- example, see Method Capsule 8.4. With the appli-
persed in an amorphous state in a cation of FSD technology to the spray-dried for-
pharmaceutically acceptable polymer by spray mulation of telaprevir, significant improvement in
drying. Examination of the application reveals manufacturing efficiency of the final tablet dos-
that the preferred compositions contain the amor- age form is achieved. Thus, in the case of
phous drug in a single-phase system with a cellu- telaprevir, spray-drying technologies improved
losic polymer, i.e., hypromellose or hypromellose not only bioavailability but also drug product
acetate succinate, along with a solubility- manufacturing.
enhancing surfactant, i.e., sodium lauryl sulfate
or vitamin E TPGS. Figure 10.19 shows results of
Ivacaftor
a formulation screening study conducted in rat
Ivacaftor (VX-770), developed by Vertex
models which appears to indicate that the
Pharmaceuticals, is indicated for the treatment of
HPMC/vitamin E TPGS formulation yields the
cystic fibrosis and was approved by the FDA
greatest exposure. This patent application claims
under the name Kalydeco®. The molecule
enhanced bioavailability with the amorphous dis-
exhibits solubility-limited absorption with an
persion composition relative to the crystalline
oral bioavailability of only 3–6% in rats when
drug, and although the extent of this improvement
administered in a crystalline form (Hurter et al.
is not known, the progression of the molecule into
2011). A solubility-enhanced form of ivacaftor is
phase III clinical trials is a good indication that
therefore required to achieve therapeutic efficacy
the formulation has enabled therapeutically effec-
at a reasonable dose.
tive oral delivery of telaprevir.
A series of formulation approaches were
Additionally, US patent application 2010/
evaluated to enhance the bioavailability of
0011610 describes FSD of telaprevir solid
ivacaftor, including a spray-dried amorphous dis-
dispersions. The FSD process produced dry mate-
persion concept. Solubility studies conducted in
rial with significantly greater particle sizes and
FaSSIF media revealed that amorphous ivacaftor
bulk powder densities over traditional spray dry-
produced a solubility value of 67.4 μg/ml, which
ing, thus enabling direct compression of the end
marks a significant solubility improvement over
product (Bittorf et al. 2010). For further
crystalline polymorph form B (1 μg/ml). The
Table 10.22 Pharmacokinetic parameters of various forms of ivacaftor in rats

Drug form Dose (mg/kg) AUC (μg*h/mL) Tmax (h) %F
85% amorphous drug 50 135.5 27.6 6.0 0.0 95.0 20.0
200 371.9 46.1 6.0 0.0 61.0 7.0
Crystalline drug in aqueous vehicle 50 8.0 1.2 4.0 0.0 5.5 0.8
200 16.9 3.0 4.7 1.2 3.1 0.3
Crystalline drug in PEG 50 135.1 43.0 5.5 1.0 74.0 23.0
200 431.5 101.1 14.5 11.0 67.0 16.0
Solid dispersion 25 90.1 8.1 6.0 0.0 111.0 10.0
100 260.8 28.4 6.0 0.0 109.0 12.0
Adapted from Hurter et al. (2011)
ASDD concept utilizing HPMCAS as a carrier formulation was produced by spray drying from
was then developed to employ the solubility solution in acetone (Beyerinck et al. 2005). This
improvement provided by the amorphous com- “solubility-improved” form of torcetrapib was
pound in a stable and viable formulation. As incorporated into an osmotic pump-controlled
shown in Table 10.22, the solubility enhancement release tablet and coated with an immediate
provided by the amorphous form of the molecule release layer of atorvastatin, as described in
resulted in a substantial increase in oral bioavail- WIPO filing WO/2006/082500 (Berchielli et al.
ability over the crystalline API in aqueous and 2006). This combination therapy was found to be
PEG vehicles. The ASDD formulation of effective in lowering cholesterol levels during
ivacaftor in HPMCAS generated exposures in clinical trials; however, due to increased mortality
excess of 100% relative bioavailability at both rates and cardiovascular events, the program was
25 and 100 mg/kg doses. This marked a signifi- halted late in development.
cant improvement in oral absorption over the
crystalline compound in both vehicles and a mod- Elbasvir and Grazoprevir
erate improvement over amorphous drug, partic- The combination of elbasvir and grazoprevir was
ularly at higher doses. This study revealed that the developed to treat hepatitis C virus as direct-
ASDD formulation concept substantially acting antivirals targeting the NS5B inhibitors
increased exposure of ivacaftor and enabled ther- and NS3/4 protease inhibitors, respectively. The
apy at lower doses with a stabilized amorphous oral absorption of the drugs was enabled by
form of the drug. The successful commercializa- spray-dried dispersions. The drugs are formulated
tion of this molecule demonstrates yet again the individually in spray drying using different
value of spray drying with respect to generating polymers and then mixed for the drug product in
viable drug products from poorly water-soluble tablet form (McKelvey and Kesisoglou 2019). In
molecules. the case of grazoprevir, the spray-dried dispersion
was formulated using copovidone and a surfac-
Torcetrapib tant, SLS. The effect of drug concentration on
Torcetrapib is another example of a developing oral absorption was studied in grazoprevir solid
drug molecule whose therapeutic efficacy was dispersion comparing with neat amorphous API
enabled by formulation as an amorphous solid (Table 10.23). According with the filing patent
dispersion system by spray drying. Torcetrapib US 2016/0339074, the oral adsorption of
is a cholesteryl ester transfer protein inhibitor grazoprevir in neat amorphous showed to be
which was being developed by Pfizer Inc. before 25% lower than the solid dispersion comprising
the program was terminated during phase III clin- 33% of API (used as reference) (Marota et al.
ical trials. As described in Example I of WIPO 2015). Furthermore, the concentration of API
filing WO/2005/011636, a torcetrapib/HPMCAS present in SDD was found to be critical: for an
(1:3 w/w) amorphous solid dispersion increase from 33% to 40%, the exposure is
Table 10.23 Human pharmacokinetic results from bio-comparison study of grazoprevir formulations at different drug
concentrations (Marota et al. 2015; McKelvey and Kesisoglou 2019)
Drug load in spray-dried Drug load in drug Relative Relative
Spray-dried formulation intermediate product AUC0,1 Cmax
Amorphous drug 100% 70% 0.15 0.05
Amorphous drug 100% 69% 0.23 0.11
Solid dispersion in copovidone 40% 15% 0.56 0.52
and SLS
Solid dispersion in copovidone 33% 10% 1 (reference) 1
and SLS (reference)
Table 10.24 Chemical stability comparison of the SDD systems at 40 C and 75% relative humidity in open glass vials
for 8-day period (Fry et al. 2017)
HPLC area %
Tested SDD with 30% drug load 0 day 4 days 8 days
Solid dispersion in PVP-VA 99.45 99.21 99.35
Solid dispersion in Eudragit L100 98.63 96.10 93.16
Solid dispersion in HPMCP H-55 97.30 93.03 86.63
Solid dispersion in CAP 95.45 90.89 87.15
reduced approximately by half which shows an polymer selection, solid dispersion with 30%
impact of increasing drug loading in SDD. was formulated with different polymers in a
The elbasvir compound was classified as BCS lab-scale unit to assess the SDD stability and
class IV with a weakly basic nature. The preva- properties. Similar solvent system (mixture of
lence of acid-reducing medication in hepatitis C THF and methanol) was used for the comparison
virus patients turns the formulation development at 5% w/w solids concentration. The stability
critical when considering the API properties. The assessment was performed in accelerated
application of solid dispersion was considered to conditions of 40 C and 70% of relative humidity
overcome such challenge without restrictions of under open conditions for a period of 8 days. The
acid-reduction agents. The elbasvir spray-dried systems produced remained amorphous for
dispersion formulated with HPMC also enabled period analyzed. In terms of chemical stability,
to reduce the food effect observed in neat amor- it was detected an impurity growth (as shown in
phous drug, as further discussed in McKelvey and Table 10.24), more specifically carbamate impu-
Kesisoglou (2019). rity which may be linked to the acidic properties
The combo developed by MSD was approved of some polymers (Fry et al. 2017). A negligible
by FDA in 2016 and is commercially available by impact is observed for copovidone (PVP-VA)-
brand name Zeptier®. based SDD which was chemically and physically
stable for the considered period.
Tucatinib Tucatinib is commercially available by brand
In 2020, FDA approved tucatinib for patients name Tukysa®.
with HER2-positve metastatic breast cancer. Dur-
ing the development of tucatinib, it was identified
at least 29 polymorphic forms and other 7 solvates 10.10 Spray Drying for Lung Delivery
forms (Gala et al. 2020). Nonetheless, the drug
was found to be thermodynamically stable. Pulmonary drug delivery has long been employed
According to the application US 2017/0136022, for the treatment of respiratory diseases. It is
the drug was then developed based in solid dis- generally accepted as the default route of admin-
persion produced by spray drying. As part of istration for the treatment of asthma and chronic
obstructive pulmonary diseases. The lung is very diameter. Given that low-density or hollow
attractive for drug delivery due to its large surface particles are advantageous for inhalation, several
area (up to 100 m2) and low enzymatic activity formulation technologies were developed to pro-
which provides a controlled environment for sys- duce such materials by spray drying. Among
temic absorption of medications (Labiris and those, PulmoSphere® is of particular interest.
Dolovich 2003). This is especially relevant for This technology involves spray drying a
proteins that are often subjected to rapid enzy- fluorocarbon-in-water emulsion stabilized by a
matic degradation. Therefore, research in this long-chain phospholipid, e.g.,
field has attracted many pharmaceutical dipalmitoylphosphatidylcholine – pulmonary
companies, and many advances have been made endogenous surfactant (DPPC) (Tarara et al.
in all areas of development related to this field, 2007; Weers 2000). The sudden removal of the
i.e., particle engineering, formulation fluorocarbon constituent leaves a hole in the par-
technologies, and delivery devices. Despite the ticle surface. In this way, perforated hollow
lack of commercial success of Pfizer’s product particles are formed as depicted in Fig. 10.20.
Exubera® (inhalable insulin), its development Their density is very low (less than 0.05 g/ml),
was a major breakthrough with respect to particle and their aerodynamic properties are quite
engineering and formulation for inhalation amendable to inhalation (Vehring 2008; Weers
products. This section will be focused on parti- 2000).
cle/powder characteristics that are important to Other technology to improve aerodynamic
inhalation and how spray drying can be a useful performance and porosity is proposed by Civitas
technology with respect to particle design. Therapeutics with application of levodopa drug
The performance of powders for inhalation is product as described in US 7,008,644 (Batycky
related mainly to (1) aerodynamic particle size et al. 2002). The method proposes the use of two
and (2) powder dispersibility (Weiler et al. streams in which the one stream comprises the
2010). Product performance can be assessed by dissolved drug. Both streams are then combined
in vitro determination of fine particle fraction using a static mixer located immediately before
(FPF), for instance, by the use of an Andersen the nozzle. In the tailored static mixer, the hydro-
cascade impactor. The typical particle size for phobic and hydrophilic components from the two
inhalation products is in the range of 1–5 μ. How- streams may be mixed avoiding problems of
ever, more important than the absolute particle material compatibilities. This is further used to
size is the aerodynamic diameter which takes optimize the porosity of the product by the use of
into account particle density and shape. For fur- volatile salts. The salt carbonation of components
ther details on particle design for inhalation and as ammonium carbonate results in the nucleation
the application of spray-dried particles, see of carbon dioxide inducing gas and liquid phase
Vehring (2008). within the droplet. The heterogeneous phase
Particles produced by spray drying possess results in high-porosity particles after the drying
inherently better properties than particles pro- completeness.
duced by common top-down technologies Besides particle size, spray drying also enables
(micronization processes). They tend to be spher- the formation of different particle morphologies,
ical, having less points of contact between namely, by tuning the corrugation of the particle
particles when compared to the flat surfaces of surface. By tuning particle corrugation, a smaller
micronized particles. Therefore, with spray-dried radius of curvature in the contact zone between
powders, cohesive forces are lower, dispersibility particles can be achieved, thereby reducing cohe-
is better, and free particle fractions (FPF) are sion forces and improving powder dispersibility
greater (Weiler et al. 2010). Also, spray-dried (Weiler et al. 2008). This optimization may be
powders tend to present densities less than performed by adjusting some spray-drying pro-
0.5 g/mL and, in some cases, even lower than cess parameters such as drying temperature and
0.1 g/mL which improves the aerodynamic feed concentration that have a direct impact on
Fig. 10.20 PulmoSphere™ solid foam particles. Reproduced from Vehring (2008) with permission from Springer
the Peclet number (namely, on the evaporation nozzles are unable to produce the required droplet
rate). An increase in both parameters will acceler- size distribution, even for very high pressures
ate the evaporation rate and shell formation which (200 bar). However, in many cases even using
will lead to a more spherical and less corrugated high atomization ratios, the particle-size distribu-
particle. tion obtained is too coarse, and feed concentration
In order to control particle size, density, and must be tuned. It is not uncommon to use feed
morphology by spray drying, all components concentrations between 1 and 2% w/w in spray-
must be in solution. This typically forms an amor- drying processes for inhalable products. Obvi-
phous product; thus, the use of stabilizing ously, the throughput will be reduced with such
excipients (e.g., sugars, polymers) is required in low feed concentrations, but that is an unavoid-
many cases. The spray-dried particles of able compromise.
Exubera® exemplify such a case. They consisted The collection of these small and low-density
of insulin and a buffered composition of particles is also a challenge with traditional
stabilizers comprising sodium citrate, glycine, cyclone systems as yields lower than 60% are
and mannitol (Vehring 2008). Therefore, the common. For high-value products, namely,
concepts of spray-dried solid dispersion stabiliza- proteins, this is unacceptable from a commercial
tion are also applicable to spray-dried particles for perspective. Therefore, optimized cyclone
inhalation. However, the selection of stabilizers designs, or even the use of filter bags as a collec-
must take into account the specificities of lung tion system, are required.
delivery, namely, approval for this mode of
administration and safe dosing levels.
The most challenging aspects in the develop-
10.11 Emerging Technology Trends
ment of a spray-drying process for an inhalable
and Applications
powder are (1) the generation of very fine
particles and (2) process yield. Regarding the
10.11.1 Innovative Approaches
former, the use of two-fluid nozzles or even
high-pressure two-fluid nozzles with very high
Although spray drying is a mature technology in
atomization ratios (in many cases above 10) is
pharmaceutical field, continuous innovation has
usually necessary. In this respect, pressure
been promoted to overcome the process then fed and atomized using a flash nozzle
limitations and challenge. (described in section “Atomizer selection”). The
Spray drying is a continuous operation, never- application of coprecipitations principles is an
theless, enclosed by two batch operations: solu- alternative approach. As described in US
tion preparation and secondary drying. The 10,821,375, a continuous process of
implementation of a continuous solution prepara- coprecipitation using microfluidization enables
tion may be key not only to debottleneck the SD to control the particle attributes and micro-
feed system but also to address and/or mitigate and/or nano-range material. The solvent, in
potential solution stability concerns and ensure which the drug and polymer are dissolved, and
product quality. Vicente et al. proposed a method the antisolvent are mixed in reaction chamber.
to enable the continuous solution preparation in The resulting stream appears as a suspension
which the solids and solvents are mixed through which passes through a filtration system prior to
microfluidization (Vicente et al. 2018). The use of the spray drying wherein the product is isolated in
microreactors intensifies the mixture enhancing solid form. An advantage of the proposed appli-
the dissolution kinetics of the materials by cation is the concentration step by a filtration
particle-size reduction. Moreover, the method system prior to the spray drying which decreases
enables to increase the temperature of solution the costs and energy requirements during
without applying any external heat energy source. isolation.
On the other hand, the application of a contin-
uous process for secondary drying step may be
challenging. This is due to the low solvent content 10.11.2 Spray Congealing
required by the ICH guideline and the diffusion
limits of the particles. Different methods have Spray congealing consists of atomizing a molten
been used to accelerate the drying rate as the mixture into fine droplets followed by contact
application of a stripping gas stream. In fact, US with a cold gas to force rapid solidification. The
9,265,731 method proposed the use of a process is similar to spray drying, and, in fact, the
humidified stream to enhance the molecular equipment required to perform both processes is
mobility and consequently expedite the drying basically the same. The advantage of the spray-
process (Ray et al. 2005). congealing process over spray-drying one is that
Other challenge of spray drying is in the appli- it is solvent-free and provides higher throughputs.
cation of compound known as brick-dust The final product also presents clear advantages.
compounds. Such compounds are characterized The absence of solvent evaporation results in
by a high melting temperature (over 200 C) and nonporous, smooth spherical particles, as seen in
poor solubility in aqueous and organic systems. Fig. 10.21. Consequently, the product typically
Consequently, the drug formulation into an ASD contains large particles (250–2000 μ) and high
is not feasible by hot-melt extrusion and not cost- densities resulting in free flowing powders suit-
effective by spray drying. Temperature-shit pro- able for downstream processing, such as direct
cess may be a solution for such compounds since compression (Patel and Chen 2001).
it increases the process solids concentration and Spray congealing consists of preparing a mol-
therefore solids throughput. In such method ten mixture comprising the drug and suitable
described in US 9,248,584 (Friesen et al. 2011), excipients in a vessel by homogenization at
a suspension using an organic solvent is prepared temperatures at least 10 C above the melting
at room temperature and then heated over 100 C point or glass transition temperature. The molten
through an in-line heater at high pressure. The material is then fed to a congealing chamber
increase of temperature and pressure enables the where it is atomized into fine droplets by a suit-
full dissolution of the drug. The low residence able atomizer similar to those used in conven-
time of the fluid from heater to nozzle reduces tional spray drying (e.g., rotary, two-fluid, or
the chance of drug degradation. The solution is pressure nozzle). The molten droplets are then
Fig. 10.21 Spray-

congealed particles
obtained in a pilot-scale
equipment. Reproduced
Hovione FarmaCiencia SA
solidified on contact with a cocurrent cold process solidification of the melt and the production of
gas stream. The solid particles are separated from spherical particles with a defined particle-size
the gas using a cyclone or filter bag in a similar range that is achieved in a single-step unit opera-
manner to spray drying. tion. In this way, secondary processing is avoided
The melt may be prepared by (1) dissolving like spheronization or milling that is often
the drug in the molten excipients, (2) suspending required following the melt extrusion process.
the drug in the molten excipients, (3) dissolving A challenge of spray congealing is the require-
the excipients in the molten drug, or ment for atomization of high-viscosity feeds into
(4) suspending the excipients in the molten fine droplets. In order that atomization can be
drug. Amorphous solid dispersions are only performed, the viscosity of the melt should be
formed in the case where the drug is in the liquid below 20,000 cP, preferably less than
state, either dissolved in the excipients or melted 10,000 cP. The feeding line and atomizing
(cases 1 and 3). nozzles should be kept at a temperature equal to
The preparation of the melt is one of the most or above the melt point to prevent clogging. All
challenging aspects of the spray-congealing pro- atomizing nozzles suitable for spray drying are
cess. The solubility of drugs in polymers is often theoretically suitable for spray congealing; how-
limited which leads to low drug loads. Moreover, ever, due to higher viscosities, rotary nozzles are
their stability at relatively high temperatures is, in preferred.
many cases, insufficient. In order to overcome Regarding the particular aspects of the formu-
this issue for heat-sensitive products, melting lation, all excipients that are solid at room tem-
and homogenization of the molten feed may be perature and melt without degradation may be
performed in line and in a continuous manner by used in spray congealing. However, those with
using a screw extrusion system (single screw or relatively moderate glass transition temperatures
twin screw) and feeding directly to the atomiza- or melting points are preferred. Excipients used in
tion nozzle (Appel et al. 2005). In this way, the hot-melt extrusion are also well suited for spray
drug is subjected to high temperatures for a much congealing: polymers like Eudragit E100 and
shorter duration. PEGs; surfactants like Gelucire 44/14, polyethyl-
Spray congealing offers a key advantage over ene glycol fatty acid esters, and poloxamers;
a typical hot-melt extrusion process with respect solubilizers such as alcohols like benzyl alcohol,
to downstream processing. Specifically, sorbitol, and mannitol; and esters like ethyl oleate
and ethyl caprylate (Patel and Chen 2001). The Wall Materials
function of the solubilizers is to increase the drug The selection of wall materials depends on the
solubility in the melt, thereby increasing the drug objective of the microencapsulation process and
load in the formulation. the physicochemical properties of the active
ingredient. They play an important role in
flowability, mechanical stability, and shelf life
(Ré 1998). They should be highly soluble in the
10.11.3 Spray Drying
process solvent and have adequate film-forming
for Microencapsulation
ability. The most common wall materials used in
microencapsulation by spray drying are the
Microencapsulation can be defined as the entrap-
following:
ment of an active ingredient by a protective wall.
The wall may be designed to protect the active • Carbohydrates (starches, maltodextrins, and
ingredient from oxidation, light, pH, enzymatic sucrose).
degradation, etc. The wall can also be used for • Gums and gelatins (arabic gum, xanthine, and
taste masking, to improve powder attributes (e.g., gelatin).
flowability) or to produce controlled-release • Cellulose-based polymers (hydroxypropyl
particles (Ré 1998; Shaidi and Han 1993). It has methylcellulose acetate succinate, cellulose
been widely applied in the food and pharmaceuti- acetate phthalate, and hydroxypropyl methyl
cal industries mainly for protective or controlled- cellulose).
release formulation applications. • Polylactic acid (PLA) and polylactide-co-
There are several methods to produce glycolide (PLGA) (biodegradable polymers).
microencapsulated materials, such as coacerva- • Acrylate-based polymers (polymethacrylate
tion, molecular inclusion, emulsification, and in and polymethylmethacrylate).
situ polymerization among others. Spray drying
can be used with several of these techniques as a These materials are chosen to provide high
microencapsulation method or merely as drying/ solubility in the spray-drying solvent, emulsifica-
isolation process. As a continuous and gentle tion properties, film-forming ability, good perfor-
process amenable to heat-sensitive materials, mance during spray-drying process (related to
spray drying is particularly useful as a physical glass transition temperature and stickiness), and
method for microencapsulation. It is therefore the low to moderate viscosities at relatively high
most common and cost-effective way to produce concentrations (close to 50% w/w).
microencapsulated materials in the pharmaceuti-
cal and food industries (Gharsallaoui et al. 2007).
Carbohydrates
In terms of process, microencapsulation by
Among the carbohydrates, maltodextrins and
spray drying is achieved by suspending or
modified starches constitute good candidates for
emulsifying the active ingredient in a solution/
microencapsulation by spray drying. They are
emulsion containing the dissolved wall materials.
highly soluble in water and yield solutions with
This solution may then be homogenized to
low viscosity. They also have good spray-drying
decrease the suspension/emulsion particle size
properties, namely, high glass transition
before feeding to the spray drier. The feed is
temperatures, little tendency to stick to equipment
then atomized in the nozzle, and during drying
walls, and produce relatively dense powders.
the dissolved materials precipitate to form a shell
However, carbohydrates have poor film-forming
around the active ingredient. In this section, wall
capacity which is important for encapsulation
materials and the spray-drying process
efficiency and especially in the retention of vola-
parameters that impact microencapsulation effi-
tile constituents. This can be solved by adding
ciency will be discussed.
gums (e.g., gelatins, arabic gum) to the solution.
The mass ratio between carbohydrate and gum
varies, but it is typically greater than 1 (Bayram an optimum temperature for the microencapsula-
et al. 2005). Another benefit of using gum is the tion process. Increasing the solution temperature
improvement of emulsification properties in the was found to improve the atomization process by
formulation which can lead to enhanced bioavail- decreasing the solution viscosity. The use of very
ability (Bruschi et al. 2003). high drying temperatures led to the breakage of
the particle shell due to the increase in vapor
Gums and Gelatins pressure of the entrapped solvent. This ultimately
Spray-drying pure gums may be challenging due resulted in poor encapsulation efficiency. There-
to their poor drying behavior associated with low fore, the selection of drying conditions, namely,
glass transition temperatures. They are typically temperatures, is of critical importance not only for
used together with other wall materials, namely, process performance (yield) but also for product
carbohydrates, as discussed previously. quality and encapsulation efficiency. In the latter
aspect, the visualization of particles by scanning
PLGA/PLA electron microscopy provides a fast and reliable
Poly(lactic-co-glycolic acid) is a biodegradable method to assess the effect of the drying process
and biocompatible copolymer. It degrades in the on the integrity of the microparticles.
body to form lactic and glycolic acid monomers Another important feature when spray drying
which are by-products of natural metabolic suspension for microencapsulation purposes is
pathways. This biodegradable polymer has been the ratio between the particle size of the
the subject of intense research in drug delivery in suspended material and the droplet size generated
recent years (Langer 1990; Putney and Burke in the atomization process. Obviously, the droplet
1998). The use of PLGA/PLA and associated needs to be larger than the material to be
microencapsulation techniques are well described encapsulated. In this particular case, the increase
in the scientific literature (Wischke and in solution viscosity due to the use of polymers
Schwendeman 2008). may be beneficial because it enables the genera-
Examples of marketed PLGA tion of larger droplets (Shu et al. 2006).
microencapsulated drugs produced by spray dry- It is important to note that the presence of
ing are buserelin acetate marketed as Suprecur suspended solid particles must be taken into
MP® (Japan) from Mochida Pharmaceutical account when selecting the atomization nozzle.
(licensed fomo Aventis) and bromocriptine For obvious reasons, the dimension of the
marketed as Parlodel® from Sandoz (Wischke suspended particles or agglomerates must be
and Schwendeman 2008). smaller than the minimum passage of the nozzle.
Given that nozzle blockage may have a cata-
Particular Features of the Spray-Drying strophic effect on the spray-drying process,
Process for Microencapsulation in-line filters are usually employed. Their pore
Microencapsulated particle formation depends on size is designed to be smaller than the orifice,
the balance between the solvent evaporation rate but sufficiently large to allow passage of a great
and the diffusion rate of wall materials in the majority of the suspended particles.
liquid phase during the drying process. This bal- Another important aspect is abrasion of the
ance can be quantified by the Peclet number nozzle caused by the expulsion of suspended
(Vehring 2008). According to their influence on particles at high velocities. This causes additional
this balance, spray-drying process parameters erosion of the nozzle and a corresponding
effect particle size, density, morphology, residual decrease in atomization efficiency over time.
moisture, and microencapsulation efficiency. Therefore, careful selection of nozzle type and
A study on microencapsulation of lycopene material of construction is recommended along
(Shu et al. 2006) demonstrated that feed and with routine verification of nozzle performance.
spray-drying temperatures influenced encapsula- In this particular situation, two-fluid nozzles rep-
tion efficiency. The authors found that there was resent a more robust solution than pressure
nozzles and are preferred in cases where the easily control the droplet size and with a low
suspended particles are large or very abrasive. shear while compared with the pressure nozzle
The pressure nozzle is also used in some cases (Ziaee et al. 2019). Such process has been suc-
for suspensions; however, the material of con- cessfully applied with approved products in the
struction must be sufficiently hard, such as tung- market. The first approved biomolecule by spray
sten carbide. The compatibility of the suspension drying was Exubera®, in 2006, for lung delivery
with the associated high-pressure pump must also (as described in Sect. 10.10). Later, Raplixa®
be assessed. produced by aseptic spray drying was submitted
and approved in 2015. FDA’s approval of
Raplixa® was a milestone which supported the
10.11.4 Spray Drying for Large industry to explore aseptic spray drying as a
Molecules manufacturing option. Nevertheless, the aseptic
spray drying has nowadays limited availability
The research of pharmaceutical biologics such as in processing plants of companies and contract
proteins have been increasing in the recent years, manufacturers.
especially in the dry powder formulations. Con- As alternative, the spray freeze drying
trarily to the aqueous dosage forms, the dry pow- combines the principles of spray drying and
der formulations ideally increase the material freeze drying. Similar to those, the spray freeze
stability and shelf time facilitating the storage drying is divided into three main steps: atomiza-
and transport requirements (as refrigerated tion, fast freezing, and drying by sublimation
conditions). The powder form of biologics is tra- (Emami et al. 2018). Like the spray-congealing
ditionally obtained by freeze drying. In this process, the dried particle has similar particle size
approach, the solution is firstly frozen followed as the frozen droplets. The physical attributes of
by a sublimation stage in which the solvent is the product as particle size and morphology
removed. Freeze drying is a well-established pro- emerge from the freezing step and not in drying
cess for marketed biopharmaceuticals with over as SD process. Different spray freezing
400 approved products by the FDA (Pinto et al. techniques may be applied such as
2021). Nevertheless, the technology presents sev- (i) atmospheric spray, (ii) spray freezing with
eral challenges, mainly (i) lack of ability for par- compressed carbon dioxide, (iii) freezing by
ticle engineering, (ii) scalability, and (iii) resource spraying into vapor over a cryogenic liquid, and
intensive. (iv) spray freezing into liquid. A comprehensive
The particle engineering capabilities, scalabil- comparison of these techniques and sublimation
ity, and the ability of continuous processing are drying is detailed and explored in Wanning
the main advantages and drivers for the applica- et al. (2015).
tion of spray drying in the biomolecules. Table 10.25 summarizes and compares the
The application of spray drying for the isola- technologies discussed. In the processes
tion and particle engineering of large molecules described, the biomolecules are subject to stress
has increased in the recent years with the conditions which may affect the physical and
strengthening of fundamental understanding. chemical stability of the compound. Therefore,
The two major concerns for spray drying are stabilizers are used even if the protein is stabled
normally related with the stress from nozzle in solid form. The selection of stabilizer is then
shear and the high temperature which may impact crucial for the successful formulation develop-
on product stability. Modeling tools as CFD have ment of a solid dosage form. The stabilizers may
been applied to further confirm the temperature be divided into five groups: polyols, proteins,
profile within the chamber as well as low resi- carbohydrates, amino acids, and others (including
dence time decreasing the chance of denaturation salts and surfactants) (Mensink et al. 2017). The
or degradation. The preferred atomization stabilizers may further enhance the glassy state
systems have been the two-fluid nozzle able to and work as a water replacer.
Table 10.25 Comparison of stress conditions, advantages, and challenges of drying technologies for biomolecules
(Emami et al. 2018; Ziaee et al. 2019)
Stress
Technology circumstances Advantages Disadvantages
Freeze Crystallization. Stabilized process. No particle engineering capabilities: Need
drying pH changes, Minimal product losses. of downstream processing.
Dehydration Minimizes denaturation by avoiding Scalability.
stress. high temperatures. Resource intensive (time and power).
Well-known aseptic processing. Large footprint.
Spray Atomization/ Control of particle attributes, viz., size, Product losses larger than freeze drying.
drying shear stress. density, and morphology. Larger temperatures required.
Thermal stress. Continuous process with low
Dehydration residence time.
stress. Target aerosolization profiles.
Spray Atomization/ Avoids elevated temperatures, Complex process and optimization.
freeze shear stress. minimizing denaturation. Requires liquid nitrogen.
drying Freezing Particle engineering capabilities. High-cost operations.
stress.
Dehydration
stress.
technology to inhaled drug delivery was

10.12 Summary
discussed, as well as emerging applications of
the process pertinent to poorly soluble
Spray drying is a mature technology that has been
compounds, i.e., spray congealing and microen-
used industrially since the mid-nineteenth century
capsulation. In conclusion, this chapter has
and long used in the pharmaceutical industry for
demonstrated the applicability of spray-drying
API and excipient manufacturing. In this chapter,
technology to the formulation of poorly water-
the process fundamentals, process components,
soluble drugs at all development stages. The
and equipment options have been reviewed to
recent market and late-stage clinical success of
provide the reader with the fundamental under-
drug products based on spray drying described
standing necessary to begin practicing this tech-
herein indicate the increasing importance of the
nology. The application of spray-drying
technology to the pharmaceutical industry with
technology to the formulation of poorly water-
regard to enhancing therapies with poorly water-
soluble drugs was reviewed with particular focus
soluble drugs.
on ASDD systems. The development process for
ASDD formulations was discussed along with
Method Capsule 1
considerations for final dosage form production
pertinent to the various stages of new drug devel- Preparation and Characterization of Spray-
opment. Examples of ASDD systems from the Dried Microparticles
pharmaceutical literature were reviewed along Based on the method reported by De Jaeghere
with industrial examples of marketed or close- et al. (2000)
to-market products. These examples not only pro- Objective
vide the reader with practical case studies to con-
sider regarding the use of spray drying for the • Produce and characterize amorphous spray-
formulation of poorly water-soluble drugs but dried microparticles of CGP 70726 with
also illustrate the commercial viability of the pro- Eudragit L100-55 for enhanced oral
cess. Finally, the application of spray-drying absorption.
Equipment and Materials an emulsion–diffusion method in the fed

state.
• CGP 70726.
• Eudragit L100-55 (methacrylic acid Method Capsule 2
copolymer).
Polymer Selection for an ASDD System by
• Methanol.
Supersaturated In Vitro Dissolution
• Büchi Mini Spray Dryer, Model 190.
Screening
• Malvern Mastersizer® 2000.
Based on the method reported by Curatolo
• PW 1729 X-ray generator.
et al. (2009).
Method
Objective
• Feed preparation: In 300 g of methanol,
• Employ a supersaturated in vitro dissolu-
dissolve CGP 70726 (1% w/w) and
tion test (micro-centrifuge method) to rank
Eudragit L100-55 (4% w/w) with stirring.
order the performance of polymers as
• Spray-drying parameters:
carriers for compound 5 in ASDD systems.
– Nozzle type: pneumatic two-fluid.
– Nozzle diameter: 0.5 mm.
– Collection system: cyclone.
• Compound 5 (API), pure crystalline and
– Drying gas: nitrogen.
amorphous.
– Feed: methanolic solution, 5% (w/w)
• Polymers: HPMCAS-MF, CAP, CAT,
solids.
HPMC, HPMCP, PVP.
– Feed rate: 4 mL/min.
• Phosphate-buffered saline, pH 6.5.
– Inlet temperature: 50 C.
• Mobile phase: 60/40 1.7% ammonium
– Outlet temperature: 37–42 C.
ascorbate/acetonitrile.
– Aspirator setting: 15.
• Controlled temperature box at 37 C.
• Morphology assessment: PXRD.
• Micro-centrifuge tube (polypropylene,
• Particle-size analysis: laser light diffraction.
Sorenson Bioscience Inc.)
• In vivo performance: single-dose, oral
• Vortex mixer (Fisher Vortex Genie 2).
administration in beagle dogs (n ¼ 4).
• Micro-centrifuge, Marathon, Model
Results
Micro A.
• Pipette (Gilson Pipetman P-100).
• The average production yield of the spray-
• HPLC (Hewlett Packard 1090 HPLC,
dried microparticles for two runs was 67%.
Phenomenex Ultracarb ODS 20 analytical
• The spray-dried microparticles were deter-
column, absorbance measured at 215 nm
mined to be amorphous by PXRD.
with a diode array spectrophotometer).
• The mean particle diameters of two spray-
Method
dried microparticle batches were
10.0 1.5 μm and 9.2 1.3 μm as deter-
• Amorphous solid dispersion formulations
mined by laser light diffraction.
of compound 5 (10% w/w) and the various
• The amorphous spray-dried microparticles
polymer carriers were produced by spray
substantially enhanced the oral absorption
drying from organic solution.
of CGP 70726 over the crystalline drug
• In a controlled temperature box at 37 C,
(plasma levels not quantifiable).
4.0 mg of each ASDD powder was weighed
• The amorphous spray-dried microparticles
into an empty micro-centrifuge tube. Then,
produced exposures that were greater than
2.0 mL of phosphate-buffered saline
twofold that of nanoparticles produced by
(pH 6.5) was added to the tube (theoretical Objective

maximum drug concentration 200 μg/mL).
• The tube was then closed, timer started, and • To prepare amorphous solid dispersions of
the tube was mixed continuously for 60 s AMG-517 in HPMCAS-MF at 15% and
with a vortex mixer on the highest speed. 50% drug loading.
• The tube was transferred to the centrifuge Equipment and Reagents
and allowed to stand for 6 min and then
centrifuged at 13,000 g for 60 s. • AMG-517, micronized free base (API).
• A 25 μL sample was removed from the • HPMCAS-MF (AQOAT AS-MF, Shin-
supernatant using a pipette at 10 min after Etsu Chemical Company).
the timer was started. • Ethyl acetate (99.5% minimum).
• The solids in the centrifuge tube were then • Büchi Mini Spray Dryer, Model B290.
resuspended by vortex mixing for 30 s. • Malvern Mastersizer 2000 equipped with a
• The centrifuge tube was returned to the Hydro 2000 μP wet dispersion cell.
centrifuge and allowed to stand undisturbed • Phillips automated X-ray powder diffrac-
until the next sample time point. tometer, X’Pert PRO.
• At each time point (5, 10, 20, 40, 90 min), • Modulated DSC, Q1000 by TA
the tube was centrifuged, supernatant sam- Instruments.
pled, and solids resuspended as described. • 40 C/75% RH stability chamber
Results Method
• All ASDD formulations showed significant • Feed preparation: HPMCAS-MF was

initial supersaturation relative to the pure dissolved in ethyl acetate at a concentration
crystalline and amorphous compound of 2% (w/w). AMG-517 was then dissolved
5. The rank order of Cmax was as follows: in the polymer solution to a concentration
HPMCAS-MF > CAP > CAT > HPMC > of 0.353% or 2% (w/w) to produce drug–
HPMCP > PVP > pure API (crystalline and polymer ratios of 15:85 and 50:50.
amorphous). • Spray-drying parameters:
• Subsequent precipitation of supersaturated – System: open cycle.
compound 5 was seen for all ASDD – Nozzle type: 48 KHz ultrasonic
formulations. The rank order of extent of atomizing nozzle, 2 W power supply.
supersaturation for the formulations was as – Atomizing air: focusing nitrogen at
follows: HPMCAS-MF > CAP > CAT > 30 SLPM.
HPMC > HPMCP > PVP. – Collection system: cyclone.
• From this study, HPMCAS-MF was – Drying gas: nitrogen.
identified as the optimum polymer carrier – Drying gas flow rate: 300 SLPM.
in an ASDD system with compound 5 with – Feed: ethyl acetate solution, 2.353% and
respect to in vitro dissolution performance. 4% (w/w) solids.
– Feed rate: 0.75 mL/min.
Method Capsule 3 – Inlet temperature: 75 C.
– Aspirator setting: bypassed.
Spray Drying of AMG-517 with HPMCAS-
• Morphology assessment: PXRD.
MF
• Determination of Tg’s: modulated DSC.
Based on the method reported by Kennedy
et al. (2008).
• Stability assessment: accelerated storage at Equipment and Materials

40 C/75% RH. • VX-950 (telaprevir).
• In vivo performance: single-dose, oral • HPMCAS (AQOAT, Shin-Etsu).
administration in cynomolgus monkeys • Dichloromethane.
(n ¼ 6). • 8000 L stirred tank reactor
Results • Niro PSD-4 spray dryer (drying capacity
1250 kg/h) configured in FSD mode (closed
• Powder yields for the 15% and 50% drug cycle).
load AMG-517:HPMCAS-MF • Pressure nozzle (Spraying Systems MFP
formulations were 80.5% and 50%, (Maximum Free Passage) SK Series
respectively. SPRAYDRY® Nozzles Series variety, ori-
• Median particle size (volume-based fice 52 with core 27).
diamber d50) was 34.75 and 40.7 μm for Method
the 15% and 50% drug load ASDDs,
respectively. • Feed preparation: Charge VX-950 (85%
• At both drug loadings, the ASDD w/w) and HPMCAS (15% w/w) to the
formulations were determined by PXRD stirred tank reactor. Then add sufficient
to be amorphous and found by MDSC to dichloromethane to achieve a solid content
be single-phase systems with Tg’s of of 20% (w/w). Stir until a clear solution is
106 and 98 C for the low and high drug obtained while keeping the reactor at 20 C.
load formulations, respectively. • Fluidized spray-drying parameters:
• The ASDD systems were found to remain – Drying gas: nitrogen, cocurrent.
PXRD amorphous after 6-month storage at – Feed: dichloromethane solution, 20%
40 C/75% RH. (w/w) solids.
• The 15% drug load ASDD formulation in a – Feed rate: 151 kg/h.
capsule (with 5% SDS) yielded 163% – Feed pressure: 22 bar.
greater exposure in monkeys as compared – Inlet temperature: 75 3 C.
to micronized crystalline AMG-517 in – Outlet temperature: 35 5 C.
aqueous suspension. – ΔPcyclone: 10–12 mm H2O
– Fines return position: middle.
Method Capsule 4 – Fluid Bed 1 temperature set point:
40 C.
Fluidized Spray Drying of VX-950
– Fluid Bed 2 temperature set point:
(Telaprevir) with HPMCAS
35 C.
Based on the method reported by Bittorf
– Fluid Bed 3 temperature set point:
et al. (2010).
30 C.
Objective
Results
• Employ fluidized spray drying (FSD) to
• Product properties:
produce an ASDD product composed of
– Bulk density: 0.32 g/mL.
VX-950 and HPMCAS suitable for direct
– Tap density: 0.41 g/mL.
compression, i.e., greater average particle
– d10: 16.47 μm,
size and bulk density versus traditional
– d50: 60.03 μm,
spray drying.
– d90: 151.05 μm, Equipment and Materials

– Span: 2.24.
– Distribution: unimodal. • Milled itraconazole (Dv90 < 10 μ).
• Product obtained was suitable for direct • Gelatin.
tablet compression. • Sucrose.
• Deionized water.
Method Capsule 5 • Niro Mobile Minor.
Method
Spray Congealing of Vitamin E
Objective
• Feed preparation: a solution of gelatin was
prepared in hot water (60 C) at a concen-
• Employ spray congealing to produce amor-
tration of 2% w/w. Sucrose was added
phous vitamin E.
(sucrose/gelatin 4:1). After sucrose dissolu-
tion itraconazole (10% of total solids) was
added.
• Vitamin E.
• Spray drying:
• Büchi Mini Spray Dryer, Model B290
– Drying gas: nitrogen, cocurrent.
equipped with spray congealing features.
– Feed: suspension of itraconazole.
Method
– System: closed cycle.
– Nozzle type: two-fluid nozzle.
• Feed preparation: Charge vitamin E in a
thermoregulated vessel and heat up to
60 C (keep temperature above at least
– Drying gas flow rate: 80 kg/h.
10 C).
– Atomization gas flow rate: 2.6 kg/H.
• Spray congealing:
– Feed flow rate: 1.3 kg/h.
– Drying gas: nitrogen, cocurrent.
– Feed: vitamin E melted.
– Outlet temperature: 80 C.
– System: open cycle.
• Morphology assessment: SEM.
– Nozzle type: two-fluid nozzle.
Results
– Drying gas flow rate: fan at 100%.
• Powder yield: 85%.
• Median particle size (volume-based
• Morphology assessment: SEM.
diamber d50) was 25 μm.
• Particle morphology: corrugated spheres.
Results
• Powder yields 80%.

• Median particle size (volume-based References
diamber d50) was 40 μm.
• Smooth spherical particles. Al-Khattawi A, Bayly A, Phillips A, Wilson D. The design
and scale-up of spray dried particle delivery systems.
Expert Opinion Drug Del. 2018;15(1):47–63. https://
Method Capsule 6 doi.org/10.1080/17425247.2017.1321634.
Ambike AA, Mahadik KR, Paradkar A. Spray-dried amor-
Microencapsulation of Itraconazole phous solid dispersions of simvastatin, a low Tg drug:
Objective in vitro & in vivo evaluations. Pharm Res. 2005;22(6):
990–8.
• Microencapsulation of itraconazole by Appel LE, Crew MD, Friesen DT, et al. Spray-congeal
process using an extruder for preparing multipar-
spray drying. ticulate crystalline drug compositions containing
preferably a poloxamer and a glyceride. World Intel- (rapidly accelerated fibrosarcoma) inhibitor. Int J
lectual Property Organization WO/2005/ Pharm. 2013;454(1):241–8.
053656. 2005. Curatolo W, Nightingale J, Herbig S. Utility of
Assunção Y. Fluidized bed spray drying in the pharma- hydroxypropylmethylcellulose acetate succinate
ceutical industry. Masters thesis, Universidade de (HPMCAS) for initiation and maintenance of drug
Coimbra. 2011. supersaturation in the GI milieu. Pharm Res. 2009;26
Batycky RP, Jackson BC, Johnston LP, Mintzes JD, (6):1419–31.
Penachio EE. Method and apparatus for producing De Jaeghere F, Allémann E, Kubel F, et al. Oral bioavail-
dry particles. Patent No. US7008644. 2002. ability of a poorly water soluble HIV-1 protease inhib-
Bayram ÖA, Bayram M, Tekin AR. Spray drying of itor incorporated into pH-sensitive particles: effect of
sumac flavour using sodium chloride, sucrose, glucose the particle size and nutritional state. J Control Release.
and starch as carriers. J Food Eng. 2005;69(2):253–60. 2000;68(2):291–8.
Bercea M, Eckelt J, Wolf BA. Vapor pressures of polymer Dobry DE, Settell DM, Baumann JM, Ray RJ, Graham LJ,
solutions and the modeling of their composition depen- Beyerinck RA. A model-based methodology for spray-
dence. Ind Eng Chem Res. 2009;48(9):4603–6. https:// drying process development. J Pharm Innov. 2009;4
doi.org/10.1021/ie801965h. (3):133–42. https://doi.org/10.1007/s12247-009-
Berchielli A, Eisenhart EK, Herbig SM, et al. Dosage 9064-4.
forms providing controlled and immediate release of Emami F, Vatanara A, Park E, Na D. Drying technologies
cholesteryl ester transfer protein inhibitors and imme- for the stability and bioavailability of biopharma-
diate release of HMG-CoA reductase inhibitors. World ceuticals. Pharmaceutics. 2018;10(3):131. https://doi.
Intellectual Property Organization org/10.3390/pharmaceutics10030131.
WO2006082500. 2006. Engers D, Teng J, Jimenez-Novoa J, et al. A solid-state
Berndl G, Rosenberg J, Liepold B, et al. Solid pharmaceu- approach to enable early development compounds:
tical dosage formulations. US Patent Application selection and animal bioavailability studies of an
200,810,181,948 A1. 2008. itraconazole amorphous solid dispersion. J Pharm Sci.
Beyerinck RA, Dobry DE, Friesen DT, et al. Spray drying 2010;99(9):3901–22.
processes for forming solid amorphous dispersions of Fernandes DA, Barros R, Moura C, Pereira J, Corvo LM,
drugs and polymers. World Intellectual Property Orga- Costa E. Ultrasonic versus two-fluid nozzle in a spray
nization WO2005011636. 2005. drying process: a comparative study for the production
Bittorf KJ, Katstra JP, Gaspar F. Fluidized Spray Drying. of dry powder inhaler formulations for biopharmaceu-
US Patent Application 20,100,011,610. 2010. tical delivery. Drug delivery to the lungs. 2017.
Broadhead J, Rouan SKE, Rhodes CT. The spray drying of Finotello G. Droplet collision dynamics in a spray dryer:
pharmaceuticals. Drug Dev Ind Pharm. 1992;18 experiments and simulations [Technische Universiteit
(11–12):1169–206. Eindhoven]. 2019. www.tue.nl/taverne
Bruschi ML, Cardoso MLC, Lucchesi MB, et al. Gelatin Finotello G, De S, Vrouwenvelder JCR, Padding JT, Buist
microparticles containing propolis obtained by spray- KA, Jongsma A, Innings F, Kuipers JAM. Experimen-
drying technique: preparation and characterization. Int tal investigation of non-Newtonian droplet collisions:
J Pharm. 2003;264(1–2):45–55. the role of extensional viscosity. Exp Fluids. 2018;59
Carstensen JT, Zoglio MA. Tray drying of pharmaceutical (7):1–16. https://doi.org/10.1007/s00348-018-2568-2.
wet granulations. J Pharm Sci. 1982;71(1):35–9. Finotello G, Padding JT, Buist KA, Jongsma A, Innings F,
Celik M, Wendel SC. Spray drying and pharmaceutical Kuipers JAM. Droplet collisions of water and milk in a
applications. In: Parikh DM, editor. Handbook of phar- spray with Langevin turbulence dispersion. Int J Mul-
maceutical granulation technology. 2nd ed. Boca tiphase Flow. 2019;114:154–67. https://doi.org/10.
Raton: Taylor and Francis; 2005. 1016/j.ijmultiphaseflow.2019.03.003.
Chal B, Mogalian E, Oliyai R, et al. Combination formu- Frake P, Greenhalgh D, Grierson SM, et al. Process control
lation of two antiviral compounds. US Patent Applica- and end-point determination of a fluid bed granulation
tion US14/168,264. 2014. by application of near infra-red spectroscopy. Int J
Chiou WL, Riegelman S. Pharmaceutical applications of Pharm. 1997;151(1):75–80.
solid dispersion systems. J Pharm Sci. 1971;60(9): Friesen DT, Shanker R, Crew M, et al. Hydroxypropyl
1281–302. methylcellulose acetate succinate-based spray-dried
Crew MD, Curatolo WJ, Friesen DT, et al. Pharmaceutical dispersions: an overview. Mol Pharm. 2008;5(6):
compositions of cholesteryl ester transfer protein 1003–19.
inhibitors. US Patent 7,235,259. 2007. Friesen DT, Newbold DD, Baumanm JM, DuBose DB,
Cui Y, Murphy M, Dinehart K, et al. Pharmaceutical Millard DL. High-temperature spray drying process
compositions. US Patent Application and apparatus. Patent No. US 9,248,584 B2. 2011.
20,060,089,385. 2006. Fry DS, Lindemann CM, Preigh M, Bloom CJ, Smithey
Cui Y, Chiang PC, Choo EF, et al. Systemic in vitro and CDC, Dubose DB, Gautschi J, Smithey D. Solid
in vivo evaluation for determining the feasibility of dispersions of a erb2 (her2) inhibitor Patent
making an amorphous solid dispersion of a B-Raf No. US20170136022A1. 2017.
Gala UH, Miller DA, Williams RO. Harnessing the thera- supersaturation screening study. Eur J Pharm
peutic potential of anticancer drugs through amorphous Biopharm. 2008b;69(1):158–66.
solid dispersions. BBA Rev Cancer. 2020;1873(1): Jasuja A. Atomization of crude and residual fuel oils.
188319. https://doi.org/10.1016/j.bbcan.2019.188319. Trans ASME J Eng Power. 1979;101:250–8.
Gaspar F, Vicente J, Neves F, Authelin J-R. Spray drying: Jiang YJ, Umemura A, Law CK. An experimental investi-
scale-up and manufacturing. In: Shah N, Sandhu H, gation on the collision behaviour of hydrocarbon
Choi D, Chokshi H, Malick A, editors. Amorphous droplets. J Fluid Mech. 1992;234(1):171–90. https://
solid dispersions. Advances in delivery science and doi.org/10.1017/S0022112092000740.
technology. Springer; 2014. p. 261–302. https://doi. Jung J-Y, Yoo SD, Lee S-H, et al. Enhanced solubility and
org/10.1007/978-1-4939-1598-9_8. dissolution rate of itraconazole by a solid dispersion
Gharsallaoui A, Roudaut G, Chambin O, et al. technique. Int J Pharm. 1999;187(2):209–18.
Applications of spray-drying in microencapsulation Kakuda TN, Schöller-Gyüre M, Workman K, et al. Single-
of food ingredients: an overview. Food Res Int. and multiple-dose pharmacokinetics of etravirine
2007;40(9):1107–21. administered as two different formulations in HIV-1-
Gilead. Gilead Announces U.S. FDA Priority Review infected patients. Antivir Ther. 2008;13:655–61.
Designation for Sofosbuvir/Velpatasvir for Treatment Kennedy M, Hu J, Gao P, et al. Enhanced bioavailability
of All Genotypes of Chronic Hepatitis C Infection. of a poorly soluble VR1 antagonist using an amor-
2016. https://www.gilead.com/news/press-releases/ phous solid dispersion approach: a case study. Mol
2016 Pharm. 2008;5(6):981–93.
Gorman E, Mogalian E, Oliyai R et al. Solid dispersion Kiekens FRI, Voorspoels JFM, Baert LEC Process for
formulation of an antiviral compound. US Patent preparing spray dried formulations of TMC125.
Application 14/168,313. 2015. World Intellectual Property Organization WO/2007/
Hansen OE, Ullum TU Air disperser for a spray dryer and 141308. 2007.
a method for designing an air disperser. US Patent Kim EH, Chen XD, Pearce D. On the mechanisms of
Application 20090008805. 2009. surface formation and the surface compositions of
Harvoni. Package Insert. 2015. http://www.harvoni.com/ industrial milk powders surface compositions of indus-
Hede PD, Bach P, Jensen AD. Two-fluid spray trial milk powders. Dry Technol. 2003;21(2):265–78.
atomisation and pneumatic nozzles for fluid bed coat- https://doi.org/10.1081/DRT-120017747.
ing/agglomeration purposes: a review. Chem Eng Sci. Kim MS, Kim JS, Cho W, et al. Supersaturatable
2008;63(14):3821–42. https://doi.org/10.1016/j.ces. formulations for the enhanced oral absorption of
2008.04.014. sirolimus. Int J Pharm. 2013;445(1):108–16.
Heigoldt U, Sommer F, Daniels R, et al. Predicting in vivo Kirkpatrick PA, Schulman M, Lehmann DM, et al. Three-
absorption behavior of oral modified release dosage fluid atomizing nozzle and method of utilization
forms containing pH-dependent poorly soluble drugs thereof. US Patent 4,610,760. 1986.
using a novel pH-adjusted biphasic in vitro dissolution Kleinebudde P. Roll compaction/dry granulation: pharma-
test. Eur J Pharm Biopharm. 2010;76(1):105–11. ceutical applications. Eur J Pharm Biopharm. 2004;58
Hoe S, Ivey JW, Boraey MA, Shamsaddini-Shahrbabak A, (2):317–26.
Javaheri E, Matinkhoo S, Finlay WH, Vehring R. Use Kohri N, Yamayoshi Y, Xin HE, et al. Improving the oral
of a fundamental approach to spray-drying formulation bioavailability of albendazole in rabbits by the solid
design to facilitate the development of multi- dispersion technique. J Pharm Pharmacol. 1999;51(2):
component dry powder aerosols for respiratory drug 159–64.
delivery. Pharm Res. 2014;31(2):449–65. https://doi. Kwong E, Higgins J, Templeton AC. Strategies for bring-
org/10.1007/s11095-013-1174-5. ing drug delivery tools into discovery. Int J Pharm.
Hurter P, Rowe W, Young CR, et al. Solid forms of 2011;412:1–7.
N-[2,4-BIS(1,l-D1methylethyl) d-hydroxyphenyli-14- Labiris NR, Dolovich MB. Pulmonary drug delivery part I:
dihydro-4-oxoquinoline-3-carboxamide. US Patent physiological factors affecting therapeutic effective-
Application 20,110,064,811. 2011. ness of aerosolized medications. Br J Clin Pharm.
International Conference on Harmonization. Impurities: 2003;56(6):588–99.
guideline for residual solvents, Q3C(R5). 2011. Langer R. New methods of drug delivery. Science.
Janssens S, Humbeeck JV, Van den Mooter G. Evaluation 1990;249(4976):1527–33.
of the formulation of solid dispersions by co-spray Law D, Krill SL, Schmitt EA, et al. Physicochemical
drying itraconazole with Inutec SP1, a polymeric sur- considerations in the preparation of amorphous
factant, in combination with PVPVA 64. Eur J Pharm ritonavir–poly(ethylene glycol) 8000 solid dispersions.
Biopharm. 2008a;70(2):500–5. J Pharm Sci. 2001;90(8):1015–25.
Janssens S, Nagels S, Armas HN, et al. Formulation and Law D, Schmitt EA, Marsh KC, et al. Ritonavir–PEG
characterization of ternary solid dispersions made up of 8000 amorphous solid dispersions: in vitro and
Itraconazole and two excipients, TPGS 1000 and in vivo evaluations. J Pharm Sci. 2004;93(3):563–70.
PVPVA 64, that were selected based on a Le HP. Progress and trends in ink-jet printing technology.
J Imaging Sci Technol. 1998;42(1):49–62.
Lee E-J, Lee S-W, Choi H-G, et al. Bioavailability of Miller D, DiNunzio J, Yang W, et al. Targeted intestinal
cyclosporin A dispersed in sodium lauryl sulfate- delivery of supersaturated itraconazole for improved
dextrin based solid microspheres. Int J Pharm. oral absorption. Pharm Res. 2008a;25(6):1450–9.
2001;218(1–2):125–31. Miller DA, DiNunzio JC, Yang W, et al. Enhanced in vivo
Lefebvre A. The prediction of Sauter mean diameter for absorption of itraconazole via stabilization of supersat-
simplex pressure-swirl atomisers. Atomisation Spray uration following acidic-to-neutral pH transition. Drug
Technol. 1987;3:37–51. Dev Ind Pharm. 2008b;34(8):890–902.
Lefebvre A, McDonell V. Atomization and sprays. In: Mizoe T, Ozeki T, Okada H. Application of a four-fluid
Boca Raton: CRC Press. 2nd ed. CRC Press; 2017. nozzle spray drier to prepare inhalable rifampicin-
https://doi.org/10.1016/0009-2509(90)87140-N. containing mannitol microparticles. AAPS
Lefebvre AH, Wang XF. Mean drop sizes from pressure- PharmSciTech. 2008;9(3):755–61.
swirl nozzles. J Propuls Power. 1987;3(1):11–8. Moser JD, Broyles J, Liu L, et al. Enhancing bioavailabil-
https://doi.org/10.2514/3.22946. ity of poorly soluble drugs using spray dried solid
Lowinger M, Tatavarti AS, Marsac P et al. Composition of dispersions: Part I. Am Pharm Rev (September/
a Non-Nucleoside Reverse Transcriptase Inhibitor WO October), 2008a; p. 68–71.
Patent Application WO/2015/077273. 2015. Moser JD, Broyles J, Liu L, et al. Enhancing bioavailabil-
Marasini N, Tran TH, Poudel BK, et al. Fabrication and ity of poorly soluble drugs using spray dried solid
evaluation of pH-modulated solid dispersion for dispersions: Part II. Am Pharm Rev (November/
telmisartan by spray-drying technique. Int J Pharm. December), 2008b;60–78.
2012;441(1):424–32. Mujumdar AS. Principles, classification, and selection of
Marota MJ, McKelvey C, Birringer N, Kuiper J, Harmon dryers. In: Mujumdar AS, editor. Handbook of indus-
PA, Socia AJ, Marsac PJ, Conway SL. Pharmaceutical trial drying. 3rd ed. Boca Raton: CRC Press; 2007.
composition of selective hcv ns3/4a inhibitors. Patent Overhoff KA, Moreno A, Miller DA, et al. Solid
No. US20160339074A1. 2015. dispersions of itraconazole and enteric polymers
Martin J. Transcript of Q3 2015 Gilead Sciences Inc. made by ultra-rapid freezing. Int J Pharm. 2007;336
earnings conference call. 2015. http://investors.gilead. (1):122–32.
com/phoenix.zhtml?p¼irol-eventDetails&c¼69964& Paradkar A, Ambike AA, Jadhav BK, et al. Characteriza-
eventID¼5204510 tion of curcumin-PVP solid dispersion obtained by
Masters K. Spray drying handbook. In: AIChE Journal. spray drying. Int J Pharm. 2004;271(1–2):281–6.
3rd ed. Wiley; 1979. https://doi.org/10.1002/aic. Parikh DM, Mogavero M. Fluidization theory. In: Parikh
690260430. DM, editor. Handbook of pharmaceutical granulation
Masters K. Spray drying handbook. New York: technology. 2nd ed. Boca Raton: Taylor &
Wiley; 1985. Francis; 2005.
Masters K. Spray drying in practice. Denmark: Spray Dry Patel MV, Chen F Solid carriers for improved delivery of
Consult Internations ApS; 2002. active ingredients in pharmaceutical compositions. US
Matsumoto T, Zografi G. Physical properties of solid Patent 6,248,363. 2001.
molecular dispersions of indomethacin with poly Paudel A, Van den Mooter G. Influence of solvent compo-
(vinylpyrrolidone) and poly(vinylpyrrolidone-co- sition on the miscibility and physical stability of
vinyl-acetate) in relation to indomethacin crystalliza- Naproxen/PVP K 25 solid dispersions prepared by
tion. Pharm Res. 1999;16(11):1722–8. cosolvent spray-drying. Pharm Res. 2012;29(1):
McKelvey CA, Kesisoglou F. Enabling an HCV treatment 251–70. https://doi.org/10.1007/s11095-011-0539-x.
revolution and the frontiers of solid solution formula- Paudel A, Van Humbeeck J, Van den Mooter
tion. J Pharm Sci. 2019;108(1):50–7. https://doi.org/ G. Theoretical and experimental investigation on the
10.1016/j.xphs.2018.11.003. solid solubility and miscibility of naproxen in poly
Mensink MA, Frijlink HW, van der Voort Maarschalk K, (vinylpyrrolidone). Mol Pharm. 2010;7(4):1133–48.
Hinrichs WLJ. How sugars protect proteins in the solid Percy SR. Improvement in Drying and Concentrating Liq-
state and during drying (review): mechanisms of stabi- uid Substances by Atomizing. US Patent 125:
lization in relation to stress conditions. Eur J Pharm 406. 1872.
Biopharm. 2017;114:288–95. https://doi.org/10.1016/ Pinto JT, Faulhammer E, Dieplinger J, Dekner M,
j.ejpb.2017.01.024. Makert C, Nieder M, Paudel A. Progress in spray-
Mezhericher M, Levy A, Borde I. Theoretical drying drying of protein pharmaceuticals: literature analysis
model of single droplets containing insoluble or of trends in formulation and process attributes. Dry
dissolved solids theoretical drying model of single Technol. 2021:1–32. https://doi.org/10.1080/
droplets containing insoluble or dissolved solids. Dry 07373937.2021.1903032.
Technol. 2007;25:1035–42. https://doi.org/10.1080/ Pisecky J, Krag J, Sorensen IH. Process for producing an
07373930701394902. agglomerated powdery milk product. US Patent
Miller RW. Roller compaction technology. In: Parikh DM, 4,490,403. 1984.
editor. Handbook of pharmaceutical granulation tech- Poozesh S, Bilgili E. Scale-up of pharmaceutical spray
nology. 3rd ed. New York: Informa Healthcare; 2010. drying using scale-up rules: a review. Int J Pharm.
2019;562:271–92. https://doi.org/10.1016/j.ijpharm. Shu B, Yu W, Zhao Y, et al. Study on microencapsulation

2019.03.047. of lycopene by spray-drying. J Food Eng. 2006;76(4):
Poozesh S, Setiawan N, Arce F, Sundararajan P, Rocca JD, 664–9.
Rumondor A, Wei D, Wenslow R, Xi H, Zhang S, Singh A, Van den Mooter G. Spray drying formulation of
Stellabott J, Su Y, Moser J, Marsac PJ. Understanding amorphous solid dispersions. Adv Drug Deliv Rev.
the process-product-performance interplay of spray 2016;100:27–50. https://doi.org/10.1016/j.addr.2015.
dried drug-polymer systems through complete struc- 12.010.
tural and chemical characterization of single spray Sotthivirat S, McKelvey C, Moser J, et al. Development of
dried particles. Powder Technol. 2017;320:685–95. amorphous solid dispersion formulations of a poorly
https://doi.org/10.1016/j.powtec.2017.07.042. water-soluble drug, MK-0364. Int J Pharm. 2013;452
Porfirio T, Galindo-Rosales FJ, Campo-Deaño L, (1):73–81.
Vicente J, Semião V. Rheological characterization of Tanno F, Nishiyama Y, Kokubo H, et al. Evaluation of
polymeric solutions used in spray drying process. Eur J hypromellose acetate succinate (HPMCAS) as a carrier
Pharm Sci. 2021;158:105650. https://doi.org/10.1016/ in solid dispersions. Drug Dev Ind Pharm. 2004;30(1):
j.ejps.2020.105650. 9–17.
Putney SD, Burke PA. Improving protein therapeutics Tarara ET, Weers JG, Kabalnov A, et al. Engineered
with sustained-release formulations. Nat Biotechnol. Particles and methods of use. US Patent
1998;16(2):153–7. 7,306,787. 2007.
Qian J, Law CK. Regimes of coalescence and separation in Vandecruys R, Peeters J, Verreck G, et al. Use of a screen-
droplet collision. J Fluid Mech. 1997;331:59–80. ing method to determine excipients which optimize the
https://doi.org/10.1017/S0022112096003722. extent and stability of supersaturated drug solutions
Radcliffe A. High speed aerodynamics and jet propulsion, and application of this system to solid formulation
vol XI, Fuel injection. Princeton: Princeton University design. Int J Pharm. 2007;342(1–2):168–75.
Press; 1960. Vehring R. Pharmaceutical particle engineering via spray
Ray RJ, Newbold DD, Beyerinck RA, Dobry DE, Grove drying. Pharm Res. 2008;25(5):999–1022. https://doi.
KD. Drying of drug-containing particles. Patent org/10.1007/s11095-007-9475-1.
No. US9265731B2. 2005. Vicente J. Modeling and simulation of the spray drying
Ré MI. Microencapsulation by spray drying. Dry Technol. processes, based on quality by design principles. Uni-
1998;16(6):1195–236. versity of Lisbon; 2014.
Salcedo RL, Pinho MJ. Pilot- and industrial-scale experi- Vicente J, Pinto J, Menezes J, Gaspar F. Fundamental
mental investigation of numerically optimized analysis of particle formation in spray drying. Powder
cyclones. Ind Eng Chem Res. 2002;42(1):145–54. Technol. 2013;247:1–7. https://doi.org/10.1016/j.
Santos JL, Temtem M, Gaspar F. Improved spray drying powtec.2013.06.038.
process for production of powder with enhanced Vicente J, Sá Couto C, Temtem M, Ferreira R. A spray
properties. Patent No. WO2015150763. 2015. https:// drying process with continuous preparation of spray
patentscope.wipo.int/search/en/detail.jsf? solution. Patent No. WO 2019162688-
docId¼WO2015150763&redirectedID¼true 20190829. 2018.
Scholler M, Hoetelmans R, Beets G, et al. Substantial Vig B, Morgen M. Formulation, process development, and
improvement of oral bioavailability of TMC125 using scale-up: spray drying amorphous solid dispersions for
new tablet formulations in healthy volunteers in inter- insoluble drugs. In: Qiu Y, Chen Y, Zhang GGZ, Yu L,
national AIDS conference 2005. Rio de Janeiro: Mantri RV, editors. Developing solid oral dosage
Brazil; 2005. forms. 2nd ed. Elsevier; 2017. p. 793–820. https://
Serajuddin ATM, Sheen P-C, Mufson D, et al. Effect of doi.org/10.1016/B978-0-12-802447-8.00030-3.
vehicle amphiphilicity on the dissolution and bioavail- Wanning S, Süverkrüp R, Lamprecht A. Pharmaceutical
ability of a poorly water-soluble drug from solid spray freeze drying. Int J Pharm. 2015;488(1–2):
dispersions. J Pharm Sci. 1988;77(5):414–7. 136–53. https://doi.org/10.1016/j.ijpharm.2015.
Shaidi F, Han XQ. Encapsulation of food ingredients. Crit 04.053.
Rev Food Sci Nutr. 1993;33:501–47. Weers J. Dispersible powders for inhalation applications.
Shanbhag A, Rabel S, Nauka E, et al. Method for screen- Innov Pharm Technol. 2000;1:111–6.
ing of solid dispersion formulations of low-solubility Weiler C, Egen M, Trunk M, et al. Dispersibility of jet
compounds – miniaturization and automation of sol- milled vs. spray dried powders. Resp Drug Deliv.
vent casting and dissolution testing. Int J Pharm. 2008;2:571–6.
2008;351(1–2):209–18. Weiler C, Egen M, Trunk M, et al. Force control and
Shi Y, Gao P, Gong Y, et al. Application of a biphasic test powder dispersibility of spray dried particles for inha-
for characterization of in vitro drug release of immedi- lation. J Pharm Sci. 2010;99(1):303–16.
ate release formulations of celecoxib and its relevance Wicks ZW. Free volume and the coatings formulator. J
to in vivo absorption. Mol Pharm. 2010;7(5):1458–65. Coating Technol. 1986;58:22–32.
Wildfong PLD, Samy A-S, Corfa J, et al. Accelerated fluid
bed drying using NIR monitoring and
phenomenological modeling: method assessment and Wu WD, Patel KC, Rogers S, et al. Monodisperse droplet
formulation suitability. J Pharm Sci. 2002;91(3): generators as potential atomizers for spray drying tech-
631–9. nology. Boca Raton: Taylor & Francis; 2007.
Wischke C, Schwendeman SP. Principles of encapsulating Yamashita K, Nakate T, Okimoto K, et al. Establishment
hydrophobic drugs in PLA/PLGA microparticles. Int J of new preparation method for solid dispersion formu-
Pharm. 2008;364(2):298–327. lation of tacrolimus. Int J Pharm. 2003;267(1–2):
Witschi C, Doelker E. Residual solvents in pharmaceutical 79–91.
products: acceptable limits, influences on physico- Yoshioka M, Hancock BC, Zografi G. Crystallization of
chemical properties, analytical methods and indomethacin from the amorphous state below and
documented values. Eur J Pharm Biopharm. 1997;43 above its glass transition temperature. J Pharm Sci.
(3):215–42. 1994;83(12):1700–5.
Wong SM, Kellaway IW, Murdan S. Enhancement of the Ziaee A, Albadarin AB, Padrela L, Femmer T, O’Reilly E,
dissolution rate and oral absorption of a poorly water Walker G. Spray drying of pharmaceuticals and
soluble drug by formation of surfactant-containing biopharmaceuticals: critical parameters and experi-
microparticles. Int J Pharm. 2006;317(1):61–8. mental process optimization approaches. Eur J Pharm
Sci. 2019;127(April 2018):300–18. https://doi.org/10.
1016/j.ejps.2018.10.026.
Pharmaceutical Cryogenic Technologies
11
Sawittree Sahakijpijarn, Chaeho Moon,
Abstract authors of this chapter from the first and sec-

Poor bioavailability associated with poorly ond editions. This current third edition chapter
water-soluble compounds remains a challeng- is a revision and update of the previous
ing issue in drug development. Particle engi- edition.
neering may be used to improve the
physicochemical properties of poorly water- Keywords
soluble compounds, thereby enhancing the Particle engineering · Spray freeze Drying
bioavailability. Cryogenic technologies, (SFD) · Spray freezing into liquid (SFL) · Thin
including spray freeze drying (SFD), spray film freezing (TFF) · Amorphous
freezing into liquid (SFL), and thin film freez- nanostructures · Dissolution rates ·
ing (TFF), are “bottom-up” precipitation pro- Supersaturation · Nucleation · Particle growth
cesses to generate amorphous, nanostructured arrest · Polymers · Solvents · Dry powder
aggregates with significantly enlarged surface inhalation
area, higher dissolution rates, and supersatura-
tion, via rapidly inducing nucleation followed
by particle growth arrest through stabilization 11.1 Introduction
via polymers and solidification of the solvent.
This chapter provides detailed description of 11.1.1 Therapeutic Shortfalls of Poorly
each cryogenic process, formulation Water-Soluble Drugs
guidelines, and characterization analyses.
Finally, examples of cryogenically engineered In modern drug discovery processes, routine use
drug compositions with improved in vitro and of high-throughput screening, combinatorial
in vivo macroscopic performance are provided chemistry, and computer-aided drug design
to illustrate the potential benefits of cryogenic appear to result in a higher prevalence of lead
technologies, especially TFF. The current compounds of increased molecular weight and
authors would like to thank and acknowledge lipophilicity, despite the high efficiency of the
the significant contribution of the previous automated processes. About 40% of approved
drugs, 90% of new chemical entities, and 75%
S. Sahakijpijarn · C. Moon · R. O. Williams III (*) of compounds under development exhibit poor
Division of Molecular Pharmaceutics and Drug Delivery, aqueous solubility (Rodriguez-Aller et al. 2015;
College of Pharmacy, The University of Texas at Austin, Kalepu and Nekkanti 2015). Among which, for
Austin, TX, USA those with high permeability through
454 S. Sahakijpijarn et al.
biomembranes, classified as Biopharmaceutical membranes to exert therapeutic effects (Martin

Classification System (BCS) Class II drugs, the et al. 1993). Salt forms can improve wettability
poor dissolution rate limits drug molecules and bioavailability of drugs. For example,
released into biological fluid contacting the albendazole salts exhibited better wettability due
absorbing mucosa (Rasenack and Muller 2002). to the hydrophilic and ionic nature of their
Basically, it is in the form of an aqueous solution crystals. Microenvironmental pH changes also
that a drug can be absorbed into the systemic affect the solubility of the salt forms, therefore
circulation and exert its therapeutic effect. Conse- some salt forms are superior to others. Particle
quently, the poorly water-soluble drugs often size reduction typically helps increase solubility
result in low and highly variable bioavailability, of the drug. However, in some cases, this is not
and sub-optimal therapeutic effects in patients, possible due to the poor wettability (Paulekuhn
particularly when delivered via the oral adminis- et al. 2013).
tration (Muller et al. 2001; Patravale et al. 2004). Mechanical milling was generally used to
reduce particle size. However, it generates
particles with irregular shape and a wide range
11.1.2 Solid Dispersion/Solution, of size distribution. Moreover, thermo-
Supersaturation degradation is another issue associated with mill-
ing process. Spray drying is also not an ideal
To improve the bioavailability of poorly water- method of choice due to only 50% dry product
soluble drugs, defined as the rate and extent of the recovery (Esclusa-Diaz et al. 1996). Leleux and
drug that reaches systemic circulation, new Williams (2014) recently reviewed mechanical
technologies and innovative formulations and reduction methods regarding particulate systems.
drug delivery systems were explored to improve There are various bottom-up and top-down
dissolution properties of poorly water-soluble techniques utilized to produce crystalline drugs
drugs throughout the past decade. These include in the micron size range. The pros and cons of
more conventional techniques such as the use of both approaches were discussed in terms of their
surfactants, cyclodextrin inclusion complexation, contribution to the pharmaceutical field. How-
emulsification processes, co-solvency, salt forma- ever, the top-down methods have demonstrated
tion, powder milling, and spray drying. However, greater efficiency in the industrial scale. The
these attempts have been of limited success, and authors stated that micronization of some
each was found with inherent problems of their extremely low solubility drugs did not have a
efficacy or stability in the final product. substantial impact on their solubility. However,
The increased amount of excipients required to size reduction to the nanoscale particle size range
formulate the poorly water-soluble drugs may was possibly an effective method to enhance sol-
potentially increase side effects, resulting in low ubility of poorly soluble drugs.
patient compliance. Alternatively, invasive dos- To overcome these shortcomings, novel
age forms such as parenteral formulations have to technologies such as hot-melt extrusion, particle
be developed to address the challenges being engineering by use of supercritical fluid,
presented. However, with even less nanomilling, and solution-based micro-/nanopar-
pharmaceutically acceptable excipient options, ticle precipitation (Betageri and Makarla 1995;
solubilization of drug is practically limited (Liu Mawson et al. 1997; Rogers et al. 2001; Sarkari
2000). et al. 2002; Hu et al. 2004b; Matteucci et al. 2007)
Salt formation is one of the commonly used have been developed to enhance saturation solu-
means to increase aqueous solubility of poorly bility of poorly water-soluble drugs. It may be
water-soluble drugs. Despite the unionized form practical to increase apparent solubility and/or
being much less soluble than its salt, of further the dissolution rate via alteration of the solid-
interest, therapeutically, it is the unionized form state form.
that more readily penetrates biological
11 Pharmaceutical Cryogenic Technologies 455
Solid dispersion typically refers to systems in surface area, it does not increase equilibrium sol-
which drug particles are homogeneously ubility. Often, for drugs with very low aqueous
distributed throughout a solid matrix of excipi- solubility, the achieved increase in dissolution
ent(s). This system provides the possibility of rate is limited and insufficient to provide signifi-
reducing the particle size of drugs to nearly a cant enhancement of bioavailability (Muller et al.
molecular level in order to transform the drug 2001). However, when the drug particle sizes
from the crystalline to partially amorphous mor- were deducted in the 100-nm range, they dissolve
phology. A solid solution results when the drug is more quickly and achieve supersaturation versus
molecularly dispersed throughout a solid matrix, the micronized drug particles, as described by the
i.e., complete amorphous morphology (Kapsi and Noyes–Whitney and Ostwald–Freundlich
Ayres 2001), in which the particle size of the drug equations (Grant and Brittian 1995). Both particle
has been reduced to its absolute minimum with- dissolution kinetics and solubility are size depen-
out any crystalline drug domains (Leuner and dent. Thus, the dissolution of drug nanoparticles
Dressman 2000). The amorphous form of a drug in vivo is usually accompanied by an increase in
has a higher thermodynamic chemical potential bioavailability (Hintz and Johnson 1989; Borm
than its crystalline counterpart. et al. 2006).
Additionally, exposure area of drug to the Furthermore, novel investigation of supersatu-
dissolution media was also greatly enhanced due ration levels of drug nanoparticles in aqueous
to significantly increased surface areas of drug media demonstrated significantly higher values
compositions obtained in this way. Therefore, than the much larger microparticles of drug com-
poorly water-soluble drugs in the solid solution/ position, which have the potential to crystallize
dispersion exhibit higher dissolution rates and during slow dissolution (Matteucci et al. 2007).
higher saturation concentration than their intrinsic Amorphous nanostructured formulations of
solubility of crystalline form of drug, i.e., they poorly water-soluble drugs have also been devel-
can produce supersaturated solutions (Betageri oped to enhance therapeutic effectiveness (Yang
and Makarla 1995). Generation of supersaturation et al. 2008a).
provides higher number of free drug molecules in An emulsion-freeze-drying technique was
the solution available for absorption, thereby used to prepare nanoparticles of poorly water-
leading to enhanced bioavailability. soluble drugs. In situ formation of poorly water-
soluble drug nanoparticles within a porous hydro-
philic polymer (PVA) scaffold by freeze-drying
11.1.3 Solubility Advantage o/w emulsions has been accomplished. The pore
of Nanoparticles structure maintained the drug nanoparticles and
prevented agglomeration. The nanoparticles of
Production of drug-loaded nanoparticles for the the model poorly water-soluble drug loaded in
poorly water-soluble drugs is an alternative and porous polymer rapidly dissolved in water
promising approach to overcome their low aque- (Grant and Zhang 2011).
ous solubilities and the consequential low He et al. (2013) reported that the dissolution of
bioavailabilities (Muller et al. 2001). a model poorly water-soluble drug, indometha-
Nanoparticles are currently defined as single cin, was significantly improved due to the
particles with a diameter less than 100 nm. resulting nanoparticles prepared by a cryogenic
Agglomerates of nanoparticles can be larger technique. However, preventing nanoparticles
than 100 nm in diameter but may be from aggregation and agglomeration is also chal-
de-agglomerated either with weak mechanical lenging. Nanosuspensions of indomethacin with
forces or by dispersing in a solvent. food proteins as novel stabilizers were prepared
Although micronization can increase dissolu- by a nanoprecipitation–ultrasonication method
tion rate of the poorly water-soluble drugs by followed by freeze drying. The nanosuspensions
reduction of particle size and thereby increased were lyophilized, and the original particle size
and particle-size distribution were maintained. inducing a rapid change in solute solubility to
The protein stabilizers physically stabilized indo- generate solid particles.
methacin nanosuspensions via a combination of Another example of novel bottom-up process
two mechanisms: electrostatic repulsion and ste- using freeze drying to improve the dissolution
ric stabilization. behavior of poorly water-soluble drugs was
Published literature reviews have discussed reported by De Waard et al. (2008).
nanoparticle technologies. For example, a review Nanocrystalline particles of fenofibrate produced
of nanoparticle engineering processes for the by the “controlled crystallization during freeze
enhancement of dissolution rates of poorly drying” process significantly increased dissolu-
water-soluble drugs was published by Hu and tion of the drug compared to tablets containing
co-workers (2004b). This overview focused on the physical mixture. The freeze-dried formula-
the commercially viable nanoparticle engineering tion contained fenofibrate, a drug with low Tg,
processes available for increasing the dissolution and mannitol as a carrier since it crystallizes.
properties of poorly water-soluble drugs. Cryo- Freezing rate and the ratio of water-to-tertiary
genic spray processes and spray freezing into butyl alcohol (TBA) affected the nucleation rate
liquid (SFL) were proposed as alternative that controlled the size of crystals and dissolution
approaches to prepare nanoparticles of poorly performance. The challenge is how to apply the
water-soluble drugs. Examples using these process to high Tg drugs.
techniques were also included. Besides, bottom- There are some recently reported examples of
up technologies utilized to prepare the bottom-up particle engineering based upon
nanosuspensions of poorly water-soluble drugs cryogenic technologies. For example, Yasmin
were recently reviewed by Du et al. (2015). et al. (2014) successfully utilized lyophilization
Spray freeze drying and freeze drying were used to prepare silica–lipid hybrid (SLH)
as the “solidification methods for the microparticles from submicron o/w emulsions
nanosuspension prepared by the bottom-up stabilized by silica nanoparticles. The authors
approach.” The article included other solidifica- reported that the performance of microparticles
tion techniques, advantages, manufacturing pro- fabricated from this technique were comparable
cesses, the corresponding characterization to those produced using spray drying. Hence this
methods, drug dosage forms, and limitations of technology is another alternative method to man-
commercial drug products. ufacture SLH formulations of poorly water-
soluble drugs which are thermally labile and
also other challenging drugs to process through
spray drying.
11.1.4 Overview of Cryogenic
To the contrary, a comparison of two
Technologies
processing approaches (spray- and freeze-drying)
to manufacture solid phospholipid nanoparticles
Cryogenic particle engineering technologies were
of celecoxib, a BCS Class II drug, was reported.
developed to improve the solubility and dissolu-
Celecoxib with phospholipid E80 and trehalose
tion properties by creating nanostructured amor-
was formulated in various drug-to-excipients
phous particles with dramatically enlarged
ratios. Spherically shaped, amorphous
surface area at very low temperature conditions,
nanoparticles (average diameters <1 μm) were
in contrast to micronized crystalline form of
produced by spray-drying; while larger particles
poorly water-soluble drugs (Hu et al. 2002,
of the matrix-like structure of solid amorphous
2004a, b, c; Rogers et al. 2003a; Overhoff et al.
phospholipid dispersion were prepared by freeze-
2007a). Cryogenic technologies basically can be
drying. Both products significantly improved the
categorized into micro-/nanoparticle precipitation
dissolution of the drug. The apparent and molec-
technologies or the so-called “bottom-up” particle
ular solubility of the drug from the spray-dried
engineering technologies, with the mechanism of
formulation in phosphate buffer (pH 6.5) and in
biorelevant fasted state simulated intestinal fluid in vivo performance will be provided and
(pH 6.5) were considerably higher than the analyzed. It can be readily recognized that
freeze-dried powder. This probably resulted applications of these cryogenic technologies
from the difference in particle size and surface may be applied to other freezing processes.
morphology (Fong et al. 2015).
As well known, solubility is largely deter-
mined by the bonding interactions between the
11.1.5 Commonly Used Cryogens
solute and solvent on a molecular level; it is also
heavily influenced by external factors, including
The most commonly used cryogen in cryogenic
temperature, pressure, polarity of the solvent, and
processes is liquid nitrogen. Liquid nitrogen is a
pH. Sudden shift of one of these factors can
colorless, odorless, and nonflammable liquid that
generate strong driving forces to induce nucle-
boils at a temperature of approximately 195 C.
ation leading to particle formation. Cryogenic
Due to its natural abundance in the atmosphere,
technologies utilize cryogens, such as liquid
liquid nitrogen is relatively cheap and is readily
nitrogen, to induce abrupt temperature change of
available in large quantities throughout the world.
a system containing solubilized poorly water-
Compared to other liquid cryogens, it is relatively
soluble drug molecules alone or along with excip-
safe and is already accepted for use in certain
ient molecules. The rapid cooling rates of up to
medical applications. The primary disadvantages
1.0 106 K/s may produce stable amorphous
of liquid nitrogen are that it can pose a safety
nanostructured particles with significantly
hazard as an asphyxiant and it also exhibits a
enlarged surface areas to facilitate rapid and
behavior known as the Leidenfrost effect,
higher-level dissolution in biological fluids,
which, in some instances, may actually result in
thereby enhancing bioavailability.
decreased freezing rates for droplets that come
Generally, cryogenic technologies involve the
into contact with the liquid nitrogen reservoir
rapid freezing of single solvent or co-solvent-
but are shielded by a vapor layer for some amount
based solution containing drug alone or along
of time before direct contact can be made with the
with stabilizer or solubility enhancer. The
liquid nitrogen.
generated frozen material is then lyophilized to
Liquid argon is the most common alternative
remove the solvent by sublimation like traditional
to liquid nitrogen for use in SFD. It has many of
lyophilization and atmospheric freeze drying
the same advantages as liquid nitrogen and also
(ATMFD) (Derle et al. 2010), thus yielding a
suffers from some of the same disadvantages such
dry powder of high surface area. Cryogenic pro-
as being an asphyxiating agent. It boils at nearly
cesses are defined by the type of injection device
the same temperature as liquid nitrogen at
(capillary, rotary, pneumatic, and ultrasonic noz-
185 C; however, it is typically more expensive
zle), location of nozzle (spraying onto or into a
than liquid nitrogen and less widely available.
cryogenic liquid, or applying the solution onto a
Another group of alternatives is liquid
cryogenic substrate), and the composition of
hydrocarbons, such as liquid propane, pentane,
cryogenic liquid, such as liquid hydrofluor-
and hexane. They have relatively higher boiling
oalkanes, liquid nitrogen, liquid argon, com-
point and wider availability than liquid argon.
pressed fluid carbon dioxide, and organic
However, these systems are considerably more
solvents.
dangerous to work with due to their extreme
In this chapter, spray freeze drying (SFD),
combustibility. Additionally, these materials
SFL, and thin film freezing (TFF) cryogenic
have not been accepted or tested in use with
technologies for pharmaceutical applications
pharmaceutical products and could be a potential
will be discussed in detail. Examples of recent
source of contamination when brought into direct
studies using the cryogenic technologies with
contact with drug products.
step-by-step procedures to engineer poorly
water-soluble drugs for improved in vitro and
11.2 Cryogenic Technologies improvement in in vitro dissolution and oral bio-

availability compared to unprocessed baicalein
11.2.1 Spray Freeze Drying (SFD) (He et al. 2011).
SFD has been used not only to improve the
SFD has been used in pharmaceutical research for solubility and dissolution of poorly water-soluble
over 60 years and is likely the oldest of the drugs but also used to enhance the permeability of
cryogenic pharmaceutical processing BCS class IV compounds. Oleanolic acid, a BCS
technologies. One of the first published papers Class IV compound, showed a poor oral bioavail-
in the literature describing this technique was in ability of only 0.7% in rats. Compared to the
1948, which was used as a means to produce commercial tablet, the kinetically stable, amor-
protein powders with varying surface areas for phous solid dispersions of the oleanolic acid,
subsequent absorption isothermal analysis with a stabilizer, a wetting agent, and a penetra-
(Benson and Ellis 1948). SFD has historically tion enhancer, produced by SFD technique was
been used as a method to process thermally labile superior in terms of in vitro dissolution and
compounds such as proteins and peptides as well uniform absorption. Inter-individual variability
as even larger biological molecules in both the in oral absorption is common for the BSC Class
pharmaceutical and food industries (Costantino IV drugs. This SFD-processed formulation not
et al. 2000, 2002; Anandharamakrishnan et al. only improved drug absorption but also reduced
2010; Ishwarya et al. 2015) because, unlike the large variability of intestinal permeability,
spray drying, no heat is required to obtain the which is the critical problem of the compounds
final powder formulation. In addition to proteins in this class (Tong et al. 2011).
and peptides, SFD has also been applied to differ- Encapsulation has been used to improve the
ent poorly water-soluble drugs in order to bioavailability of lipophilic bioactive compounds
enhance their solubility and non-pharmaceutical (Turchiuli et al. 2014). Unlike other techniques,
applications such as food processing SFD does not require heat and high temperature,
(Mumenthalera and Leuenberger 1991; Zijlstra making this technique more suitable for heat-
et al. 2007; Tong et al. 2011; Niwa et al. 2012, sensitive bioactive compounds like vitamin
Niwa and Danjo 2013; Wanning et al. 2015). E. Parthasarathi and Anandharamakrishnan
SFD has been applied to fabricate kinetically (2016) compared the properties of vitamin E
stable, amorphous solid dispersions. Adeli et al. microcapsules preparing using three different
reported the significant improvement of dissolu- techniques: spray drying, freeze-drying, and
tion rate of the solid dispersion of azithromycin spray freeze-drying. Although there is no signifi-
(BCS class II drug) and polyvinyl alcohol cant difference in encapsulation efficiency,
prepared using spray freeze drying, which in vivo bioavailability study in rats demonstrated
contributes to an increase in saturation solubility that SFD microcapsules enhanced the oral
of dispersion SFD compared to that unprocessed absorption of vitamin E relatively higher than
drug (Adeli 2017). Another study from He’s the other two techniques. This is possibly related
group also reported that baicalein, a BCS class to the porous structure of SFD microcapsules,
II drug, was in the amorphous state when which can improve the dissolution and absorption
prepared a solid dispersion with Pluronic F68, a rate in the gastrointestinal tract (Parthasarathi and
low Tg and Tm carrier, by spray freeze drying, Anandharamakrishnan 2016).
while the drug was recrystallized in the solid SFD has also been applied in the pulmonary
dispersion containing same compositions delivery of poorly water-soluble drugs. A solu-
prepared by the solvent evaporation method. tion of the poorly water-soluble drug, cyclospor-
The amorphous state of baicalein provided ine, with mannitol was processed through the
benefits in physical and chemical stability over SFD technique to prepare a dry powder for inha-
storage at 40 C up to six months, and lation. SFD produced porous microparticles with
a high specific surface area. The drug was in an
amorphous state while mannitol crystalized dur- the liposome vehicle can entrap the drug
ing the freeze-drying process. The hydrophilic molecules, which helps to sustain the drug release
property of mannitol promoted dissolution of the in the pulmonary system and reduces the potential
drug. Moreover, drug-mannitol co-formulation of systemic adverse effects (Rudokas et al. 2016).
exhibited better aerosol dispersion since mannitol Ye et al. (2017) reported ultrasonic spray freeze
successfully improved adhesive and cohesive drying is a suitable method to prepare inhalable
behavior between the drug particles (Niwa et al. clarithromycin liposomal dry powders. The use of
2012). mannitol and sucrose as co-lyoprotectants
The formation of inhalable micronized porous resulted in high entrapment efficiency up to 80%
particles containing the poorly water-soluble cor- and a narrow size distribution. Additionally, the
ticosteroid, budesonide, and mannitol using the formulations were stable after 3-month storage at
SFD technique was investigated. The excipients, 25 C and exhibited high aerosolization effi-
hydroxypropyl beta-cyclodextrin and/or ciency (emitted dose >85%, 43–50% fine particle
l-leucine, were used in the formulation. The fraction) (Ye et al. 2017).
results demonstrated that both excipients at a Several studies combined spray freeze drying
suitable ratio influenced the particle shape and with other techniques to produce an inhalable
morphology as well as improved aerosol perfor- powder. Niwa and Danjo (2013) developed a
mance and dissolution of the particles method combining the wet milling and the SFD
(Parthasarathi and Anandharamakrishnan 2016). techniques to generate a novel product for poorly
Yu et al. (2016) recently compared the water-soluble drugs. The suspension was
properties of the coarse-carrier-free dry powder prepared by wet milling of drug in an aqueous
inhalation formulations of ciprofloxacin nanoplex medium of polymer as a dispersing agent and
by spray drying (SD) and SFD. The formulations then sprayed via SFD to yield a dry nanosized
contained D-mannitol as a drying adjuvant and powder. The drug in the porous network structure
L-leucine as an aerosol dispersing agent. The spontaneously released into the nanoscale sus-
SFD powder exhibited superior aerodynamic pension and rapidly dissolved in both acidic and
properties as it showed a higher emitted dose neutral media. The SFD formulation exhibited a
(92% vs. 66%), higher fine particle fraction better dispersibility and more efficient dispersing
(29% vs. 23%), and smaller mass median aerody- into a nanosuspension than the product prepared
namic diameter (3 μm vs. 6 μm). The higher by spray drying (Niwa and Danjo 2013).
aerosol performance of SFD powder is possibly Recently, a microfluidic reactor has been cou-
related to low density, high porosity of SFD pow- pled with ultrasonic spray freeze drying to pre-
der, and higher leucine content in the SFD formu- pare inhalable powder. Saboti et al. (2017)
lation, compared to SD formulation. It was prepared budesonide particles using an
reported that higher content of leucine resulted antisolvent precipitation method. The drug solu-
in the crumpled particles instead of spherical tion in organic solvent was mixed with aqueous
particles, thereby limiting the amount of leucine solution in a T-junction microfluidic chip, and
in SD formulations. Although the SFD powders then form a drug suspension. Precipitated
that contained higher leucine content showed particles in suspension were continuously
poorer aqueous reconstitution as compared to atomized via an ultrasonic atomization probe
the SD powder, which may negatively affect the into a container containing liquid nitrogen. This
mucus penetrating ability in bronchiectasis lungs, strategy can avoid the additional step of homoge-
there is no difference in the antibiotic release nization and the need for a stabilizer. The in vitro
profile and antimicrobial activity between both aerodynamic testing showed that fine crystalline
engineered powders (Yu et al. 2016). BDS powders were aerosolizable with optimal
SFD has also been applied to develop aerosol performance (47.6% 2.8% to
liposomal dry powder inhalation. Liposomes for 54.9% 1.8% fine particle fraction [Saboti
pulmonary delivery have gained more interest as et al. 2017]).
In addition to poorly water-soluble drugs, SFD aspect to consider is the type of nozzle and atom-
has also been applied to vaccine delivery of ization parameters utilized. To achieve atomiza-
biologics. Aluminum salts such as aluminum tion, one of three different types of atomizing
hydroxide, aluminum phosphate, or alum are nozzles is typically used, including hydraulic
insoluble compounds that are widely used as (pressure) nozzles, pneumatic nozzles, ultrasound
adjuvants since clinical studies have or vibration nozzles, and monodisperse droplet
demonstrated that they can enhance the antigenic- generators. The primary trade-off in these nozzles
ity and improve the efficacy of some vaccines is particle-size control versus liquid
such as diphtheria and tetanus toxoids (Baylor processing rate.
et al. 2002). It was reported that slow freezing Hydraulic nozzles transform the pressure into
caused alum particles to coagulate and agglomer- kinetic energy, which forces the fluid to pass
ate, leading to the stability problem (Maa et al. through an orifice. Subsequently, the liquid is
2003). Maa et al. compared the effect of freeze broken down and dispersed into small droplets.
drying, spray drying, air drying, and SFD on The droplet size depends on the feed rate, viscos-
preserving alum’s adjuvant effect in the alum- ity, and spraying pressure (Adali et al. 2020).
hydroxide-adjuvanted hepatitis B surface antigen Pneumatic nozzles require the atomization
and the alum phosphate-adjuvanted diphtheria energy of a compressed gas flow to provide
and tetanus toxoids. It was found that the fast shear force to the liquid and generate a wide
freezing rate of SFD decreased the tendency of range of droplet sizes. The two-fluid nozzles are
alum coagulation as it produced homogeneous the most popular configuration of pneumatic
suspensions upon reconstitution. In contrast, nozzles for pharmaceutical applications. They
other processes caused alum coagulation and a allow the highest processing rates up to 15 L/
significant loss in immunogenicity. min and suitable for viscous solutions. Although
Wanning et al. (2015) and Vishali et al. (2019) the particle-size distribution created by these
recently summarized several applications of SFD nozzles can easily span several orders of magni-
in drug delivery. SFD exhibited advantages not tude, these nozzles better control the particle size
only for poorly water-soluble small molecule distribution than hydraulic nozzles (Adali et al.
drugs but also for delivery of biologics through 2020). The particle-size distribution of a given
various routes of administration such as pulmo- spray created using a two-fluid nozzle is primarily
nary, nasal, epidermal, and intradermal ballistic controlled by the properties of the liquid formula-
routes (Vishali et al. 2019; Wanning et al. 2015). tion (i.e., the surface tension and viscosity), noz-
zle geometry, and the flow rate of the liquid and
Process of SFD (Fig. 11.1) atomizing gas. Two-fluid nozzles are preferred
SFD is generally described as a three-step process when fast processing is required but droplet-size
involving the atomization of a drug feed solution distribution, hence final particle-size distribution,
or suspension, freezing the atomized droplets, and is not critical. Another disadvantage of the
sublimation of solvent from the frozen material to two-fluid nozzle system is that the large volume
obtain a final, typically amorphous, dry powder of atomizing gas utilized by the nozzle can
composition. The equipment and experimental decrease the efficiency and effectiveness of the
set-up used in SFD are fairly straightforward; cryogenic vapor into which the droplets and
however, there are a few choices to consider at atomizing gas are being sprayed. This can
each of the three steps of the process. increase the cost of the overall process and the
In the first step of SFD, solutions or rate at which cryogen is consumed.
suspensions of drug alone or along with excipi- Ultrasonic nozzles allow for relatively high
ent(s) are prepared and atomized into small processing rates, potentially up to 100 mL/min,
droplets using specialized fluid nozzles or vibrat- with a better control of particle-size distribution,
ing orifice droplet generators, over a cryogenic and more importantly, do not utilize large
vapor to achieve rapid freezing. The primary volumes of atomizing gas to generate droplets.
Fig. 11.1 Schematic diagram of the spray freezing drying process. (Reproduced from Vishali et al. (2019) with
permission from Elsevier)
The particle-size distribution of a spray created reservoir, which can further ensure that the freez-
using these nozzles is primarily controlled by the ing process is completed. The most commonly
properties of the liquid formulation, the properties used cryogen in SFD is liquid nitrogen vapor.
of nozzle (i.e., orifice size and atomizing surface When processing poorly water-soluble drugs
area), and the frequency of nozzle vibration. using SFD, organic solvents are typically used
Ultrasonic nozzles are preferred for applications to prepare the drug in solution/suspension. Very
where both control of particle size and reasonably low vapor temperatures are therefore required to
high processing rates are needed. ensure the freezing point of the solvents reached.
Monodisperse droplet generators are used Once the droplets of drug in solution/suspen-
when extremely precise controls of droplet and sion have been frozen, the final step is removal of
particle-size distributions are required. These the solvents by sublimation to obtain dry powder
systems utilize the same technology present in form of the engineered drug composition. Tradi-
ink-jet printing systems to create a controlled tional lyophilization or ATMFD is typically
monodisperse droplet size; however, the primary employed to sublime the frozen solvents. In either
disadvantage of these systems is that they have case, process conditions during sublimation must
very low processing rates of about 0.1 mL/min. In be controlled precisely to ensure that no melting
addition to slow processing rates, these systems occurs, which could potentially undo any advan-
are also more prone to clogging and can process tageous physical properties imparted during the
only very low viscosity solutions. rapid freezing of SFD process.
The second step in SFD is the freezing of the
atomized droplets of drug solution using a cryo-
Conventional Lyophilization
genic vapor. In many cases, the cryogenic vapor
Conventional lyophilization is conducted at
is created over a cryogenic liquid reservoir. When
reduced pressure with vacuum level of few hun-
the atomized droplets fall through the vapor
dred millitorr and low shelf temperature to main-
phase, they then encounter the cryogenic liquid
tain the frozen SFD-processed material. The
lyophilization is broken into two stages known as Leuenberger 1991; Rogers et al. 2003a). Its use in
the primary and secondary drying stages. During the food industry has been entirely focused on the
primary drying, the temperature and pressure removal of water as a solvent and for this process
inside the lyophilizer are typically kept below it has shown great promise. In specific cases,
that of the triple point of the solvent. This especially with freezing processes that result in
promotes the sublimation of the frozen solvent more porous or discrete ice chips/particles (e.g.,
from the solid phase directly to the gaseous SFD, SFL, and TFF), ATMFD has actually been
phase, without allowing any melting. After all or proven to be a much faster and more energy-
the majority of the bulk solvent has been efficient sublimation process due to the vastly
removed, the process is then shifted to the sec- increased heat and mass transfer afforded by the
ondary drying stage, where the shelf temperature use of a fluid-bed configuration (Mumenthalera
is typically brought up to room temperature or and Leuenberger 1991). However, in the case of
higher to remove the molecularly bounded sol- poorly soluble compounds, the process is compli-
vent. In most cases, convention lyophilization is cated by the use of cosolvent systems. As with
utilized as it is well studied and accepted for traditional lyophilization techniques, cosolvent
pharmaceutical process. systems using organic solvents with lower melt-
The enhancement of dissolution rate and oral ing points may prove to be problematic due to the
bioavailability of solid dispersions of valsartan, a low processing temperatures required as well as
poorly water-soluble drug, prepared by a freeze- the large volumes of gas separation that may be
drying technique was reported. The alkaline solu- required. When removing water as a solvent,
tion without organic solvent containing the API, standard refrigeration techniques can be used on
hydrophilic polymers, an alkalizer, and a surfac- the large scale to dry and recirculate desiccated air
tant was lyophilized. Hydrogen bonds between to ensure that the partial pressure of water vapor
drug and polymer were detected by FTIR. The in the air is low, but when the solvent phase is
resulting amorphous solid dispersions showed something other than water it may be more diffi-
greater dissolution rate and significantly higher cult to remove these vapors from the recirculated
oral bioavailability as compared with the bulk dry gas. Theoretically, it should be possible to
drug substance (Xu et al. 2014). utilize ATMFD for the sublimation and drying
of drug products created using cosolvent systems,
Atmospheric Freeze Drying (ATMFD) but currently, very little research exists on this
ATMFD is conducted at atmospheric pressure topic.
conditions and thus does not require a vacuum Rahman and Mujumdar (2012) reported
system. In this case, a cold desiccated gas (typi- advantages and limitations of the ATMFD pro-
cally air or nitrogen) is circulated in and around a cess, a comparison between vacuum freeze drying
mass of frozen product in a fluid-bed-style con- and ATMFD, a novel approach to ATMFD, types
figuration. For these systems to work properly, of ATMFD and their applications. In addition,
the circulation gas needs to be colder than the mathematical models for ATMFD process opti-
melting point of the solvent to ensure that no mization were also presented. The models were
melting occurs. Besides low temperature, the cir- related to energy, heat, and mass transfer of the
culation gas needs to have a very low partial drying process. Studies of the drying process
pressure of the solvent vapor so that a mass trans- sought to overcome the limitations of ATMFD.
fer driving force is available to allow for sublima-
tion of the frozen solvent. ATMFD has been Continuous SFD Process
utilized and well-studied in the food industry Several continuous SFD processes have been
(Meryman 1959; Boeh-Ocansey 1983; Rahman suggested for manufacturing a large scale of pow-
& Mujumdar 2012) but only recently, in the past der. The first design was suggested by Rey et al
15–20 years, it has begun to gain some traction in (Fig. 11.2a). The liquid is dropped into the cham-
the pharmaceutical literature (Mumenthalera and ber that is filled with cold air stream. Then the
frozen spherical samples are dried and transported The new design, LYnfinity, uses a single drop
on the heated conveyors along a vacuum cham- atomizer instead of a multiple-channel atomizer,
ber. Finally, the dried powder is continuously which can minimize atomization shear stress due
transferred and filled in vials. Later, this design to no requirement for high process feed pressure
was modified by the IMA Group (Fig. 11.2b). (Adali et al. 2020).
Fig. 11.2 Schematic

diagram of continuous
spray freezing drying; (a)
Rey’s concept; (b)
LYnfinity; (c) ULVAC
technologies and Azbil
Telstar Technologies; (d)
Meridion Technologies; (e)
Stirred Freeze-Drying.
Figure 11.2. (a), (b), (c),
and (d) are reproduced from
Adali et al. (2020) with
permission from MDPI.
Figure 11.2. (e) is
reproduced from Pisano
et al. (2019) with
The second proposed design is developed by manufacture monoclonal antibody microspheres

ULVAC technologies and Azbil Telstar at an industrial scale. This design is combined a
Technologies (Fig. 11.2c). This design is a fully spray freezing chamber (SprayCon) and a cylin-
closed system that combines the mass-production drical rotary drum (LyoMotion) in a fully
system and the fine-spray freeze-drying technol- contained continuous line (Fig. 11.2d). A nozzle
ogy. A nozzle sprays the liquid into droplets in a on the top of the chamber is used to spray and
vacuum chamber. The droplets are self-freezing break up the liquid into a droplet. Then the
while falling in the vacuum chamber. The frozen droplets are frozen while falling into the freezing
samples are collected and continuously dried on chamber cooled with gaseous and/or liquid nitro-
the heating/cooling belt in the vacuum chamber gen. The frozen droplets are continuously trans-
(Adali et al. 2020). ferred to the precooled drum of the rotary dryer.
For the third design, Meridion Technologies In the drying chamber, the frozen droplets are
(Germany) introduced a continuous process to gently mixed while dynamic freeze-drying under
vacuum by the sublimation energy from radiating enhanced control is due to the fact that geometric
sources and temperature-controlled surfaces of particle size is fixed during droplet formation and
the drum. Subsequently, the homogenous dried subsequent freezing so that a specific final particle
powder is obtained after the drying cycle and can size can be selected by controlling the size of
be continuously transferred to a container (Adali droplet that is produced during atomization,
et al. 2020). whereas with spray drying, significant droplet
Another proposed design is called Stirred shrinkage and deformation occur due to surface-
Freeze-Drying or Active Freeze Drying tension forces imparted as the droplet dries. Aero-
(Fig. 11.2e). It is designed to address the forma- dynamic diameter, on the other hand, is a function
tion of lumps during the conventional tray-type of both geometric particle size and density and, in
drying. The liquid is fed and frozen in a freezing this case, density can also be controlled indepen-
medium and/or jacket cooling. Then the frozen dently from geometric size by controlling the
sample is dried under controlled pressure in the solids loading and formulation parameters used
drying chamber with a stirrer. The heat required to prepare the drug solution that is subsequently
for sublimation is obtained from the heat from the processed via SFD (Maa et al. 2004).
jacketed wall, which is transferred and distributed
by a stirrer in the chamber. The continuous Disadvantages of SFD Process
mixing helps to enhance the heat transfer rate Although several reports have presented the
and distribution over the conventional tray-type unique features and applications of SFD, this
freeze dryer, which has the potential to shorten technique have some limitations. Compared to
the drying cycle. Moreover, the final dried sample other cryogenic techniques such as shelf freeze-
is more homogenous than a conventional freeze drying, SFL, and TFF, SFD process can result in
dryer (Pisano et al. 2019). decreased activity of biological compounds when
excipients are not used. SFL and TFF result in
Advantages of SFD Process even higher level of biological activity, because
One potential advantage of SFD over other cryo- the large droplets used in these processes have
genic processes is that it typically results in the relatively less air–water interface. Whereas SFD
highest rate of freezing at around 106 K/s when utilizes atomized fine droplets, a larger air–water
appropriately small droplet sizes are prepared interfacial region is created where absorption and
(Engstrom et al. 2007b). This rapid freezing can denaturation of proteins can occur. However, sev-
be useful when working with poorly water- eral recent studies have shown that this effect can
soluble drugs that may have very rapid rates of be minimized with the incorporation of excipient
recrystallization, or when very high surface area (s) (stabilizer, complexing agents,
powders are needed. cryoprotectants, and surfactants) (Maa et al.
The most common usage of SFD to date has 1999; Costantino et al. 2000, 2002; Yu et al.
been in the area of the preparation of proteins and 2006).
peptides for inhalation as a competing and gentler SFD is relatively expensive from a
alternative to traditional spray-drying approaches. manufacturing perspective compared to freeze
In contrast to spray drying, SFD process lacks the drying and spray drying since it requires low
usage of high and potentially damaging heat to pressure, low temperature, and cryogenic liquid.
dry powders and provides very high production The use of cryogenic liquid such as liquid nitro-
yields of more than 95% versus about 50% for gen requires additional safety protocols. Cur-
spray drying (Maa et al. 1999). rently, a few companies have tried to scale up
Additionally, the SFD process allows for inde- the manufacturing of SFD powder. For example,
pendent control over both the aerodynamic and Meridion Technologies manufactured a produc-
geometric particle size of prepared powders, tion scale of monoclonal antibodies using a semi-
which are critical parameters in the development continuous process of SFD (Wanning et al. 2015),
of formulations for inhalation delivery. This while Amgen manufactured a pilot scale of
protein powder using SFD (Nguyen et al. 2004). upon contacting the cryogen. A schematic dia-
Additionally, several proposed designs have the gram of SFL process is shown in Fig. 11.3.
potential to be converted for the continuous pro- Then the produced frozen materials were col-
duction of pharmaceutical products. Despite these lected and subsequently dried by lyophilization
potentials, SFD is still challenging for full com- or ATMFD as depicted above, to obtain dry and
mercial use as these factors must be validated for flowable powders of drug compositions.
the transition. The SFL-processed drug powders were gener-
ally characterized by micronized structure with
amorphous morphology, high surface area, and
11.2.2 Spray Freezing into Liquid (SFL) improved wettability in aqueous media,
indicating enhanced dissolution properties of the
SFL, one of the novel cryogenic particle engi- poorly water-soluble drugs (Hu et al. 2002, 2003;
neering technologies, was developed and pat- Rogers et al. 2003a, b).
ented (Williams et al. 2003) and subsequently The benefits of SFL process result from
commercialized by The Dow Chemical intense atomization in conjunction with high
Company. freezing rates. Liquid nitrogen has been typically
employed as the cryogenic liquid due to nearly
Process of SFL instantaneous freezing of the atomized feed liquid
The SFL process involves preparation of a feed resulting from the low boiling point of liquid
liquid, such as an aqueous, organic, or aqueous– nitrogen. The nozzle that used to spray feed liquid
organic cosolvent solution, aqueous–organic in SFL process is composed of an insulating
emulsion, or suspension containing a drug along material such as poly-ether-ether-ketone (PEEK)
with pharmaceutical excipient(s) or drug alone: tubing to prevent premature freezing of the feed
Spray the drug feed liquid under pressure through liquid. As the liquid exits the nozzle, liquid–liq-
a small diameter insulating nozzle directly into a uid impingement occurs between the pressurized
liquid cryogen, such as compressed fluid carbon feed liquid exiting the nozzle and the cryogenic
dioxide, helium, propane, ethane, liquid nitrogen, liquid, such as liquid nitrogen. The estimated
liquid argon, or hydrofluoroethers (Williams et al. cooling rates are strongly related to droplet parti-
2003). The rapid cooling leads to immediate cle size of feed liquid, with higher freezing rates
freezing of the atomized droplets of feed liquid observed with smaller droplet sizes due to higher
Fig. 11.3 Schematic

diagram of spray freezing
into liquid (SFL) process.
(Reproduced from Rogers
et al. (2002a, b) with
surface area available for heat transfer (Maa and surface areas less than 1 m2/g. The SFL process
Prestrelski 2000; Engstrom et al. 2007a, b). The can minimize exposure to the gas–liquid interface
cooling rates of SFL process for two different of droplets containing protein, as the spray nozzle
cryogens, iso-pentane and liquid nitrogen, were was immersed under the surface of the cryogenic
calculated to be 8.9 106 and 1.1 105 K/s, liquid. Thus, the SFL process can reduce protein
respectively (Engstrom et al. 2007a, b). SFL into adsorption, denaturation, and aggregation, and,
iso-pentane produced faster cooling rates despite consequently, lead to higher enzymatic activities
having a higher temperature (160 C) compared than that processed by SFD (Yu et al. 2006;
to liquid nitrogen (196 C). This was attributed Engstrom et al. 2007a, b). Although the cooling
to the boiling of liquid nitrogen around the rate in SFL is about 103 K/s, 3 orders of magni-
inserted spray nozzle and/or sprayed feed fluid tude less than that in SFD, it is sufficiently fast to
formed an insulating layer, known as the arrest the growth of submicron protein particles
Leidenfrost effect (Sitte et al. 1987). (Engstrom et al. 2009).
Pros of SFL Process Cons of SFL Process

The advantages of the SFL process result from Increase in the drug/excipient(s) concentrations
intense atomization of drug feed liquid and the of feed liquid normally leads to increases in vis-
high freezing rates. SFL process used liquid nitro- cosity of the feed liquid. The relatively high vis-
gen nearly exclusively as the cryogen. A very cosity of the feed liquid can limit the application
high degree of atomization is achieved by of SFL, as it can inhibit liquid jet breakup,
spraying directly into the cryogenic liquid as in resulting in slower cooling rates and larger parti-
contrast to spraying into the vapor phase above cle sizes and eventually fibers (Barron et al.
the cryogenic liquid, because liquid–liquid 2003).
impingement occurs between the pressurized Moreover, removal of solvent from the col-
feed solution exiting the nozzle and the cryogenic lected frozen materials by lyophilization is costly
liquid (Rogers et al. 2002b). Thus, high freezing for the equipment (lyophilizers) and is a time- and
rates can be achieved in SFL process due to the energy-intensive process that could take days or
low intrinsic temperature of liquid nitrogen and even weeks to finish (Franks 1992).
the high surface area of atomized droplets of the
drug feed liquid.
Consequently, amorphous morphology of 11.2.3 Thin Film Freezing (TFF)
drug compositions can be formed by SFL pro-
cess, as a result of the ultra-rapid freezing- TFF, also known as cold metal block freezing,
induced simultaneous vitrification of the feed was initially used to cool approximately 100-μm-
solution. The high degree of atomization and thick tissue samples at rates between 100 and
ultra-rapid freezing (URF) rates led to the forma- 10,000 K/s (Gilkey and Staehelin 1986) for
tion of amorphous highly porous nanostructured nonpharmaceutical application. Impingement
particles (Hu et al. 2002). and solidification of liquefied droplets onto a
Moreover, it is advantageous to use SFL pro- cold solid surface have also been used in the
cess to make stable submicron protein drug electrical and semi-conductor industries to add
particles. As discussed previously, SFD has the thin layers of frozen material onto a surface.
potential to cause protein aggregation due to the TFF was also referred as ultra-rapid freezing
large gas–liquid interface in the spraying step. On (URF), spray forming, thermal spray coating,
the other hand, slow cooling by lyophilization splat cooling, slat quenching solidification,
(about 1 K/min) can produce stable protein plasma or powder spray deposition, etc.
particles; however, the particle size was found to (Overhoff et al. 2009). Recently, TFF has been
be a minimum of a few microns in diameter with used as a particle engineering technique to
improve the dissolution profiles of poorly water- freezing rates of the droplets fallen on the drum.
soluble drugs and produce powder formulations To minimize the formation of water-vapor con-
for dry powder inhalation and dry powder form of densation and ice on the steel surface, it is better
vaccines. to place the TFF apparatus in a dry box or
humidity-controlled environment with relative
humidity of less than 15%. Moreover, a blade
TFF Process
made of stainless steel or Teflon is mounted
In a similar manner to the SFD and SFL pro-
along the rotating drum surface to remove the
cesses, the first step in TFF process is a prepara-
ice immediately before the droplets of feed liquid
tion of a feed liquid containing a drug along with
impacting the drum.
pharmaceutical excipient(s) or drug alone. Then,
The surface temperature of the drum can be
droplets of the feed liquid, which are released
monitored by using a surface-moving probe ther-
from a funnel, syringe, or a pump/dripper system
mocouple attachment. When the temperature on
with a controlled flow rate fall from a given height
the steel surface reached a proper level, various
and impact, spread, and freeze on a cooled surface
feed liquids can be applied to the rotating steel
of a cryogenic substrate, as depicted in a sche-
drum drop-wise from a height of approximately
matic diagram in Fig. 11.4. A cryogenic substrate
10 cm. Upon impacting on the cryogenic surface,
is commonly selected from materials with a ther-
the feed liquid droplets (about 2–4 mm in diame-
mal conductivity k between 10 and 20 W/(m.K).
ter) deformed into thin films (about 100–400 μm
A rotating cylindrical stainless-steel drum
in thickness) of disk shape and rapidly cooled
approximately 17 cm in length and 12 cm in
until frozen on timescales of 70–1000 ms, which
diameter with a mirror-polished surface was
corresponds to a cooling rate of about 102 K/s
employed to serve as the cryogenic substrate.
(Fukai et al. 2000; Pasandideh-Fard et al. 2002),
The drum is hollow with 0.7-cm-thick walls and
as illustrated in Fig. 11.5 (Engstrom et al. 2009).
was filled with cryogen such as dry ice or liquid
The frozen disk is scraped off the stainless-steel
nitrogen on the inside. As a result of thermal
surface with a blade prior to one full revolution
conductivity through the steel, the equilibrium
and falls in a collecting pan that is filled with dry
drum surface temperatures were measured to be
ice or liquid nitrogen to maintain the frozen disk.
223 K or 133 K for dry ice and liquid nitrogen,
After processing a batch, the collecting pan
respectively (Engstrom et al. 2009). The exposure
containing the frozen disk is transferred to a
of the cold drum to the atmosphere allowed a thin
lyophilizer where the solvent is removed by sub-
layer of ice to condense on the drum surface,
limation (Overhoff et al. 2007a). The cooling
which may affect the conductivity of the cryo-
rates in TFF and SFL processes are comparable.
genic substrate, and consequently may affect the
Fig. 11.4 Schematic diagram of thin film freezing (TFF) process. (Modified from Sahakijpijarn et al. (2020b) with
permission from MDPI)
Fig. 11.5 Diagram of the a T = 298 K

thin film freezing process Surface Area/Volume
displaying the falling 17 cm–1
droplet (a), spreading after 3.6 mm
impact on the stainless-steel
surface (b), and during
cooling and freezing as a T = 223 K
thin film (c) (drawn to
scale). (Reproduced from
Engstrom et al. (2008) with
permission from Springer) b
Freezing Front Unfrozen Liquid

t < 10 ms
c
Surface Area/Volume
216 μm 46 cm–1
12 mm
Because of rapid conductive heat transfer, temperature. In the case of voriconazole

resulting in high supersaturation and nucleation nanoaggregates produced by TFF, the lower cryo-
rates, TFF process can create powders with high genic surface temperature (150 C) resulted in
surface area by creating nanostructured particles smaller nanoparticles than those processed at the
and enhanced dissolution properties by producing higher temperature (150 C) (Moon et al. 2019).
amorphous solid dispersions, similar to those pro- The lower initial solids content formulation gave
duced by other rapid freezing technologies. As in a better FPF result, possibly due to the increased
other freezing technologies, the rapid freezing of fragility by generating smaller particles (Wang
the feed liquid is critical in preventing phase et al. 2014a; Beinborn et al. 2012; Moon et al.
separation of solute and solvent during freezing, 2019). Moreover, the solvent system composition
allowing for the compositions to molecularly dis- used for TFF significantly affected powder
perse with each other. properties due to different viscosity of the solvent
To tailor the properties of the respirable brittle system and possible cryo-phase separation (Moon
matrix powder, the processing parameters, such et al. 2019). When a different ratio between water
as processing temperature, solid loading, and and 1,4-dioxane in a binary solvent system was
processing temperature, must be considered used to produce voriconazole powder
(Moon et al. 2019; Sahakijpijarn et al. 2020a). formulations with either PVP K12 or PVP K30
The processing parameters used in the TFF pro- by TFF, different crystallinity, specific surface
cess potentially affected the physicochemical and area, and aerodynamic properties were observed
aerodynamic properties of the resulting matrix (Beinborn et al. 2012). Voriconazole
powders. Product with improved properties nanoaggregates made with mannitol also con-
(e.g., greater specific surface area, higher poros- firmed that a different ratio of water and acetoni-
ity, smaller particles, and lower density) was trile in a binary solvent system can affect
prepared at a higher freezing rate by controlling aerodynamic properties (Moon et al. 2019).
the temperature of the cryogenic surface
Fig. 11.6 Scanning electron micrographs of TFF-processed powders containing mannitol and lysozyme, processed at:
(a) 50 C, (b) 100 C. (Reproduced with permission)
A combination of TFF and another technique 2003). The cooling rate of the thin films in TFF
(i.e., template emulsion technique) was also used can be controlled readily by varying the tempera-
to produce an amorphous form of the poorly ture of the metal surface, and the surface temper-
water-soluble drug. The hydrophobic drug was ature of the cryogenic substrate may be measured
encapsulated into the internal phase of the emul- directly. The cooling rate of the samples
sion through the template emulsion method processed by TFF can be controlled, and this
before rapid freezing through TFF process. For- resulted in faster nucleation, induced production
mulation prepared using the TFF-template emul- of the nanoaggregates of voriconazole with
sion technology showed significant improvement smaller nanoparticles at a lower temperature of
of wetting and dissolution properties of the drug the metal surface (Moon et al. 2019). The SEM
(Lang et al. 2014). images in Fig. 11.6 presents that a lower
processing temperature at 100 C generated
Advantages of TFF Process smaller ice channels by rapid nucleation and par-
Although the cooling rate in TFF (102 K/s) is ticle growth prevention.
lower, compared to those of SFD (106 K/s) and For SFL and SFD processes, the complex
SFL (103 K/s), it is still sufficient to produce rapid geometry of the turbulent spray in the liquid
nucleation and to prevent significant particle nitrogen combined with the Leidenfrost effect
growth during freezing. In TFF process, the size can be somewhat difficult to control and monitor
of the unfrozen channels was sufficiently thin and (Sitte et al. 1987). In TFF, more concentrated and
the increase in the viscosity of the unfrozen solu- thus more viscous solutions may be processed, as
tion was sufficiently fast to be able to achieve the droplets are not atomized. However, the thick-
similar particle sizes and morphologies as for ness profile of the film along the radius of the
the moderately faster process, SFL, and the frozen disk may change with viscosity (Overhoff
much faster process, SFD (Engstrom et al. et al. 2009).
2008). The rapid freezing rate of TFF was used TFF process can provide a high yield of
to successfully manufacture amorphous form of products. In TFF, collection of the frozen thin
meloxicam, which could be considered a chal- films of the feed liquid droplets leads to nearly
lenging drug to form amorphous solid dispersion, 100% yields, whereas in SFD process, yields
without excipients (Jermain et al. 2019). were about 80% because of the result of entrain-
TFF on a cold metal surface bypasses the need ment of uncaptured particles in the atomized
to maintain aseptic conditions of a liquid cryogen, aqueous stream, particles attaching to the walls
for example, liquid nitrogen (Gosselin et al. of collection vessels, and inefficient separation of
the cryogen from the 10–100-μm frozen particles Disadvantages of TFF Process
(Johnson 1997; Overhoff et al. 2009). First, maintenance of a low humidity for TFF
TFF process offers the flexibility of the process increases costs for facility design, equip-
amounts of drugs to be processed. By using ment, and operation, especially for commercial
TFF, it is feasible to accommodate either small productions.
quantities (<1 mL) of drug feed solution due to Second, with all freezing processes, the
the high efficiency in the collection of frozen quantities and quality of cryogen required for
films or large-scale production by adding multiple manufacturing production-scale batch sizes
drippers to make droplets in parallel and increas- could also add to production costs. To date, it is
ing the length of the drum. The rotating drum of not sure which of the aforementioned cryogenic
TFF apparatus offers scale-up advantages over processes is the most cryogenically efficient
other cryogenic particle engineering technologies (Overhoff et al. 2009).
by becoming more of a continuous freezing pro- Additionally, similar to the SFL process,
cess. Thus, TFF process is not limited by the removal of solvent from the collected frozen
amount of drug to be processed. It is feasible films by lyophilization is costly, for both equip-
from the early-stage screening of drug in milli- ment (lyophilizers) and energy consumption.
gram quantity to commercial product
manufacturing at a scale of kilograms to tons.
As of 2021, two drug substances for dry powder
11.2.4 Storage of Dried Powders
inhalation developed using TFF are in phase I
clinical trials (National Library of Medicine
The effect of moisture on the properties of
(U.S.), NCT04576325; TFF Pharmaceuticals
low-density microparticles was reported (Watts
2020).
et al. 2013; Sahakijpijarn et al. 2020b). Moisture
TFF process can be scaled suitably for clinical
sorption on the particle surface affects particle
trials while still providing homogeneous powder
morphology, geometric particle size, surface
products. While shelf freeze-dried mannitol pro-
area of brittle matrix microparticles, which subse-
duced heterogeneous polymorph mixtures of α,
quently affect the particle dispersibility and aero-
β, and δmannitol, TFF produced homoge-
solization. For example, tacrolimus dry powder
neous polymorph mixtures of them (Moon et al.
containing 5% w/w of lactose showed a signifi-
2016). Additionally, Sahakijpijarn et al. presented
cant decrease in specific surface area, geometric
homogeneity of the TFF powder formulation of
particle size, which agreed with a collapse of
tacrolimus containing lactose by combining high
brittle matrix structure observed by SEM. The
mass resolution Tof-SIMS and high spatial reso-
change in powder physical properties led to a
lution Tof-SIMS mapping (Sahakijpijarn et al.
significant decrease in aerosol performance after
2020a).
1-month storage at 40 C/75%RH. However, it
In addition to the advantage of being a simple,
was reported that storage of powder in hermeti-
efficient, and robust process for freezing, TFF
cally sealed glass containers with desiccants
also renders improvement in the stability of the
under dry nitrogen helped to minimize the change
protein product due to the minimized gas–liquid
in physical and aerodynamic properties of TFF
interface of the feed liquid, in comparison to SFD
powder. The stability study demonstrated that the
and SFL. It was found that minimizing gas–liquid
aerosol performance, geometric particle size, and
interface can improve protein stability by limiting
specific surface area of TFF tacrolimus/lactose
the amount of protein that can absorb to the
(95/5 w/w) encapsulated powder did not signifi-
interface. The surface area to volume ratio of the
cantly change over six months storage at 25 C/
gas–liquid interface in TFF was 2 orders of mag-
60%RH and only slightly changed over six
nitude lower than in SFD, leading to much less
months storage at 40 C/60%RH (Sahakijpijarn
protein adsorption and aggregation (Engstrom
et al. 2020a). Hence, it is suggested that samples
et al. 2008).
should be transferred in a packaging area, where the supercooling is extremely high (rapid freezing
the humidity is controlled, typically to less than rate), the formation of a vitrified solution may
30% RH. occur in which the nucleation of crystals may be
minimized or fully prevented, leading to an amor-
phous material (Yu 2001; Overhoff et al. 2009).
11.2.5 Mechanism of Rapid During freezing, the supersaturation of the sol-
Freezing-Induced Particle ute in the unfrozen domains as a function of the
Formation phase diagram establishes a driving force for pre-
cipitation to occur. Particle growth under this
Solubility is heavily influenced by external condition can occur through condensation of indi-
properties, including temperature, pH, polarity vidual molecules onto a growing nucleus, coagu-
of the solvent, and pressure. Sudden shift of one lation of two growing particles, or Ostwald
of these properties can induce nucleation, leading ripening, which is the growth of larger structures
to particle formation. Nanoparticles may be at the expense of smaller structures (van de Witte
formed by maximizing the supersaturation to et al. 1996). Rapid nucleation at high supersatu-
induce precipitation instantaneously and then ration all at one time period will produce uni-
arresting growth (Matteucci et al. 2007). Gener- formly sized particles and lower Ostwald
ally, faster nucleation relative to particle growth ripening (Overhoff et al. 2009).
leads to a smaller median particle size and more As depicted in Fig. 11.7, a greater rapid
uniform particle-size distribution. cooling rate will produce a larger number of
Rapid freezing can be categorized as a precip- nuclei and more solid particles separated by thin-
itation technology, where most of the solvent is ner ice domains than in the case of slow cooling
separated from the solutes to form ice and the and slow nucleation. As the cooling domains
solute phase becomes highly concentrated. Upon vitrify, the high viscosity inhibits the further
initiation of freezing a homogeneous solution, the growth of the particles (Engstrom et al. 2007a, b).
formation of frozen solvent particles and a drug/
polymer-rich phase begin to appear (Tang and
Pikal 2004). The rate of cooling in conjunction 11.3 General Guidelines
with other factors, such as solute concentration, for Cryogenic Technology
plays a key role in determining the final particle
size and structure of the solid powders (Overhoff 11.3.1 Selection of Solvent/Cosolvent
et al. 2007a). The rate of growth and number of Systems for Cryogenic
solvent crystals in a freezing solution are deter- Processes
mined by the degree of supercooling. Higher
supercooling results in more/smaller ice crystals The first step in the cryogenic processes is to
and larger ice-specific surface area (Jiang and make feed liquids. Despite the diversity of feed
Nail 1998). As different freezing methods can liquids (solution, suspension, and emulsion) that
produce different supercooling effects, freezing can be employed for the cryogenic processes, a
with liquid nitrogen basically can provide the complete solution of poorly water-soluble drug
highest supercooling, while solutions subjected alone or along with pharmaceutical excipient
to slow cooling rates, for example, freezing with (s) was mostly reported.
the precooled shelf method, give the lowest
supercooling. The solvent in the supercooled Solubility
solution nucleates and forms crystalline solvent Solubility may be defined as the maximum con-
particles which grow during freezing. Increased centration of a substance that may be completely
supercooling, in turn, increases the nucleation dissolved in a given solvent at a given tempera-
rate of frozen solvent particles while minimizing ture and pressure. Solute molecules are held
the time for frozen solvent particle growth. When together by certain intermolecular forces (dipole–
Fig. 11.7 Frozen morphologies of dilute solution with Amorphous ice particles are represented as white domains
high supercooling (a), concentrated solution with high and solute precipitate as solid dots or gray regions.
supercooling (b), dilute solution with low supercooling (Reproduced from Engstrom et al. (2007a) with permis-
(c), and concentrated solution with low supercooling (d). sion from Elsevier)
dipole, induced dipole–induced dipole, ion–ion, chemical nature of the participating molecules.
etc.), as are the molecules of the solvent. In order The dissolution of hydrophobic materials, which
for dissolution to occur, these cohesive forces of can dissolve readily in nonpolar organic solvents,
like molecules must be broken and adhesive differs from that of hydrophilic excipient
forces between solute and solvent must be (s) which tend to dissolve in polar aqueous phase.
formed. The dissolution process of solids in Generally, hydrophilic excipient(s) would be
liquids involves three steps: (1) the removal of a incorporated into compositions containing poorly
molecule from the solute; (2) creation of a hole in water-soluble drug to improve its wettability. To
the solvent; and (3) insertion of the solute mole- make a feed solution accommodating both hydro-
cule into the solvent (i.e., solute–solvent interac- phobic API and hydrophilic excipient molecules
tion) (Hildebrand and Scott 1950). This in a fully dissolved state, it is critical to choose a
interaction between the solute and the solvent is proper solvent system. One approach is to mix
obviously dependent on the physical and solvents of different polarities to form a solvent
Table 11.1 The physical properties of the commonly used solvents for cryogenic technologies
Vapor
Boiling Melting pressure Viscosity
Molecular point point Density (mmHg) (mPas)
Solvent Formula weight ( C) ( C) (g/mL) Miscibility 25 C 20 C
Water H2O 18.02 100 0 0.997 Miscible 23.78 1
Acetone C3H6O 58.08 56.2 94.3 0.786 Miscible 240 0.33
Acetonitrile C2H3N 41.05 81.6 46 0.786 Miscible 73 0.34
t-Butyl alcohol C4H10O 74.12 82.2 25.5 0.786 Miscible 41.25 3.62
1,4-Dioxane C4H8O2 88.11 101.1 11.8 1.033 Miscible 30 1.54
1,3-Dioxolane C3H6O2 74.08 75.6 95 1.06 Miscible 70 0.66
Ethyl acetate C4H8O2 88.11 77 83.6 0.895 8.7 14 0.51
Methanol CH4O 32.04 64.6 98 0.791 Miscible 13.4 1.04
Tetrahydrofuran C4H8O 72.11 66 108.4 0.886 30 170 0.55
system of optimum polarity to dissolve the Fluid Dynamics

solutes of different polarities. This method is For SFL process, viscosity of the solvent is an
referred to as solvent blending or cosolvency. important factor that needs to be considered for
When talking about liquids, the term miscibility preparation of the feed solution to be sprayed
rather than solubility may be used to describe the through the nozzle into liquid cryogens. Addi-
affinity between the liquids (Conventional 2000). tionally, the melting point of the solvent(s) used
Liquids that form a homogenous system when in SFL process is better not higher than 0 C;
mixed in any proportion are said to be miscible otherwise, the feed solution may freeze prema-
(e.g., water and ethanol). Those in which only turely within the atomizing nozzle before sprayed
certain volume ratios produce homogenous into cryogens. Examples of case are tert-butanol
mixtures are said to be miscible in certain and 1,4-dioxane, in which many poorly water-
proportions (e.g., water and chloroform). Immis- soluble drugs show good solubility. However,
cible liquids will not produce a homogenous solu- due to their relatively high viscosity, 3.62 and
tion in any proportion (e.g., water and olive oil). 1.54 cP, respectively, compared to other organic
For the cosolvent used to make feed solution, the solvents, and high melting point for liquid, 25 and
solvents must, obviously, be miscible. For exam- 12 C, respectively (see Table 11.1.), they are not
ple, tetrahydrofuran/water co-solvent was used to proper candidates for SFL. Because the TFF tech-
form feed solutions for SFL process, because of nology applies the droplets directly onto the cryo-
the ability to dissolve both a poorly water-soluble genic substrate, premature freezing is not a
drug and hydrophilic excipients (Hu et al. 2002; concern and solvents with high melting point
Rogers et al. 2002b). Some poorly water-soluble may be used.
Nevertheless, TFF process also requires con-
drugs have relatively low solubility in the tetra-
sideration of the fluid dynamics for the solvent
hydrofuran/water co-solvent; later on, acetoni-
during spreading and freezing. The ability of an
trile, acetone, methanol, methylene chloride,
impinging droplet on the cryogenic substrate to
1,3-dioxolane, tert-butanol, and 1,4-dioxane
spread into a thin large-diameter disk prior to
were used to greatly increase the solubility of
freezing influences the cooling rate of the thin
drugs. However, the percentage of methylene
disk. Danazol/polyvinylpyrrolidone (PVP)
chloride in cosolvent should be less than 1% to
powders were produced under the same TFF pro-
be miscible. Worthy to note is that for the use of
cess except for using acetonitrile and tert-butanol,
low-melting-point solvents, a liquid nitrogen cold
respectively, as the solvent. It was observed that
trap is necessary to capture the solvent vapor
the powder from acetonitrile solution exhibited
before sucked into vacuum pump during
more uniform nanostructured surface morphol-
lyophilization.
ogy compared to the powder from tert-butanol
solution. The morphological difference was employed in studies of TFF process, which has
attributed to the cooling rate of the two solvents less concern of premature freezing compared to
(Overhoff et al. 2007a, 2009). Moreover, solvents SFL. Tert-butanol was often used in combination
of the feed solution with greater thermal with low-melting-point solvent system for its
diffusivities are more desirable for rapid heat easy-to-freeze and good lyophilization
transfer. The rapid heat transfer in TFF process characteristics. As a result of its high vapor pres-
is the result of intimate contact between the solu- sure (Table 11.1.) and its crystal morphology, the
tion and cryogenic substrate. sublimation rate of tert-butanol is greater than 2.5
times that of water. 1,4-dioxane/water co-solvent
Ease of Lyophilization was extensively used to make feed solutions for
Freeze drying, i.e., lyophilization, is the com- TFF process (Overhoff et al. 2007a, b; DiNunzio
monly used means to obtain dry powder from et al. 2008; Yang et al. 2008a, b), mainly due to
the cryogenically generated materials by remov- its relatively high freezing point and vapor pres-
ing the solvent(s). The ideal solvent for freeze sure, which make the freezing and lyophilization
drying has the following properties: a high processes easily manageable. These solvents
vapor pressure, a melting point either below or prove beneficial by reducing the lyophilization
slightly above room temperature, a high viscosity, time (Ni et al. 2001) or eliminating the solvent-
and a low toxicity. It must provide a stable envi- removal process altogether as some of these
ronment for freeze drying and be rapidly and solvents sublime at ambient conditions or higher
completely removed to produce dry material (Tesconi et al. 1999). Moreover, they have low
(Ni et al. 2001). biological toxicity and basically no harmful
A variety of different types of organic solvents impact to the environment.
were used in cryogenic processes, as mentioned According to these criteria, a solvent with
in Table 11.1. Most of these water-miscible these ideal properties may not exist for cryogenic
organic solvents have freezing points below processes.
75 C. The frozen disk made by the low- Formation of co-solvent systems can mitigate
melting-point solvents tends to melt during limitations of certain solvents (solubility) while
lyophilization and it makes a very challenging maintaining some of the ideal fluid dynamics and
task for the lyophilizer condenser to catch the lyophilization characteristics. Combinations of
sublimed vapor of these solvents. A cold trap 1,3-dioxolane (solubility enhancer) and tert-
between lyophilizer sample chamber and vacuum butanol (ideal freezing and lyophilization
pump is necessitated to prevent the solvent vapor properties), tetrahydrofuran and water (both are
from being sucked into the vacuum pump. solubility enhancers), acetonitrile (solubility
Accordingly, organic solvents with higher enhancer and freezing and lyophilization
melting point are of great interest for selecting properties) and water, and 1,4-dioxane (solubility
proper solvent for cryogenic process. Acetoni- enhancer and freezing and lyophilization
trile, having a melting point of 45 C, viscosity properties) and water have been used to generate
of 0.34 cP, good heat transfer, and the unique nanostructured poorly water-soluble drug
ability to dissolve both a hydrophobic drug and compositions using cryogenic technologies.
hydrophilic excipient(s), was widely employed
for many compositions containing poorly water-
soluble drug in both SFL and TFF processes to 11.3.2 Selection of Excipients,
increase drug loading with reduced risk of liquid– Concentrations for Drug
liquid phase separation. More importantly, the Formulations Using Cryogenic
high vapor pressure of 73 mmHg at 25 C eases Processes
the removal of the solvent by lyophilization.
Subsequently, high-melting-point solvents The ultimate goal of cryogenic particle engineer-
1,4-dioxane and tert-butanol were often ing technologies is to improve the dissolution
properties of poorly water-soluble drug by polyvinyl pyrrolidone-covinyl acetate

making an amorphous nanoparticulate form of (PVP/VA), hydroxypropylmethylcellulose ace-
the drug. However, a primary concern of the tate succinate (HPMCAS), high-molecular-
formed amorphous material is the inherent insta- weight polyethylene glycol (PEG), polyvinyl
bility, due to the higher energy state. Amorphous alcohol (PVA), polylactic acids (PLA) and
material may automatically convert back to a polylactic-co-glycolic acid (PLGA), etc. Catalytic
low-energy crystalline state. Recrystallization of pretreated softwood cellulose (CPSC) isolated
the amorphous drug may be avoided by inclusion from pine wood (Pinus sylvestris) was recently
of stabilizing hydrophilic excipients in the com- used as a stabilizing carrier polymer of the cryo-
position. Principally, stabilizing excipients com- genic co-ground spray-dried formulation of
bat recrystallization of amorphous drug particles poorly water-soluble drug (Penkina et al. 2015).
by steric hindrance and/or the formation of hydro- Besides polymers, nonpolymeric natural
gen bonds with drug molecules (Khougaz and products (e.g., lecithin) and small-molecule
Clas 2000; Vasanthavada et al. 2005). These hydrophilic excipients (e.g., sugars) can also be
stabilizing excipients must be hydrophilic for employed to enhance wettability and solubility of
improving the wetting properties of drug poorly water-soluble drugs, with better safety
particles; however, the hydrophilic excipients profiles. Sugars such as lactose, sucrose, treha-
should not be hygroscopic. Otherwise, they lose, and, recently, fructose oligomer inulin were
absorb moisture easily when exposed to ambient frequently used as an excipient to stabilize amor-
environment, which can lead to morphological phous drugs, peptides, and proteins during drying
instability of the excipient-stabilized amorphous and subsequent storage (Davies and Feddah
drug particles by either displacing drug molecules 2003; Van Drooge et al. 2004). The addition of
from hydrogen-bonding sites or plasticizing poly- sugars has been shown to extend the shelf life of
meric stabilizers (Forster et al. 2001; amorphous systems by preventing crystallization
Vasanthavada et al. 2004). (Eriksson et al. 2003).
As morphological instability and poor wetta- On the contrary, drugs with low Tg were
bility are the substantial limitations to particle important for freeze-drying of poorly soluble
engineering of poorly water-soluble drugs, drugs. To fabricate nanocrystals, fenofibrate, a
concessions need to be made in selecting appro- low Tg, poorly water-soluble drug, was mixed
priate stabilizing excipient(s), with respect to with mannitol. Excipient such as mannitol was
rigidity for steric hindrance or hydrophilicity. selected for this process since it was susceptible
Many readily soluble and/or wettable excipients to crystallize. The produced nanocrystalline
do not provide adequate steric hindrance of particles of poorly water-soluble drug also drasti-
recrystallization due to low melting points or cally enhanced dissolution of the poorly water-
glass transition temperature (Tg). On the contrary, soluble drugs (De Waard et al. 2008).
excipients with high melting points or Tg’ s typi- Worthy to note is that excipient to drug ratio is
cally have relatively poor wetting properties. also an important consideration in formulation
Therefore, polymers with high glass transition design. First, excipient to drug ratio may affect
temperature (Tg) and hydrophilicity, such as the morphology of the engineered drug composi-
PVP and hydroxypropylmethylcellulose tion. The Tg of the overall drug composition is a
(HPMC), are popular candidates to formulate function of the fraction of each component. If
poorly water-soluble drugs for improved aqueous there is more amorphous material present in the
dissolution. To expand the choices of polymers, drug composition, more stabilizing polymer
combinations of rigid and malleable excipients would be needed to be present; otherwise, it
such as surfactants (e.g., poloxamer, sodium may lead to a greater risk of recrystallization.
dodecyl sulfate, and Tween 80) often can be Second, excipient to drug ratio affects hydrophi-
used to achieve reasonable stability and wettabil- licity of the engineered drug compositions. It has
ity. Other commonly used polymers include been demonstrated that as the potency of a poorly
water-soluble drug composition increased, the from Fig. 11.8a and b, while decreasing the solid
wettability is often decreased, particularly for loading to 0.2% formed discrete nanoparticles, as
the matrix system of a drug composition where shown in Fig. 11.8c and d. At 0.2% solid loading,
drug and excipient molecules distribute ran- low polymer loading showed amorphous string-
domly. Hydrophilic excipients act at the surface like structures (Fig. 11.8c), whereas increasing
of drug composition particles to reduce the solid– the polymer ratio resulted in more spherical
liquid interfacial tension between the hydropho- nanoparticles (Fig. 11.8d), indicating that the sur-
bic drug and aqueous media (Sinswat et al. 2005). face morphologies are strongly influenced by the
As potency is increased, a greater proportion of percentage of polymer in the composition. By
the particle surface area is occupied by drug manipulating the solid loading and polymer por-
molecules; hence, the surface is rendered more tion in the feed solution for cryogenic processes,
hydrophobic, i.e., less wettable. Thus, it is often drug compositions with desired characteristics
difficult to achieve high-potency drug particles may be tailored. For example, low-density
with acceptable wetting properties. However, by powders of drug compositions intended for pul-
carefully selecting excipient(s) and controlling monary delivery can be produced from a feed
the proportions of drug and excipient(s) in the solution with drug loading commonly in the
cryogenically processed drug compositions, it range of 0.5–1% (w/v). To make drug composi-
was also possible to achieve high-potency drug tion powders to be further processed into tablets
compositions with improved wettable surface and or capsule dosage forms, the feed solution may be
high stability against recrystallization. An made to above 5% (w/v) if solubility permitted,
SFL-processed danazol/PVP K-15 composition so that much denser powders can be obtained to
with high potency of 91% was reportedly press into tablets or fill into capsules.
maintained in its amorphous structure and rapid
dissolution characteristics after one month of
cycled stability conditions (5 to 40 C every 11.3.3 Properties of Pharmaceutical
3 h) (Hu et al. 2004a, b, c). Powders Made by Cryogenic
In addition to the considerations for Processes
characteristics and stability of drug composition,
formulation design also needs to comply with Engineered High Surface Area Powder
regulatory requirements for final dosage forms. for Oral Drug Delivery
It is important to judiciously select excipients Based on the extensive studies of the cryogenic
that have been accepted/approved by FDA for particle engineering technologies, drug
specific administration routes. Especially for par- compositions processed by these technologies
enteral and pulmonary deliveries, only very lim- have been shown to create nanostructured amor-
ited excipients are approved for use, mainly phous particles with dramatically enlarged sur-
including biocompatible and biodegradable face area and improved dissolution properties, in
materials. contrast to crystalline bulk drug. Additionally,
Moreover, the total solid loading (drug plus these processes allow molecular incorporation of
excipients), particularly the excipient to drug ratio hydrophilic pharmaceutical excipient(s) to drug
in the feed solution for cryogenic processes, compositions, leading to improvement of the wet-
affects morphology and surface area of the cryo- ting characteristics of the bulk poorly water-
genically engineered drug compositions. Typi- soluble drug powder (Hu et al. 2002, 2004a, b;
cally, a higher solid loading infers denser Rogers et al. 2003a, b; Overhoff et al. 2007a;
engineered drug composition powder, accord- Purvis et al. 2007; Yang et al. 2008b). The cryo-
ingly, the lower the surface areas of the powder. genic technologies are primarily precipitation,
Itraconazole and hydroxypropylmethyl cellulose i.e., “bottom-up,” processes, allowing for a reduc-
phthalate (HP-55) at 2.0% solid loading tion in the particle size of drug particles without
demonstrated a porous matrix structure as seen degradation that is induced by heating or
Fig. 11.8 Scanning electron micrographs of TFF-processed powders containing ITZ and hydroxypropylmethyl cellu-
lose phthalate (HP-55) at various solid loading and drug polymer ratios: (a) 4:1HP55 (2%), (b) 1:4HP55 (2%), (c) 4:
1HP55 (0.2%), (d) 1:4HP55 (0.2%). (Reproduced from Overhoff et al. (2007b) with permission from Elsevier)
mechanical process. As discussed in Sect. 12.2.5, medium, C is the concentration of drug in the
amorphous morphology and uniform small parti- medium at time t, and h is the thickness of the
cle size result from the very rapid freezing rates of diffusion boundary layer. With dramatically
these processes. Both the spray of drug feed increased surface area and decreased particle
solutions to produce tiny droplets and impinging size, the dissolution rate can be increased. Fur-
of the feed solution drop on a solid substrate to thermore, the formation of metastable amorphous
form a very thin spread layer of liquid to enable form yields higher energy states for the drug and
dramatic enlargement of surface area, which, in thus a greater thermodynamic driving force for
turn, enhances the rapid freezing process. dissolution. Therefore, both dissolution rate and
According to the Noise–Whitney equation the extent of dissolution are improved, i.e.,
achieves supersaturation of the drug.
dM=dt ¼ ADðC s C Þ=h,
The very rapid freezing of the cryogenic pro-
where dM/dt is the rate of dissolution, A is the cesses enables production of amorphous
surface area available for dissolution, D is the nanostructured pharmaceutical powders with rel-
diffusion coefficient of the compound, Cs is the atively small amounts of excipient(s) to achieve
solubility of the compound in the dissolution high drug loadings of commonly 50–86% drug/
total solids, while maintaining high dissolution A novel way recently patented is utilization of
rates. hot-melt extrusion process to prepare extrudates
Besides the excellent performance of the cryo- employing relatively low-melting/softening-tem-
genically engineered pharmaceutical powders in perature polymer or a blend of polymers as carrier
pulmonary delivery, rapid-release tablet formula- and cryogenically engineered pharmaceutical
tion containing SFL-processed danazol powders (Miller et al. 2008). It is critical to con-
micronized compositions by direct compression duct the extrusion at a temperature approximating
was also reported (Hu et al. 2004a, b, c). The high or above the softening or melting temperature of
surface area, high porosity, and amorphous struc- the carrier polymer and below the point of solubi-
ture of the SFL-processed danazol compositions lization of drug-containing particles in the carrier
contribute to their significantly enhanced dissolu- system, and below the recrystallization point of
tion profiles. However, these properties may be amorphous drug in the drug-containing particles.
changed during the process of incorporation of Therefore, high Tg excipient(s) need to be
other tablet excipients and tableting compression. employed in the preparation of cryogenically
By selection of suitable excipients and direct engineered pharmaceutical powders. The
compression without involving water in the obtained extrudate containing the cryogenically
tableting, it was found that over 90% of danazol engineered pharmaceutical powder particles can
released in only 10 min from tablets containing be milled for further tableting or capsule filling
SFL-processed danazol composition (danazol/ for oral administration.
PVP K-15/SLS ¼ 4:1:1), similar to the dissolu- Compared to the conventional dosage forms,
tion profiles of the SFL composition itself. The the lower amount of excipients and drug in cryo-
amorphous morphology was also maintained. genically processed drug compositions to be
Utilizing high Tg excipient in the SFL-processed delivered to patients can provide increased patient
composition contributed to prevent recrystalliza- compliance, safety, and therapeutic efficacy of the
tion of the SFL-processed danazol during the poorly water-soluble drugs (Yang et al. 2008b).
tableting process.
An alternative means is to fill the cryogeni- Brittle Matrix Powder for Pulmonary Drug
cally engineered pharmaceutical powders into Delivery
capsules for oral delivery. A solid solution of In addition, cryogenically engineered pharmaceu-
itraconazole (ITZ) and enteric polymer cellulose tical powders with the properties described above
acetate phthalate (CAP) (ITZ:CAP ¼ 1:2, w/w) also offer the versatility of being further assem-
was created by TFF process (DiNunzio, Miller bled into diverse dosage forms for application,
et al. 2008). In vitro supersaturated dissolution such as oral, parenteral, and pulmonary delivery.
results demonstrated significantly lower levels of The most extensively explored application is pul-
supersaturation in acidic media and greater monary delivery, including application of the
extents of supersaturation in neutral media com- engineered pharmaceutical powders directly in
pared to Sporanox pellets, the marketed product powder form to various dry powder inhaler
of itraconazole. Both the ITZ:CAP ¼ 1:2 and (DPI) devices, making aqueous dispersion of the
Sporanox pellets were filled into size-9 capsules pharmaceutical powder for nebulizers, and dis-
and dosed to rats. The pharmacokinetics study persing the pharmaceutical powders in
indicated that ITZ:CAP ¼ 1:2 achieved 2 times propellants such as hydrofluoroalkanes (HFA)
higher oral bioavailability and more rapid onset of for pressurized metered dose inhalers (pMDI).
action versus Sporanox® pellets. The amorphous Pulmonary drug delivery is of great interest
nature of TFF-engineered ITZ:CAP ¼ 1:2, the due to its targeted delivery of drugs to the patho-
intestinal targeting, and the increased durations logical sites. However, for inhalation delivery,
of supersaturation rendered by enteric polymers particle-size distribution and morphology of the
were needed to contribute to the improved drug-containing particles have pronounced
bioavailability. effects on drug deposition in the respiratory tract
and therapeutic effects. There are some and FPF of 83.3% were achieved. Pharmacoki-
drawbacks of nebulization such as reconstitution, netic results revealed that the brittle powder
long dosing times, and inefficient targeting of the exhibited higher pulmonary bioavailability but
lung. Moreover, drug loss during patient exhala- lower systemic concentration than the crystalline
tion of more than 50% resulted in a low efficient dry powder formulation. Moreover, perhaps due
therapeutic of drug delivery via vibrating mesh to the decreased clearance, the TFF powder
nebulizer and led to the development of dry pow- retained in the lung longer than the other.
der inhalation formulation (Watts et al. 2013). Inhaled rapamycin was also prepared as amor-
TFF can produce low-density (0.01–0.03 g/ phous brittle matrices using TFF process. The
mL), high-surface-area amorphous brittle matri- optimized TFF formulation with lactose
ces composed of sub-500-nm primary structures. monohydrate was selected to compare to the
The tremendously low-density (<0.01 g/cm3) milled microparticles of the physical mixture. In
microparticles provide some benefits for pulmo- terms of aerosolization performance, both
nary drug delivery since they can be easily formulations were appropriate for pulmonary
sheared apart into smaller microparticles by high delivery. The in vitro aerodynamic properties of
shear velocity in a passive commercial DPI the brittle matrix powder were superior to the
device (Watts et al. 2013). Watts et al. reported micronized crystalline powder. However, the
that TFF-engineered dry powders containing in vivo study of the single-dose administration
tacrolimus and nonpolymer excipient (e.g., showed the lower lung deposition, higher sys-
sugars) exhibited optimal aerosol performance temic concentration of the fine low-density
for deep lung delivery by filling the powders particles as compared with the crystalline powder.
into size-3 capsules and used in Handihaler® These resulted from the enhancement of drug
and Easyhaler® DPI devices. Shear stress from solubility in the amorphous formulation and the
the Handihaler® DPI device affected the diameter FPF at a more distal part of the lungs (Carvalho
of the respirable brittle powders. Although lactose et al. 2014).
and raffinose exhibited better aerosol perfor- Similarly, beclomethasone dipropionate dry
mance than mannitol formulations, both powders that were engineered by TFF process
saccharides were hygroscopic material. Hence, demonstrated proper aerodynamic profiles for
the formulations required a moisture barrier pack- inhalation when applied to Handihaler® device,
age (Watts et al. 2013). suggesting that TFF-generated low-density,
In addition to Watts’ work on tacrolimus, a nanostructured micron-sized pharmaceutical
high-potency tacrolimus dry powder for inhala- powders are suitable for application in DPI.
tion was prepared by TFF (Sahakijpijarn et al. Recently, remdesivir powder formulations for
2020a). The dry powder containing 95% (w/w) inhalation were developed by TFF. These
of tacrolimus presented high SSA (74 m2/g) as a remdesivir brittle matrix powders made with
nanostructured brittle matrix powder. When either Captisol®, mannitol, lactose, or leucine
aerosolized by high resistant RS01 monodose performed highly aerosolizable, with MMAD of
dry powder inhaler, it resulted the mass median up to 0.7 μm and over 90% FPF (Sahakijpijarn
aerodynamic diameter (MMAD) of 2.6 μm, and a et al. 2020b). The in vitro pharmacokinetic study
fine particle fraction (FPF) of 69%. demonstrated that pulmonary administration can
The in vitro and in vivo performance of the deliver remdesivir to the lung and subsequently
TFF brittle powder and crystalline micronized can be converted to GS-441524, a nucleoside
formulations of tacrolimus were reported by analogue, both in the lungs and plasma of
Wang et al. (2014). Miat® monodose inhaler hamsters. The plasma level of remdesivir and
was used to deliver the low-density particles of GS-441524 was higher than EC50 over 20 h,
tacrolimus to the deep lung. MMAD of 2.26 μm
which was sufficiently high to provide the of poorly water-soluble drugs by pulmonary
antiviral activity (Sahakijpijarn et al. 2021). delivery of nanostructured aggregates was
Niclosamide was also repurposed as an reported in a murine model. An amorphous
antiviral therapy for the treatment of coronavirus nanostructured composition containing
disease (COVID-19). Niclosamide dry powder itraconazole:mannitol:lecithin (1:0.5:0.2, w/w/w)
formulation containing mannitol and leucine made by TFF process illustrated dramatically
prepared by TFF exhibited high aerosol perfor- enhanced supersaturation dissolution profile and
mance (76.6 1.4% FPF of recovered dose, aerodynamic properties suitable for deep lung
1.11 0.07 μm MMAD). The histopathology delivery in vitro. Inhalation of the nebulized col-
analysis demonstrated that dry powder adminis- loidal dispersion of this composition by mice for
tration of niclosamide was safe after an acute 20 min produced significantly high drug deposi-
three day and multi-dose administration in rats. tion in lung and effective therapeutic concentra-
Additionally, the vivo pharmacokinetic study in tion in blood. The observed dramatic
hamsters showed that dry powder administration improvement in bioavailability of the
by inhalation can deliver niclosamide to the lungs TFF-processed itraconazole was mainly
and maintained the drug level above the required attributed to the amorphous nature, smaller
IC50 and IC90 levels for at least 24 h after a nanoscaled particle size, and the presence of leci-
single administration in a Syrian hamster model thin, which is biocompatible to lung and acts as a
(Jara et al. 2021). permeation enhancer to facilitate ITZ to permeate
In addition to single drug formulations, TFF through the lung epithelium (Yang et al. 2008b).
was also applied to prepare dry powder Another promising application of cryogeni-
formulations of a fixed-dose combination of cally engineered pharmaceutical powders is their
salmeterol xinafoate and mometasone furoate incorporation into pMDIs. Suspension stability in
was prepared as a brittle matrix powder. The pMDIs is a significant challenge because
composite particles were studied using a surfactants that have been approved by FDA for
marketed DPI device for inhaled drug delivery. inhalation are insoluble in propellants of pMDI.
The in vitro aerodynamic testing indicated that Commercial pMDI products are therefore per-
brittle matrix powder enhanced aerodynamic plexed by poor suspension stability with settling
properties as compared to the micronized powder observed within minutes. Consequently, it
blend. Both drugs co-deposited when they were requires patients to shake the pMDI device right
together in the composite particles. On the other before inhalation. A novel rod-shaped nanoparti-
hand, the deposition and dose uniformity of the cle (nanorods) of bovine serum albumin that was
blend of two micronized drugs were not consis- used as a model protein drug, created by TFF
tent. Regarding the respirable dose delivered and process, was found capable of forming stable
stability results, the formulation without suspensions in HFA against settling for one year
stabilizing materials was chosen for the in vivo and produced optimal aerodynamic properties
study. Pharmacokinetic study showed that the (Engstrom et al. 2009). The excellent suspension
lung concentration of both drugs from the brittle stability of the protein particles in HFA was
matrix formulation was greater than the crystal- attributed to their anisotropic geometry. Spherical
line drugs blend. The systemic drug levels of the particles pack together in a more efficient manner
TFF-processed powder were also higher than the than anisotropic particles, such as rods [86].
physical mixture due to the rapid absorption and Therefore, open flocs composed of nanorods
the higher lung tissue exposure (Liu et al. 2014). take up a larger volume and have lower density,
Pulmonary drug delivery targeted to the compared to flocs containing spherical particles.
alveoli for systemic absorption has become an TFF process is applicable to a wide variety of
increasingly attractive route of administration drugs to make nanorods for pMDI without the
for poorly water-soluble drugs (Courrier et al. need of stabilizing the primary particles.
2002). The concept of improving bioavailability
11.3.4 Analytical Methods Used PXRD is one of the most widely used quanti-
to Characterize Pharmaceutical fication techniques because of its simplicity and it
Powders Made by Cryogenic measures differences in periodicities of atoms/
Processes molecules in a powder sample (Stephenson et al.
2001). It provides important insight, based on the
Solid-State Characterization degree of long-range order present, into the extent
The solid state of the drug and excipient are and nature of the crystallinity and microstructure.
important aspects of the physical and chemical PXRD patterns of crystalline forms show strong
stability, as well as pharmaceutical and therapeu- diffraction peaks, whereas amorphous states
tic performance of the drug product. exhibit diffuse and halo diffraction patterns.
Amorphous state, a disordered phase, having X-ray procedures for the estimation of degree
similar mechanical and physical properties of a of crystallinity are based upon the measurement
supercooled liquid existing at temperatures below of X-ray scattering from the entire sample includ-
its thermodynamic crystallization temperature but ing the crystalline and amorphous region of the
has not been given sufficient time to anneal and sample. The experimentally measured crystalline
crystallize to its thermodynamically stable and amorphous intensities are proportional to the
ordered phase, inherently has a higher degree of crystalline and amorphous fraction of the sample.
molecular mobility (Hancock et al. 1995). Even a Quantification of amorphous material by PXRD
small amount of crystalline form of the drug can can be achieved by three methods: (1) measuring
significantly affect the in vivo performance of the the characteristic crystalline peak intensities,
amorphous drug (Hancock and Parks 2000). (2) measuring the integrated peak areas of the
Therefore, it is important to monitor and charac- principal crystalline peaks, and (3) measuring
terize the extent of crystallinity or disorder during the intensity of characteristic region of amor-
formulation development, manufacturing, and phous scattering; of physical mixtures of known
over the intended shelf life of pharmaceutical crystallinity to yield a calibration curve which is
product to ensure a robust and safe formulation used for further quantification studies (Shah et al.
by understanding the behavior of these amor- 2006).
phous systems. Limit of detection (LOD) of crystallinity in
Various analytical techniques have been amorphous drug compositions of X-ray diffrac-
reported for quantifying amorphous or crystalline tion is 5–10%, and the limit of quantitation
phase in solids. The classical methods of (LOQ) is 2–5% (Nagapudi and Jona 2008). The
evaluating the solid state are powder X-ray dif- specificity and accurate quantitative nature of this
fractometry (Newman and Byrn 2003; Shah et al. nondestructive technique make it the first line
2006) and thermal analyses (Clas et al. 1999). choice for studying crystallinity of pharmaceuti-
cal materials.
Wide-angle X-ray scattering (WAXS) is an
X-Ray Diffractometry (XRD)
X-ray diffraction technique that provides the
Diffraction techniques are perhaps the most defin-
crystallographic structure, atomic positions, and
itive method of detecting and quantifying molec-
sizes in a unit cell, as well as chemical compo-
ular order in any system. Conventional, wide-
sition. WAXS is similar to small-angle X-ray
angle and small-angle diffraction techniques
scattering (SAXS), but it has a shorter distance
have all been used to study order in systems of
from the sample to the detector, which can
pharmaceutical relevance (Salekigerhardt et al.
observe the diffraction maxima at larger angles.
1994). Diffraction is defined as a scattering phe-
WAXS refers to the analysis of Bragg peaks
nomenon in which the incident X-rays, depending
scattered to wide-angle (Ѳ >10 ), which provide
upon the phase difference, are reinforced to form
the ability to measure subnanometer-sized struc-
diffracted beams (Suryanarayan 1995).
ture (Lamba 2016). WAXS has been applied to
determine the crystalline structure and of drug
substances and excipients. Additionally, WAXS absence of moisture (Clas et al. 1999). By using
is used to quantify the crystallinity in unknown scanning calorimetry and thermo-mechanical
samples using a calibration curve of the physical methods, it was found that in order for the average
mixture of pure amorphous and crystalline relaxation time constants to significantly exceed
compounds at known ratios (Dong and Boyd the projected shelf life for a pharmaceutical prod-
2011). The detection limit of crystallinity is uct (approximately three years), it was generally
less than 1% (Wang et al. 2000). necessary to store the amorphous materials as
much as 50 C below the glass transition region
Thermal Analysis (Hancock et al. 1995; Hancock 2002).
Analyses based on thermal energy principle have The LOD and LOQ of crystallinity often tend
been widely employed to characterize amorphous to be better using DSC, with levels of 1–5% and
pharmaceutical systems. less than 1%, respectively (Nagapudi and Jona
Crystallization from the amorphous state can 2008). DSC also requires much less material
be induced by the thermo-analytical techniques to than PXRD.
produce an exothermic change whose magnitude
is then quantitatively related to the extent of crys- Spectroscopy
tallization occurring. This can then be used to Spectroscopic methods such as Raman, infrared
determine the crystallinity of a partially amor- (IR), and near-infrared (NIR) have also been
phous sample provided total crystallization of reported for crystallinity quantitation (Head and
this sample is known to occur (Salekigerhardt Rydzak 2003; Brown et al. 2007). They provide
et al. 1994). Differential scanning calorimetry chemically resolved information with small
(DSC) in both conventional and modulated amount of material requirements. However, data
modes has been used to quantify the extent of interpretation from spectroscopic methods can
crystallinity. run into problems because of their inability to
Amorphous materials can be characterized by unambiguously separate peaks from different
their glass transition temperature (Tg). By DSC, phases in the sample.
Tg is characterized by a change in heat capacity, Solid-state nuclear magnetic resonance
which is seen as a change in the baseline. The Tg (ssNMR) is another powerful technique which
may be important in determining the relative has been recently applied for the characterization
chemical and physical stability of formulations of pharmaceutical solids. ssNMR is a
containing amorphous drugs (Yoshioka et al. non-destructive and non-invasive analysis that
1994). requires relatively large amounts of material
From a pharmaceutical perspective, it was (60–200 mg), compared to XRD and DSC
thought that below the Tg the molecular mobility (Nagapudi and Jona 2008). ssNMR has been
was very low, and long-term product stability can used to quantify the amount of crystalline and
be achieved by storing amorphous amorphous content in a sample. The primary
pharmaceuticals at sub-Tg temperatures. In the advantage of ssNMR lies in its selectivity and
amorphous form of a hypoglycemic agent for ability to probe a variety of nuclei. Among the
diabetes mellitus, with a glass transition tempera- above-mentioned methods, ssNMR is the only
ture of 71 C, no recrystallization was found after technique that does not require a pure reference
a 4-month storage at room temperature in the standard for phase quantitation (Offerdahl et al.
absence of moisture. However, crystallization 2005). LOQ downs to 0.25% can be achieved
occurred after storage at 50 C for two months. when 1H, 31P, or 19F nuclei are used for quantita-
The extent of recrystallization increased with tion, while LOQ of about 3% can be achieved
increasing storage temperature. Some amorphous when using 13C. Additionally, ssNMR has been
drugs with high Tg can remain stable for extended applied to detect intra- and inter-molecular
times. For example, an API with a Tg of 125 C interactions between drug and excipient (Nie
does not crystallize from the solid state in the et al. 2016). The spin-lattice relaxation time (T1)
and the rotating frame-spin lattice relaxation time Atomic Force Microscopy (AFM)
(T1ρ) were also used to evaluate the phase separa- AFM, a powerful surface and nanoimaging ana-
tion and miscibility between drug and excipient in lytical technique, offers a unique opportunity to
amorphous solid dispersions (Yuan et al. 2014; Li examine surface structure of a variety of materials
et al. 2021). with mesoscopic-scale resolution (10–6–10–9 m)
and quantify the individual particle and excipient
Surface Analyses interaction by direct force measurement in a vari-
ety of environmental conditions (Jalili and
Scanning Electron Microscopy (SEM) Laxminarayana 2004). Especially for inhalation
SEM is recognized as a unique tool in the visual drug delivery, AFM is very useful to provide
examination of solid-state drug compositions and tailored investigations of particle–particle
their surfaces. The resolution is of the order of interactions within DPIs, particle–DPI wall
nanometers (magnifications in the range interactions, and also to perform in vivo
20–100,000x). A fine beam of electrons of simulations of inhaled particle–pulmonary surfac-
medium energy (5–50 keV) scans a gold–palla- tant interactions (Sindel and Zimmermann 2001).
dium-coated sample producing secondary AFM can also be used for phase imaging.
electrons, backscattered electrons, light or Surface amorphous domains of sorbitol were
cathodoluminescence, and X-rays. The latter reported to be identified and mapped by using
allow for X-ray microanalysis for specific both AFM and Raman microscopy (Ward et al.
elements. SEM is routinely used for imaging 2005). Also, AFM and Raman microscopy were
particles in the micron and smaller size range utilized for screening of miscibility of drug-
and for examining the surfaces of larger particles. excipient and stability of solid dispersions. AFM
The resolution allows identification of specific assay technique is able to achieve the characteri-
surface geometric features that are indicative of zation of the miscibility and stability of molecu-
structural phenomena. larly disperse mixtures in just hours or a couple
days instead of weeks or months, which is bene-
Energy Dispersive X-Ray Spectroscopy (EDX) ficial for product development (Lauer et al. 2011).
EDX is a non-destructive technique that is used Weiss et al. (2015) recently reviewed the use
for the elemental analysis or chemical characteri of AFM as a powerful tool to characterize DPI
zation of a sample. EDX can provide an overall formulations. This review included theories of
mapping for the analysis of the drug and excipient adhesion and cohesion forces, summary of the
distribution on the surface of powder via SEM AFM applications for drug particle and formula-
images (Scoutaris et al. 2014). SEM/EDX can tion characterization, and particularly emphasized
detect the element with a detection limit of its use as a colloidal probe. Assessment of
100–3000 ppm and depth resolution of interparticulate forces was also discussed.
~0.5–3 μm. Recently, Moon et al. used
SEM/EDX to determine the chemical
Time-of-Flight Secondary-Ion Mass
compositions of two differently observed particle
Spectrometry (ToF-SIMS)
morphologies that were observed on the surface
Tof-SIMS is a highly sensitive surface analytical
of TFF voriconazole/mannitol (50/50). Due to
technique that has increasingly been applied for
micron-scale depth resolution, the spot analysis
the surface analysis of pharmaceutical powders
was applied to avoid the overlapping between the
due to a detection limit (the partis-per-billon
signals of two molecules. Therefore, SEM-EDX
range in some cases), high spatial resolution
was able to determine the location of
(~200 nm), and high surface sensitivity (Prestidge
voriconazole on a micron-sized particle by
et al. 2007). In a ToF-SIMS experiment, a pulsed
detecting fluorine, oxygen, and nitrogen, as well
primary ion beam is accelerated and focused on
as the location of mannitol on porous matrix by
the sample surface under ultra-high vacuum,
detecting only oxygen atoms (Moon et al. 2019).
resulting in a bombarding particle on the surface. pharmaceutical powders. The principle, based on
Then a sputter plume containing positive or neg- the Washburn model, consists of registering the
ative secondary particles such as atoms, whole volume of pores penetrated at each intrusion pres-
molecules, and fragments of molecules are sure, which can be easily transformed into pore
ejected from the surface. The charge on the sec- size via the Washburn equation (Washburn 1921)
ondary ions is separated based on their mass-to- to give a complete pore-size distribution (Carli
charge ratio in a time-of-flight region. Since dif- and Motta 1984).
ferent mass fragments have a different kinetic Both of these techniques can provide reliable
energy and velocity, the secondary ions’ time to information about pore-size/volume distribution,
reach the detector can be referred to the mass-to- particle-size distribution, and specific surface area
charge ratio and mass spectrum for every detector for porous solids regardless of their nature and
pixel. ToF-SIMS can determine the distribution shape.
of an active compound within a matrix by the
extremely short analysis depth (~1 nm) (Barnes Contact Angle
et al. 2011). Contact angle is a quantitative measure of the
wetting of a solid by a liquid. It is defined geo-
Specific Surface Area Measurement metrically as the angle formed by a liquid at the
The surface area of a solid material is the total three-phase boundary where a liquid, gas, and
surface of the sample that is in contact with the solid intersect. Pharmaceutically, wetting is not
external environment. It is expressed as square an end in itself but is the preliminary step in
meters per gram of dry sample. Poorly water- another process, e.g., dispersion or dissolution,
soluble drugs are often rendered more available both in vitro and in vivo. The improvement in
for absorption by reducing the particle size, i.e., wettability of a hydrophilic character probably
increasing the surface area. Surface area is was responsible for the increased dissolution
strongly related to the particle sizes, pore size, rate of hydrophobic drugs (Lerk et al. 1976).
and the pore volume. The smaller the particle Contact angle is therefore often determined as a
size and pore size, the higher the surface. The measure of the surface energetics of drug
larger is the pore volume, the larger is the surface. substances.
The surface area results from the contribution of Optical tensiometry (goniometry) is a com-
the internal surface of the pores plus the external monly used method to measure contact angles of
surface of the pharmaceutical powders. The phar- drug substances, which are compressed to form
maceutical powders generated by cryogenic smooth and flat-faced tablets. Analysis of the
technologies are generally fluffy and porous shape of a sessile drop of test liquid placed on
from visual and microscopic observations. For the flat-faced tablets is the basis for optical tensi-
such porous systems, the contribution of the ometry. Contact angle can be assessed directly by
external surface to the total is very limited. measuring the angle formed between the solid
Physical and chemical gas adsorption and mer- and the tangent to the drop surface, as shown in
cury intrusion porosimetry are the most widely Fig. 11.9.
used techniques to characterize powders and solid
materials. Gas adsorption porosimetry typically Other Techniques
can be performed using the Brunauer–Emmett–
Teller (BET) theory (Brunauer et al. 1938) which Particle Size
allows for multilayer adsorption. With nitrogen Particle size is one of the physicochemical
gas adsorption, depending on the equipment used, properties influencing the performance of drug
pore diameter in the range of 0.3–300 nm, i.e., product and its manufacturing processability and
mesopores and macropores, can be determined. quality attributes (http://www.ich.org/). The
Mercury intrusion porosimetry is another com- influence includes dissolution rate, drug release
monly used method to measure surface area of profiles, and bioavailability; in vivo particle
θ = 90° θ > 90° θ < 90°
Fig. 11.9 Contact angles formed by a liquid at the liquid, gas, and solid three-phase intersections. Contact angle can be
assessed directly by measuring the angle formed between the solid plane and the tangent to the drop surface using optical
tensiometry (goniometry)
distribution and deposition, absorption rate, and is the rate-limiting step to absorption.
clearance time; aerosolization behavior and per- Amorphization of poorly water-soluble drugs
formance of respiratory formulations; content and can increase dissolution rates and achieve super-
dose uniformity; and flow and packing properties, saturation, thereby the bioavailability (Yamashita
mixing and segregation of powders, etc. et al. 2003). The amorphous forms, due to higher
There are various principles and techniques molecular mobility as compared to the
used for particle-size measurement. Among the corresponding crystalline form, may have
techniques most commonly used in the pharma- enhanced dissolution rate and this difference can
ceutical research and development, microscopy is then be used to estimate the degree of amorphous
often applied as an absolute particle-sizing content in a given sample. Although the amor-
method because it is the only method where the phous form will have a higher dissolution rate
individual particles can be observed, measured, because of higher free energy, and higher surface
and their shape determined. Static and dynamic area after certain particle engineering processes,
light scattering, Coulter counter (electrical zone there is an inherent risk of recrystallization in the
sensing), time of flight (TOF), and cascade dissolution fluid. Nevertheless, the amount
impactions are also widely used methods dissolved from a drug composition versus that
(Shekunov et al. 2007). from the pure crystalline drug, i.e., extent of
Pharmaceutical powders generated by cryo- supersaturation, has been used to quantify crys-
genic technologies were found suitable for pul- tallinity in drug solid dispersion/solution systems.
monary delivery (Vaughn et al. 2006; Yang et al.
2008b). The recent trend of systemic pulmonary
Density Measurement
drug delivery makes it very important to under-
Solid density is a fundamental physical property
stand the correlation between the aerodynamic
of pharmaceutical powders. Generally, crystalline
diameter, determined by in vitro measurements
state of materials has a higher density than their
or in vivo lung deposition studies, and different
amorphous counterparts because the atoms in the
geometric diameters measured by a variety of
crystal lattice are located at a minimum possible
nonaerodynamic techniques. Andersen cascade
distance from each other. An increase in lattice
impactor (ACI), next-generation impactor
disorder (i.e., increase in amorphous phase) usu-
(NGI), and time-of-flight aerodynamic particle
ally results in an increase in volume and therefore
sizer (APS) are commonly employed to measure
a decrease in density (Suryanarayan 1985).
aerodynamic particle sizes.
Hence, density can also be used as an alternative
parameter for investigating the crystallinity of
Dissolution pharmaceutical powders. The cryogenic
It is well recognized that the low dissolution rate technologies discussed in this chapter can pro-
of poorly water-soluble drugs in biological fluids duce amorphous, low-density pharmaceutical
powders. Bulk, tapped and true density Brittleness Measurement

measurements, as defined by US Pharmacopeia, Mechanical testing by the texture analyzer can
are commonly used to characterize these pharma- measure the force versus the compaction of a
ceutical powders. given amount of TFF powders. By calculating
the slope of the resulting forces over a given
amount of the powder in mass, the brittleness
Dynamic Vapor Sorption (DVS)
can be determined. Wang et al. compared brittle-
DVS provides accurate gravimetric data in con-
ness of TFF RHOMAN, rhodamine B and man-
junction with a control of humidity and tempera-
nitol, at various solid loadings between 0.5% and
ture. It does this by varying the vapor
10%, and reported that the TFF RHOMAN pow-
concentration surrounding the sample and mea-
der at higher solid loading resulted in a higher
suring the change in mass which this produces in
value of the slope in the unit of N/g, representing
a humidity-controlled microbalance system. Gas
that higher energy to obtain brittle fracture of the
and vapor absorption occur into the disordered
powders (Wang et al. 2014).
regions located at the surface, and absorption
phenomenon is known to accelerate where disor-
dered surface regions are present. Water vapor is
most commonly used, though it is possible to use 11.4 Examples of Poorly
a wide range of organic solvents. Water vapor Water-Soluble Drug
sorption is particularly a useful method to assess Compositions Made by
the presence of amorphous material either as a the Cryogenic Processes
single component or in combination (Costantino with Improved Dissolution
et al. 1998; Stubberud and Forbes 1998). Properties
In view of the above description of cryogenic

Inverse Gas Chromatography (IGC)
technologies, step-by-step examples of using
IGC is a vapor sorption technique for studying
cryogenic technologies to engineer poorly
solids using gas chromatography principles, in
water-soluble drug compositions to improve
which the powder is packed into or coated onto
their dissolution profiles are provided below.
a chromatography column and a series of known
The examples include the development of small
nonpolar and polar probe gases are eluted.
molecule compounds for oral and pulmonary
Interactions between the gas molecules and the
drug delivery, and development of biologics
stationary phase (powder) result in a characteris-
vaccines that contain insoluble alum adjuvants.
tic net retention volume, which is used in the
determination of free energy of adsorption, pow-
der crystallinity, and other thermodynamic sur-
face parameters (Hickey et al. 2007a). 11.4.1 Cyclosporin
IGC is nondestructive and considers only the
outermost layer of the sample (Feeley et al. 1998), CsA, a potent immunosuppressant, is a BCS
i.e., the region of the substance directly involved Class II drug with aqueous solubility of 3.69 μg/
during interactions. It is a sensitive technique and mL at 37 C and log P value of 3.0. Previous
recently has found its use in pharmaceuticals, research studies have shown that the administra-
such as in characterization of DPI (Sethuraman tion of CsA to the lungs in addition to base
and Hickey 2002) and pMDI (Traini et al. 2005) therapy can increase the overall survival rate of
formulations, for which adhesional properties are lung transplant patients (Iacono et al. 1997;
considered crucial for their aerodynamic perfor- Costantino et al. 2004). However, due to its
mance. IGC can be used for the determination of poor solubility, previous studies have utilized
surface energy and surface acid/base properties nebulized CsA in propylene glycol solution for
which directly influence adhesional properties. inhalation to increase its bioavailability (Burkart
et al. 2003). The inhalation of propylene glycol • Primary lyophilization was conducted with
resulted in the removal of a number of patients a stage temperature of 35 C and a vac-
from the study due to extreme irritation of the uum pressure of 165 mTorr for 24 h.
airways despite the use of local anesthesia. The • Secondary lyophilization was conducted
use of rapidly dissolving, amorphous dry powder with a stage temperature of 20 C and a
of CsA would alleviate the issues associated with vacuum pressure of 37.5 mTorr for 24 h to
the use of propylene glycol while still allowing obtain the final dry powder formulation.
for improved dissolution and bioavailability. Results
Additionally, the higher deposition efficiency • After SFD processing, both pure CsA and
obtained by using a DPI would also allow for CsA/inulin formulations showed high spe-
lower metered doses and lead to improved patient cific surface areas ranging from 145 to
compliance. SFD-processed amorphous dry 185 m2/g for CsA/inulin formulations and
powders of CsA and inulin were prepared to 40 m2/g for pure CsA.
examine its viability as a pulmonary delivery • Analysis of the secondary structure of the
formulation. SFD-processed CsA formulations using
FTIR confirmed the amorphous nature of
Method Capsule 1 these formulations.
• Cascade impaction studies revealed that
Preparation of a Cyclosporine A Solid Disper-
SFD-processed CsA formulations had high
sion for Inhalation by SFD
respirable (>75%) and fine particle
Based on the method reported by Zijlstra et al.
fractions (50%), making them ideal
(2007)
candidates for inhalation delivery.
Objective
• Dissolution testing showed that
• To obtain a dry powder formulation of CsA
SFD-processed CsA formulations had
for pulmonary delivery with enhanced
superior wetting and dissolution perfor-
chemical and physical properties, such as
mance compared to physical mixtures of
wetting, dissolution rate, and aerodynamic
bulk CsA and inulin.
performance
• Liquid nitrogen
• CsA, inulin and tert-butyl alcohol This study highlighted the use of SFD as a
• SFD apparatus equipped with a heater method to prepare an amorphous dry powder
two-fluid nozzle (0.5 mm orifice) formulation of CsA and inulin for pulmonary
• Christ model Alpha 2–4 stage lyophilizer delivery. By utilizing SFD, a high surface, wetta-
Method ble, and rapid dissolving formulation was
• CsA and inulin were first dissolved into prepared that had superior aerosol performance
tert-butyl alcohol and water, respectively, and enhanced dissolution. Based on these results,
and then mixed at a 40/60 volume ratio with SFD process is a viable alternative to prepare CsA
a constant 5% (w/v) solid loading. formulations for pulmonary delivery in lung
• Formulations with CsA to inulin ratios of transplant patients, compared to the propylene
10, 20, 30, 50, and 100% were prepared. glycol-based CsA formulation.
• After mixing, solutions were sprayed into
liquid nitrogen vapor at a flow rate of
6 mL/min with an atomizing airflow of 11.4.2 Carbamazepine
500 L/h.
• After spraying, frozen droplets were then Carbamazepine was the first poorly water-soluble
collected and placed into a precooled drug used as a model for SFL process. Carba-
(35 C) stage lyophilizer. mazepine is an anticonvulsant and mood-
stabilizing drug used primarily in the treatment of Method

epilepsy. Although the drug has been in use for • Prepare carbamazepine feed solution by
more than 20 years, oral administration of carba- dissolving the drug in THF, dissolving the
mazepine encounters multiple challenges, includ- hydrophilic excipient SLS in purified water,
ing low aqueous solubility (17.7 μg/mL at 25 C, as listed in the table below, then mix the
log P value of 2.45) with high dosage required for organic and aqueous solutions to form a
therapeutic effect (more than 100 mg/day), a one-phase cosolvent solution with a solid
narrow therapeutic window, and dissolution- loading of 0.44% (w/v).
limited bioavailability (Bertilsson and Tomson
1986). Carbamazepine is classified as a BCS Component Solvent
Class II drug (Amidon et al. 1995). The low Organic Carbamazepine, THF, 29.80 g
rate of dissolution in aqueous biological media phase 0.20 g
is considered to be a likely cause of the irregular Aqueous SLS, 0.20 g Purified water,
and delayed absorption issues encountered with phase 59.6 g
the oral delivery of carbamazepine (Moneghini
et al. 2001). In vivo pharmacokinetic studies • Fill a clean large 4-liter insulated beaker
revealed a strong correlation between oral bio- with liquid nitrogen.
availability and the physical form and formula- • Spray and atomize the carbamazepine feed
tion of carbamazepine (Hickey et al. 2007a, b), solution beneath liquid nitrogen surface at
suggesting that improvement of the dissolution 5000 psi constant pressure through a
properties of carbamazepine can result in 10-cm-long, 63.5-μm ID PEEK nozzle
improved pharmacokinetics and bioavailability. into the beaker by a syringe pump.
Rogers and co-workers (Rogers et al. 2002a, b) • Monitor liquid nitrogen level in the beaker
developed a novel composition of carbamazepine to ensure that the PEEK nozzle is kept
made by SFL process to improve the dissolution beneath liquid nitrogen level during the
profile. spray.
• Collect the frozen material on a 150-mesh
Method Capsule 2 sieve or wait till the liquid nitrogen
evaporated after the spray is completed.
Preparation of Engineered Carbamazepine
• Transfer the collected frozen material to a
Compositions by SFL
precooled shelf in a tray lyophilizer to
Based on the method reported by Rogers et al.
maintain the frozen material and remove
2002a, b
solvents.
Objective
• Store the obtained flowable dry powder of
• To enhance the dissolution rate of poorly
carbamazepine composition at room tem-
water-soluble drug carbamazepine by using
perature under vacuum in the desiccator.
a novel SFL process to engineer the drug
composition.
Equipment and Materials Results
• SFL apparatus • SEM images demonstrated that porous
• Liquid nitrogen micron-sized particles consisting of carba-
• Carbamazepine, sodium lauryl sulfate mazepine and SLS matrix system were cre-
(SLS), tetrahydrofuran (THF), and purified ated by SFL process.
water • XRD indicated that the highly crystalline
• Bench-top tray lyophilizer carbamazepine bulk powder was
• Desiccator and desiccant
transferred to complete amorphous state by Method Capsule 3

the SFL process.
Preparation of Engineered High-Potency
• SFL-processed carbamazepine composition
Danazol Compositions by SFL
was wetted and dissolved immediately in
Based on the study reported by Hu et al.
purified water; almost 100% carbamaze-
(2004a, b, c)
pine was released within 10 min, signifi-
Objective
cantly greater than carbamazepine bulk
• To investigate the use of organic solvents in
powder and the physical mixture of carba-
the SFL particle-engineering process to
mazepine and SLS.
make rapid-dissolving high-potency dana-
zol powders and to examine their particle
size, surface area, and dissolution rate
Danazol/PVP K-15 ratio Danazol PVP Acetonitrile DCM Potency Solid

(w/w) (g) K-15 (g) (mL) (mL) (%) loading (%)
1:2 0.2 0.4 70 – 33 0.86
1:1 0.2 0.2 70 – 50 0.57
2:1 0.4 0.2 70 – 66 0.86
3:1 0.6 0.2 65 5 75 1.14
10:1 1 0.1 55 15 91 1.57
This study was a proof of concept that SFL Equipment and Materials
process can significantly improve the dissolution • SFL apparatus
properties of poorly water-soluble drug. • Liquid nitrogen
• Danazol micronized powder, polyvinylpyr-
rolidone (PVP) K-15, sodium lauryl sulfate
11.4.3 Danazol (SLS), acetonitrile, dichloromethane
(DCM), and purified water
Danazol is a synthetic steroid derived from • Bench-top tray lyophilizer
ethisterone that has low aqueous solubility • Desiccator and desiccant
(0.5 μg/mL at 37 C) with high permeability Method
across biological membranes (log P value of • Prepare danazol feed solutions by
4.53) (Bakatselou et al. 1991; Badawya et al. dissolving the drug and PVP K-15 in vari-
1996). Therefore, danazol is classified as BCS ous weight ratios in acetonitrile or acetoni-
Class II drug. The oral bioavailability of danazol trile/dichloromethane mixtures to form
is dissolution rate-limited. Many techniques have solutions with total solid loading up to
been utilized to enhance the dissolution of dana- 1.6% (w/v), as listed in the table below.
zol. However, because of the practically insoluble • Fill a clean insulated container with liquid
nature of crystalline danazol, the improvement of nitrogen.
aqueous solubility was limited. • Spray and atomize the danazol feed
The several different danazol formulations solutions beneath liquid nitrogen surface,
engineered by SFL process were reported with a constant pressure of 2000 psi to
(Hu et al. 2002; Rogers et al. 2002a, b). After provide a flow rate of 50 mL/min for the
initial proof-of-concept studies, formulations feed solution to spray through PEEK tubing
with high potency and enhanced stability of of 127 μm ID into liquid nitrogen using a
amorphous state (i.e., achieved high Tg), besides syringe pump.
high dissolution rates, were designed.
• Monitor liquid nitrogen level in the beaker with smooth primary particle size of about
to ensure that the PEEK nozzle is kept 100 nm in diameter as shown in Fig. 11.10c, d,
beneath liquid nitrogen level during the in contrast to the micron-sized crystalline bulk
spray. danazol in Fig. 11.10a. Because SFL-processed
• Collect the frozen material on a 150-mesh powders have high surface areas and contain
sieve or wait till the liquid nitrogen has amorphous danazol, enhanced dissolution of the
evaporated after the spray is completed. poorly water-soluble drug in aqueous media was
• Transfer the collected frozen material to a achieved.
precooled shelf in a tray lyophilizer to A subsequent stability study was conducted
maintain the frozen material and remove for the SFL-processed danazol/PVP K-15 powder
solvents. A cold trap is connected to the (75% potency) at cycle conditions (5 to 40 C
lyophilizer for the compositions prepared every 3 h). It was found that the amorphous
with DCM. structure and rapid dissolution characteristics
• Store the obtained dried powders in glass were maintained after one month of cycled stabil-
vials at room temperature under vacuum in ity conditions (Hu et al. 2004a, b, c). The high
the desiccator. stability of amorphous SFL powders was partially
attributed to the selection of PVP K-15 as stabi-
lizer in the composition. PVP K-15 has a Tg of
Results 146 C; the interaction of danazol with PVP K-15
• XRD indicated that the SFL-processed may lead to a reduction in the molecular mobility
micronized danazol/PVP K-15 of danazol in the formed solid solution/dispersion
compositions with potencies of 50–91% of danazol/PVP K-15 powder.
were all amorphous.
• Surface areas of these SFL-processed dana- Method Capsule 4
zol/PVP K-15 powders were in the range of
28–115 m2/g, in a reversed order to the Preparation of Amorphous High-Potency
increasing solid loading of the SFL feed Danazol Compositions Using SFL
solutions. Processed o/w Emulsions
• Contact angles of these SFL-processed Based on the study reported by Rogers et al.
danazol/PVP K-15 powders were in the (2003a, b)
range of 22–35 , much reduced compared Because of the innate hydrophobicity of the
to 57 of the unprocessed bulk danazol. poorly water-soluble drugs, there was an
Among different SFL-processed upper limit for the drug concentrations
compositions, contact angles increased with that could be dissolved in solvent/cosolvent
increasing potencies of danazol in the system. Low drug concentrations in the
compositions. feed solution for SFL process resulted. To
• SFL-processed danazol compositions further increase drug loading in feed solu-
exhibited significantly enhanced dissolution, o/w emulsions with higher danazol
tion rates with 95% of danazol dissolved and excipient concentrations were
in only 2 min for the high-potency compo- formulated for SFL process (Rogers et al.
sition, whereas the micronized bulk danazol 2003b), as the total concentration of drug in
dissolved slowly with only 30% of the the o/w emulsions could be much larger
danazol released in the same time frame. than in the cosolvent because of the high
solubility of hydrophobic drug in the inter-
nal organic phase of the emulsion.
The rapid freezing of SFL process produced Objective
porous, nanostructured aggregates of the danazol/ • To further increase drug loading in feed
PVP K-15 compositions as seen in Fig. 11.10b, solution, o/w emulsions with higher
Fig. 11.10 Representative SEM images of SFL-processed danazol/PVP K-15 compositions: Bulk micronized danazol
(a); SFL danazol/PVP K-15 (50% potency) powder at magnification of 20 K (b); at magnification of 60 K (c); at
magnification of 200 K (d). (Adapted from Hu et al. (2004) with permission from Elsevier)
danazol and excipient concentrations were dichloromethane ethylacetate (DCM), and

formulated for SFL process. purified water
Equipment and Materials • Rotor-and-stator homogenizer, high-
• SFL apparatus pressure homogenizer
• Liquid nitrogen • Bench-top tray lyophilizer
• Danazol, poly(vinyl alcohol) (PVA, MW • Desiccator and desiccant
22000), poloxamer, PVP K15, Method
Formulations Component wt. ratio Solvent Potency

Solution Danazol 2 THF/water 40%
PVA (MW 22 000) 1
Poloxamer 407 1
PVP K-15 1
Emulsion Danazol 20 Ethyl acetate/ water, or 87%
PVA (MW 22 000) 1 DCM/water
Poloxamer 407 1
PVP K-15 1
• Prepare danazol SFL solution by dissolving Results

the drug and excipients in THF and water, • Total drug/excipients concentrations of
respectively, then mix the organic and 5.75–7.5% were achieved in the emulsion
aqueous phases to form feed solution with formulations, compared to that of 0.55% in
a potency of 40%, as listed in the table the cosolvent solution composition.
below. • XRD indicated that the lyophilized danazol
• Prepare oil-in-water (o/w) emulsions of SFL-solution composition and
danazol for SFL process by high-pressure SFL-emulsion compositions were amor-
homogenization. Dissolve danazol and phous, even for SFL-emulsion
excipients in organic solvents and water, compositions with high API-to-excipients
respectively. Slowly pour the organic ratio of 20:3.
phase into the aqueous phase under con- • Surface areas increased with increasing
stant mixing, then blend for 1 min using a drug and excipient concentrations, ranging
high-speed rotor-and-stator homogenizer. from 8.9 m2/g of the SFL-solution compo-
Further homogenize the emulsion for ten sition to 83.1 m2/g of the SFL-emulsion
cycles at 20,000 PSI (138 MPa) using a compositions.
high-pressure homogenizer to reduce the • Both danazol SFL-solution composition
oil droplets to < 1 μm in diameter. and SFL-emulsion compositions wetted
• Fill clean insulated containers with liquid and dissolved rapidly (100% in 2 min)
nitrogen for SFL process. than the slowly frozen counterpart and
• Spray and atomize the danazol feed solu- bulk danazol (50% in 2 min). Even for the
tion and emulsions beneath liquid nitrogen SFL-emulsion compositions of high API-
surface at 5000 psi (34.5 MPa) at a flow rate to-excipients ratios, >90% of the danazol
of 20 mL/min through a 127 μm ID PEEK dissolved within 5 min. The SFL-emulsion
nozzle measuring 15 cm in length by a compositions retained the high dissolution
syringe pump. rates that were achieved from SFL-solution
• Monitor and top up liquid nitrogen level in composition.
the container to ensure that the PEEK noz-
zle is immersed under liquid nitrogen level
all the time. High-potency formulations with high drug-to-
• Collect the frozen material on a 150-mesh excipient ratios and rapid dissolution rates would
sieve or wait till the liquid nitrogen has be advantageous in increasing dosages and in
evaporated after the spray is completed. ameliorating side effects attributed to less
• Lyophilize the collected frozen material excipients needed.
using a tray lyophilizer equipped with a
liquid nitrogen cold trap to condense
DCM or ethyl acetate, of which the low 11.4.4 Tacrolimus
melting points exceed the capture capacity
of condenser. Maintain vacuum of Tacrolimus is a hydrophobic macrolide antibiotic
100 mTorr throughout the lyophilization that is used as a potent immunosuppressive agent,
cycle. and has superior immunosuppressive effect com-
• Store the obtained dried powders under pared to cyclosporine A. However, the erratic oral
vacuum in the desiccator at room absorption profiles of tacrolimus have limited its
temperature. therapeutic efficiency. It was reported that the oral
bioavailability of tacrolimus ranges from 4% to
93% with a mean value of 25% (Wallemacq and tip 10 cm above the drum. The feed solu-
Verbeeck 2001). Various innovative formulations tion drops spread and form thin layer frozen
and technologies have been used to improve the disk simultaneously upon impacting on
bioavailability of tacrolimus, with the theory that the drum.
increasing the solubility through polymorphism • Place a metal pan filled with liquid nitrogen
can have a significant increase in biopharmaceu- under the drum to collect and maintain the
tical activity. generated frozen material.
• Transfer immediately the frozen material
Method Capsule 5 into a bench-top tray lyophilizer with shelf
precooled to 60 C after evaporation of
Preparation of Tacrolimus Solid Dispersions
excessive liquid nitrogen in the collecting
for Oral Delivery Using TFF
pan, and start lyophilization.
Based on the study reported by Overhoff et al.
• Store the obtained dried powders of TAC
(2008), TFF technology was also employed
compositions at room temperature under vac-
to make engineered tacrolimus
uum in the desiccator until characterization.
compositions with hydrophilic stabilizers.
Objective
• To investigate tacrolimus solid dispersions Results
containing various stabilizers prepared by • XRD indicated that the TFF-engineered
TFF process and to determine the effect on TAC compositions were all amorphous.
their ability to form supersaturated • SEM displayed highly porous network of
solutions in aqueous media and on enhanc- nanostructured aggregates of the
ing transport across biological membranes TFF-engineered TAC compositions, in
Equipment and Materials contrast to the plate-shaped crystalline
• TFF apparatus bulk TAC with particle sizes ranging from
• Liquid nitrogen a few microns to over 120 μm in diameter.
• Tacrolimus (TAC), poly(vinyl alcohol) • Supersaturation dissolution testing
(PVA), poloxamer 407 (P407), and sodium demonstrated the three TFF-engineered
dodecyl sulfate (SDS) TAC compositions, as well as the commer-
• Bench-top tray lyophilizer cial product Prograf®, showed rapid dissolu-
• Desiccator and desiccant tion reaching their maximum supersaturation
• Dissolution tester within 2 h and achieved supersaturation rela-
Method tive to the solubility of bulk crystalline TAC.
• Prepare feed solutions of TAC compositions However, only the TFF-processed TAC/SDS
by dissolving TAC and excipient(s) at 1:1 ¼ 1/1 composition had higher supersatura-
(w/w) ratio and 1.0% solid loading in a tion than Prograf®.
cosolvent acetontrile/water (60/40, v/v), as
listed in the table below. In this study, the selected stabilizers, including
• Apply the feed solutions to the rotating partially hydrolyzed PVA, which has been shown
stainless-steel drum that precooled to to increase drug concentration in vivo (Suzuki
150 C drop-wisely from a glass funnel and Sunada 1998), P407 which has been shown
Acetonitrile/water Potency
TAC/excipient(s) ratio (w/w) Excipient(s) (v/v) (%) Solid loading (%)
1:1 SDS 60/40 50 1.0
1:1 PVA/P407 (1/1) 60/40 50 1.0
1:1 P407 60/40 50 1.0
140 80
70
TAC Whole blood Conc. (ng/ml)

TAC Whole blood Conc. (ng/ml) 120 60
50
100
40
30
80
20
60 10
0
0 1 2 3 4 5 6
40 Time (hr)
20
0
0 5 10 15 20
Time (hr)
Fig. 11.11 Mean whole blood absorption levels of tacrolimus/SDS ( filled square), TFF-processed
tacrolimus compositions produced using the TFF process tacrolimus/PVA:P407 ( filled triangle), TFF-processed
compared to Prograf®. Powders were given in gelatin tacrolimus/P407 ( filled circle), Prograf® capsule powder
capsule containing 1.5 mg equivalent tacrolimus (5 mg/ ( filled diamond). (Reproduced from Overhoff et al. (2008)
kg) dosed via oral gavage to a rat model: TFF-processed with permission from Springer)
to alter surface properties of crystals, and SDS to make respirable low-density

which is a nonpolymeric anionic surfactant. SDS microparticles of tacrolimus.
may be used to facilitate wetting and dissolution Objective
rates. TAC has good solubility in organic solvent • To compare performance of dry powder
and is readily dissolvable in the cosolvent. inhalation formulations of tacrolimus
Single-dose pharmacokinetic study of orally (TAC) prepared using TFF to that of pro-
administered TAC compositions in a rat model duced by micronization in the in vitro and
was conducted to determine the in vivo perfor- in vivo studies
mance of the TFF-processed compositions. The Equipment and Materials
results suggested that TFF-processed TAC/P407 • TFF apparatus
¼ 1/1 achieved the greatest absorption with a 1.5- • Liquid nitrogen
fold increase in AUC and higher Cmax compared • Tacrolimus monohydrate, mannitol
to Prograf®. All the TFF-engineered tacrolimus (MAN), Polysorbate 80, acetonitrile
compositions had a shorter Tmax compared to • Bench-top tray lyophilizer
Prograf®, as seen in Fig. 11.11. It is therefore • Desiccator and desiccant
concluded that the enhanced physico-chemical • Next generation pharmaceutical impactor
properties of TFF-engineered TAC compositions (NGI)
led to enhanced in vivo absorption over the cur- Method
rent commercial product of TAC. • Dissolve MAN in purified water and dis-
solve TAC in acetonitrile
Method Capsule 6 • Prepare a co-solvent mixture of ACN and
water (60:40 v/v) containing TAC and
Dry Powder Inhalation Formulations of
MAN (1:1 w/w) with total solids content
Tacrolimus Prepared by TFF
of 0.75% w/v
Based on the study reported by Wang et al.
(2014), TFF technology was also employed
• Freeze the co-solvent solution rapidly on a lung delivery when incorporated into a commer-
cryogenically cooled rotating stainless steel cially available DPI. ATR-FTIR results showed
surface (50 3 C). that there was no interaction between TAC and
• Remove resulting thin film from the surface MAN molecules. As a result, drug molecules did
using a scraper not have an influence from seeding of crystalline
• Keep frozen samples in liquid nitrogen mannitol. Regarding in vivo study, the TFF for-
• Sublimate solvents in lyophilizer over 48 h mulation exhibited higher AUC and slower
at pressures below 200 mTorr, clearance rate in the lung tissue compared to
• Ramp the shelf temperature up from 40 to the micronized formulation due to the faster
25 C dissolution rate and the less susceptible of the
• Purge dry nitrogen into the chamber to TFF formulation to the clearance mechanisms of
equilibrate to atmospheric pressure before the lung.
removing sample from the lyophilizer
• Store final product in a vacuum desiccator Method Capsule 7
at room temperature
Development of High-Potency Tacrolimus for
• Prepare a physical mixture of micronized
Dry Powder Inhalation Using TFF
TAC and micronized MAN (1:1 by weight,
Based on the study reported by Sahakijpijarn
micronized TACMAN) by blending the
et al. (2020a)
two powders in a tubular mixer
Objective
Results
• To investigate the feasibility of preparing
• DSC and XRD results indicated that TAC
high potency tacrolimus dry powder inha-
in the TFF TACMAN was amorphous,
lation formulations using TFF
while MAN in the TFF TACMAN (1:1)
exhibited crystallinity.
• TFF apparatus
• TAC in the TFF TACMAN formulation
• Liquid nitrogen
remained in an amorphous form for up to
• Tacrolimus, lactose monohydrate, manni-
six months.
tol, trehalose
• SEM images showed a highly porous struc-
• Bench top tray lyophilizer
ture of the TFF formulations.
• High-resistance and low-resistance
• Figures 11.12 and 11.13 presented a highly
Monodose RS01 dry powder inhalers
porous structure and a relatively small aero-
• HPMC capsules (size 3)
dynamic size distribution (GSD ¼ 2.08 μm)
of the TFF TACMAN formulation.
• Next generation pharmaceutical impactor
• Almost 80% of the TFF formulation was
(NGI)
delivered to the lung using Miat®
Method
Monodose inhaler.
• Tacrolimus and lactose were dissolved in a
• In vivo study presented that the AUC of the
co-solvent mixture of ACN and water (60:
TFF formulation was higher than the
40 v/v) containing different amount of lac-
micronized TAC while the lung clearance
tose with total solids content of 0.75%, or
rate of the micronized particles was a lot
2.5% w/v
faster than the aerosolized TFF particles.
• Freeze the co-solvent solution rapidly on a
• The TFF TACMAN formulation could
cryogenically cooled rotating stainless-steel
avoid clearance mechanisms of the lung
surface (70 3 C).
and deliver TAC on pulmonary epithelium,
• Keep frozen samples in liquid nitrogen.
while minimizing systemic concentration.
• Primary dry at 40 C below 100 mTorr
According to the particle size distribution, the over 20 h.
TFF formulation of TAC is appropriate for deep
Fig. 11.12 The SEM images of the micronized (a) and the TFF (b) of the TACMAN at low (2.70K X, top) and high
(25.00K X, bottom) magnifications. (Reproduced from Wang et al. (2014) with permission from Springer)
40
Formulations TED (%) FPF of Emitted Dose(%) MMAD (μm) GSD (μm)
35
Percentage of Metered Dose (%)
TFF TACMAN TFF TACMAN 94.2±0.7 83.3±1.7 2.26±0.13 2.08±0.10
30
25
20
15
10
0
Capsule Device Neck Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Stage 6 Stage 7 MOC
Fig. 11.13 NGI results of aerodynamic diameter distribution of TFF TACMAN formulation released from a Miat®
monodose inhaler at the flow rate of 90 L/min. (Reproduced from Wang et al. (2014) with permission from Springer)
• Ramp the shelf temperature up from 40 to • Store bulk powder and capsules filled with
25 C over 20 h, and hold at 25 C below powder in a borosilicate glass vial.
100 mTorr over 20 h. • Dry samples in vials in a lyophilizer at
25 C and 100 mTorr for 5 h, then
backfilled with nitrogen gas before closing minimized while maintaining good aerosolization
the rubber stopper inside the lyophilizer. and enhancing the drug dissolution. The brittle
Results matrix of nanostructured aggregates can be
• DSC and XRD results indicated that TAC aerosolized by an off-the-shelf dry powder inhaler
and LAC were amorphous. device with less flow rate dependence.
• XRD diffractogram showed TAC in
TAC/LAC (95/5 w/w) remained in an
amorphous form after storage at 40 C/
11.4.5 Itraconazole
75%RH and 25 C/60%RH for up to six
months.
Itraconazole is a broad-spectrum antimycotic
• SEM images showed TFF bulk powders
triazole used for both prophylaxis and treatment
generally exhibited a brittle matrix mor-
of invasive fungal diseases for the last two
phology of highly porous particles.
decades. Itraconazole has pH-dependent solubil-
• TFF TAC/LAC (95/5) exhibited optimal
ity with extremely low value of approximately
aerosol performance (69.31 3.45% fine
1 ng/mL at neutral pH and approximately 4 μg/
particle fraction of recovered dose, 2.61
mL at pH 1 (Peeters et al. 2002). Given the high
0.25 μm mass medium aerodynamic
log P value of 6.2, itraconazole is classified as a
diameter).
BCS class II drug (Amidon et al. 1995).
• The aerosol performance of TFF TAC/LAC
Sporanox® oral capsule is a currently available
(95/5) remained unchanged after six
marketed oral dosage form of itraconazole. How-
months of storage at 25 C/60%.
ever, the oral absorption of ITZ in a subset of
• ssNMR indicated no intermolecular
immunocompromised patients was not optimal,
interactions between tacrolimus and lactose
and the pharmacokinetics varied considerably
are observed; however, the system has good
among patients (Poirier et al. 1997). To treat
miscibility between tacrolimus and lactose
invasive fungal infection, especially Aspergillus
at a domain size of ~100 nm.
spp. infections, itraconazole levels of greater than
• TFF TAC/LAC (50/50) and TAC/LAC
0.5 μg/g of lung tissue, or 0.5 μg/mL of blood, are
(95/5) exhibited higher and faster dissolu-
required (Sobel 2000).
tion profiles than physical mixture of
A novel composition containing itraconazole:
unprocessed TAC and LAC powder.
mannitol:lecithin (1:0.5:0.2, w/w/w) for pulmo-
• The FPFs (of recovered dose) of F6 dis-
nary delivery was made by TFF technology
persed using high-resistance RS01
(Yang et al. 2008b). Here, we provide a step-by-
Plastiape® DPIs at 1, 2, and 4 kPa were
step procedure for a standard batch of the
63.24 2.09%, 65.78 2.68%, 69.31
itraconazole composition.
3.45%, respectively.
• The FPFs (of recovered dose) of F6 dis-
Method Capsule 8
persed using low-resistance RS01
Plastiape® DPIs at 1, 2, and 4 kPa were Preparation of Amorphous Nanoparticulate
52.59 2.66%, 52.19 4.31%, 53.72 Itraconazole Composition for Pulmonary
1.82%, respectively (Figs. 11.14 Delivery Using TFF
and 11.15). Based on the study reported by Yang et al.
(2008a, b)
Objective
This study highlighted the use of TFF as a
• To develop amorphous nanoparticulate
method to prepare a dry powder formulation of
formulations of itraconazole for improved
TAC containing high drug loading for pulmonary
bioavailability following pulmonary
delivery. By ultra-rapid freezing, the amount of
administration
excipient required in a formulation can be
Fig. 11.14 The SEM images of the bulk powder of F1 (TAC/LAC; 50/50) and F6 (TAC/LAC; 95/5) (25KX, top) and
aerosolized powder of F1 (TAC/LAC; 50/50) and F6 (TAC/LAC; 95/5) after dispersing by a DP4 insufflator (50KX,
bottom). (Modified from Sahakijpijarn et al. (2020a) with permission from Elsevier)

profile of TFF TAC/LAC
(50/50), TAC/LAC (95/5),
and physical mixture of
unprocessed powder under
sink condition in phosphate
buffer with 0.2% tween 80.
(Reproduced from
Sahakijpijarn et al. (2020a)
Elsevier)
Equipment and Materials • Dissolution testing revealed TFF-processed

• TFF apparatus itraconazole composition achieved about
• Liquid nitrogen five times higher supersaturation than a
• Itraconazole, mannitol, lecithin, crystalline Wet-milled itraconazole.
1,4-dioxane, and purified water
• Desiccator and desiccant For comparison, a crystalline itraconazole-
Method Cryogenic technologiespoorly water-soluble
• Prepare itraconazole feed solution: dissolve drug compositionsitraconazoleItraconazole nano-
lecithin (118 mg) in 200 mL co-solvent of particle composition was made by wet ball mill-
1,4-dioxane and purified water (65/35, v/v); ing process (named Wetmilled itraconazole).
Subsequently dissolve itraconazole SEM images showed that Wet-milled
(588 mg) and mannitol (294 mg) to form a itraconazole was composed of fractured,
solution of 0.5% (w/v) solid loading with irregularshaped particles with various sizes, rang-
itraconazole: mannitol: lecithin ¼ 1:0.5:0.2 ing from about 150 to 600 nm in length, as shown
weight ratio. in Fig. 11.16a. In contrast, TFF-processed
• Precool the stainless-steel drum of TFF itraconazole composition exhibited a highly
apparatus to 70 C. porous structure with more regularly roundshaped
• Apply the feed solution to the cold drum of particles in aggregated network, as shown in Fig.
TFF apparatus from a glass funnel set 11.16b. Both the milling and TFF process dra-
10 cm above the top surface of the rotating matically reduced itraconazole particles to
stainless-steel drum. nanosize range compared to bulk micron-sized
• Collect the generated frozen material in a itraconazole particles shown in Fig. 11.16.
container filled with liquid nitrogen. Their corresponding mean particle size in
• Transfer immediately the frozen material aqueous dispersion was 230 and 570 nm, respec-
into a bench-top tray lyophilizer with shelf tively. Dissolution testing revealed
precooled to 20 C after evaporation of TFF-processed itraconazole composition
excessive liquid nitrogen, and start achieved about 27 times higher supersaturation
lyophilization. versus itraconazole equilibrium solubility, and
• Store the lyophilized dry powder at room 5 times higher dissolved itraconazole than the
temperature in the desiccator under Wet-milled itraconazole, as seen from Fig. 11.17
vacuum. (Yang et al. 2010).
Results A subsequent in vivo single-dose 24 h phar-
• XRD and DSC confirmed that macokinetic study of inhaled nebulized colloidal
TFF-processed itraconazole:mannitol:leci- dispersions of the TFF-processed amorphous
thin ¼ 1:0.5:0.2 was amorphous. itraconazole composition and the crystalline
• SEM images showed the TFF-processed Wet-milled itraconazole (equivalent to 20 mg
itraconazole composition has a highly itraconazole/mL) were conducted in a rat model.
porous structure with more regularly The results demonstrated a significantly higher
round-shaped particles in aggregated systemic absorption and Cmax of itraconazole in
network. blood of rats that inhaled TFF-processed amor-
• Aqueous colloidal dispersion of the phous itraconazole composition than those that
TFF-processed itraconazole composition inhaled crystalline Wet-milled itraconazole Cryo-
has a mean particle size of 230 nm. genic technologies poorly water-soluble drug
compositions it raconazoleI traconazole (Fig.
Fig. 11.16 SEM images of (a) Wet-milled itraconazole magnification of 20 k, and (c) bulk ITZ as received at a
(wet milling–processed pure ITZ) powder at a magnifica- magnification of 10 k. (Reproduced from Yang et al.
tion of 20 k, (b) URF–ITZ (URF-processed ITZ/mannitol/ (2010) with permission from Elsevier)
lecithin ¼ 1:0.5:0.2, weight ratio) powder at a

profiles of Wet-milled
itraconazole colloidal
dispersion (ITZ/mannitol/
lecithin ¼ 1:0.5:0.2) and
TFF-processed
itraconazole/mannitol/
lecithin ¼ 1:0.5:0.2
colloidal dispersion in
simulated lung fluid (pH ¼
7.4) at supersaturation
conditions (i.e., 100 times
equilibrium solubility of
micronized crystalline
itraconazole was added)
and 37 C. (Reproduced
from Yang et al. (2010)
Elsevier)
Fig. 11.18 (a) Plasma

concentration of
itraconazole in rats, (b) lung
deposition of itraconazole
in rats at 0 and 24 h post
inhalation of a single-dose
nebulized aqueous
Wet-milled itraconazole
colloidal dispersion
(ITZ/mannitol/lecithin ¼ 1:
0.5:0.2) and TFF-processed
itraconazole/mannitol/
lecithin ¼ 1:0.5:0.2
colloidal dispersion. (Data
are presented as mean
SD. *p < 0.05, **p < 0.01.
Adapted from Yang et al.
(2010) with permission
from Elsevier)
11.18a), although the lung depositions of as a BCS class II drug with about 30% of oral
itraconazole were comparable for both inhaled bioavailability (Zhang et al. 2012). Several
compositions (Fig.11.18b). It is concluded from strategies have been used to improve the solubil-
this study that using TFF technology to make ity and bioavailability of FB. This example
amorphous nanoparticulate composition of describes how to improve the oral bioavailability
poorly water-soluble drugs is an alternative and of fenofibrate by TFF.
promising approach to overcome the low aqueous
solubility issues by providing higher dissolution Method Capsule 9
rate and apparent solubility, and subsequently
Preparation of Amorphous Fenofibrate Solid
higher bioavailability.
Dispersions by TFF
Based on the study reported by Zhang et al.
(2012)
11.4.6 Fenofibrate Objective
• To prepare amorphous fenofibrate solid
Fenofibrate (FB) is a pro-drug of fenofibric acid dispersions using the TFF method and to
used in the therapy of hypertriglyceridemia and incorporate the solid dispersions into
dyslipidemia. FB is a lipophilic drug (log P ¼ 5.2) pharmaceutically acceptable dosage forms
and practically insoluble in water. FB is classified for improving the bioavailability
Equipment and Materials • Morphology of amorphous aggregates of

• TFF apparatus samples using SEM illustrated a sponge-
• Liquid nitrogen Fenofibrate (FB) , like structure resulting in the creation of
fenofibric acid, Methocel® E5, Soluplus®, high surface area and the formation of
Hydroxypropyl methylcellulose phthalate loose connection between particles.
NF (HP55), Hydroxypropyl methylcellu- • Both PXRD and MDSC results revealed the
lose acetate succinate LF (HPMCAS-LF), drug was molecularly dispersed in
1,4-dioxane, and water polymers.
• Bench top tray freeze dryer • There were no crystals of FB observed in
• Particle-size analyzer by laser light TFF samples.
scattering • Supersaturation dissolution of FB solid
• BET apparatus dispersions prepared by TFF method
• Differential scanning calorimeter (DSC) showed success in enhancing drug release
• Powder X-ray diffraction (XRD) that exceeded the equilibrium solubility and
• Fourier-transform infrared spectroscopy- maintained FB in solution.
attenuated total reflectance (FTIR-ATR) • In vivo study demonstrated that levels of
• Dissolution tester plasma drug concentration of the amor-
Method phous FB prepared using TFF processing
• Prepare 1% w/v solids content of FB:poly- were drastically higher than the bulk FB.
mer at designed ratio of 1:4, 1:6 or 1:8 in
1,4-dioxane or its mixture with water at
ratio of 8:2 v/v as shown in the table below: The morphology of the porous aggregates of
amorphous FB solid dispersion from the SEM
FB:Polymer images agreed with those surface area
Formulation ratio (%w/w) Solvent measurements using BET. The specific surface
FB: 1:4. 1:6. 1:8 1,4-dioxane area of the TFF formulations increased as com-
Soluplus pared with the bulk FB. Preparation of homoge-
FB:HPMC 1:4. 1:6. 1:8 1,4-dioxane: neous solutions prior to the rapid freezing
E5 water (8:2, v/v) procedure resulted in no crystalline structure
FB: 1:4. 1:6. 1:8 1,4-dioxane observed in TFF formulations. Polymers possibly
HPMCAS prevented crystal growth of FB due to the fact that
FB:HP55 1:4. 1:6. 1:8 1,4-dioxane they were adsorbed on the drug crystal interface
and the crystal nucleation was inhibited.
• Apply solutions of FB-excipient onto the Dissolution of amorphous FB significantly
pre-cooled cryogenic substrate (45 C) increased after TFF processing. FB-HP-55 formu-
• Collect frozen solids and lyophilize using lation had the largest surface area as compared to
freeze dryer in which the temperature is the other formulations. Although, it unexpectedly
increased from 40 to 25 C over 48 h. exhibited the worst dissolution results under
• Store the dry powders in desiccator at room supersaturation conditions and in the dissolution
temperature under vacuum condition. media the recrystallization rate of this formulation
was fastest. This resulted from the decrease of FB
solubility in acidic solution. HP-55 contains a
Results
larger number of acidic groups compared to
• FB-polymer solid dispersions prepared by
other polymers, which results in the poorest dis-
TFF technique provided extremely high
solution performance of FB. After the amorphous
surface area, microstructure, and
TFF powders were compressed into slug tablets,
wettability.
120
100
Bulk FB
Plasma fenofibric acid conc. (μg/ml)

FB-HPMC E5 (1:6) Granule
FB-HPMCAS (1:6) Granule
80
60
40
20
0
0 10 20 30 40 50
Time (h)
Fig. 11.19 Pharmacokinetic profile of drug plasma concentration of fenofibric acid after single dose oral administration
of different formulations to rats. (Reproduced from Zhang et al. (2012) with permission)
compression triggered the recrystallization of FB, Method Capsule 10

which resulted in lower dissolution.
Development of Voriconazole Dry Powder
Figure 11.19 showed the success of the
Inhalation Produced by TFF
enhancement of drug absorption of FB solid dis-
Based on the method reported by Beinborn
persion. This significant improvement of oral bio-
et al. (2012)
availability was achieved in these non-surfactant,
Objective
non-porous excipients processed through the TFF
• To manufacture voriconazole formulations
approach.
using TFF to enhance properties for dry
powder inhalation and to study the process
parameters which impact the morphology
11.4.7 Voriconazole and aerodynamic properties of the resulting
formulations
Voriconazole (VCZ) is a second-generation Equipment and Reagents
triazole antifungal agent. Although the solubility • Liquid nitrogen
of VCZ is slightly lower than the boundary for • Voriconazole (VRC) , lactose
BCS I drugs, VCZ is considered a BCS Class II monohydrate, polyvinylpyrrolidone K30
drug with a moderate lipophilicity (log D7.4 ¼ (Povidone K-30, USP), polyvinylpyr-
1.8). To improve the drug delivery to the target rolidone K12 (Plasdone® K12),
site of pulmonary aspergillosis, the examples hydroxypropyl methylcellulose (HPMC)
below describe the development of voriconazole K3 (Methocel™ K3), 1,4-dioxane, and
dry powder for inhalation prepared by TFF. methanol
• TFF apparatus
• Bench top tray freeze dryer
Method
• Prepare VRC TFF formulations in various
The engineered particles were successfully
process parameters, including the effect of
generated using the TFF process to enhance pow-
stabilizing excipient, drug to excipient
der properties. VRC TFF formulation without
ratio, percentage of solids content, and sol-
stabilizing excipients produced microstructured,
vent composition.
crystalline aggregate particles. There was no
• Dissolve VRC and excipients in
notable influence of solvent composition or per-
1,4-dioxane or deionized water.
centage of solids content on the solid-state
• Drop the solutions onto a precooled rotat-
properties and the aerodynamic performance of
ing cryogenic, steel surface (approximately
the formulation. Drug-excipient ratio had an
40 C) to produce frozen thin films.
effect on preparation of the nanostructure of
• Remove thin films from the steel surface
amorphous solid dispersion formulations using
using a scraper and keep in frozen mass in
the TFF technique.
liquid nitrogen.
All investigated parameters, including type or
• Sublimate the solvents by lyophilization to
grade of stabilizing excipient, drug-excipient
dry powders (perform the lyophilization
ratio, solvent composition, and percentage of
over 48 h at pressures below 200 mTorr,
solids content, in the liquid feed solution signifi-
while the shelf temperature was gradually
cantly affected physicochemical properties and
ramped from 40 to 25 C).
morphology of the resulting formulations. Sol-
• Transfer and store the dried powders in
vent composition in the TFF liquid feed solution
vacuum desiccators at room temperature.
also influenced the spreading of the droplets on
Results
the cryogenic surface which appeared in the spe-
• Microstructured, crystalline low-density
cific surface area and macroscopic powder of the
aggregate particles with the specific surface
TFF formulations. Two solid-state properties
area of approximately 10 m2/g of VCR
were the most significant factors affecting the
were obtained from the TFF formulation
aerodynamic properties of VRC TFF powders.
without stabilizing excipients.
• Nanostructured, amorphous low density
Method Capsule 11
aggregate particles were obtained from
TFF VRC–PVP solutions. The specific sur- Development of High Potency Nanoaggregates
face area of formulations (ranging from of Voriconazole Dry Powder Inhalation
15 to 180 m2/g) depended on the solvent Based on the method reported by Moon et al.
system compositions, grade of polymer, (2019)
and drug-polymer ratio. Objective
• The lowest specific surface area for VRC • To develop high potency nanoaggregates of
formulations were obtained from crystalline voriconazole composition for
formulations manufactured with dry powder inhalation
1,4-dioxane, with and without PVP K12, Equipment and Reagents
however, exhibited the best aerodynamic • Liquid nitrogen
properties. • Voriconazole (VRC), Pearlitol® PF,
• Microstructured crystalline TFF–VRC and acetonitrile
nanostructured amorphous TFF–VRC– • TFF apparatus
PVP K12 (1:2) demonstrated total emitted • Bench top shelf freeze dryer
fractions of 80.6% and 96.5%, fine particle Method
fractions of 43.1% and 42.4%, and mass • Dissolve various ratio of voriconazole
median aerodynamic diameters of 3.5 and (30–100% w/w) and mannitol in a mixture
4.5 μm, respectively, when delivery using a of water and acetonitrile (50:50 v/v)
Handihaler® dry powder inhaler (DPI). • Drop the solution onto a rotating cryogeni-
cally cooled to 60 C stainless steel drum
Fig. 11.20 Fine Particle Fractions of voriconazole with different amounts of mannitol. (Reproduced from Moon et al.
(2019) with permission from American Chemical Society)
• Collect the frozen samples in a stainless • To design a production design space and
steel container filled with liquid nitrogen scale-up of voriconazole nanoaggregates
• Remove the solvents by a shelf lyophilizer Equipment and Reagents
• Store the powder formulations at room tem- • Liquid nitrogen
perature with desiccants • Voriconazole (VRC), Pearlitol® PF,
Results acetonitrile
• The drug loading of voriconazole dry pow- • TFF apparatus
der in the range of 90–97% w/w showed • Bench top shelf freeze dryer.
optimal aerosol perfomance aerosolization Method
of voriconazole potency was obtained • Dissolve voriconazole (95% w/w) and
between 90% and 97% (w/w) (Fig. 11.20). mannitol (5% w/w) in a mixture of water
• At the optimum ratio of voriconazole and and acetonitrile (30:70, 50:50, or 70/30 v/v)
mannitol for high aerosolization, mannitol at a solid content in the solution of 13%
was acting as a surface texture-modifying (w/v)
agent to reduce cohesive-adhesive energy • Drop the solution onto a rotating cryogeni-
between particles (Fig. 11.20). cally cooled (60 or 150 C) stainless
• By TFF, voriconazole was manufactured as steel drum
nanoaggregates, consisting of a few • Collect the frozen samples in a stainless
nanometers of nanoparticles steel container filled with liquid nitrogen
• Crystalline mannitol was phase-separated • Remove the solvents by a shelf lyophilizer
from crystalline voriconazole • Store the powder formulations at room tem-
perature with desiccants
Method Capsule 12 Results
• The voriconazole nanoaggregates
Processing Design Space for Voriconazole
processed at 150 C consisted of smaller
Nanoaggregates for Dry Powder Inhalation
size of nanoparticles, compared to those
Based on the method reported by Moon et al.
processed at 60 C
(2019)
Objective
• The faster nucleation with ultra-rapid • To study the in vivo behavior and pharma-
supercooling was observed at 150 C, cokinetic profiles of crystalline and amor-
compared to TFF process at 60 C phous rapamycin when administered
• The binary solvent system of water/aceto- through dry powder inhalation to the lungs
nitrile (30/70 v/v) generated 20μm porous of rats
mannitol particles and that the mannitol did Equipment and Materials
not work as a surface texture-modifying • TFF apparatus
agent for voriconazole nanoaggregates due • Liquid nitrogen
to possible cryo-phase separation • Rapamycin (Sirolimus), lactose
• When the solid load was increased from 1% monohydrate (Lactohale® LH 200), aceto-
to 3% (w/v), FPF was decreased nitrile, and water
• The lower processing temperature • Bench top tray freeze dryer
(150 C) produced more aerosolizable • Wet ball milling
voriconazole nanoaggregates, providing • Particle-size analyzer by laser light
enhanced FPF, compared to that produced scattering
at higher processing temperature (60 C) • BET apparatus
• High resistance RS00 device resulted in • Differential scanning calorimeter (DSC)
more consistent and higher FPF of the • Powder X-ray diffraction (XRD)
voriconazole nanoaggregates at various • Next generation cascade impactor (NGI)
flow rates or loading doses Method
• Dissolve rapamycin:lactose in a ratio of 1:1
by weight in a co-solvent mixture of aceto-
nitrile and water (3:2) with different final
11.4.8 Rapamycin
solids content 0.40% and 0.75% (w/v)
• Freeze the solution rapidly onto the cryo-
Rapamycin (Sirolimus, RAPA) is a carboxylic
genically cooled stainless steel surface
lactone-lactam macrolide antibiotic currently
(80 C)
approved for prophylaxis of rejection for kidney
• Collect frozen films in a container filled
allograft patients. RAPA is poorly water-soluble
with liquid nitrogen
(2.6 μg/mL). The oral bioavailability of RAPA
• Transfer the frozen formulation to a 70 C
oral solution (Rapamune®) is only about 14%.
freezer until liquid nitrogen is completely
Due to low oral bioavailability, a higher dose of
evaporated
RAPA is required to achieve the therapeutic win-
• Lyophilize the frozen material in the lyoph-
dow; however, this also increases the potential of
ilizer over 24 h at 40 C at pressure
side effects. Therefore, an alternative route of
400 mTorr and increase the temperature to
administration would be another way to improve
25 C over 24 h with a pressure below
the bioavailability and efficacy of this drug. The
200 mTorr, and hold at 25 C for 24 h
example below describes the preparation of an
• Prepare micronized crystalline rapamycin
RAPA amorphous solid dispersion for pulmonary
and lactose at 1:1 weight ratio using wet
delivery.
ball milling and lyophilize at 80 C in the
lyophilizer for 48 h
Method Capsule 13
• Keep all dry powder products in a desicca-
Manufacture of Amorphous Rapamycin Solid tor under vacuum condition at room
Dispersions for Dry Powder Inhalation temperature
Using TFF Results
Based on the study reported by Carvalho et al. • An absence of crystalline peaks from the
(2014) XRD diffractogram indicated the
Objective
amorphous state of both TFF formulations • Resulting GSD indicated that the physical
of rapamycin:lactose with solids content mixture had a wider aerodynamic particle
0.40% and 0.75%. Also, modulated DSC size distribution than the TFF formulations.
indicated that both formulations existed in • In vivo systemic bioavailability of
the amorphous form. rapamycin from TFF formulation was
• A resulting TGA profile showed the TFF greater than its crystalline counterpart. The
formulations initially decomposed at a tem- dissolution rate of the crystalline formula-
perature of about 180 C. Minimal solvent tion was low and this resulted in poor
evaporation of both formulations was only absorption from the lung fluid and low sys-
3.2% through the range of temperature temic bioavailability. The level of drug in
from 45 to 110 C. the lung from 12–24 h indicated that TFF
• The specific surface area of TFF powder powder and crystalline powder stayed in
formulations dramatically increased com- the lung for the same period of time. How-
pared to the bulk and wet milling materials ever, the TFF formulation exhibited a
(see table below). higher level of systemic bioavailability.
The percentage of FPF at a more distal
Specific surface area part of the lung also increased.
Formulation SD (m2/g) • The wet ball milling reduced the particle
Lactose LH200 0.34 0.02 size of bulk rapamycin and lactose to
Rapamycin 0.57 0.08 sub-micron range. The irregular to round-
Wet ball milled 10.76 0.01 shaped particles of the drug adhered to the
lactose lactose surface were confirmed by SEM
Wet ball milled 14.29 0.25 images. The improvement of aerosolization
rapamycin during inhalation of the milled samples was
RapaLac_0.75% 69.57 6.25 expected. Nonetheless, both TFF powders
RapaLac_0.40% 85.72 19.86 gave better aerosolization characteristics
compared to the milled formulation. This
• Diluted solid loading solution (0.40%) study further investigated the influence of
resulted in a less dense TFF powder with solids loading concentration on aerosoliza-
greater specific surface area. Therefore, this tion properties of the rapamycin TFF for-
TFF powder exhibited better performance mulation. The specific surface area of TFF
using the dry powder inhaler device by powder formulations was more increased
generating greater FPF values with smaller for the formulation containing a solids con-
MMAD and GSD values than the least tent of 0.40% than for the 0.75% formula-
porous powder. tion. SEM pictures also showed the lower
• In vitro aerosol performance was density of the lower solids content
demonstrated using NGI. The results are formulation.
presented as follows:
Total emitted Fine particle Geometric

dose – TED fraction – FPF Mass media aerodynamic standard
Formulation (%) (%) diameter – MMAD (μm) deviation – GSD
RapaLac 78.92 36.79 1.81 4.26
physical mixture
RapaLac_0.40% 97.14 72.11 2.1 2.25
RapaLac_0.75% 94.71 61.29 2.43 2.76
• The TFF powder formulation showed supe- 11.4.9 Fixed-Dose Combination Drug
rior in vitro aerosol performance to the Formulation
physical mixture because the brittle powder
matrix of TFF samples is easy to disinte- Method Capsule 14
grate to low density aerosolized particles.
Pulmonary Delivery of a Fixed-Dose Combi-
The reduction of electrostatic and van der
nation Drug Formulation Prepared Using
Waals forces is likely to result in low adhe-
TFF
sion between particles.
Based on the study reported by Liu et al.
• Rapamycin in the TFF formulation was an
(2015)
amorphous material, which basically
Objective
enhances solubility and dissolution rate of
• To develop a fixed dose combination for-
its crystalline form. In vivo study exhibited
mulation and evaluate its suitability as an
a 24-h pharmacokinetic profile determined
inhaled product by in vitro and in vivo
in BAL, lung tissue, and blood. The amount
studies
of drug per gram of lung tissue detected in
the lung was more than that detected in
• TFF apparatus
broncho-alveolar lavage (BAL). The rate
• Liquid nitrogen
of absorption of rapamycin from the TFF
• Salmeterol xinafoate (SX), mometasone
formulation into the systemic circulation is
furoate (MF), alpha-lactose monohydrate,
drastically greater than the crystalline phys-
mannitol, d-Trehalose anhydrous, glycine,
ical mixture since dissolution is the rate-
acetonitrile, and methanol
limiting step for the crystalline sample.
This resulted in high drug levels in BAL,
low absorption from lung fluid, and low
• NGI
systemic bioavailability.
• Dynamic vapor sorption (DVS)
• Amorphous rapamycin from the TFF pro-
Method
cess appeared to be eliminated from the
• Prepare a combination of SX and MF (ratio
lung more rapidly than the crystalline phys-
of 50:220 by weight) with or without other
ical mixture. This is possibly due to the
pharmaceutical excipients (lactose, manni-
hygroscopic property of the TFF formula-
tol, trehalose, or glycine) (1:1 molar ratio)
tion and the hydrophilic characteristics of
in a co-solvent mixture of tertiary butanol,
the lactose, which facilitates the water sorp-
1,4-dioxane, acetonitrile, and purified water
tion of the formulation. Hygroscopic TFF
(2:1:3:3, v/v)
processed rapamycin particles may deposit
• Apply the solution to a rotating cryogeni-
more in the conducting zone and be
cally cooled steel surface of the thin film
eliminated by the mucociliary clearance.
apparatus (90 3 C)
Amorphous rapamycin dissolved in the lin-
• Remove thin frozen films using a scraper
ing fluid may generate a supersaturation
• Collect and maintain the frozen samples in
environment, which precipitated and
a container filled with liquid nitrogen
increased susceptibility of clearance
• Lyophilize samples in a freeze dryer
mechanisms.
• Collect dry powders and store in a vacuum
desiccator at room temperature
• Prepare the micronized physical blends of
SX, MF, and excipients in the same ratio as
used in the TFF formulations for
comparison
Fig. 11.21 Sorption (-) 40

and desorption (---) BMP SXMF-Sorption
BMP SXMF-Desorption
isotherms of BMP 35 BMP SXMFLac-Sorption
formulations; (□) BMP BMP SXMFLac-Desorption
BMP SXMFMan-Sorption
SXMF, (⋄) BMP 30 BMP SXMFMan-Deportion
Change Mass (%)

SXMFLac, (Δ) BMP BMP SXMFGly-Sorption
SXMFMan, (x) BMP 25 BMP SXMFGly-Desorption
BMP SXMFTre-Sorption
SXMFGly and (Ο) BMP SXMFTre-Desoption
20
SXMFTre after the full
cycle of 0% RH to 90% 15
RH. (Reproduced from Liu
et al. (2015) with 10
5
0
0 10 20 30 40 50 60 70 80 90
Target (%)
Results • About 50% of the loaded dose from the

• The TFF product of neat SX crystallized excipient-free BMP was delivered in the
during the lyophilization process. respirable range.
• The Brittle matrix powder (BMP) of • This formulation was stable in an amor-
co-deposition was successfully formulated phous state at ambient conditions for up to
using the TFF technology. six months.
• DSC thermograms and PXRD patterns • After the single dose administration, the
agreed in the presence of the pharmacokinetic study in the lung exhibited
co-amorphous state of the BMP combina- that the concentration of drugs from the
tion formulations. BMP formulation was much higher than
• BMP fixed-dose combinations exhibited a that of the crystalline physical mixture
sponge-like matrix consisting of large (Fig. 11.22).
porous, homogeneous, and brittle particles.
In contrast, the micronized drug physical
The TFF formulation of neat SX was not able to
mixtures were not homogeneous.
maintain its amorphous morphology during
• The specific surface area of the BMP was
lyophilization as it had a low Tg. The absence
approximately 30-fold greater than that of
of melting peaks of drugs, the single Tg(s) and the
the micronized formulations.
re-crystallization peak in the BMP formulations
• Figure 11.21 showed the sorption and
confirmed the formation of the drug-drug homo-
desorption isotherms of the BMP fixed-
geneous amorphous phase. One drug was
dose combinations using DVS. The BPM
dissolved in the other drug (or the excipients).
of lactose- and trehalose-based
The co-amorphous solids improved solubility,
formulations were hygroscopic matrix. On
stability, and drug concentration. Hence, to retain
the other hand, the mannitol- and glycine-
these improvements, the formulations must with-
based formulations were non-hygroscopic.
stand a propensity for recrystallization. Stability
• Aerosolized BMP combination
study exhibited no changes in amorphous state of
formulations exhibited the co-deposition
the most BMP samples after six months at ambi-
of each drug and the uniformity of the
ent condition. The formulations with higher Tg
delivered dose.
(s) had lower molecular movement which can
reduce crystallization. Likewise, the hygroscopic
formulations are supposed to be less stable in
Fig. 11.22 In vivo pharmacokinetic study of SX and MF normalized with respect to the target delivered SX and MF
in (a) plasma, (b) lung tissue following dry powder insuf- dose of 450 μg/kg and 1980 μg/kg, respectively.
flation; (□) BMP SXMF formulation and (O) micronized (Reproduced from Liu et al. (2015) with permission from
SXMF formulation (n ¼ 3). The plot for lung tissue was Elsevier)
co-amorphous BMP formulations. FPFs of the of COVID-19. It is poorly water-soluble and

formulations with non-hygroscopic excipients or available in intravenous dosage forms, which is
without excipient were higher than the hygro- limited to use only for hospitalized patients. To
scopic powders. Owing to the stability and aero- direct administration to the primary site of infec-
dynamic performance, the formulation without tion, the lungs, the example below describes the
excipient was selected for the in vivo study. Phar- development of inhalable remdesivir powder
macokinetic data, the absorption rate and plasma prepared by TFF.
AUC0–24h of both drugs from TFF formulations
were significantly higher than those of the Method Capsule 15
micronized powders due to the higher dissolution
Dry Powder Inhalation Formulations of
and absorption. However, the enhancement of
Remdesivir Prepared by TFF
absorption resulted in a higher drug concentra-
Based on the study reported by Sahakijpijarn
tion. Hence, the adverse effect should be taken
(2020b, 2021)
into account.
Objective
• To develop remdesivir inhalable
formulations and invesitgate their in vivo
11.4.10 Remdesivir pharmacokinetic profile in rodent models
Remdesivir (REM) is a broad-spectrum antiviral • TFF apparatus
agent developed for the treatment of Ebola virus • Liquid nitrogen
and has currently been approved for the treatment
• Remdesivir (REM), GS-441524, lactose and REM/LEU (80/20) remained in an

monohydrate (LAC), mannitol (MAN), amorphous form after one month of storage
l-leucine (LEU), Sulfobutylether-beta- at 25 C/60%RH.
cyclodextrin (SBECD, Captisol®, CAP), • TFF REM formulations containing 80%
acetonitrile, trifluoracetic acid drug loading with an excipient (e.g.,
• Bench top tray lyophilizer Captisol mannitol, lactose, leucine)
• High-resistance Monodose RS00 dry pow- exhibited desirable aerosol performance
der inhalers (up to 82.71 2.54% FPF of recovered
• HPMC capsules (size 3) dose, 1.45 0.07 μm MMAD).
• Desiccator and desiccant • The aerosol performance of REM/LAC
• Next generation pharmaceutical impactor (80/20 w/w), REM/CAP (80/20 w/w) and
(NGI) REM/LEU (80/20 w/w) did not signifi-
Method cantly change after one month of storage
• Prepare REM with or without other phar- at 25 C/60% RH.
maceutical excipients (lactose, mannitol, • Amorphous REM prepared by TFF
leucine, Captisol®) (20%, 50%, 80%, and increased it solubility and subsequently
100% drug loading) in a co-solvent mixture improved the drug dissolution of
of acetonitrile and purified water (1:1). remdesivir in simulated lung fluid.
• Apply the solution to a rotating cryogeni- • An in vivo pharmacokinetic study in rats
cally cooled steel surface of the thin film showed that TFF REM/CAP (80/20 w/w)
apparatus (100 10 C). showed faster absorption of REM at initial
• Collect and maintain the frozen samples in time point (5 and 15 mins) than TFF
a container filled with liquid nitrogen. REM/LEU (80/20 w/w), resulting in higher
• Primary dry at 40 C below 100 mTorr AUC0–24 of REM and its metabolite,
for over 20 h. GS-441524, were detected in the plasma.
• Ramp the shelf temperature up from 40 to • REM/LEU (80/20 w/w) was slowly
25 C for over 20 h and hold at 25 C below absorbed into the systemic circulation,
100 mTorr over 20 h. allowing REM retained in the lungs, and
• Store bulk powder and capsules filled with subsequently metabolized to GS-441524
powder in a borosilicate glass vial. in the lungs. As a result, the levels of
• Dry samples in vials in a lyophilizer at REM and GS-441524 following dry pow-
25 C and 100 mTorr for 5 h, then der administration of TFF REM/LEU
backfilled with nitrogen gas before closing (80/20 w/w) were higher than those of
the rubber stopper inside the lyophilizer. TFF REM/CAP (80/20 w/w).
• Seal the vials with aluminum caps before • An in vivo pharmacokinetic study in
packing in an aluminum foil bag with a hamsters showed that TFF REM/LEU
desiccant. (80/20 w/w) required lower total drug expo-
Results sure in order to achieve a high effective
• TFF REM formulations exhibited a brittle concentration against SAR-CoV-2, as it
matrix consisting of highly porous exhibited a greater Cmax with shorter
particles. Tmax and lower AUC0–24 of
• DSC and XRD results indicated REM, lac- GS-441524 (Fig 11.23).
tose, Captisol® were amorphous in all • The plasma levels of REM and GS-441524
formulations, while mannitol and leucine were higher than reported EC50s of both
were crystalline. remdesivir and GS-441524 (in human epi-
• XRD diffractogram showed REM in thelial cells) over 20 h after dry powder
REM/LAC (80/20), REM/CAP (80/20) administration of both TFF REM/CAP
Fig. 11.23 Plasma

concentration-time profiles
of TFF REM/CAP 80/20 w/
w (F10) and TFF
REM/LEU 80/20 w/w (F13)
after a single inhalation
administration in rats; (a)
remdesivir; (b) GS-441524.
(Reproduced from
Sahakijpijarn et al. (2020)
Elsevier)
(80/20 w/w) and TFF REM/LEU (80/20 greatly variable plasma concentration. Due to the
w/w) (Fig. 11.24). narrow therapeutic window of CBZ, the drug
level needs to be routinely monitored to prevent
seizures and minimize the potential of adverse
This study is another example that highlighted
events. To address these problems, this example
the application of TFF to prepare dry powder for
describes the development of controlled release
inhalation that contained high drug loading of
formulations of CBZ prepared by TFF.
poorly water-soluble drug for the treatment of
respiratory diseases.
Method Capsule 16
Development of Modified-Release Amorphous
Solid Dispersions of CBZ by TFF
11.4.11 Carbamazepine
Based on the study reported by Li et al. (2021)
Objective
Carbamazepine (CBZ) is an antiepileptic drug
• To improve the bioavailability of carba-
that has been commercialized as an oral dosage
mazepine (CBZ) via modified-release
form. CBZ, a BCS class II drug, exhibits erratic
oral absorption, suboptimal bioavailability and a
Fig. 11.24 Plasma

concentration-time profiles
of TFF REM/CAP 80/20
w/w and TFF REM/LEU
80/20 w/w after a single dry
powder insufflation in
hamsters; (a) Plasma REM
level; (b) Plasma
GS-441524 level. Dash line
and dot line represent EC50
of remdesivir and
GS-441524 in human
epithelial cells.
(Reproduced with
(Sahakijpijarn et al. 2021))
amorphous solid dispersions prepared by Methods

TFF • TFF was used to prepare CBZ formulations
Equipment and Material with immediate-, delayed-, and controlled-
• TFF apparatus release properties. CBZ was combined with
• Liquid nitrogen HPMC E3 (hydrophilic), L100-55 (enteric),
• Cabamazepine (CBZ), METHOCEL* E3 and cellulose acetate (CA, lipophilic),
premium LV hydroxypropyl methylcellu- defined as CBZ-ir-ASD, CBZ-dr-ASD,
lose (HPMC E3), Eudragit® L100-55 and CBZ-cr-ASD, respectively.
(L100-55), cellulose acetate (CA), silicified • CBZ and excipients (HPMC E3, L100-55,
microcrystalline cellulose (SMCC) magne- CA) in a co-solvent mixture of 1,4-dioxane
sium stearate (MgSt), and sodium starch and purified water or pure 1,4-dioxane.
glycolate, 1,4-dioxane, acetonitrile, metha- • Apply the solution to a rotating cryogeni-
nol, and trifluoroacetic acid cally cooled steel surface of the thin film
• Bench top tray lyophilizer apparatus (50 C).
• Hydraulic press
Fig. 11.25 In vitro

dissolution profiles of three
formulations of CBZ-mr-
AS: CBZ-ir-ASD
(immediate release), CBZ-
dr-ASD (delayed release),
and CBZ-cr-ASD
(controlled release) (n ¼ 3).
(Reprinted from Li et al.
(2020) with permission
from MDPI)
• Dry samples in a lyophilizer. Ramp the formulations were more over 90% within
shelf temperature up from 50 to 25 C 15 mins after the pH shift from 1.2 to
over 72 h. 6.8 (Fig. 11.25).
• The CBZ-TFF powder was then mixed with • Single-dose 24 h pharmacokinetic studies
SMCC90, magnesium stearate, and sodium in rats showed that all CBZ formulations
starch glycolate. exhibited higher plasma concentration,
• A dry granulation method was used to pre- Cmax and AUC0-t of CBZ and its metabo-
pare capsules. The mixture was compressed lite, CBZ-E, compared to crude
into slug tablets with a hydraulic press. CBZ (Fig.11.26).
Then the tablets were ground into granules • The delayed release formulation had a lon-
and sieved with 40-mesh sieve before fill- ger Tmax than crude CBZ, indicating a
ing in a capsule. slower absorption and elimination
Results (Fig.11.26).
• SEM demonstrated that TFF produced
microparticle agglomerates that are
connected to form a sponge-like porous This study is another example that highlighted
matrix structure, and no crystal CBZ was the application of TFF to prepare modified-
observed. release amorphous solid dispersion of poorly
• XRD and DSC showed that all three water-soluble drug, which can improve its oral
formulations of CBZ-ASD (immediate-, bioavailability.
delayed, and controlled-release
formulations) were amorphous.
• The dissolution testing under sink
conditions demonstrated that over 90% of
11.4.12 Aluminum Salt-Adjuvant
the drug in CBX-ir-ASD (immediate Vaccines
release) was dissolved within 120 mins in
Aluminum salts such as aluminum hydroxide and
an acidic medium (pH 1.2), while less than
50% of drug in the controlled and delayed- aluminum phosphate are insoluble substances
that have been used as human vaccine adjuvants.
release formulation were released in an
A major challenge in developing alum-based
acidic medium. However, when the pH of
vaccines is the particle aggregation during freez-
dissolution medium was changed to 6.8, the
ing, which subsequently affects the immunoge-
amount of drug accumulated release from
nicity of the vaccines. The examples below
controlled and delayed-release
Fig. 11.26 Mean plasma-drug concentrations of CBZ (a) and CBZ-E (b) in male Sprague–Dawley rats after single-dose
oral administration of CBZ-ir-ASD capsules (immediate release), CBZ-dr-ASD capsules (delayed release), and CBZ-cr-
ASD capsules (controlled release) (n ¼ 6). (Reprinted with permission from MDPI (Li et al. 2020))
showed the use of TFF to develop stable alum- • Collect and maintain the frozen samples in
based vaccines as a dry powder. a container filled with liquid nitrogen.
• Dry samples in a lyophilizer. Ramp the
Method Capsule 17 shelf temperature up from 40 to 26 C
over 72 h below 200 mTorr.
Development of Vaccines Containing Alumi-
Results
num Salts Prepared by TFF
• The microscopic images demonstrated no
Based on the study reported by Li et al. (2015)
significant aggregation was observed after
and Thakkar et al. (2017)
reconstitution of OVA-absorbed aluminum
Objective
hydroxide dry powder prepared by TFF,
• To investigate the feasibility of TFF to pre-
while aggregation was observed in
pare vaccines that are adjuvants with alumi-
OVA-absorbed aluminum hydroxide sus-
num salts
pension after shelf-freezing at 80 or
20 C (Fig. 11.27).
• TFF apparatus
• No significant difference between particle
• Liquid nitrogen
size of the reconstituted OVA-absorbed
• Aluminum hydroxide, Aluminum phos-
aluminum hydroxide and the freshly
phate, ovalbumin (OVA), tetanus toxoid
prepared OVA-absorbed aluminum
vaccine
hydroxide suspensions (9.7 2.5
μm vs. 9.4 1.7 μm, respectively),
Method
indicating TFF was able to prepare a dry
• OVA was used as a model antigen to absorb
powder without significant change in
different aluminum-containing compounds
particle size.
including dried aluminum hydroxide gel,
• Low amount of trehalose (2%w/v) in the
2% hydrogel, and aluminum phosphate.
formulation was able to stabilize the
• Add 2%w/v of trehalose in the suspensions.
OVA-absorbed aluminum hydroxide and
• Apply the suspensions to a rotating cryo-
produce dry powder by TFF without caus-
genically cooled steel surface of the thin
ing particle aggregation.
film apparatus.
Fig. 11.27 OVA-absorbed aluminum hydroxide pre- combined with 2% trehalose before (b), after TFF and
pared by TFF; comparison of particle size distribution reconstitution (c), after shelf-freezing at 20 C (d), and
before and after TFF and reconstitution (a); microscopic after shelf-freezing at 20 C (e). (Reproduced from Li
images of OVA- OVA-absorbed aluminum hydroxide et al. (2015) with permission from Elsevier)
• The immunogenicity test in mice showed were immunized with OVA-absorbed alu-
that no significant difference in the serum minum hydroxide following TFF and
anti-OVA IgG level between mice that reconstitution and mice that were
immunized with freshly prepared changed after TFF and reconstitution

OVA-absorbed aluminum hydroxide. (Fig. 11.28).
• Tetanus toxin binding inhibition test • The stability study showed that the particle
showed that the activity of tetanus toxoid size distribution of reconstituted TFF
(TT) vaccine did not significantly change OVA-absorbed aluminum hydroxide dry
after the TFF and reconstitution. powder containing 2% w/v trehalose did
• The in vivo toxin neutralization assay not significantly change after one month
demonstrated that the immunogenicity of of storage at room temperature, 30 C, and
the TT vaccine was not significantly 40 C.
• The particle aggregation was slightly
detected after three months of storage only
at 40 C, however; the noticeable amounts
of particle aggregates were detected after
six months of storage at room temperature,
30 C, and 40 C.
• No significant difference in the anti-OVA
IgG levels between mice that were
immunized with OVA/aluminum hydrox-
ide vaccine reconstituted from dry powder
that was stored in various temperatures for
one month or three months (Fig. 11.29).
This example highlighted another application

of TFF to prepare a stable dry powder of vaccines
containing aluminum salts. The use of TFF with
the combination of an optimal amount of treha-
Fig. 11.28 Anti-tetanus toxin IgG level in mice serum lose can prevent particle aggregation during prep-
after immunization with freshly prepared TT vaccine and aration, which can maintain the activity and
reconstituted suspension of TFF TT vaccine powder. immunogenicity of vaccines. This would be a
(Reproduced from Li et al. (2015) with permission from
promising approach to minimize the limitation
Elsevier)
Fig. 11.29 Anti-OVA toxin IgG level in mice serum after immunization with TFF OVA/aluminum hydroxide vaccine
powder that was stored at room temperature, 30 C and 40 C for one month (a), or for three months (b), and then
reconstituted. (Reproduced from Thakkar et al. (2017) with permission from Taylor & Francis)
of storage conditions and the requirement of cold- • Apply the suspensions to a rotating cryo-
chain storage of vaccines (Figs. 11.27 and 11.28). genically cooled steel surface of the thin
film apparatus.
Method Capsule 18 • Collect and maintain the frozen samples in
a container filled with liquid nitrogen.
Nasal Delivery of Aluminum Salt-Adjuvant
• Dry samples in a lyophilizer. Gradually
Dry Powder Vaccine Prepared by TFF
ramp the shelf temperature up from 40
Based on the study reported by Thakkar et al.
to 26 C over 72 h below 200 mTorr.
(2018)
Results
Objective
• SEM/EDX showed a homogenous distribu-
• To investigate the feasibility of
tion of Al, O, Na, and Cl, indicating vaccine
administrating aluminum salt-adjuvant dry
was uniformly distributed in the dry
powder vaccines prepared by TFF via the
powder.
nasal route
• The particle size analysis by laser diffrac-
tion (Malvern’s Spraytec) demonstrated
• TFF apparatus
that the median diameter of dry powder
• Liquid nitrogen
was 12.55 4.69 μm.
• Aluminum hydroxide (Alhydrogel®; ~10
• In vitro nasal deposition testing at 15 L/min
mg/mL aluminum), ovalbumin, trehalose
showed that 90% of the powder was depos-
ited in the nasal cavity, and only about 10%
• A nasal dry powder delivery device that is
of recovered dose was deposited in the
composed of a housing reservoir (e.g., the
post-nasal fraction (Fig. 11.30).
square-shaped base of an oral gavage nee-
• The vaccine powder was intranasally dosed
dle) and a syringe plunger
at 20 μg/rat on day 0, 14, and 28 days. The
Method
immunogenicity test which was analyzed
• Mix 10 ml of 10 mg/mL aluminum
28 days following the third immunization
(Alhydrogel®) with 10 ml of mg/mL of
showed that intranasal administration
OVA solution in 0.9% normal saline.
induced the anti-OVA IgG in rat serum
• Add 2% w/v of trehalose.
Fig. 11.30 (a) A diagram of nasal cast used that is consisted of the vestibule (V) and nasal valve area make up the
anterior region, the upper (U), middle (M), and lower (L) turbinate regions of the posterior nasal cavity, and posterior the
nasopharynx (N); (b) % deposition of the OVA-Alhydrogel® dry powder vaccine in nasal casts operated at 15 L/min.
(Reproduced from Thakkar et al. (2018) with permission from Elsevier)
Fig. 11.31 Serum anti-OVA IgG titers, mucosal IgA titers, and IgA titer in BAL in rats immunized with
OVA-Alhydrogel® vaccine by intranasal or subcutaneous administration. (Reproduced from Thakkar et al. (2018)
with permission from Elsevier)
and IgA level in rat nose and BAL samples infection, this study aimed to develop
(Fig. 11.31). niclosamide dry powder for inhalation by TFF.
• No significant difference in the aluminum
levels in rat brain tissues among four Method Capsule 19
groups of rats (e.g., saline, intranasal pow-
Development of Niclosamide Dry Powder for
der, intranasal liquid, and subcutaneous liq-
Inhalation by TFF
uid) after four weeks of the third
Based on the study reported by Jara et al.
immunization, indicating intranasal admin-
(2021)
istration has less potential to affect the level
Objective
of aluminum in the brain as compared to
• To develop niclosamide dry powder for
subcutaneous administration at the same
inhalation formulations for the treatment
dose.
of COVID-19
This study highlighted the feasibility to deliver • TFF apparatus
an aluminum salt-adjuvant dry powder vaccine • Liquid nitrogen
prepared by TFF via intranasal administration. • Niclosamide, mannitol (MAN), l-leucine
The optimal amount of TFF vaccine powder can (LEU), t-butanol, acetonitrile, formic acid
be delivered and retained in the nose, which can • Bench top tray lyophilizer
induce the specific mucosal and systemic immune • High-resistance Monodose RS00 dry pow-
responses after immunizations. der inhalers
• HPMC capsules (size 3)
• Next generation pharmaceutical impactor
11.4.13 Niclosamide
(NGI)
Method
Niclosamide is an antihelmintic drug with broad
• Prepare niclosamide with mannitol and leu-
spectrum antiviral activity. Recent studies
cine (22/73/5 w/w) in a co-solvent mixture
demonstrated that it has a specific pharmacologi-
of t-butanol and purified water (4:1)
cal effect against SARS-CoV-2 infection. It is a
• Apply the solution to a rotating cryogeni-
poorly water-soluble drug with poor and erratic
cally cooled steel surface of the thin film
oral bioavailability. To improve the bioavability
apparatus (120 10 C)
and efficacy of the treatment of pulmonary
Fig. 11.32 Histological examination of lung tissue; (a) normal alveolar region (magnification of 100x); (b) bronchial
inflammation of a 1 mm airway section in Rat 4 after three days of dosing niclosamide inhalation powder (200 μg of
niclosamide per animal). (Reproduced from Jara et al. (2021) with permission from Elsevier)
• Collect and maintain the frozen samples in inflammation and a few to moderate
a container filled with liquid nitrogen eosinophils in the rat lung
• Primary dry at 40 C below 100 mTorr tissue (Fig. 11.32).
over 20 h • The in vivo pharmacokinetics study on
• Ramp the shelf temperature up from 40 to Syrian hamsters showed that a single dose
25 C for over 20 h and hold at 25 C below pulmonary administration of niclosamide at
100 mTorr over 20 h both 145 μg/kg and 290 μg/kg was able to
Results achieve lung concentrations above the
• The in vitro aerodynamic testing required IC50 and IC90 levels for at least
demonstrated that the formulation 24 h (Fig. 11.33).
containing 73% w/w mannitol and 5%
w/w leucine exhibited high aerosol perfor-
This example also highlighted the application
mance (1.11 μm MMAD, 76.6% FPF of
of TFF to develop poorly water-soluble drug as
recovered dose).
inhalable dry powder in order to avoid the limited
• SEM showed TFF produced nanostructured
oral bioavailability and enhance the efficacy of
brittle matrix of niclosamide and
the treatment of respiratory diseases.
excipients.
• DSC and XRD demonstrated niclosamide
and mannitol remained crystalline after
TFF process. 11.5 Summary
• In vivo pharmacokinetic study in rats
showed that a multi-dose administration of Significant advances have been made in the past
inhaled niclosamide is slowly cleared from few decades toward understanding and tackling
the lungs as the amount of drug in the lungs the poor bioavailability issues associated with
and plasma were still detectable 24 h fol- poorly water-soluble drugs. Among various
lowing administration of the last dose. novel methods developed to improve dissolution
• A multi-dose administration of inhaled properties for poorly water-soluble drugs, crea-
niclosamide was safe with mild pulmonary tion of amorphous pharmaceutical materials holds
toxicity as the histopathological examina- a lot of promise/demonstrated macroscopic
tion demonstrated mild signs of
Fig. 11.33 Lung concentration-time profile of TFF for SARS CoV 2 is 0.09156 μg/g of wet tissue.
niclosamide after a single dry powder insufflation in ham- (Reproduced from Jara et al. (2021) with permission
ster at 145 μg/kg and 290 μg/kg. Assuming that the density from Cold Spring Harbor Laboratory)
of wet lung tissue is 1 g/mL. Estimated niclosamide IC50
performance advantages versus crystalline TFF process is more cost effective and scalable
counterpart. for manufacturing production.
Cryogenic particle engineering technologies By selecting proper excipient(s) and drug-to-
are “bottom-up” precipitation processes to gener- excipient ratio in formulation and appropriate
ate amorphous nanostructured aggregates with packaging and storage, cryogenic technologies
significantly enlarged surface area, higher disso- can be used to manufacture stable amorphous
lution rates, and supersaturation, by rapidly drug dosage forms for diverse routes of adminis-
inducing nucleation followed by particle growth tration, such as pulmonary, parenteral, and oral,
arrest through stabilization via polymers and with potentially reduced drug dose and side
solidification of the solvent. The improved disso- effects.
lution properties of poorly water-soluble drugs
were achieved by (1) reducing particle size,
thereby increasing surface area; (2) creating
References
amorphous morphology; and (3) intimately
mixing the drug with hydrophilic excipients. Adali MB, Barresi AA, Boccardo G, Pisano R. Spray
Moreover, without introducing mechanical force freeze-drying as a solution to continuous
and heat, cryogenic processes are specifically manufacturing of pharmaceutical products in bulk.
suitable for thermal-labile drugs. Compared to PRO. 2020;8(6):709.
Adeli E. The use of spray freeze drying for dissolution and
SFD and SFL, TFF in particular has been shown oral bioavailability improvement of Azithromycin.
to create stable amorphous compositions of Powder Technol. 2017;319:323–31.
poorly water-soluble drugs with significantly Amidon GL, Lennernas H, et al. A theoretical basis for a
improved bioavailability and protein biopharmaceutic drug classification: the correlation of
in vitro drug product dissolution and in vivo bioavail-
nanoparticles with high activity. Additionally, ability. Pharm Res. 1995;12(3):413–20.
Anandharamakrishnan C, Rielly CD, et al. Spray-freeze- rapamycin dry powder via pulmonary administration in
drying of whey proteins at sub-atmospheric pressures. rats. Eur J Pharm Biopharm. 2014;88(1):136–47.
Dairy Sci Technol. 2010;90(2–3):321–34. Clas SD, Dalton CR, et al. Differential scanning calorime-
Badawya SIF, Ghorabb MM, et al. Characterization and try: applications in drug development. Pharm Sci
bioavailability of danazol-hydroxypropyl Technolo Today. 1999;2(8):311–20.
β-cyclodextrin coprecipitates. Int J Pharm. 1996;128 Conventional, U. P. The United States pharmacopoeia.
(1–2):45–54. Rockville: The United States Pharmacopoeia Conven-
Bakatselou V, Oppenheim RC, et al. Solubilization and tional; 2000. p. 8.
wetting effects of bile salts on the dissolution of Costantino HR, Curley JG, et al. Water sorption behaviour
steroids. Pharm Res. 1991;8(12):1461–9. of lyophilised protein-sugar systems and implications
Barnes TJ, Kempson IM, Prestidge CA. Surface analysis for solid-state interactions. Int J Pharm. 1998;166:211–
for compositional, chemical and structural imaging in 21.
pharmaceutics with mass spectrometry: a ToF-SIMS Costantino HR, Firouzabadian L, et al. Protein spray-
perspective. Int J Pharm. 2011;417(1):61–9. freeze drying. Effect of atomization conditions on par-
Barron MK, Young TJ, et al. Investigation of processing ticle size and stability. Pharm Res. 2000;17(11):
parameters of spray freezing into liquid to prepare 1374–83.
polyethylene glycol polymeric particles for drug deliv- Costantino HR, Firouzabadian L, et al. Protein spray
ery. AAPS PharmSciTech. 2003;4:1–13. freeze drying. 2. Effect of formulation variables on
Baylor NW, Egan W, Richman P. Aluminum salts in particle size and stability. J Pharm Sci. 2002;91(2):
vaccines–US perspective. Vaccine. 2002;20(Suppl 3): 388–95.
S18–23. https://doi.org/10.1016/s0264-410x(02) Costantino HR, Johnson OL, et al. Relationship between
00166-4. encapsulated drug particle size and initial release of
Beinborn NA, Lirola HL, et al. Effect of process variables recombinant human growth hormone from biodegrad-
on morphology and aerodynamic properties of able microspheres. J Pharm Sci. 2004;93(10):2624–34.
voriconazole formulations produced by thin film freez- Courrier HM, Butz N, et al. Pulmonary drug delivery
ing. Int J Pharm. 2012;429(1):46–57. systems: recent developments and prospects. Crit Rev
Benson SW, Ellis DA. Surface areas of proteins; surface Ther Drug Carrier Syst. 2002;19(4–5):425–98.
areas and heats of absorption. J Am Chem Soc. Davies NM, Feddah MR. A novel method for assessing
1948;70(11):3563–9. dissolution of aerosol inhaler products. Int J Pharm.
Bertilsson L, Tomson T. Clinical pharmacokinetics and 2003;255(1–2):175–87.
pharmacological effects of carbamazepine and carba- De Waard H, Hinrichs WL, Frijlink HW. (2008). A novel
mazepine-10,11-epoxide. An update. Clin bottom–up process to produce drug nanocrystals: con-
Pharmacokinet. 1986;11(3):177–98. trolled crystallization during freeze-drying. Journal of
Betageri GV, Makarla KR. Enhancement of dissolution of controlled release. 128(2):179–83.
glyburide by solid dispersion and lyophilization Derle D, Patel J, et al. Particle engineering techniques to
techniques. Int J Pharm. 1995;126(1–2):155–60. enhance dissolution of poorly water soluble drugs. Int J
Boeh-Ocansey O. A study of the freeze drying of some Curr Pharm Res. 2010;2(1):10–5.
liquid foods in vacuo and at atmospheric pressure. Dry DiNunzio JC, Miller DA, et al. Amorphous compositions
Technol. 1983;2:389–405. using concentration enhancing polymers for improved
Borm P, Klaessig FC, et al. Research strategies for safety bioavailability of itraconazole. Mol Pharm. 2008;5(6):
evaluation of nanomaterials, part V: role of dissolution 968–80.
in biological fate and effects of nanoscale particles. Dong Y-D, Boyd BJ. Applications of X-ray scattering in
Toxicol Sci. 2006;90(1):23–32. pharmaceutical science. Int J Pharm. 2011;417(1):
Brown SC, Claybourn M, et al. Optimizing raman spec- 101–11.
troscopy to quantify polymorphic forms of a drug Du J, Li X, et al. Nanosuspensions of poorly water-soluble
molecule. Am Pharm Rev. 2007;10(58):60–7. drugs prepared by bottom-up technologies. Int J
Brunauer S, Emmett P, et al. Adsorption of gases in Pharm. 2015;495(2):738–49.
multimolecular layer. J Am Chem Soc. 1938;60:309– Engstrom JD, Simpson DT, et al. Stable high surface area
19. lactate dehydrogenase particles produced by spray
Burkart GJ, Smaldone GC, et al. Lung deposition and freezing into liquid nitrogen. Eur J Pharm Biopharm.
pharmacokinetics of cyclosporine after aerosolization 2007a;65(2):163–74.
in lung transplant patients. Pharm Res. 2003;20(2): Engstrom JD, Simpson DT, et al. Morphology of protein
252–6. particles produced by spray freezing of concentrated
Carli F, Motta A. Particle size and surface area solutions. Eur J Pharm Biopharm. 2007b;65(2):
distributions of pharmaceutical powders by microcom- 149–62.
puterized mercury porosimetry. J Pharm Sci. 1984;73 Engstrom JD, Lai ES, et al. Formation of stable submicron
(2):197–203. protein particles by thin film freezing. Pharm Res.
Carvalho SR, Watts AB, et al. Characterization and phar- 2008;25(6):1334–46.
macokinetic analysis of crystalline versus amorphous
Engstrom J, Tam J et al (2009) Templated open flocs of He W, Lu Y, et al. Food proteins as novel nanosuspension
nanorods for enhanced pulmonary delivery with stabilizers for poorly water-soluble drugs. Int J Pharm.
pressurized metered dose inhalers. Pharm Res 26 2013;441(1):269–78.
(1):101–117 Head T, Rydzak J. Chemometric models using diamond
Eriksson JHC, Hinrichs WLJ, et al. Investigations into the attenuated total reflectance IR and Raman spectros-
stabilization of drugs by sugar glasses: III. The influ- copy to characterize and quantitate polymorphs in
ence of various high-pH buffers. Pharm Res. 2003;20: pharmaceuticals. Am Pharm Rev. 2003;6:78–84.
1437–43. Hickey AJ, Mansour HM, et al. Physical characterization
Esclusa-Diaz MT, Guimaraens-Mendez M, et al. Charac- of component particles included in dry powder
terization and in vitro dissolution behaviour of inhalers. I Strategy review and static characteristics. J
ketoconazole/β- and 2-hydroxypropyl-β-cyclodextrin Pharm Sci. 2007a;96(5):1282–301.
inclusion compounds. Int J Pharm. 1996;143:203–10. Hickey MB, Peterson ML, et al. Performance comparison
Feeley JC, York P, et al. Determination of surface of a co-crystal of carbamazepine with marketed prod-
properties and flow characteristics of salbutamol uct. Eur J Pharm Biopharm. 2007b;67(1):112–9.
sulphate, before and after micronisation. Int J Pharm. Hildebrand JH, Scott RL. Solubility of nonelectrolytes.
1998;172:89–96. New York: Reinhold; 1950, 11–13, 47, 160, 175–197
Fong SYK, Ibisogly A, et al. Solubility enhancement of Hintz RJ, Johnson KC. The effect of particle-size distribu-
BCS Class II drug by solid phospholipid dispersions: tion on dissolution rate and oral absorption. Int J
spray drying versus freeze-drying. Int J Pharm. Pharm. 1989;51(1):9–17.
2015;496:382–91. Hu J, Rogers TL, et al. Improvement of dissolution rates of
Forster A, Hempenstall J, et al. Characterization of glass poorly water soluble APIs using novel spray freezing
solutions of poorly water-soluble drugs produced by into liquid technology. Pharm Res. 2002;19(9):
melt extrusion with hydrophilic amorphous polymers. 1278–84.
J Pharm Pharmacol. 2001;53(3):303–15. Hu J, Johnston KP, et al. Spray freezing into liquid (SFL)
Franks F. Freeze-drying: from empiricism to predictabil- particle engineering technology to enhance dissolution
ity. The significance of glass transitions. Dev Biol of poorly water soluble drugs: organic solvent versus
Stand. 1992;74:9–18, discussion 19 organic/aqueous co-solvent systems. Eur J Pharm Sci.
Fukai J, Ozaki T, et al. Numerical simulation of liquid 2003;20(3):295–303.
droplet solidification on substrates. J Chem Eng Jpn. Hu J, Johnston K, et al. Rapid release tablet formation of
2000;33:630–7. micronized danazol powder produced by spray freez-
Gilkey JC, Staehelin LA. Advances in ultrarapid freezing ing into liquid (SFL). J Drug Deliv Sci Technol.
for the preservation of cellular ultrastructure. J Electron 2004a;14(4):305–11.
Microsc Tech. 1986;3:177–210. Hu J, Johnston KP, et al. Nanoparticle engineering pro-
Gosselin PM, Thibert R, et al. Polymorphic properties of cesses for enhancing the dissolution rates of poorly
micronized carbamazepine produced by RESS. Int J water soluble drugs. Drug Dev Ind Pharm. 2004b;30
Pharm. 2003;252(1–2):225–33. (3):233–45.
Grant DJW, Brittian HG. Physical characterisation of Hu J, Johnston KP, et al. Rapid dissolving high potency
pharmaceutical solids. New York: Marcel danazol powders produced by spray freezing into liq-
Dekker; 1995. uid process. Int J Pharm. 2004c;271(1–2):145–54.
Grant N, Zhang H. Poorly water-soluble drug Iacono AT, Smaldone GC, et al. Dose-related reversal of
nanoparticles via an emulsion-freeze-drying approach. acute lung rejection by aerosolized cyclosporine. Am J
J Colloid Interface Sci. 2011;356(2):573–8. Respir Crit Care Med. 1997;155(5):1690–8.
Hancock BC. Disordered drug delivery: destiny, dynamics Ishwarya SP, Anandharamakrishnan C, et al. Spray-
and the Deborah number. J Pharm Pharmacol. 2002;54 freeze-drying: a novel process for the drying of foods
(6):737–46. and bioproducts. Trends Food Sci Tech. 2015;41(2):
Hancock BC, Parks M. What is the true solubility advan- 161–81.
tage for amorphous pharmaceuticals? Pharm Res. Jalili N, Laxminarayana K. A review of atomic force
2000;17(4):397–404. microscopy imaging systems: application to molecular
Hancock BC, Shamblin SL, et al. Molecular mobility of metrology and biological sciences. Mechatronics.
amorphous pharmaceutical solids below their glass 2004;14(8):907–45.
transition temperatures. Pharm Res. 1995;12(6): Jara MO, Warnken ZN, Sahakijpijarn S, Moon C, Maier
799–806. EY, Christensen DJ, et al. Niclosamide inhalation pow-
He X, Pei L, Tong HHY, Zheng Y. Comparison of spray der made by thin-film freezing: pharmacokinetic and
freeze drying and the solvent evaporation method for toxicology studies in rats and hamsters. bioRxiv,
preparing solid dispersions of baicalein with Pluronic 2021.2001.2026.428293. 2021; https://doi.org/10.
F68 to improve dissolution and oral bioavailability. 1101/2021.01.26.428293.
AAPS PharmSciTech. 2011;12(1):104–13. Jermain SV, Miller D, Spangenberg A, Lu X, Moon C,
Su Y, Williams RO 3rd. Homogeneity of amorphous
solid dispersions – an example with KinetiSol. Drug
Dev Ind Pharm. 2019;45(5):724–35. https://doi.org/10. Maa YF, Zhao L, Payne LG, Chen D. Stabilization of
1080/03639045.2019.1569037. alum-adjuvanted vaccine dry powder formulations:
Jiang S, Nail SL. Effect of process conditions on recovery mechanism and application. J Pharm Sci. 2003;92(2):
of protein activity after freezing and freeze-drying. Eur 319–32.
J Pharm Biopharm. 1998;45(3):249–57. Maa YF, Ameri M, et al. Influenza vaccine powder formu-
Johnson KA. Preparation of peptide and protein powders lation development: spray-freeze-drying and stability
for inhalation. Adv Drug Deliv Rev. 1997;26(1):3–15. evaluation. J Pharm Sci. 2004;93(7):1912–23.
Kalepu S, Nekkanti V. Insoluble drug delivery strategies: Martin A, Swarbrick J, et al. Physical pharmacy: physical
review of recent advances and business prospects. Acta chemical principles in the pharmaceutical sciences.
Pharm Sin B. 2015;5(5):442–53. Philadelphia: Lippincott Williams & Wilkins; 1993,
Kapsi SG, Ayres JW. Processing factors in development of 125–142, 212–250, 329–334
solid solution formulation of itraconazole for enhance- Matteucci ME, Brettmann BK, et al. Design of potent
ment of drug dissolution and bioavailability. Int J amorphous drug nanoparticles for rapid generation of
Pharm. 2001;229(1–2):193–203. highly supersaturated media. Mol Pharm. 2007;4(5):
Khougaz K, Clas SD. Crystallization inhibition in solid 782–93.
dispersions of MK-0591 and poly(vinylpyrrolidone) Mawson S, Yates MZ, et al. Stabilized polymer
polymers. J Pharm Sci. 2000;89(10):1325–34. microparticles by precipitation with a compressed
Lamba D. Wide-Angle X-Ray Scattering (WAXS). In: fluid antisolvent.2. Poly(propylene oxide)- and poly
Drioli E, Giorno L, editors. Encyclopedia of (butylene oxide)-based copolymers. Langmuir.
membranes. Berlin/Heidelberg: Springer; 2016. 1997;13(6):1519–28.
p. 2040–2. Meryman H. Sublimation freeze drying without vacuum.
Lang B, Chow KT, et al. Thin film freezing-template Science. 1959;130:628–9.
emulsion of itraconazole to improve the dissolution Miller DA, McConville JT, et al. Stabilized HME compo-
properties of poorly water-soluble drugs. J Drug sition with small drug particles. USPTO. USA, Board
Deliv Sci Tec. 2014;24(2):205–11. of the regents, The University of Texas at Austin
Lauer ME, Grassmann O, et al. Atomic force microscopy- System. US 2008/0274194 Al; 2008.
based screening of drug-excipient miscibility and sta- Moneghini M, Kikic I, et al. Processing of carbamazepine-
bility of solid dispersions. Pharm Res. 2011;28(3): PEG 4000 solid dispersions with supercritical carbon
572–84. dioxide: preparation, characterisation, and in vitro dis-
Leleux J, Williams RO. Recent advancements in mechani- solution. Int J Pharm. 2001;222(1):129–38.
cal reduction methods: particulate systems. Drug Dev Moon C, Watts AB, Williams RO III. Thin film freezing
Ind Pharm. 2014;40(3):289–300. for production of an amorphous mannitol/lysozyme
Lerk CF, Schoonen AJ, et al. Contact angles and wetting formulation for inhalation: a comparison of freezing
of pharmaceutical powders. J Pharm Sci. 1976;65(6): techniques. Respir Drug Deliv. 2016;3:611–6.
843–7. Moon C, Watts AB, Lu X, Su Y, Williams RO 3rd.
Leuner C, Dressman J. Improving drug solubility for oral Enhanced Aerosolization of high potency
delivery using solid dispersions. Eur J Pharm nanoaggregates of Voriconazole by dry powder inha-
Biopharm. 2000;50(1):47–60. lation. Mol Pharm. 2019;16(5):1799–812.
Li X, Thakkar SG, Ruwona TB, Williams RO 3rd, Cui Muller RH, Jacobs C, et al. Nanosuspensions as particulate
Z. A method of lyophilizing vaccines containing alu- drug formulations in therapy rationale for development
minum salts into a dry powder without causing particle and what we can expect for the future. Adv Drug Deliv
aggregation or decreasing the immunogenicity follow- Rev. 2001;47(1):3–19.
ing reconstitution. J Control Release. 2015a;204:38– Mumenthalera M, Leuenberger H. Atmospheric spray-
50. freeze drying: a suitable alternative in freeze-drying
Li M, Xu W, Su Y. Solid-state NMR spectroscopy in technology. Int J Pharm. 1991;72(2):97–110.
pharmaceutical sciences. TrAC Trends Anal Chem. Nagapudi K, Jona J. Amorphous active pharmaceutical
2021;135:116152. ingredients in preclinical studies: preparation, charac-
Liu R. Water-insoluble drug formulation. Englewood: terization, and formulation. Curr Bioact Compd.
Interpharm Press; 2000. 2008;4:213–24.
Liu S, Watts AB, et al. Formulation of a novel fixed dose National Library of Medicine (U.S.). Pharmacokinetic
combination of salmeterol xinafoate and mometasone profile of voriconazole inhalation powder in adult
furoate for inhaled drug delivery. Eur J Pharm subjects with asthma. Identifier NCT04576325.
Biopharm. 2015;96:132–42. 2020.; https://clinicaltrials.gov/ct2/show/
Maa YF, Prestrelski SJ. Biopharmaceutical powders: par- NCT04576325
ticle formation and formulation considerations. Curr Newman AW, Byrn SR. Solid-state analysis of the active
Pharm Biotechnol. 2000;1(3):283–302. pharmaceutical ingredient in drug products. Drug
Maa YF, Nguyen PA, et al. Protein inhalation powders: Discov Today. 2003;8(19):898–905.
spray drying vs spray freeze drying. Pharm Res. Nguyen XC, Herberger JD, Burke PA. Protein powders for
1999;16(2):249–54. encapsulation: a comparison of spray-freeze drying
and spray drying of Darbepoetin Alfa. Pharm Res. co-grinding of piroxicam with carrier polymers. Drug
2004;21(3):507–14. Dev Ind Pharm. 2015;42:1–11.
Ni N, Tesconi M, et al. Use of pure t-butanol as a solvent Pisano R, Arsiccio A, Capozzi LC, Trout BL. Achieving
for freeze-drying: a case study. Int J Pharm. 2001;226 continuous manufacturing in lyophilization:
(1–2):39–46. technologies and approaches. Eur J Pharm Biopharm.
Nie H, Su Y, Zhang M, Song Y, Leone A, Taylor LS, 2019;142:265–79.
Marsac PJ, Li T, Byrn SR. Solid-state spectroscopic Poirier JM, Hardy S, et al. Plasma itraconazole
investigation of molecular interactions between concentrations in patients with neutropenia:
Clofazimine and Hypromellose Phthalate in amor- advantages of a divided daily dosage regimen. Ther
phous solid dispersions. Mol Pharm. 2016;13(11): Drug Monit. 1997;19(5):525–9.
3964–75. Prestidge CA, Barnes TJ, Skinner W. Time-of-flight
Niwa T, Danjo K (2013) Design of self-dispersible dry secondary-ion mass spectrometry for the surface char-
nanosuspension through wet milling and spray freeze- acterization of solid-state pharmaceuticals. J Pharm
drying for poorly water-soluble drugs. Eur J Pharm Sci Pharmacol. 2007;59(2):251–9.
50 (3-4):272-281. https://doi.org/10.1016/j.ejps.2013. Purvis T, Mattucci ME, et al. Rapidly dissolving
07.011. repaglinide powders produced by the ultra-rapid freez-
Niwa T, Mizutani D, Danjo K. Spray freeze-dried porous ing process. AAPS PharmSciTech. 2007;8(3):E58.
microparticles of a poorly water-soluble drug for respi- Rahman SMA, Mujumdar AS. 7 atmospheric freeze
ratory delivery. Chem Pharm Bull. 2012;60(7):870–6. drying. In: Progress in Food Preservation. Oxford:
Offerdahl TJ, Salsbury JS, Dong Z, Grant DJ, Schroeder Wiley-Blackwell; 2012. p. 143.
SA, Prakash I, Gorman EM, Barich DH, Munson Rasenack N, Muller BW. Dissolution rate enhancement by
EJ. Quantitation of crystalline and amorphous forms in situ micronization of poorly water-soluble drugs.
of anhydrous neotame using 13C CPMAS NMR spec- Pharm Res. 2002;19(12):1894–900.
troscopy. J Pharm Sci. 2005;94(12):2591–605. Rodriguez-Aller M, Guillarme D, Veuthey J-L, Gurny
Overhoff KA, Engstrom JD, et al. Novel ultra-rapid freez- R. Strategies for formulating and delivering poorly
ing particle engineering process for enhancement of water-soluble drugs. J Drug Deliv Sci Technol.
dissolution rates of poorly water-soluble drugs. Eur J 2015;30:342–51.
Pharm Biopharm. 2007a;65(1):57–67. Rogers TL, Johnston KP, et al. Solution-based particle
Overhoff KA, Moreno A, et al. Solid dispersions of formation of pharmaceutical powders by supercritical
itraconazole and enteric polymers made by ultra-rapid or compressed fluid CO2 and cryogenic spray-freezing
freezing. Int J Pharm. 2007b;336(1):122–32. technologies. Drug Dev Ind Pharm. 2001;27(10):
Overhoff KA, McConville JT, et al. Effect of stabilizer on 1003–15.
the maximum degree and extent of supersaturation and Rogers TL, Hu J, et al. A novel particle engineering
oral absorption of tacrolimus made by ultra-rapid technology: spray-freezing into liquid. Int J Pharm.
freezing. Pharm Res. 2008;25(1):167–75. 2002a;242(1–2):93–100.
Overhoff KA, Johnston KP, et al. Use of thin film freezing Rogers TL, Nelsen AC, et al. A novel particle engineering
to enable drug delivery: a review. J Drug Deliv Sci technology to enhance dissolution of poorly water
Technol. 2009;19(2):89–98. soluble drugs: spray-freezing into liquid. Eur J Pharm
Parthasarathi S, Anandharamakrishnan C. Enhancement of Biopharm. 2002b;54(3):271–80.
oral bioavailability of vitamin E by spray-freeze drying Rogers TL, Nelsen AC, et al. Enhanced aqueous dissolu-
of whey protein microcapsules. Food Bioprod Process. tion of a poorly water soluble drug by novel particle
2016;100:469–76. engineering technology: spray-freezing into liquid with
Pasandideh-Fard M, Chandra S, et al. A three dimensional atmospheric freeze-drying. Pharm Res. 2003a;20(3):
model of droplet impact and solidification. Int J Heat 485–93.
Mass Transf. 2002;45(11):2229–42. Rogers TL, Overhoff KA, et al. Micronized powders of a
Patravale VB, Date AA, et al. Nanosuspensions: a poorly water soluble drug produced by a spray-
promising drug delivery strategy. J Pharm Pharmacol. freezing into liquid-emulsion process. Eur J Pharm
2004;56(7):827–40. Biopharm. 2003b;55(2):161–72.
Paulekuhn GS, Dressman JB, et al. Salt screening and Rudokas M, Najlah M, Alhnan MA, Elhissi A. Liposome
characterization for poorly soluble, weak basic delivery systems for inhalation: a critical review
compounds: case study albendazole. Pharm Int J highlighting formulation issues and anticancer
Pharm Sci. 2013;68(7):555–64. applications. Med Princ Pract. 2016;25(Suppl 2):
Peeters J, Neeskens P, et al. Characterization of the inter- 60–72.
action of 2-hydroxypropyl-beta-cyclodextrin with Saboti D, Maver U, Chan H-K, Planinšek O. Novel
itraconazole at pH 2, 4, and 7. J Pharm Sci. 2002;91 budesonide particles for dry powder inhalation
(6):1414–22. prepared using a microfluidic reactor coupled with
Penkina A, Semjonov K, et al. Towards improved solubil- ultrasonic spray freeze drying. J Pharm Sci. 2017;106
ity of poorly water-soluble drugs: cryogenic (7):1881–8.
Sahakijpijarn S, Moon C, Ma X, Su Y, Koleng JJ, Suryanarayan R. X-ray powder diffractometry. In:

Dolocan A, Williams RO 3rd. Using thin film freezing Brittain H, editor. Physical characterization of pharma-
to minimize excipients in inhalable tacrolimus dry ceutical solids. New York: Marcel Dekker; 1995.
powder formulations. Int J Pharm. 2020a;586:119490. p. 187–221.
Sahakijpijarn S, Moon C, Koleng JJ, Christensen DJ, Suzuki H, Sunada H. Influence of water-soluble polymers
Williams RO 3rd. Development of Remdesivir as a on the dissolution of nifedipine solid dispersions with
dry powder for inhalation by thin film freezing. combined carriers. Chem Pharm Bull (Tokyo).
Pharmaceutics. 2020b;12(11):1002. 1998;46(3):482–7.
Sahakijpijarn S, Moon C, Warnken ZN, Maier EY, Tang X, Pikal MJ. Design of freeze-drying processes for
DeVore JE, Christensen DJ, Koleng JJ, Williams III pharmaceuticals: practical advice. Pharm Res. 2004;21
RO. (2021). In vivo pharmacokinetic study of (2):191–200.
remdesivir dry powder for inhalation in hamsters. Tesconi MS, Sepassi K, et al. Freeze-drying above room
International Journal of Pharmaceutics: X 3:100073. temperature. J Pharm Sci. 1999;88(5):501–6.
Salekigerhardt A, Ahlneck C, et al. Assessment of disorder TFF Pharmaceuticals, Inc. TFF Pharmaceuticals updates
in crystalline solids. Int J Pharm. 1994;101(3):237–47. progress of human clinical trial with Tacrolimus Inha-
Sarkari M, Brown J, et al. Enhanced drug dissolution using lation powder. https://ir.tffpharma.com/news-releases/
evaporative precipitation into aqueous solution. Int J news-release-details/tff-pharmaceuticals-updates-prog
Pharm. 2002;243(1–2):17–31. ress-human-clinical-trial
Scoutaris N, Vithani K, Slipper I, Chowdhry B, Thakkar SG, Ruwona TB, Williams RO 3rd, Cui Z. The
Douroumis D. SEM/EDX and confocal Raman micros- immunogenicity of thin-film freeze-dried, aluminum
copy as complementary tools for the characterization salt-adjuvanted vaccine when exposed to different
of pharmaceutical tablets. Int J Pharm. 2014;470(1–2): temperatures. Hum Vaccin Immunother. 2017;13(4):
88–98. 936–46.
Sethuraman V, Hickey A. Powder properties and their Thakkar SG, Warnken ZN, Alzhrani RF, Valdes SA,
influence on dry powder inhaler delivery of an Aldayel AM, Xu H, Williams RO 3rd, Cui
antitubercular drug. AAPS PharmSciTech. 2002;3: Z. Intranasal immunization with aluminum salt-
E28. adjuvanted dry powder vaccine. J Control Release.
Shah B, Kakumanu VK, et al. Analytical techniques for 2018;292:111–8. https://doi.org/10.1016/j.jconrel.
quantification of amorphous/crystalline phases in phar- 2018.10.020.
maceutical solids. J Pharm Sci. 2006;95(8):1641–65. Tong HHY, Du Z, Wang GN, Chan HM, Chang Q, Lai
Shekunov BY, Chattopadhyay P, et al. Particle size analy- LCM, Chow AHL, Zheng Y. Spray freeze drying with
sis in pharmaceutics: principles, methods and polyvinylpyrrolidone and sodium caprate for improved
applications. Pharm Res. 2007;24(2):203–27. dissolution and oral bioavailability of oleanolic acid, a
Sindel U, Zimmermann I. Measurement of interaction BCS Class IV compound. Int J Pharm. 2011;404(1):
forces between individual powder particles using an 148–58.
atomic force microscope. Powder Technol. 2001;117: Traini D, Rogueda P, et al. Surface energy and interparticle
247–54. forces correlations in model pMDI formulations.
Sinswat P, Gao X, et al. Stabilizer choice for rapid Pharm Res. 2005;22(5):816–25.
dissolving high potency itraconazole particles formed Turchiuli C, Munguia MJ, Sanchez MH, Ferre HC,
by evaporative precipitation into aqueous solution. Int Dumoulin E. Use of different supports for oil encapsu-
J Pharm. 2005;302(1–2):113–24. lation in powder by spray drying. Powder Technol.
Sitte H, Edelmann L, et al. Cryofixation without 2014;255:103–8.
pretreatment at ambient pressure. In: Steinbrecht RA, van de Witte P, Dijkstra PJ, et al. Phase separation pro-
Zierold K, editors. Cryotechniques in biological elec- cesses in polymer solutions in relation to membrane
tron microscopy. Berlin: Springer; 1987. p. 87–113. formation. J Memb Sci. 1996;117:1–31.
Sobel JD. Practice guidelines for the treatment of fungal Van Drooge DJ, Hinrichs WLJ, et al. Incorporation of
infections. For the Mycoses Study Group. Infectious lipophilic drugs in sugar glasses by lyophilization
Diseases Society of America. Clin Infect Dis. 2000;30 using a mixture of water and tertiary butyl alcohol as
(4):652. solvent. J Pharm Sci. 2004;93(3):713–25.
Stephenson GA, Forbes RA, et al. Characterization of the Vasanthavada M, Tong WQ, et al. Phase behavior of
solid state: quantitative issues. Adv Drug Deliv Rev. amorphous molecular dispersions I: determination of
2001;48(1):67–90. the degree and mechanism of solid solubility. Pharm
Stubberud L, Forbes RT. The use of gravimetry for the Res. 2004;21(9):1598–606.
study of the effect of additives on the moisture induced Vasanthavada M, Tong WQ, et al. Phase behavior of
recrystallization of amorphous lactose. Int J Pharm. amorphous molecular dispersions II: role of hydrogen
1998;163:145–56. bonding in solid solubility and phase separation kinet-
Suryanarayan R. Evaluation of two concepts of crystallin- ics. Pharm Res. 2005;22(3):440–8.
ity using calcium gluceptate as a model compound. Int
J Pharm. 1985;24:1–17.
Vaughn JM, McConville JT, et al. Single dose and multi- Yang W, Peters JI, et al. Inhaled nanoparticles – a current
ple dose studies of itraconazole nanoparticles. Eur J review. Int J Pharm. 2008a;356(1–2):239–47.
Pharm Biopharm. 2006;63(2):95–102. Yang W, Tam J, et al. High bioavailability from nebulized
Vishali DA, Monisha J, Sivakamasundari SK, Moses JA, itraconazole nanoparticle dispersions with biocompati-
Anandharamakrishnan C. Spray freeze drying: ble stabilizers. Int J Pharm. 2008b;361(1–2):177–88.
emerging applications in drug delivery. J Control Yang W, Johnston KP, et al. Comparison of bioavailability
Release. 2019;300:93–101. of amorphous versus crystalline itraconazole
Wallemacq PE, Verbeeck RK. Comparative clinical phar- nanoparticles via pulmonary administration in rats.
macokinetics of tacrolimus in paediatric and adult Eur J Pharm Biopharm. 2010;75(1):33–41.
patients. Clin Pharmacokinet. 2001;40(4):283–95. Yasmin R, Tan A, et al. Lyophilized silica lipid hybrid
Wang ZG, Hsiao BS, Sirota EB, Srinivas S. A simulta- (SLH) carriers for poorly water-soluble drugs: physi-
neous small- and wide-angle X-ray scattering study of cochemical and in vitro pharmaceutical investigations.
the early stages of melt crystallization in polyethylene. J Pharm Sci. 2014;103(9):2950–9.
Polymer. 2000;41(25):8825–32. Ye T, Yu J, Luo Q, Wang S, Chan H-K. Inhalable
Wang Y-B, Watts AB, et al. Effect of processing clarithromycin liposomal dry powders using ultrasonic
parameters on the physicochemical and aerodynamic spray freeze drying. Powder Technol. 2017;305:63–
properties of respirable brittle matrix powders. J Drug 70.
Deliv Sci Tec. 2014a;24(4):390–6. Yoshioka M, Hancock BC, et al. Crystallization of indo-
Wang Y-B, Watts AB, et al. In vitro and in vivo perfor- methacin from the amorphous state below and above
mance of dry powder inhalation formulations: compar- its glass transition temperature. J Pharm Sci. 1994;83
ison of particles prepared by thin film freezing and (12):1700–5.
micronization. AAPS PharmSciTech. 2014b;15(4): Yu L. Amorphous pharmaceutical solids: preparation,
981–93. characterization and stabilization. Adv Drug Deliv
Wanning S, Süverkrüp R, Lamprecht A. Pharmaceutical Rev. 2001;48(1):27–42.
spray freeze drying. Int J Pharm. 2015;488(1):136–53. Yu Z, Johnston KP, et al. Spray freezing into liquid versus
Ward S, Perkins M, et al. Identifying and mapping surface spray-freeze drying: influence of atomization on pro-
amorphous domains. Pharm Res. 2005;22(7): tein aggregation and biological activity. Eur J Pharm
1195–202. Sci. 2006;27(1):9–18.
Washburn EW. The dynamics of capillary flow. Phys Rev. Yu H, Teo J, Chew JW, Hadinoto K. Dry powder inhaler
1921;17:273–83. formulation of high-payload antibiotic nanoparticle
Watts AB, Wang Y-B, et al. Respirable low-density complex intended for bronchiectasis therapy: spray
microparticles formed in situ from aerosolized brittle drying versus spray freeze drying preparation. Int J
matrices. Pharm Res. 2013;30(3):813–25. Pharm. 2016;499(1):38–46.
Weiss C, McLoughlin P, et al. Characterisation of dry Yuan X, Sperger D, Munson EJ. Investigating miscibility
powder inhaler formulations using atomic force and molecular mobility of Nifedipine-PVP amorphous
microscopy. Int J Pharm. 2015;494(1):393–407. solid dispersions using solid-state NMR spectroscopy.
Williams RO, Hu J et al. Process for production of Mol Pharm. 2014;11(1):329–37.
nanoparticles and microparticles by spray freezing Zhang M, Li H, Lang B, O’Donnell K, Zhang H, Wang Z,
into liquid. U. S. Patent 20030041602. 2003. Dong Y, Wu C, Williams RO 3rd. Formulation and
Xu WJ, Xie HJ, et al. Enhanced dissolution and oral delivery of improved amorphous fenofibrate solid
bioavailability of valsartan solid dispersions prepared dispersions prepared by thin film freezing. Eur J
by a freeze-drying technique using hydrophilic Pharm Biopharm. 2012;82(3):534–44.
polymers. Drug Deliv. 2014;0:1–8. Zijlstra GS, Rijkeboer M, et al. Characterization of a
Yamashita K, Nakate T, et al. Establishment of new prep- cyclosporine solid dispersion for inhalation. AAPS
aration method for solid dispersion formulation of J. 2007;9(2):E190–9.
tacrolimus. Int J Pharm. 2003;267(1–2):79–91.
Precipitation Technologies
for Nanoparticle Production 12
Tuangrat Praphawatvet and Robert O. Williams III
Abstract bioavailability (Lipinski 2001, 2002). Conse-

Precipitation technologies have been widely quently, the majority of poorly water-soluble
studied for nanoparticle production because drugs fail to reach the market because their
they provide more control over particle size, absorption into the body is limited by their slow
shape, and morphology than mechanical pro- dissolution rates in bodily fluids (Gardner et al.
cesses, such as milling and homogenization. 2004; Rabinow 2004; Kipp 2004; Crison 2000).
Several precipitation processes are discussed Traditional approaches to improving drug disso-
in this chapter, with special attention to exper- lution rates have focused on increasing the drug’s
imental parameters and typical particle solubility, often utilizing solubilizing excipients
attributes. The chapter also touches on novel (CREMPHOR EL1 (polyethoxylated castor oil) is
nanoparticle recovery techniques that may be added to TAXOL)2, complexing agents
coupled with precipitation processes to enable (cyclodextrins and polyethylene glycols), coordi-
these precipitation technologies to be scaled nation complexes (water-soluble silver-ammonia
for commercial applications. complex ([Ag(NH3)2]+ and EDTA)), or
cosolvents (ethanol–water solvent mixtures)
Keywords (Rabinow 2004; Kipp 2004; Muller et al. 2001;
Loftsson 2017). However, the success of these
Nanoparticle precipitation · Nanoparticle approaches has been limited due to the large
production · Compressed fluids · Supercritical quantities of excipients required to achieve suffi-
fluids · Antisolvent precipitation · Precipitation cient solubilities, which increase the likelihood of
into acid · Nanoparticles · Dissolution adverse side effects in patients and limits drug
improvement loading (Rabinow 2004; Kipp 2004; Muller
et al. 2001). For example, the marketed product
SPORANOX IV3 requires 400 mg of
12.1 Introduction 2-hydroxypropyl-β-cyclodextrin to solubilize
10 mg of the active ingredient, itraconazole
It has been reported that 40% or more of newly
discovered drug candidates are poorly water sol-
1
uble, often resulting in poor and/or erratic CREMPHOR EL is a registered trademark of BASF
Corporation.
2
T. Praphawatvet · R. O. Williams III (*) TAXOL (paclitaxel) is a registered trademark of Bristol-
College of Pharmacy, The University of Texas at Austin, Myers Squibb Company.
3
Austin, TX, USA SPORANOX IV is a registered trademark of Janssen
e-mail: [email protected] Pharmaceutical Products, LP.
530 T. Praphawatvet and R. O. Williams III
(Sporanox Package Insert). Similar limitations Nanoparticles may be produced by top-down

impact the utilization of lipid-based formulations, or bottom-up approaches. Top-down approaches
which employ liposomes and emulsions to refer to mechanical processes, such as milling and
address solubility issues (Rabinow 2004). Solubi- homogenization, and use high-impact forces to
lization of drugs using lipid-based methods leads break large particles into smaller particles.
to drug loadings well below 50% w/w (Matteucci Particles with median diameters of 300–400 nm
et al. 2006) and often below ~10% w/w, espe- are commonly produced by these methods, and
cially for high-melting-point compounds, thus particles smaller than 200 nm have been reported
restricting their use in high-dose formulations for the poorly water-soluble drugs danazol (aver-
(Rabinow 2004; Kipp 2004; Muller et al. 2001). age particle diameter of 169 nm) (Liversidge and
Consequently, only a small number of Cundy 1995) and atovaquone (average particle
commercialized pharmaceutical products are diameter of 100 nm) (Dearn 1994; Westesen and
based on these strategies (Muller et al. 2001). Siekmann 1998) using top-down methods. How-
An alternative approach to enhancing the dis- ever, the high-energy inputs required to achieve
solution rates of poorly water-soluble drugs has these levels of size reduction may subject the drug
been to formulate the drugs as nanoparticles, to chemical degradation, often through thermal
loosely defined in the pharmaceutical industry as degradation, due to the considerable amount of
structures with a diameter less than 1 μm. heat that is often generated during milling and
According to the Noyes–Whitney equation, homogenization processes (Jacobs et al. 2000;
which is based on Fick’s first law of diffusion, Liversidge et al. 2003; Muller and Bohm 1997).
dissolution rates of drug particles may be These high-energy methods are also prone to
enhanced by increasing the drug’s solubility in producing partially amorphous drug domains,
aqueous media (CEq) and/or by reducing particle complicating control of crystalline morphology,
size, which increases the surface area for adsorp- and thus drug stability (Chan and Chew 2003).
tion (A) and decreases the boundary layer thick- Moreover, these methods often require lengthy
ness (h) (Noyes and Whitney 1897): processing times, risk contamination with
impurities, and are subject to low process yields
∂M DA
¼ C Eq CBulk , ð12:1Þ (Muller et al. 2001). In contrast, bottom-up
∂t h
approaches refer to solution-based precipitation
where M is the mass of undissolved drug, t is the techniques that induce phase separation of the
time, D is the average diffusion coefficient, and drug (originally in solution) from the solvent.
CBulk is the drug concentration in the bulk solu- Precipitation is driven by a deviation from phase
tion. Nanoparticle formulations offer several equilibrium conditions, where typical supersatu-
advantages over lipid-based solubilization ration driving forces are gradients in concentra-
methods for improving drug dissolution rates. tion or temperature. This chapter will focus on
Unlike lipid-based techniques, particle formation precipitation processes where supersaturation of a
processes are more amenable to compounds that drug solution produces nucleation and growth
have low solubility in both water and oils, which under controlled conditions to influence particle
is often the case for high-energy crystals formation. Supersaturation, S, is defined as the
(Rabinow 2004). Additionally, by circumventing solute concentration (Cdrug) relative to that under
the need to deliver a dissolved compound, the equilibrium conditions (S ¼ Cdrug/Ceq). Freezing-
drug’s preferred crystalline state may be pre- induced nanoparticle precipitation methods
served during delivery and storage (Rabinow (based on a thermal driving force) are discussed
2004). Furthermore, solid nanoparticles facilitate in detail in review articles by Overhoff et al.
higher drug loadings than solubilized (2007a, 2009).
formulations, which is crucial for high-dose In nonfreezing-based precipitation techniques,
compounds. the poorly water-soluble drug is typically
dissolved in a solvent, and precipitation of the
12 Precipitation Technologies for Nanoparticle Production 531
!
drug is initiated by a reduction in solvent power, 16πγ 3 V M 2 N A
by either addition of an antisolvent or solvent B / exp
0
, ð12:2Þ
3ðRT Þ3 ½ ln ð1 þ SÞ2
evaporation. A reduction in solvent powder
leads to supersaturation of the drug and drives where γ is the interfacial tension, VM is the molar
nucleation of drug particles. Once nucleation volume, NA is Avogadro’s number, R is the ideal
occurs, the particles grow by condensation, in gas law constant, and T is the temperature.
which dissolved drug molecules diffuse to the According to (12.2), nucleation rates increase as
particle surface and integrate into a solid particle, the degree of supersaturation increases. However,
and/or by coagulation, where multiple particles supersaturation levels decrease as particles grow
collide and aggregate to form larger particles by condensation due to a reduction in the solute
(Fig. 12.1) (Weber and Thies 2002). Stabilizers mass in solution. Hence, condensation competes
may be added to the system to arrest particle with nucleation. Furthermore, coagulation
growth. Moreover, they can stabilize drug competes with condensation by reducing the
particles that suffer from instability in the physio- total number of particles, and thus surface area,
logical environment. As a result, stabilizers can in the system (Matteucci et al. 2006). Therefore,
reduce particle sizes and improve particle final particle properties, including size distribu-
dispersibility during precipitation processes tion and morphology, are heavily impacted by the
(Tang et al. 2020). processing parameters that influence nucleation
Particle nucleation and growth are competing and growth rates (e.g., solvent choice, stabilizer
processes. The degree of supersaturation, S, sig- selection, and mixing rates). Rapid generation of
nificantly impacts nucleation rates, as seen in the supersaturation over a narrow time interval
equation describing primary nucleation rate, B0 facilitates more narrow PSDs. Therefore, faster
(Sohnel and Garside 1992): nucleation rates, relative to growth, favor the
production of uniformly smaller particles.
Fig. 12.1 Mechanism of Coagulation

precipitation. Adapted from
Weber and Thies (2002)
Nucleation Condensation
Relative to top-down approaches (milling and advantages over more conventional liquid
homogenization processes), precipitation antisolvent processes. The antisolvent may be
technologies are typically more controlled, in completely removed via pressure reduction to
terms of consistently producing particles with the gaseous phase, resulting in improved product
similar morphologies, and offer the ability to purities and reduced environmental and toxicity
achieve higher drug loadings (Matteucci et al. concerns. An SCF is a fluid that has been com-
2006, 2007; Overhoff et al. 2007a, b; Engstrom pressed beyond its critical pressure (Pc) or heated
et al. 2007, 2008; Rasenack and Muller 2002; above its critical temperature (Tc). An important
Rogers et al. 2004; Shoyele and Cawthorne feature of SCFs is that their densities can change
2006; Vaughn et al. 2005; Young et al. 2000). significantly with small changes in pressure.
Precipitation processes not only increase drug These density changes give rise to variations in
loading but also improve the dissolution rate of diffusivity and viscosity, as well as the solubility
drug particles and formulations compared to of other solvents and small solutes. Typical
top-down processes (Tang et al. 2020; Zhou diffusivities of SCFs are on the order of 103
et al. 2020; Essa et al. 2017). Moreover, precipi- cm2/s (~100 times greater than that for liquids),
tation processes are often easier to scale up and and their viscosities are on the order of 104
require less particle handling than milling and g/cm/s (~100 times lower than that for liquids).
homogenization operations, resulting in higher These favorable mass transfer properties facilitate
process yields and lower impurity risks, as well rapid diffusion of the SCF antisolvent into a liq-
as simplified cleaning and sterilization procedures uid solvent, enabling rapid supersaturation and
(Rogers et al. 2001a). Additionally, precipitation nucleation, thus favoring the production of small
technologies may be operated as continuous or particles.
semicontinuous processes, whereas milling and Several commonly used SCFs are listed in
homogenization operations are batch processes Table 12.1 (Sekhon 2010). Of these fluids, carbon
(Rogers et al. 2001a). The continuous processes dioxide (CO2) is the most prevalently used in
used in precipitation technologies improve drug pharmaceutical applications because it is inex-
solubility and therapeutic benefit (Chen et al. pensive, nonflammable, and nontoxic. Figure 12.2
2018). This chapter will focus on several different shows how small changes in pressure and temper-
approaches to nanoparticle precipitation that are ature result in significant changes in the density of
relevant to pharmaceutical drug development, CO2, which in turn, largely affects the solubility
highlighting key advances and important of small molecules in CO2 (Fig. 12.3). The den-
processing parameters for various active pharma- sity of CO2, as a function of pressure (P) and
ceutical ingredients (API). Precipitation processes temperature (T), may be obtained from the
utilizing compressed or supercritical fluids, as NIST Standard Reference Database (http://
well as aqueous media, as antisolvents, will be webbook.nist.gov) or calculated as follows
summarized. Furthermore, novel modifications to (Jouyban et al. 2002):
conventional precipitation processes will be pffiffiffiffi
1
discussed, in addition to several techniques that ρCO2 ¼ exp ð27:091 þ 0:609 T
44
have been used to harvest the resultant
3966:170 3:445P
nanoparticles after precipitation, including þ
T T
flocculation-based processes. pffiffiffi
þ 0:401 PÞ, ð12:3Þ
where ρCO2 is in moles/mL, T is in Kelvin, and

12.2 Precipitation Processes
P is in bars.
Utilizing Compressed
In addition to CO2 density, a drug’s solubility
or Supercritical Fluids
in supercritical CO2 (scCO2) is dependent upon
the drug’s vapor pressure and drug–CO2 interac-
Compressed fluid and supercritical fluid (SCF)
tion. The following empirical correlation for drug
antisolvent precipitation processes offer several
Table 12.1 Critical constants for select supercritical fluids (SCF)

SCF Tc ( C) Pc (bar) Safety hazard
Trifluoromethane (fluoroform) 25.9 47.5
Chlorotrifluoromethane 28.9 39.2
Ethane 32.3 48.8 Flammable gas
Carbon dioxide 31.1 73.7
Dinitrogen monoxide (laughing gas) 36.5 72.6 May enhance combustion of other substances
Sulfur hexafluoride 45.5 37.6
Chlorodifluoromethane (HCFC 22; R22) 96.4 49.1 Combustible under certain conditions
Propane 96.8 43.0 Extremely flammable
Ammonia 132.4 112.7 Flammable and toxic
Trichlorofluoromethane (CFC 11; R11) 198.0 44.1
Water 374.0 220.5
Methane 82.6 45.9 Flammable gas
Ethylene 9.2 50.4 Extremely flammable gas and toxic
Propylene 91.85 46.2 Extremely flammable gas and toxic
Xenon 16.55 58.7 May cause rapid suffocation
Toluene 318.55 41.1 Flammable and toxic
Ethanol 240.75 6.14 Highly flammable
Acetone 234.95 4.70 Flammable and eye irritated
Adapted from Gupta (2006), Li and Xu (2019)
Fig. 12.2 Dependency of

CO2 density on pressure
and temperature. Data from
NIST Standard Reference
Database (http://webbook.
nist.gov/chemistry)
solubility in CO2 was developed by Mendez- Table 12.2). Accurate knowledge of a drug’s
Santiago and Teja (1999): solubility in CO2 is necessary to reliably produce
adequate process yields.
106 A Bρ1
γ2 ¼ exp þ þC , ð12:4Þ SCF precipitation techniques fall into three
P T T major categories: (1) gas antisolvent precipitation
(GAS), (2) precipitation with a compressed
where γ 2 is in μmol/mol, P is in bars, T is in
antisolvent (PCA), and (3) rapid expansion from
Kelvin, ρ1 is the CO2 density in mol/mL, and
supercritical solutions (RESS). PCA processes
the constants A, B, and C are empirical constants
are also commonly referred to as aerosol solvent
(values for various drug molecules are listed in
extraction system (ASES), solution-enhanced
Fig. 12.3 Solubility of

several drug compounds in
CO2 at varying pressures
and temperatures. Data
adapted from Gupta (2006)
Table 12.2 Values of empirical constants used to determine drug solubility in CO2 using (12.4)
Drug A B C
7-Azaindole 8412 87,110 20.66
Behenic acid 4473 61,240 6.8
Biphenyl 10,200 132,800 25.75
Brassylic acid 10,860 146,100 21.01
Capsaisin 7172 70,830 19.54
Cholecalciferol 9784 172,500 18.42
Diphenylamine 18,720 397,100 33.4
Eicosanoic acid 15,990 161,600 36.97
1-Eicosanol 14,530 122,500 36.15
Endrin 9912 167,800 20.29
Ergocalciferol 1092 173,500 21.51
Flavone 11,430 110,100 27.38
d()-Fructose 871.2 10,740 4.29
d(+)-Glucose 847.1 2471 9.12
3-Hydroxyflavone 9746 81,530 21.31
Ketoprofen 12,090 157,500 24.72
Medroxyprogesterone acetate 10,270 186,100 17.77
Methoxychlor 12,670 184,100 27.38
Monocrotaline 10,440 8057 20.28
Mystiric acid 17,250 173,100 44.84
Naproxen 9723 122,900 18.11
Narasin 8,529 124,900 13.86
Nifedipine 10,020 168,500 15.92
Nimesulide 13,820 186,900 28.14
Nitrendipine 9546 151,400 15.91
Octacosane 19,860 123,000 52.555
(continued)

Drug A B C
1-Octadecanol 17,290 141,000 45.32
Palmityl behenate 8378 59,180 18.44
Penicillin V 6459 73,730 13.29
Phenylacetic acid 13,730 14,450 35.78
Piroxicam 10,560 18,130 17.57
Progesterone 12,090 21,040 23.43
t-Retinol 8717 168,900 16.6
Salinomycin 18,990 185,500 42.05
Stigmasterol 13,010 169,000 25.23
Testosterone 14,330 238,300 26.42
Theobromine 7443 114,000 8.31
Theophylline 6957 94 760
Triacontane 22,965 199,800 57.22
Trioctylphosphine oxide 9378 211,900 17.65
Vanillin 7334 136,500 14.53
Data from Gupta (2006)
dispersion by supercritical fluids (SEDS), and Excessive solubility of the API in CO2 would
supercritical antisolvent (SAS). The differences facilitate particle growth. Under optimal
between these different precipitation techniques operating conditions, CO2’s high solubility and
are discussed in the following sections of this favorable transport properties in the organic sol-
chapter. vent facilitate homogeneous supersaturation
conditions more rapidly than can be achieved
using liquid antisolvents. When precipitation is
complete, the CO2–organic solvent solution is
12.2.1 Precipitation with a Gaseous
flushed from the system and the precipitate, i.e.,
Antisolvent (GAS)
the drug powder, remains in the precipitation
vessel. The drug powder may then be washed
GAS precipitation is a batch process in which the
with fresh CO2 to remove the excess organic
SCF antisolvent, often CO2, is added to an
solvent. A schematic of the GAS precipitation
organic solution containing dissolved API. Typi-
system is shown in Fig. 12.4. The primary draw-
cal operating pressures for this process are 5–8
back of the GAS precipitation process is the diffi-
MPa, in the range where CO2 is highly soluble in
culty in harvesting the precipitate drug particles
most organic solvents (Martin and Cocero 2008).
from the organic solvent solution while
As the CO2 dissolves into the solute-rich liquid
minimizing particle growth and agglomeration.
phase, the solvent strength decreases. Conse-
Furthermore, in cases where elevated
quently, the API’s solubility in the solvent
temperatures are needed to sufficiently expand
decreases, which generates supersaturation of
the SCF into the organic solvent, thermal degra-
the API and promotes nucleation and precipita-
dation of the API may occur.
tion. In some cases, additional excipients may
Because GAS precipitation is driven by the
also be dissolved in the organic drug solution to
antisolvent capabilities of CO2 in the organic
precipitate the API within an excipient matrix. In
solution, appropriate processing conditions may
order to induce rapid drug nucleation, which
be selected based on optimizing thermodynamic
favors the production of small particles, CO2
criteria, specifically by understanding the volu-
must be readily soluble in the organic solvent
metric expansion of the organic solvent due to
and the API must have low solubility in CO2.
Fig. 12.4 Schematic of CO2 pump

GAS process. Schematic Solvent
adapted from Martin and recovery
Cocero (2008) and vent
Precipitator
CO2 reservoir
Solution
CO2 solubilization, and thus solubility of the sol- concentration is increased will likely not yield
ute in the solvent–CO2 mixture. In a study by de small, uniform particles, as precipitation will
la Fuente et al., volumetric expansion of the take place continuously as CO2 is fed into the
organic solvent was correlated to the difference precipitation vessel. de la Fuente et al.
between the partial molar volumes (v) of the hypothesized that optimum GAS precipitation
organic solvent under operating conditions versus conditions exist at the minimum of the solvent’s
atmospheric pressure, as shown in the following volumetric expansion curve, as defined in (12.5).
equation (de la Fuente Badilla et al. 2000): According to Fig. 12.5, naphthalene was
predicted to be successfully precipitated using
Δv vðT, PÞ v0 ðT, P0 Þ
¼ , ð12:5Þ the GAS technique while phenanthrene was not
v v0 ðT, P0 Þ (de la Fuente Badilla et al. 2000). This model has
where T is the operating system’s temperature, P0 also been verified experimentally for a salicylic
is the atmospheric pressure, and v0 is the partial acid–propanol–CO2 system (Shariati and Peters
molar volume at the operating system’s tempera- 2002).
ture and atmospheric pressure. Studies have Typical particle sizes of poorly water-soluble
shown that (12.5) adequately predicts drug drugs prepared by GAS precipitation are on the
solubilities in a solvent–CO2 solution for naph- order of 1–10 μm (Martin and Cocero 2008),
thalene and phenanthrene in a toluene–CO2 sys- although submicron particle sizes have been
tem, and thus it is capable of predicting their achieved in some cases (Turk 2009). To develop
success in forming satisfactory particles by GAS GAS precipitation, high pressure, low drug con-
precipitation (Fig. 12.5) (Martin and Cocero centration, and T-junction micromixer were
2008; de la Fuente Badilla et al. 2000; de la applied to produce submicron particles as small
Fuente et al. 2004). The likelihood of a specific as 0.5 μm (Arora et al. 2020). GAS precipitation
solute to successfully form small, uniform processes have also been reported to be success-
particles by GAS precipitation is indicated by a fully scaled from a 300-mL to 1-L batch size
steep decrease in its solubility at some CO2 con- (Muhrer et al. 2003; Muhrer and Mazzotti
centration. More specifically, a high sensitivity of 2003). However, when processes are scaled to
the solute’s solubility to CO2 concentration larger volumes, a stirrer was needed to improve
indicates that precipitation will occur rapidly mixing between the organic solvent and CO2
and homogeneously once a critical concentration (Martin and Cocero 2008). Key process
is reached. On the other hand, systems that dem- parameters that control final particle size and
onstrate only a slow decrease in solubility as CO2 morphology include the pressure and temperature
Fig. 12.5 Relative

volumetric expansion of
toluene, defined as the
difference between the
partial molar volumes (v) of
toluene under operating and
atmospheric conditions,
and solute solubility in a (a)
CO2–toluene–naphthalene
and (b) CO2–toluene–
phenanthrene system.
Reprinted from Martin and
Cocero (2008). Copyright
(2008), with permission
from Elsevier
of the precipitation process, solvent selection, and size distribution (PSD) was unimodal for “slow”
the CO2 addition rate to the organic solution (QA 0.04 min–1) and “fast” (QA 1.54 min–1)
(Muhrer et al. 2003; Fusaro et al. 2004; CO2 addition rates but was bimodal for “interme-
Subramaniam et al. 1997; Mueller et al. 2000). diate” addition rates (0.1 QA 0.5 min–1)
As mentioned previously, changes in pressure (Muhrer et al. 2003). In another example where
and temperature largely influence the mass trans- paracetamol (aqueous solubility ~12 mg/mL) was
fer properties of CO2. Solvent selection and the precipitated from an acetone solution by GAS, the
rate of CO2 addition affect supersaturation levels mean particle size decreased threefold
and, thus, nucleation and crystallization rates. In a (250–87 μm) with an increase in QA by a factor
study by Muller et al., GAS precipitation of a of three (0.1–3.33 min–1) (Fusaro et al. 2004).
proprietary poorly water-soluble drug yielded Similarly, nanoparticles of 5-fluorouracil
amorphous spheres when precipitated from etha- precipitated from dimethylsulfoxide were highly
nol, whereas a crystalline form was obtained dependent on the processing variables such as
when acetone or acetonitrile was chosen as the antisolvent addition rate, pressure, temperature,
solvent, even though all other operating and solution concentration (Esfandiari et al.
conditions were identical (Mueller et al. 2000). 2013a, b). In contrast, GAS precipitation of lyso-
Likewise, an amorphous solid dispersion of zyme from dimethyl sulfoxide (DMSO) did not
oridonin stabilized with PVP K17 exhibiting a demonstrate a significant change in particle size
dramatic increase in bioavailability was obtained with varying CO2 addition rates (Muhrer and
from ethanol solution (Li et al. 2011). Additional Mazzotti 2003). Additionally, for the studies
studies by Muller et al. reported that the average using paracetamol and lysozyme, a unimodal
particle size of their proprietary poorly water- PSD was obtained regardless of QA, which was
soluble drug, when precipitated from an ethanol varied from “slow” to “fast” (Fusaro et al. 2004;
solution, could be reproducibly adjusted to sizes Muhrer and Mazzotti 2003). In terms of pressure
between ~200 nm and 10 μm by varying the effects, a high-pressure platform was used with
addition rate of CO2 over two orders of magni- CO2. Griseofulvin was precipitated by the high-
tude (Muhrer et al. 2003; Mueller et al. 2000). pressure microfluid platform inserted into a water
The CO2 addition rate (QA) was defined as the tank. Pressurized CO2 used as an antisolvent and
ratio between the CO2 flow rate and the initial pressurized solvent (griseofulvin solution) were
volume of organic solution in order to normalize provided by syringe pumps at the controlled tem-
for different batch sizes. Moreover, the particle- perature (25.0 0.1 C) and pressure (40 bar).
The different drug concentrations and micromixer PSDs may be tuned by controlling CO2 addition
types affected the particle size of the processed rates. An increase in QA elevates supersaturation
griseofulvin. The results showed that higher drug levels, facilitating higher nucleation rates and
concentration and lower driving pressure thus promoting the formation of more nuclei,
provided larger particle formation. The particles which results in a larger population of smaller
prepared with 0.1% drug concentration had the particles. The relationship, as determined by
smallest size at 500 nm. In terms of micromixer Muhrer’s model (Muhrer et al. 2002), between
type, a T-junction mixer provided higher supersaturation ratio, S, and average particle
antisolvent concentration. Thus, the smaller size, as a function of QA, is illustrated in
particles were obtained at the smallest size of Fig. 12.6. The supersaturation ratio was calcu-
400 nm (Arora et al. 2020). lated as the ratio of the fugacity of the solute in
In light of these conflicting reports relating the liquid phase to the fugacity of the pure solid.
experimental parameters to final particle Muhrer’s model also demonstrated that in cases
properties, a better understanding of the GAS where secondary nucleation is dominant, the
process, specifically the sensitivity of CO2 addi- mean particle size is largely unaffected by
tion rates on resultant particle size, has been changes in the rate of CO2 addition, whereas
sought through the development of theoretical systems with intermediate secondary nucleation
models to describe the GAS process. Muhrer rates were predicted to be moderately affected by
et al. presented a model that couples population variations in QA and possessed bimodal
balance theory with thermodynamic equilibrium distributions. Good quantitative agreement with
to relate nucleation rates to final particle size this model was obtained in two studies where
(Muhrer et al. 2002). Solution thermodynamics phenanthrene was micronized using GAS precip-
and particle formation and growth are accounted itation (Muhrer et al. 2002; Bakhbakhi et al.
in the model based on assumptions of isothermal 2005). However, because this model was devel-
conditions and instantaneous vapor–liquid phase oped primarily to explain the effect of QA on final
equilibrium upon addition of the antisolvent, thus particle size, minor deviations between the model
neglecting any mass transfer resistance. Particle and experimental results were observed when
growth, however, is described by an empirical examining the role of initial drug concentration
correlation, which does not discern between the on particle size for a poorly water-soluble drug–
different mechanisms of condensation and coag- ethanol–CO2 system. The discrepancies were
ulation (Martin and Cocero 2008; Dodds et al. attributed to the fact that the model neglects
2007). In a study recently published by Kikic mass-transfer resistances and thus did not account
et al., the impact of the organic solvent selection, for the increasing viscosity of the organic solution
the ratio of CO2/solution, and pressure on drug due to higher drug concentration. In response,
solubility was estimated using Peng–Robinson’s Elvassore proposed a population balance model
equation of state. Ternary diagrams were that accounted for particle nucleation, growth,
obtained, enabling an initial screening for optimal aggregation, and settling, where nucleation and
processing conditions. These estimations are also growth were described by the McCabe model
valid for the SAS process detailed in Sect. 12.2.2 (Elvassore et al. 2003, 2004). The model was
(Kikic et al. 2010). validated with experimental measurements for
In the Muhrer et al. model, systems in which the GAS precipitation of poly(L-lactide) acid
primary nucleation (generation of nuclei resulting (PLLA). While a good correlation was achieved,
from supersaturation, in the absence of drug several model parameters could not be experi-
crystals) is dominant to secondary nucleation mentally determined and were assumed in order
(occurs in the presence of existing drug crystals) to fit the model to the experimental data. The
tend to be more susceptible to variations in CO2 results of this model indicated that aggregation
addition rates. Therefore, in systems dominated rates should not be neglected and that they
by primary nucleation, average particle sizes and strongly influence the attainment of unimodal
Fig. 12.6 Effect of CO2

addition rate, QA, on the
supersaturation ratio, S, and
the average size of particles
produced by GAS
precipitation for a model
phenanthrene–toluene–CO2
system. Reprinted with
permission from Fusaro
et al. (2005). Copyright
(2005) American Chemical
Society
(low aggregation rates) versus bimodal (high evolution of size and aspect ratios of crystalline
aggregation rates) distributions, in contrast to drugs. Carbamazepine (CBZ), used as a model
Muhrer’s model which did not account for aggre- drug, was produced by different rates of com-
gation rates. Dodds et al. developed another pressed CO2. A population balance model with
model that used solution thermodynamics and two length dimensions that produced the crystal-
crystallization kinetics to examine particle growth line structure of CBZ was incorporated with the
in GAS processes (Dodds et al. 2007). The Dodds thermodynamic model to forecast the vapor con-
et al. model showed good agreement with experi- tent of the equilibrium solvent at different pres-
mental results for GAS precipitation of naphtha- sure and CO2 percentage. The model and
lene, phenanthrene, cholesterol, and methodology are used to predict the crystalliza-
beclomethasone dipropionate (Dodds et al. tion process and behavior. For example, the study
2007). In a recent study by Esfandiari et al., math- used this model to understand the favorable size-
ematical modeling of the GAS process was used dependent crystal growth, the aspect ratio of crys-
to determine nucleation and growth rate tal evaluation, and the solute concentration
parameters. The model was validated by compar- (Muthancheri et al. 2020). While all of the models
ison with experimental data and was successful in contributed to an enhanced understanding of the
predicting particle size distribution (Esfandiari et underlying mechanisms driving GAS precipita-
Ghoreishi 2013a, b). Also, Erriguible et al. tion, further validation is required to understand
published an approach to model a case of their applicability to additional drug–solvent
co-crystallization with naproxen and nicotin- systems. It should be noted that predicting physi-
amide as co-former. Their modeling accounted cal properties of particles produced by GAS pre-
for the liquid vapor equilibrium and its impact cipitation has not been trivial and currently
on the solubility of naproxen and nicotinamide appears to be highly dependent on a specific
and also the nucleation and growth of the system due to the complexities that arise from
co-crystal. The experimental size distribution multiple interactions within the system (drug–sol-
was in agreement with the one predicted vent, solvent–CO2, and drug–solvent/CO2 solu-
(Erriguible et al. 2015). Recently, to improve the tion). It is also important to note that GAS
problem of CO2-antisolvent modeling, precipitation generally does not produce
Muthancheri et al. developed a model that was nanoparticles, as it is typically limited by the
incorporated with GAS to predict the dynamic mixing and thus nucleation rates that can be
Fig. 12.7 Schematic of the

PCA process. Reprinted
Martin and Cocero (2008).
Copyright 2008 with
achieved in this system. For example, shown in Fig. 12.7. The PCA process typically
nanoparticles of s-nitrosoglutathione, a therapeu- operates at 9–15 MPa, slightly higher than GAS
tic agent, cannot be produced by GAS because its processes, in order to achieve higher supersatura-
solubility is too high in a CO2/DMSO mixture, tion values and sufficient mixing between the
while the SAS process showed the ability to pro- CO2 and organic solution feed streams (Martin
duce nanoparticles at high yield (72%) and high and Cocero 2008).
purity (>80%) (Zhou et al. 2020). Atomization of the drug–CO2 solution into the
antisolvent, as opposed to bubbling the CO2 solu-
tion, facilitates more rapid mass transfer between
12.2.2 Precipitation the drug solution and the antisolvent, which
with a Compressed Liquid makes the PCA process more conducive to the
or Supercritical Fluid (PCA, production of smaller particles compared to GAS
ASES, SEDS, and SAS) precipitation (Rogers et al. 2001a; Martin and
Cocero 2008; Fusaro et al. 2004). Moreover, the
The physical properties of drug powders pro- PCA process increases the surface area of
duced by precipitation methods are greatly particles. Thus, the PCA process improved the
influenced by the process arrangement. In con- dissolution rate of drugs (Park et al. 2017). The
trast to GAS precipitation, the PCA process high surface area of atomized droplets increases
atomizes the drug solution into the SCF the area of intimate contact between the drug
antisolvent. In PCA, the organic solution solution and the antisolvent to facilitate mixing,
containing the API is atomized into a vessel that thus promoting rapid supersaturation and precipi-
has been pressurized with the compressed liquid tation. Upon atomization, the organic solvent
or SCF, often CO2. Unlike the batch GAS pro- diffuses into the CO2 phase and the CO2 diffuses
cess, PCA is a semicontinuous technique because into the organic droplets, resulting in a more
the scCO2 is continuously fed throughout the efficient, bidirectional mass transfer of CO2 and
atomization process to promote more rapid organic phase, in contrast to the unidirectional
mixing with the organic solvent. Upon removal mass transfer in GAS precipitation (Rogers et al.
of the residual solvent, the pressure in the vessel 2001a; Martin and Cocero 2008).
is reduced to atmospheric pressure, and the drug The mass-transfer efficiency between the CO2
particles are collected by a filter at the bottom of and organic solvent phase may be further
the vessel. Similar to the GAS process, additional increased by adjusting operating parameters of
excipients may also be dissolved in the organic the PCA process, including increasing the
drug solution to produce composite API/excipient miscibility between the solvent and CO2 or by
particles. A schematic of the PCA system is tuning the degree of atomization of the organic
solution into the CO2 phase. Increased miscibility zero over a distance shorter than that of charac-
between the solvent and CO2 and more intense teristic jet break-up lengths, calculated based on
atomization, which yields higher surface area classic jet break-up theory (Lengsfeld et al. 2000).
droplets, enhance mass-transfer efficiency Thus, distinct droplets do not form, and the sol-
(Rogers et al. 2001a; Fusaro et al. 2004). For vent stream forms more of a gaseous plume
systems in which the solvent and CO2 are fully (Bristow et al. 2001). Therefore, atomization for
miscible (supercritical conditions), experimental miscible fluids was analyzed using gaseous
parameters that affect mixing rates between the mixing theory and mixing rates, using mixing
solvent and CO2 streams, such as degree of atom- length scales for turbulent mixing (Shekunov
ization, are less likely to influence precipitation et al. 1999; Jarmer et al. 2003).
results for some nozzle designs, thus suggesting Atomization intensity may be increased using
that mixing rates between the solvent and CO2 are ultrasonic dispersion devices, coaxial nozzles, or
faster than precipitation rates (Reverchon et al. two-component jet nozzles to enhance the inter-
2003a, b, 2007). However, for systems where action between the solvent and antisolvent in a
solvent and antisolvent are only partially miscible mixing chamber prior to atomization. Schematics
(subcritical conditions), mixing parameters sig- of different nozzle types are shown in Fig. 12.8.
nificantly influence precipitation results. A recent Ultrasonic dispersion devices vibrate at an ultra-
study demonstrated that the number of nozzles in sonic frequency to enhance mass-transfer effi-
the PCA process also affected the size and mor- ciency by increasing mixing rates between the
phology of particles. For example, coprecipitation solvent and antisolvent, as well as atomizing the
of substrates and polymers showed submicron feed solution into smaller droplets. Final particle
particle size and amorphous morphology when sizes may be tuned by controlling the vibration
using a single nozzle. On the other hand, the intensity of the dispersion device. For coaxial
precipitated microparticles were obtained when (or two-fluid) nozzles (Fig. 12.8a), the organic
using double nozzles (García-Casas et al. drug solution is fed through one axis and the
2017b). Furthermore, changes in particle mor- scCO2 is fed through the other. As the two feeds
phology, as well as an increased propensity for meet, intense mixing of the two streams facilitates
particle agglomeration, are frequently observed at rapid nucleation and particle precipitation upon
subcritical conditions, indicating that mixing of atomization from the nozzle. Primary particle
the CO2 and solvent is not complete and occurs sizes may be controlled by adjusting the relative
simultaneously with precipitation during droplet velocities of the two streams, which regulates the
formation (Martin and Cocero 2008). An increase intensity of mixing between the solvent and
in atomization intensity facilitates solvent–CO2 antisolvent phase. Several configurations for
mixing during droplet formation. For subcritical coaxial nozzles have been utilized, with optimal
conditions, the degree of atomization may be designs heavily dependent on the particular drug
quantified by the Weber number, NWe, a dimen- system. For example, curcumin and PVP were
sionless ratio of inertial to surface tension forces, dissolved in an acetone/ethanol mixture, and
which is given by then the solution was processed by SAS with a
coaxial nozzle to obtain nanoparticle size with
N We ¼ ρA v2 Ddrop =σ, ð12:6Þ spherical morphology. Nanoparticles prepared at
the SCCO2 operating temperature of 40 C
where ρA is the antisolvent density, v is the rela-
showed an average particle size of 82 nm. More-
tive velocity, Ddrop is the droplet diameter, and σ
over, a higher PVP concentration added to the
is the interfacial tension. Higher-intensity atomi-
solution resulted in a higher dissolution rate and
zation is characterized by larger NWe values for a
aqueous solubility (Machmudah et al. 2020). In
given Reynolds number (Re) (Lengsfeld et al.
some cases, a converging–diverging nozzle is
2000). However, for supercritical conditions, the
employed to rapidly disperse the liquid feed dur-
surface tension of the organic solvent decreases to
ing atomization to facilitate nanoparticle
Drug
a b Drug c Drug d
SCF solution solution solution
Drug SCF
solution SCF SCF
SCF
Swirl insert Swirl

vane
Swirl chamber
Nozzle exit
Two fluid nozzle
Fig. 12.8 Schematics of different nozzles used in SCF et al. 2005; Jarmer et al. 2003). Reprinted with permission
precipitation processes: (a) coaxial nozzle (Okamoto and from Okamoto and Danjo (2008) (Copyright 2008 with
Danjo 2008), (b) coaxial nozzle with a converging– permission from Elsevier), Fusaro et al. (2005) (Copyright
diverging annulus (Fusaro et al. 2005), and (c, d) two 2005 American Chemical Society), and Jarmer et al.
configurations for a two-component jet nozzle (Fusaro (2003) (Copyright 2003 with permission from Elsevier)
production (Fig. 12.8b). In two-component jet laboratory and pilot plant scales, under conditions
nozzles, the antisolvent is introduced at a sharp of partial solubility of CO2 in the solvent, was
angle into the mixing chamber to enhance turbu- heavily influenced by nozzle design. While PCA
lence of the fluids during mixing (Fig. 12.8c, d). processes operating at higher Re are more likely
Studies have shown that turbulent mixing of the to be successfully scaled up, maintenance of a
solvent and antisolvent greatly impacts supersat- constant Re or constant jet velocities at the
uration homogeneity, allowing for more control antisolvent inlet does not guarantee scalability
of the PSD during PCA by tuning precipitation between laboratory and pilot plant batches
kinetics (nucleation and growth rates) (Jarmer (Jarmer et al. 2006). One criterion that enables
et al. 2003). Primary particle sizes ranging from process scalability is the maintenance of a con-
200 to 1000 nm for poorly water-soluble drugs stant energy dissipation rate in the nozzle. Nozzle
and as low as 50 nm for water-soluble molecules design significantly influences the propagation of
have been achieved using these technologies secondary nucleation mechanisms and thus
(Table 12.3) (Gupta 2006). impacts energy dissipation rates during solvent
Scalability of the PCA technology has been atomization. Another option to achieve scalability
demonstrated for the production of paracetamol is to target a constant suspension density and
particles at laboratory scales (1–8 10–4 kg/s residence time within the mixing chamber by
CO2 + ethanol + paracetamol flowrates) to small adjusting solvent flow rates through the nozzle,
manufacturing plant scales (0.9–1.5 10–2 kg/s which maintains mixing quality and, thus,
CO2 + ethanol + paracetamol flowrates) (Baldyga promotes comparable nucleation and growth
et al. 2010). In terms of batch sizes, 1 kg rates. When either of these conditions was met,
nanoparticles/day have been produced at the PLLA particles with similar PSDs were obtained
pilot plant scale using PCA (Gupta 2006). How- at both laboratory and pilot scales of production
ever, it is important to note that strategies for (Martin and Cocero 2008; Jarmer et al. 2006).
scaling up PCA processes differ when operating When operating in the complete miscibility
under subcritical or supercritical regimes. Sub- regime, PCA precipitation of amoxicillin
critical operating conditions exhibit higher conducted at both laboratory and pilot plant
sensitivities to certain parameters, such as nozzle scales yielded very similar results, in terms of
design. PCA precipitation of PLLA at both particle size and morphology, regardless of
Table 12.3 Drug nanoparticles produced by PCA

Atomization Particle
Drug Solvent P (bar) T (K) conditions size (nm)
Albumin (Bustami et al. 2000) Water/ethanol N/A N/A Coaxial nozzle 50–500
Amoxicillin (Reverchon et al. 2000, N- 150 313 Coaxial nozzle with 120–1200
2003b; Reverchon and Della Porta Methylpyrrolidone coaxial injector
1999)
Gentamicin/PLA (Falk et al. 1997) Methylene 85 308 US nozzle, vibrating 200–1000
chloride at 120 kHz
Hydrocortisone (Weber et al. 1999) Dimethyl 100 308 600
sulfoxide
Ibuprofen (Weber et al. 1999) Dimethyl 100 308 500–1000
sulfoxide
Insulin (Bustami et al. 2000) Water/ethanol Coaxial nozzle 50–500
Naltrexen/l-PLA (Weber et al. Methylene 85 308 US nozzle, vibrating 200–1000
1999) chloride at 120 kHz
Nicotinic acid (Falk et al. 1997) Ethanol Coaxial nozzle 400–750
RhDNase (Hanna and York 1998) Ethanol Coaxial nozzle 50–500
Salbutamol (Bustami et al. 2000) Methanol/acetone 100 333 Coaxial nozzle 500
Naloxone/l-PLA (Hanna and York Methylene 85 308 US nozzle, vibrating 200–1000
1998) chloride at 120 kHz
Dexamethasone phosphate (Falk Methanol 102 313 US nozzle, 175
et al. 1997) power ¼ 90 W
Griseofulvin (Thote and Gupta Dichloromethane 96.5 308 US nozzle, 310–510
2005) power ¼ 90–180 W
Griseofulvin (Chattopadhyay and Tetrahydrofuran 96.5 308 US nozzle, 200–280
Gupta 2001a) power ¼ 120–180 W
Lysozyme (Chattopadhyay and Dimethyl 96.5 310 US nozzle, 190–730
Gupta 2001a) sulfoxide power ¼ 12–90 W
Lysozyme (Rodrigues et al. 2009) Ethanol 180–250 318–333 Coaxial nozzle 100–5000
Tetracycline (Chattopadhyay and Tetrahydrofuran 96.5 310 US nozzle, 110–270
Gupta 2001b) power ¼ 30–120 W
Quercetin (García-Casas et al. Acetone 90 313.15 Nozzle diameter 84–145
2017a) 100 μm
Rifampicin (Djerafi et al. 2017) Ethyl acetate 100 308.15- N/A 190–230
333.15
Curcumin (Machmudah et al., 2020) Acetone/ethanol 8–12 313–333 Coaxial nozzle 50–125
US ultrasonic. Data adapted from Gupta (2006)
nozzle design and residence time in the precipita- precipitation vessel. This subcritical operating
tion vessel (Martin and Cocero 2008; Reverchon condition resulted in ~200-μm acetaminophen
et al. 2003b). The same trends were observed in a particles. The large particle sizes were attributed
study by Wubbolts et al., where acetaminophen to the growth of nucleated crystals in the solvent-
and ascorbic acid particles were produced by rich phase at the bottom of the vessel. In contrast,
PCA under subcritical versus supercritical an ascorbic acid–ethanol–CO2 system under
conditions (Wubbolts et al. 1999). When an supercritical conditions yielded ~1–5-μm
acetaminophen–ethanol solution was atomized particles, in which particle size was virtually
into CO2 under subcritical conditions, the insensitive to temperature and pressure changes
droplets did not fully evaporate and a solvent- while in the supercritical regime. Recently, math-
rich region was observed at the bottom of the ematical modeling was developed to predict the
particle size and the nucleation rate of solute in the organic solvent resulted in larger
nanoparticles produced by the SAS process for particles (Reverchon et al. 2007). In fact,
laboratory scale-up. In the SAS process, the flow Fig. 12.9 shows that, for the entire range of
and jet dispersion of organic solvents in compounds that were tested, average particle
pressurized CO2 were studied by their hydrody- sizes scaled linearly with the relative concentra-
namic behavior to apply to a computational tion of solute in the feed organic solvent, CR, for a
model of fluid dynamics. Since different produc- given operating temperature, pressure, and mole
tion scales affected the particle precipitation on fraction of antisolvent, where CR ¼ Cdrug/Ceq and
the SAS process, this model was used to predict Cdrug is the concentration of solute in the organic
particle size and nucleation rate of different pro- solvent. This linear relationship between feed
duction scales. Moreover, the computational drug concentration and particle size indicates
model of fluid dynamics incorporated with that the particle sizes produced by PCA depend
mixed suspension-mixed product removal primarily on the differential between the solute
(MSMPR) crystallizer could be the method to concentration in organic solvent and the saturated
understand how the nanoparticles’ precipitation concentration, not necessarily on the properties of
by SAS happened (Cardoso et al. 2019). a specific solute. Additionally, wider PSDs were
Reverchon et al. further investigated the span observed for higher solute concentrations.
of particle properties produced by PCA when Extrapolation of the linear relationship between
operating under supercritical conditions particle size and relative solute feed concentration
(Reverchon et al. 2007). More than suggests that the smallest average diameter of
20 compounds, spanning a wide range of particles produced by PCA is 45 nm, which is in
materials including superconductor and catalyst accordance with what has been observed in the
precursors, dye pigments, polymers, and literature. The smallest average particle sizes
pharmaceuticals, were examined in this study reported for PCA processes are on the order of
(Reverchon et al. 2007). Nanoparticles were 40–50 nm for several compounds, including lyso-
formed only under supercritical conditions and zyme, rifampicin, and polyvinyl alcohol (PVA).
when the solute was virtually insoluble in the Growth of particle sizes from systems with higher
antisolvent–solvent mixture. In agreement with feed solute concentrations may be attributed to an
previous studies, particle size was not dependent increased concentration of nuclei, which increase
on nozzle design for these experiments, which collisions rates. In the cases where the SC fluid is
were operated at supercritical conditions (Martin not a strong antisolvent, it was reported that
and Cocero 2008; Reverchon et al. 2003a, b; another driving force for recrystallization could
Wubbolts et al. 1999). Rifampicin and ethyl cel- be obtained by operating in non-isothermal
lulose were coprecipitated to form submicron- conditions (e.g., solution warmer than SC fluid).
sized particles. The formulation was prepared by Indeed, due to solubility increasing with temper-
SAS. A solvent mixture including ethyl acetate ature, a higher supersaturation level was achieved
and dimethyl sulfoxide prevented crystalline for- when solution and SC fluid were mixed
mation and improved the encapsulation efficiency (Erriguible et al. 2013). Moreover, the total con-
of coprecipitated rifampicin. Moreover, the centration of drugs in the mixture affected the
coprecipitated particles, which showed nanoparticle size. For example, zein and diclofenac
sized at 190–230 nm, maintained amorphous mixtures that were prepared with different
morphology for up to 6 months after storage at concentrations (5–50 mg/ml) produced different
room temperature and 75% relative humidity particles, including microparticles, sub-micron
(Djerafi et al. 2017). However, the initial solute particles, and nanoparticles. At the same temper-
concentration in the organic solvent did influence ature and pressure condition of the SAS process,
final particle size. Increased concentrations of the the lower the total concentration that was used,
Fig. 12.9 Mean particle 200

diameter, as a function of
relative solute
175
concentration in feed, CR
(CR ¼ Cdrug/Ceq), of
Mean particle diameter (nm)

various materials, including 150
metal acetates,
pharmaceuticals, polymers, 125
and dye pigments,
processed by PCA under
supercritical conditions 100
(P ¼ 150 bar and
T ¼ 40 C). Reprinted from 75
Reverchon et al. (2007).
Copyright 2007 with
50
25
0
0.00 0.05 0.10 0.15 0.20
CR
the smaller the particles that were obtained. How- polygonal sheets and elongated rectangular
ever, adding zein to formulations prefers to pro- prisms were demonstrated when using DMFA
duce microparticles and sub-microparticles than and methanol, respectively. Interestingly, the
nanoparticles. As a result, a diclofenac concentra- particles had a higher dissolution rate of
tion of 20 mg/ml demonstrated an ability to pro- 20–30% compared with the original form
duce nanoparticles. Furthermore, the drugs and (Kudryashova et al. 2018).
excipients ratio also affected particle size. The The PCA manufacturing technique coupled
increasing ratio of zein led to producing larger with an appropriate formulation (usually an amor-
particles. In terms of temperature effects, increas- phous state stabilizer) can change the crystalline
ing the temperature of the SAS process also state of the drug (Lim et al. 2010). Indeed, amor-
decreased microparticles to sub-micron particles. phous solid dispersion nanoparticles of sirolimus,
The optimum conditions that produced PVP, and surfactant were produced, and they
nanoparticles were 0:1 of zein/diclofenac (w/w), exhibited improved solubility, dissolution
total drug concentration of 20 mg/ml, 40 C, and properties, and stability. These results were con-
90 bar. The obtained nanoparticles demonstrated firmed in vivo in mice where enhanced bioavail-
amorphous morphology and particle size of ability of sirolimus nanoparticles was observed
138 nm (Franco et al. 2018). Different organic (Kim et al. 2011). HPMC/PVP was shown to
solvents used in the SAS process also affected the decrease the dissolution rate of amorphous
size and morphology of particles. telmisartan due to a gel layer formation; therefore,
Sub-microparticles of moxifloxacin were devel- a balance between amorphous state stabilization
oped by the SAS process with different solvents and dissolution rate must be defined (Park et al.
including N,N-dimethylformamide (DMFA), 2013). Recently, coprecipitation of hydrochloro-
methanol, and acetic acid. This technique pro- thiazide and PVP was produced by the PCA pro-
duced particle sizes in the range of 470–900 nm. cess. The nanoparticles obtained from this
The smallest particles with irregular, rounded process demonstrated amorphous morphology,
corner shapes were prepared by acetic acid, spherical shape, and size of 52.3 nm without
while the largest particles were produced by extensive agglomeration. Coprecipitated hydro-
DMFA. Different morphologies including chlorothiazide nanoparticles showed an increase
in the dissolution rate up to 3.2 times compared the solvent droplet initially swells due to the
with the unprocessed hydrochlorothiazide. Drug diffusion of antisolvent into the droplet. As the
concentration and solvent also affected particle pressure in the system is increased, the lifetime of
size. The lower drug concentration (10 mg/ml) the solvent droplet decreases because the droplet
decreases particle size to 52.3 nm compared shrinks as the CO2–solvent mixture evaporates to
with 173.2 nm of the high concentration (30 mg/ induce precipitation. However, as the system
ml). Moreover, the increasing ratio of acetone/ tends toward near-critical conditions, the lifetime
DMSO as solvents showed a reduction of particle of the solvent droplet increases drastically
size (Park et al. 2017). Rossman et al. because CO2 diffusivity tends toward zero near
demonstrated that paracetamol crystal polymor- the critical point. Longer droplet lifetimes may
phism could be modified by varying the ethanol/ lead to larger particle sizes because droplet coa-
acetone content in the drug solution. It was also lescence, and thus particle growth, is more likely.
found that varying the solvent led to primary or Because distinct droplets do not form under
secondary structure of paracetamol. Low levels of supercritical conditions, a hypothetical interface,
supersaturation led to larger crystals due to the based on the density gradient between the
prolonged crystal growth phase (Rossmanna et al. solvent-rich and antisolvent-rich regions, was
2013). assumed in the model. Modeling results indicated
PCA processing was also successfully used to that solvent droplets would swell if the density of
produce sub-micron co-crystals of several drug the organic solvent was higher than that of the
models, demonstrating the ability of this tech- antisolvent. Likewise, the solvent droplets would
nique to rapidly screen pharmaceutical shrink if the solvent density was lower than that
co-crystals (Padrela et al. 2010; Djerafi et al. of the antisolvent. The extent of droplet swelling
2017). or shrinking is dependent on the system’s temper-
In order to gain a more fundamental under- ature and pressure, as it affects density and diffu-
standing of how different operating conditions sivity differences between solvent-rich and
influence particle properties, several theoretical antisolvent-rich domains. In systems near their
models have been developed to describe particle critical point, solvent droplets undergo greater
formation and growth in the PCA process. Many swelling and experience longer lifetimes and are
of the models focus on calculating the rate of more sensitive to operating conditions than
mass transfer of antisolvent into the solvent systems far from the critical point. Elvassore
phase because this is believed to be a key factor et al. expanded upon Werling and Debenedetti’s
dictating particle size and morphology. Werling model by including the effects of the solute on the
and Debenedetti proposed a model for two-way diffusivity and density of the SCF into the mass-
mass transfer between a droplet of organic solvent transfer calculations (Elvassore et al. 2004). Fur-
and compressed antisolvent that accounts for both thermore, the swelling test of drugs indicates the
subcritical and supercritical conditions (Werling drug solubility in each solvent. Thus, the swelling
and Debenedetti 1999, 2000). The model assumes test can be used to select an appropriate solvent
that the droplet of organic solvent is stagnant; for the PCA process (Park et al. 2017). The
thus, only mass transfer by diffusion is consid- assumption of a stagnant droplet of organic sol-
ered. Machado et al. suggested that the critical vent is maintained, and the diffusion flux in the
point of solvent–CO2 system or solute–solvent– solute–solvent–antisolvent system was calculated
CO2 system was a significant factor for the pre- using Maxwell–Stefan relationships. In this
cipitation process. High mass transfer and droplet model, slowly diffusing solutes, such as
supersaturation generally succeeded at a super- polymers, were found to increase droplet
critical phase. Thus, coprecipitation of the lifetimes by as much as one order of magnitude
particles preferably occurred at temperature and for high solute concentrations, compared to
pressure that were higher than the critical point solutes with faster diffusivities. The extent of a
(Machado et al. 2018). For subcritical conditions, solute’s solubility in the solvent–antisolvent
mixture also influenced the particles’ CO2 (730 kg/m3) decreased the average diameter
morphologies, as the evolution of the precipita- of hydrochlorothiazide/PVP nanoparticles to
tion front was found to be significantly different 154.2 nm, while a higher density (830 kg/m3)
for highly soluble and poorly soluble compounds provided an average diameter at 205.8 nm (Park
(Elvassore et al. 2004). In a study of et al. 2017). Reverchon and De Marco proposed
coprecipitated anthocyanin and PVP by the PCA an explanation of morphology and particle size
process, the lower polymer concentration induced for differentiating nanoparticles and spherical
the droplet expansion process. As a result, the microparticles. For instance, they explained sur-
saturation and precipitation of solute occur face tension vanishing at supercritical conditions
slowly, and smaller particles are obtained. More- and liquid jet breakup; two precipitation
over, the PVP concentration not only affected mechanisms in competition influenced the
particle size but also affected particle morphol- morphologies and final particle size. They
ogy. The lowest concentration of PVP produced demonstrated that two mechanisms were involved
spherulite-shaped particle in layers, while the in crystal formation: (1) droplet drying followed
highest concentration of PVP produced irregular by fast crystallization, which led to spherically
particles on a plain surface (Machado et al. 2018). shaped crystals; and (2) precipitation from an
Perez de Diego et al. proposed a model that expended liquid, which led to different
accounted for the convective motion of CO2 morphologies and particle size depending on the
(Perez de Diego et al. 2006). Martin et al. has interaction with the solvent. This knowledge
adapted the mass-transfer model developed by allows the selection of the particle size of the
Werling and Debenedetti (1999, 2000) to simu- precipitated particles (Reverchon and De Marco
late the formation of protein particles by PCA 2011). Additionally, it is important to design
(Martin et al. 2007). More recently, a numerical systems away from the critical point of the
model utilizing computational fluid dynamics antisolvent because the near-zero diffusivities at
(CFD) calculations (Martin and Cocero 2004) this condition lead to droplets with longer
more accurately modeled supercritical systems lifetimes, which have a propensity to result in
using a turbulent, gaseous plume to simulate the larger particle sizes. However, current models
organic feed stream, instead of the hypothetical are still not able to universally quantify the depen-
spherical droplet used by Werling and dence of particle size on process parameters for a
Debenedetti (1999, 2000). While each new range of drug–solvent–antisolvent systems. As
model includes an additional degree of the PCA mentioned previously, multiple interactions
process’s complexity to impart further insight, all within the system (drug–solvent, solvent–CO2,
of the models express similar trends. Droplet and drug–solvent/CO2 solution) significantly
lifetimes are shorter for supercritical systems affect thermodynamics, hydrodynamic, mass
than subcritical systems, and shorter growth transfer, and mixing and precipitation kinetic
periods are more likely to lead to smaller particle behavior, thus making it difficult to generalize
sizes. When operating in the supercritical condi- results for a wide range of systems.
tion, the most important mechanism affecting
final particle size is primary nucleation, and thus
process parameters that facilitate more rapid and 12.2.3 Rapid Expansion
higher nucleation rates tend to form smaller of Supercritical Solutions (RESS)
particles. Furthermore, both at subcritical and
supercritical conditions in the PCA process In contrast to GAS and PCA processes, the rapid
showed the same correlation between the density expansion of supercritical solution (RESS) pro-
of CO2 and particle size. The results indicated that cess utilizes the SCF as a solvent, not an
the higher the density that was used, the smaller antisolvent. The solute is dissolved directly into
the particles that were produced. For instance, at the SCF phase in the extraction unit. Then, the
the supercritical condition, a lower density of system is depressurized across a nozzle into a
Fig. 12.10 Schematic of Preheater

the RESS process.
Schematic adapted from Heated
Martin and Cocero (2008) nozzle
CO2
pump Precipitator
Saturator
CO2 reservoir Vent
collection chamber at atmospheric conditions. atomization, termed as pre-expansion tempera-

The sudden depressurization causes evaporation ture (Tpre-exp) and pressure (Ppre-exp), respectively,
of the SCF, resulting in a significant reduction in as well as post-expansion temperature (Tpostexp)
solvent power, and thus promotes rapid nucle- and pressure (Ppostexp). These conditions deter-
ation and precipitation of the solute. As with the mine the process path along the pressure–temper-
other particle formation techniques discussed pre- ature (P–T ) diagram for the SCF. The P–T
viously, additional excipients may be dissolved in diagram for CO2 is shown in Fig. 12.11.
the SCF, typically CO2, to produce composite Depending on initial P–T conditions, the expan-
particles of drug and excipients (Turk 2009). A sion pathway may intersect the vapor–liquid sat-
schematic of the RESS process is shown in uration line, resulting in significant changes in
Fig. 12.10. Intense atomization of the drug–CO2 particle morphology (Martin and Cocero 2008).
stream is desirable to achieve nanoparticles from When the expansion path intersects the solid–
the RESS process. Therefore, depressurization of liquid saturation line, solid, frozen CO2 forms
the CO2 feed stream from the nozzle is designed during atomization, requiring the nozzle to be
to be extremely rapid, with typical CO2 flow rates heated during operation to prevent clogging. Noz-
exiting the nozzle at the speed of sound, creating zle design is another parameter that has reportedly
supersaturation levels on the order of 105–106 influenced final particle properties, as the geome-
within a time frame of 10–6–10–4 s (Debenedetti try of the nozzle influences the timescale over
et al. 1993). The intense turbulence generated by which depressurization occurs and, thus, the
rapid depressurization of CO2 distributes the degree of atomization (Martin and Cocero 2008;
newly generated supersaturation regions almost Rogers et al. 2001a). Recently, the operating
instantaneously and homogeneously throughout conditions of the RESS process, including Ppre-
the fluid, which facilitates the production of exp, Tpre-exp, Dnoz, and spray distance, were
small particles with narrow PSDs. This rapid dis- investigated for their effect on the particle size
sipation of energy is highly endothermic, and thus of nanoparticles. The higher the Ppre-exp was used,
the nozzle is generally heated to prevent freezing the lower the mean particle size was obtained.
of CO2 during atomization, which can cause Since Ppre-exp enhances the solubility of drugs in
clogging. scCO2 that provides higher supersaturation, the
Several process parameters of RESS that have nucleation rate is improved. Thus, this condition
been reported to affect final particle shows the increase of the particle growth time
characteristics include the temperature and pres- inside the nozzle. In terms of Tpre-exp effects,
sure in the extraction unit and the temperature and higher temperature leads to a decrease in mean
pressure of the SCF–drug solution just before particle size. Besides, temperature also affects the
Fig. 12.11 P–T diagram P

of CO2. Dashed lines
illustrate pathways that may
be taken during CO2 7.3 MPa
depressurization from the
nozzle in PCA. Reprinted
from Martin and Cocero
(2008). Copyright 2008
Elsevier 0.5 MPa
211 K 304 K T
density of CO2. The higher the Tpre-exp was used, To date, however, a definitive relationship
the lower the density of CO2 was shown. As a between experimental process parameters and
result, the sublimation vapor pressure was particle properties has not been established and,
increased, and the solute solubility was improved. in some cases, experimental results have been
Moreover, the higher temperature could enhance inconsistent. For example, for a given Ppre-exp,
nucleation and crystallization rate because the an increase in Tpre-exp from 350 to 425 K resulted
higher temperature enhances the drug concentra- in an increase in the size of benzoic acid particles
tion at the saturated condition (Sodeifan et al. produced by RESS (diameter increased from 0.2
2019a). Furthermore, the nozzle types and nozzle to 1.3 μm), while the particle size of cholesterol
diameters also appear as critical parameters in the remained unchanged (~0.25 μm) (Fig. 12.12).
RESS process. A capillary nozzle or a laser- Similarly, an increase in Ppre-exp resulted in a
drilled orifice nozzle is typically used to prepare decrease in particle size of benzoic acid, while
fine particles. For instance, a laser-drilled orifice the size of cholesterol particles again remained
nozzle with different diameters affected the parti- essentially constant. Numerous studies have been
cle size produced by the RESS process. The conducted to better understand which process
reduction of the nozzle diameter parameters most strongly and consistently influ-
(0.45–0.35 mm) increased the supersaturation ence final particle size. The RESS process com-
and nucleation rate at the nozzle tip. As a result, monly produces particles in the 1–5-μm size
the smaller precipitated particles were produced. range, although submicron particles have been
Spray distances also affect particle characteristics produced under specific operating conditions
because the nucleation and particle growth pro- (Gupta 2006). Another example, ipriflavone,
ceed in the post-expansion phase. For example, also showed the smaller particles were obtained
the spray distance increasing 1–3 cm did not while using process parameters including higher
affect particle size, while the spray distance of Ppre-exp and Tpre-exp, lower Tpost-exp, and spraying
5 cm produced the smaller particles significantly. distance. The RESS process demonstrated an
A reason was that a sufficiently long distance of ability to reduce particle sizes and alter particle
spraying provided a longer time to break apart shapes of ipriflavone particles. The processed
large particles into fine particles. However, when particles had a smaller size as low as 4.4 μm
increasing to 7 cm, the larger particle size was compared with 30.9 μm of the unprocessed
obtained. This phenomenon reportedly occurred particles (Wang and Su 2020). Several RESS
because the exceedingly long distance prolonged studies have been highlighted in Table 12.4.
the residence time to increase particle growth. Clearly, process temperatures and pressures and
Thus, spray distance appeared as a parameter nozzle geometry significantly influence particle
that should be optimized (Sodeifan et al. 2019a). size and shape, and, in some cases, morphology
(Turk and Bolten 2010). For instance, the
Fig. 12.12 Influence of

pre-expansion conditions
on particle sizes of benzoic
acid and cholesterol
prepared by RESS.
Reprinted from Turk
(2000). Copyright 2000
Elsevier
increasing of pre-expansion temperature and noz- • Influence of Ppre-exp: An increase in Ppre-exp

zle length created a higher particle size and parti- typically leads to smaller particle sizes because
cle size distribution (Bagheri et al. 2019). a higher pressure results in an increased mass
Relatively small adjustments to just one of these flow rate of CO2, which decreases the resi-
operating conditions may significantly impact dence time of the particles in the expansion
particle diameter by an order of magnitude, as chamber, reducing the time for particle
well as completely alter the particle shape from growth. The reduction in residence time also
a sphere to a needle shape, as seen in the cases for facilitates the production of more spherical
salicylic acid and griseofulvin particles produced particle shapes by limiting the time available
by RESS (Table 12.4). Based on reports from the for additional growth along one axis.
literature, including those listed in Table 12.4, • Influence of nozzle: Nozzles with smaller
several trends in processing conditions have length-to-diameter (L/Dnoz) ratios have been
been identified to facilitate nanoparticle produc- found to produce smaller particles, as larger
tion (Turk 2009). nozzle diameters facilitate increased CO2 mass
flow rates (for a given Tpre-exp and Ppre-exp).
Additionally, nozzles with smaller L/Dnoz
• Influence of Tpre-exp: An increase in Tpre-exp
ratios allow for the pressure drop to occur
typically leads to larger particle sizes. For a
closer to the free jet (Rogers et al. 2001a;
given operating pressure, even though ele-
Weber et al. 2002). As the L/Dnoz ratio is
vated temperatures may increase drug
increased, there is an increased propensity for
solubilities and thus increase supersaturation
an initial burst of particle nucleation to occur
and nucleation rates (Liu and Nagahama
near the nozzle exit. A second round of nucle-
1996), the higher temperatures also tend to
ation occurs upon full expansion of the SCF,
increase turbulence within the mixing cham-
resulting in larger particles as well as broader
ber, leading to higher instances of particle
PSDs. Typically, nozzle diameters range from
coagulation (Franklin et al. 2001). The
10 to 50 μm i.d., and length-to-diameter ratios
increased rate of particle coagulation appears
range from 5 to 100 (Young et al. 2000).
to outweigh the benefits of enhanced nucle-
ation rates achieved by elevated Tpre-exp
conditions. It is important to note that these reported trends
reflect a considerable portion of the studies in the
Table 12.4 Drug particles produced by RESS

Tpre-exp Tpost-exp Particle size Nozzle parameters L/
Drug P (bar) (K) (K) (μm)a Dnoz, Dnoz (μm)
Aspirin (Domingo et al. 1997) 160–200 403 NRb Nonspherical: 5, 40
2–5
Caffeine (Ksibi et al. 1995) 150 380 300 Needles: 3–5/1 20, 220
150 380 350 Needles: 15–20/ 20, 220
1
Ibuprofen (Kayrak et al. 2003; 150 361 298 Nonspherical: 44.4, 180
Charoenchaitrakool et al. 2000) 2–9
190 308 298 Nonspherical: 20–40, 50
2.5–5
Cholesterol (Turk 2009) 200–300 353–422 298 Nonspherical: 7.8, 45
0.2–0.3
Salicylic acid (Reverchon et al. 1993; 200 373 263–273 Spheres: 1–5 20, 40
Turk and Lietzow 2008) 200 373 293–303 Needles: 5–30/ 20, 40
1–3
200 328 298 Spheres: 1, 50
0.13–0.23
Griseofulvinc (Reverchon and Pallado 200 423 298 Spheres: 0.9–1.4 20, 40
1996; Turk et al. 2002) 200 323 298 Needles: 13–36/ 20, 40
1.0–1.3
200 348–418 298 Spheres: 0.25 1, 50
β-sitosterol (Turk et al. 2002) 200–300 348–418 298 Spheres: 1, 50
0.18–0.23
Paclitaxel (Yildiz et al. 2007) 150–250 323 283 Nonspherical: 70, 50
0.3–2.8
Naproxen (Turk 2009) 200 363 NRb Shape not NRb
reported: 0.66
Fenofibrate (Hiendrawan et al. 2014) 200 308 NR 3.04 Dnoz: 200
Raloxifene (Keshavarz et al. 2012) 177 323 NR Spheres: 0.016 Dnoz: 30
Ipriflavone (Wang et al. 2020) 100–200 453–473 283–303 Shape not Dnoz: 50
reported:
4.4–8.6
Coumarin-7 dye (Sodeifan et al. 150–330 308–338 NR Spheres: Dnoz: 150–450
2019a) 0.224–0.0214
Disulfiram (Tang et al. 2020) 250 328.15 NR Needle: NR
0.27–1.53
Unless otherwise noted, CO2 was the solvent
aFor needle-shaped particles: length/diameter. For spherical particles: diameter. Nonspherical refers to particles that do
not possess an aspect ratio typical of needle-shaped particles but deviate significantly from an aspect ratio of unity (i.e.,
rectangle with aspect ratio ~2). Sizes for nonspherical particles correspond to effective diameters
b
NR indicates a value that was not reported
c
Solvent was CHF3
literature but are not exclusively observed. In response to the seemingly conflicting exper-
Deviations from these observed trends, as in the imental results surrounding the RESS process,
case of cholesterol particles produced by RESS, several theoretical models have been postulated
have been associated with extremely low solute to gain fundamental knowledge about the RESS
solubilities in SCF and/or solutes that signifi- process in order to better target optimal process
cantly influence the surface tension of the SCF parameters suitable for nanoparticle production.
(Turk 2000). Many of the models focus on the expansion of the
SCF in the nozzle to qualify the impact of nozzle using a modified definition for supersaturation,
design on final particle characteristics. The S, which was adjusted to account for the highly
models show that sonic velocities are achieved nonideal behavior of SCF by including fugacity,
at the nozzle outlet, and the resultant supersonic f, as a thermodynamic correction factor.
jet exiting the nozzle immediately experiences a
Cdrug f T, P, C drug
steep drop in pressure and temperature, causing S¼ : ð12:7Þ
solute precipitation. Thus, nucleation occurs pri- C eq f T, P, Ceq
marily during free jet expansion. Calculations
estimate that nuclei formed in the free jet are as Helfgen et al. applied the modified supersatu-
small as 5–10 nm for poorly water-soluble drugs ration term in conjunction with the general
(Gupta 2006; Reverchon and Pallado 1996; Turk dynamic equation for aerosols, commonly used
et al. 2002). However, intense turbulence within to describe particle growth in aerosols (Pratsinis
the supersonic free jet often results in significant 1988) to predict particle nucleation and growth
coagulation between particles before the SCF in rates in RESS. Results from the model indicated
the droplets completely evaporates (Franklin et al. that the majority of particle precipitation and
2001; Helfgen et al. 2003). Thus, controlling- growth took place in the free jet and that turbulent
expansion conditions may be tuned to facilitate coagulation in the free jet is the primary mecha-
SCF evaporation and minimize droplet coagula- nism of particle growth (Franklin et al. 2001;
tion. For example, expansion chamber geometries Helfgen et al. 2003). Relatively good agreement
that minimize the formation of turbulent eddies between the model and experimental results were
are desirable to lower the probability of particle demonstrated for the production of benzoic acid,
coagulation (Helfgen et al. 2003). Additionally, griseofulvin, and β-sitosterol by RESS (Helfgen
the introduction of an air flow jet into the expan- et al. 2003). While trends relating particle size to
sion chamber resulted in smaller particles by experimental parameters such as nozzle design
reducing the residence time of the drug particles and pre- and postexpansion conditions identified
in the expansion chamber (Helfgen et al. 2003). by various models have been in accordance with
Recently, a population balance model was devel- experimental observations, quantitative determi-
oped using the RESS process to produce small nation of nucleation and growth rates for a wide
particles with narrow particle size distribution. range of drug systems remains challenging
The numeric simulations, including mass, because reasonable values for some model
momentum, and energy balance, were parameters cannot be determined experimentally
investigated to understand hydrodynamic and must be assumed.
behaviors and particle formations. The driving Recently, Mullers et al. used RESS as a
force of the particle formation appears as super- method to combine micronization and co-crystal-
saturation. The high value of the supersaturation, lization in a single manufacturing step. Pure
which increases homogeneous nucleation, occurs co-crystals of ibuprofen and nicotinamide were
under low density in an expansion phase. More- obtained due to the very fast precipitation
over, the coagulation induces particle growth conditions and the absence of organic solvents.
resulting in higher particle size distribution The solubility difference between ibuprofen and
(Bagheri et al. 2019). In addition to the work nicotinamide in the supercritical fluid was a con-
focused on nozzle design, other models have cern because it influences supersaturation and
examined particle formation and growth within thus nucleation. As previously reported
an SCF. The theories used to describe particle (Vemavarapu et al. 2009), the authors stated that
growth in gaseous and liquid phases were also simultaneous precipitation of both components
found to be applicable, with minor adjustments, was plausible due to the high affinity of the
for supercritical precipitation processes. co-former for the drug compared to the solvent.
Debenedetti (1990) and Turk (2000) calculated The dissolution rate of ibuprofen was signifi-
nucleation rates achieved in the RESS process cantly increased and was explained by the higher
surface area (Mullers et al. 2015). Moreover, the coprecipitated particles because stabilizers are
RESS process converted the crystalline morphol- involved in the stabilization and monodispersing
ogy of unprocessed drugs to the amorphous mor- of drugs and polymers. For example, a small
phology of processed drugs resulting in the amount of PVP (0.5%) significantly improved
increase of the dissolution rate and bioavailabil- the % drug loading of disulfiram and disulfiram–
ity. For example, nanoparticles of disulfiram and copper complex by nearly 100%. However,
disulfiram–copper, which were coated with a increasing the concentration of polymer did not
polymer such as PVP or methoxy b-poly show significant improvement in drug loading
(L-lactide) 2000-poly (ethylene glycol) 2000 (Tang et al. 2020). However, this strategy is not
(mPLLA-PEG), were successfully produced by always recommended as it may lead to solubiliza-
the RESS process. The nanoparticles tion of the particles in the cosolvent. Additional
demonstrated amorphous morphology and methods to increase process yields and reduce
smaller particle size. However, disulfiram–copper particle coagulation for RESS-based techniques
nanoparticles coated with PVP of 2.0% are discussed in the next section.
demonstrated lower particle distribution com-
pared with the particles coated with mPLLA-
PEG of 2.0%. According to the physical
12.2.4 Modified RESS Processes
properties of the coated disulfiram nanoparticles
including smaller size, amorphous morphology,
RESS into Aqueous Solutions: Rapid Expansion
and higher dispersibility, the nanoparticles
from Supercritical to Aqueous Solution
showed a faster drug-release profile compared
(RESAS) and Rapid Expansion of a Supercrit-
with the unprocessed particles (Tang et al.
ical Solution into a Liquid Solvent (RESOLV)
2020). RESS does not require organic solvents,
does not involve milling, and may be operated at
To address the significant particle growth that
moderate temperatures (typically below 80 C).
occurs in the RESS process due to particle
However, the primary drawback of RESS is low
collisions during free jet expansion, the process
process yields. Most organic solids possess low
was modified by directing the atomized drug–
solubility in scCO2 due to the low polarizability
SCF solution into an aqueous solution to provide
of CO2. Therefore, large amounts of SCF are
a barrier against particle growth. This modified
required to produce relevant batch sizes. For
RESS process was coined RESAS, also known as
example, the solubility of griseofulvin in scCO2
RESOLV. In RESAS/RESOLV, the supercritical
is only 18 ppm. Therefore, the production of
solution is atomized through a nozzle directly into
18 moles (~6 kg) of griseofulvin by RESS
an aqueous solution containing a stabilizer, typi-
would require one million moles (~44,000 kg)
cally a surfactant. A schematic for the RESAS/
of CO2. In order to overcome the low throughput
RESOLV process is shown in Fig. 12.13. The
rates due to the low solubility of drugs in the SC
nozzle is placed below the surface of the aqueous
fluid, closed-loop recirculation of the fluid could
solution to promote intimate contact between the
be incorporated in the manufacturing process.
newly formed nuclei exiting the nozzle and the
Recovery of the resultant particles is also chal-
stabilizers dissolved in the aqueous media.
lenging, as efficient filtration is required to
Because the turbulent expansion of CO2 in a
remove such large volumes of solvent (Gupta
surfactant solution produces considerable
2006). To increase drug loading, extraction
amounts of foam, nitrogen is streamed above the
temperatures and pressures may be increased.
aqueous solution to disrupt the foam and facilitate
The addition of cosolvents, such as methanol,
drainage back into the bulk liquid phase (Young
acetone, and ethanol, to scCO2 has also been
et al. 2000).
used to increase drug solubility. Moreover, the
By atomizing the SCF stream into a surfactant
selection of polymers as stabilizers shows the
solution, particle growth in the free jet may be
potential to increase drug loading and yield of
Solution
Cell P
Prepressurization line
Carbon
dioxide
T
supply T
N2
Fig. 12.13 Schematic of the RESAS process; inset is a photograph of the spray of the CO2 solution stream expanding
through a tapered elliptical nozzle with a flow rate of 2.5 mL/min at 345 bar. Adapted from Young et al. (2000, 2003)
arrested by the rapid adsorption of surfactant et al. 2004, 2006), naproxen (64 nm when
molecules to the newly formed particle surfaces. stabilized by PVP40K) (Turk 2009), and paclitaxel
Young et al. demonstrated the ability of the (200–530 nm when stabilized by PVP40K or
RESAS process to produce ~500-nm particles of PVP360K) (Pathak et al. 2007). However, for
the poorly water-soluble drug, cyclosporin A these cases, the drug/polymer ratio was typically
(CsA), using Tween 80 as a stabilizer (Young <<1, ~0.08–0.2. To better understand how to
et al. 2000). In contrast, CsA particles produced efficiently increase the drug potency of RESAS
by RESS, where the scCO2 solution was sprayed particles while still maintaining submicron parti-
into air instead of a Tween 80 solution, were cle sizes, the critical processing parameters for the
3–50 μm in diameter. As a control, the scCO2 RESAS process were investigated by Young et al.
solution was also sprayed into water containing (2000, 2003, 2004). Experimental parameters
no surfactant to validate the role of Tween 80 in such as surfactant selection, temperature of the
impeding particle coagulation and growth. Resul- aqueous reservoir, and final drug concentration,
tant particle sizes ranged between 0.23 and in addition to the operating parameters known to
4.10 μm. Therefore, inhibited CsA particle influence particle properties in RESS, were varied
growth in the RESAS process is attributed to the to manipulate the efficiency of surfactant
rapid diffusion of Tween 80 to particle surfaces molecules to stabilize nanoparticles (Young
and its ability to provide steric stabilization to the et al. 2000). Nonionic surfactants, Pluronic F127
particles. (also known as poloxamer 407) and Myrj 52, in
The successful production of drug addition to Tween 80, were explored in efforts to
nanoparticles by RESAS/RESOLV has also stabilize CsA particles. CsA particles stabilized
been demonstrated for ibuprofen (40–80 nm in by Pluronic F127 and Myrj 52 were about twice
diameter when stabilized by Tween 80) (Turk as large (>840 nm in diameter) as those stabilized
2009) or polyvinylpyrrolidone (PVP40K) (Pathak by Tween 80 (500 nm in diameter) when
produced at similar operating conditions, rate at the nozzle occurs, leading to the produc-
emphasizing the importance of selecting tion of smaller particles. Increasing temperature
stabilizers with sufficient affinity for the drug decreases particle size because higher tempera-
particle surface and adequate chain length to pro- ture increases the pressure of solutes sublimation
vide steric repulsion. In contrast, a phospholipid- that enhances drug solubility in scCO2. As a
based surfactant produced CsA particles with a result, higher supersaturation improves nucle-
mean diameter of 220 nm, about half the size of ation rate and reduces particle size. Thus, smaller
the Tween 80-stabilized particles produced by particles were obtained at a higher temperature. In
RESAS at similar operating conditions. However, terms of the nozzle dimensions, a lower nozzle
higher amounts of phospholipid were necessary diameter provided a lower particle size because of
to stabilize the smaller CsA particles compared to higher pressure and temperature drop at the noz-
Tween, only achieving a drug/surfactant ratio of zle leading to improve supersaturation and nucle-
0.1 compared to 0.65 for Tween-stabilized ation rate. Thus, lower particle growth occurred
particles. In the case of phospholipids, the bulk (Sodeifian et al. 2019b).
of the surfactant arranges to form vesicles. The Moreover, modified-RESS process was devel-
aggregation number of surfactant molecules is oped by incorporating an auxiliary injection
much larger for vesicles than for micelle-forming device to continue the solution flow feeding
surfactants such as Tween, which explains the with scCO2 at the supercritical condition. The
lower drug/surfactant ratios observed for phos- results showed that this process had the potential
pholipid stabilizers. The temperature of the aque- to produce lonidamine nanoparticles size as small
ous reservoir is also a key parameter for the as 72.67 nm compared with the unprocessed par-
RESAS process, as it influences the surfactant ticle size that showed over 100 μm (Chen et al.
assembly and thus the rate at which the surfactant 2018).
is able to reach the particles’ surface. Phospho- Recently, to modify RESAS and Rapid Expan-
lipid stabilizers are especially sensitive to temper- sion of a Supercritical Solution into a Liquid
ature because vesicles tend to become rigid at Solvent (RESOLV) processes, an ultrasonic sys-
temperatures below 25 C. Hence, phospholipids tem was incorporated into processes, namely
are more effective stabilizers when heated to US-RESAS and US-RESOLV, respectively. The
higher temperatures and facilitate the stabilization rapid expansion was completed through the noz-
of smaller particles. Under optimized conditions zle inserted into a sample collector. The ultrasonic
(Taqueous bath ¼ 80 C, CsA concentration in system is also connected to the sample collector.
CO2 ¼ 54 mg/mL, CO2 flow rate through noz- Thus, the expansion of the supercritical CO2 solu-
zle ¼ 2.5 mL/min, and pressure drop across noz- tion occurred while sonicating the solution. At the
zle ¼ 345 bar), a phospholipid surfactant mixture optimum condition (30 MPa, 338.2 K, 5 mm), the
stabilized ~500-nm CsA particles (31% w/w smallest nanoparticles were obtained from
drug) at drug suspension concentrations up to US-RESAS with an average particle size of
5.4% w/w (Young et al. 2004). The increase in 26.0 nm compared with nanoparticles produced
drug suspension concentration resulted in slightly by RESAS that had the highest average particle
increased particle sizes, compared to the 220-nm size of 123 nm. In the US-RESAS process, the
CsA particles when suspension concentrations effects of temperature, pressure, nozzle diameter,
were held to 1.3% w/w (Young et al. 2004). and nozzle length on particle size demonstrated
Process parameters including temperature, pres- the same correlation as the RESAS process. Fur-
sure, nozzle diameter, and nozzle length also thermore, the power of the ultrasonic system also
affected particle size. Increasing pressure reduced affected the particle size because the ultrasonic
the average particle size of loratadine because system could break particle agglomeration. This
higher pressure increased the scCO2 density and phenomenon improved the homogeneity and
drug concentrations in the supercritical CO2. As a nucleation rate of nanoparticles. Thus, the highest
result, a higher supersaturation and nucleation power at 70 W/L produced the smallest
Fig. 12.14 (a) Schematic

of the RESS process and (b)
RESS-SC process. Circles
represent drug particles and
stars represent solid-
cosolvent particles.
Reprinted with permission
from Thakur et al. (Thakur
and Gupta 2005).
Copyright (2005) American
Chemical Society
nanoparticles compared with no ultrasonic primary limitation of RESAS, as in the RESS

incorporated (Sodeifian et al. 2019b). In terms process, is that the solute must possess moderate
of US-RESOLV, amiodarone hydrochloride solubility in an SCF.
nanoparticles were prepared with polymeric
stabilizers. The ultrasonic system was equipped
Rapid Expansion of Supercritical Solutions
with the sample collection part the same as
with Solid Cosolvents (RESS-SC)
US-RESAS. The results showed that
In the RESS-SC process, a cosolvent that
US-RESOLV produced the smaller nanoparticles
solidifies upon atomization from the nozzle is
at 47.9 nm compared with the unprocessed
used to enhance the solubility of solutes in
amiodarone particles (75.4 μm) and amiodarone
scCO2, as well as providing a barrier for coagula-
nanoparticles produced with the RESOLV pro-
tion in the free jet during scCO2 expansion
cess (101.3 nm). The US-RESOLV process
(Thakur and Gupta 2005). In contrast to RESS,
demonstrated an ability to alter the crystalline
where the nuclei tend to coagulate during free jet
morphology of the unprocessed particles to the
expansion, the excess amounts of solid cosolvent
amorphous morphology of the processed
added during the RESS-SC process surround the
particles. Moreover, incorporation of an ultra-
nuclei to create a physical barrier to reduce coag-
sonic system into the RESOLV process also
ulation. The cosolvent may be removed later by
improved the dissolution rate of amiodarone
lyophilization. A schematic representing the
hydrochloride compared with the RESOLV pro-
RESS-SC technique, in contrast to RESS, is
cess and was higher than the unprocessed
shown in Fig. 12.14.
amiodarone up to 13.2-folds (Sodeifian and
In addition to the typical operating parameters
Sajadian 2019).
that are important in RESS, clearly, the selection
The RESAS process was shown to success-
of the solid cosolvent is a key parameter in the
fully produce smaller particles of water-insoluble
RESS-SC process. The solid cosolvent must be
materials than was achieved by RESS due to
nonreactive with the drug and CO2, possess good
particle stabilization within an aqueous surfactant
solubility in scCO2, be in the solid state at the
solution. In the case of mild particle aggregation
nozzle exit, have a reasonably high vapor pres-
after RESAS precipitation, a high-pressure
sure to facilitate removal by sublimation, and,
homogenization step has been added to the end
preferably, be inexpensive since excess amounts
of the RESAS process to promote more uniform
are needed to maintain submicron particle sizes.
PSDs and to break up any aggregates that may
Thus far, menthol has been the most prevalently
have formed. This process train has been patented
used solid cosolvent for RESS-SC applications.
by RTP Pharmaceuticals Inc. and was later
Menthol is a natural product extracted from mint-
licensed by Baxter Healthcare Corporation for
flavored plants, possesses a melting point of
incorporation into their NANOEDGE technology
42 C, and satisfies all the criteria listed above.
(Hu et al. 2004; Keck and Mueller 2006). The
Menthol enhanced the solubility of the poorly
water-soluble drug griseofulvin 28-fold in effect of the mean particle diameter was the tem-
sc-CO2, enabling the production of 50–250-nm perature (Sodeifian et al. 2018). The RESS-SC
particles by RESS-SC, which is an order of mag- technique broadens the applicability of the
nitude smaller than those produced by RESS, at a RESS process to more drugs, as well as facilitates
28-fold increase in payload. Moreover, menthol, the production of higher payloads compared to
known as solid cosolvent for the RESS-SC pro- RESS. Yet, stability and reproducibility of the
cess, improved the solubility of an anticancer nanoparticles are a concern; Uchida et al. success-
drug. Letrozole nanoparticles, which was known fully overcame poor particle size and morphology
as a poorly water-soluble drug, were developed reproducibility occurring with menthol
by the RESS-SC process. This process produced co-solvent by replacing it with vanilline (Uchida
letrozole nanoparticles that showed higher solu- et al. 2015). However, not all drugs exhibit
bility of 7.1-folds than nanoparticles produced increased solubility with the presence of solid
without menthol in supercritical CO2 (Sodeifian co-solvents, and thus RESS-SC is not a universal
and Sajadian 2018). Lonidamine, which inhibits solution for all drug systems.
glycolysis in cancer cells, was improved to
increase the drug solubility by the RESS-SC pro- Particles from Gas Saturated Solution
cess using methanol as a solid cosolvent. The (PGSS) Process
study showed that this technique altered the phys- The PGSS process flow is similar to RESS but
ical state of crystalline to amorphous morphology differs in the case that CO2 does not act as a
without chemical compositions changes. As a solvent. In PGSS, the CO2 is dissolved in a
result, the dissolution rate of lonidamine melted solid, and the mixture is depressurized
nanoparticles was higher than that of the unpro- through a nozzle. Expansion of the dissolved
cessed lonidamine (Chen et al. 2018). Recently, CO2 results in intense atomization and cooling
the RESS-SC process with menthol reduced the of the molten solid and thus precipitation of
particle size of aprepitant nanoparticles to particles. This process is suitable for materials
23.0 nm compared to 25.6 μm of the original with a large solubility in CO2, such as PEGs and
aprepitant. The dissolution rate of aprepitant oils (Martin et al. 2010; Perrut et al. 2005).
nanoparticles was 8.2 folds higher than that of Benefits of this process are that it consumes less
the original aprepitant (Sodeifian et al. 2018). CO2 than the previously discussed SCF
Aminobenzoic acid (80-nm mean diameter) and technologies, it may be operated under moderate
phenytoin (120-nm mean diameter) particles have pressures (10–15 MPa), and solubility of the drug
also been produced using the RESS-SC process in the CO2 is not necessary to achieve high pro-
(Thakur and Gupta 2005, 2006a, b). Moreover, cess yields, as the drug can be dispersed in the
the mean particle diameter of coumarin-7 was melted solid (Martin et al. 2010; Perrut et al.
reduced to 21.37 nm at the supercritical tempera- 2005). Therefore, this precipitation process is
ture at 338 K, the pressure at 330 bar, nozzle optimal for polymer encapsulation and is capable
diameter at 0.15 mm, and spraying distance at of particle micronization, typically yielding
5 cm (Sodeifian et al. 2019a). Besides, tempera- micron-sized particles, larger than achieved by
ture, pressure, solid cosolvent, and spray distance RESS (~3–60 μm for theophylline and PEG
also influenced the size and morphology of 6000) (Martin et al. 2010; Rodrigues et al.
particles. The temperature effect showed the 2004). Precipitated phenylalanine particles pro-
highest impact of size-reduction of nanoparticles. duced by the mixture of enteric polymer Eudragit
For example, as the temperature was increased L100, supercritical carbon dioxide, and ethanol
from 318.2 to 338.2 K, the mean particle diameter with PGSS process showed size-reduction from
decreased from 107.6 to 20.6 nm (Sodeifian and 200.9 20.5 to 140.6 40.1 μm. The drug
Sajadian 2018). In another study of RESS release profile presented sustained at low pH
parameters, the greatest effect on the mean parti- conditions (pH 2.1) and higher drug release at
cle size was the nozzle diameter, while the least high pH conditions (pH 6.8 and 8.0) (Tokunaga
Table 12.5 Comparison of micronization techniques using compressed fluids

Organic
Temperature solvent Compressed Compressed fluid Yields poorly water-
Process ( C) required fluid as solvent as antisolvent soluble nanoparticles
GAS 25–80 Yes No Yes Yes
PCA/SAS/ 25–80 Yes No Yes Yes
ASES/SEDS
RESS 100 No Yes No Yes
RESAS 25–80 No Yes No Yes
PGSS ~25–40 No No No No
Adapted from Rogers et al. (2001a) and Perrut et al. (2005)
et al. 2021). Theoretical models that describe the predominantly descriptive, rather than predictive,
PGSS process, which were built upon existing with the conclusions heavily dependent upon the
RESS models, suggest that the larger particles materials and conditions of that specific study.
produced by PGSS compared to RESS are due The inability to develop generalized models that
to significant coagulation in the free jet region accurately predict final particle sizes with respect
(Martin and Cocero 2008; Li et al. 2005). to different operating parameters over a wide
Recently, PGSS has been used as a range of drug systems is due to the simultaneous
manufacturing technique for lipid-based influence of the operating parameters on multiple
microparticles in order to improve their dissolu- particle formation and growth factors, such as
tion. Fenofibrate solid dispersion in Gelucire thermodynamics, fluid mechanics, mass transfer,
50/13 was obtained and exhibited an improved and mixing and precipitation kinetic behavior.
dissolution profile (Pestiau et al. 2015). More- Despite case-specific results, general attributes
over, the PGSS process could improve drug of the different processes may be identified to
performances. For example, lidocaine-loaded provide general guidelines as to the capabilities
solid lipid microparticles (SLMPs) were proof each process. Comparisons between the differ-
duced by glyceryl monostearate with the PGSS ent SCF micronization options are shown in
process. In this most recent study, SLMPs Table 12.5.
showed a sustained release of the dissolution pro- Generally, GAS processes produce larger
file that could prolong the duration of local anes- particles than PCA processes, primarily due to
thesia. Furthermore, SLMPs containing 10%w/ the higher mass-transfer rates achieved in PCA.
w lidocaine showed promising results of pain Characteristic mass-transfer times (τmt) for GAS
relief and anti-infection activity (Lopez-Iglesias and PCA processes have been calculated based on
et al. 2020). models developed by Lin et al. and Werling and
Debenedetti (Fusaro et al. 2005; Werling and
Debenedetti 1999, 2000; Lin et al. 2003).
12.2.5 Comparison of Precipitation τGAS
mt ¼ M 0 =M CO2 , ð12:8Þ
Processes Utilizing Supercritical
Fluids τPCA
mt ¼ t max V 0 =ðV max V 0 Þ: ð12:9Þ
SCF precipitation technologies have where M0 is the initial amount of solvent, M CO2 is
demonstrated the ability to produce submicron the CO2 addition rate, tmax is the time for the
particles of poorly water-soluble drugs. However, solvent droplet in the PCA process to swell to
the creation of submicron particles is not consid- its maximum diameter, V0 is the original volume
ered typical for any of the processes, as 1–5-μm of the solvent droplet prior to swelling, and Vmax
particles are commonly produced. The experi- corresponds to the volume of the solvent droplet
mental research in this area has been at its maximum diameter. In the GAS process, the
mass-transfer rate is a function of the CO2 addi- In contrast to GAS and PCA, precipitation by
tion rate. In PCA, the mass-transfer rate is RESS results from a sudden change in pressure,
correlated to the change in volume of the solvent which causes a decrease in solvent power, and
droplet due to the mass transfer of CO2 into the thus prompts nucleation and precipitation of
droplet. Figure 12.15 illustrates the effect of char- particles. Depressurization of CO2 during RESS
acteristic mass-transfer times on particle size, has been reported to occur at the speed of sound,
based on Lin’s model. The estimated range of corresponding to timescales on the order of 10–6–
mass-transfer times for PCA is about 2 orders of 10–4 s (Debenedetti et al. 1993). Because the
magnitude smaller than that for the GAS process, timescale over which depressurization occurs
further validating the theory that the primary dif- may be correlated to the timescale during which
ference between these two processes is the mass- nucleation occurs, one may expect RESS to be
transfer rates that can be achieved. These mass- capable of producing smaller nanoparticles, com-
transfer rates correlate directly with rates of gen- pared to PCA and GAS. However, collisions in
eration of supersaturation and thus give an indi- the free jet lead to particle growth and similar
cation of characteristic nucleation times. These particle sizes unless a solvent containing a stabi-
estimates were confirmed experimentally by the lizer is utilized as in RESAS and RESOLV or
precipitation of the poorly water-soluble drug RESS-SC. Additionally, RESS does not utilize
paracetamol using both GAS and PCA. Mean organic solvents and therefore minimizes envi-
particle sizes ranging from 90 to 250 μm were ronmental and toxicity concerns regarding resid-
produced by GAS precipitation, in comparison to ual solvent levels. PGSS requires neither organic
1.3–2.5 μm for PCA. Corresponding mass- solvents nor the solute to possess high solubility
transfer times were 20–900 s and 0.04–0.12 s in CO2, thus facilitating large process yields. The
for GAS and PCA, respectively, in the range primary drawback to PGSS is that significant
predicted in Fig. 12.15. coagulation between primary particles occurs
Fig. 12.15 Average size of particles produced by directly with characteristic nucleation times, then the
SCF-based precipitation technologies (GAS and PCA) as RESS process may also be quantitatively compared to
a function of the characteristic mass-transfer time calcu- GAS and PCA. Adapted from Fusaro et al. (2005). Copy-
lated using the model presented in Lin et al. (2003). If the right 2005 with permission from Elsevier
characteristic mass-transfer time is believed to correlate
during processing, resulting in typical particle Minimizing the miscibility of the organic solvent
sizes greater than several microns. with the aqueous solution reduces particle growth
via Ostwald ripening and limits the tendency of
the organic solvent to interfere with the
12.2.6 Evaporative Precipitation into capabilities of the surfactant to coat the particles
Aqueous Solution (EPAS) and provide steric stabilization. For CsA particles
stabilized with Pluronic F127, smaller particles
To address the solubility restrictions that have were produced when DCM was chosen as the
limited the applicability of SCF precipitation organic solvent versus diethyl ether (mean parti-
technologies for nanoparticle production, the cle diameter of 423 nm versus 1218 nm using
evaporative precipitation into aqueous solution DCM and diethyl ether, respectively) (Chen
(EPAS) process was developed based upon simi- et al. 2002). Both solvents possess similar
lar operating principles as RESAS. In EPAS, the volatilities and heats of vaporization (Carl
drug is dissolved in an organic solvent and then 1999). However, at the aqueous reservoir temper-
atomized into a heated aqueous solution. ature of 75 C, the solubility of DCM in water is
Stabilizers may be incorporated into the organic 4 mg/mL, compared to 12 mg/mL for diethyl
or aqueous phase, or both. A schematic ether.
representing the EPAS process is shown in Due to the similar particle formation
Fig. 12.16. The elevated temperature of the aque- mechanisms of EPAS and RESAS, key EPAS
ous solution facilitates rapid evaporation of the operating parameters also include nozzle design,
organic solvent, which induces supersaturation process temperature, stabilizer selection and con-
and subsequent nucleation of the drug. The large centration, and final suspension concentration in
interfacial area produced by the nucleating the aqueous phase. The nozzles used in EPAS
surfaces provides a strong driving force for the processes are similar to those for RESAS,
adsorption of the stabilizers to the newly formed targeting intense atomization of the organic solu-
particles. Passivation of the particle surface by tion into the aqueous bath to facilitate rapid nucle-
stabilizers hinders particle growth via condensa- ation, as well as rapid diffusion of the stabilizers
tion and coagulation. The resultant particles may to the particle surfaces. In terms of process tem-
be harvested by filtration, lyophilization, or spray perature, the organic solution is often heated to
drying of the drug dispersion (Sarkari et al. 2002). improve the solute solubility in the organic solu-
Because organic compounds generally possess tion, in addition to promoting more rapid evapo-
significantly higher solubilities in organic ration of the solvent and, thus, supersaturation
solvents as compared to SCFs, particularly CO2, and nucleation. For similar reasons, the aqueous
the EPAS process is more amenable to a wider reservoir is also typically heated to accelerate
variety of APIs and can achieve higher process evaporation and, thus, nucleation rates. Higher
yields compared to RESAS. temperatures in the aqueous reservoir also pro-
The key operating parameters that impact par- mote the diffusion of the stabilizers to the particle
ticle size and morphology in the EPAS process surface. Chen et al. (2002) showed that the size of
are similar to those mentioned for RESAS, as polyvinylpyrrolidone (PVP)-stabilized CsA
EPAS parallels RESAS in many aspects. How- particles decreased from 1354 to 803 nm when
ever, the evaporation of CO2 droplets is more the temperature of the aqueous solution was
rapid than for an organic solvent. Droplet forma- increased from 55 to 85 C. However, the oppo-
tion is well defined in RESAS because CO2 is site trend was observed for CsA particles
only slightly miscible with water. In EPAS, stabilized with Tween 80 where, under the same
dichloromethane (DCM) has been chosen as the operating conditions, mean particle size increased
organic solvent because of its similar low from 308 to 774 nm for the same temperature
miscibility with water, in addition to its ability increase. In the case of ethoxylated surfactants,
to solubilize a variety of organic compounds. such as Tween 80, high temperatures weaken the
heating jacket
T
P
N2 water
T
pure solvent
reservoir
T
drug solution H
reservoir
T–thermocouple;
P–pressure regulator;
H–HPLC pump
Fig. 12.16 Schematic of EPAS. Reprinted from Chen et al. (2002). Copyright 2002 with permission from Elsevier
hydrogen bonding between the ethylene oxide Typical particle sizes generated by EPAS, as
groups and, thus, hinder steric stabilization determined by light scattering, are in the range of
(Blankschtein et al. 1986). High temperatures 1–10 μm. However, analysis of the particles by
also have an adverse effect on some triblock microscopy and Brunauer–Emmett–Teller (BET)
copolymers, such as Pluronic F127, in which the surface-area measurements suggests that the
solution viscosity increases for elevated micron-sized entities are actually aggregates of
temperatures, resulting in longer diffusion times smaller, submicron particles (Vaughn et al.
(Sinswat et al. 2005). Therefore, the effect of 2005; Sinswat et al. 2005). As the solvent
temperature on stabilizer performance should be evaporates and nucleation occurs, the nuclei
considered during stabilizer selection. Another become more concentrated as the organic droplet
parameter that must be addressed due to its influ- shrinks, which increases the probability of coag-
ence on supersaturation levels is the drug concen- ulation. Thus, effective stabilization after nucle-
tration in the feed solution. Unlike RESAS, in ation is necessary to maintain small particle sizes.
which the drug concentration is limited by low The selected stabilizer should diffuse rapidly in
solubility in CO2, feed concentration may be the appropriate solvent, have sufficient chain
varied in EPAS due to the larger solubilities of length to provide steric stabilization, and have a
drugs in organic solvents. When the feed CsA high affinity for adsorption of the drug surface. In
concentration was increased from 1 to 5% w/v, general, larger-molecular-weight
the average size of CsA particles decreased by at (MW) stabilizers take longer to diffuse to particle
least 40%, down to submicron particles, when surfaces, compromising final particle size. On the
stabilized by several different surfactants (Chen other hand, the greater radius of gyration provides
et al. 2002). The higher drug concentrations in the better steric stabilization. Therefore, the need for
feed generated higher degrees of supersaturation rapid diffusion to the particle surface must be
during solvent evaporation, leading to smaller balanced with the need for a surfactant with suffi-
particles. cient MW to provide effective steric stabilization.
Several studies have compared the steric
capabilities of various surfactants. For example, organic and the aqueous phases, even smaller
CsA particles prepared by EPAS using Tween 80 particles were created, about 40% lower than
or Myrj 52 were ~500–600 nm in diameter, com- when Pluronic was added only to the aqueous
pared to ~1100 nm when stabilized using higher phase. The ability to reduce particle sizes by
MW PVP 40 T under the same operating including stabilizers in both the organic and aque-
conditions. However, PVP 40 T was found to be ous phases during EPAS precipitation was also
a better stabilizer for danazol particles produced demonstrated for danazol and itraconazole (Itz)
by EPAS, compared to lower MW surfactants particles, where particle-size reductions up to one
such as Pluronic F127, sodium lauryl sulfate order of magnitude were achieved, down to sub-
(SLS), and sodium deoxycholic acid (DCA). micron levels, depending on the selected combi-
PVP-stabilized danazol particles were 10–17 μm nation of stabilizers (Vaughn et al. 2005; Sinswat
in diameter, compared to 22–30-μm particles et al. 2005).
when stabilized by the lower MW surfactants Interestingly, high-potency particles, greater
(Chen et al. 2004b). It is interesting to note the than 50% w/w drug, may still be produced by
large disparity between the sizes of the CsA and EPAS despite increasing the concentration of
danazol particles stabilized with PVP 40 T under stabilizers in both the organic and aqueous
similar operating conditions. PVP-stabilized CsA phases. As mentioned previously, particle growth
particles produced by EPAS ranged between is impeded by the adsorption of stabilizers to the
600 and 1100 nm in diameter for final dispersion particle surface. Because the hydrophilic portions
concentrations between 1 and 5% w/v, whereas of the surfactant favor the drug–water interface,
PVP-stabilized danazol particles were 10–17 μm relative to the hydrophobic particle interior, the
when prepared at a 2% w/v aqueous dispersion surfactant selectively orients itself at the particle
(Chen et al. 2002, 2004b). This difference in size surface (Matteucci et al. 2007). Upon passivation
highlights the fact that stabilizer selection is of the particle surface, the loading of surfactant is
highly dependent upon the affinity of a stabilizer limited by the equilibrium adsorption. Therefore,
to adsorb on a particular drug surface, in addition the most effective strategy for stabilizing particles
to growth rates for particular drugs. produced by EPAS is to accelerate surfactant
In EPAS, another factor to consider when adsorption to the nucleating surfaces through sur-
selecting the appropriate surfactant is whether to factant selection and placement and to use excess
incorporate the stabilizer into the organic phase, amounts of surfactant. Unadsorbed surfactants
in addition to the aqueous phase. The addition of may be removed by centrifugation after precipita-
effective amounts of stabilizers to the SCF phase tion (Vaughn et al. 2005; Sinswat et al. 2005).
was not plausible in RESAS due to the low solu- EPAS production of Itz yielded particles with a
bility of many stabilizers, especially high-MW BET surface area of 6.31 m2/g (~731 nm in diam-
polymers, in SCFs. The addition of a stabilizer eter, assuming a spherical geometry) and 93.8%
to the organic phase in EPAS has enabled the w/w potency when the stabilizer, Pluronic F127,
production of smaller particles, compared to was added to both the organic and aqueous phases
systems where the stabilizer is only present in (Sinswat et al. 2005; Chen et al. 2004a). When Itz
the aqueous phase, because less time is required was stabilized using PVP-K15 by EPAS, particles
for the surfactant to diffuse to the particle surface, as small as 500 nm in diameter were achieved for
as it does not need to cross the aqueous/solvent a drug-to-excipient ratio of 0.79 (Chen et al.
boundary. The average diameter of carbamaze- 2004a). It should be noted, however, that slightly
pine (CBZ) particles prepared by EPAS stabilized larger particle sizes are generally observed as the
using Pluronic F127 was ~20% lower when the drug/surfactant ratio (i.e., drug potency) is
Pluronic was integrated into the organic phase increased (Chen et al. 2002, 2004a, 2006). For
versus the aqueous phase (mean diameter of the precipitation of CsA particles with Tween 80
13 and 16 μm, respectively) (Sarkari et al. as the stabilizer, the mean particle size increased
2002). When stabilizers were added to both the from 338 to 523 nm to 921 nm when the drug-to-
excipient ratio was increased from 0.33 to 0.72 to et al. 2005; Sarkari et al. 2002; Sinswat et al.
2.50, respectively (Chen et al. 2002). 2005). However, for higher suspension
Contact-angle measurements verified that the concentrations, larger particle sizes, as well as
hydrophilic stabilizer sufficiently coated the sur- broader PSDs, are observed. When higher sus-
face of the particles, as expected, given the col- pension concentrations are desired, it is often
loidal stability. For these measurements, the drug beneficial to increase the surfactant concentration
dispersions produced by EPAS were centrifuged in the system to ensure sufficient coverage of the
to remove unadsorbed surfactant, then they were drug particles. An additional benefit of the well-
dried, and the resultant powder was compacted stabilized EPAS particles is that after drying the
into a tablet. The contact angle observed for a drug dispersion, the powders have been shown to
droplet of water on the tablet surface was then redisperse to sizes similar to those present in the
measured. The contact angle for Itz tablets original dispersion, indicating good stabilizer
prepared from EPAS powder was ~32% smaller coverage of the particles (Chen et al. 2004b).
than that for a tablet prepared from a physical Moreover, the EPAS process was shown to pro-
mixture of the identical composition, validating duce both crystalline and amorphous particles,
the claim that the EPAS process tends to orient depending on the stabilizers chosen. Rapid stabi-
the stabilizers to the particle surface (Sinswat lization of particles, before molecules are able to
et al. 2005). Additional studies have verified rearrange into the crystalline structure, leads to
these results, where high-potency particles of car- higher amorphous content in the particle.
bamazepine and danazol produced by EPAS, Table 12.6 summarizes some particle properties
composed of at least 50% w/w drug, possess achieved through precipitation by EPAS.
smaller contact angles than identical formulations The advantages of EPAS and emulsion
prepared as a physical mixture (Vaughn et al. templating have been combined in order to
2005; Sarkari et al. 2002). The lower contact develop a more robust precipitation process
angles of the EPAS particles also indicate called Advanced EPAS. Indeed, replacing the
enhanced wettability over the physical mixtures, organic solution with an oil-in-water emulsion
which is especially important to achieve favorable allowed for better control over the particle size
dissolution rates for poorly water-soluble drugs. than compared with an organic solution as used in
Because the EPAS process preferentially EPAS. The evaluation of the influence of
concentrates the surfactant at the particle surface, processing parameters demonstrated indepen-
where steric and wetting capabilities are dence with regards to particle size, proving the
maximized, only small amounts of surfactant are robustness of the process. Furthermore,
required to stabilize particles with high drug Advanced EPAS overcomes the limitation of
potency and to improve wettability for enhanced EPAS in terms of scale-up as the requirement
dissolution. EPAS has also demonstrated for consistent atomizing nozzle is eliminated
advantages with respect to quercetin chemical (Bosselman et al. 2012).
stability in nanosuspension compared to solution.
This was explained by surface coverage by the
surfactant coupled with a nanosuspension, which 12.3 Antisolvent Precipitation Using
offered only the outer surface for degradation. It Organic Solvents (APs)
was also mentioned that due to the protection of
the stabilizer layer, even the outer surface was The AP process, in which organic solvents make
protected from degradation (Gao et al. 2011). up the solvent phase, is one of the most common
Due to the ability of EPAS to produce particles bottom-up approaches for particle formation. AP
with good surfactant coverage, high drug suspen- processes are relatively simple, cost effective, and
sion concentrations can be obtained, typically may be operated continuously, facilitating scale-
between 15 and 50 mg/mL, which is highly up. The scalability of AP processes has been
attractive for parenteral applications (Vaughn demonstrated by Novartis for the production of
Table 12.6 Drug particles produced by EPAS

Dispersion
API/ loading
Drug Stabilizer stabilizer Mean particle diameter (μm) (mg/mL) Amorphous
Itz (Chen et al. PVP-K15 0.9–0.7 ~0.51 10 No
2004a; Sinswat Pluronic F127 (BET: 9 m2/g)
et al. 2005) Combination
Tween80 9.5–15 ~730–1500 15 No
PVP-K15 (BET: 3.1–6.3 m2/g)
Pluronic F127
Combination
CsA (Chen et al. l-α-Phosphatidyl- 0.14–0.35 0.25–0.47 14–35 Yes
2002) choline
Brij 0.3–0.5 0.033–1.04 5–50 Yes
Myrj
Tween
PEGs 0.3–0.5 0.077–1.39 5–50 Yes
PVPs
CBZ (Sarkari et al. Dooxycholic 0.45–0.83 12–19 9–40 Yes
2002) acid
PVP-K15
Sodium dodecyl
sulfate
Pluronic F127
Danazol (Chen PVPs 0.5–4 12–30 5 No, but
et al. 2004b; Sodium dodecyl ~20%
Vaughn et al. 2005) sulfate reduced
Dooxycholic crystallinity
acid
Pluronic F127
Combination
PVP-K15 1 0.62 (BET: 7.41 m2/g)/ Not Yes
~0.15–0.5 μm drug domains in reported
7–10 μm aggregates
Riccardin D (Liu Poloxamer 188 0.33 ~ 0.18 100 No
et al. 2012) HPMC
PVP K30
Combination
hydrosols and by Soliqs/Abbott for Nanomorph drives nucleation. Unlike EPAS, an organic sol-
products (Keck and Mueller 2006). Similar to vent that is miscible with the antisolvent is
EPAS, larger process yields are generally selected to facilitate mixing between the two
obtained from AP operations, as compared to phases. Diffusion of the organic solvent into the
SCF-based techniques, because organic aqueous phase tends to spread the nuclei apart,
compounds possess higher solubilities in organic reducing coagulation rates compared to EPAS, in
solvents than in SCFs. Additionally, many of the which droplet shrinkage during solvent evapora-
same particle formation mechanisms discussed tion tends to promote coagulation of nuclei.
for EPAS apply to AP technologies. In AP, the Stabilizers may be added to the solvent or
poorly water-soluble drug is first dissolved in an antisolvent phases to further mitigate particle
organic solvent, and then the drug solution is growth by condensation and coagulation. The
mixed with an antisolvent, often water. As the hydrophilic segments of the stabilizer preferen-
two phases mix, the drug solubility decreases, tially extend toward the aqueous environment,
resulting in supersaturation of the drug, which and, thus, the stabilizer adsorbs at the drug–
Fig. 12.17 Schematic of

antisolvent precipitation Solvent Phase Antisolvent Phase
(AP) of drug particles in the
presence of amphiphilic
stabilizers. Reprinted with
permission from Matteucci
Society
Nuclei
Stabilizer
Drug molecule
water interface and not within the interior of the large Da) lead to low, and often nonuniformly
particle. Passivation of the particle surface by the distributed, levels of supersaturation, which sub-
stabilizers hinders particle growth. The selective sequently result in slower nucleation rates relative
orientation of the stabilizer at the particle surface to particle growth rates. These poor mixing
facilitates the production of stable, high-potency conditions tend to produce large polydisperse
drug particles with a minimum stabilizer to drug particles.
ratio. An illustration of the driving mechanism for Favorable operating conditions promote rapid
the AP process is shown in Fig. 12.17. mixing and thus facilitate the production of
In the previously discussed precipitation-based smaller particles, as characterized by Da values
particle formation techniques, micron-sized near unity. A reduction in Da may be accom-
particles are more commonly produced than sub- plished by generating greater supersaturation via
micron particles. In AP, process and formulation more rapid nucleation (to reduce τmix) and/or by
parameters can often be manipulated to yield extending the time for condensation and coagula-
submicron particles. A contributing factor to the tion via the addition of stabilizers (to increase
higher propensity for AP to form nanoparticles is τprec). When Da is equal to unity, the particle
the miscibility between the solvent and formation process is insensitive to further
antisolvent, which facilitates both rapid supersat- reductions in mixing time. Figure 12.18 illustrates
uration as well as efficient adsorption of this concept, where the size of β-carotene
stabilizers to nucleating drug particles. Addition- particles was found to decrease with increased
ally, atomization of a partially miscible drug solu- jet velocity, which influences mixing intensity,
tion into the aqueous phase is not necessary to until a threshold value was reached (Johnson
achieve small particles because the solvent and et al. 2006). Above this threshold, the particle
antisolvent are fully miscible. Therefore, a critical size remained constant with further increases in
determinant of final particle size is the efficiency jet velocity. Therefore, when Da > 1, the particle
of mixing between the antisolvent and solvent formation process is “transport controlled,”
phases. The impact of mixing on particle forma- signifying that mixing times may be optimized
tion may be described by the Damkohler number to achieve smaller particles. However, if
(Da), defined as the ratio of mixing time (τmix) to conditions correspond to a Da 1, the process
precipitation time (τprec). conditions have already been optimized to mini-
mize particle size, and only a reduction in drug
Da ¼ τmix =τprec : ð12:10Þ
concentration or a change in solvent or stabilizer
τprec is a function of condensation time, τcond, and selection may offer further improvements for par-
coagulation time, τcoag (refer to Fig. 12.1). Poor ticle size reduction.
mixing conditions (i.e., large τmix, resulting in a
Fig. 12.18 Diameter of

β-carotene particles
produced by AP, as a
function of the stream
velocity of the organic drug
solution into an aqueous
antisolvent. An increase in
stream velocity results in a
decrease in particle size
until the break point is
reached. Adapted and
reprinted with permission
from Johnson et al. (2006).
Copyright (2006) American
Chemical Society
The limits of AP processing parameters were larger particle sizes during AP unless stabilizer
thoroughly investigated by Matteucci et al. to levels were also increased accordingly. However,
gain a solid understanding of the impact of pro- upon reaching a threshold stabilizer level, Itz drug
cess parameters on the mechanisms driving parti- loadings up to 86% w/w were produced at mini-
cle formation and stabilization, primarily for mal cost to particle size. In fact, the PSD did not
cases where high drug loadings are desired. In change significantly when the solid loading in the
the study, Itz, the model poorly water-soluble final suspension was varied between 1.8 and
drug, was stabilized using the amphiphilic poly- 8.9 mg/mL. Matteucci et al. calculated nucleation
mer Pluronic F127 (P127). Process parameters and growth rates using a population balance
including stabilizer concentration, the phase in model in conjunction with the mixed-suspension,
which the stabilizer is added, process tempera- mixed-product-removal crystallization (MSMPR)
ture, and mixing intensity between the organic model to characterize nucleation and growth
and aqueous phases were examined. As in kinetics (Jarmer et al. 2004) in order to justify
EPAS, higher concentrations of stabilizers the differences in PSD for different experimental
resulted in smaller particles when prepared by parameters.
AP, as expected. Additionally, for a given amount The temperature of the aqueous bath, into
of stabilizer, smaller particles were obtained when which the organic drug solution is introduced,
the stabilizer was added to the organic versus the was also shown to heavily influence final particle
aqueous phase. Therefore, for P127-stabilized-Itz size. Matteucci et al. reported that the average
particles of similar sizes, a 50% w/w drug loading particle size of Itz particles stabilized with P127
can be achieved when P127 is added to the increased only 15% when the temperature was
organic phase, compared to only a 25% w/w Itz raised from 3 to 10 C. However, particle sizes
loading when P127 is incorporated in the aqueous increased by a factor of 40 when the temperature
phase. Another important process parameter for of the aqueous reservoir was set near room tem-
both EPAS and AP is the final suspension con- perature, at 20 C. The operating temperature
centration. As in EPAS, increased suspension influences several aspects of the particle forma-
concentrations, which were achieved by increas- tion process. While higher temperatures generally
ing feed drug concentrations, generally led to increase diffusion rates to allow stabilizers to
quickly reach growing particle surfaces, they also tuning another process parameter toward a more-
tend to cause an increase in drug solubility in the optimal setting allows a targeted Da condition of
solvent/water mixture, which reduces supersatu- unity to be achieved and thus facilitates the pro-
ration and nucleation rates and increases the production of submicron particles even at lower
pensity for Ostwald ripening, all of which leads to mixing intensities, which require lower energy
larger particle sizes. Additionally, elevated inputs (Matteucci et al. 2006). A summary of
temperatures tend to desolvate amphiphilic how different process parameters can compensate
molecules due to weakened hydrogen bonding for lower Re conditions to yield submicron
with water, which may reduce steric stabilization. particles is shown in Table 12.7.
Unlike EPAS, where higher temperatures are Matteucci et al. further explored the range of
needed to facilitate solvent evaporation and, mixing intensities capable of producing Itz
thus, subsequent supersaturation, smaller particle nanoparticles stabilized with P127, ranging from
sizes are expected when the aqueous reservoir is simply pouring the organic into the aqueous solu-
maintained at lower temperatures during tion to dropwise addition to syringe injection, in
antisolvent precipitation. For Itz particles addition to the use of high-velocity jets. In all
stabilized with P127, Matteucci et al. cases, the aqueous phase was mixed using a mag-
recommended operating at a precipitation temper- netic stir bar (~500 rpm) to enhance heat and
ature of 3 C (Matteucci et al. 2006). Recently, mass transfer during mixing of the organic and
AP conditions including temperature, pressure, aqueous phases. As shown in Table 12.8, submi-
and solvents can control particle size and mor- cron particles may still be produced when the
phology of products. For instance, 1,3-bis organic phase is poured or added dropwise into
(9-carbazolyl) benzene (mCP) was studied in dif- the aqueous phase, although a sizable percentage
ferent organic solvents and temperatures from of micron-sized particles were also obtained.
283.15 to 313.15 K. The result showed that These lower-energy, and thus poorer-mixing,
higher temperature improved the solubility of conditions likely produce smaller degrees of
mCP and 1-methyl-2-pyrrolidinone provided the local supersaturation, resulting in slower nucle-
highest solubility. Moreover, the precipitation ation compared to the syringe and high-velocity
products had the lowest size distribution jet addition techniques. This hypothesis is
(33 nm) in methanol and water condition of corroborated by the lower calculated nucleation
antisolvent (Zou et al. 2019). rates for the lower-energy mixing techniques, as
To examine the role of mixing energies, determined using the MSMPR/population model.
characterized by Re, on final particle size, the Interestingly, the addition of the organic solution
intensity by which the organic solution was by syringe at a high organic flow rate of 130 mL/
introduced to the aqueous phase was adjusted by min yielded particles of comparable size to those
varying nozzle diameters and jet velocities. Not produced using high-velocity jets, where the
surprisingly, smaller nozzle diameters and higher organic flow rate was 10 mL/min. Additionally,
jet velocities, which create higher Re conditions, because the same solvent/stabilizer system was
tended to yield smaller particle sizes. However, used in all cases, growth rates were relatively
Matteucci et al. found that particles with sizes similar, with slightly lower calculated growth
similar to those produced under high Re rates for the syringe and high-velocity jet
conditions could still be produced under low or techniques, attributed to more efficient particle
moderate mixing energies by adjusting other stabilization due to enhanced diffusion of
experimental parameters to push the Da back stabilizers to the particle surface.
toward unity, such as increasing the flow rate of The ability of the AP process to form
the organic solution (decreases τmix) and/or nanoparticles using low-energy mixing intensity
increasing the stabilizer concentration in the methods has been further demonstrated in several
aqueous bath (increases τprec). Therefore, com- other reports. Rasenack and Muller (2002)
pensation for a nonoptimal mixing intensity by showed that, despite the organic drug solution
Table 12.7 Compensation variables that may be adjusted to maintain a low Damkohler number
PSDa
Nozzle Organic flow rate Stabilizer concentration (% Stabilizer (μm)
type (mL/min) Re w/w) location D50/D90
Organic flow rate vs. Re
0.04700 i.d. 130 3400 75 Aqueous 0.24/0.56
Crimpedb 10 6300 50 Aqueous 0.23/0.52
Stabilizer concentration vs. Re
0.002500 i. 10 5000 14 Aqueous 0.27/28
d.
0.0300 i.d. 10 410 67 Aqueous 0.27/2.9
0.0300 i.d. 10 410 75 Aqueous 0.23/0.69
Crimpedb 10 6300 14 Aqueous 0.23/0.52
Stabilizer concentration vs. location
0.0300 i.d. 10 410 75 Aqueous 0.23/0.69
0.0300 i.d. 10 410 50 Organic 0.24/0.59
Stabilizer location vs. Re
0.04700 i.d. 10 410 14 Organic 0.29/4.4
0.0300 i.d. 130 3400 14 Aqueous 0.24/0.56
0.002500 i. 10 5000 14 Aqueous 0.27/28
d.
Adapted from Matteucci et al. (2006)
a
D50 and D90 refer to the diameter at which the cumulative sample volume was under 50% and 90%, respectively
b
Crimped nozzle refers to a 0.0300 i.d. stainless-steel tubing that was crimped and then filed at the cut end until a stable
atomized flow was achieved, as described in Young et al. (2000)
Table 12.8 Impact of the method by which the organic phase is introduced to the aqueous phase on the size of Itz
particles prepared by AP and stabilized with P127
Organic flow PSD (μm)
Organic introduction rate D10/D50/ %< Nucleation rate: Growth rate:
technique Re (mL/min) D90 1 μm 10–20 n0 102 Gτ
Pouring Low ~340 0.12/0.39/ 67 1.6 6.2
8.4
Dropwise addition Low ~11 0.14/0.83/ 52 1.6 6.2
14
Syringe (0.04700 i.d.) 3400 ~130 0.1/0.24/ 97 2.0 6.0
0.56
High-velocity jet (0.002500 i. 5300 10 0.13/0.27/ 86 1.8 6.1
d.) 28
In each case, an Itz loading of 86% w/w and a suspension concentration of 8.9 mg/mL were achieved, with the P127
placed in the aqueous phase only (1.67 mg/mL). Adapted from Matteucci et al. (2006)
being merely poured into the aqueous phase, Itz maintain submicron particle sizes. On the other
nanoparticles were produced when the appropri- hand, more hydrophilic stabilizers, such as dex-
ate stabilizer was selected. In the case of Itz, tran, polyvinylalcohol, polyvinylpyrrolidone,
stabilizing agents containing cellulose ethers hydroxyethyl starches (HES), and polar
with alkyl-substituents, such as methyl cellulose substituted cellulose ethers (hydroxyethyl
(MC), methylhydroxyethylcellulose (MHEC), celluloses (HEC) and hydroxypropyl celluloses
and hydroxypropylmethylcellulose (HPMC), (HPC)), yielded micron-sized particles under sim-
effectively protected against particle growth to ilar operating conditions (Fig. 12.19). These
100
90
Particle Size
80
70
60
(mm)
50
40 D99
30 D90
20 D50
10 D10
0
er
n
tin
EC
ES
PC
00
EC
al C
te
in
P
ra
PV
PV
PM
ct
na
at
la
40
H
H
H
xt
pe
w
ge
gi
M
de
H
C
a
N
M
m
diu
so
% HPMC 4000
Fig. 12.19 Particle size of itraconazole (Itz) particles produced by AP stored as a dispersion 24 h after precipitation.
Concentration of stabilizer in water was 0.025% w/w. Data from Rasenack and Muller (2002)
Fig. 12.20 Influence of

different concentrations of
HPMC 4000 on the size of
itraconazole (Itz) particles
produced by AP. Particles
were stored as a dispersion,
and sizes were measured
24 h after precipitation.
Data from Rasenack and
Muller (2002)
results indicate that the stabilizer must interact of 0.025% in the aqueous reservoir was required
sufficiently with the newly formed surface of the to stabilize 600-nm Itz particles under the experi-
poorly water-soluble compound in order to pro- mental conditions used by Rasenack and Muller
vide an efficient barrier to particle growth. In the (~30/1 HPMC/Itz, organic solution poured into
cases of MHEC and MC, the methoxyl and aqueous solution). Higher HPMC concentrations
hydroxypropyl groups adsorb onto hydrophobic did not further reduce particle sizes due to the
surfaces (Daniels and Barta 1994). Although nature by which particles are stabilized in AP,
HPMC is relatively hydrophilic, it is sufficiently where passivation of the particle surface indicates
hydrophobic to facilitate adsorption onto hydro- maximum stabilization (Fig. 12.20). Similar
phobic particle surfaces (Chang and Gray 1978). results were observed when HPMC was used to
The ability of HPMC to efficiently stabilize Itz stabilize ketoconazole particles produced by AP
nanoparticles was further explored by examining under similar operating parameters (Rasenack
the range of HPMC concentrations required for and Muller 2002). Fenofibrate (~320 nm in diam-
stabilization. A minimum HPMC concentration eter) stabilized by a combination of sodium
Fig. 12.21 The effect of

aging time on the size of 2500
freshly precipitated
Z-average Diameter (nm)

fenofibrate drug particles in
dispersion under stirring 2000
conditions at the rate of
600 rpm. Reprinted from
1500
Hu et al. (2011). Copyright
2011 with permission from
Elsevier 1000
500
0
0 5 10 15 20 25 30
Time (min)
dodecyl sulfate (SDS) and HPMC (Hu et al. by-design using resveratrol as a model drug. Res-
2011) and spironolactone nanoparticles veratrol nanosuspension produced by the AP pro-
(200–400 nm in diameter) stabilized with cess based on a quality by design showed the
HPMC have also been prepared by AP (Dong lowest mean particle diameter of 46.3 nm. PVP
et al. 2009), where the organic solution was rap- VA64, PVP K12, and SLS were selected as
idly injected into the aqueous phase using a stabilizers because these stabilizers demonstrated
pipette or syringe. In both studies, the organic an ability to produce nanoparticles and the higher
and aqueous phases were also pumped into a precipitation inhibition ability compared with
static mixer, which consists of a chamber HPMC and HPC (Fig 12.22). In a pharmacoki-
containing several baffles to facilitate mixing netic study, the maximum plasma concentration
between entering fluids (Hu et al. 2011; Dong (Cmax) and the area under the plasma concentra-
et al. 2010; Gassmann et al. 1994). Resultant tion (AUC0-12h) were higher than those with the
fenofibrate and spironolactone particles were of unprocessed resveratrol (Kuk et al. 2019). The
comparable size, although slightly larger than particle growth may be driven by condensation
those prepared when the organic phase was of dissolved drug molecules and/or Ostwald rip-
introduced by injection (~330 and 500 nm for ening due to the organic solvent still present in the
fenofibrate and spironolactone, respectively), aqueous suspension. In response to these
indicating relatively efficient mixing within the challenges, modifications to the AP process have
static mixer. In the case of the fenofibrate been developed to facilitate nanoparticle produc-
particles, continued monitoring of particle size tion and maintain particle size after precipitation
showed that the freshly precipitated particles by minimizing particle growth and are discussed
grew over time if left in the suspension, up to in the next section.
four times the initial size in just 10 min
(Fig. 12.21) (Hu et al. 2011). Recently, PVP
was incorporated to stabilize precipitated particles
12.3.1 Recent Trends in AP Processes
produced by the AP process. The results showed
that the particle size of 1,3-bis(9-carbazolyl) ben-
Flash Nanoprecipitation (FN) Process
zene was reduced from 1 μm to 500 nm and its
To facilitate the production of amorphous
morphology was altered from elliptic shape to
nanoparticles for enhanced dissolution of poorly
lumpish shape (Zou et al. 2019). Furthermore,
water-soluble drugs, flash nanoprecipitation (FN)
the AP process was developed with a quality-
aims to minimize mixing times between the
Fig. 12.22 Inhibitory effects of PVP, HPC, and HPMC on resveratrol precipitation. Reprinted with permission from
Kuk et al. (2019). Copyright (2019) Pharmaceutic, MDPI
solvent and antisolvent, down to millisecond Prud’homme, optimal performance of the FN

timescales, using a custom-designed confined process is achieved when τmix values are less
impinging jet (CIJ) mixer. In a CIJ mixer, a sol- than 100 ms (Johnson et al. 2006). These low
vent stream and an antisolvent stream are τmix values promote nanoparticle production, as
introduced into a mixing chamber at turbulent seen in (12.10), as well as narrow PSDs. In FN
jet velocities in such a manner that the two studies, Johnson and Prud’homme stress the
streams are collinear and thus collide with each importance of not only comparing τprec of the
other. Mixing within impinging jets produces a drug to τmix but also matching τprec of the drug
region of high-energy dissipation, as the kinetic with τprec of the stabilizer, especially in the case of
energy of the jet streams is converted to turbulent polymeric stabilizers. CIJ mixers have the ability
motion through collision and redirection of fluid to generate high supersaturation and superior
flow within a confined volume. These high- nucleation rates. Thus, CIJ mixers can produce
energy dissipation regions rapidly reduce the nanoparticles with high drug loading and high
scale of segregation between the two fluid encapsulation efficiency. CIJ mixers were used
streams, thus facilitating rapid nucleation. The to produce various drug nanoparticles
mixing chambers in the FN process must be (Fig. 12.24). For example, clofazimine, which is
large enough for the high-energy dissipation a lipophilic riminophenazine antibiotic, was stud-
regions to form but limited in volume to avoid ied using the FN process with CIJ mixers.
significant bypassing of any fluid from intense Stabilizers including hypromellose acetate succi-
mixing (Johnson et al. 2006; Johnson and nate (HPMCAS) and lecithin were included to
Prud’homme 2003a, b, c). A schematic of the produce clofazimine nanoparticles. The
FN process is shown in Fig. 12.23. nanoparticles produced by HPMCAS
Mixing energies characterized by Re up to demonstrated a mean particle diameter of
3000 have been reported for CIJ mixers, 90 nm, while nanoparticles produced by lecithin
corresponding to a characteristic mixing time, showed a higher mean particle diameter of
τmix, of 5 ms, when jet diameters, jet velocities 170 nm. Moreover, the dissolution of clofazimine
of the fluid stream, and chamber size have been nanoparticles increased in the range of 50–90-
optimized (Johnson and Prud’homme fold compared to the unprocessed clofazimine
2003a, b, c). According to Johnson and (Zhang et al. 2017). The flow rate of liquid jets

confined impinging jet
(CIJ) apparatus: A solvent
jet, in which the poorly
water-soluble drug and impingement
stabilizers are dissolved, mixer
and an antisolvent jet
containing stabilizers are
impinged against each other
to facilitate mixing of the hydrophilic polymer hydrophobic compoumd & polymer
two solutions. High- in antisolvent (water) in water miscible organic solvent
velocity impingement
promotes rapid mixing
within the chamber to
facilitate rapid particle
precipitation. Reprinted
with permission from Zhu
Society
block copolymer-protected nanoparticles
Fig. 12.24 Application of CIJM in drug nanoparticles. Reprinted with permission from Tao et al. (2019). Copyright
(2018) with permission from Elsevier
is one of the critical parameters of the FN process. a nucleation initiator to further control nanoparti-
A study of florfenicol-loaded poly(lactide-co- cle production. Characteristic precipitation times
glycolide) and poly-caprolactone nanoparticles may be adjusted by tuning stabilizer properties,
demonstrated that the higher flow rate produced such as molecular weight (MW) and the size ratio
smaller particle sizes (Turino et al. 2018). When of hydrophilic to hydrophobic moieties, as well as
the precipitation times for the drug and stabilizer the drug feed concentration. To demonstrate the
are manipulated to match one another, the hydro- importance of stabilizer selection in FN,
phobic portion of the stabilizer is designed to β-carotene particles stabilized by polystyrene
precipitate onto the surface of the drug particle (PS) (2 K)-b-polyethylene glycol (PEG) (5 K)
at the onset of nucleation, thus deterring further and polycapralactone (PCL) (3.6 K)-b-PEG
particle growth beyond nucleation sizes. More- (6 K) were compared. Particles stabilized with
over, the proper selection of stabilizers can act as PS-b-PEG, possessing a drug potency of 66%
w/w, were ~100 nm in diameter, while those importance of drug insolubility to prevent
stabilized with PCL-b-PEG required higher stabi- interparticulate migration. They also mentioned
lizer levels to achieve an average particle size of the use of log P to predict particle stability,
100 nm, reducing the drug potency to 18% enabling therefore a fast and easy pre-clinical
w/w. The reduced effectiveness of the PCL-b- drug screening (Zhu 2014).
PEG polymer to stabilize the β-carotene particles The necessity of hand operation for the FNP
was attributed to the lower melting point of the process was eliminated by a modification of the
PCL, which may have facilitated aggregation design of the CIJ mixer and by adding a second
between the particles (Zhu et al. 2007). Polyelec- antisolvent dilution step. These two changes
trolytes such as poly (ethylene imine) or chitosan allowed for fast quenching with a high antisolvent
can also stabilize nanoparticles of β-carotene to concentration enhancing therefore nanoparticle
an average diameter of <100 nm and a drug stability. Stable and reproducible nanoparticles
loading of >80%. Besides steric stabilization, (55 nm) of β-carotene were obtained using CIJ
polyelectrolytes also provided electrostatic stabi- with dilution. Because it overcomes the equal
lization of the amorphous nanoparticles as volume ratios of the original CIJ design, CIJ-D
demonstrated by zeta potential measurements. enables a decrease in the volume needed, making
The amorphous state of the nanoparticles was it an inexpensive technique (Han et al. 2012).
due to fast precipitation (Zhu et al. 2010). FN Recent research on the FN process has focused
has also reported the successful production of on the addition of a multi-inlet-vortex mixer
CsA nanoparticles (~300 nm) stabilized by a (MIVM) to FN (schematic shown in Fig. 12.25)
combination of dextrose monohydrate and leci- to allow for efficient mixing of multiple streams
thin at a drug potency of 30% w/w (Chiou et al. with unequal flow rates, which has been found to
2008a, b). Methods, such as vacuum distillation be a requirement for some systems to achieve
or spray drying, or dialysis was required to optimal nanoprecipitation conditions. Moreover,
remove the organic solvent from the final suspen- MIVM can overcome the limitation of CIJ mixers
sion in order to minimize aggregation of particles and provide rapid mixing and scalability of the
after precipitation Putsulka et al. investigated on continuous production of nanoparticles. MIVM
the impact of block copolymer, solute, and API has higher adjustability of flow rate and
on the FN process. The results were in agreement antisolvent mixer that achieves different levels
with a model developed by Johnson and of supersaturation, nucleation, and growth time
Prud’homme (2003). They demonstrated that scale to control particle size. Thus, this mixer
log P was a good indication of the stability of device showed steadier function and broader
hydrophobic drugs in nanoparticles.
Manufacturing of 100 nm particles with at least
50% drug loading was reported preferable with
small molecules having a log P < 6 and > 1%
solubility in water-miscible solvent. Also, 90% of
drug-loaded particles were achievable if drug log
P was ~10, but the size of the resulting particles
was around 200 nm showing the impact of load-
ing on the particle size. The equal mass ratio of
small molecule and polymer led to a particle size
of ~100 nm independent of solid concentration
(Pustulka et al. 2013). Log P has also been
correlated with particle stability; indeed, log Fig. 12.25 Schematic of a multi-inlet-vortex mixer
(MIVM), used to facilitate efficient mixing of multiple
P > 12 showed good stability and log P < 2 was
feed streams with unequal flow rates into the confined
very difficult to generate nanoparticles due to its impinging jet (CIJ) mixer. Reprinted from Liu et al.
high solubility. Later, Zhu et al. confirmed the (2008). Copyright 2008 with permission from Elsevier
applications than CIJ mixers for poorly water- of 170 nm. After spray drying, the final
soluble drug nanoparticles, including polymeric nanoparticles showed diameters between
micelles and nanoparticles, polyelectrolyte com- 100 and 200 nm (Szymusiak et al. 2016). To
plex, nanocrystal compounds, and solid lipid achieve a scalable process, MIVM was integrated
nanoparticles (Tao et al. 2019). To validate that into spray drying or lyophilization to scale up
sufficient mixing is achieved within the MIVM, production. For example, OZ439, which is a
Liu et al. developed a computation fluid dynamics new active against drug-resistant malaria, was
(CFD) simulations program to emulate flow developed to produce nanoparticles by the FN
behavior within the MIVM (Liu et al. 2008; process, including MIVM and CIJ mixers.
Shen et al. 2011). The simulation was validated MIVM incorporated with spray drying or lyophi-
using a model experiment, involving a parallel, lization successfully produced the smallest
competing reaction system, in which product nanoparticles containing OZ439 and HPMCAS.
yields were measured to give an indication of As mentioned, the nanoparticles produced by
mixing intensity. An excellent correlation MIVM showed a lower diameter of 100 nm,
between the simulation and experimental data while CIJ produced larger nanoparticles of
was found, providing validation of mixing perfor- 150 nm. Moreover, MIVM incorporated into the
mance within the MIVM, as well as a useful tool spray drying improved the dissolution kinetics
to optimize process parameters for nanoparticle compared with MIVM incorporated the lyophili-
production in future studies (Liu et al. 2008). zation. In a stability study, the nanoparticles pro-
Shen et al. confirmed the rapid micromixing and duced by MIVM incorporated spray drying
high supersaturation leading to nanoparticle for- demonstrated stable amorphous morphology
mation. High Re was important to produce after storage at 50 C and 75% RH in uncapped
nanoparticles with a controlled size distribution, vials for one month and 40 C and 75% RH in
as previously mentioned by Liu et al. (Liu et al. capped vials for 6 months (Ristroph et al. 2019).
2008). MIVM demonstrated the ability to encap- The scalability of this new mixer was
sulate at high efficiency a wide variety of demonstrated for different chamber geometries
materials such as model drug, b-carotene, and and sizes (Johnson et al. 2006). The benefit of
hydrophilic charged polymers (Shen et al. highly stabilized nanoparticles of curcumin
2011). Appropriate stabilizers selected to the processed by FN (mean particle size < 80 μm)
MIVM achieve nanosized drug HPMCAS compared to unprocessed particles was
particles. In the case of the production of pacli- demonstrated in vivo in mice with improved clin-
taxel nanoparticles by FN, submicron particles ical efficacy at a low dose and much higher bio-
could not be produced using only PCL(3 K)-b- availability (Cheng et al. 2012). Recently, Chow
PEG (5 K) as the stabilizer. However, the intro- et al. also compared the performance of MIVM
duction of a PCL homopolymer in addition to the with a two-stream confined impinging jet with a
block copolymer using the MIVM-modified CIJ dilution mixer (CIJ-D-M). They showed that both
mixer yielded particles ranging from 80 to mixers enable particle sizes below 100 nm and
145 nm in diameter, depending on the MW of high encapsulation efficiency (>99.9%). Particles
the PCL homopolymer (Johnson et al. 2006). To produced with MIVM displayed better short-term
produce curcumin nanoparticles by MIVM, stability as well as roughly spherical particles
PLGA (stabilizer) reduced the particle size at an compared to the irregular nanoaggregates
appropriate ratio. The weight ratio of API and obtained with CIJ-D-M. However, smaller parti-
stabilizers at 1:1 produced fine particles cle size was obtained with the CIJ-D-M; these
exhibiting diameters between 50 and 250 nm results were explained by the differences in the
(Szymusiak et al. 2016). In the study of cyclo- configuration of the mixing chambers (Chow
sporine nanoparticles, lecithin and lactose were et al. 2014). In order to overcome the short-term
used as stabilizers. MIVM produced the suspen- stability, they co-formulated the product with
sion of nanoparticles that had an average diameter polyvinylpyrolidone (PVP); PVP surrounded the
particle in a protective barrier, thus preventing the with the process to remove excess solvent imme-
leaking of curcumin and extended the stability of diately after precipitation instead of in a separate
the nanosuspension (Chow et al. 2015). Recently, step, substantial levels of particle growth may be
d-α-tocopheryl polyethylene glycol 1000 succi- minimized. In CP, solvent removal is typically
nate (TPGS) stabilized itraconazole performed within a wiped film evaporator to max-
nanosuspension over 15 days because TPGS imize the available surface area over which sol-
decreased Ostwald ripening in the precipitation vent evaporation may occur, as well as to reduce
medium. The itraconazole nanoparticles also any foaming that might have occurred during
showed reproducibility with a mean particle size processing, given the high amounts of surfactants
below 100 nm (Wan et al. 2018). Moreover, required for stabilization in some cases (Rogers
MIVM also produced highly disordered amor- et al. 2004; Hitt et al. 2003, 2006). Rapid removal
phous nanoparticles that had high free energy. of a significant portion of the solvent markedly
Thus, this mixer improved not only the drug reduces the solubility of the drug in the mixed
dissolution rate but also water solubility that was solvent. Thus, particle growth by condensation of
useful for achieving higher bioavailability dissolved drug molecules and Ostwald ripening
(Chopra et al. 2016; Tao et al. 2019). However, may be greatly minimized.
metastable amorphous drug nanoparticles Danazol and naproxen particles prepared by
occurred and were transformed into stable crys- CP were well below 1 μm in diameter, 200 and
talline form in water. Thus, several studies 270 nm, respectively, as long as the operating
reported using this mixer with proper excipients, temperature was kept at 3 C (Rogers et al.
and incorporating freeze drying or spray drying to 2004). Measured residual solvent levels, metha-
remove the water medium improved the stability nol for both cases, in the aqueous suspension after
of drug nanoparticles. For instance, nano-matrix the solvent removal step ranged between 70 and
particles for inhalation of cyclosporine were pro- 380 ppm, well below the International Confer-
duced by MIVM combined with spray-dried and ence on Harmonization (ICH) guidelines for
included lecithin and lactose as excipients. The pharmaceuticals for human use (Rogers et al.
results showed that it was capable of producing an 2004). Lower precipitation temperatures
amorphous state of nanoparticles that improved corresponded to lower residual methanol levels,
the dissolution of cyclosporine nanoparticles (Tao in addition to favoring nanoparticle production. A
et al. 2019; Szymusiak et al. 2016). significant increase in particle size (up to an order
of magnitude) and polydispersity (from unimodal
to bi- and tri-modal distributions) was observed
Controlled Precipitation (CP) Process
when the precipitation temperature was increased
To minimize particle growth after precipitation,
to 25 and 50 C (Rogers et al. 2004), consistent
the controlled precipitation (CP) process
with the AP work by Matteucci et al. (2006).
incorporates a semicontinuous solvent removal
Additionally, the ability of the CP process to
step, such as vacuum distillation, after the solvent
produce high-potency nanoparticles has been
and antisolvent are mixed together (Fig. 12.26).
demonstrated for several poorly water-soluble
By incorporating the solvent removal step in line

the controlled precipitation
process. Adapted from
Rogers et al. (2004)
drugs, including Itz stabilized by HPMC (up to demonstrated the ability to produce flurbiprofen
94% w/w drug potency for particles 90–355 nm nanoparticles that showed amorphous morphol-
in diameter) (Matteucci et al. 2007), cyclosporin ogy and smaller particles compared with the
A (CsA) stabilized using Tween 80 (91% w/w unprocessed flurbiprofen. The tablet formulations
drug potency for particles 300 nm in diameter) prepared with the flurbiprofen nanoparticles were
(Tam et al. 2008), and repaglinide (REP) studied for dissolution. The results indicated a
stabilized with HPMC (50% w/w drug potency significantly higher dissolution rate and dissolu-
for particles 650 nm in diameter) (Sinswat et al. tion efficiency compared to the physical mixture
2007), as expected since the particle formation and the unprocessed drug (Essa et al. 2017).
and stabilization mechanism for CP are similar
to those of AP. The scalability of the CP process
was also demonstrated by the successful produc- 12.4 Precipitation into Acid:
tion of a 1-kg batch of naproxen particles (Rogers Microprecipitated Bulk Powder
et al. 2004). The impingement mixer can be com- (MBP) Formulations
bined with an online spray-dryer for a continuous
process suitable for industrial-scale production of Scientists at Roche have recently presented a
nanoparticles (Dong et al. 2010). modified approach to AP, in which a poorly
Another important aspect of the CP process is water-soluble drug compound is stabilized by an
its high propensity to produce amorphous ionic polymer via precipitation in an acid bath
particles due to the rapid nucleation and stabiliza- (Albano et al. 2002; Shah et al. 2012). A
tion rates generated during the precipitation pro- simplified process flow diagram is illustrated in
cess. As seen in the Noyes–Whitney equation Fig. 12.27. The setup, which is patent protected
(12.1), higher metastable solubilities of high- (Desai et al. 2010) and described by Shah et al.,
energy amorphous compounds, Csat, relative to consists of two vessels with temperature control,
crystalline compounds, provide a larger concen- one containing a cooled acidic aqueous phase
tration gradient to drive particle dissolution. Pro- (+5 C) and the other one containing an organic
duction and stabilization of an amorphous solution where a drug and a polymer are
morphology require rapid particle stabilization dissolved. Both vessels contain automatic stirrers;
during the precipitation process, similar to the the acidic aqueous phase is circulated in a closed
conditions required for stabilization of small par- loop and passes through a high shear mixing unit
ticle sizes. The same principles are applied to where the organic phase is added via an injection
control particle morphology, as the goal is to nozzle at a defined flow rate. The addition of the
stabilize the particle before the molecules can organic solution to the cooled acid phase where
arrange into a crystal structure. Controlled precip- both drug and polymer are highly insoluble
itation of Itz (Matteucci et al. 2007), CsA (Tam initiates nucleation of the drug particles, and a
et al. 2008), and REP (Sinswat et al. 2007) have precipitate, which is a mixture of API and ionic
yielded nanoparticles possessing an amorphous polymer, is formed. After complete addition of
morphology, which contributed to their enhanced the drug–polymer phase, the suspension is passed
dissolution rates over the bulk, crystalline drug through the shear-mixing unit to adjust the parti-
particles. Furthermore, a solid carrier improved cle size. The suspension is then filtered and
the dissolution rate of drug formulations such as washed with the acidic aqueous phase and then
tablets. In a controlled precipitation process, with water to remove the organic solvent. Once
Aerosil is a solid carrier dispersed in a drug solu- washed, the precipitate is dried to a desired water
tion containing flurbiprofen and hydrophilic poly- content (e.g., 2% w/w). The ionic nature of the
mer (PVP, HPMC, or PEG). After adding stabilizer is critical for effective “micropreci-
acidified water (0.01 N HCl) as an antisolvent, pitation” of “nanosized” drug domains, which
flurbiprofen nanoparticles precipitated on the are claimed to be molecularly dispersed through-
solid carrier’s surface. As a result, adding Aerosil out a polymer matrix. Therefore, polymers such
Fig. 12.27 Process train

for production of
microprecipitated bulk
powder (MBP)
formulation. Adapted and
published with permission
from Shah et al. (2010)
as hypromellose acetate succinate, Both compounds were characterized as amor-

polymethacrylate, polymethylmethacrylate, phous after processing. However, under the
hypromellose phthalate, polyvinylphthalate, conditions tested, only one of them was in a
Eudragit, hydroxypropyl methylcellulose phthal- single-phase system, as demonstrated by the sin-
ate (HPMCP), poly(methacrylic acid/ethyl acry- gle Tg observed by DSC. These results
late) (EL100), and cellulose acetate phthalate are demonstrated that depending on the interaction
applicable. The resultant powder has been termed and miscibility between the drug and polymer, a
a “microprecipitated bulk powder” (MBP). dispersion of amorphous drug in the polymer or
Amorphous morphologies have also been formed molecular dispersion of the drug could be
under sufficiently rapid stabilization conditions. obtained. The dissolution of solid dispersions
Ionic polymers with a MW of at least 80,000 Da manufactured with MBP was dependent on the
and a glass transition temperature, Tg, > 50 C lipophilicity of the drug but was optimized using
have been found to promote efficient stabilization additional excipients and led to rapid and com-
of submicron, amorphous drug domains by this plete dissolution of the compounds with up to
precipitation technique. The acidic aqueous bath 20-fold supersaturation in an aqueous environ-
is likely needed to elicit the desired charge on the ment. These in vitro results were confirmed in
polymer to promote a strong interaction between beagle dogs where a 20-fold increase in bioavail-
the polymer and the newly formed drug surfaces ability compared to the micronized crystalline
(Shah et al. 2010). Shah et al. described the drug was demonstrated. Accelerated stability
manufacturing of two proprietary compounds studies indicated assay values higher than 99%
using MBP technology. HPMC-AS and and a maintained amorphous state. The high Tg of
Eudragit® L100 were investigated as carriers. the polymer, its low affinity for water, and a good
miscibility with the drug were responsible for commercialization under the trade name of
product stability. Sorafenib-loaded MBP was Zelboraf® (EMA assessment report, 2012).
developed to improve oral bioavailability. Cat-
ionic Eudragit E PO and anionic Eudragit S100
were used to prepare MBP incorporated 12.5 Nanoparticle Recovery
ursodeoxycholic acid (UDCA). Sorafenib-loaded
MBP was developed to improve oral bioavailabil- Whereas a wide variety of techniques have been
ity. Cationic Eudragit E PO and anionic Eudragit developed to produce aqueous dispersions of
S100 were used to prepare MBP incorporated nanoparticles, the recovery of the nanoparticles
ursodeoxycholicacid (UDCA). Sorafenib and in the solid state remains a formidable challenge.
Eudragit were dissolved in DMA and methanol Common techniques for solvent removal include
and then were added UDCA. The mixture was spray drying, freeze drying (i.e., lyophilization),
dropped into 1N NaOH as an antisolvent for and ultrafiltration (Limayem et al. 2004; Torino
Eudragit E PO and 0.1 N HCl as an antisolvent et al. 2010; Matteucci et al. 2008). Particle growth
for Eudragit S100. After the MBP suspension was may occur in these processes, as the nanoparticles
centrifuged, the MBP pellets were collected, are concentrated during solvent removal. Concen-
washed with an antisolvent, and dried by freeze tration of nanoparticles may occur by various
drying (Park et al. 2020). Recently, Bhujbal et al. pathways, depending on the state of solvation of
developed precipitation into the acid process with the polymeric stabilizer during solvent removal,
acidic polymers including hydroxypropyl meth- as shown in Fig. 12.28. Process conditions, such
ylcellulose phthalate (HPMCP), HPMC-AS, poly as temperature, salinity of the nanoparticle disper-
(methacrylic acid/ethyl acrylate) (EL100), and sion, or rate of solvent removal, may be
cellulose acetate phthalate (CAP). Lumefantrine manipulated to influence the flocculation behav-
and an acidic polymer were dissolved in ior of the nanoparticles. In spray drying, dense
dimethylacetamide (DMA) and sprayed into flocs, which do not redisperse well back to pri-
ice-cold acidified water (0.02 N HCl). mary nanoparticles, may be produced because the
Precipitated lumefantrine nanoparticles exhibited increase in nanoparticle concentration within the
amorphous morphology. Furthermore, a higher shrinking, evaporating droplet raises collision
ratio of polymer including HPMC-AS and rates and the propensity for Ostwald ripening.
HPMCP provided a higher dissolution profile of Additionally, the high temperatures required for
nanoparticles, while CAP showed the lowest drug sufficient solvent evaporation, typically greater
dissolution profiles (Bhujbal et al. 2020). This than 90 C, have been shown to desolvate some
manufacturing process appears to be a good alter- polymeric stabilizers and can further facilitate the
native when hot-melt-extrusion and spray-drying formation of large, dense flocs (Heimenz and
technologies cannot be used to improve the solu- Rajagopalan 1997; Larson 1999; Napper 1983).
bility of poorly water-soluble compounds. Moreover, rapid recovery of nanoparticle
Vemurafenib falls into this category; indeed, its suspensions into solid-state produced
high melting point and low solubility in organic nanoparticles that had long-term stability. In a
solvent render hot–melt-extrusion or spray drying study of clofazimine-loaded nanoparticles, flash
unsuitable techniques. Consequently, solvent- nanoparticles combined with spray drying
controlled precipitation was investigated, and appeared as a nanomanufacturing platform for
precipitated amorphous solid dispersion in nanoparticle recovery. The results showed spray
HPMC-AS was manufactured. Vemurafenib drying and lyophilization provided dry powders
amorphous MBP was incorporated into tablets that had long-term stability over 2 months at
and exhibited enhanced stability and clinical effi- 40 C and 70% relative humidity (RH). More-
cacy in humans (Shah et al. 2013). over, clofazimine-loaded nanoparticles showed a
This technology has demonstrated scalability, higher dissolution rate compared with unpro-
and vemurafenib has recently been approved for cessed particles (Feng et al. 2018). Similarly, for
Fig. 12.28 Floc structure

as a function of polymer
solvation and particle
volume fraction, Φ.
Polymer solvency
diminishes with an increase
in salinity or temperature.
Adapted and reprinted with
kind permission from
Springer Science+Business
Media: Matteucci et al.
(2008), copyright 2008
freeze drying and ultrafiltration, the increase in polymer-coated nanoparticles flocculate. Solva-
nanoparticle concentration and potential changes tion of polyethylene oxide (PEO)-, PVP-, and
in solvent quality with solvent removal may also HPMC-based stabilizers are known to decrease
produce dense flocs with the same limitations. with an increase in salinity or temperature (Pandit
These particle recovery techniques are energy et al. 2000; Xu et al. 2006; Pang and Englezos
intensive and may require long processing times 2002). Sodium sulfate (Na2SO4) has been used to
(freeze drying and ultrafiltration). flocculate crystalline naproxen nanoparticles
An alternative approach to nanoparticle recov- stabilized by PVP- or PEO-based polymers
ery is to form large, open flocs of primary (Chen et al. 2009) and Itz nanoparticles stabilized
nanoparticles that may be more efficiently filtered by mixtures of poloxamer 407 (P407) and HPMC
than isolated primary particles, dried, and then (Matteucci et al. 2008). This process is best
redispersed upon dosing (Matteucci et al. 2008; illustrated by “Path A” in Fig. 12.28, where the
Chen et al. 2009; Miller et al. 2012). Aqueous addition of salt rapidly decreases the polymer
suspensions (500 mg drug in 50 ml solution) of solvation and sticky collisions between
large, open flocs can be filtered in minutes to unstabilized nanoparticles produce open flocs.
obtain a dry powder (Matteucci et al. 2008; The strong van der Waals attraction between
Chen et al. 2009), compared to hours for recovery particles “locks in” the open floc structure and
of primary nanoparticles by filtration (typically inhibits rearrangement of the particles, as
~0.03 mL/min cm2) (Matteucci et al. 2008). Floc- indicated by microscopy images (Fig. 12.29)
culation of primary particles may be induced by and calculated fractal dimensions < 2. The flocs
adding a salt to raise the ionic strength of the are essentially formed by diffusion-limited col-
solution (Matteucci et al. 2008; Chen et al. loid aggregation. In contrast, slow induction of
2009) or by changing the pH (Miller et al. polymer desolvation (“Path B”) allows particles
2012), in each case to desolvate the polymer to rearrange into more energetically favorable
stabilizer on the particle surface. In the case of dense flocs. Recovery methods in which nanopar-
flocculation with salt, the loss of hydration of the ticle volume fractions increase via solvent reduc-
polymer leads to a loss in steric stabilization of tion, as in spray (“Path D”) and freeze drying
the nanoparticles. At the cloud point of the poly- (“Path C”) and ultrafiltration (“Path C”), also
mer, steric stabilization becomes weak and the tend to form denser flocs that may not readily
Fig. 12.29 Scanning electron microscopy (SEM) images of salt-flocculated Itz nanoparticle dispersions stabilized with
different amounts of excipient. Reprinted with kind permission from Springer Science+Business Media: Matteucci et al.
(2008), copyright 2008
redisperse to primary particles after drying, as et al. 2008, 2009; Chen et al. 2009), and drug
compared to flocs produced by “Path A.” Besides, yields after filtration were as high as 99% (w/w
salt flocculation is a method of nanocomplex recovered/input drug), compared to typical
recovery. Curcumin and chitosan were formed recoveries of 50–70% for spray-dried materials
as positively charged nanocomplex suspension (Nguyen et al. 2004; Maa et al. 1999). Further-
and added into Na2PO3 salt solution (pH 13.05). more, salt-flocculated Itz particles maintained
The recovered nanoparticles were filtrated and their amorphous morphology from the original
washed with deionized water to remove the precipitated dispersions, as shown by differential
excess salt (Fig. 12.30). The obtained scanning calorimetry (DSC) and dissolution stud-
nanoparticles were amorphous and in the nano- ies, whereas the Itz particles crystallized when
meter size range of 303–332 nm (Dong et al. recovered by spray drying (Matteucci et al. 2008).
2020). Moreover, flocculate formation may be By a similar principle, electrosteric stabiliza-
tuned to balance requirements for drug loading tion of nanoparticles coated with charged
(drug/excipient ratio within the particles) versus polymers may be manipulated by adjusting pH
process yields by controlling the nature and com- to neutralize a sufficient fraction of the charges.
position of the polymer stabilizers and the amount Itz nanoparticles stabilized by adjusting
of salt used to induce flocculation, which was pH-sensitive methacrylate-based polymer,
demonstrated during flocculation of naproxen Eudragit L100-55, were flocculated rapidly by
(Chen et al. 2009) and Itz nanoparticles lowering the solution pH to 2.5 with hydrochloric
(Matteucci et al. 2009). acid (HCl) to protonate the carboxylate groups on
Upon redispersion of the flocs in a good sol- the polymer (Miller et al. 2012). As described
vent, or during drug administration in physiologi- above, the rapid and strong increase in interparti-
cal fluids, the primary nanoparticles within the cle attraction due to the pH reduction under con-
floc are highly accessible to the solvent and thus stant volume fraction resulted in relatively large
readily redisperse, a behavior indicative of loose, open flocs (Matteucci et al. 2008; Chen et al.
open flocs. Itz and naproxen powders that were 2009). As observed for salt flocculation of Itz
dried after salt flocculation have been shown to nanoparticles, crystallization of Itz was minimal
redisperse to their original freshly precipitated since the large flocs were rapidly filtered at room
particle sizes (~300 nm diameter) (Matteucci temperature (Matteucci et al. 2008, 2009). Upon
Fig. 12.30 Illustration of salt-induced flocculation and filtration recovery of curcumin–chitosan nanocomplex.
Reprinted with permission from Dong et al. (2020). Copyright 2020 with permission from Elsevier
redispersion at pH 6.8, solvation of the enteric induced flocculation was used as a model drug.
polymer resulted in only a slight increase in the The results showed that the flocculation-and-fil-
size of the EL100-55-stabilized nanoparticles tration platform provided 14–17% higher produc-
(Miller et al. 2012). The combination of selected tion yield with significantly lower energy
polymer and surfactant such as hydroxypropyl requirement compared with ultracentrifugation.
cellulose (HPC) and sodium dodecyl sulfate Moreover, flocculation-and-filtration and ultra-
(SDS) provided synergistic electrosteric stabiliza- centrifugation also produced amorphous powders
tion of drug nanoparticles. Itraconazole that showed a higher dissolution rate compared
nanoparticles containing HPC-SDS had higher with the unprocessed drug (Dong et al. 2020).
wettability and the smallest particle size. As a Another advantage of this technique is that large
result, this formulation showed the fastest amounts of the stabilizer may be used in the
itraconazole release (Li et al. 2018). Preservation particle formation stage to promote the produc-
of the amorphous morphology after the tion of nanoparticles without impacting final drug
pH-flocculation process was verified by the loadings because any excess, unadsorbed
achievement of higher in vivo bioavailability in stabilizers are removed during filtration, thus
rats upon oral administration of the flocculated facilitating high drug/polymer ratios in the pow-
powders relative to a commercial Itz solid disper- der, with drug loadings in the range of 80–99%
sion (Sporanox) (Miller et al. 2012). An advan- (Matteucci et al. 2008, 2009; Chen et al. 2009;
tage of the pH flocculation process, relative to Miller et al. 2012). Thus, the flocculation/filtra-
previous salt flocculation studies, is the potential tion recovery process offers a simple, efficient
for a decrease in salt impurities in the final alternative to traditional nanoparticle recovery
product. techniques capable of yielding nanoparticle
For the flocculation/filtration process, the abil- assemblies with high drug loadings and process
ity to operate at low temperatures and constant yields while preserving amorphous
particle volume fractions (flocculation induced morphologies.
without solvent reduction), as well as rapid
removal of solvent, inhibited both growth and
crystallization of the amorphous primary
12.6 Conclusion
nanoparticles and promoted the preservation of
the highly open floc structures. Recently, the
Precipitation processes possess several
flocculation and filtration platform was compared
advantages for nanoparticle production over con-
with ultracentrifugation to improve drug solubil-
ventional top-down approaches, such as milling
ity and dissolution. Amorphous drug–polyelec-
and homogenization, as they offer enhanced con-
trolyte nanoparticle complex including curcumin
trol of morphology, particle size, and PSD with
and Na3PO4/Na2SO4 salts produced by salt-
minimal complications of contamination and
product degradation. Recently, several precipita- Materials and Equipment

tion processes have been demonstrated to achieve • Poorly water-soluble drug (proprietary)
enhanced stability. The ability of different precip- – MW ¼ 600 g/mol
itation processes, in which SCFs and organic – Tmelt ¼ 200 C
solvents were utilized as both solvents and – Solubility (T ¼ 25 C): 3 mg/mL in
antisolvents, to reproducibly yield nanoparticles water; 18.5 mg/mL in acetone or aceto-
for a wide range of pharmaceutical materials has nitrile; insoluble in CO2
been demonstrated. Each precipitation process • Solvent: ethanol
involves special techniques that influence particle • Antisolvent: CO2
production. Thus, the strengths and weaknesses • Cryostat to subcool CO2 from reservoir
of the different precipitation processes have also tank
been highlighted to aid in screening the suitability • Pump to transfer CO2 from reservoir tank
of a particular process for different drug systems. to precipitator
Another important trend in precipitation research – Gilson HPLC pump (low/intermediate
that has been highlighted in this chapter is the flow rates)/Haskel pneumatic piston
emphasis on understanding the fundamental pump (high flow rates)
mechanisms that drive these precipitation pro- • Heater coil (via water bath) to preheat CO2
cesses in order to facilitate successful scale-up feed to process temperature before entering
of the precipitation techniques and promote their precipitator
utility in commercial settings. A product utilizing • Precipitation vessel (1 L)
a solvent-controlled precipitation technology has – Equipped with a mechanical stirrer
recently been commercialized. Increased knowl- – Sinter metal filter connected to the outlet
edge of these precipitation technologies benefits tube
not only the microparticle/nanoparticle produc- • Pump to transfer organic solution to precip-
tion fields but also areas of drug encapsulation itation vessel
(Rodrigues et al. 2004; Li et al. 2005; Young et al. • Oil bath (T ¼ 80 C) to heat the fluid line
1999) and cocrystallization (Padrela et al. 2009) exiting the precipitation vessel (to avoid
and will be of general interest for all sectors blockage)
involving particle engineering. Novel particle Method
recovery processes, based on controlled floccula- • The poorly water-soluble drug was
tion/filtration of primary nanoparticles, have also dissolved in 75–150 mL of ethanol at
been discussed as an efficient means to harvest concentrations between 50 and 90% of the
nanoparticles after precipitation, making precipi- solubility at 25 C.
tation processes more attractive and feasible for • Drug/ethanol solution was pumped into the
industrial production. precipitator.
• CO2 (preheated to 25 C) was fed into the
Method Capsule 1 precipitator at 18–360 mL/min until the
precipitator was full.
Precipitation by GAS
• Contents in precipitator stirred at 500 rpm
during CO2 filling and for 30 min after the
Based on the method reported by Muhrer precipitator was filled.
et al. (2003). • Exit valve on precipitator was opened to
Objective flush out solvents, and fresh CO2 was
• To obtain nanoparticles using GAS pumped through (20 mL/min) for at least
precipitation 5 h to remove any residual organic solvent
from powder.
• Dry powder harvested from metal filter.
Results • Coaxial injector: internal tube i.d. ¼ 3 mm

• SEM indicated that spherical particles with fitted with a 0.5-mm diameter nozzle; annu-
a moderate level of agglomeration were lus i.d. ¼ 8.5 mm
produced. • Liquid separator (13 L) heated with a water
– Lower CO2 flow rates (2–18 mL/min): jacket to separate and collect solvent and
Produced bimodal particle-size antisolvent
distributions (PSD) Method
1–6-μm diameter • PCA operation was conducted in a continu-
– Higher CO2 flow rates (240–360 mL/ ous mode, where CO2 exiting the precipi-
min): tator was recirculated and the solvent
Produced unimodal PSDs exiting the liquid separator was purged
500–720-nm diameter and stored.
• X-ray powder diffraction indicated that the • CO2 was pumped into the precipitator at a
resultant particles were amorphous. constant flow rate (0.6–2.0 kg/h), passing
• Gas chromatography determined residual through a heat exchanger to heat the CO2 to
solvent levels were below 0.01 wt%. 40 C until steady-state conditions were
achieved. The precipitator pressure was set
Method Capsule 2 to 150 bar.
• Pure NMP was fed into the chamber
Precipitation by PCA
through the coaxial injector for a few
minutes.
Based on the method reported by Reverchon • ~100 mL of the amoxicillin/NMP solution
et al. (2003b) (20–100 mg/mL amoxicillin) was then fed
Objective into the chamber at the same flow rate as the
• To obtain nanoparticles using PCA pure NMP.
precipitation • Pure CO2 continued to flow through the
Materials and Equipment precipitator for a predetermined amount of
• Amoxicillin time, correlating to the calculated time
• Solvent: N-methyl 2-pyrrolidone (NMP) required for 99 wt% elimination of NMP.
• Antisolvent: CO2 • Dry powder was harvested from the internal
• Shell-and-tube heat exchanger to cool feed basket inside the precipitator.
to CO2 reservoir Results
• Diaphragm pump to transfer CO2 from res- • 90 wt% drug recovery, as calculated com-
ervoir to dryer pared to initial drug amount in organic feed
• CO2 dryer (113 L) with silica gel as the solution.
drying medium to remove trace amounts of • Amorphous particles were produced.
water from CO2 reservoir, prior to entering • SEM indicated that spherical particles were
precipitator produced.
• Heat exchanger to heat CO2 to supercritical – Lower drug feed concentrations (20 mg/
conditions mL):
• Piston pump to transfer organic solution to Produced more narrow particle-size
precipitator distributions (PSD)
• Precipitation vessel (5.2 L) 200–600-nm diameter
– Internal stainless steel basket to collect – Higher drug feed concentrations
powder (100 mg/mL):
– Temperature maintained by a water Produced more broad PSDs
jacket 500–1800-nm diameter
Method Capsule 3 Objective

• To obtain nanoparticles using EPAS
Precipitation by RESS
precipitation
Materials and Equipment
Based on the method reported by Turk • Cyclosporin A (CsA).
et al. (2002). • Stabilizers: Myrj 52, Tween 80, polyvinyl-
Objective pyrrolidone (PVP 40 T).
• To obtain nanoparticles using RESS • Solvent: dichloromethane (DCM).
precipitation • Antisolvent: deionized water.
Materials and Equipment • EPAS apparatus: stainless-steel coiled tube
• β-sitosterol (3 m length, 1/1600 o.d 0.03000 i.d.)
• Solvent: CO2 housed within a plastic water jacket (2400
• Extraction column length, 1–1/200 o.d.).
• Diaphragm pump to transfer CO2 into • Temperature controller to circulate and heat
extraction column water through the EPAS apparatus’s water
• Water bath to heat extraction column jacket.
• Capillary nozzle (i.d. of 50 μm, length of • Nozzle: stainless-steel tube (1000 length,
50 μm)—heated to 115–145 C to prevent 1/1600 o.d 0.03000 i.d.) that was cut with
clogging during atomization a wire cutter to produce a thin, elliptical slit.
Method The tapered section of the orifice was
• CO2 from the reservoir was pressurized to a ~0.5 mm in length, and the tip was filed
pre-expansion pressure of 20–30 MPa. down (to adjust the thickness of the slit)
• The scCO2 was then pumped through a until desired atomization was achieved,
water bath heated to 75–145 C generally characterized by a pressure drop
(pre-expansion temperature) to the extrac- across the nozzle orifice of ~20 MPa for
tion column. flow rates of 1 mL/min.
• The extraction column, which was also • HPLC pump to feed solvent into the
immersed in the water bath antisolvent.
(T ¼ 75–145 C), was packed with the • Water bath to heat aqueous surfactant
drug and the CO2 solution became solution.
saturated with the drug. • Separatory funnel (125 mL).
• The scCO2–drug solution was then • Lyophilizer.
expanded through a heated, capillary noz- Method
zle into an expansion chamber for powder • Solutions of 1–5% w/v CsA in DCM and
collection. aqueous solutions of 1% w/v surfactant
Results were prepared.
• SEM indicated that roughly spherical • The aqueous surfactant solution (50 mL)
particles with moderate agglomeration was poured into the separatory funnel,
were produced. Primary particles were which was then submerged in a water bath
~150 nm in diameter. (T ¼ 75 C).
• Particle size was independent of • The nozzle of the EPAS apparatus was
pre-expansion temperature. submerged ~2 cm below the surface of the
aqueous solution.
Method Capsule 4 • The organic drug solution was fed into the
aqueous surfactant solution at a flow rate of
Precipitation by EPAS 1 mL/min until a drug concentration of 1%
w/v CsA (10 mg/mL) was achieved in the
Based on the method reported by Chen et al. aqueous suspension.
(2002)
– Turbulence from the atomization of the • Cyclosporin A (CsA)

organic solution was sufficient to facili- • Stabilizers: lecithin and dextrose
tate mixing with the aqueous phase. monohydrate
– To suppress the surfactant foam pro- • Solvent: ethanol
duced by the intense mixing of the • Antisolvent: deionized (DI) water
organic and aqueous phases, nitrogen • Two syringe pumps (50 mL syringe)
was blown across the top of the • Confined liquid impinging jet (CLIJ) mixer
separatory funnel. Method
• To dry the particles, the suspensions were • CsA (0.7 g) was dissolved in ethanol
flash-frozen in liquid nitrogen and (10 mL) and loaded into one of the syringe
lyophilized to powders. pumps.
Results • An aqueous solution (30 mL) of lecithin
• Particle sizing results were determined by (0.3 g) and dextrose monohydrate (1.5 g)
dynamic light scattering. was prepared and loaded into the other
– Block copolymer (Myrj 52, Tween 80) syringe pump.
versus homopolymer (PVP 40 T) • The syringe pumps containing the organic
stabilizers: and aqueous solutions were fed into the
Myrj and Tween-stabilized particles CLIJ mixer at 40 mL/min and 120 mL/
were about half the size of min, respectively, until a total of 10 mL of
PVP-stabilized particles (530–- organic solution and 30 mL of aqueous
630 nm vs. 1080 nm, respectively). solution had been dispensed.
– 1% w/v versus 5% w/v drug concentra- • The resultant aqueous suspension was
tion in organic feed: quenched in 50 mL of DI water.
Higher drug concentrations in the feed Results
yielded smaller particles (~340 nm • Particle sizing results were determined by
for Mryj and Tween-stabilized laser light scattering.
particles and ~600 nm for – Average particle diameter: 294 nm (span
PVP-stabilized particles). 1.017, GSD 1.46).
• 86–96% drug recoveries were achieved, as – After drying, the particles were a mean
calculated compared to the initial drug diameter of approximately 260 nm.
amount in feed solution. • Scanning electron microscopy
• X-ray powder diffraction indicated that the demonstrated that the particles were
resultant particles were amorphous. spherical.
• Gas chromatography determined that resid-
ual solvent levels in powders were below Method Capsule 6
0.0004 wt%.
Precipitation by CP
Method Capsule 5
Based on the method reported by Matteucci
Flash Nanoprecipitation (FN) et al. (2007)
Objective
Based on the method reported by Chiou et al.
• To obtain nanoparticles using CP
(2008a)
Materials and Equipment
Objective • Itraconazole (Itz)
• To obtain nanoparticles using FN • HPMC E5
Materials and Equipment • Solvent: 1,3 dioxolane
• Antisolvent: deionized water
• Syringe to inject organic solution into a • X-ray powder diffraction and differential
mixer scanning calorimetry indicated that the
• Mixing apparatus to mix organic and aque- resultant particles were amorphous.
ous phase
• Vacuum distillation apparatus equipped Method Capsule 7
with a wiped film evaporator
Salt Flocculation for Nanoparticle Recovery
• Pump to transfer aqueous suspension from
mixing apparatus to vacuum distillation
apparatus Based on the method reported by Matteucci
• Lyophilizer et al. (2008)
Method Objective
• Solutions of Itz (3.3% w/w) in 1,3 • To recover nanoparticles produced by pre-
dioxolane and aqueous solutions of cipitation processes using salt flocculation
HPMC (various concentrations) were Materials and Equipment
prepared. • Aqueous suspension of itraconazole (Itz)
• The aqueous phase was maintained at nanoparticles, prepared by controlled pre-
Tprecip ¼ 3 C. cipitation, stabilized with a combination of
• The organic phase was rapidly introduced both Pluronic F127 (P127) and HPMC E5
to the aqueous phase using a mixing appa- (8:1:2 Itz:P127:HPMC)
ratus (may be accomplished by using a • Sodium sulfate salt, anhydrous (Na2SO4)
syringe to inject the organic phase into the • Type P2 filter paper (area: 95 cm2,
aqueous phase, as the aqueous phase is pore size: 1–3 μm)
being stirred by a magnetic stir bar). – Filter paper cut into a circle with a
• The newly formed aqueous suspension was 11-cm diameter
pumped to the vacuum distillation appara- • Vacuum pump
tus, where the methanol content in the Method
slurry was reduced. • At room temperature, a 1.5-M solution of
• To dry the drug particles, the aqueous Na2SO4 was added to an aqueous nanopar-
suspensions were frozen in liquid nitrogen ticle dispersion (10 mg/mL Itz) at a ratio of
and then lyophilized to powders. 12:5 v:v (salt solution:dispersion) and
Results allowed to sit for 3 min.
• Particle sizing results were inferred from – Within seconds, flocs formed and were
BET surface area measurements. observed to take up the entire volume of
– Lower drug/stabilizer ratios resulted in the nanoparticle dispersion/salt solution
smaller particle sizes. mixture.
Average particle sizes were 90, 200, – After 3 min, larger flocs formed and the
270, and 355 nm for a 1/2, 1/1, 2/1, flocs creamed, taking up ~20% of the
and 4/1 Itz/HPMC ratio, respectively. original volume.
• Scanning electron microscopy • The flocculated suspension was filtered
demonstrated that the particles were spheri- through the filter paper under vacuum
cal and confirmed particle-size estimates until water was no longer observed on top
from BET surface area measurements. of the filter cake (typically <8 min for
• Contact-angle measurements demonstrated ~200 mL of the nanoparticle suspension/
that the HPMC was primarily concentrated salt solution mixture).
on the particle surface and not within the • An aqueous HPMC solution (30 mL at the
interior of the particle. same concentration as the aqueous phase
during nanoparticle precipitation) was Materials and Equipment

cooled in an ice bath. Immediately after • Poorly water-soluble drug (proprietary)
filtration, the chilled HPMC solution was – MW ¼ 536.6 g/mol
used to rinse the filter cake. – Tmelt ¼ 120 C
– For the 8:1:2 Itz:P127:HPMC particles, – Solubility: <0.05 mg/mL in water,
a 2.5-mg/mL HPMC solution was used. >200 mg/mL in dimethylacetamide
• The filter cake was dried at room tempera- • Stabilizing polymer: HPMC-AS-LF
ture and atmospheric pressure overnight. • Solvent: dimethylacetamide
• Dried powders were gently scraped off the • Antisolvent: aqueous solution maintained
filter paper with a spatula. between pH 1 and 3 and temperature at
Results 5 C 2 C
• Static light scattering results showed that • Vessel containing cooled, pH controlled
the flocculated nanoparticles redispersed antisolvent
back down to near-original particle sizes • Vessel containing drug and polymer
in DI water. solution
– Before flocculation: D10/D50/D90 were • Vacuum filtration capabilities
110/340/2260 nm. • Forced air oven or fluid bed dryer
– After salt flocculation and redispersion Method
in water with 5 min of sonication: • A solution containing 20% (w/w) API and
D10/D50/D90 were 120/370/1480 nm. polymer (ratio of API to polymer 4:6) was
• Scanning electron microscopy images con- dissolved in dimethylacetamide.
firm the ~300-nm primary particle sizes • The solution was then added to the chilled
reported by light scattering (see Fig. 12.29). aqueous acidic vessel (ratio of solvent to
• 94 wt% drug loading (% of drug in dried antisolvent to be maintained as 1:10 during
powder). precipitation) allowing for rapid
• Drug yields of ~90% were obtained, as coprecipitation.
calculated compared to the amount of drug • The precipitate was washed using the same
in the initial dispersion. acidic aqueous solution followed by water
• Contact-angle measurements indicated that washings.
the stabilizers were concentrated on the • Once washed, the precipitate was collected
particle surface, not within the particle as a wet cake by a vacuum filtration device.
interior. • The wet cake was dried in a forced-air oven
• Temperature-modulated differential scan- or fluid bed dryer at 45 C 5 C.
ning calorimetry (mDSC) indicated that pri- • MBP was then de-lumped using a hammer
mary nanoparticles remained amorphous mill to achieve the desired particle size.
after salt flocculation. Results
• Differential scanning calorimetry indicated
Method Capsule 8 that the compound exhibited a single Tg,
indicating a single-phase system
Microprecipitated Bulk Powder (MBP)
demonstrating a molecular dispersion of
the drug within the carrier.
Based on the method reported by Shah et al. • XRD demonstrated the amorphous nature
(2012) of the MBP product.
Objective • Transmission electron microscopy showed
• To obtain stable amorphous particles using uniform material across an area of 0.2 μm.
MBP technology
• The dissolution testing demonstrated a fast • Solvent CO2 was added to the reaction
release rate and extent of supersaturation vessel.
with a 20-fold supersaturation maintained. • Once the pressure reached the desired
• Animal studies in rats confirmed the in vitro value, the drug solution and CO2 fluid
results with exposures ~40-fold higher as were co-sprayed via a co-axial nozzle.
compared to nanosuspensions of the drug. • After co-spraying, the CO2 fluid containing
• Stability studies demonstrated the unusu- IMC and ethanol was expanded from the
ally stable character of this amorphous for- reaction vessel to an aqueous media of
mulation under accelerated conditions. 30 mL using the backpressure regulator.
• Critical process conditions leading to a • Once the expansion was finished, the sus-
robust process have been identified as pension was dispersed by sonication.
follows: • The suspension was added to a 200-mL
– Precipitation rate stainless steel container and freeze-dried at
– Solvent /antisolvent ratio 120 C for 72 h.
– Temperature Results
– Hydrodynamic conditions • Production of spherical nanoparticles of
– Washing cycles indomethacin was achieved.
– Drying • Pressure and temperature affected yield
• Drug yields of 90% were calculated. (26.7–47.7%) and particle size.
• 25 MPa and 40 C appeared as optimal
Method Capsule 9 conditions.
• In the high-pressure vessel, the particle size
Rapid Expansion from Supercritical to Aque-
increases with increasing temperature. Noz-
ous Solution
zle geometry, solubility, and nature of
• Based on the method reported by Tozuka
solute–solvent interaction affect particle
et al. (2010)
properties.
• Objective
• Freeze-dried samples reproduced submi-
• To obtain nanoparticles using RESAS
cron particles when dispersed in water.
precipitation
• Scanning electron microscopy indicated
• Materials and Equipment
that spherical shape particles were obtained
• Indomethacin
with a pressure of 25–40 MPa.
• Ethanol
• X-ray diffraction demonstrated low crystal-
• Polyvinyl alcohol
linity in the samples with an amorphous
• RESAS apparatus comprising
component of indomethacin.
• CO2 pump
• The stable Υ form and metastable α crystal-
• Solution pump
line forms are obtained depending on
• Pressure regulator
processing parameters.
• Reaction vessel
• Dissolution profile indicated that scCO2-
• Precipitation unit
treated samples dissolved 90% of the drug
• Freeze drier
in 10 min, which was significantly faster
• Stainless steel container
than commercial indomethacin.
Method
• This was attributed to the nanorange
• Indomethacin was dissolved into a 10-mL
particles leading to an increase in surface
ethanol to a 20-mg/mL solution.
area and the formation of a metastable form
• Liquefied CO2 was added to the vessel at
having higher solubility.
14 ml/min.
Method Capsule 10 • After precipitation occurred, the sample

was collected.
Supercritical Antisolvent (SAS) Process
• The collected samples were washed by pure
scCO2 through the high-pressure vessel to
Based on the method reported by Djerafi et al. remove the remaining solvent.
(2017) • After the washing step, particles were
Objective recovered on a metal frit filter while slowly
• To obtain coprecipitated nanoparticles depressurizing the autoclave.
using SAS precipitation
Materials and Equipment Results
• Rifampicin
• Production of rifampicin nanoparticles was
• Ethylcellulose
achieved.
• Ethyl acetate
• Drug/polymer ratio and different solvent
• DMSO
affected yield (27–85%), particle size, and
• Phosphate-buffered saline
morphology
• SAS apparatus comprising
• 1:4 ratio of rifampicin/ethyl cellulose
• CO2 high-pressure pump
dissolved with EtAc/DMSO (70/30) produced
• CO2 frit filter
quasi-spherical particles with the highest yield
• High-pressure precipitator
at 85%.
• Organic solution
• 1:4 ratio of rifampicin/ethyl cellulose
• Capillary tube
dissolved with EtAc produced quasi-spherical
• High-pressure liquid pump
particles with the smallest particle size at
• Metal frit filter
139 nm and 70% yield.
• Backpressure regulator
• Scanning electron microscopy indicated quasi-
• Solvent trap
spherical shape of particles was produced by
• Autoclave
the coprecipitated process.
• Stainless steel container
• X-ray diffraction demonstrated coprecipitated
rifampicin, and ethyl cellulose prepared with
Method ethyl acetate/DMSO had amorphous
• Rifampicin and ethyl cellulose were morphology.
dissolved in an organic solvent. • Physical stability of the amorphous particles
• The autoclave was used to reach the opti- was achieved at up to 6 months of storage at
mum temperature (35oC) and the con- room temperature, 75%RH.
trolled pressure (10 MPa) by the • Dissolution profile indicated that
backpressure regulator. coprecipitated rifampicin exhibited sustained
• Preheated CO2 was loaded into the auto- release of rifampicin.
clave with controlled flow rates (67 g/min
and 17 g/min). Method Capsule 11
• The organic solvent was dispersed by a Rapid Expansion of Supercritical Solutions
capillary tube, while scCO2 stream was with Solid Cosolvents (RESS-SC)
configurated to the desirable solvent/CO2 • Based on the method reported by Sodeifan
molar ratio. and Sajadian (2018)
• The organic solvent was replaced by the Objective
rifampicin and ethyl cellulose mixture. • To obtain nanoparticles using RESS-SC
• The pressure of 0.02 MPa and temperature precipitation
of 1oC were controlled during the precipi-
tation process.
Materials and Equipment • The longer spray distance created smaller

• Letrozole particles.
• Methanol • Different solid and co-solvent
• Carbon dioxide (CO2) concentrations affected the solubility of
• RESS-SC apparatus comprising the drug in CO2: the optimum solid
• High-pressure CO2 pump co-solvent concentration for letrozole was
• Extraction cell was controlled temperature 7 %w/w.
in an oven • Scanning electron microscopy indicated
• Orifice nozzle that spherical shape particles were obtained
• Stainless steel tubing with pressures of 18–36 MPa.
• Fine needle valve • Dynamic light scattering showed the lowest
• Filter mean particle diameter of 19.0 nm when
• Precipitation unit produced at the optimum conditions.
• Stainless steel container • X-ray diffraction demonstrated the lower
• Vacuum oven degree of crystallinity of letrozole
Method nanoparticles.
• CO2 was introduced into the extraction cell • Dissolution profile indicated that the
by a high-pressure pump to reach a desir- processed letrozole has a higher dissolution
able pressure (18–36 MPa). rate of 92% in 75 min compared with 51%
• Letrozole 2 g was dissolved into 1 g meth- dissolution rate of the unprocessed
anol, then loaded into the extraction cell. letrozole.
• CO2 pumped to the saturated extraction cell • Difference factors that measured the differ-
was heated to a desirable temperature ence of processed and unprocessed
(318.2–338.2 K) in the oven. letrozole at each time point were
• The saturated letrozole–CO2 solution demonstrated at 17.10 and 184.80,
flowed at a rate of 2.40 L/h into an orifice respectively
nozzle located inside the collection vessel.
• Letrozole nanoparticles were collected by a Acknowledgments The current authors would like to
filter in the expansion chamber. thank and acknowledge the significant contribution of the
previous authors of this chapter from the second edition.
• After particle collection, the co-solvent was
This current third edition chapter is a revision and update
removed by a vacuum oven at 0.5 mbar, of the previous edition.
318 K, until no additional weight loss
occurred.
Results References
• Production of spherical nanoparticles of
Albano AA, Phuapradit W, Sandhu HK, Shah NH. Stable
letrozole was achieved. complexes of poorly soluble compounds in ionic
• Pressure, temperature, and spray distance polymers. US 6,350,786. 2002.
affected particle size. Arora D, Sedev R, Beh CC, Priest C, Foster
• 36 MPa and 338.2 K appeared as optimal NR. Precipitation of drug particles using a gas
antisolvent process on a high-pressure microfluidic
conditions. platform. Ind Eng Chem Res. 2020;59(25):11905–13.
• Spray distance of 7 cm appeared as optimal Bagheri H, Hashemipour H, Ghader S. Population balance
conditions. modeling: application in nanoparticle formation
• In the high-pressure vessel, the particle size through rapid expansion of supercritical solution.
Comput Part Mech. 2019;6(4):721–37.
decreased with increasing temperature. Bakhbakhi Y, Rohani S, Charpentier PA. Micronization of
• The higher pressure leading to increased phenanthrene using the gas antisolvent process: part
super-saturation produced smaller particles. 2. Theoretical study. Ind Eng Chem Res. 2005;44(19):
7345–51.
Baldyga J, Czarnocki R, Shefeunov BY, Smith Chen X, Vaughn JM, Yacaman MJ, Williams RO,
KB. Particle formation in supercritical fluids – scale- Johnston KP III. Rapid dissolution of high-potency
up problem. Chem Eng Res Des. 2010;88:331–41. danazol particles produced by evaporative precipita-
Bhujbal SV, Pathak V, Zemlyanov DY, Taylor LS, Zhou tion into aqueous solution. J Pharm Sci. 2004b;93(7):
QT. Physical stability and dissolution of lumefantrine 1867–78.
amorphous solid dispersions produced by spray anti- Chen X, Lo CY-L, Sarkari M, Williams RO, Johnston KP
solvent precipitation. J Pharm Sci. 2020; III. Ketoprofen nanoparticle gels formed by evapora-
Blankschtein D, Thurston GM, Benedek tive precipitation into aqueous solution. AICHE
GB. Phenomenological theory of equilibrium thermo- J. 2006;52(7):2428–35.
dynamic properties and phase separation of micellar Chen XX, Matteucci ME, Lo CY, Johnston KP, Williams
solutions. J Chem Phys. 1986;85(12):7268–88. RO. Flocculation of polymer stabilized nanocrystal
Bosselmann S, Nagao M, Chow KT, Williams RO, 3rd. suspensions to produce redispersible powders. Drug
Influence of formulation and processing variables on Dev Ind Pharm. 2009;35(3):283–96.
properties of itraconazole nanoparticles made by Chen B-Q, Kankala RK, Wang S-B, Chen A-Z. Continu-
advanced evaporative precipitation into aqueous solu- ous nanonization of lonidamine by modified-rapid
tion. AAPS PharmSciTech. 2012;13(3):949–60. expansion of supercritical solution process. J Supercrit
Bristow S, Shekunov T, Shekunov BY, York P. Analysis Fluids. 2018;133:486–93.
of the supersaturation and precipitation process with Cheng KK, Yeung CF, Ho SW, Chow SF, Chow AHL,
supercritical CO2. J Supercrit Fluids. 2001;21(3): Baum L. Highly stabilized curcumin nanoparticles
257–71. tested in an in vitro blood–brain barrier model and in
Bustami R, Chan H-K, Dehghani F, Foster N. In Genera- alzheimer’s disease Tg2576 mice. AAPS J. 2012;15
tion of protein microparticles using high pressure (2):324–36.
modified carbon dioxide. In International symposium Chiou H, Chan H-K, Prud’homme RK, Raper
on supercritical fluids, Atlanta, GA. 2000. JA. Evaluation on the use of confined liquid impinging
Cardoso FAR, Vogel EM, Souza MF, Cardozo-Filho jets for the synthesis of nanodrug particles. Drug Dev
L. Mathematical modeling to predict the size and Ind Pharm. 2008a;34(1):59–64.
nucleation rate of micro and nanoparticles using the Chiou H, Chan H-K, Heng D, Prud’homme RK, Raper
scale-up process with supercritical CO2. J Supercrit JA. A novel production method for inhalable cyclo-
Fluids. 2019;154 sporine A powders by confined liquid impinging jet
Carl LY. Chemical properties handbook. New York: precipitation. J Aerosol Sci. 2008b;39(6):500–9.
McGraw-Hill; 1999. Chopra M, Jain R, Dewangan AK, Varkey S, Mazumder
Chan H-K, Chew NYK. Novel alternative methods for the S. Design of curcumin loaded polymeric nanoparticles-
delivery of drugs for the treatment of asthma. Adv optimization, formulation and characterization. J
Drug Deliv Rev. 2003;55(7):793–805. Nanosci Nanotechnol. 2016;16(9):9432–42.
Chang SA, Gray DG. The surface tension of aqueous Chow HF, Sun CC, Chow AHL. Assessment of the rela-
hydroxypropyl cellulose solutions. J Colloid Interface tive performance of a confined impinging jets mixer
Sci. 1978;67:255–65. and a multi-inlet vortex mixer for curcumin nanoparti-
Charoenchaitrakool M, Dehghani F, Foster NR, Chan cle production. Eur J Pharm Biopharm. 2014;88(2):
HK. Micronization by rapid expansion of supercritical 462–71.
solutions to enhance the dissolution rates of poorly Chow SF, Wan KY, Cheng KK, Wong KW, Sun CC,
water-soluble pharmaceuticals. Ind Eng Chem Res. Baum L, Chow AHL. Development of highly
2000;39(12):4794–802. stabilized curcumin nanoparticles by flash nanopreci-
Chattopadhyay P, Gupta RB. Production of griseofulvin pitation and lyophilization. Eur J Pharm Biopharm.
nanoparticles using supercritical CO2 antisolvent with 2015;94:436–44.
enhanced mass transfer. Int J Pharm. 2001a;228(1–2): Crison JR. Biopharmaceutical aspects of water-insoluble
19–31. drugs for oral drug delivery. In: LIU R, editor. Water-
Chattopadhyay P, Gupta RB. Production of antibiotic insoluble drug formulation. Boca Raton: CRC
nanoparticles using supercritical CO2 as antisolvent Press; 2000.
with enhanced mass transfer. Ind Eng Chem Res. Daniels R, Barta A. Pharmacopoeial cellulose ethers as oil-
2001b;40(16):3530–9. in-water emulsifiers I: interfacial properties. Eur J
Chen X, Young TJ, Sarkari M, Williams RO, Johnston Pharm Biopharm. 1994;40:128–33.
KP. Preparation of cyclosporine A nanoparticles by de la Fuente Badilla JC, Peters CJ, de Swaan AJ. Volume
evaporative precipitation into aqueous solution. Int J expansion in relation to the gas-antisolvent process. J
Pharm. 2002;242(1–2):3–14. Supercrit Fluids. 2000;17(1):13–23.
Chen X, Ill Benhayoune Z, Williams RO, Johnston de la Fuente JC, Shariati A, Peters CJ. On the selection of
KP. Rapid dissolution of high potency itraconazole optimum thermodynamic conditions for the GAS pro-
particles produced by evaporative precipitation into cess. J Supercrit Fluids. 2004;32(1–3):55–61.
aqueous solution. J Drug Deliv Sci Technol. Dearn AR. Atovaquone pharmaceutical compositions.
2004a;14(4):299–304. WO 9414426, 23 Dec 1993. 1994.
Debenedetti PG. Homogeneous nucleation in supercritical Erriguible A, Neurohr C, Revelli AL, Laugier S,
fluids. AICHE J. 1990;36(9):1289–98. Fevotte G, Subra-Paternault. Cocrystallization induced
Debenedetti PG, Tom JW, Kwauk X, Yeo SD. Rapid by compressed CO2 as antisolvent: simulation of a
expansion of supercritical solutions (RESS): batch process for the estimation of nucleation and
fundamentals and applications. Fluid Phase Equilibr. growth parameters. J Supercrit Fluids. 2015;98:194–
1993;82:311–21. 203.
Desai D, Diodone R, Go Z, Ibrahim PN, Iyer R, Mair HJ, Esfandiari N, Ghoreishi SM. Synthesis of 5-Fluorouracil
Sandhu HK, Shah NH, Visor G, Wyttenbach N, nanoparticles via supercritical gas antisolvent process.
Lauper S, Pudewell J, Wierschem F. Compositions J Supercrit Fluids. 2013a;84:205–10.
and uses thereof. US20100310659 A1 (03.31/ Esfandiari N, Ghoreishi SM. Kinetics modeling of ampi-
2010). 2010. cillin nanoparticles synthesis via supercritical gas
Djerafi R, Swanepoel A, Crampon C, Kalombo L, antisolvent process of supercritical fluids. Chem Eng
Labuschagne P, Badens E, Masmoudi Technol. 2013b;81:119–27.
Y. Supercritical antisolvent co-precipitation of rifam- Essa EA, Elmarakby AO, Donia AMA, El Maghraby
picin and ethyl cellulose. Eur J Pharm Sci. 2017;102: GM. Controlled precipitation for enhanced dissolution
161–71. rate of flurbiprofen: development of rapidly
Dodds S, Wood JA, Charpentier PA. Modeling of the disintegrating tablets. Drug Dev Ind Pharm. 2017;43
gas-antisolvent (GAS) process for crystallization of (9):1430–9.
beclomethasone dipropionate using carbon dioxide. European Medicines Agency. Zelboraf. Assessment
Ind Eng Chem Res. 2007;46:8009–17. report. 2012.
Domingo C, Berends E, van Rosmalen GM. Precipitation Falk R, Randolph TW, Meyer JD, Kelly RM, Manning
of ultrafine organic crystals from the rapid expansion MC. Controlled release of ionic compounds from poly
of supercritical solutions over a capillary and a frit (L-lactide) microspheres produced by precipitation
nozzle. J Supercrit Fluids. 1997;10(1):39–55. with a compressed antisolvent. J Control Release.
Dong Y, Ng WK, Shen S, Kim S, Tan RBH. Preparation 1997;44(1):77–85.
and characterization of spironolactone nanoparticles by Feng J, Zhang Y, McManus SA, Ristroph KD, Lu HD,
antisolvent precipitation. Int J Pharm. 2009;375(1–2): Gong K, White CE, Prud'homme RK. Rapid Recovery
84–8. of Clofazimine-Loaded Nanoparticles with Long-Term
Dong Y, Ng WK, Hu J, Shen S, Tan RBH. A continuous Storage Stability as Anti-Cryptosporidium Therapy.
and highly effective static mixing process for ACS Appl Nano Mater. 2018;1(5):2184–94.
antisolvent precipitation of nanoparticles of poorly Franco P, Reverchon E, De Marco I. Zein/diclofenac
water-soluble drugs. Int J Pharm. 2010;386(1–2): sodium coprecipitation at micrometric and nanometric
256–61. range by supercritical antisolvent processing. J CO2
Dong B, Cheow WS, Hadinoto K. Recovery of amorphous Util. 2018;27:366–73.
drug-polyelectrolyte nanoparticle complex by salt- Franklin RK, Edwards JR, Chernyak Y, Gould RD,
induced flocculation to circumvent the need for ultra- Henon F, Carbonell RG. Formation of perfluoro-
centrifugation. Adv Powder Technol. 2020;31(8): polyether coatings by the rapid expansion of supercrit-
3102–9. ical solutions (RESS) process. Part 2: numerical
Elvassore N, Parton T, Bertucco A, Di Noto V. Kinetics of modeling. Ind Eng Chem Res. 2001;40(26):6127–39.
particle formation in the gas antisolvent precipitation Fusaro F, Mazzotti M, Muhrer G. Gas antisolvent recrys-
process. AICHE J. 2003;49(4):859–68. tallization of paracetamol from acetone using com-
Elvassore N, Cozzi F, Bertucco A. Mass transport pressed carbon dioxide as antisolvent. Cryst Growth
modeling in a gas antisolvent process. Ind Eng Chem Des. 2004;4(5):881–9.
Res. 2004;43(16):4935–43. Fusaro F, Haenchen M, Mazzotti M, Muhrer G,
Engstrom JD, Simpson DT, Cloonan C, Lai ES, Williams Subramaniam B. Dense gas antisolvent precipitation:
RO III, Kitto GB, Johnston KP. Stable high surface a comparative investigation of the GAS and PCA
area lactate dehydrogenase particles produced by spray techniques. Ind Eng Chem Res. 2005;44(5):1502–9.
freezing into liquid nitrogen. Eur J Pharm Biopharm. Gao L, Liu G, Wang X, Liu F, Xu Y, Ma J. Preparation of
2007;65(2):163–74. a chemically stable quercetin formulation using
Engstrom JD, Lai ES, Ludher B, Chen B, Milner TE, Kitto nanosuspension technology. Int J Pharm. 2011;404
GB, Williams RO III, Johnston KP. Formation of sta- (1–2):231–7.
ble submicron protein particles by thin film freezing. García-Casas I, Montes A, Pereyra C, Martinez de la Ossa
Pharm Res. 2008;25(6):1334–6. EJ. Generation of quercetin/cellulose acetate phthalate
Erriguible A, Laugier S, Laté M, Subra-Paternault systems for delivery by supercritical antisolvent pro-
P. Effect of pressure and non-isothermal injection on cess. Eur J Pharm Sci. 2017a;100:79–86.
re-crystallization by CO2 antisolvent: solubility García-Casas I, Montes A, Pereyra C, Martínez de la Ossa
measurements, simulation of mixing and experiments. EJ. Co-precipitation of mangiferin with cellulose ace-
J Supercrit Fluids. 2013;76:115–25. tate phthalate by Supercritical antisolvent process. J
CO2 Util. 2017b;22:197–207.
Gardner CR, Walsh CT, Almarsson O. Drugs as materials: precipitation with a compressed-fluid antisolvent. J
valuing physical form in drug discovery. Nat Rev Drug Supercrit Fluids. 2006;37(2):242–53.
Discov. 2004;3(11):926–34. Johnson BK, Prud'homme RK. Flash NanoPrecipitation of
Gassmann P, List M, Schweitzer A, Sucker H. Hydrosols – Organic Actives and Block Copolymers using a Con-
alternatives for the parenteral application of poorly fined Impinging Jets Mixer. Australian Journal of
water-soluble drugs. Eur J Pharm Biopharm. 1994;40 Chemistry. 2003;56(10):1021–4.
(2):64–72. Johnson BK, Prud’homme RK. Flash nanoprecipitation of
Gupta RB. Nanoparticle technology for drug delivery, vol. organic actives and block copolymers using a confined
53. 1st ed. New York: Taylor & Francis; 2006. impinging jets mixer. Aust J Chem. 2003a;56(10):
p. 53–84. 1021–4.
Han J, Zhu Z, Qian H, Wohl AR, Beaman CJ, Hoye TR, Johnson BK, Prud’homme RK. Chemical processing and
Macosko CW. A simple confined impingement jets micromixing in confined impinging, jets. AIChE
mixer for flash nanoprecipitation. J Pharm Sci. J. 2003b;49(9):2264–82.
2012;101(10):4018–23. Johnson BK, Prud’homme RK. Mechanism for rapid self-
Hanna M, York P. Method + apparatus for the formation assembly of block copolymer nanoparticles. Phys Rev
of particles. WO 9836825, 20 Feb 1998 Lett. 2003c;91(11):118302/1–4.
Heimenz PC, Rajagopalan R. Principles of colloid and Johnson BK, Saad W, Prud’homme
surface chemistry. New York: Marcel Dekker; 1997. RK. Nanoprecipitation of pharmaceuticals using
Helfgen B, Turk M, Schaber K. Hydrodynamic and aero- mixing and block copolymer stabilization. ACS Sym
sol modelling of the rapid expansion of supercritical Ser. 2006;924:278–291; Polymeric Drug Delivery II
solutions (RESS-process). J Supercrit Fluids. 2003;26 Jouyban A, Rehman M, Shekunov BY, Chan H-K, Clark
(3):225–42. BJ, York P. Solubility prediction in supercritical CO2
Hiendrawan S, Veriansyah B, Tjandrawinata R. Prepara- using minimum number of experiments. J Pharm Sci.
tion of Fenofibrate Microparticles Using Top-down 2002;91(5):1287–95.
and Bottom-up Processes. Procedia chemistry. Kayrak D, Akman U, Hortaçsu Ö. Micronization of Ibu-
2014;9:257–64. profen by RESS. J Supercrit Fluids. 2003;26(1):17–31.
Hitt JE, Tucker CJ, Evans JC, Curtis CA, Svenson Keck CM, Mueller RH. Drug nanocrystals of poorly solu-
S. Process to precipitate drug particles. US ble drugs produced by high pressure homogenization.
20030049323, 08/27/2002. 2003. Eur J Pharm Biopharm. 2006;62(1):3–16.
Hitt JE, Rogers TL, Gillespie IB, Scherzer BD, Garcia PC, Keshavarz A, Karimi-Sabet J, Fattahi A, Golzary A,
Beck NS, Tucker CJ, Young TJ, Hayes DA, Williams Rafiee-Tehrani M, Dorkoosh FA. Preparation and
RO III, Johnston KP, McConville JT, Peters JI, characterization of raloxifene nanoparticles using
Talbert R, Burgess D. Enhanced delivery of pharma- Rapid Expansion of Supercritical Solution (RESS).
ceutical compositions to treat life threatening The Journal of Supercritical Fluids. 2012;63:169–79.
infections. WO 2006026502, 08/26/2005. 2006. Kikic I, De Zordi N, Moneghini M, Solinas D. Solubility
Hu J, Johnston KP, Williams RO III. Nanoparticle engi- estimation of drugs in ternary systems of interest for
neering processes for enhancing the dissolution rates of the antisolvent precipitation processes. J Supercrit
poorly water soluble drugs. Drug Dev Ind Pharm. Fluids. 2010;55(2):616–22.
2004;30(3):233–45. Kim MS, Kim JS, Park HJ, Cho WK, Cha KH, Hwang
Hu J, Ng W-K, Dong Y-C, Shen S-C, Tan RBH. Continu- SJ. Enhanced bioavailability of sirolimus via prepara-
ous and scalable process for water-redispersible tion of solid dispersion nanoparticles using a supercrit-
nanoformulation of poorly aqueous soluble APIs by ical antisolvent process. Int J Nanomed. 2011;6:2997–
antisolvent precipitation and spray-drying. Int J Pharm. 3009.
2011;404(1–2):198–204. Kipp JE. The role of solid nanoparticle technology in the
Jacobs C, Kayser O, Muller RH. Nanosuspensions as a parenteral delivery of poorly water-soluble drugs. Int J
new approach for the formulation for the poorly solu- Pharm. 2004;284(1–2):109–22.
ble drug tarazepide. Int J Pharm. 2000;196:161–4. Ksibi H, Subra P, Garrabos Y. Formation of fine
Jarmer DJ, Lengsfeld CS, Randolph TW. Manipulation of powdersof caffeine by RESS. Adv Powder Technol.
particle size distribution of poly(l-lactic acid) 1995;6(1):25–33.
nanoparticles with a jet-swirl nozzle during precipita- Kudryashova EV, Sukhoverkov KV, Deygen IM, Vorobei
tion with a compressed antisolvent. J Supercrit Fluids. AM, Pokrovskiy OI, Parenago OO, Presnov DE,
2003;27(3):317–36. Egorov AM. Moxifloxacin micronization via supercrit-
Jarmer DJ, Lengsfeld CS, Randolph TW. Nucleation and ical antisolvent precipitation. Russian Journal of Phys-
growth rates of poly(L-lactic acid) microparticles dur- ical Chemistry B. 2018;11(7):1153–62.
ing precipitation with a compressed-fluid antisolvent. Kuk DH, Ha ES, Ha DH, Sim WY, Lee SK, Jeong JS, Kim
Langmuir. 2004;20(17):7254–64. JS, Baek IH, Park H, Choi DH, Yoo JW, Jeong SH,
Jarmer DJ, Lengsfeld CS, Randolph TW. Scale-up criteria Hwang SJ, Kim MS. Development of a resveratrol
for an injector with a confined mixing chamber during nanosuspension using the antisolvent precipitation
method without solvent removal, Based on a Quality
by Design (QbD) approach. Pharmaceutics. 2019;11 Liversidge EM, Liversidge GG, Cooper ER. Nanosizing: a
(12) formulation approach for poorly-water-soluble
Larson RG. The structure and rheology of complex fluids. compounds. Eur J Pharm Sci. 2003;18:113–20.
New York: Oxford University Press; 1999. Loftsson T. Drug solubilization by complexation. Int J
Lengsfeld CS, Delplanque JP, Barocas VH, Randolph Pharm. 2017;531(1):276–80.
TW. Mechanism governing microparticle morphology Lopez-Iglesias C, Quilez C, Barros J, Velasco D, Alvarez-
during precipitation by a compressed antisolvent: Lorenzo C, Jorcano JL, Monteiro FJ, Garcia-Gonzalez
atomization vs nucleation and growth. J Phys Chem CA. Lidocaine-loaded Solid Lipid Microparticles
B. 2000;104(12):2725–35. (SLMPs) produced from gas-saturated solutions for
Li K, Xu Z. A review of current progress of supercritical wound applications. Pharmaceutics. 2020;12(9)
fluid technologies for e-waste treatment. J Clean Prod. Maa Y-F, Nguyen P-A, Sweeney T, Shire SJ, Hsu
2019;227:794–809. CC. Protein inhalation powders: spray drying vs
Li J, Rodrigues M, Paiva A, Matos HA, Gomes de spray freeze drying. Pharm Res. 1999;16(2):249–54.
Azevedo E. Modeling of the PGSS process by crystal- Machado APDF, Rueda M, Barbero GF, Martín Á, Cocero
lization and atomization. AIChE J. 2005;51(8): MJ, Martínez J. Co-precipitation of anthocyanins of
2343–57. the extract obtained from blackberry residues by
Li S, Liu Y, Liu T, Zhao L, Zhao J, Feng N. Development pressurized antisolvent process. J Supercrit Fluids.
and in-vivo assessment of the bioavailability of 2018;137:81–92.
oridonin solid dispersions by the gas anti-solvent tech- Machmudah S, Winardi S, Wahyudiono KH, Goto
nique. Int J Pharm. 2011;411(1-2):172–7. M. Formation of fine particles from curcumin/PVP by
Li M, Suriel I, Vekaria J, Proske J, Orbe P, Armani M, the supercritical antisolvent process with a coaxial
Dave RN, Bilgili E. Impact of dispersants on dissolu- nozzle. ACS Omega. 2020;5(12):6705–14.
tion of itraconazole from drug-loaded, surfactant-free, Martin A, Cocero MJ. Numerical modeling of jet hydro-
spray-dried nanocomposites. Powder Technol. dynamics, mass transfer, and crystallization kinetics in
2018;339:281–95. the supercritical antisolvent (SAS) process. J Supercrit
Lim RTY, Ng WK, Tan RBH. Amorphization of pharma- Fluids. 2004;32(1–3):203–19.
ceutical compound by co-precipitation using supercrit- Martin A, Cocero MJ. Micronization processes with super-
ical anti-solvent (SAS) process (Part I). J Supercrit critical fluids: fundamentals and mechanisms. Adv
Fluids. 2010;53(1-3):179–84. Drug Deliv Rev. 2008;60(3):339–50.
Limayem I, Charcosset C, Fessi H. Purification of nano- Martin A, Bouchard A, Hofland GW, Witkamp GJ,
particle suspensions by a concentration/diafiltration Cocero MJ. Mathematical modeling of the mass trans-
process. Sep Purif Technol. 2004;38:1–9. fer from aqueous solutions in a supercritical fluid dur-
Lin C, Muhrer G, Mazzotti M, Subramaniam B. Vapor- ing particle formation. J Supercrit Fluids. 2007;41(1):
liquid mass transfer during gas antisolvent recrystalli- 126–37.
zation: modeling and experiments. Ind Eng Chem Res. Martin A, Pham H, Kilzer A, Kareth S, Weidner
2003;42(10):2171–82. E. Micronization of polyethylene glycol by PGSS
Lipinski CA. Avoiding investment in doomed drugs. Is (particles from gas saturated solutions)-drying of aque-
poor solubility an industry wide problem? Curr Drug ous solutions. Chem Eng Process. 2010;49:1259–66.
Discov. 2001:17–9. Matteucci ME, Hotze MA, Johnston KP, Williams RO III.
Lipinski C. Poor aqueous solubility- an industry wide Drug nanoparticles by antisolvent precipitation:
problem in drug discovery. Am Pharm Rev. 2002;5: mixing energy versus surfactant stabilization.
82–5. Langmuir. 2006;22(21):8951–9.
Liu G-T, Nagahama K. Application of rapid expansion of Matteucci ME, Brettmann BK, Rogers TL, Elder EJ,
supercritical solutions in the crystallization separation. Williams RO, Johnston KP. Design of potent amor-
Ind Eng Chem Res. 1996;35:4626–34. phous drug nanoparticles for rapid generation of highly
Liu G, Zhang D, Jiao Y, Zheng D, Liu Y, Duan C, et al. supersaturated media. Mol Pharm. 2007;4(5):782–93.
Comparison of different methods for preparation of a Matteucci ME, Paguio JC, Miller MA, Williams RO,
stable riccardin D formulation via nano-technology. Int Johnston KP III. Flocculated amorphous nanoparticles
J Pharm. 2012;422(1–2):516–22. for highly supersaturated solutions. Pharm Res.
Liu Y, Cheng C, Liu Y, Prud’homme RK, Fox 2008;25(11):2477–87.
RO. Mixing in a multi-inlet vortex mixer (MIVM) for Matteucci ME, Paguio JC, Miller MA III, Williams RO,
flash nano-precipitation. Chem Eng Sci. 2008;63(11): Johnston KP. Highly supersaturated solutions from
2829–42. dissolution of amorphous itraconazole microparticles
Liversidge GG, Cundy KC. Particle size reduction for at pH 6.8. Mol Pharm. 2009;6(2):375–85.
improvement of oral bioavailability of hydrophobic Mendez-Santiago J, Teja AS. The solubility of solids in
drugs: I. Absolute oral bioavailability of supercritical fluids. Fluid Phase Equilibr. 1999;158–
nanocrystalline danazol in beagle dogs. Int J Pharm. 160:501–10.
1995;125(1):91–7. Miller MA, DiNunzio J, Matteucci ME, Ludher BS III,
Williams RO, Johnston KP. Flocculated amorphous
itraconazole nanoparticles for enhanced in vitro super- Overhoff KA, Johnston KP, Tam J, Engstrom J, Williams
saturation and in vivo bioavailability. Drug Dev Ind RO III. Use of thin film freezing to enable drug deliv-
Pharm. 2012;38(5):557–70. ery: a review. J Drug Deliv Sci Technol. 2009;19(2):
Mueller M, Meier U, Kessler A, Mazzotti M. Experimental 89–98.
study of the effect of process parameters in the recrys- Padrela L, Rodrigues MA, Velaga SP, Matos HA, de
tallization of an organic compound using compressed Azevedo EG. Formation of indomethacin-saccharin
carbon dioxide as antisolvent. Ind Eng Chem Res. cocrystals using supercritical fluid technology. Eur J
2000;39(7):2260–8. Pharm Sci. 2009;38(1):9–17.
Muhrer G, Mazzotti M. Precipitation of lysozyme Padrela L, Rodrigues MA, Velaga SP, Fernandes AC,
nanoparticles from dimethyl sulfoxide using carbon Matos HA, Gomes de Azevedo E. Screening for phar-
dioxide as antisolvent. Biotechnol Prog. 2003;19(2): maceutical cocrystals using the supercritical fluid
549–56. enhanced atomization process. J Supercrit Fluids.
Muhrer G, Lin C, Mazzotti M. Modeling the gas 2010;53(1–3):156–64.
antisolvent recrystallization process. Ind Eng Chem Pandit N, Trygstad T, Croy S, Bohorquez M, Koch
Res. 2002;41(15):3566–79. C. Effect of salts on the micellization, clouding, and
Muhrer G, Mazzotti M, Muller M. Gas antisolvent recrys- solubilization behavior of pluronic F127 solutions. J
tallization of an organic compound. Tailoring product Colloid Interface Sci. 2000;222:213–20.
PSD and scaling-up. J Supercrit Fluids. 2003;27(2): Pang P, Englezos P. Phase separation of polyethylene
195–203. oxide (PEO)-water solution and its relationship to the
Muller RH, Bohm BHL. Colloidal drug carriers expert flocculating capability of the PEO. Fluid Phase
meeting, 3rd meeting, Berlin, Germany, May 29–31, Equilibr. 2002;194–197:1059–66.
1997. In: Mueller RH, Benita S, Boehm BHL, editors. Park J-H, Baek M-J, Lee J-Y, Kim K-T, Cho H-J, Kim D-
Nanosuspensions. Berlin: Medpharm Scientific; 1997. D. Preparation and characterization of sorafenib-
p. 149–74. loaded microprecipitated bulk powder for enhancing
Muller RH, Jacobs C, Kayser O. Nanosuspensions as oral bioavailability. International journal of
particulate drug formulations in therapy. Rationale for pharmaceutics. 2020;589:119836.
development and what we can expect for the future. Park J, Cho W, Cha KH, Ahn J, Han K, Hwang
Adv Drug Deliv Rev. 2001;47(1):3–19. SJ. Solubilization of the poorly water soluble drug,
Müllers K, Paisana M, Wahl MA. Simultaneous formation telmisartan, using supercritical anti-solvent (SAS) pro-
and micronization of pharmaceutical cocrystals by cess. Int J Pharm. 2013;441(1-2):50–5.
Rapid Expansion of Supercritical Solutions (RESS). Park HJ, Yoon TJ, Kwon DE, Yu K, Lee Y-W.
Pharm Res. 2015;32(2):702–13. Coprecipitation of hydrochlorothiazide/PVP for the
Muthancheri I, Long B, Ryan KM, Padrela L, dissolution rate improvement by precipitation with
Ramachandran R. Development and validation of a compressed fluid antisolvent process. J Supercrit
two-dimensional population balance model for a super- Fluids. 2017;126:37–46.
critical CO2 antisolvent batch crystallization process. Pathak P, Meziani MJ, Desai T, Sun Y-P. Nanosizing drug
Advanced Powder Technology. 2020;31(8):3191–204. particles in supercritical fluid processing. J Am Chem
Napper DH. Polymeric stabilization of colloidal Soc. 2004;126(35):10842–3.
dispersions. New York: Academic Press; 1983. Pathak P, Meziani MJ, Desai T, Sun Y-P. Formation and
Nguyen XC, Herberger JD, Burke PA. Protein powders for stabilization of ibuprofen nanoparticles in supercritical
encapsulation: a comparison of spray-freeze drying fluid processing. J Supercrit Fluids. 2006;37(3):
and spray drying of darbepoetin alfa. Pharm Res. 279–86.
2004;21(3):507–14. Pathak P, Prasad GL, Meziani MJ, Joudeh AA, Sun Y-P.
Noyes AA, Whitney WR. The rate of solution of solid Nanosized paclitaxel particles from supercritical car-
substances in their own solutions. J Am Chem Soc. bon dioxide processing and their biological evaluation.
1897;19:930–4. Langmuir. 2007;23(5):2674–9.
Okamoto H, Danjo K. Application of supercritical fluid to PerezdeDiego Y, Wubbolts FE, Jansens PJ. Modelling
preparation of powders of high-molecular weight drugs mass transfer in the PCA process using the Maxwell–
for inhalation. Adv Drug Deliv Rev. 2008;60(3): Stefan approach. J Supercrit Fluids. 2006;37(1):53–62.
433–46. Perrut M, Jung J, Leboeuf F. Enhancement of dissolution
Overhoff KA, Engstrom JD, Chen B, Scherzer BD, Milner rate of poorly-soluble active ingredients by supercriti-
TE, Johnston KP, Williams RO. Novel ultra-rapid cal fluid processes: part I: micronization of neat
freezing particle engineering process for enhancement particles. Int J Pharm. 2005;288(1):3–10.
of dissolution rates of poorly water-soluble drugs. Eur Pestieau A, Krier F, Lebrun P, Brouwers A, Streel B,
J Pharm Biopharm. 2007a;65(1):57–67. Evrard B. Optimization of a PGSS (particles from gas
Overhoff KA, Moreno A, Miller DA, Johnston KP, saturated solutions) process for a fenofibrate lipid-
Williams RO. Solid dispersions of itraconazole and based solid dispersion formulation. Int J Pharm.
enteric polymers made by ultra-rapid freezing. Int J 2015;485(1-2):295–305.
Pharm. 2007b;336(1):122–32.
Pratsinis SE. Simultaneous nucleation, condensation, and spray-freezing technologies. Drug Dev Ind Pharm.
coagulation in aerosol reactors. J Colloid Interface Sci. 2001a;27(10):1003–15.
1988;124(2):416–27. Rogers TL, Gillespie IB, Hitt JE, Fransen KL, Crowl CA,
Pustulka KM, Wohl AR, Lee HS, Michel AR, Han J, Hoye Tucker CJ, Kupperblatt GB, Becker JN, Wilson DL,
TR, McCormick AV, Panyam J, Macosko CW. Flash Todd C, Broomall CF, Evans JC, Elder
nanoprecipitation: particle structure and stability. Mol EJ. Development and characterization of a scalable
Pharm. 2013;10(11):4367–77. controlled precipitation process to enhance the dissolu-
Rabinow BE. Nanosuspensions in drug delivery. Nat Rev tion of poorly water-soluble drugs. Pharm Res.
Drug Discov. 2004;3(9):785–96. 2004;21(11):2048–57.
Rasenack N, Muller BW. Dissolution rate enhancement by Rossmanna M, Braeuerb A, Leipertz A, Schluecker
in situ micronization of poorly water-soluble drugs. E. Manipulating the size, the morphology and the
Pharm Res. 2002;19(12):1894–900. polymorphism of acetaminophen using supercritical
Reverchon E, De Marco I. Mechanisms controlling super- antisolvent (SAS) precipitation. J Supercrit Fluids.
critical antisolvent precipitate morphology. Chem Eng 2013;82:230–7.
J. 2011;169(1-3):358–70. Sarkari M, Brown J, Chen X, Swinnea S, Williams RO,
Reverchon E, Della Porta G. Production of antibiotic Johnston KP. Enhanced drug dissolution using evapo-
micro- and nano-particles by supercritical antisolvent rative precipitation into aqueous solution. Int J Pharm.
precipitation. Powder Technol. 1999;106(1–2):23–9. 2002;243(1–2):17–31.
Reverchon E, Pallado P. Hydrodynamic modeling of the Sekhon BS. Supercritical fluid technology: an overview of
RESS process. J Supercrit Fluids. 1996;9(4):216–21. pharmaceutical applications. Int J PharmTechnol Res.
Reverchon E, Donsi G, Gorgoglione D. Salicylic acid 2010;2(1):810–26.
solubilization in supercritical CO2 and its Shah N, Sandhu H, Phuapradit W, Lyer R, Albano A,
micronization by RESS. J Supercrit Fluids. 1993;6 Desai D, Choi D, Tang K, Tian H, Chokshi H, Go Z,
(4):241–8. Malick W, Radinov R, Shankar A, Wolff S, Mair
Reverchon E, Della Porta G, Falivene MG. Process H. Solid complexes with ionic polymers. Pharm
parameters and morphology in amoxicillin micro and Technol. 2010;32(12):46–7.
submicro particles generation by supercritical Shah N, Sandhu H, Phuapradit W, Pinal R, Iyer R,
antisolvent precipitation. J Supercrit Fluids. 2000;17 Albano A, Chatterji A, Anand S, Choi DS, Tang K,
(3):239–48. Tian H, Chokshi H, Singhal D, Malick
Reverchon E, Caputo G, De Marco I. Role of phase W. Development of novel microprecipitated bulk pow-
behavior and atomization in the supercritical der (MBP) technology for manufacturing stable amor-
antisolvent precipitation. Ind Eng Chem Res. phous formulations of poorly soluble drugs. Int J
2003a;42(25):6406–14. Pharm. 2012;438(1-2):53–60.
Reverchon E, De Marco I, Caputo G, Della Porta G. Pilot Shah N, Iyer RM, Mair HJ, Choi DS, Tian H, Diodone R,
scale micronization of amoxicillin by supercritical Fahnrich K, Pabst-PAvot A, Tang K, Scheubel E,
antisolvent precipitation. J Supercrit Fluids. 2003b;26 Grippo JF, Moreira SA, Go Z, Mouskountakis J,
(1):1–7. Louie T, Ibrahim PN, Sandhu H, Rubia L,
Reverchon E, De Marco I, Torino E. Nanoparticles pro- Chokshi H, Singhal MW. Improved human bioavail-
duction by supercritical antisolvent precipitation: a ability of Vemurafenib, a practically insoluble drug,
general interpretation. J Supercrit Fluids. 2007;43(1): using an amorphous polymer-stabilized solid disper-
126–38. sion prepared by a solvent-controlled coprecipitation
Ristroph KD, Feng J, McManus SA, Zhang Y, Gong K, process. J Pharm Sci. 2013;102(3):967–81.
Ramachandruni H, White CE, Prud'homme RK. Spray Shariati A, Peters CJ. Measurements and modeling of the
drying OZ439 nanoparticles to form stable, water- phase behavior of ternary systems of interest for the
dispersible powders for oral malaria therapy. J Transl GAS process: I. The system carbon dioxide +
Med. 2019;17(1):97. 1-propanol + salicylic acid. J Supercrit Fluids.
Rodrigues M, Peiriço N, Matos H, Gomes de Azevedo E, 2002;23(3):195–208.
Lobato MR, Almeida AJ. Microcomposites theophyl- Shekunov BY, Hanna M, York P. Crystallization process
line/hydrogenated palm oil from a PGSS process for in turbulent supercritical flows. J Crystal Growth.
controlled drug delivery systems. J Supercrit Fluids. 1999;198/199(Pt 2):1345–51.
2004;29(1–2):175–84. Shen H, Hong S, Prud’homme RK, Liu Y. Self-
Rodrigues MA, Li J, Padrela L, Almeida A, Matos HA, de assembling process of flash nanoprecipitation in a
Azevedo EG. Anti-solvent effect in the production of multi-inlet vortex mixer to produce drug-loaded poly-
lysozyme nanoparticles by supercritical fluid-assisted meric nanoparticles. J Nanopart Res. 2011;13(9):
atomization processes. J Supercrit Fluids. 2009;48(3): 4109–20.
253–60. Shoyele SA, Cawthorne S. Particle engineering techniques
Rogers TL, Johnston KP, Williams RO III. Solution-based for inhaled biopharmaceuticals. Adv Drug Deliv Rev.
particle formation of pharmaceutical powders by 2006;58(9–10):1009–29.
supercritical or compressed fluid CO2 and cryogenic
Sinswat P, Gao X, Yacaman MJ, Williams RO, Johnston Tao J, Chow SF, Zheng Y. Application of flash nanopreci-
KP. Stabilizer choice for rapid dissolving high potency pitation to fabricate poorly water-soluble drug
itraconazole particles formed by evaporative precipita- nanoparticles. Acta Pharm Sin B. 2019;9(1):4–18.
tion into aqueous solution. Int J Pharm. 2005;302 Thakur R, Gupta RB. Rapid expansion of supercritical
(1–2):113–24. solution with solid cosolvent (RESS-SC) process: for-
Sinswat P, Matteucci ME, Johnston KP, Williams RO III. mation of griseofulvin nanoparticles. Ind Eng Chem
Dissolution rates and supersaturation behavior of Res. 2005;44(19):7380–7.
amorphous repaglinide particles produced by con- Thakur R, Gupta RB. Rapid expansion of supercritical
trolled precipitation. J Biomed Nanotechnol. 2007;3 solution with solid cosolvent (RESS-SC) process: for-
(1):18–27. mation of 2-aminobenzoic acid nanoparticle. J
Sodeifian G, Sajadian SA. Solubility measurement and Supercrit Fluids. 2006a;37(3):307–15.
preparation of nanoparticles of an anticancer drug Thakur R, Gupta RB. Formation of phenytoin
(Letrozole) using rapid expansion of supercritical nanoparticles using rapid expansion of supercritical
solutions with solid cosolvent (RESS-SC). J Supercrit solution with solid cosolvent (RESS-SC) process. Int
Fluids. 2018;133:239–52. J Pharm. 2006b;308(1–2):190–9.
Sodeifian G, Sajadian SA. Utilization of ultrasonic- Thote AJ, Gupta RB. Formation of nanoparticles of a
assisted RESOLV (US-RESOLV) with polymeric hydrophilic drug using supercritical carbon dioxide
stabilizers for production of amiodarone hydrochloride and microencapsulation for sustained release.
nanoparticles: optimization of the process parameters. Nanomedicine. 2005;1(1):85–90.
Chem Eng Res Des. 2019;142:268–84. Tokunaga S, Ono K, Ito S, Sharmin T, Kato T, Irie K,
Sodeifian G, Sajadian SA, Daneshyan S. Preparation of Mishima K, Satho T, Harada T, Aida TM, Mishima
Aprepitant nanoparticles (efficient drug for coping K. Microencapsulation of drug with enteric polymer
with the effects of cancer treatment) by rapid expan- Eudragit L100 for controlled release using the particles
sion of supercritical solution with solid cosolvent from gas saturated solutions (PGSS) process. J
(RESS-SC). J Supercrit Fluids. 2018;140:72–84. Supercrit Fluids. 2021;167
Sodeifian G, Saadati Ardestani N, Sajadian SA, Soltani Torino E, Marco ID, Reverchon E. Organic nanoparticles
Panah H. Experimental measurements and thermody- recovery in supercritical antisolvent precipitation. J
namic modeling of Coumarin-7 solid solubility in Supercrit Fluids. 2010;55:300–6.
supercritical carbon dioxide: Production of Tozuka Y, Miyazaki Y, Takeuchi H. A combinational
nanoparticles via RESS method. Fluid Ph Equilibria. supercritical CO2 system for nanoparticle preparation
2019a;483:122–43. of indomethacin. Int J Pharm. 2010;386(1–2):243–
Sodeifian G, Sajadian SA, Saadati Ardestani N, 248.
Razmimanesh F. Production of Loratadine drug Turino LN, Stella B, Dosio F, Luna JA, Barresi
nanoparticles using ultrasonic-assisted Rapid expan- AA. Nanoparticles obtained by confined impinging
sion of supercritical solution into aqueous solution jet mixer: poly(lactide-co-glycolide) vs. Poly-epsilon-
(US-RESSAS). J Supercrit Fluids. 2019b;147:241–53. caprolactone. Drug Dev Ind Pharm. 2018;44(6):
Sohnel O, Garside J. Precipitation: basic principles and 934–41.
industrial applications. Newton: Butterworth- Turk M. Influence of thermodynamic behaviour and solute
Heinemann; 1992. properties on homogeneous nucleation in supercritical
Sporanox Package Insert. (Janssen Pharmaceutica solutions. J Supercrit Fluids. 2000;18(3):169–84.
Products, L.P.) Turk M. Manufacture of submicron drug particles with
Subramaniam B, Rajewski RA, Snavely enhanced dissolution behaviour by rapid expansion
K. Pharmaceutical processing with supercritical carbon processes. J Supercrit Fluids. 2009;47(3):537–45.
dioxide. J Pharm Sci. 1997;86(8):885–90. Turk M, Bolten D. Formation of submicron poorly water-
Szymusiak M, Hu X, Leon Plata PA, Ciupinski P, Wang soluble drugs by rapid expansion of supercritical solu-
ZJ, Liu Y. Bioavailability of curcumin and curcumin tion (RESS): results for naproxen. J Supercrit Fluids.
glucuronide in the central nervous system of mice after 2010;55:778–85.
oral delivery of nano-curcumin. Int J Pharm. 2016;511 Turk M, Lietzow R. Formation and stabilization of submi-
(1):415–23. cron particles via rapid expansion processes. J
Tam JM, McConville JT, Williams RO, Johnston KP III. Supercrit Fluids. 2008;45:346–55.
Amorphous cyclosporin nanodispersions for enhanced Turk M, Hils P, Helfgen B, Schaber K, Martin HJ, Wahl
pulmonary deposition and dissolution. J Pharm Sci. MA. Micronization of pharmaceutical substances by
2008;97(11):4915–33. the rapid expansion of supercritical solutions (RESS): a
Tang H-X, Cai Y-Y, Liu C-G, Zhang J-T, Kankala RK, promising method to improve bioavailability of poorly
Wang S-B, Chen A-Z. Sub-micronization of disulfiram soluble pharmaceutical agents. J Supercrit Fluids.
and disulfiram-copper complexes by Rapid expansion 2002;22(1):75–84.
of supercritical solution toward augmented anticancer Uchida H, Nishijima M, Sano K, Demoto K, Sakabe J,
effect. J CO2 Util. 2020; 39 Shimoyama Y. Production of theophylline
nanoparticles using rapid expansion of supercritical
solutions with a solid cosolvent (RESS-SC) technique. of hydroxypropyl methylcellulose solution and matrix
The Journal of Supercritical Fluids. 2015;105:128–35. tablet. J Appl Polym Sci. 2006;102:4066–74.
Vaughn JM, Gao X, Yacaman M-J, Johnston KP, Yildiz N, Tuna S, Döker O, Çalimli A. Micronization of
Williams RO. Comparison of powder produced by salicylic acid and taxol (paclitaxel) by rapid expansion
evaporative precipitation into aqueous solution of supercritical fluids (RESS). J Supercrit Fluids.
(EPAS) and spray freezing into liquid (SFL) 2007;41(3):440–51.
technologies using novel Z-contrast STEM and com- Young TJ, Johnston KP, Mishima K, Tanaka
plimentary techniques. Eur J Pharm Biopharm. H. Encapsulation of lysozyme in a biodegradable poly-
2005;60(1):81–9. mer by precipitation with a vapor-over-liquid
Vemavarapu C, Mollan MJ, Needham TE. Coprecipitation antisolvent. J Pharm Sci. 1999;88(6):640–50.
of pharmaceutical actives and their structurally related Young TJ, Mawson S, Johnston K. Rapid expansion from
additives by the RESS process. Powder Technol. supercritical to aqueous solution to produce submicron
2009;189(3):444–53. suspensions of water-insoluble drugs. Biotechnol Prog.
Wan KY, Wong KW, Chow AHL, Chow SF. Impact of 2000;16:402–7.
molecular rearrangement of amphiphilic stabilizers on Young TJ, Johnston KP, Pace GW, Mishra
physical stability of itraconazole nanoparticles AK. Phospholipid-stabilized nanoparticles of cyclo-
prepared by flash nanoprecipitation. Int J Pharm. sporine A by rapid expansion from supercritical to
2018;542(1-2):221–31. aqueous solution. AAPS PharmSciTech. 2003;5(1):
Wang B-C, Su C-S. Solid solubility measurement of 1–16.
ipriflavone in supercritical carbon dioxide and micro- Young TJ, Johnston KP, Pace GW, Mishra
particle production through the rapid expansion of AK. Phospholipid-stabilized nanoparticles of cyclo-
supercritical solutions process. J CO2 Util. 2020;37: sporin A by rapid expansion from supercritical to
285–94. aqueous solution. AAPS PharmSciTech. 2004;5(1):
Weber M, Thies M. Understanding the RESS process. In: 70–85.
Sun Y-P, editor. Supercritical fluid technology in Zhang Y, Feng J, McManus SA, Lu HD, Ristroph KD,
materials science and engineering. New York: Marcel Cho EJ, Dobrijevic EL, Chan HK, Prud'homme
Dekker; 2002. p. 387–437. RK. Design and solidification of fast-releasing
Weber A, Weiss C, Tschernjaew J, Kummel R. Gas clofazimine nanoparticles for treatment of cryptospo-
antisolvent crystallization. From fundamentals to ridiosis. Mol Pharm. 2017;14(10):3480–8.
industrial applications. Oberhausen: Fraunhofer Zhou Y, Gaucher C, Fries I, Hobekkaya MA, Martin C,
Institut Umwelt- Sicherheits- Energietechnik; 1999. Leonard C, Deschamps F, Sapin-Minet A, Parent
p. 235–8. M. Challenging development of storable particles for
Weber M, Russell LM, Debenedetti PG. Mathematical oral delivery of a physiological nitric oxide donor.
modeling of nucleation and growth of particles formed Nitric Oxide. 2020;104–105:1–10.
by the rapid expansion of a supercritical solution under Zhu Z. Flash nanoprecipitation: prediction and enhance-
subsonic conditions. J Supercrit Fluids. 2002;23(1): ment of particle stability via drug structure. Mol
65–80. Pharm. 2014;11(3):776–86.
Werling JO, Debenedetti PG. Numerical modeling of mass Zhu Z, Anacker JL, Ji S, Hoye TR, Macosko CW,
transfer in the supercritical antisolvent process. J Prud’homme RK. Formation of block copolymer-
Supercrit Fluids. 1999;16(2):167–81. protected nanoparticles via reactive impingement
Werling JO, Debenedetti PG. Numerical modeling of mass mixing. Langmuir. 2007;23(21):10499–504.
transfer in the supercritical antisolvent process: misci- Zhu Z, Margulis-Goshen K, Magdassi S, Talmon Y,
ble conditions. J Supercrit Fluids. 2000;18(1):11–24. Macosko CW. Polyelectrolyte stabilized drug
Westesen K, Siekmann B (1998) Solid lipid particles, nanoparticles via flash nanoprecipitation: a model
particles of bioactive agents and methods for the man- study With b-carotene. J Pharm Sci. 2010;99(10):
ufacture and use thereof. US 5785976, 12 April 1994 4295–306.
Wubbolts FE, Bruinsma OSL, van Rosmalen Zou Y, Huang H, Pu Y, Wang J-X, Wang D, Chen J-F.
GM. Dry-spraying of ascorbic acid or acetaminophen Solubility, solubility modeling, and antisolvent precip-
solutions with supercritical carbon dioxide. J Crystal itation of 1,3-Bis(9-carbazolyl)benzene in organic
Growth. 1999;198/199(Pt 1):767–72. solvents. J Chem Eng Data. 2019;64(10):4349–56.
Xu XM, Song YM, Ping QN, Wang Y, Liu XY. Effect of
ionic strength on the temperature-dependent behavior
Emerging Technologies to Increase
the Bioavailability of Poorly 13
Water-Soluble Drugs
Daniel A. Davis Jr., Rishi Thakkar, Mohammed Maniruzzaman,

Dave A. Miller, and Robert O. Williams III
Abstract the previous authors’ significant contribution

The poor aqueous solubility and dissolution of from the first and second edition. This current
new chemical entities (NCEs) can be the third edition chapter is a revision and update of
driving factor for eliminating pharmaco- the original authors’ work from the first and
logically superior leads. Reconsidering the second editions.
chemistry of these compounds can be econom-
ically burdensome and may result in the loss of Keywords
activity. Moreover, a considerable percent of KinetiSol® Dispersing (KSD) · 3D-Printing ·
drug substances in the market also face solu- Selective laser sintering (SLS) · Fused-
bility issues. Solving solubility issues can deposition modeling (FDM) · Electrostatic
improve the efficiency of the dosage forms, spinning · Dielectric-induced heating · Hot-
the efficacy of the molecules, and open melt extrusion (HME) · Multilayer tablets ·
possibilities for repurposing these molecules Tablet-in-tablet · Ordered mesoporous silica
for other ailments. Novel processes and (OMS)
formulation-based approaches that have over-
come traditional techniques’ limitations can
prove to be effective and economically feasi- 13.1 Introduction
ble for these high-potential leads and drug
substances. The purpose of this chapter is to With the advent of high-throughput screening
describe emerging technologies for solubility techniques in drug discovery, the number of
enhancement, allowing the reader to gain an compounds reaching the formulation develop-
understanding of their utility. The current ment stage has increased drastically (Lipinski
authors would like to thank and acknowledge et al. 2001; Lipinski 2004). Consequently, this
has given rise to a dramatic increase in the num-
Authors Daniel A. Davis Jr. and Rishi Thakkar have ber of compounds that exhibit poor water solubil-
equally contributed to this chapter. ity. These drug substances may also exhibit
D. A. Davis Jr. (*) · R. Thakkar · M. Maniruzzaman · instabilities or other properties that severely
R. O. Williams III limit standard processing and formulation
Pharmaceutics Division, College of Pharmacy, University techniques. The development of new techniques
of Texas at Austin, Austin, TX, USA to improve the bioavailability of compounds such
as these has become a major point of interest in
D. A. Miller the pharmaceutical industry. In recent years, a
DisperSol Technologies, Georgetown, TX, USA
number of novel processing and formulation solvents, which must ultimately be recovered
techniques have emerged that allow for the suc- and properly disposed of. Dispersions prepared
cessful manufacture and delivery of these by spray drying often require extended drying
compounds. steps to reduce residual solvent levels to those
Emerging technologies that utilize specific outlined in International Conference on
processes primarily offer novel methods of Harmonization guidelines. Furthermore, the
forming amorphous solid dispersions (ASD) spray-drying process results in dispersions with
with traditional materials. These technologies relatively low bulk densities requiring additional
offer alternatives to methods such as spray drying downstream processing such as roller
and hot-melt extrusion (HME). Specifically, compaction.
KinetiSol® Dispersing (KSD), 3D printing, and The HME process may not be feasible when
electrostatic spinning are emerging techniques the drug substance is thermally labile or shear
that have recently been developed for the produc- sensitive. One major disadvantage of this technol-
tion of ASDs. Emerging formulation techniques ogy is that high temperatures and prolonged resi-
include using mesoporous materials as carriers for dence times are normally required to facilitate the
adsorbed drugs and the preparation of transition from the thermodynamically stable
co-amorphous mixtures. The following sections crystalline state to the high-energy amorphous
discuss the background of each of these state. Many drug substances are known to
techniques, as well as specific examples that degrade at elevated temperatures (Murphy and
exhibit the advantages of each bioavailability Rabel 2008; Repka et al. 1999, 2003; Follonier
enhancement technique. et al. 1994), and degradation due to thermal expo-
sure has been reported for many pharmaceutically
relevant polymers (El’Darov et al. 1996; Crowley
13.2 Process-Based Techniques et al. 2002; Capone et al. 2007). Simply reducing
processing temperature to limit degradation may
Solid dispersions have received a significant not be feasible due to viscosity and glass-transi-
amount of interest in the scientific literature as a tion temperature limitations. To circumvent this
method to improve the oral bioavailability of issue, liquid or solid-state plasticizers may be
poorly water-soluble compounds (Serajuddin incorporated into dispersions as a processing aid
1999; Leuner and Dressman 2000; Breitenbach (Repka et al. 1999; Follonier et al. 1994; Zhu
2002; Crowley et al. 2007). These systems, origi- et al. 2002, 2006). While incorporation of a plas-
nally described by Sekiguchi and Obi, contain at ticizer does allow for processing at decreased
least one drug substance dispersed within an inert temperatures, it may result in a physically unsta-
carrier in the solid-state (Sekiguchi 1961; Chiou ble system due to reduced glass transition temper-
and Riegelman 1971). ASDs, in which the drug is ature, Tg, and increased molecular mobility
molecularly dispersed in a single-phase system, (Hancock et al. 1995; Hancock 2002).
may be prepared by methods that can be broadly Residence times at elevated temperatures in
classified as being solvent- or fusion-based the HME process will vary with processing
techniques. Traditional high-throughput conditions but can be expected to fall within the
manufacturing processes such as spray drying 1–2 min range and, in some cases, as long as
and HME have been used with great success for 10 min (Verreck et al. 2006; Kumar et al. 2008).
the production of many solid dispersion systems. In order to reduce the residence time in a hot-melt
However, each technique has inherent extruder, the level of mixing can often be adjusted
disadvantages that may prevent the successful by the addition or removal of mixing elements.
processing of some compounds. While reducing the number of mixing elements
While the spray-drying process has the advan- will, in most cases, decrease residence time, spe-
tage of utilizing traditional manufacturing equip- cific shear input is reduced and drug substances
ment, it relies on the use of potentially toxic may not fully transition to the amorphous state.
13 Emerging Technologies to Increase the Bioavailability of Poorly Water-Soluble Drugs 601
13.2.1 KinetiSol® Dispersing KSD compounders can be operated in batch

mode or as a semi-continuous process. The batch-
KinetiSol Processing and Equipment mode KSD compounder (Model: KC254B) is
Kinetisol® (KSD) is a fusion-based processing shown in Fig. 13.2, and the semi-continuous com-
technique for the rapid production of polymeric pounder (Model: KC254C) is shown in Fig. 13.3;
amorphous solid dispersions. In this high-energy the semi-continuous unit enables automated feed-
mixing process, a shaft with protruding blades ing and quenching operations every 0.5–2 min
rotates at speeds of up to 3500 rpm within a with a product throughput of 20–30 kg/h. The
sealed processing chamber containing a drug– proof of concept work with KSD for pharmaceu-
polymer blend. A cross-sectional view of the tical ASD systems was first evaluated on
processing chamber is illustrated in Fig. 13.1. industrial-scale compounders with batch sizes
Through a combination of kinetic and thermal from 1 to 10 kg (Miller 2007); however, it has
energy, compositions are processed into a molten since been scaled down to accommodate pharma-
mass without the need for external heat input ceutical development with further reduced mate-
(DiNunzio et al. 2010c). Along with the molten rial needs.
transition, KSD rapidly and thoroughly mixes the
active pharmaceutical ingredient (API) with the KinetiSol’s Advantage Over Hot-Melt
excipient(s) on a molecular level to achieve Extrusion
single-phase, homogenous ASD systems. A The following case studies highlight the benefits
computer-control module monitors the that arise from the robust KSD processing
composition’s real-time temperature, and upon capabilities. Examples are then compared to
reaching the ejection temperature set point, the HME to give perspective to the relative
molten material is ejected. Total processing times processing advantage. Highlights are KinetiSol’s
are typically less than 20 s, and elevated ability to process high molecular weight viscous
temperatures are generally maintained for less polymers without the need for surfactants, which
than 5 s, thus significantly reducing the time the compromise ASD stability, and the ability to pro-
drug–polymer system is exposed to elevated cess challenging APIs (e.g., high-melting point,
temperatures. thermally labile APIs).
DiNunzio et al. conducted early evaluations of
the KSD technology to prepare solid dispersions
of itraconazole (ITZ) (DiNunzio et al. 2010b, c, d;
DiNunzio 2008). Compositions of ITZ with
hypromellose (hydroxypropyl methylcellulose,
HPMC) and Eudragit® L100-55 prepared using
KSD showed comparable or more rapid drug
release in vitro those prepared by HME; this
finding did not show a significant difference
in vivo using a rat model (DiNunzio 2008;
DiNunzio et al. 2010c). However, KSD was
shown to prepare more homogenous dispersions,
as seen by a single Tg value with thermal analy-
sis, and more physically stable, as they did not
require a plasticizer for processing, as is often the
case with HME (DiNunzio et al. 2010b, c). The
authors showed that 1:2 compositions of ITZ:
Eudragit® L100-55, an enteric polymer with con-
Fig. 13.1 Cross-sectional view of the KinetiSol® Dis-
persing processing chamber. (Reproduced with permission centration enhancing properties, could not be
from DiNunzio 2008) effectively processed by HME due to the high
Fig. 13.2 Kinetisol®

KC254B Compounder for
batch-mode operation.
(Courtesy of Dispersol
Technologies)
Fig. 13.3 Kinetisol®

KC254C Compounder for
semi-continuous operation.
(Courtesy of Dispersol
Technologies)
viscosity and decomposition of the polymer incorporation of TEC. Milled samples were
(DiNunzio et al. 2010b). Triethyl citrate (TEC) placed at accelerated stability conditions (40 C/
was incorporated into HME compositions at 20% 75% RH) for 6 months to evaluate the physical
(by dry polymer weight) as a processing aid. stability of compositions prepared by each
Conversely, compositions containing a 1:2 ratio processing method. As shown in Fig. 13.4, crys-
of ITZ: Eudragit® L100-55 were successfully talline peaks identified as those characteristic of
processed by KSD, primarily due to the high ITZ were found in HME compositions at the 3-
torque output inherent to this technology. All and 6-month time points, indicating physical
samples were determined to be amorphous by instability due to increased molecular mobility.
powder X-ray diffraction (PXRD). Furthermore, Compositions prepared by KSD exhibited physi-
modulated DSC analysis showed that the HME cal stability over the same 6-month testing period.
and KSD compositions exhibited glass transition Since these studies, the application of KSD has
temperatures of 54.2 C and 101.3 C, respec- evolved to increase the formulation design space
tively; the lower Tg of the HME due to the for amorphous solid dispersions, including the
Fig. 13.4 PXRD patterns

of HME and KSD ITZ:
Eudragit L100-55 solid
dispersions measured over
6 months’ accelerated
stability at 40 C/75%
RH. Samples were stored in
30 cc HDPE induction
sealed bottles. (Reproduced
DiNunzio et al. 2010b)
ability to process high melting point and ther- which allows processing at temperatures below
mally sensitive materials and high viscosity poly- the melting point of the drug. Because of this, the
mer carriers, which are particularly difficult to production of ASD systems with high melting
process using HME and spray-drying processes. point materials (Tm > 200 C) is possible with
KSD processing. Hughey et al. demonstrated the
ability to process high melting point drugs with
High Melting Point APIs
the KSD process, using meloxicam (MLX) with a
With HME processing, temperatures must be suf-
melting point of 270 C as a model compound
ficiently high to produce molten material to
(Hughey et al. 2011). This study utilized
enable proper mixing and dispersion and/or solu-
Soluplus® as the polymer carrier. KSD, at all
bilization of the drug into the polymer carrier; this
rotor speeds and ejection temperatures studied,
is usually 15–30 C above the melting point of the
rendered all samples amorphous; however, higher
drug substance and/or polymer. Thus, with high
ejection temperatures yielded more MLX degra-
melting point drugs, high processing
dation. With 3000 rpm rotor speed and an ejection
temperatures would be required, which could be
temperature of 110 C, 98.6% MLX recovery was
detrimental to the chemical stability of the drug
obtained from the final dispersion. Extrusion was
and/or polymer (Lakshman et al. 2008;
carried out at temperatures up to 175 C; how-
LaFountaine et al. 2015b). Alternatively, a sub-
ever, dispersions still exhibited a low degree of
stantial amount of plasticizer would be required to
residual crystallinity when the molten material
reduce processing temperatures, which could
was recirculated for 2 min. Additionally, the lon-
reduce the solid dispersion’s physical stability
ger residence time led to degradation, with only
(Verreck 2012). The high shear rates imparted
87.1% MLX recovery. More recently, KSD was
by KSD processing accelerates the solubilization
utilized to prepare amorphous solid dispersions of
kinetics of a compound in a molten polymer,
acetyl-11-keto-b-boswellic acid (AKBA) with a ITZ supersaturation was seen with increasing vis-
melting point of 295 C, a compound unable to be cosity, albeit up to a point (Hughey et al. 2012).
extruded due to its high melting point and thermal HPMC E50LV was evaluated as one of the
liability (Bennett et al. 2013). KSD rendered promising viscous polymers for ITZ supersatura-
AKBA amorphous with four different polymeric tion; however, at HME processing temperatures
carriers: HPMCAS-LF, HPMCAS-MF, above its Tg, HPMC has been shown to brown/
Eudragit® L100-55, and Soluplus® in combina- darken and chemically degrade (LaFountaine
tion with Eudragit® L100-55 with a potency et al. 2015b; Coppens et al. 2004). Hughey et al.
greater than 99%. evaluated the effect of processing conditions,
shown in Fig. 13.5, and found that the polymer
High Molecular Weight, Viscous Polymers degradation was related to both shear and thermal
In addition to processing high melting point drug energy input, as well as the interaction between
substances, the short processing times and lower the two suggesting the polymer chains may be
ejection temperatures with KSD were shown to more susceptible to shearing forces at higher
allow processing of thermally labile drugs temperatures. KSD was able to render the ITZ:
(DiNunzio et al. 2010a; Hughey et al. 2010). HPMC E50LV compositions amorphous, includ-
The short residence times of the KSD process ing with increasing drug load of up to 50%w/w,
can also benefit the polymer carrier’s thermal with limited polymer degradation and no notice-
stability; additionally, the high torque outputs able browning at processing speeds of 2250 rpm
allow for the ability to process high viscosity, and ejection temperature of 120 C, significantly
high molecular weight carriers – both of these lower than the melting point of both the drug and
increasing the options for polymeric carriers of polymer.
amorphous solid dispersion systems. Hughey LaFountaine et al. systematically investigated
et al. showed that within a given chemistry of the processability of increasing polymer molecu-
cellulosic polymers, the higher the degree of lar weight and thus viscosity using HPMC (E5,
Fig. 13.5 KSD temperature profiles for 1:2 ITZ:HMPC E50LV compositions. (Reproduced with permission from
Hughey et al. 2012)
E15, and E50) and polyvinylpyrrolidone (PVP) release, as the highly hydrolyzed grades are
(K17, K30, and K90) using HME and KSD with more water-soluble.
griseofulvin (GRIS) as a model drug
(LaFountaine et al. 2015a). As seen with previous Short Chain, Low Molecular Weight
studies, HPMC E15, HPMC E50, PVP K30, and Oligomers
K90 were not amenable to HME processing with- As seen with high-molecular weight, viscous
out a plasticizer (Fousteris et al. 2013; polymers, KinetiSol processing relies on the gen-
Ghebremeskel et al. 2006). Additionally, the eration of frictional and shear stresses to promote
HME dispersions prepared with HPMC E5 and thermal fusion of the drug–polymer system.
PVP K17 were phase-separated at higher drug Long-chain, high molecular weight polymers cre-
loads, as shown in Fig. 13.6. However, KSD ate greater frictional resistance than short-chain,
was able to prepare amorphous dispersions for low molecular weight oligomers. Additionally, in
all polymer grades at drug loads of 10%, 20%, a KinetiSol system, the shear stress is directly
and 40%, except for a small degree of residual proportional to the fluid’s consistency
crystallinity seen in the PVP K30 40% drug load (Gopakumar and Page 2005); therefore, the con-
formulation. Although, GRIS, a proton acceptor, sistency of the oligomer melt being less than the
is expected to interact with PVP, a proton accep- polymer melt, in addition to the lower molecular
tor, no chemical interaction was seen via FTIR or weight, makes the processing of oligomers (e.g.,
FT-Raman analysis. After 6 months at open dish Hydroxypropyl B cyclodextrin) not apparently
accelerated stability conditions (40 C/75%RH), obvious.
all formulations showed no physical change via Gala et al. demonstrated, for the first time, the
PXRD. With no chemical interactions, the stabi- feasibility and benefit of processing oligomers,
lization of these systems was attributed solely to specifically hydroxypropyl B cyclodextrin
decreased molecular mobility. (HPBCD), during the KinetiSol process. In this
The ability of KSD to process materials not study, a 10% w/w abiraterone and 90% w/w
previously amenable to thermal processing or HPBCD formulation was ejected at 160 C. The
spray drying due to melting point and viscosity final product was less agglomerated and more
limitations has opened the door to evaluating brittle than a traditional long-chain linear poly-
polymers that were thus far not considered for mer; nonetheless, the formulation was rendered
ASDs. Polyvinyl alcohol (PVA) is a hydrophilic, entirely amorphous without significant degrada-
non-toxic, semi-crystalline polymer often utilized tion. When evaluated using a non-sink, pH-shift
as a stabilizing agent in emulsions, a viscosity dissolution method, the HPBCD ASD released
increasing agent or lubricant in ophthalmic and abiraterone faster and achieved higher levels of
topical solutions, as well as in composite biode- supersaturation in the acidic phase than binary
gradable materials (Rowe et al. 2012). The abiraterone ASD formulations using traditional
polymer’s high viscosity and melting point limits polymers (i.e., HPMC E3, HPMC E5, PVP K30,
its use in HME as it requires up to 40% plasticizer and PVAP). However, upon transition to the neu-
to reduce processing temperatures, particularly tral media, FaSSIF, the HPBCD failed to stabilize
with PVA grades with a high degree of hydrolysis the supersaturated drug, and rapid crystallization
(De Jaeghere et al. 2015). Brough et al. evaluated occurred. Therefore, a ternary component
PVA grades with varying degrees of hydrolysis HPMCAS 126G, a pH-dependent polymer, was
with ITZ, showing that KSD processing could selected to stabilize the supersaturated abiraterone
render the ITZ amorphous, while the PVA when transitioning from the acidic phase to the
retained its crystalline structure (Brough et al. neutral phase. It was determined, the addition of
2015), as shown in Fig. 13.7. The PVA 4-88 10% w/w HPMCAS 126G prevented recrystalli-
and 4-75 grades, indicative of 88% and 75% zation in the neutral phase when compared to the
hydrolysis, respectively (Note: 98% is considered binary HPBCD ASD. The ternary HPBCD ASD
fully hydrolyzed), showed rapid and full drug and the Binary HPBCD ASD showed a 13.8- and
Fig. 13.6 PXRD profiles (from bottom to top) of GRIS K90 ( ) (from 10% to 40% drug load) in (b) and
crystalline drug ( ), 10% drug load physical mixtures KSD processed dispersions of HPMC E15 ( ) and
( ), HME processed dispersions ( ), and KSD HMPC E50 ( ) (from 10% to 40% drug load) in (d).
processed dispersions ( ) (from 10% to 40% drug (Reproduced with permission from LaFountaine et al.
load) in (a) (PVP K17) and (c) (HPMC E5). KSD 2015a)
processed dispersions of PVP K30 ( ) and PVP
12.4-fold increase in bioavailability in beagle the use of plasticizers. As described by Jermain

dogs when compared to the generic abiraterone et al. processing ASDs with HPMCAS has the
acetate tablets, respectively (Gala et al. 2020). following advantages: (i) a high glass transition
The high-complexing ability of cyclodextrins, temperature in the solid-state, improving physical
along with the rapid release and pronounced stability during shelf-life; (ii) the succinyl and
supersaturation effects, offers a unique alternative acetyl groups present have strong interactions
to traditional high molecular weight polymers. with weakly basic drugs, which promotes
Further investigation of the long-term stability in-solution stabilization; (iii) in the ionized form,
and applicability to other APIs is warranted. polymer aggregation is minimized, providing sta-
bility of amorphous drug–polymer colloids; and
(iv) the amphiphilic nature allows for improved
Anionic Polymers and Weakly Basic APIs
drug–polymer interactions and maintains stable
The dynamic processing ability of KinetiSol
colloids in aqueous solutions. Additionally, a
processing enables the successful formulation of
major concern when weakly basic APIs are
anionic polymers, specifically HPMCAS grades,
formulated into ASDs is overcoming
which are challenging to extrude in HME without
ITZ:PVAL 4-98 (1:4)
ITZ:PVAL 4-88: PVAL 40-88 (1:2:2)
ITZ:PVAL 4-88 (1:4)
ITZ:PVAL 4-75 (1:4)
ITZ:PVAL 4-38 (1:4)
Bulk Itraconazole
5 10 15 20 25 30 35 40 45 50
Angle (2-theta)
Fig. 13.7 PXRD profiles of (from bottom to top) bulk ITZ, ITZ: PVA 4-38 (1:4), ITZ: PVA 4-75 (1:4), ITZ: PVA 4-88
(1:4), ITZ: PVA 4-88: 40-88 (1:2:2) and ITZ 4-98 (1:4). (Reproduced with permission from Brough et al. 2015)
crystallization when transitioning from the stom- formulation is exposed to elevated temperatures
ach, an ionized high-solubility state, to the small (e.g., above 40 C) for 5.2 s, achieving an amor-
intestine, a unionized low-solubility state. Two phous composition with an impurity profile simi-
case studies are presented, each highlighting, in lar to that of the commercial product. The
a different way, the benefit of using a weakly pH-shift dissolution study used equivalent
basic API and an anionic polymer during the amounts of ritonavir, 100 mg, to evaluate the
KinetiSol process. release of the 30% w/w formulation to that of
In the first study, Ellenberger et al. show the the commercial product. In the acidic phase, the
benefit of processing a weakly basic API, ritona- crushed Norvir tablet released ritonavir faster and
vir, with the anionic polymer, HPMCAS, instead achieved higher concentrations than the KSD for-
of the commercial polymer, copovidone. In the mulation utilizing HPMCAS. However, upon
marketed product Novir®, the weakly basic API, pH-transition to the neutral phase, the KSD for-
ritonavir, is present at a 15% drug load (w/w) and mulation, composed of the pH-dependent poly-
is processed into an ASD by HME. The extruded meric carrier, rapidly released ritonavir; whereas,
ASD (i.e., ritonavir, sorbitan monolaurate, and the commercial product rapidly recrystallized.
Copovidone) represents 667 mg of the 789 mg When equivalent ritonavir doses were tested in
tablet, making the final tablet large for patient beagle dogs, the area under the curve (AUC)
consumption. To minimize pill burden for value for the KSD formulation was within 2%
patients, the ritonavir drug load was increased to of the Norvir reference tablet and achieved
30% w/w and processed with the anionic polymer lower Cmax values, Fig. 13.8 (Ellenberger et al.
HPMCAS (i.e., Affinisol 912 granular grade) by 2018).
the KinetiSol process. During processing, the
Fig. 13.8 Bioavailability comparison of ritonavir plasma concentration of the marketed product, Norvir, to increased
drug load KinetiSol formulation. (Reproduced with permission form Ellenberger et al. 2018)
The authors demonstrated the ability to use there are no observable differences in molecular
KinetiSol processing to reformulate a poorly properties. However, upon pH-shift dissolution,
water-soluble drug, with similar levels achieved the SDD achieves rapid supersaturation in the
in vivo despite using an enteric polymer that acidic phase followed by immediate recrystalliza-
modified the drug release compared to the com- tion in the neutral phase, compared to the KSD,
mercial product. In future studies, a pharmacoki- which slowly releases BI-667 in the acidic phase
netic study with sufficient power can be done to followed by a rapid release in the neutral phase.
determine the bioequivalence of the KSD formu- Considering the samples are highly similar at a
lation to the Norvir product. molecular level, differences in particle morphol-
In the second study, Jermain et al. make a ogy, size, and surface area are responsible for the
head-to-head comparison showing the improved significant differences during in vitro testing. The
bioavailability of an ASD containing a weakly particles generated by spray-drying contain
basic drug and an anionic polymer (i.e., BI-667 crevices and folds that increase the specific sur-
and HPMCAS, respectively) formulated by face area (e.g., 3.37 m2/g) and have a D50 value of
KinetiSol processing compared to an ASD of 6.1 microns. The particles generated by KinetiSol
identical composition formulated by spray drying processing are smooth and dense, decreasing the
(Jermain et al. 2020). By 1D solid-state nuclear specific surface area (e.g., 0.21 m2/g). They can
magnetic resonance (ssNMR), the KinetiSol ASD be milled to various sizes with D50 values ranging
(i.e., KSD) and the spray-dried ASD (i.e., SDD), from 102 to 578 microns for this study. A phar-
both formed miscible systems with similar macokinetic study in male Beagle dogs deter-
domain sizes, along with the 2D 13C-1H mined there was no significant difference
HETCOR spectra being highly identical, suggests between the SDD formulation and the physical
mixture. In contrast, the KSD formulation steady-state, which is controlled by the selected
demonstrated a 3.07-fold bioavailability enhance- RPM, allows for prolonged mixing (i.e., times
ment compared to the physical mixture. The dras- >20 s) within a narrow temperature range. The
tic difference in bioavailability is attributed to the composition will remain in this steady-state until
SDD particles precipitating almost immediately the temperature gradient between the surround-
after oral administration. For the SDD, the ings (e.g., processing chamber) and composition
smaller particle size and increased specific sur- diminishes, decreasing the energy’s driving force
face area ultimately expose excessive API on the out of the composition and ultimately causes a
surface, causing the rapid release of BI-667 in further increase in energy.
acid despite the weakly basic drug being spray Davis et al. incorporated 3% of a highly con-
dried with an enteric polymer. The dense KSD ductive material, candurin®, into a physical mix-
granules have less exposure of unprotected API ture of drug and polymer (BI-667 and HPMCAS-
on the surface, allowing the polymer’s enteric HMP, respectively) to get a relative xx% increase
nature to minimize drug release in the acidic in thermal conductivity. The authors achieved an
phase, while maximizing bioavailability by amorphous composition at lower processing
releasing the API in the small intestine (Jermain temperatures, attributed to the increased mixing
et al. 2020). This study highlights that though two time achieved by the prolonged mixing of the
processes might make similar ASDs at a molecu- composition with increased thermal conductivity,
lar level, the particle morphology, size, and spe- Fig. 13.9.
cific surface area are equally important and vary Processing at a lower temperature for a slightly
by the formulation process; in the case of the longer period potentially offers an alternative
weakly basic drug, BI-667, and the enteric poly- processing profile to overcome degradation
mer, HPMCAS, the increased surface area and associated with processing at higher temperatures
small particle size proved detrimental to the when trying to achieve an amorphous composi-
spray dried process. tion (Davis et al. 2020).
Prolong Mixing with Thermal Conductivity High Drug Loading Reduces Pill Burden
Modification With the ability to prepare higher drug loads,
A recently reported unique application for KSD offers a significant advantage to reduce pill
KinetiSol processing is its ability to achieve a size and pill burden. Hughey et al. investigated a
state of prolonged mixing within a narrow tem- 1:3 carbamazepine (CBZ): Soluplus® formulation
perature range (Davis et al. 2020). The authors at a 70% dispersion loading in tablet and found
suggest that increasing the composition’s thermal that the drug release was slowed after compres-
conductivity enables the formulation to more effi- sion, with less than 80% release in 2 h compared
ciently transfer energy in and out of the system. to the 100% drug release in less than 30 min from
Traditionally, as described in the above sections, the dispersion alone in 0.1 N HCl dissolution
at a critical RPM that is unique to each composi- media (Hughey et al. 2013). The addition of
tion, a rapid increase in temperature occurs until potassium bicarbonate at levels as low as 1% to
the composition reaches the ejection temperature. the tablet formulation allowed for rapid disinte-
The formulation’s inability to dissipate the energy gration to obtain a release profile similar to that of
transferred from the blades resultantly causes an the dispersion alone. In another study, Hughey
increase in the formulation’s temperature. et al. incorporated a low-substituted grade of
Improving the formulation’s conductivity enables hydroxypropyl cellulose (HPC), a compressible
the formulation to better dissipate the energy super-disintegrant, in the dispersion composition
input by the blades to the surrounding chamber. for nifedipine (NIF): HPMCAS to improve disin-
In the case where the energy input is not more tegration properties after tablet compression,
than the composition’s ability to dissipate the finding that the intragranular disintegrant enabled
energy, a dynamic steady-state is formed. This higher drug release in the acid phase due to
Fig. 13.9 The KSD–TCE profile represents the prolonged mixing achieved when the thermally conductive excipient is
present, compared to a traditional KSD processing profile
improved wetting (Hughey et al. 2014). Alterna- scenario, the dose delivered is even more critical,
tively, Keen et al. evaluated the use of KSD to as toxic doses are significantly lower. Further-
prepare monolithic tablet’s direct shaping/mold- more, homogeneity of solid dispersions is vital
ing from the ejected KSD molten mass (Keen for the physical stability of the ASD, as drug-rich
et al. 2015). To prepare a controlled release tablet or polymer-rich regions compromise the long-
formulation of ITZ, high molecular weight PVP term stability of the ASD.
K30 was utilized as the polymer carrier and glyc- Jermain et al. investigated KinetiSol
eryl behenate (Compritol 888 ATO, C888), a processing’s ability to achieve adequate
hydrophobic lipid, as the release modifier and processing of drug–polymer systems at low drug
plasticizer. When compared to ITZ: PVP K30 loadings (e.g., 1% w/w). Despite the very short
formulations with and without C888 prepared by processing time of KSD processing (e.g., less
HME, followed by milling and tablet compres- than 40 s), drug–polymer systems as low as 1%
sion, the monolithic tablet exhibited a zero-order w/w were shown to produce homogenous drug–
release with a slowed release from the C888 polymer mixtures, as evaluated using content uni-
containing formulations, while the compressed formity, blend homogeneity, and scanning elec-
tablets showed a plateau in drug release as tron microscopy/energy-dispersive X-ray
shown in Fig. 13.10, attributed to water ingress spectroscopy (SEM/EDS). The ability to achieve
into tablet pores that enabled local precipitation. a homogenous ASD at 1% w/w drug loading in
less than 40 s is attributed to the high degree of
distributive and dispersive mixing present during
Homogenous Production of ASDs
the KinetiSol process. The authors demonstrate
with Low Drug Loading for Highly
the utility of KinetiSol processing for highly
Potent APIs
potent API and the ability to make homogenous
The pharmaceutical industry is continually push-
ASDs during the short processing window
ing the boundaries to maximize drug loading
(Jermain et al. 2019).
within ASDs, while the inverse, low-drug loading
in ASDs, is rarely mentioned. Formulating ASDs
with a low-drug loading becomes particularly
important when a highly potent drug with a nar-
row therapeutic index and poor water solubility is
introduced into the development pipeline. In this
dosage form manufacturing platforms and opened

up the possibilities of 3D printing personalized
medicine (Zhang et al. 2018; Trivedi et al. 2018).
Current research has demonstrated 3D printing of
dosage forms with a wide range of active phar-
maceutical ingredients (API) with different
compositions. Moreover, researchers have
investigated the impact of the structure and the
3D printing parameters on the 3D printed dosage
forms’ performance (Thakkar et al. 2020; Zhang
et al. 2020; Ilyés et al. 2019). It has been observed
that simple changes in the design or printing
parameters can tune the release of the API from
the dosage form and, if needed, can also trans-
form the solid-state of the API to improve its
solubility and dissolution performance (Alhijjaj
et al. 2019; Smith et al. 2018; Arafat et al. 2018)
The application of 3D printing in pharmaceutical
manufacturing has been made possible by the
frequent use of polymers as pharmaceutical
excipients. This is because most polymers
respond to the different mechanisms and stimuli
exploited by the 3D printing platforms outlined in
Fig. 13.11. Current 3D printing platforms can be
broadly divided into two major categories (i.e.,
material (solid or liquid) bed based, and material
deposition based) (Azad et al. 2020; Davis et al.
2020; Zhang et al. 2021a, b). Material bed–based
Fig. 13.10 ITZ dissolution rates measured by United 3D printing employs either a powder bed (powder
States Pharmacopeia (USP) Method II at 50 rpm at 37 C bed fusion, binder jetting) or a liquid photopoly-
in 0.1 N HCl: (a) ITZ and PVP solid dispersions at 0.2 mg/ mer in a vat (vat photopolymerization), where the
mL at ITZ concentration; (b) ITZ, PVP, and C888 solid
material is selectively combined (sintered, fused,
dispersions at 0.2 mg/mL ITZ concentration. (Reproduced
with permission from Keen et al. 2015) joined, or cured) by means of heat energy, laser
source, or simply a liquid bonding agent (Davis
et al. 2020; Healy et al. 2019). The FDA approved
13.2.2 Additive Manufacturing 3D printed tablet, Spritam® takes advantage of
the powder bed based binder jetting technology,
Additive manufacturing can be defined as the where the tablet’s inherently porous structure
process of joining materials to build objects, usu- improves the disintegration and thereby absorp-
ally layer upon layer, by using a three- tion of the levetiracetam, which has a
dimensional (3D) model file (STL, OBJ, FBX, disintegration-limited absorption (Jacob et al.
COLLADA, 3DS, IGES, STEP, and VRML/ 2016). The second category involves material
X3D) (Gioumouxouzis et al. 2019). 3D printing deposition, where the material is dispensed (jetted
boasts capabilities for dispensing low volumes or extruded) by means of pressure and/or temper-
with accuracy, spatial control, and layer-by-layer ature in a systematic ‘x, y, z’ fashion (Aho et al.
assembly to prepare complex compositions and 2019). The two main underlying process
geometries. The flexibility of designing complex categories involve material extrusion and material
3D parts has further expanded the versatility of jetting technologies. Although material jetting
Fig. 13.11 AM processes for compositions with polymeric components
has not been used frequently for pharmaceutical with thermoplastic filament(s), Fig. 13.12. These
manufacturing, material extrusion is one of the filaments are introduced to the extrusion head via
most widely investigated additive manufacturing driving motor rollers; the filament is converted to
(AM) techniques for developing customized dos- its molten form by heating elements/liquefier. The
age forms with modified performance profiles and molten state allows for the extrusion through the
improved solubilities of different APIs. Based on nozzle tip. The nozzle tip deposits the extruded
the principle of AM, fusion-(heat or laser)based polymer on a heated platform in a systematic
processes can induce a solubility advantage in the fashion as per the design of the 3D model and
3D printed dosage forms. Two such techniques the slicing parameters. The heated nozzle’s tem-
have found application in manufacturing dosage perature is set above the glass transition tempera-
forms with improved dissolution behavior (i.e., ture of the feed material to facilitate extrusion.
material extrusion-based fused filament fabrica- Based on the type of the polymer and the thickness
tion (FFF)/fused deposition modeling (FDM) 3D of the filament, a higher extrusion temperature
printing, and powder bed fusion-based selective may be required to facilitate the homogeneous
laser sintering 3D printing) (Alhijjaj et al. 2016; melting of the feedstock. The temperature of the
Melocchi et al. 2016; Govender et al. 2020; Davis platform is set below the glass transition tempera-
et al. 2020). Their underlying principles and rele- ture of the polymer. The platform is heated to
vant case studies have been discussed in-depth in prevent the breaking of filaments due to immedi-
this chapter. ate cooling and facilitate the polymer’s adherence
to the build platform. A high platform temperature
can lead to the deformation of the part from the
Fused Deposition Modeling
base of the prototype. The extrusion temperature
Fused deposition modeling (FDM), also referred
is responsible for the amorphous conversion of the
to as fused filament modeling (FFM) or fused
API post FDM processing. Some polymers, such
filament fabrication (FFF), is a material
as polylactic acid (PLA), PVA, and acrylonitrile
extrusion-based additive manufacturing platform.
butadiene styrene (ABS), are commonly used with
The conventional FDM platform includes a spool
Fig. 13.12 Schematic of fused deposition modeling (FDM) printer: (1) filament spool (2), feedstock (filaments),
(3) z-axis movement screw, (4) extrusion head, (5) extrusion nozzle, (6) part/prototype on the heated build platform,
(7) control settings, (8) driving motor, (9) heating chamber, and (10) extrusion nozzle with molten feedstock
the FDM process and are commercially available and residual crystallinity, most probably due to
as pre-processed filaments. The drug is loaded in the drying rate, incubation may not be optimal for
these preprocessed filaments using solvent- the preparation of amorphous solid dispersion
mediated surface absorption of the API on the filaments for use with FDM.
filaments. Moreover, filaments for pharmaceutical Goyanes et al. prepared drug-loaded PVA
application of FDM have been manufactured filaments with each acetaminophen (APAP) and
using hot-melt extrusion (HME) technology. caffeine (CAF) using HME to prepare filaments
Thermoplastic polymers are combined with API for use with FDM (Goyanes et al. 2015). Fixed-
and plasticizers along with other excipients using dose combination (FDC) tablets were prepared
HME to attain sufficient tensile strength and flexi- with multilayered and tablet-in-tablet geometries
bility (elongation) of the filaments to prevent them containing both drugs, as shown in Fig. 13.13.
from breaking during the AM process. Another APAP was rendered amorphous during the HME
key parameter for the filaments is the diameter to process, but the CAF remained dispersed in its
ensure uniform heat transfer throughout the crystalline state. The multilayered tablets
printing. exhibited synchronous release of both drugs,
Skowyra et al. incubated commercially avail- while the tablet-in-tablet geometry showed
able PVA filaments in a methanol solution of delayed release of up to 135 min of the inner
prednisolone (PDN) for 24 h (Skowyra et al. component when the outer layer was APAP and
2015). The methanol was chosen to both solubi- up to 90 min delayed-release when the outer layer
lize the poorly water-soluble PDN and reduce was CAF, shown in Fig. 13.14. A longer delayed-
swelling of the PVA filament. After drying, a release was observed when the drug loading was
drug load of approximately 1.9% was achieved, lower, as the PVA levels are higher and control
with the adsorbed PDN mostly in the amorphous the erosion-based drug release. This study
state from the evaporation of the methanol. The showed the ability to make complex geometries
drug-loaded filament was then used with an FDM to control drug release using HME to prepare
printer to prepare tablets of varying size and, thus, ASD as feed filament.
varying doses, 2–10 mg. With a low drug loading
Fig. 13.13 Schematic of (a) multilayer and (b) tablet-in-tablet geometry for acetaminophen and caffeine FDC tablets.
(Reproduced with permission from Goyanes et al. 2015)
Powder Bed Fusion capacity of the polymer with respect to the drug
Powder bed fusion employs thermal energy, species, and secondly, the low degree of mixing
which selectively fuses regions of the powder observed during the SLS 3D printing process.
bed. Selective laser sintering, in particular, uses The first aspect is similar to other methods used
a laser source emitting laser of a particular for ASD manufacturing and are heavily based on
wavenumber and power, a schematic is shown the formulation composition and selection of
in Fig. 13.15. The build surface and build cham- excipients. The second aspect (i.e., the degree of
ber are maintained at threshold temperatures to mixing) depends heavily on the process. The
initiate sintering of the selected powder bed observed degree of mixing is tunable and higher
region while preventing phase-transformation in processes such as KinetSol® and HME, and
before laser exposure. The laser then sinters hence they are more suitable for formulations
regions as per the computer-aided design (CAD) with a higher drug load. For SLS 3D printing,
file and parameters set (Shen et al. 2018). The the thermally conductive laser absorbing species
build surface is set below the composition’s glass absorbs the heat generated when stimulated by
transition temperature, whereas the chamber tem- the laser and dissipates this heat to the
perature is set the same as the build surface or surrounding particles, predominantly composed
lower. Higher chamber temperatures can lead to of polymers. The polymer melts and dissolves
premature melting, fusion, or agglomeration, hin- the surrounding API molecule while bridging
dering the smooth flow of the feedstock and with the neighboring polymeric particle when
eventually leading to print failure (Davis et al. cooled instantly, as depicted in Fig. 13.16. The
2020). stimuli applied by the laser is very short-lived, but
Current studies have shown that formulations sufficient for these series of events to occur
with a laser absorbing species and a thermoplastic (Davis et al. 2020; Thakkar et al. 2021).
polymer with low drug loads (i.e., less than 20% The first-ever case study of 3D printing ASD
w/w) can successfully form amorphous solid using the PBF AM platform was demonstrated by
dispersions with a high degree of mixing of the Davis et al. For this study, a poorly water-soluble
drug with the polymer. Increasing the drug load antiretroviral API ritonavir was selected as a
can lead to the amorphous conversion of the drug, model drug. The drug was blended with a ther-
but might not lead to the formation of solid moplastic polymer (vinylpyrrolidone-vinyl ace-
dispersions with the polymer and might require tate copolymers) and a laser absorbing,
further formulation considerations. There are two thermally conductive excipient (potassium alumi-
key factors responsible for these observations; num silicate-based pearlescent pigment). This
firstly, the solubilization and stabilization physical blend was used as a feedstock for
Fig. 13.14 Drug release

from multilayered tablets
(top) and tablet-in-tablet
(DuoCaplet) with APAP
outer layer (middle) and
CAF outer layer (bottom)
prepared with varying drug
load. (Reproduced with
permission from Goyanes
et al. 2015)
Fig. 13.15 A schematic of processing amorphous solid dispersions using a Powder Bed Fusion-(PBF)based AM
technology
Fig. 13.16 Schematic outlining the sintering process leading to ASD formation
manufacturing tablets in one step. It was observed 13.2.3 Electrostatic Spinning

post-processing that the drug was converted to its
amorphous form using solid-state analysis. Upon Electrostatic spinning is a technique that has been
advanced characterization, it was seen that the used successfully in the polymer industry for
drug existed as an amorphous solid dispersion many years to produce a variety of products,
when compared with hot-melt extruded such as separation membranes (Doshi and
references. The manufactured tablets exhibited a Reneker 1993; Reneker and Chun 1996). As
21-fold solubility advantage compared to the applied in the polymer industry, a polymer in
physical mixture (feedstock), where the drug solution is drawn through a capillary tube that is
was present in its crystalline form, Fig. 13.17. subjected to an electric field. As the electric field
This study expanded the application of intensity is increased, the solution forms a Taylor
PBF-based AM platforms in pharmaceutical cone at the tip of the capillary tube. Once the
manufacturing (Davis et al. 2020). electric field overcomes the force of surface ten-
sion, the solution is ejected in the form of an
Fig. 13.17 Data comparing amorphous solid dispersions samples. (c) PXRD overlay of candurin, ritonavir (RTV),
manufactured using powder bed fusion (F1-P4-10; Kollidon Va64, physical mixture, and SLS/HME based
F3-P7S-20) with HME (a) 13C–1H HETCOR spectra of formulations. (d) in vitro drug release testing for SLS–
HME and SLS samples. (b) 1D13C CP-MAS spectra of ASD with 10% and 20% drug loads
crystalline RTV, physical mixtures, HME, and SLS
electrically charged jet. As the solvent evaporates, In one of the earlier studies, Verreck et al.
narrow unwoven filaments are formed with investigated the use of electrostatic spinning to
diameters ranging from 50 nm to 5 μm (Doshi prepare hydrophilic amorphous solid solutions
and Reneker 1993). An illustration of a simple containing ITZ (Verreck et al. 2003). HPMC-
electrostatic spinning process is shown in based compositions containing 20% (w/w) and
Fig. 13.18. Factors affecting the overall diameter 40% (w/w) ITZ were prepared in an ethanol and
of the electrostatically spun fibers include, but are methyl chloride co-solvent system. A solids con-
not limited to, polymer solution viscosity, surface centration of 12% (w/w) was chosen based on
tension of the polymer solution, electric field optimal viscosity. A voltage of 16 kV or 24 kV
strength, feed rate, and dielectric constant was applied, and unwoven fibers were collected.
(Deitzel et al. 2001). This technique was then Fibers containing 40% ITZ processed at 16 kV
applied to the pharmaceutical industry for both and 24 kV were found to have diameters ranging
controlled and immediate release applications from 1 to 4 μm and 300 to 500 nm, respectively,
(Baldoni et al. 2001; Igantious and Sun 2004; as shown in Fig. 13.19a, b. However, as the drug
Kenawy et al. 2002; Verreck et al. 2003). In loading was decreased to 20% and processed at
pharmaceutical applications, a drug substance 24 kV, fiber diameters ranged from 500 nm to
and polymer are dissolved in a solvent system, 3 μm, as shown in Fig. 13.19c. These findings
as would be required in a spray-drying process or demonstrated that drug concentration within the
the physical mixture is melted and processed fiber and the electric potential applied can both
using melt electrospinning in a solvent free sys- have a significant impact on the morphology of
tem if the components are not thermolabile, and the resulting fiber.
the resultant molten mass is of optimum viscosity Unmilled fibers were found to be amorphous
(Ignatious et al. 2010). in nature and showed no signs of crystalline
Fig. 13.18 Schematic of Collector

an electrostatic spinning
system in a (a) vertical
set-up and (b) horizontal
set-up. (Reproduced with
Syringe Polymer solution
permission from Bhardwaj,
Spinneret
Kundu 2010)
Fibers
High Voltage
Fig. 13.19 Scanning electron micrograph of (a) 500). In the insert (magnification 8500), the bar
ITZ/HPMC 40:60 w/w electrostatically spun fibers at represents 3 μm, and (c) ITZ/HPMC 20:80 w/w electro-
16 kV (magnification 2000), (b) ITZ/HPMC 40:60 statically spun fibers at 24 kV (magnification 2000).
w/w electrostatically spun fibers at 24 kV (magnification (Reproduced with permission from Verreck et al. 2003)
character, but the presence of a solid solution was electrostatically spun in a PVP matrix at a con-
not verified. It was found that when milled, the centration of 40% and demonstrated by DSC and
composition containing 40% ITZ exhibited a XRPD analyses to be amorphous. The nanofibers
small degree of crystallinity. An in vitro dissolu- were dosed to fasted adult male beagle dogs to
tion analysis of the unmilled fibers in 0.1 N HCl assess their absorption. Similarly, compressed
showed complete ITZ release; however, the rate nanofibers (pellets), non-milled compound I, and
was slower than that of solvent cast films and wet-bead-milled compound I were accessed. The
HME-processed compositions. plasma AUC values demonstrated that wet-bead-
Ignatious et al. further demonstrated the utility milled compound I provided the highest degree of
of electrostatic spinning for the preparation of absorption, followed by the nanofibers, nanofiber
solid dispersions (Ignatious et al. 2010). In this pellets, and non-milled compound I. The
work, the researchers evaluated three compounds researchers stated that the nanofiber composition
in poly (ethylene oxide) (POLYOX®), could be optimized to further improve absorption.
polyvinylpyrrolidone-vinyl acetate copolymer In a second example by Ignatious et al. another
(PVPVA), and poly(vinyl pyrrolidone) (PVP) proprietary compound (compound II) was
matrices. In one example by Ignatious et al. a formulated by electrostatic spinning in both
proprietary compound (Compound I) was POLYOX® and Eudragit® L100-55 matrices.
A
40
Non-milled Compound II
35
Electrospun Compound II in Eudragit L100-55
30 Nano-milled Compound II
Electrospun Compound II in POLYOX
% Dissolved
25
20
15
10
0
0 5 10 15 20 25 30 35 40
Time (min)
B
50
Non-milled Compound II
45
Electrospun in Eudragit L100-55
40
Nano-milled Compound II
35 Electrospun Compound II in POLYOX
% Dissolved
30
25
20
15
10
0
0 5 10 15 20 25 30 35 40
Time (min)
Fig. 13.20 In vitro dissolution profiles of Compound II at (a) pH 1.0 and (b) pH 7.5. (Reproduced with permission from
Ignatious et al. 2010)
The in vitro dissolution rates were studied at pH milled formulation and the POLYOX®-based
values of 1.0 and 7.5, as illustrated in Fig. 13.20. fibers. This dissolution rate increase was primar-
At pH 1.0, the matrix utilizing POLYOX® ily due to the amorphous nature of compound II
exhibited release rates similar to a nano-milled and/or the concentration enhancing properties of
formulation. The dissolution rate of Eudragit® Eudragit® L100-55.
L100-55-based matrices was very low due to its More recently, coaxial electrospinning was
poor solubility at this pH value. At a dissolution utilized to prepare a multi-component fiber sys-
medium pH of 7.5, the Eudragit® L100-55-based tem (Yu et al. 2011). For this system, a core
electrostatically spun formulation provided a sig- formulation was prepared using acyclovir and
nificantly faster dissolution rate than the nano- PVP, with a surrounding sheath consisting of
Fig. 13.21 CAR drug 110

release in 900 mL 0.1 M
HCl, USP II at 50 rpm. 100
(MES melt electrospun 90
fibers, SES solvent-based
electrospun fibers, EX HME 80
particles, CAR unprocessed
70
Dissolution (%)
crystalline carvedilol).
(Reproduced with 60
permission from Nagy et al.
2013) 50
40
30
20
10
0
0 1 2 3 4 5
t (min)
MES SES EX CAR
PVP, sodium dodecyl sulfate, and sucralose. With 0.7 μm, respectively, which also resulted in lower
a high melting point (257 C) and poor solubility specific surface area with the MES fibers. The
in water, as well as many organic solvents, acy- HME particles were milled to 20 μm mean particle
clovir was especially challenging to process using size, and they were all evaluated for drug release
conventional HME or spray-drying methods. against the crystalline CAR, as shown in
However, using coaxial electrospinning, the core Fig. 13.21. Surprisingly, the MES material, with
solution of 10%w/v PVP and 2%w/v acyclovir in larger fiber diameter and lower surface area,
a dimethylacetamide: ethanol (4:6), not typically exhibited a faster release profile than the
amenable to electrospinning, was able to be electrospun fibers; while both fibers showed faster
processed as the sheath solution, consisting of release than the HME particles. A similar drug
10%w/v PVP, 0.5%w/v SDS, and 0.2%w/v release trend was seen with CAR from a
sucralose in a water:ethanol (2:8) acted as a copovidone (PVPVA 64) carrier (Balogh et al.
guide, enabling the formation of core-sheath 2015a, b). As with HME, plasticizers can be
nanofibers. These prepared fibers resulted in added to reduce processing temperatures as
rapid drug release, 100% within 1 min due to needed (Balogh et al. 2014).
the amorphous nature of the drug and high sur- Electrostatic and MES are promising pro-
face area, compared to the 40% drug release in cesses to produce ASDs that exhibit very large
60 min from micronized (<100 μm) acyclovir surface areas due to the production of nano- and
particles. micro-sized fibers. The rapid drying of the sol-
Nagy et al. conducted a study to evaluate the vent-based electrospinning process is able to pro-
properties of a dispersion of carvedilol (CAR) duce amorphous drug molecules dispersed in the
with Eudragit® E prepared using HME, polymer matrix and allow for higher drug loading
electrospinning, and MES (Nagy et al. 2013). All than HME (Nagy et al. 2012; Yu et al. 2011).
methods produced amorphous solid dispersions A current study conducted by Yildiz Z.I. and
with a 20% drug load; however, the mean fiber colleagues utilized electrospinning to manufac-
diameter of the MES material was significantly ture polymer-free inclusion complexes of
larger than that from electrospinning, 250 versus sulfisoxazole using sulfobutyl ether7-beta-
cyclodextrin (SBE7-β-CD, Captisol®). For this generate heat due to the high-frequency recipro-
study, a highly concentrated aqueous solution of cation of dipole molecules. Unlike other methods
the components were electrospun to produce a which use conduction, convection, or radiation,
free-standing and handy nano-fibrous web, with where the heat transfer is not uniform and mixing
an average fiber diameter of 650 290 nm. It was is essential to prevent over or uneven heating,
further observed that the produced nanofiber microwave-based techniques depict uniform and
depicted an enhanced solubility for the loaded simultaneous heating of the mass without mixing.
drug. This was attributed to the increase in surface Microwaves have been used in the rapid solvent-
area and the inclusion complexation as a result of free synthesis of organic molecules because of
the process (Yildiz et al. 2017). Similar studies their ability to induce uniform heating and
for manufacturing inclusion complexes have been recently gained attention in the manufacturing of
conducted using the electrospinning process amorphous solid dispersions (Zhang et al.
(Yildiz et al. 2018; Daghrery et al. 2020; Aytac 2021a, b). Previous studies have shown a strong
et al. 2019; Balogh et al. 2015a, b). correlation between the microwaving time and
Although the above case study describes poly- the amorphous conversion of the drug in the
mer-free electrospinning of poorly water-soluble formulation at different drug loads. Edinger
drugs, based on the drug candidate, advanced M. and colleagues conducted a study to quantify
formulation strategies may be required to develop the impact of microwaving time and drug load on
a formulation. In another study by Domokos the amorphous conversion of celecoxib in the
A. and colleagues, they developed an orally polyvinylpyrrolidone (PVP) matrix. They kept
disintegrating web for carvedilol. Initial studies the power constant at 1000 W and inspected the
showed that the inclusion of the drug with samples’ solid-state using varying drug loads
hydroxypropyl-β-cyclodextrin (HPβCD) was not (10–50%) using Raman spectroscopy every min-
sufficient to provide the required solubility ute for a total time of 10 min. The partial least
(6.25 mg CAR in 20 mL) in the oral cavity, square regression model observed a strong corre-
even in the presence of citric acid to ionize the lation between independent variables (i.e., the
basic drug. However, PVP K30, along with citric microwaving time) and the drug load on the
acid, achieved notable supersaturation of the dependent variable (i.e., amorphous drug con-
drug. The formulation was prepared by layering tent). It was observed that over 90% of the drug
the nanofibers containing drug and polymer onto was amorphous post-processing for the 10% drug
pullulan, which is a well-soluble polysaccharide load, whereas only 30% of the drug was
film carrying citric acid. This double-layered for- converted for a 50% drug load, results shown in
mulation showed ultrafast disintegration and dis- Fig. 13.22. This partial conversion can be
solution and met regulatory requirements for attributed to the polymer’s solubilization capacity
orodispersible dosage forms (<30 s). On further and the absence of mixing during the process
assessment, the drug in the formulation was pres- (Edinger et al. 2018).
ent in its amorphous state, which may be partially Another study by Doreth M. and colleagues
responsible for the enhanced dissolution rate investigated the impact of polymer (PVP K
(Domokos et al. 2019). 12, 17, and 25) molecular weight on microwave-
assisted amorphization of indomethacin at 300 W
for 5 and 10 min. They observed that an increase
13.2.4 Microwave-Induced in polymer molecular weight reduces the amor-
Dielectric-Induced Heating phous conversion; whereas, an increase in input
energy aids the amorphous conversion. The latter
Dielectric- or microwave-induced heating is in agreement with the research published by
employs electromagnetic radiation, including Edinger M. and colleagues. Moreover, Doreth
radio waves, or microwaves which are absorbed M. and colleagues also concluded that
by the dielectric material. These materials plasticization and, thereby, Tg of the polymer

the relative amorphous
fraction of celecoxib as a
function of microwaving
time for 10% (red), 20%
(black), 30% (green), 40%
(blue), and 50% (magenta)
CCX tablets (error bars
indicate standard deviation,
n ¼ 3). (From Edinger et al.
2018)
also play a crucial role in the drug’s amorphous Microwave-assisted dielectric heating can be
conversion (Doreth et al. 2018). used to convert directly compressed tablets with
Zhang J. and colleagues demonstrated the an optimized formulation into amorphous solid
pharmaceutical applicability of microwave- dispersion at the point of care or by the patient.
assisted dielectric heating and manufacturing of This could circumvent the storage stability and
amorphous solid dispersions thereof. In this recrystallization issues observed in other ASD
research, directly compressed indomethacin manufacturing platforms. Although restrictions
tablets were compared with ASD formulation such as the type of polymers acceptable, the
using different polymers (Soluplus, HPC HF, drug load, and the absence of mixing during
HPMC, HPMCAS, EPO, and PEO) having a processing may lead to uncertainties in the per-
drug loading of 30%. The drug’s amorphous con- formance of the final dosage forms; moreover, the
version was monitored using the yellowness and energy provided by microwave radiation could
an IR sensor, where the complete amorphous also lead to unwanted drug-excipient or drug–
conversion was determined by the end time for drug interactions; hence extensive research is
each pair. This complete amorphous conversion required to optimize this otherwise promising
was also confirmed using DSC and XRD analy- process.
sis. The processing time depended on the
polymer’s glass transition temperature, where
Soluplus depicted a complete conversion in less
13.3 Formulation-Based Techniques
than 15 min and HPMCAS MG required over
30 min for complete conversion. Zhang J. and
While the formation of ASDs is an effective tech-
colleagues observed that Soluplus, HPMCAS
nique to enhance poorly water-soluble
MG, and PEO matrices demonstrated a quicker
compounds’ bioavailability, various other
and complete release of the drug (>80%), whereas
formulation-based techniques can achieve the
other polymers exhibited a better release in com-
same goal. Many of these techniques are covered
parison to the directly compressed tablets, but the
in previous chapters, including the formation of
release was not complete (<40%), results shown
salts, co-crystals, co-solvent or complexed
in Fig. 13.23 (Zhang et al. 2021a, b).
systems, and lipid-based or self-emulsifying
Fig. 13.23 In vitro drug release profiles of the IND from the DC and ASD tablets (n ¼ 3). (From Zhang et al. 2021a, b)
drug delivery systems. However, some of these such as catalysis and optics; however, by 2001,
systems have limitations, such as toxicity limits these high surface areas and highly porous
with surfactants and other lipids; thus, formula- materials were being evaluated as carriers for
tion techniques continue to be developed for bio- drug delivery (Vallet-Regi et al. 2001). With the
availability enhancement. recent clinical approval of the Cornell Dots
(C dots), the safety of these inorganic materials
is being evaluated, opening the door for future
13.3.1 Mesoporous Materials clinical studies for OMS for drug delivery
(Bradbury et al. 2015).
Mesoporous materials, defined as having pore OMS materials can have a wide range of
diameters from 2 to 50 nm, have emerged as mesostructures from 1D lamellar to tetragonal
promising carriers for poorly soluble compounds. (Garcia-Bennett et al. 2005); however, the most
The distribution of nanosized pores provides a commonly used OMS materials for drug delivery
high surface area for drug loading and thin are characterized as having a 2D hexagonal or 3D
channels that prevent the formation of highly cubic mesostructures, as shown in Fig. 13.24.
ordered crystals, maintaining deposited drug in Materials are available in a wide range of
its metastable amorphous state (Wang et al. mesopore diameters (2 nm < Dp < 50 nm) and
2010). Additionally, the material’s pore size or have relatively large pore volumes of 0.6–1.0 mL/
surface chemistry can be readily tailored to allow g (Mamaeva et al. 2013). Two of the most com-
for immediate, sustained, or controlled release of monly used OMS materials investigated in recent
the adsorbed drug (Wang et al. 2015). years are MCM-41 (Mobil Composition of Matter
No. 41) and SBA-15 (Santa Barbara Amorphous
Mesoporous Silica No. 15) (Mellaerts et al. 2008b; Wang et al.
Ordered mesoporous silica (OMS) materials were 2013). The pore diameter ranges between 2 and
discovered in the early 1990s by the Mobil Cor- 10 nm for MCM-41 and 5 and 15 nm for SBA-15
poration for use in a broad range of applications (Zhao et al. 1998). Note that SBA-15 can be
Fig. 13.24 TEM image (left top) and model of unit cell (left bottom) of 2D hexagonal mesostructure and TEM image
(right top) and model unit cell (right bottom) of 3D cubic mesostructure. (Reproduced with permission from Fadeel and
Garcia-Bennett 2010)
prepared to exhibit a 2D hexagonal or 3D cubic addition of a surfactant as a structure or pore-

structure (Kruk et al. 2000). In addition to small forming agent. A washing or calcination step is
pore diameters and high pore volumes, these used to remove the surfactant, leaving behind
materials have a high surface area empty pores (Kresge et al. 1992; Inagaki et al.
(600–1000 m2/g) and a dense distribution of 1993). The particle size, pore size and volume,
silanol groups (2–3 groups/nm) to serve as and mesostructures can be readily tailored by
adsorption sites or as sites for complexation with processing parameters, such as surfactant type
other functional groups, such as carboxylic acids and temperature (Hu et al. 2011). Drug loading
or amines (Mamaeva et al. 2013; Wang et al. and addition of functional groups typically occur
2015). post-synthesis. Methods for drug loading include
OMS materials are prepared by a sol–gel or adsorption from solution, electrostatic adsorption,
ion reaction method in acid or alkaline conditions incipient wetness impregnation, the melt method,
to produce homogeneous-size particles with the or co-spray-drying (Mellaerts et al. 2008a; Shen
et al. 2010). Adsorption from solution is the most loading of 50.7%, 61.2%, and 60.3% drug load-
common method for poorly soluble compounds, ing (Zhang et al. 2010b). These loading efficiency
whereby the OMS particles are immersed in a of the OMS was attributed to the OMS silanol
concentrated organic solution of the active com- groups forming hydrogen bonding with the drug,
pound and the resulting suspension is dried to albeit weak bonds, as seen by the rapid release of
remove residual solvent, leaving behind the the drug from the carrier during in vitro release
adsorbed drug. Notably, the polarity of the sol- testing. The release behavior was consistent with
vent can affect drug-loading efficiency, as polar the pore diameter, with 83.5%, 95.4%, and 100%
solvents will compete for the adsorption sites release seen at 30 min with increasing pore diam-
(Xu et al. 2013). eter. Zhang et al. prepared OMS carriers with
Vallet-Regi et al. (2001) first evaluated OMS 28.3 nm pore diameter, also referred to as a
as a reservoir for controlled drug-delivery mesocellular foam, to increase drug release rate
systems (Vallet-Regi et al. 2001). In this initial of poorly soluble drug simvastatin (SIM) for
study, the researchers utilized MCM-41 with two 100% release by 60 min, compared to 62% from
different pore sizes for the delivery of ibuprofen, SBA-15 with a 6.5 nm pore diameter (Zhang et al.
whose molecular size was in the range of the 2011). Wang et al. showed that the preparation of
mesopores. In order to load the OMS, substrates 3D continuous and interconnected macroporous
were submerged in a solution of ibuprofen and (pore size 200 and 500 nm) silica particles
hexane (33 mg/mL). Upon removal, substrates resulted in faster dissolution of adsorbed indo-
contained 30% (w/w) ibuprofen, demonstrating methacin (IND) when compared to conventional
that ibuprofen was absorbed into the mesopores, OMS carriers, MCM-41 and SBA-15 OMS.
which was further confirmed by BET analysis. In However, this led to reduced physical stability,
vitro dissolution analysis of the ibuprofen-loaded showing recrystallization after 3 months at
OMS exhibited rapid initial release followed by a accelerated stability conditions (40 C/75%RH)
sustained release, with 80% release reached after (Wang et al. 2013), demonstrating the trade-off
3 days. This type of two-phase release behavior between rapid dissolution and physical stability
has been seen across various OMS studies due to as a function of pore size.
the release of the surface adsorbed drug followed Mellaerts et al. stored itraconazole-(ITZ)-
by the diffusion of drug from the pores (Thomas loaded SBA-15 particles (10 and 20%w/w drug
et al. 2010); the desorption of drugs in the pores in loading) at 25 C at 0,52 and 97% RH for up to
the presence of water is explained by the affinity 12 months to determine the long-term stability of
OMS has for water over hydrophobic compounds adsorbed amorphous ITZ (Mellaerts et al. 2010).
and is diffusion controlled (Mellaerts et al. Storage at the high humidity (52 and 97% RH)
2008a, b; Andersson et al. 2004). While the conditions led to increased adsorbed water, up to
focus of initial studies conducted by Vallet-Regi 30% water content in the 10% ITZ formulation.
et al. was on controlled drug-delivery systems, a The majority of this water was only physically
major focus of OSM research has shifted to adsorbed, while a small percentage (1.39–1.64%)
immediate release systems for the delivery of was determined to be chemically bound to the
poorly water-soluble compounds; this section SBA-15 surface via hydroxylation. This yielded
will highlight the recent work on the use of a faster and more complete release of ITZ from
OMS for the delivery of poorly soluble the SBA-15. These changes in drug release were
compounds. attributed to the hydroxylation of the SBA-15
For immediate release applications of OMS surface from the bound water, rendering the sur-
materials, the dissolution rate of the drug sub- face more hydrophilic and more prone to
stance can be tailored by the pore size and surface liberating the ITZ when placed in an aqueous
area. Zhang et al. studied the loading of poorly environment, as shown in Fig. 13.25. The physi-
water-soluble telmisartan (TEL) onto OMS using cal stability of the ITZ was maintained, most
pore sizes of 4, 8, and 13 nm, obtaining drug probably due to the unchanged or decreasing
drug release, particularly at high compression

forces. With the addition of microcrystalline cel-
lulose, a plastically deforming filler, and sodium
crosscarmellose sodium, a disintegrant, there was
some recovery of drug release seen when com-
pressed at 120 MPa. In another study, Vialpando
et al. evaluated the stability of drug loaded
ITZ-OMS particles during a small-scale wet gran-
ulation process to improve bulk powder
properties (Vialpando et al. 2012). Using a 20%
PVP aqueous solution, granules were prepared
without an impact to the OMS pore diameter
and volume, indicating physical integrity of the
particles. Subsequent compression at 120 MPa
compression force did produce a drop in the
pore diameter and volume; however, the addition
Fig. 13.25 Adsorbed water from humid storage of a disintegrant to the tablet formulation enabled
conditions causes hydroxylation of silanol surface, leading
to increased release rate of drug upon contact with aqueous comparable drug release when compared to the
fluids. (Reproduced with permission from Mellaerts et al. granules alone.
2010) In addition to the loading and stabilizing
benefits from the nanosized pores of OMS
pore volume that prevented molecular mobility materials, many of these materials are often
for crystal formation, further emphasizing the precipitated as nanosized particles or mesoporous
benefits of mesoporous materials for use with silica nanoparticles (MSN). This offers an obvi-
poorly soluble compounds. ous advantage to increase cellular uptake, as seen
As with amorphous solid dispersion systems, by its early adoption and continuing development
the use of a precipitation inhibitor can be advan- for parenteral formulations of cancer drugs, and
tageous to maintain supersaturated particularly, poorly soluble cancer agents such as
concentrations. Van Speybroeck et al. were able paclitaxel (Wang et al. 2015; Meng et al. 2010;
to show improved maintenance of supersaturation Lu et al. 2007; Slowing et al. 2006; Mekaru et al.
of ITZ in FaSSIF media with the inclusion of 2015). Zhang et al. evaluated the use of MSNs for
HPMC AS and a small improvement with the oral delivery of TEL and showed not only
HPMC, when physically mixed with the drug- increased cellular uptake when compared to
loaded SBA-15 particles (Van Speybroeck et al. mesoporous silica microparticles (MSM) but
2010). Interestingly, the improvement from also a reduced rate of efflux and thus greater
HPMC AS was not seen in vivo due to its low permeability when compared to a control TEL
solubility at pH < 5.5; however, the HPMC was solution (Zhang et al. 2012). This increased per-
able to produce a 60% increase in bioavailability meability correlates well with previous findings
when compared to the OMS (SBA-15)-ITZ alone of increased permeability of other poorly soluble
(AUCsum 14,937 1617 versus compounds, such as GRIS, from MSNs (Bimbo
8987 2726 nM h). et al. 2011). In vivo studies conducted on beagles
Vialpando et al. evaluated the downstream (n ¼ 6) showed tablets prepared using ITZ-MSNs
processing of itraconazole-(ITZ)loaded OMS in and MSMs performed better than the commercial
an oral tablet formulation (Vialpando et al. 2011). tablet, Micardis®, with relative bioavailabilities of
This study showed that compression of neat OMS 154.4% and 129.1%, respectively. This higher
and ITZ-loaded OMS resulted in decreased pore bioavailability of the MSN formulation was
volume and surface area due to the collapse of the attributed to the improved absorption of the
pore walls, which subsequently led to reduced
Fig. 13.26 Loading and functionalization possibilities using OMS for drug delivery. (Reproduced with permission from
Rosenholm et al. 2010)
amorphous TEL from the MSN and the reduced release from OMS could be significantly retarded
rate of P-gp-mediated efflux. (Zhang et al. 2010b). The grafted aminopropyl
With a high concentration of silanol groups on groups exhibited stronger bonding with the TEL
the mesoporous silica wall surfaces, functiona- than seen with the OMS silanol groups, slowing
lization with many types of organic molecules to drug release significantly. In vitro drug release
modulate drug-release kinetics is possible, as testing showing 100% release by 24 h. These
shown in Fig. 13.26 (Ukmar and Planinšek results correlate well with previous studies
2010; Yang et al. 2012). Functionalized OMS conducted with aminopropyl group modifications
systems are especially desirable from a drug to MCM-41 carriers for the delivery of IBU
delivery standpoint, allowing for targeted release (Muñoz et al. 2003).
and/or increased protection of the API from harsh As an alternative to the synthesis of OMS
biological environments. Researchers have materials, existing pharmaceutical grade silicas,
reported the use of a variety of functional i.e., fumed or colloidal silica (Aerosil®) or mag-
moieties for sustained, controlled (Zhang et al. nesium aluminum silicate (Neusilin®), have also
2010b; Muñoz et al. 2003; Song et al. 2005; been utilized as carriers for poorly soluble
Wang et al. 2009; Qu et al. 2006; Tang et al. compounds (Kinnari et al. 2011; Kim 2013).
2006, 2010; Bernardos et al. 2008; Xu et al. These materials tend to exhibit a more
2008), or pH-dependent release (Peng et al. non-ordered structure and can be driven mostly
2013; Yang et al. 2005). With the addition of an by surface adsorption as opposed to adsorption
aminopropyl group, Zhang et al. showed that TEL within the pore walls of the mesostructures
previously described; therefore, are often materials have shown an increased production of
prepared by using adsorption of the poorly solu- reactive oxygen species (ROS), specifically the
ble compound with a polymer, precipitation superoxide O2.-, in Caco-2 cells at a threshold
inhibitor, and/or solubilizer (Kim 2013). This is concentration of 1 mg/mL (Heikkilä et al. 2010;
the premise of the Spray-dried NanoAdsorbate Bimbo et al. 2011). Increased levels of ROS can
technology developed by Bend Research®, lead to weakened cell membranes, diminished
whereby the dissolved drug and polymer and cell metabolism and increased apoptotic signal-
suspended silica carrier are spray dried to form ing. The production of these species was
solid particles with adsorbed amorphous drug– attributed to the silanol groups along the surface
polymer dispersion on the surface of the silica of these OMS materials. Based on this, develop-
particles. These particles are then amenable for ment of inert ordered mesoporous carbon (OMC)
downstream processing, such as encapsulation or materials was initiated for use in drug delivery
tableting. (Kim et al. 2008; Oh et al. 2010; Saha et al.
Recently, OMS materials have been 2014a). Cytotoxicity studies with OMC materials
incorporated within the HME process to improve in Caco-2 cells showed no apparent toxicity at
solubility, stability, and drug loading of poorly concentrations up to 800 μg/mL (Zhao et al.
water-soluble drugs. During processing by HME, 2012b). At higher concentrations, cell survival
benefits have been seen with both binary was reduced, although it remained above 80%,
compositions (i.e., OMS and API) and ternary due to carrier agglomeration inside the cell. Stud-
compositions (i.e., OMS, API, and polymer). ies in HeLa cells also showed positive cell viabil-
For example, Maniruzzaman et al. demonstrated ity for concentrations up to 500 μg/mL, with
the ability to use the inorganic excipient, magne- slight toxicity seen with the highest surface area
sium aluminometasilicase, within HME to render OMC tested (~70% survival) at the highest con-
indomethacin amorphous within the porous net- centration (Gencoglu et al. 2014).
work at various drug loadings (i.e., 20%, 30%, As with OMS, OMC materials are also utilized
and 40%) while retaining physical stability at in other systems and processes, including cataly-
elevated conditions for 12 months sis and separations. Additionally, porous carbon
(Maniruzzaman et al. 2015). More recently, (or activated charcoal) has been administered to
Hanada et al. demonstrated the benefit of treat acute poisoning in an emergency setting due
incorporating amorphous itraconazole (ITZ) into to its highly absorptive properties (Kulig et al.
a high-viscosity polymer (HPMC AF4M) that is 1985; Neuvonen and Olkkola 1988;
efficiently loaded into mesoporous silica (XDP) Yachamaneni et al. 2010). The preparation of
pours during the extrusion process. All grades of OMC materials involves hard-templating or soft-
HPMC successfully incorporated the components templating, also referred to as self-assembly
within the porous structure and rendered ITZ (Liang et al. 2008; Ma et al. 2013). Hard-
amorphous at a 50% drug loading. During templating involves preparation of an OMS struc-
in vitro pH-shift dissolution, the formulation that ture, followed by the incorporation of a carbon
incorporated the highest molecular weight of source, i.e., sucrose; the particles are then washed
HPMC, best maintained supersaturation after to remove the silica template. Soft-templating,
transitioning from the acidic to the neutral phase shown in Fig. 13.27, involves cross-linking a
(Hanada et al. 2020). suitable organic resin in the presence of pore
forming surfactants; suitably those that do not
Mesoporous Carbon have a char yield. The cross-linking is followed
The physical liabilities with OMS materials may by pyrolysis, which removes the surfactant, and
limit its pharmaceutical application, including the lastly, the carbonization of the matrix. OMC
hydroxylation of its silanol groups upon aging materials have been shown to exhibit larger pore
and the ease of collapse upon compression. Addi- volumes and greater surface areas than their silica
tionally, safety studies conducted with OMS counterparts (Zhang et al. 2013).
Fig. 13.27 Schematic of soft-templating for the preparation of OMC particles. (Reproduced with permission from Saha
et al. 2014b)
Fig. 13.28 Adsorption 2.0

isotherms showing IBU
solution concentration C-70*
impact on OMC drug
loading with the following C-100*
sample name (pore 1.5
diameter, pore volume):
C-70 (4.1 nm, 0.97 cm3/g), C-70
n/w (mmol/g)
C-70 (4.1 nm, 0.97 cm3/g),

C-70* (3.9 nm, 1.42 cm3/ C-100
1.0
g), C-100 (3.5 nm,
0.64 cm3/g), C-100*
(3.4 nm, 1.11 cm3/g).
(Reproduced with
permission from Wang 0.5
et al. 2011a)
0.0
0 20 40 60 80 100 120 140
c (mmol/l)
Wang et al. showed the effect of ibuprofen from the pores; however, in the absence of the
(IBU) solution concentration and pore volume silanol groups, the release of the drug from the
on OMC drug loading, revealing adsorption OMC carrier is much faster. Zhao et al. evaluated
isotherms which plateaued as shown in the use of OMC for delivery of a poorly soluble
Fig. 13.28, due to saturation of adsorption sites drug, using lovastatin (LOV) as a model com-
(Wang et al. 2011a). The drug release from the pound (Zhao et al. 2012b). 2D hexagonal and
OMC is similar to that of OMS materials in that 3D cubic structures, also referred to as
there is a rapid release of surface adsorbed mesoporous carbon spheres (MCS), of OMC
materials, followed by a slower release of drug materials were prepared with varying pore sizes

from pure crystalline LOV
(a), 3D OMC with pore size
6.0 nm (b), 5.1 nm (c),
3.8 nm (d), and 2D OMC
with pore size 5.0 nm (e),
4.4 nm (f), 7.0 nm (g).
(Reproduced with
permission from Zhao et al.
2012b)
Table 13.1 Properties of MSM and c-MCM particles, empty and loaded with CAR
Sample SBET (m2/g) Vt(cm3/g) wBJH (nm) Vmi (cm3/g) Drug loading (%)
MSMs 893.1 8.5 1.03 0.04 9.1 1.8 0.1 0.02
c-MCMs 1736.2 13.3 1.82 0.1 7.8 1.9 0.37 0.09
CAR-MSMs 271.8 9.5 0.36 0.04 3.4 1.5 0 30.1 0.4
CAR-c-MCMs 141.0 23.7 0.15 0.11 2.6 0.3 0 41.6 1.2
Reproduced with permission from Zhang et al. (2013)
and loaded with LOV using a 60 mg/mL organic oral bioavailability of carvedilol (CAR) (Zhang
solvent solution. The 3D cubic material produced et al. 2013). The carboxylation was incorporated
higher pore volumes and surface areas than the to increase wetting of the MCM upon ingestion.
2D hexagonal materials. The 3D structure also The c-MCMs showed significantly higher pore
showed less residual crystallinity of the adsorbed volume and surface area to MSMs prepared, as
LOV than seen with the 2D cubic materials; how- shown in Table 13.1, even with only a small
ever, XPS analysis showed that approximately difference in pore diameter. These properties
53% of the drug was present on the surface of enabled a higher drug loading with the c-MCM
the 3D structure (10–15 nm depth of analysis), particles (41.6%) than with the MSM particles
compared to the 38% on the surface of the 2D (30.1%). The c-MSM formulation showed a sig-
hexagonal material. The in vitro drug release, nificant increase in bioavailability in vivo in bea-
shown in Fig. 13.29, showed rapid release of gle dogs against the commercial Dilatrend® tablet
LOV from the 3D OMC with complete release at a 200 mg dose with AUC0-24h of
in around 90 min for all pore sizes. The release 784.79 56.29 ng h/mL vs. 420.27.2
from the 2D structure showed in initial rapid 41.22 ng h/mL of the control. This relative
release, attributed to the surface adsorbed LOV, bioavailability of nearly 180% was attributed to
followed by a slow release of LOV, attributed to the amorphous state of the CAR, the highly dis-
release of LOV adsorbed on the pore walls, with persed nature of the drug along the high surface
only 60–70% release by 90 min. Although, the area of the carrier, as well as the limited crystal
rate of release trended with pore size, the small size restricted by the mesopore size.
range used (3.8–7.0 nm) provided only small In addition to the benefits of increased dissolu-
differentiations between the drug release curves. tion, OMC systems can provide other advantages
Zhang et al. utilized carboxylated mesoporous as drug delivery systems. A study conducted
carbon microparticles (c-MCM) to enhance the using celecoxib (CEL) loaded MCSs with a
Caco-2 cell line showed similar suppression of concentrations (i.e., 5 M) due to the disruption
efflux as seen with MSNs (Wang et al. 2014), of electrostatically interacting polymer layers.
leading to higher bioavailability. Another study For targeted drug delivery, Zhang et al.
showed reduced mucosa irritation, a side effect of prepared glucose-based mesoporous carbon
large doses of CEL, when dosed with CEL-OMC nanospheres (MCN) stabilized by phospholipids
attributed to reduced contact of the drug with the for increased hydrodynamic stability needed for
mucosal surface as the drug is dispersed on the IV administration, as shown in Fig. 13.30 (Zhang
surface or within the pores of the OMC carrier, as et al. 2015b). The particles were able to adsorb a
well as the rapid absorption of the small, dis- drug loading of 42.7% of poorly soluble com-
persed amorphous CEL from the carrier (Zhao pound SNX-2112, an Hsp90 inhibitor to induce
et al. 2012a). apoptosis (Liu et al. 2012). When compared to a
As with OMS materials, OMC materials can ANX-2112 solution, enhanced antitumor effect
also be functionalized or coated to release drug was seen with these phospholipid stabilized
when triggered by temperature (Zhu et al. 2011), MCN due to targeting glucose transporters,
pH (Zhu et al. 2012; Zhang et al. 2015a), applied which are overexpressed in breast cancer cells,
magnetic field (Oh et al. 2010; Wang et al. for increased cellular uptake and due to the nano-
2011b), or other stimuli (Amritha Rammohan particle size of the carriers which allowed for
et al. 2013), as well as for targeted drug delivery passive accumulation due to the enhanced perme-
(Zhang et al. 2015b). As the OMC surface does ation and retention (EPR) effect seen with the
not contain the silanol groups present on OMS leaky vasculature in tumor tissues, also shown in
surfaces, chemical treatment or activation may be Fig. 13.30.
required to introduce functional groups onto the
surface for more efficient adsorption and/or Other Mesoporous Materials
functionalization. For example, Saha et al. chem- The evaluation of alternative mesoporous
ically activated the surface of their OMC materials has also gained interest for drug deliv-
materials using KOH to introduce oxygen onto ery. Inorganic mesoporous carriers have included
the carbon surface to promote mucoadhesion hydroxycarbonate apatite, titanium dioxide (Jiang
properties (Saha et al. 2015). et al. 2012). Organic mesoporous carriers have
To obtain a sustained release profile and included starch and porous polymers, such as
reduce burst release from surface adsorbed drug, PVA (Roberts and Zhang 2013; Wu et al. 2011).
Zhang et al. prepared nimeodipine (NIM) loaded Hydroxycarbonate apatite (HCA) is another
OMC with a lipid bilayer or shell (Zhang et al. material commonly employed for tissue and
2015c). This composite system was able to reduce bone engineering scaffolds. HCA is a calcium
burst release from 50% to 35% of drug release phosphate-based inorganic material that has
within 2 h and extended the release profile to often been used as artificial bone materials. Its
18 h vs. 12 h for 100% release when compared inert nature, biocompatibility, relatively high sur-
to the NIM-OMC alone. Alternatively, for a stim- face area, and uniform pore formation has led to
ulus based release, Rammohan et al. utilized the the adoption of HCA and hydroxyapatite (HA) as
slightly negative zeta potential (0.3 mV) of their drug delivery systems (Zhang et al. 2010a; Zhao
OMC particles to facilitate a layer by layer (LbL) et al. 2012c; Ye et al. 2010; Mizushima et al.
polymer coating using positively charged poly- 2006). Zhao et al. prepared mesoporous HCA
electrolyte, polyethylene imine (PEI), followed particles using a hard-template method with
by negatively charged polystyrene sulfonates CaCO3 as the sacrificial template and CTAB as
(PSS), and lastly, by positively charged the pore-forming agent (Zhao et al. 2012c). The
polyallylamine hydrochloride (PAH) (Amritha particles with the largest pore volume and surface
Rammohan et al. 2013). In this system, the drug area (0.175 mL/g and 121.9 m2/g, respectively)
is released in the presence of high salt were then loaded with CAR using carrier to drug
ratios of 3:1, 2:1, and 1:1 to obtain drug loading
Fig. 13.30 Glucose based MCN stabilized by phospholipids for IV administration for targeted drug delivery to tumor
site. (Reproduced with permission from Zhang et al. 2015b)
of 23%, 31%, and 48%, respectively; however, techniques, including injection molding, in situ
the higher drug load particles (31% and 48%) polymerization, or using a solvent-exchange
showed a higher degree of residual crystallinity. method to produce a porous gel (Nakamatsu
The 23% drug loaded formulation showed a faster et al. 2006). Wu et al. evaluated the use of porous
and more complete drug release with over 90% starch as a carrier for poorly soluble drug, LOV
release seen by 20 min when compared to the (Wu et al. 2011). Although the large and irregular
other drug loads and raw drug, attributed to pore diameters, 20–80 nm, resulted in adsorbed
higher degree of amorphous CAR on the lower LOV present in a partially amorphous form, the
drug loaded formulation. The 23% drug loaded dissolution rates were still faster than the com-
particles were also shown to be physically stable mercially available capsules.
for when held in a desiccator at 25 C for
6 months; however, as FTIR analysis showed no
physical interaction, the stability of the adsorbed 13.3.2 Co-amorphous Mixtures
drug was attributed to the inhibition of recrystal-
lization by the restrictive pore size. The amorphous state of a drug is a high-energy,
Porous starch has been used for tissue engi- highly disordered state that enables higher appar-
neering and bone cements due to their biocompat- ent solubility and a faster dissolution rate than its
ible and biodegradable nature (Gomes et al. 2001, crystalline form (Hancock and Zografi 1997).
2002). Additionally, the availability of hydroxyl However, this is a metastable form that has a
groups on the surface of starch enables tendency to crystallize during manufacture, stor-
interactions for adsorption and/or functiona- age, or during drug release. ASDs using polymer
lization. Porous starch can be prepared by various carriers have become a preferred method to
enhance solubility and bioavailability of poorly well with theoretical values calculated by the
soluble compounds utilized in several marketed Gordon–Taylor equation, which presumes no
products including Sporanox®, Kaletra®, and interaction between the two components. How-
Incivek®. In these polymer-based systems, the ever, solid state characterization of other systems,
drug is molecularly dispersed in a polymer carrier including cimetidine (CIM)–naproxen (NAP)
and can be kinetically stabilized by high Tg and/or co-amorphous mixtures exhibited Tgs higher
high viscosity polymers to reduce molecular than the Gordon–Taylor prediction, suggesting
mobility. Additionally, chemical bonding an interaction between the imidazole ring of the
between the polymer and drug can lead to addi- CIM with the carboxyl group of the NAP, which
tional stabilization (Janssens and Van den Mooter may contribute to the formation and stability of
2009; Laitinen et al. 2013). One of the limitations the mixture (Yamamura et al. 1996; Allesø et al.
of polymer-based solid dispersions is the high 2009). This interaction was also seen with a
weight percentage of polymer that is often needed CIM–indomethacin (IND) mixture (Yamamura
to form a molecular dispersion or glassy solution; et al. 2000), while a CIM–diflunisil mixture
this can lead to a higher pill burden for drugs that showed a stronger interaction, forming an amor-
require higher doses. Additionally, many of the phous salt (Yamamura et al. 2002). Alesso et al.
polymer carriers used require a plasticizer for further evaluated the CIM–NAP co-amorphous
processing and/or are hygroscopic; plasticizer mixture prepared by co-milling, in place of the
and adsorbed water, which has a plasticizing solvent evaporation methods from the previous
effect, can lower the system Tg, leading to physi-
cal instability. Co-amorphous systems, defined as
a combination of two small molecules, have been
found to increase apparent solubility and dissolu-
tion rate, as well as maintain physical stability
with drug-small molecule ratios of 1:2, 1:1, and
2:1 (Allesø et al. 2009; Chieng et al. 2009;
Löbmann et al. 2011, 2012; Dengale et al. 2014;
Knapik et al. 2015; Teja et al. 2015), making
them a promising alternative to the conventional
polymer-based amorphous dispersion. Many of
these co-amorphous systems have been prepared
using sugars, urea, and citric acid (Lu and Zografi
1998; Ahuja et al. 2007; Masuda et al. 2012;
Ewing et al. 2015); however, much of the
research has focused on the use of drug–drug
and drug–amino acid mixtures (Löbmann et al.
2014). Preparation of these co-amorphous
mixtures use conventional manufacturing
methods, such as co-milling, shown in
Fig. 13.31, or solvent evaporation methods,
including spray drying.
Co-amorphous Drug–Drug Mixtures

First published investigations of co-amorphous
mixtures were as early as 1989, with the prepara-
tion of phenobarbital–salicin and phenobarbital–
Fig. 13.31 Preparation of co-amorphous drug–amino
antipyrine binary systems (Fukuoka et al. 1989). acid via milling. (Reproduced with permission from
Both systems showed resulting Tgs that correlated Jensen et al. 2015b)
Fig. 13.32 PXRD diffractograms of CIM–NAP co-amorphous mixtures prepared at varying ratios after 33 days of
storage in desiccators at 4 C, 25 C, and 40 C. (Reproduced with permission from Allesø et al. 2009)
studies. These studies showed that milling of showed some interaction between the carboxylic
NAP alone was not sufficient to produce amor- acid of the IND with the imine group (C¼N) of
phous NAP; however, co-milling with CIM pro- the RAN; however, these interactions were not
duced co-amorphous mixtures at 1:2, 1:1, and 2:1 strong enough to form a salt or co-crystal.
NAP–CIM. Of these, only the 1:1 ratio was found IND was also evaluated with NAP and again
to be physically stable, as shown in Fig. 13.32, showed that the presence of IND enabled the
attributed to the 1:1 molecular interaction formation of co-amorphous IND–NAP, even
between the two drug molecules. when amorphous NAP could not be prepared
In another system, also prepared by alone (Löbmann et al. 2011). It was theorized
co-milling, IND with ranitidine hydrochloride that the IND–NAP formed a heterodimer through
(RAN) formed a co-amorphous mixture, with hydrogen bonding between each drug’s carbox-
the 1:1 ratio showing a higher degree of physical ylic group, as shown in Fig. 13.33; this was later
stability after 30 days than the 1:2 and 2:1 IND– confirmed using IR and quantum mechanical
RAN mixtures on stability, although all showed calculations (Löbmann et al. 2013b). Predicted
some degree of recrystallization when held at Tg values, assuming no interaction, deviated for
40 C (Chieng et al. 2009). As with the the experimentally determined values; however,
phenobarbital–salicin and phenobarbital–antipy- when the 1:1 drug ratio was inserted as an indi-
rine co-amorphous mixtures, the IND–RAN mix- vidual component (assuming interaction) with
ture Tgs correlated well with predicted values, excess drug representing the second compound,
indicating no chemical interaction. FTIR analysis the 1:2 and 2:1 predicted Tg values matched the
Fig. 13.33 Chemical

structures and proposed
mechanism of interaction
between IND and NAP.
(Reproduced with
permission from Löbmann
et al. 2011)
Fig. 13.34 Intrinsic drug 25 NAP (co-amorphous)

release of co-amorphous IND (co-amorphous)
IND–NAP (1:1) showing
synchronized release of
both drugs. (Reproduced 20
Löbmann et al. 2011)
Release (nmol/mL)
15
10
0
0 5 10 15 20
Time (min)
experimentally determined values. The intrinsic IND–Ritonavir (RTV) co-amorphous mixtures

dissolution rate (IDR) testing of the 1:1 IND– showed good correlation with predicted Tg and no
NAP mixture showed synchronous release, evidence of bonding via FTIR analysis, indicating
shown in Fig. 13.34, with both drugs showing no heterodimer formation as seen with IND: NAP
increased drug release than their crystalline mixtures (Dengale et al. 2014). Dengale et al.
counterparts (NAP: 0.30 mg/cm2 min to showed this system to be stable for 90 days with
0.41 mg/cm2 min; IND: 0.055 mg/cm2 min to all ratios studied, 1:2, 1:1, and 2:1, and stability
0.42 mg/cm2 min). IND also showed an improve- was attributed to the miscibility of the drug–drug
ment over amorphous IND alone (0.36 mg/ mixture. Notably, even with threefold increase in
cm2 min); this finding is counter to previous the solubility of both drugs, only the RTV disso-
reports of decreased drug release in the presence lution rate was significantly enhanced from the
of another amorphous compound (Trasi, Taylor). co-amorphous dispersion.
Increased saturation solubility was also seen with In another study, ezetimib(EZB)–indapamid
talinolol (TLN) in a TLN: naringin (NRG) (IDP) co-amorphous mixtures were made at ratios
co-amorphous mixture when compared to the varying from 10:1 to 1:2 to evaluate physical
amorphous TLN alone (Teja et al. 2015). stability in the absence of any significant
molecular interaction (Knapik et al. 2015). In this and tyrosine (TYR) at 1:1 and 1:1:1 ratios
particular mixture, the EZB:IDP (10:1) was found of drug: amino acid or drug: amino acid: amino
to be stable for 72 days at room temperature when acid (Löbmann et al. 2013a). The amino acids
prepared with a low level of IDP; this was were chosen based on the binding site of the
attributed to an anti-plasticizing effect by the biological receptors of the respective drugs,
IDP on the EZB. It should be noted that the ARG and TYR for IND binding to COX-2 and
stability studies conducted with these PHE and TRP for CBZ binding to neuronal Na+
co-amorphous mixtures were done in dry channels; however, formulations were made for
conditions, so the effect of humidity on physical each drug using all four amino acids by
stability is yet to be reported. Although the exact co-milling. Calculated solubility parameters for
mechanism of formation and stabilization of these CBZ, IND, and the four amino acids were not
co-amorphous mixtures are not fully understood, significantly different with a range of 24.9 MPa
the intimate molecular mixing and interactions (CBZ) to 30.6 MPa (TRP), indicating good
between the two molecules, even if they are miscibility. Co-amorphous mixtures were formed
weak interactions, are considered barriers to pre- with IND with ARG, PHE and TRP, as well as 1:
vent recrystallization (Grohganz et al. 2013). 1:1 ratios with PHE: TRP and ARG–PHE. CBZ
was only able to form co-amorphous mixtures
Co-amorphous Drug–Amino Acid Mixtures with TRP and PHE:TRP and ARG:TRP; addi-
Although, drug–drug co-amorphous mixtures for tionally, amorphous CBZ was unable to be
FDC products may be advantageous in some prepared by milling. All of the co-amorphous
cases, co-amorphous mixtures using an inactive mixtures exhibited higher Tg values than the
small molecule offer an alternative for amorphous drug alone (CBZ was prepared by
monotherapy applications. Mura et al. showed quench cooling for this analysis), showing the
the ability of arginine (ARG) to form strong ability to improve physical stability of the amor-
interactions with NAP, produce a mostly amorphous API by increasing the mixture Tg. The
phous system upon grinding, and exhibit mixtures were shown to be physically stable for
enhanced dissolution properties (Mura et al. 6 months at room temperature, while the amor-
2003, 2005); thus, opening the door to the use phous IND and CBZ recrystallized within
of amino acids as an inactive small molecule for 1 week, further supporting the stabilizing effect
co-amorphous formulations. Amino acids, of the amino acid. The mechanism of stabilization
consisting of both an amine and carboxylic was also attributed to the molecular mixing and
group, have previously been utilized as salt- interactions seen using FTIR analysis (Löbmann
formers to improve drug solubility (Bastin et al. et al. 2013c). However, this screening showed
2000; Hirano et al. 2010). As they are organic that it was is not necessarily predictable as to
molecules, crucial to protein synthesis and other which amino acids could form co-amorphous
biological processes, they are inherently biocom- mixtures with which drug substance solely
patible (Dutta and Dey 2011). Furthermore, based on receptor binding properties. These
amino acids can facilitate enhanced and/or results are in line with results found from another
targeted drug delivery from increased permeabil- study, showing SIM and glibenclamide (GBC)
ity and cellular uptake due to amino acid co-amorphous mixture with and without a recep-
transporters along the cellular membrane, partic- tor amino acid (Laitinen et al. 2014).
ularly when those transporters are upregulated, as Jensen et al. showed the formation of the
seen with cancer cells (Tsume et al. 2011; Bhutia co-amorphous mixtures as a function of milling
et al. 2015; Gynther et al. 2008; Maeng et al. time, further demonstrating the ability of the two
2014). small molecules to facilitate conversion to amor-
Löbmann et al. screened formulations utilizing phous of molecules that cannot otherwise be read-
carbamazepine (CBZ) and IND and amino acids ily converted to the amorphous state upon milling
ARG, phenylalanine (PHE), tryptophan (TRP), (Jensen et al. 2015b). Figure 13.35 shows that
Fig. 13.35 PXRD diffractograms of milled IND, FUR, and TRP (top) and 1:1 IND–TRP and 1:1 FUR–TRP mixtures
(bottom) as a function of milling time. (Reproduced with permission from Jensen et al. 2015b)
milling of individual components, IND, furose- dried powder, also shown in Fig. 13.36, this
mide (FUR) and amino acid, TRP, was unable to seems to be advantageous for the maintenance
produce completely amorphous material within of supersaturation. Surprisingly, the PM formula-
90 min of milling; however, with the mixtures, tion showed in situ formation of IND–ARG salt
significant reduction in crystallinity was seen that led to a high degree of supersaturation; how-
after 5 min, with complete co-amorphous forma- ever, it was also quicker to recrystallize in solu-
tion by 30 min. tion, with final concentrations near that of the
Spray-drying has been demonstrated as a more crystalline IND.
scalable process to prepare co-amorphous drug– Overall, these co-amorphous mixtures provide
amino acid mixtures (Jensen et al. 2015a; Lenz promising alternative formulation techniques for
et al. 2015). Lenz et al. also evaluated the impact creating and stabilizing amorphous drug for
of tableting spray-dried co-amorphous mixtures enhanced bioavailability. Although the mecha-
on release properties and release, using an IND– nism of formation and stabilization is not
ARG co-amorphous formulation. When com- completely understood, it is presumed that the
pared to a tablet containing crystalline IND and increase in Tg, molecular mixing, and/or molecu-
physical mixture (PM) of crystalline IND and lar interactions, from hydrogen and π-π
ARG, the IND–ARG co-amorphous tablet interactions to the formation of co-crystals or
showed a longer maintenance of supersaturation, salts, contribute to the inhibition of
as shown in Fig. 13.36. Although, the release rate recrystallization.
from the tablet was slower than the neat spray-

(top) from tablets
containing spray-dried
(SD) IND–ARG, PM IND–
ARG or crystalline IND
prepared with 50 mg IND at
compaction pressures of
82 MPa and (bottom) SD
IND–ARG powder in
900 mL phosphate buffer
pH 4.5 (non-sink), 50 rpm
USP II paddle speed.
(Reproduced with
permission from Lenz et al.
2015)
processing techniques. The emerging

13.4 Summary
technologies described in this chapter offer spe-
cific advantages over traditional processing and
A number of techniques have emerged in recent
formulation techniques, which may allow certain
years that have applications in the formulation of
drugs to be formulated that would otherwise not
poorly water-soluble compounds. However,
be possible.
selection of a suitable processing or formulation
technique is highly dependent on the specific
Method Capsule 1
physicochemical properties of the drug substance
Preparation of cyclodextrin containing Solid Dis-
being studied, as well as the target release
persion: KinetiSol® Dispersing.
properties. Many drug substances exhibit
Based on the method reported by Gala
instabilities or other properties that severely
et al. (2020).
limit the number of feasible formulation and
Objective Equipment and Reagents

• To maximize solubility enhancement of • Eudragit® L100-55
abiraterone solid dispersions containing • Itraconazole
beta-cyclodextrins. • Liquid nitrogen
Equipment and Reagents • KinetiSol® Dispersing Compounder
• Kleptose® HPB (Hydroxypropyl–B– • Impact mill
cyclodextrin). • 60-mesh screen
• Abiraterone. Method
• Affinisol™ HPMCAS 126 G. • Input an ejection temperature of 158 C and
• KinetiSol® Dispersing Compounder. a rotational speed of 3000 rpm into the
• Ika Mill. control module.
Method • Mix a 1:2 blend of itraconazole:Eudragit®
• Physical mixture is made by blending L100-55 in a polyethylene bag for 1 min.
Abiraterone: HPBCD: HPMCAS 126 G • Charge the blended material into the
(10:80:10). processing chamber.
• The KinetiSol Compounder is charged with • Pre-cool steel plates with liquid nitrogen.
10 grams of the physical mixture. • Start the compounding process.
• The physical mixture is processed at 4000 • Using the data-acquisition system, monitor
RPM for 10 s followed by 5000 RPM for temperature and rotational speeds.
3.6 s to reach a temperature of 160 C. • After material is discharged, quench
• At 160 C, the composition is ejected and between chilled plates.
quenched between plates. • Grind the brittle material in an impact mill
• The brittle material is then milled to achieve and pass through a 60-mesh screen.
a particle size <250 um. Results
Results • The temperature of the blend reached
• X-ray powder diffraction indicated an 158 C in 14.1 s with exposure to
amorphous composition. temperatures greater than 100 C for only
• HPLC analysis indicated a purity of 98.1% 2 s, resulting in no chemical degradation of
with no single impurity >0.5%. Eudragit® L100-55.
• The HPBCD allowed for rapid release in • X-ray powder diffraction patterns indicated
the acidic phase, while HPMCAS allowed that the composition was amorphous.
for sustained supersaturation in the neutral • Differential scanning calorimetry
phase during a pH shift dissolution. thermograms showed the presence of a
• Compared to the generic abiraterone acetate single-phase system with no endothermic
tablet, Lot 6, processed by KinetiSol, events.
showed a 13.8-fold bioavailability for an • The plasticizer-free composition exhibited
equivalent dose in an in vivo study on a high degree of physical stability due to its
male beagle dogs. high glass transition temperature.
Method Capsule 2 Method Capsule 3

Preparation of Solid Dispersions: KinetiSol® Preparation of ritonavir containing Solid Disper-
Dispersing. sion: Powder bed fusion.
Based on the method reported by DiNunzio Based on the method reported by Davis
et al. (2010b). et al. (2020).
Objective
• To rapidly prepare plasticizer-free solid
dispersions containing Eudragit® L100-55.
Objective more basic pH (6.8) when evaluated using

• To manufacture one-step amorphous solid a pH shift dissolution study.
dispersion of ritonavir as a pharmaceutical
dosage form. Method Capsule 4
Equipment and Reagents Preparation of Solid Dispersions: Electrostatic
• Kollidon® VA64 (Polyvinylpyrrolidone- Spinning.
vinyl acetate copolymer). Based on the method reported by Verreck et al.
• Ritonavir. (2003).
• Candurin® gold sheen (potassium alumi-
Objective
num silicate-based pearlescent pigment).
• To prepare solid dispersions of itraconazole
• FujiSil® (Colloidal Silicon Dioxide).
by electrostatic spinning.
• Sintratec kit or any other similar selective
laser sintering 3D printer with a 455 nm
• Hypromellose
visible laser source.
• Itraconazole
Method
• Ethanol
• Physical mixture is made by blending rito-
• Methylene chloride
navir:Kollidon VA 64: Candurin (10:87:3)
• Electrostatic spinner
or ritonavir:Kollidon VA 64: Candurin:sili-
• Turbula mixer
con dioxide (20:76:3:1) and passed through
• Cryogenic mill
a no.170 sieve (any sieve with a mesh size
• Liquid nitrogen
smaller than 100 microns will work.)
Method
• The reservoir chamber is filled with the
• Prepare physical blends containing 20%
physical blend and the CAD file of the 3D
w/w or 40% w/w itraconazole by mixing
model for the tablets based on the dimen-
in a Turbula mixer for 10 min.
sional requirements is added to the
• Prepare a 12% w/w solution of the
software.
itraconazole: hypromellose blend in a mix-
• The dosage forms are printed at a laser
ture of ethanol and methylene chloride (40:
speed of 50–75 mm/s, surface temperature
60 ethanol: methylene chloride w/w).
of 105 C, chamber temperature of 90 C,
• Place the solution into the spinneret and
and hatch spacing of 25 μ. The powder
apply a high voltage (16–24 kV).
parameters are set to the machine’s stan-
• Optional: Mill the resulting fibers by cryo-
dard values.
genic grinding.
• The manufactured tablets are dedusted
Results
using an air gun or a brush, and the dosage
• SEM analysis demonstrated that drug con-
forms are evaluated for their dimensions,
centration and processing voltage can sig-
weight variability, and drug content.
nificantly impact fiber size and shape.
• The residual powder is collected and
recycled for the next batch.
thermograms showed that compositions
Results
containing 20% (w/w) and 40% (w/w)
• X-ray powder diffraction indicated an
itraconazole were amorphous.
amorphous composition.
• HPLC analysis indicated a purity of 98.1%
thermograms indicated that the milling pro-
with no single impurity >0.5%.
cess facilitated recrystallization of amor-
• The dosage forms show rapid disintegration
phous itraconazole.
and dissolution in the acidic medium (pH 2)
• Dissolution rates of itraconazole were
and a 21-fold increase in solubility at a
found to be highly dependent on the drug:
polymer ratio, fiber diameter, and
presentation used (nonwoven fabrics, mill- References

ing, etc.).
Aho J, Bøtker JP, Genina N, Edinger M, Arnfast L,
Method Capsule 5 Rantanen J. Roadmap to 3D-printed oral pharmaceuti-
cal dosage forms: feedstock filament properties and
Preparation of Solid Dispersions: Mesoporous characterization for fused deposition modeling. J
Silica. Pharm Sci. 2019;108(1):26–35.
Based on the method reported by Mellaerts Ahuja N, Katare OP, Singh B. Studies on dissolution
et al. (2008b). enhancement and mathematical modeling of drug
release of a poorly water-soluble drug using water-
Objective soluble carriers. Eur J Pharm Biopharm. 2007;65(1):
26–38.
• To prepare mesoporous silica loaded with
Alhijjaj M, Belton P, Qi S. An investigation into the use of
itraconazole. polymer blends to improve the printability of and reg-
Equipment and Reagents ulate drug release from pharmaceutical solid
• Itraconazole dispersions prepared via fused deposition modeling
(FDM) 3D printing. Eur J Pharm Biopharm.
• Ordered mesoporous silica (OMS)
2016;108:111–25.
• Methylene chloride Alhijjaj M, Nasereddin J, Belton P, Qi S. Impact of
• Rotary mixer processing parameters on the quality of pharmaceutical
Method solid dosage forms produced by fused deposition
modeling (FDM). Pharmaceutics. 2019;11(12):633.
• Prepare a 5 mg/mL solution of itraconazole
Allesø M, Chieng N, Rehder S, Rantanen J, Rades T,
in methylene chloride. Aaltonen J. Enhanced dissolution rate and
• Add OMS at an OMS:Itraconazole ratio of synchronized release of drugs in binary systems
75:25 and agitate for 24 h with a rotary through formulation: amorphous naproxen–cimetidine
mixtures prepared by mechanical activation. J Control
mixer.
Release. 2009;136(1):45–53.
• Remove methylene chloride by evaporation Amritha Rammohan B, Tayal L, Kumar A, Sivakumar S,
and dry powder overnight at 35 C. Sharma A. Fabrication of polymer-modified monodis-
• Heat mixture to 100 C for 5 min under perse mesoporous carbon particles by template-based
vacuum and at 40 C for 48 h to ensure approach for drug delivery. RSC Adv. 2013;3(6):
2008–16. https://doi.org/10.1039/C2RA22261B.
complete removal of methylene chloride. Andersson J, Rosenholm J, Areva S, Lindén M. Influences
Results of material characteristics on ibuprofen drug loading
• Differential scanning calorimetry and release profiles from ordered micro-and
mesoporous silica matrices. Chem Mater. 2004;16
thermograms did not exhibit glass transi-
(21):4160–7.
tion or endothermic events, indicating that Arafat B, Qinna N, Cieszynska M, Forbes RT, Alhnan
itraconazole was molecularly dispersed MA. Tailored on demand anti-coagulant dosing: an
within the OMS. in vitro and in vivo evaluation of 3D printed purpose-
designed oral dosage forms. Eur J Pharm Biopharm.
• BET analysis showed that the surface area
2018;128:282–9.
of OMS decreased from 844 to 355 m2/g Aytac Z, Ipek S, Erol I, Durgun E, Uyar T. Fast-dissolving
after loading with itraconazole. electrospun gelatin nanofibers encapsulating ciproflox-
• Dissolution studies demonstrated that acin/cyclodextrin inclusion complex. Colloids Surf B:
Biointerfaces. 2019;178:129–36.
release rates of itraconazole from OMS
Azad MA, Olawuni D, Kimbell G, Badruddoza AZM,
were significantly faster than crystalline Hossain M, Sultana T. Polymers for extrusion-based
itraconazole. 3D printing of pharmaceuticals: a holistic materials–
• Bioavailability of itraconazole-loaded process perspective. Pharmaceutics. 2020;12(2):124.
Baldoni JM, Ignatious F, Inventors. Electrospun pharma-
OMS was found to be similar to that of
ceutical compositions patent WO2001054667; 2001.
the marketed product, Sporanox®. Balogh A, Drávavölgyi G, Faragó K, Farkas A, Vigh T,
Sóti PL, et al. Plasticized drug-loaded melt electrospun
polymer mats: characterization, thermal degradation,
and release kinetics. J Pharm Sci. 2014;103(4):
1278–87. https://doi.org/10.1002/jps.23904.
Balogh A, Farkas B, Faragó K, Farkas A, Wagner I, Van Chieng N, Aaltonen J, Saville D, Rades T. Physical char-
Assche I, et al. Melt-blown and electrospun drug- acterization and stability of amorphous indomethacin
loaded polymer fiber mats for dissolution enhance- and ranitidine hydrochloride binary systems prepared
ment: a comparative study. J Pharm Sci. 2015a;104 by mechanical activation. Eur J Pharm Biopharm.
(5):1767–76. https://doi.org/10.1002/jps.24399. 2009;71(1):47–54.
Balogh A, Horváthová T, Fülöp Z, Loftsson T, Harasztos Chiou WL, Riegelman S. Pharmaceutical applications of
AH, Marosi G, et al. Electroblowing and solid dispersion systems. J Pharm Sci. 1971;60(9):
electrospinning of fibrous diclofenac sodium- 1281–302.
cyclodextrin complex-based reconstitution injection. J Coppens K, Hall M, Larsen P, Mitchell S, Nguyen P, Read
Drug Deliv Sci Technol. 2015b;26:28–34. M et al., editors. Thermal and rheological evaluation of
Bastin RJ, Bowker MJ, Slater BJ. Salt selection and pharmaceutical excipients for hot melt extrusion.
optimisation procedures for pharmaceutical new chem- AAPS annual meeting and exposition,
ical entities. Org Process Res Dev. 2000;4(5):427–35. Baltimore; 2004.
https://doi.org/10.1021/op000018u. Crowley MM, Zhang F, Koleng JJ, McGinity
Bennett RC, Brough C, Miller DA, O’Donnell KP, Keen JW. Stability of polyethylene oxide in matrix tablets
JM, Hughey JR, et al. Preparation of amorphous solid prepared by hot-melt extrusion. Biomaterials. 2002;23
dispersions by rotary evaporation and KinetiSol dis- (21):4241–8. https://doi.org/10.1016/S0142-9612(02)
persing: approaches to enhance solubility of a poorly 00187-4.
water-soluble gum extract. Drug Dev Ind Pharm. Crowley MM, Zhang F, Repka MA, Thumma S, Upadhye
2013;41(3):382–97. SB, Kumar Battu S, et al. Pharmaceutical applications
Bernardos A, Aznar E, Coll C, Martínez-Mañez R, Barat of hot-melt extrusion: part I. Drug Dev Ind Pharm.
JM, Marcos MD, et al. Controlled release of vitamin B 2007;33(9):909–26.
2 using mesoporous materials functionalized with Daghrery A, Aytac Z, Dubey N, Mei L, Schwendeman A,
amine-bearing gate-like scaffoldings. J Control Bottino MC. Electrospinning of dexamethasone/cyclo-
Release. 2008;131(3):181–9. dextrin inclusion complex polymer fibers for dental
Bhardwaj N, Kundu SC. Electrospinning: a fascinating pulp therapy. Colloids Surf B: Biointerfaces.
fiber fabrication technique. Biotechnol Adv. 2010;28 2020;191:111011.
(3):325–47. https://doi.org/10.1016/j.biotechadv.2010. Davis DA Jr, Thakkar R, Su Y, Williams RO III,
01.004. Maniruzzaman M. Selective laser sintering
Bhutia YD, Babu E, Ramachandran S, Ganapathy 3-dimensional printing as a single step process to pre-
V. Amino acid transporters in cancer and their rele- pare amorphous solid dispersion dosage forms for
vance to “glutamine addiction”: novel targets for the improved solubility and dissolution rate. J Pharm Sci.
design of a new class of anticancer drugs. Cancer Res. 2020;110(4):1432–43.
2015;75(9):1782–8. https://doi.org/10.1158/ Davis DA, Miller DA, Su Y, Williams RO. Thermally
0008-5472.can-14-3745. conductive excipient expands kinetiSol® processing
Bimbo LM, Mäkilä E, Laaksonen T, Lehto V-P, Salonen J, capabilities. AAPS PharmSciTech. 2020;21(8):1–15.
Hirvonen J, et al. Drug permeation across intestinal De Jaeghere W, De Beer T, Van Bocxlaer J, Remon JP,
epithelial cells using porous silicon nanoparticles. Vervaet C. Hot-melt extrusion of polyvinyl alcohol for
Biomaterials. 2011;32(10):2625–33. oral immediate release applications. Int J Pharm.
Bradbury MS, Pauliah M, Wiesner U. Ultrasmall fluores- 2015;492(1–2):1–9. https://doi.org/10.1016/j.ijpharm.
cent silica nanoparticles as intraoperative imaging tools 2015.07.009.
for cancer diagnosis and treatment. In: Fong Y, Deitzel J, Kleinmeyer J, Harris D, Tan NB. The effect of
Giulianotti PC, Lewis J, Groot Koerkamp B, processing variables on the morphology of electrospun
Reiner T, editors. Imaging and visualization in the nanofibers and textiles. Polymer. 2001;42(1):261–72.
modern operating room. New York: Springer; 2015. Dengale SJ, Ranjan OP, Hussen SS, Krishna BSM,
p. 167–79. Musmade PB, Gautham Shenoy G, et al. Preparation
Breitenbach J. Melt extrusion: from process to drug deliv- and characterization of co-amorphous Ritonavir–Indo-
ery technology. Eur J Pharm Biopharm. 2002;54(2): methacin systems by solvent evaporation technique:
107–17. improved dissolution behavior and physical stability
Brough C, Miller D, Keen J, Kucera S, Lubda D, Williams without evidence of intermolecular interactions. Eur J
R III. Use of polyvinyl alcohol as a solubility- Pharm Sci. 2014;62:57–64. https://doi.org/10.1016/j.
enhancing polymer for poorly water soluble drug ejps.2014.05.015.
delivery (part 1). AAPS PharmSciTech. 2015;2015: DiNunzio JC, Miller DA, Yang W, McGinity JW,
1–13. https://doi.org/10.1208/s12249-015-0458-y. Williams III RO. Amorphous compositions using con-
Capone C, Di Landro L, Inzoli F, Penco M, Sartore centration enhancing polymers for improved bioavail-
L. Thermal and mechanical degradation during poly- ability of itraconazole. Molecular pharmaceutics. 2008
mer extrusion processing. Polym Eng Sci. 2007;47 Dec 1;5(6):968–980.
(11):1813–9. DiNunzio JC, Brough C, Brown A, Williams RO,
McGinity JW. Fusion processing of itraconazole and
griseofulvin solid dipsersions by a novel high energy Fadeel B, Garcia-Bennett AE. Better safe than sorry:
manufacturing technology – KinetiSol® Dispersing. understanding the toxicological properties of inorganic
American Association of Pharmaceutical Scientists nanoparticles manufactured for biomedical
Annual Meeting; Atlanta, Georgia; 2008. applications. Adv Drug Deliv Rev. 2010;62(3):
DiNunzio JC, Brough C, Hughey JR, Miller DA, Williams 362–74. https://doi.org/10.1016/j.addr.2009.11.008.
Iii RO, McGinity JW. Fusion production of solid Follonier N, Doelker E, Cole ET. Evaluation of hot-melt
dispersions containing a heat-sensitive active ingredi- extrusion as a new technique for the production of
ent by hot melt extrusion and Kinetisol® dispersing. polymer-based pellets for sustained release capsules
Eur J Pharm Biopharm. 2010a;74(2):340–51. https:// containing high loadings of freely soluble drugs.
doi.org/10.1016/j.ejpb.2009.09.007. Drug Dev Ind Pharm. 1994;20(8):1323–39.
DiNunzio JC, Brough C, Miller DA, Williams Iii RO, Fousteris E, Tarantili PA, Karavas E, Bikiaris D. Poly
McGinity JW. Applications of KinetiSol® dispersing (vinyl pyrrolidone)–poloxamer-188 solid dispersions
for the production of plasticizer free amorphous solid prepared by hot melt extrusion. J Therm Anal Calorim.
dispersions. Eur J Pharm Sci. 2010b;40(3):179–87. 2013;113(3):1037–47.
https://doi.org/10.1016/j.ejps.2010.03.002. Fukuoka E, Makita M, Yamamura S. Glassy state of
DiNunzio JC, Brough C, Miller DA, Williams RO, pharmaceuticals. III: thermal properties and stability
McGinity JW. Fusion processing of itraconazole solid of glassy pharmaceuticals and their binary glass
dispersions by kinetisol® dispersing: a comparative systems. Chem Pharm Bull. 1989;37(4):1047–50.
study to hot melt extrusion. J Pharm Sci. 2010c;99 https://doi.org/10.1248/cpb.37.1047.
(3):1239–53. https://doi.org/10.1002/jps.21893. Gala U, Miller D, Williams RO. Improved dissolution and
DiNunzio JC, Hughey JR, Brough C, Miller DA, Williams pharmacokinetics of abiraterone through kinetisol®
RO III, McGinity JW. Production of advanced solid enabled amorphous solid dispersions. Pharmaceutics.
dispersions for enhanced bioavailability of 2020;12(4):357.
itraconazole using KinetiSol® dispersing. Drug Dev Garcia-Bennett AE, Che S, Miyasaka K, Sakamoto Y,
Ind Pharm. 2010d;36(9):1064–78. Ohsuna T, Liu Z, et al. Studies of anionic surfactant
Domokos A, Balogh A, Dénes D, Nyerges G, Ződi L, templated mesoporous structures by electron
Farkas B, et al. Continuous manufacturing of orally microscopy. In: Abdelhamid S, Mietek J, editors. Stud-
dissolving webs containing a poorly soluble drug via ies in surface science and catalysis. Elsevier; 2005.
electrospinning. Eur J Pharm Sci. 2019;130:91–9. p. 11–8.
Doreth M, Löbmann K, Priemel P, Grohganz H, Taylor R, Gencoglu MF, Spurri A, Franko M, Chen J, Hensley DK,
Holm R, et al. Influence of PVP molecular weight on Heldt CL, et al. Biocompatibility of soft-templated
the microwave assisted in situ amorphization of indo- mesoporous carbons. ACS Appl Mater Interfaces.
methacin. Eur J Pharm Biopharm. 2018;122:62–9. 2014;6(17):15068–77. https://doi.org/10.1021/
Doshi J, Reneker DH, editors. Electrospinning process and am503076u.
applications of electrospun fibers. Industry Ghebremeskel A, Vemavarapu C, Lodaya M. Use of
Applications Society Annual Meeting. Conference surfactants as plasticizers in preparing solid dispersions
Record of the 1993 IEEE; 1993. of poorly soluble API: stability testing of selected solid
Dutta P, Dey J. Drug solubilization by amino acid based dispersions. Pharm Res. 2006;23(8):1928–36. https://
polymeric nanoparticles: characterization and biocom- doi.org/10.1007/s11095-006-9034-1.
patibility studies. Int J Pharm. 2011;421(2):353–63. Gioumouxouzis CI, Karavasili C, Fatouros DG. Recent
https://doi.org/10.1016/j.ijpharm.2011.10.011. advances in pharmaceutical dosage forms and devices
Edinger M, Knopp MM, Kerdoncuff H, Rantanen J, using additive manufacturing technologies. Drug
Rades T, Löbmann K. Quantification of microwave- Discov Today. 2019;24(2):636–43.
induced amorphization of celecoxib in PVP tablets Gomes ME, Ribeiro AS, Malafaya PB, Reis RL, Cunha
using transmission Raman spectroscopy. Eur J Pharm AM. A new approach based on injection moulding to
Sci. 2018;117:62–7. produce biodegradable starch-based polymeric
El’Darov E, Mamedov F, Gol’Dberg V, Zaikov G. A scaffolds: morphology, mechanical and degradation
kinetic model of polymer degradation during extru- behaviour. Biomaterials. 2001;22(9):883–9. https://
sion. Polym Degrad Stab. 1996;51(3):271–9. doi.org/10.1016/S0142-9612(00)00211-8.
Ellenberger DJ, Miller DA, Kucera SU, Williams Gomes ME, Godinho JS, Tchalamov D, Cunha AM, Reis
RO. Generation of a weakly acidic amorphous solid RL. Alternative tissue engineering scaffolds based on
dispersion of the weak base ritonavir with equivalent starch: processing methodologies, morphology, degra-
in vitro and in vivo performance to Norvir tablet. dation and mechanical properties. Mater Sci Eng
AAPS PharmSciTech. 2018;19(5):1985–97. C. 2002;20(1–2):19–26. https://doi.org/10.1016/
Ewing AV, Biggart GD, Hale CR, Clarke GS, Kazarian S0928-4931(02)00008-5.
SG. Comparison of pharmaceutical formulations: Gopakumar T, Pagé D. Compounding of nanocomposites
ATR-FTIR spectroscopic imaging to study drug- by thermokinetic mixing. J Appl Polym Sci. 2005;96
carrier interactions. Int J Pharm. 2015;495(1):112–21. (5):1557–63.
https://doi.org/10.1016/j.ijpharm.2015.08.068.
Govender R, Abrahmsén-Alami S, Folestad S, Larsson exhibiting thermal and acidic decomposition

A. High content solid dispersions for dose window characteristics by fusion processing: a comparative
extension: a basis for design flexibility in fused depo- study of hot melt extrusion and KinetiSol® dispersing.
sition modelling. Pharm Res. 2020;37(1):1–10. AAPS PharmSciTech. 2010;11(2):760–74. https://doi.
Goyanes A, Wang J, Buanz A, Martínez-Pacheco R, org/10.1208/s12249-010-9431-y.
Telford R, Gaisford S, et al. 3D printing of medicines: Hughey JR, Keen JM, Brough C, Saeger S, McGinity
engineering novel oral devices with unique design and JW. Thermal processing of a poorly water-soluble
drug release characteristics. Mol Pharm. 2015;12(11): drug substance exhibiting a high melting point: the
4077–84. https://doi.org/10.1021/acs.molpharmaceut. utility of KinetiSol® dispersing. Int J Pharm.
5b00510. 2011;419(1–2):222–30. https://doi.org/10.1016/j.
Grohganz H, Löbmann K, Priemel P, Tarp Jensen K, ijpharm.2011.08.007.
Graeser K, Strachan C, et al. Amorphous drugs and Hughey JR, Keen JM, Miller DA, Brough C, McGinity
dosage forms. J Drug Deliv Sci Technol. 2013;23(4): JW. Preparation and characterization of fusion
403–8. https://doi.org/10.1016/S1773-2247(13) processed solid dispersions containing a viscous ther-
50057-8. mally labile polymeric carrier. Int J Pharm. 2012;438
Gynther M, Laine K, Ropponen J, Leppänen J, Mannila A, (1–2):11–9. https://doi.org/10.1016/j.ijpharm.2012.
Nevalainen T, et al. Large neutral amino acid trans- 08.032.
porter enables brain drug delivery via prodrugs. J Med Hughey JR, Keen JM, Miller DA, Kolter K, Langley N,
Chem. 2008;51(4):932–6. https://doi.org/10.1021/ McGinity JW. The use of inorganic salts to improve
jm701175d. the dissolution characteristics of tablets containing
Hanada M, Jermain SV, Thompson SA, Furuta H, Soluplus®-based solid dispersions. Eur J Pharm Sci.
Fukuda M, Williams RO III. Ternary amorphous 2013;48(4–5):758–66. https://doi.org/10.1016/j.ejps.
solid dispersions containing a high-viscosity polymer 2013.01.004.
and mesoporous silica enhance dissolution perfor- Hughey JR, Keen JM, Bennett RC, Obara S, McGinity
mance. Mol Pharm. 2020;18(1):198–213. JW. The incorporation of low-substituted
Hancock BC. Disordered drug delivery: destiny, dynamics hydroxypropyl cellulose into solid dispersion systems.
and the Deborah number. J Pharm Pharmacol. 2002;54 Drug Dev Ind Pharm. 2014;2014:1–8.
(6):737–46. Igantious F, Sun L, Inventors. Electrospun amorphous
Hancock BC, Zografi G. Characteristics and significance pharmaceutical compositions patent
of the amorphous state in pharmaceutical systems. J WO2004014304; 2004.
Pharm Sci. 1997;86(1):1–12. https://doi.org/10.1021/ Ignatious F, Sun L, Lee C-P, Baldoni J. Electrospun
js9601896. nanofibers in oral drug delivery. Pharm Res. 2010;27
Hancock BC, Shamblin SL, Zografi G. Molecular mobility (4):576–88.
of amorphous pharmaceutical solids below their glass Ilyés K, Kovács NK, Balogh A, Borbás E, Farkas B,
transition temperatures. Pharm Res. 1995;12(6): Casian T, et al. The applicability of pharmaceutical
799–806. polymeric blends for the fused deposition modelling
Healy AV, Fuenmayor E, Doran P, Geever LM, (FDM) 3D technique: material considerations–
Higginbotham CL, Lyons JG. Additive manufacturing printability–process modulation, with consecutive
of personalized pharmaceutical dosage forms via effects on in vitro release, stability and degradation.
stereolithography. Pharmaceutics. 2019;11(12):645. Eur J Pharm Sci. 2019;129:110–23.
Heikkilä T, Santos HA, Kumar N, Murzin DY, Salonen J, Inagaki S, Fukushima Y, Kuroda K. Synthesis of highly
Laaksonen T, et al. Cytotoxicity study of ordered ordered mesoporous materials from a layered
mesoporous silica MCM-41 and SBA-15 polysilicate. J Chem Soc Chem Commun. 1993;8:
microparticles on Caco-2 cells. Eur J Pharm Biopharm. 680–2. https://doi.org/10.1039/C39930000680.
2010;74(3):483–94. https://doi.org/10.1016/j.ejpb. Jacob J, Coyle N, West TG, Monkhouse DC, Surprenant
2009.12.006. HL, Jain NB. Rapid disperse dosage form containing
Hirano A, Kameda T, Arakawa T, Shiraki K. Arginine- levetiracetam. Google Patents; 2016.
assisted Solubilization system for drug substances: sol- Janssens S, Van den Mooter G. Review: physical chemis-
ubility experiment and simulation. J Phys Chem try of solid dispersions. J Pharm Pharmacol. 2009;61
B. 2010;114(42):13455–62. https://doi.org/10.1021/ (12):1571–86. https://doi.org/10.1211/jpp.61.12.0001.
jp101909a. Jensen KT, Blaabjerg LI, Lenz E, Bohr A, Grohganz H,
Hu Y, Wang J, Zhi Z, Jiang T, Wang S. Facile synthesis of Kleinebudde P, et al. Preparation and characterization
3D cubic mesoporous silica microspheres with a con- of spray-dried co-amorphous drug–amino acid salts. J
trollable pore size and their application for improved Pharm Pharmacol. 2015a; https://doi.org/10.1111/
delivery of a water-insoluble drug. J Colloid Interface jphp.12458.
Sci. 2011;363(1):410–7. https://doi.org/10.1016/j.jcis. Jensen KT, Larsen FH, Cornett C, Löbmann K,
2011.07.022. Grohganz H, Rades T. Formation mechanism of
Hughey J, DiNunzio J, Bennett R, Brough C, Miller D, coamorphous drug–amino acid mixtures. Mol Pharm.
Ma H, et al. Dissolution enhancement of a drug
2015b;12(7):2484–92. https://doi.org/10.1021/acs. Kumar A, Ganjyal GM, Jones DD, Hanna MA. Modeling
molpharmaceut.5b00295. residence time distribution in a twin-screw extruder as
Jermain SV, Miller D, Spangenberg A, Lu X, Moon C, a series of ideal steady-state flow reactors. J Food Eng.
Su Y, et al. Homogeneity of amorphous solid 2008;84(3):441–8.
dispersions–an example with KinetiSol®. Drug Dev LaFountaine J, Prasad L, Brough C, Miller D, McGinity J,
Ind Pharm. 2019;45(5):724–35. Williams R III. Thermal processing of PVP- and
Jermain SV, Lowinger MB, Ellenberger DJ, Miller DA, HPMC-based amorphous solid dispersions. AAPS
Su Y, Williams RO III. In vitro and in vivo behaviors PharmSciTech. 2015a;2015:1–13. https://doi.org/10.
of KinetiSol and spray-dried amorphous solid 1208/s12249-015-0417-7.
dispersions of a weakly basic drug and ionic polymer. LaFountaine JS, McGinity JW, Williams RO III.
Mol Pharm. 2020;17(8):2789–808. Challenges and strategies in thermal processing of
Jiang H, Wang T, Wang L, Sun C, Jiang T, Cheng G, et al. amorphous solid dispersions: a review. AAPS
Development of an amorphous mesoporous TiO2 PharmSciTech. 2015b;2015:1–13.
nanosphere as a novel carrier for poorly water-soluble Laitinen R, Löbmann K, Strachan CJ, Grohganz H, Rades
drugs: effect of different crystal forms of TiO2 carriers T. Emerging trends in the stabilization of amorphous
on drug loading and release behaviors. Microporous drugs. Int J Pharm. 2013;453(1):65–79. https://doi.org/
Mesoporous Mater. 2012;153:124–30. https://doi.org/ 10.1016/j.ijpharm.2012.04.066.
10.1016/j.micromeso.2011.12.013. Laitinen R, Löbmann K, Grohganz H, Strachan C, Rades
Keen JM, Hughey JR, Bennett RC, Jannin V, Rosiaux Y, T. Amino acids as co-amorphous excipients for simva-
Marchaud D, et al. Effect of tablet structure on con- statin and glibenclamide: physical properties and sta-
trolled release from supersaturating solid dispersions bility. Mol Pharm. 2014;11(7):2381–9. https://doi.org/
containing glyceryl behenate. Mol Pharm. 2015;12(1): 10.1021/mp500107s.
120–6. https://doi.org/10.1021/mp500480y. Lakshman JP, Cao Y, Kowalski J, Serajuddin ATM.
Kenawy E-R, Bowlin GL, Mansfield K, Layman J, Application of melt extrusion in the development of a
Simpson DG, Sanders EH, et al. Release of tetracycline physically and chemically stable high-energy amor-
hydrochloride from electrospun poly (ethylene-co- phous solid dispersion of a poorly water-soluble drug.
vinylacetate), poly (lactic acid), and a blend. J Control Mol Pharm. 2008;5(6):994–1002. https://doi.org/10.
Release. 2002;81(1):57–64. 1021/mp8001073.
Kim M-S. Soluplus-coated colloidal silica nanomatrix sys- Lenz E, Jensen KT, Blaabjerg LI, Knop K, Grohganz H,
tem for enhanced supersaturation and oral absorption Löbmann K, et al. Solid-state properties and dissolu-
of poorly water-soluble drugs. Artif Cells Nanomed tion behaviour of tablets containing co-amorphous
Biotechnol. 2013;41(6):363–7. indomethacin–arginine. Eur J Pharm Biopharm.
Kim T-W, Chung P-W, Slowing II, Tsunoda M, Yeung 2015;96:44–52. https://doi.org/10.1016/j.ejpb.2015.
ES, Lin VSY. Structurally ordered mesoporous carbon 07.011.
nanoparticles as transmembrane delivery vehicle in Leuner C, Dressman J. Improving drug solubility for oral
human cancer cells. Nano Lett. 2008;8(11):3724–7. delivery using solid dispersions. Eur J Pharm
https://doi.org/10.1021/nl801976m. Biopharm. 2000;50(1):47–60. https://doi.org/10.1016/
Kinnari P, Mäkilä E, Heikkilä T, Salonen J, Hirvonen J, S0939-6411(00)00076-X.
Santos HA. Comparison of mesoporous silicon and Liang C, Li Z, Dai S. Mesoporous carbon materials: syn-
non-ordered mesoporous silica materials as drug thesis and modification. Angew Chem Int Ed. 2008;47
carriers for itraconazole. Int J Pharm. 2011;414(1–2): (20):3696–717. https://doi.org/10.1002/anie.
148–56. https://doi.org/10.1016/j.ijpharm.2011. 200702046.
05.021. Lipinski CA. Lead-and drug-like compounds: the rule-of-
Knapik J, Wojnarowska Z, Grzybowska K, Jurkiewicz K, five revolution. Drug Discov Today Technol. 2004;1
Tajber L, Paluch M. Molecular dynamics and physical (4):337–41.
stability of coamorphous ezetimib and indapamide Lipinski CA, Lombardo F, Dominy BW, Feeney
mixtures. Mol Pharm. 2015;12(10):3610–9. https:// PJ. Experimental and computational approaches to
doi.org/10.1021/acs.molpharmaceut.5b00334. estimate solubility and permeability in drug discovery
Kresge C, Leonowicz M, Roth W, Vartuli J, Beck and development settings. Adv Drug Deliv Rev.
J. Ordered mesoporous molecular sieves synthesized 2001;46(1–3):3–26. https://doi.org/10.1016/S0169-
by a liquid-crystal template mechanism. Nature. 409X(00)00129-0.
1992;359(6397):710–2. Liu K-S, Liu H, Qi J-H, Liu Q-Y, Liu Z, Xia M, et al.
Kruk M, Jaroniec M, Ko CH, Ryoo R. Characterization of SNX-2112, an Hsp90 inhibitor, induces apoptosis and
the porous structure of SBA-15. Chem Mater. 2000;12 autophagy via degradation of Hsp90 client proteins in
(7):1961–8. human melanoma A-375 cells. Cancer Lett. 2012;318
Kulig K, David B-O, Cantrill SV, Rosen P, Rumack (2):180–8.
BH. Management of acutely poisoned patients without Löbmann K, Laitinen R, Grohganz H, Gordon KC,
gastric emptying. Ann Emerg Med. 1985;14(6):562–7. Strachan C, Rades T. Coamorphous drug systems:
https://doi.org/10.1016/S0196-0644(85)80780-0. enhanced physical stability and dissolution rate of
indomethacin and naproxen. Mol Pharm. 2011;8(5): improves the physical properties of acyclovir. Int J
1919–28. https://doi.org/10.1021/mp2002973. Pharm. 2012;422(1):160–9.
Löbmann K, Strachan C, Grohganz H, Rades T, Mekaru H, Lu J, Tamanoi F. Development of mesoporous
Korhonen O, Laitinen R. Co-amorphous simvastatin silica-based nanoparticles with controlled release capa-
and glipizide combinations show improved physical bility for cancer therapy. Adv Drug Deliv Rev. 2015;
stability without evidence of intermolecular https://doi.org/10.1016/j.addr.2015.09.009.
interactions. Eur J Pharm Biopharm. 2012;81(1): Mellaerts R, Jammaer JAG, Van Speybroeck M, Chen H,
159–69. https://doi.org/10.1016/j.ejpb.2012.02.004. Humbeeck JV, Augustijns P, et al. Physical state of
Löbmann K, Grohganz H, Laitinen R, Strachan C, Rades poorly water soluble therapeutic molecules loaded into
T. Amino acids as co-amorphous stabilizers for poorly SBA-15 ordered mesoporous silica carriers: a case
water soluble drugs – part 1: preparation, stability and study with itraconazole and ibuprofen. Langmuir.
dissolution enhancement. Eur J Pharm Biopharm. 2008a;24(16):8651–9. https://doi.org/10.1021/
2013a;85(3, Part B):873–81. https://doi.org/10.1016/j. la801161g.
ejpb.2013.03.014. Mellaerts R, Mols R, Jammaer JAG, Aerts CA, Annaert P,
Löbmann K, Laitinen R, Grohganz H, Strachan C, Van Humbeeck J, et al. Increasing the oral bioavail-
Rades T, Gordon KC. A theoretical and spectroscopic ability of the poorly water soluble drug itraconazole
study of co-amorphous naproxen and indomethacin. with ordered mesoporous silica. Eur J Pharm
Int J Pharm. 2013b;453(1):80–7. https://doi.org/10. Biopharm. 2008b;69(1):223–30. https://doi.org/10.
1016/j.ijpharm.2012.05.016. 1016/j.ejpb.2007.11.006.
Löbmann K, Laitinen R, Strachan C, Rades T, Grohganz Mellaerts R, Houthoofd K, Elen K, Chen H, Van
H. Amino acids as co-amorphous stabilizers for poorly Speybroeck M, Van Humbeeck J, et al. Aging behavior
water-soluble drugs – part 2: molecular interactions. of pharmaceutical formulations of itraconazole on
Eur J Pharm Biopharm. 2013c;85(3, Part B):882–8. SBA-15 ordered mesoporous silica carrier material.
https://doi.org/10.1016/j.ejpb.2013.03.026. Microporous Mesoporous Mater. 2010;130(1–3):
Löbmann K, Jensen KT, Laitinen R, Rades T, Strachan CJ, 154–61. https://doi.org/10.1016/j.micromeso.2009.
Grohganz H. Stabilized amorphous solid dispersions 10.026.
with small molecule excipients. In: Shah N, Sandhu H, Melocchi A, Parietti F, Maroni A, Foppoli A,
Choi DS, Chokshi H, Malick AW, editors. Amorphous Gazzaniga A, Zema L. Hot-melt extruded filaments
solid dispersions. Advances in delivery science and based on pharmaceutical grade polymers for 3D print-
technology. New York: Springer; 2014. p. 613–36. ing by fused deposition modeling. Int J Pharm.
Lu Q, Zografi G. Phase behavior of binary and ternary 2016;509(1–2):255–63.
amorphous mixtures containing indomethacin, citric Meng H, Liong M, Xia T, Li Z, Ji Z, Zink JI, et al.
acid, and PVP. Pharm Res. 1998;15(8):1202–6. Engineered design of mesoporous silica nanoparticles
https://doi.org/10.1023/A:1011983606606. to deliver doxorubicin and P-glycoprotein siRNA to
Lu J, Liong M, Sherman S, Xia T, Kovochich M, Nel A, overcome drug resistance in a cancer cell line. ACS
et al. Mesoporous silica nanoparticles for cancer ther- Nano. 2010;4(8):4539–50. https://doi.org/10.1021/
apy: energy-dependent cellular uptake and delivery of nn100690m.
paclitaxel to cancer cells. NanoBiotechnology. 2007;3 Miller DA. PhD dissertation: improved oral absorption of
(2):89–95. https://doi.org/10.1007/s12030-008- poorly water-soluble drugs by advanced solid disper-
9003-3. sion systems. Austin: The University of Texas at
Ma T-Y, Liu L, Yuan Z-Y. Direct synthesis of ordered Austin; 2007.
mesoporous carbons. Chem Soc Rev. 2013;42(9): Mizushima Y, Ikoma T, Tanaka J, Hoshi K, Ishihara T,
3977–4003. https://doi.org/10.1039/C2CS35301F. Ogawa Y, et al. Injectable porous hydroxyapatite
Maeng H-J, Kim E-S, Chough C, Joung M, Lim JW, Shim microparticles as a new carrier for protein and
C-K, et al. Addition of amino acid moieties to lapatinib lipophilic drugs. J Control Release. 2006;110(2):
increases the anti-cancer effect via amino acid 260–5. https://doi.org/10.1016/j.jconrel.2005.09.051.
transporters. Biopharm Drug Dispos. 2014;35(1): Muñoz B, Rámila A, Pérez-Pariente J, Díaz I, Vallet-Regí
60–9. https://doi.org/10.1002/bdd.1872. M. MCM-41 organic modification as drug delivery rate
Mamaeva V, Sahlgren C, Lindén M. Mesoporous silica regulator. Chem Mater. 2003;15(2):500–3. https://doi.
nanoparticles in medicine – recent advances. Adv Drug org/10.1021/cm021217q.
Deliv Rev. 2013;65(5):689–702. Mura P, Maestrelli F, Cirri M. Ternary systems of
Maniruzzaman M, Nair A, Scoutaris N, Bradley MS, naproxen with hydroxypropyl-β-cyclodextrin and
Snowden MJ, Douroumis D. One-step continuous aminoacids. Int J Pharm. 2003;260(2):293–302.
extrusion process for the manufacturing of solid https://doi.org/10.1016/S0378-5173(03)00265-5.
dispersions. International journal of pharmaceutics. Mura P, Bettinetti GP, Cirri M, Maestrelli F, Sorrenti M,
2015 Dec 30;496(1):42–51. Catenacci L. Solid-state characterization and dissolu-
Masuda T, Yoshihashi Y, Yonemochi E, Fujii K, tion properties of Naproxen–Arginine–Hydroxypro-
Uekusa H, Terada K. Cocrystallization and pyl-β-cyclodextrin ternary system. Eur J Pharm
amorphization induced by drug–excipient interaction
Biopharm. 2005;59(1):99–106. https://doi.org/10. challenges. Nanoscale. 2010;2(10):1870–83. https://

1016/j.ejpb.2004.05.005. doi.org/10.1039/C0NR00156B.
Murphy DK, Rabel S. Thermal analysis and calorimetric Rowe RC, Sheskey PJ, Cook WG, Fenton ME, American
methods for the characterization of new crystal forms. Pharmacists A. Handbook of pharmaceutical
Drugs Pharm Sci. 2008;178:279. excipients. vol Book, Whole. London/Washington,
Nagy ZK, Balogh A, Vajna B, Farkas A, Patyi G, DC/Philadelphia: Pharmaceutical Press; 2012.
Kramarics Á, et al. Comparison of electrospun and Saha D, Warren KE, Naskar AK. Controlled release of
extruded soluplus®-based solid dosage forms of antipyrine from mesoporous carbons. Microporous
improved dissolution. J Pharm Sci. 2012;101(1): Mesoporous Mater. 2014a;196:327–34. https://doi.
322–32. https://doi.org/10.1002/jps.22731. org/10.1016/j.micromeso.2014.05.024.
Nagy ZK, Balogh A, Drávavölgyi G, Ferguson J, Pataki H, Saha D, Warren KE, Naskar AK. Soft-templated
Vajna B, et al. Solvent-free melt electrospinning for mesoporous carbons as potential materials for oral
preparation of fast dissolving drug delivery system and drug delivery. Carbon. 2014b;71:47–57. https://doi.
comparison with solvent-based electrospun and melt org/10.1016/j.carbon.2014.01.005.
extruded systems. J Pharm Sci. 2013;102(2):508–17. Saha D, Moken T, Chen J, Hensley DK, Delaney K, Hunt
https://doi.org/10.1002/jps.23374. MA, et al. Micro-/mesoporous carbons for controlled
Nakamatsu J, Torres FG, Troncoso OP, Min-Lin Y, release of antipyrine and indomethacin. RSC Adv.
Boccaccini AR. Processing and characterization of 2015;5(30):23699–707. https://doi.org/10.1039/
porous structures from chitosan and starch for tissue C5RA00251F.
engineering scaffolds. Biomacromolecules. 2006;7 Sekiguchi KON. Studies on absorption of eutectic
(12):3345–55. https://doi.org/10.1021/bm0605311. mixture. I. A comparison of the behavior of eutectic
Neuvonen P, Olkkola K. Oral activated charcoal in the mixture of sulfathiazole and that of ordinary
treatment of intoxications. Dis-Manage-Health- sulfathiazole in man. Chem Pharm Bull. 1961;9(11):
Outcomes. 1988;3(1):33–58. https://doi.org/10.1007/ 866–72.
BF03259930. Serajuddin A. Solid dispersion of poorly water-soluble
Oh WK, Yoon H, Jang J. Size control of magnetic carbon drugs: early promises, subsequent problems, and recent
nanoparticles for drug delivery. Biomaterials. 2010;31 breakthroughs. J Pharm Sci. 1999;88(10):1058–66.
(6):1342–8. https://doi.org/10.1016/j.biomaterials. Shen SC, Ng WK, Chia L, Dong YC, Tan RB. Stabilized
2009.10.018. amorphous state of ibuprofen by co-spray drying with
Peng H, Dong R, Wang S, Zhang Z, Luo M, Bai C, et al. A mesoporous SBA-15 to enhance dissolution properties.
pH-responsive nano-carrier with mesoporous silica J Pharm Sci. 2010;99(4):1997–2007.
nanoparticles cores and poly(acrylic acid) shell-layers: Shen F, Yuan S, Chua CK, Zhou K. Development of
fabrication, characterization and properties for con- process efficiency maps for selective laser sintering of
trolled release of salidroside. Int J Pharm. 2013;446 polymeric composite powders: modeling and experi-
(1–2):153–9. https://doi.org/10.1016/j.ijpharm.2013. mental testing. J Mater Process Technol. 2018;254:52–
01.071. 9.
Qu F, Zhu G, Huang S, Li S, Qiu S. Effective controlled Skowyra J, Pietrzak K, Alhnan MA. Fabrication of
release of captopril by silylation of mesoporous extended-release patient-tailored prednisolone tablets
MCM-41. ChemPhysChem. 2006;7(2):400–6. via fused deposition modelling (FDM) 3D printing.
Reneker DH, Chun I. Nanometre diameter fibres of poly- Eur J Pharm Sci. 2015;68:11–7. https://doi.org/10.
mer, produced by electrospinning. Nanotechnology. 1016/j.ejps.2014.11.009.
1996;7(3):216. Slowing I, Trewyn BG, Lin VSY. Effect of surface
Repka MA, Gerding TG, Repka SL, McGinity functionalization of MCM-41-type mesoporous silica
JW. Influence of plasticizers and drugs on the nanoparticles on the endocytosis by human cancer
physical-mechanical properties of hydroxypropyl- cells. J Am Chem Soc. 2006;128(46):14792–3.
cellulose films prepared by hot melt extrusion. Drug https://doi.org/10.1021/ja0645943.
Dev Ind Pharm. 1999;25(5):625–33. Smith DM, Kapoor Y, Klinzing GR, Procopio
Repka MA, Prodduturi S, Stodghill SP. Production and AT. Pharmaceutical 3D printing: design and qualifica-
characterization of hot-melt extruded films containing tion of a single step print and fill capsule. Int J Pharm.
clotrimazole. Drug Dev Ind Pharm. 2003;29(7): 2018;544(1):21–30.
757–65. Song S-W, Hidajat K, Kawi S. Functionalized SBA-15
Roberts AD, Zhang H. Poorly water-soluble drug materials as carriers for controlled drug delivery: influ-
nanoparticles via solvent evaporation in water-soluble ence of surface properties on matrix-drug interactions.
porous polymers. Int J Pharm. 2013;447(1–2):241–50. Langmuir. 2005;21(21):9568–75.
https://doi.org/10.1016/j.ijpharm.2013.03.001. Tang Q, Xu Y, Wu D, Sun Y. A study of carboxylic-
Rosenholm JM, Sahlgren C, Linden M. Towards multi- modified mesoporous silica in controlled delivery for
functional, targeted drug delivery systems using drug famotidine. J Solid State Chem. 2006;179(5):
mesoporous silica nanoparticles – opportunities & 1513–20.
Tang Q, Chen Y, Chen J, Li J, Xu Y, Wu D, et al. Drug salicylic acid with EC using CO 2 as a temporary
delivery from hydrophobic-modified mesoporous plasticizer. Int J Pharm. 2006;327(1):45–50.
silicas: control via modification level and site-selective Vialpando M, Aerts A, Persoons J, Martens J, Van Den
modification. J Solid State Chem. 2010;183(1):76–83. Mooter G. Evaluation of ordered mesoporous silica as
Teja A, Musmade PB, Khade AB, Dengale a carrier for poorly soluble drugs: influence of pressure
SJ. Simultaneous improvement of solubility and per- on the structure and drug release. J Pharm Sci.
meability by fabricating binary glassy materials of 2011;100(8):3411–20. https://doi.org/10.1002/jps.
Talinolol with Naringin: solid state characterization, 22535.
in-vivo in-situ evaluation. Eur J Pharm Sci. 2015;78: Vialpando M, Backhuijs F, Martens JA, Van den Mooter
234–44. https://doi.org/10.1016/j.ejps.2015.08.002. G. Risk assessment of premature drug release during
Thakkar R, Pillai AR, Zhang J, Zhang Y, Kulkarni V, wet granulation of ordered mesoporous silica loaded
Maniruzzaman M. Novel on-demand 3-dimensional with poorly soluble compounds itraconazole,
(3-D) printed tablets using fill density as an effective fenofibrate, naproxen, and ibuprofen. Eur J Pharm
release-controlling tool. Polymers. 2020;12(9):1872. Biopharm. 2012;81(1):190–8. https://doi.org/10.1016/
Thakkar R, Zhang Y, Zhang J, Maniruzzaman j.ejpb.2012.01.012.
M. Synergistic application of continuous granulation Wang G, Otuonye AN, Blair EA, Denton K, Tao Z, Asefa
and selective laser sintering 3D printing for the devel- T. Functionalized mesoporous materials for adsorption
opment of pharmaceutical dosage forms with enhanced and release of different drug molecules: a comparative
dissolution rates and physical properties. study. J Solid State Chem. 2009;182(7):1649–60.
bioRxiv; 2021. Wang F, Hui H, Barnes TJ, Barnett C, Prestidge
Thomas MJK, Slipper I, Walunj A, Jain A, Favretto ME, CA. Oxidized mesoporous silicon microparticles for
Kallinteri P, et al. Inclusion of poorly soluble drugs in improved oral delivery of poorly soluble drugs. Mol
highly ordered mesoporous silica nanoparticles. Int J Pharm. 2010;7(1):227–36. https://doi.org/10.1021/
Pharm. 2010;387(1–2):272–7. https://doi.org/10.1016/ mp900221e.
j.ijpharm.2009.12.023. Wang X, Liu P, Tian Y. Ordered mesoporous carbons for
Trivedi M, Jee J, Silva S, Blomgren C, Pontinha VM, ibuprofen drug loading and release behavior. Micropo-
Dixon DL, et al. Additive manufacturing of rous Mesoporous Mater. 2011a;142(1):334–40. https://
pharmaceuticals for precision medicine applications: doi.org/10.1016/j.micromeso.2010.12.018.
a review of the promises and perils in implementation. Wang X, Liu P, Tian Y, Zang L. Novel synthesis of
Addit Manuf. 2018;23:319–28. Fe-containing mesoporous carbons and their release
Tsume Y, Hilfinger J, Amidon G. Potential of amino acid/ of ibuprofen. Microporous Mesoporous Mater.
dipeptide monoester prodrugs of floxuridine in 2011b;145(1–3):98–103. https://doi.org/10.1016/j.
facilitating enhanced delivery of active drug to interior micromeso.2011.04.033.
sites of tumors: a two-tier monolayer in vitro study. Wang Y, Zhao Q, Hu Y, Sun L, Bai L, Jiang T, et al.
Pharm Res. 2011;28(10):2575–88. https://doi.org/10. Ordered nanoporous silica as carriers for improved
1007/s11095-011-0485-7. delivery of water insoluble drugs: a comparative
Ukmar T, Planinšek O. Ordered mesoporous silicates as study between three dimensional and two dimensional
matrices for controlled release of drugs. Acta Pharma. macroporous silica. Int J Nanomedicine. 2013;8:4015–
2010;60(4):373–85. 31. https://doi.org/10.2147/IJN.S52605.
Vallet-Regi M, Rámila A, del Real RP, Pérez-Pariente J. A Wang T, Zhao P, Zhao Q, Wang B, Wang S. The mecha-
new property of MCM-41: drug delivery system. nism for increasing the oral bioavailability of poorly
Chem Mater. 2001;13(2):308–11. https://doi.org/10. water-soluble drugs using uniform mesoporous carbon
1021/cm0011559. spheres as a carrier. Drug Deliv. 2014;2014:1–9.
Van Speybroeck M, Mols R, Mellaerts R, Thi TD, Martens Wang Y, Zhao Q, Han N, Bai L, Li J, Liu J, et al.
JA, Humbeeck JV, et al. Combined use of ordered Mesoporous silica nanoparticles in drug delivery and
mesoporous silica and precipitation inhibitors for biomedical applications. Nanomedicine. 2015;11(2):
improved oral absorption of the poorly soluble weak 313–27. https://doi.org/10.1016/j.nano.2014.09.014.
base itraconazole. Eur J Pharm Biopharm. 2010;75(3): Wu C, Wang Z, Zhi Z, Jiang T, Zhang J, Wang
354–65. https://doi.org/10.1016/j.ejpb.2010.04.009. S. Development of biodegradable porous starch foam
Verreck G. The influence of plasticizers in hot-melt for improving oral delivery of poorly water soluble
extrusion. In: Hot-melt extrusion: pharmaceutical drugs. Int J Pharm. 2011;403(1–2):162–9. https://doi.
applications. London: Wiley; 2012. p. 93–112. org/10.1016/j.ijpharm.2010.09.040.
Verreck G, Chun I, Peeters J, Rosenblatt J, Brewster Xu W, Gao Q, Xu Y, Wu D, Sun Y, Shen W, et al.
ME. Preparation and characterization of nanofibers Controlled drug release from bifunctionalized
containing amorphous drug dispersions generated by mesoporous silica. J Solid State Chem. 2008;181(10):
electrostatic spinning. Pharm Res. 2003;20(5):810–7. 2837–44.
Verreck G, Decorte A, Heymans K, Adriaensen J, Liu D, Xu W, Riikonen J, Lehto V-P. Mesoporous systems for
Tomasko D, et al. Hot stage extrusion of p-amino poorly soluble drugs. Int J Pharm. 2013;453(1):
181–97. https://doi.org/10.1016/j.ijpharm.2012. properties. Int J Pharm. 2011;410(1–2):118–24.

09.008. https://doi.org/10.1016/j.ijpharm.2010.07.040.
Yachamaneni S, Yushin G, Yeon S-H, Gogotsi Y, Zhang Y, Wang J, Bai X, Jiang T, Zhang Q, Wang
Howell C, Sandeman S, et al. Mesoporous carbide- S. Mesoporous silica nanoparticles for increasing the
derived carbon for cytokine removal from blood oral bioavailability and permeation of poorly water
plasma. Biomaterials. 2010;31(18):4789–94. soluble drugs. Mol Pharm. 2012;9(3):505–13. https://
Yamamura S, Momose Y, Takahashi K, Nagatani S. Solid- doi.org/10.1021/mp200287c.
state interaction between cimetidine and naproxen. Zhang Y, Zhi Z, Li X, Gao J, Song Y. Carboxylated
Drug Stability. 1996;1:173–8. mesoporous carbon microparticles as new approach
Yamamura S, Gotoh H, Sakamoto Y, Momose to improve the oral bioavailability of poorly water-
Y. Physicochemical properties of amorphous soluble carvedilol. Int J Pharm. 2013;454(1):403–11.
precipitates of cimetidine–indomethacin binary sys- https://doi.org/10.1016/j.ijpharm.2013.07.009.
tem. Eur J Pharm Biopharm. 2000;49(3):259–65. Zhang L, Li Y, Jin Z, Chan KM, Yu JC. Mesoporous
https://doi.org/10.1016/S0939-6411(00)00060-6. carbon/CuS nanocomposites for pH-dependent drug
Yamamura S, Gotoh H, Sakamoto Y, Momose delivery and near-infrared chemo-photothermal ther-
Y. Physicochemical properties of amorphous salt of apy. RSC Adv. 2015a;5(113):93226–33. https://doi.
cimetidine and diflunisal system. Int J Pharm. org/10.1039/C5RA19458J.
2002;241(2):213–21. https://doi.org/10.1016/S0378- Zhang X, Zhang T, Ye Y, Chen H, Sun H, Zhou X, et al.
5173(02)00195-3. Phospholipid-stabilized mesoporous carbon
Yang Q, Wang S, Fan P, Wang L, Di Y, Lin K, et al. nanospheres as versatile carriers for systemic delivery
pH-responsive carrier system based on carboxylic acid of amphiphobic SNX-2112 (a Hsp90 inhibitor) with
modified mesoporous silica and polyelectrolyte for enhanced antitumor effect. Eur J Pharm Biopharm.
drug delivery. Chem Mater. 2005;17(24):5999–6003. 2015b;94:30–41. https://doi.org/10.1016/j.ejpb.2015.
Yang P, Gai S, Lin J. Functionalized mesoporous silica 04.023.
materials for controlled drug delivery. Chem Soc Rev. Zhang Y, Zhao Q, Zhu W, Zhang L, Han J, Lin Q, et al.
2012;41(9):3679–98. https://doi.org/10.1039/ Synthesis and evaluation of mesoporous carbon/lipid
C2CS15308D. bilayer nanocomposites for improved oral delivery of
Ye F, Guo H, Zhang H, He X. Polymeric micelle- the poorly water-soluble drug, nimodipine. Pharm Res.
templated synthesis of hydroxyapatite hollow 2015c;32(7):2372–83. https://doi.org/10.1007/s11095-
nanoparticles for a drug delivery system. Acta 015-1630-5.
Biomater. 2010;6(6):2212–8. Zhang J, Vo AQ, Feng X, Bandari S, Repka
Yildiz ZI, Celebioglu A, Uyar T. Polymer-free electrospun MA. Pharmaceutical additive manufacturing: a novel
nanofibers from sulfobutyl ether7-beta-cyclodextrin tool for complex and personalized drug delivery
(SBE7-β-CD) inclusion complex with sulfisoxazole: systems. AAPS PharmSciTech. 2018;19(8):3388–402.
fast-dissolving and enhanced water-solubility of Zhang J, Thakkar R, Zhang Y, Maniruzzaman
sulfisoxazole. Int J Pharm. 2017;531(2):550–8. M. Structure-function correlation and personalized
Yildiz ZI, Celebioglu A, Kilic ME, Durgun E, Uyar 3D printed tablets using a quality by design (QbD)
T. Fast-dissolving carvacrol/cyclodextrin inclusion approach. Int J Pharm. 2020;590:119945.
complex electrospun fibers with enhanced thermal sta- Zhang J, Thakkar R, Zhang Y, Maniruzzaman
bility, water solubility, and antioxidant activity. J M. Microwave induced dielectric heating for the
Mater Sci. 2018;53(23):15837–49. on-demand development of indomethacin amorphous
Yu D-G, Zhu L-M, Branford-White CJ, Yang J-H, solid dispersion tablets. J Drug Deliv Sci Technol.
Wang X, Li Y, et al. Solid dispersions in the form of 2021a;61:102109.
electrospun core-sheath nanofibers. Int J Zhang Y, Zhang J, Thakkar R, Pillai AR, Wang J, Lu A,
Nanomedicine. 2011;6:3271–80. https://doi.org/10. et al. Functions of magnetic nanoparticles in selective
2147/IJN.S27468. laser sintering (SLS) 3D printing of pharmaceutical
Zhang C, Li C, Huang S, Hou Z, Cheng Z, Yang P, et al. dosage forms; 2021b. https://doi.org/10.26434/
Self-activated luminescent and mesoporous strontium chemrxiv.13925177.v1.
hydroxyapatite nanorods for drug delivery. Zhao D, Feng J, Huo Q, Melosh N, Fredrickson GH,
Biomaterials. 2010a;31(12):3374–83. Chmelka BF, et al. Triblock copolymer syntheses of
Zhang Y, Zhi Z, Jiang T, Zhang J, Wang Z, Wang mesoporous silica with periodic 50 to 300 angstrom
S. Spherical mesoporous silica nanoparticles for load- pores. Science. 1998;279(5350):548–52.
ing and release of the poorly water-soluble drug Zhao P, Jiang H, Jiang T, Zhi Z, Wu C, Sun C, et al.
telmisartan. J Control Release. 2010b;145(3):257–63. Inclusion of celecoxib into fibrous ordered mesoporous
https://doi.org/10.1016/j.jconrel.2010.04.029. carbon for enhanced oral bioavailability and reduced
Zhang Y, Zhang J, Jiang T, Wang S. Inclusion of the gastric irritancy. Eur J Pharm Sci. 2012a;45(5):
poorly water-soluble drug simvastatin in mesocellular 639–47. https://doi.org/10.1016/j.ejps.2012.01.003.
foam nanoparticles: drug loading and release Zhao P, Wang L, Sun C, Jiang T, Zhang J, Zhang Q, et al.
Uniform mesoporous carbon as a carrier for poorly
water soluble drug and its cytotoxicity study. Eur J Zhu Y, Mehta KA, McGinity JW. Influence of plasticizer
Pharm Biopharm. 2012b;80(3):535–43. https://doi. level on the drug release from sustained release film
org/10.1016/j.ejpb.2011.12.002. coated and hot-melt extruded dosage forms. Pharm
Zhao Q, Wang T, Wang J, Zheng L, Jiang T, Cheng G, Dev Technol. 2006;11(3):285–94.
et al. Fabrication of mesoporous hydroxycarbonate Zhu S, Chen C, Chen Z, Liu X, Li Y, Shi Y, et al. Thermo-
apatite for oral delivery of poorly water-soluble drug responsive polymer-functionalized mesoporous carbon
carvedilol. J Non-Cryst Solids. 2012c;358(2):229–35. for controlled drug release. Mater Chem Phys.
https://doi.org/10.1016/j.jnoncrysol.2011.09.020. 2011;126(1–2):357–63. https://doi.org/10.1016/j.
Zhu Y, Shah NH, Malick AW, Infeld MH, McGinity matchemphys.2010.11.013.
JW. Solid-state plasticization of an acrylic polymer Zhu J, Liao L, Bian X, Kong J, Yang P, Liu
with chlorpheniramine maleate and triethyl citrate. Int B. pH-controlled delivery of doxorubicin to cancer
J Pharm. 2002;241(2):301–10. cells, based on small mesoporous carbon nanospheres.
Small. 2012;8(17):2715–20.
Scientific and Regulatory Considerations
for Development and Commercialization 14
of Poorly Water-Soluble Drugs
Zedong Dong and Hasmukh Patel
Abstract drug product development using novel phar-

This chapter focuses on the Chemistry, maceutical technologies, development of con-
Manufacturing, and Controls (CMC) from the trol strategies, etc.; (3) case studies of marketed
scientific and regulatory perspective of the drug products of poorly water-soluble drugs in
development of poorly water-soluble drugs to various dosage forms (this part uses the public
provide insights into regulatory filing from information of the approved products as
Investigational New Drug Application (IND) examples to support the discussions as outlined
to New Drug Application (NDA) submission. in part (2)); and (4) brief discussion on the
The chapter includes two primary sections to concept of Biopharmaceutics Classification
cover the two regulatory stages for CMC mod- System (BCS) in the development of poorly
ule of filing: IND and NDA. The IND section soluble drugs. The book chapter concludes
of the chapter includes the following contents: with a brief summary that emphasizes the link
(1) brief description of general filing between regulation and science.
requirements as outlined in the Code of Federal
Regulations (CFR) and relevant guidelines and Keywords
(2) discussion of potential regulatory issues for Target identification · Validation ·
developing poorly water-soluble drugs using Pharmacophore · Leads selection ·
various pharmaceutical technologies in the Investigation · Preformulation
IND stage, i.e., solid-form selection, particle- characterization · Formulation development ·
size reduction, lipid formulation, and amor- Safety · Clinical trials · Commercialization ·
phous solid dispersion. The NDA section of Drug product manufacturing ·
the chapter includes the following: (1) general Biopharmaceutics Classification System
regulatory filing requirements of an NDA (BCS) · Drug–polymer dispersion system ·
application; (2) potential regulatory issues Dissolution testing · Genotoxic impurities
associated with poorly water-soluble drugs
and detailed discussions of topics including
solid-form selection of the drug substance,
14.1 Introduction
This book chapter reflects the views of the authors and
should not be construed to represent FDA’s views or
policies. The drug-development process is a multidisci-
plinary effort, from therapeutic target identifica-
Z. Dong (*) · H. Patel tion and validation (Lowe et al. 2009), lead series
Silver Spring, MD, USA
652 Z. Dong and H. Patel
selection and structural optimization, limited due to severely compromised bioavail-

preformulation characterization (Borchardt et al. ability unless appropriate pharmaceutical
2004) and formulation development, nonclinical technologies, such as those discussed in the pre-
safety assessment, to demonstration of safety and vious chapters of this book, are utilized. With this
efficacy in clinical trials, eventual submission of a background information, this chapter focuses on
New Drug Application (NDA), and final approval the scientific and regulatory considerations for
for marketing upon review by the agency developing and commercializing poorly water-
(Guarino 2004; Rogge and Taft 2010). Similarly, soluble small molecule drugs for oral administra-
regulation of drug development and approval tion from the perspective of Chemistry,
involves expertise from many areas (chemistry, Manufacturing, and Controls (CMC).
medicine, toxicology, clinical pharmacology, sta-
tistics, etc.) to assure the safety and rights of the
subjects in all phases of an Investigational New 14.2 Investigational New Drug
Drug Application (IND) to help assure adequate Application (IND) Stage
quality of the scientific evaluation of drugs to
permit an evaluation of the drug’s safety and With the adequate preclinical characterization of
efficacy in Phases 2 and 3 clinical studies and to and data generation for a new chemical com-
ensure both the safety and efficacy of a commer- pound on its physicochemical properties, drug
cial drug product in the patients upon approval of metabolism and pharmacokinetics (DMPK), toxi-
an NDA (FDA 2017a; 21 CFR 312.22(a)). cology, and demonstration of efficacy in animal
With the advancement of science and technol- models, a sponsor may decide to bring the com-
ogy, a large number of new chemical compounds pound to further testing in human subjects
are generated in the discovery stage. Through through an IND. IND submission and mainte-
modern screening technologies, including in nance are regulated by 21 CFR Part 312. Specifi-
silico and/or high-throughput screening, the cally, the subparts that are related to the CMC
pharmacophore for the therapeutic target is portion of an IND submission are 21 CFR 312.23
identified (Florence 2009; Hou and Xu 2004; (a)(7), 312.31, and 312.33. According to the reg-
Yang 2010). With further in vitro screening/char- ulation, for each phase of the investigation, the
acterization in the late discovery/early develop- sponsor is required to submit sufficient informa-
ment stage, if satisfactory preliminary reading on tion to assure the proper identification, quality,
the safety and efficacy of the lead(s) is obtained purity, and strength of the investigational drug.
from preclinical testing in animals, the lead com- However, with the progress of the investigation
pound(s) is moved forward into clinical develop- and with more knowledge and experience gained,
ment (Salyers 2009; Rogge and Taft 2010). It is new CMC information is required to be submitted
estimated that 40–60% of these new chemical in amendments with updates on drug substance
compounds pose technical challenges to formula- and drug product to support ongoing clinical
tion scientists due to poor aqueous solubility studies as well as future NDA submission.
(Dubin 2006; Lipinski 2000; Merisko-Liversidge
and Liversidge 2008). From the literature, it
appears that the majority of new small molecule
compounds that go into clinical trials have low 14.2.1 Initial IND Submission for Phase
aqueous solubility, particularly in the therapeutic 1 Clinical Study
area of oncology. With the challenging solubility
hurdle for this type of compound, their develop- Per regulations, the initial phase 1 CMC submis-
ment into a potential drug candidate will be sion generally emphasizes the identification and
control of the raw materials, the new drug
14 Scientific and Regulatory Considerations for Development. . . 653
substance,1 and the investigational new drug acceptance criteria, and the analytical methods
product2 to ensure the safety of the subjects in used for stability studies, should be provided
the proposed clinical studies. The following is a for review.
brief summary of the documentation needed for
According to 21 CFR 312.23(a)(7)(iv)(b), the
submission. For detailed information, please refer
following information for the investigational drug
to the relevant FDA guidelines (FDA 1995).
product is required before a phase 1 study begins:
Before entering into a phase 1 clinical study, as
per 21 CFR 312.23(a)(7)(iv)(a), the following 1. “[A] list of all components, which may include
information for the new drug substance is reasonable alternatives for inactive
required to be submitted to the agency for evalu- compounds, used in the manufacture of the
ation (FDA 1995): investigational drug product, including both
those components intended to appear in the
1. “[A] description of the drug substance, includ-
drug product and those which may not appear
ing its physical, chemical, or biological
but which are used in the manufacturing pro-
characteristics.” This includes the physico-
cess.” Information on the quality of the inac-
chemical properties and preliminary character-
tive ingredients such as compendial grade
ization for the elucidation of the molecular
(USP) and GRAS (i.e., Generally Regarded
structure.
As Safe as per 21 CFR 170–199) should be
2. “[T]he name and address of its manufacturer.”
provided. For those excipients (e.g., novel
The sponsor should provide the street address
excipients) that are noncompendial or not
of the manufacturer of the drug substance for
included in the GRAS list, additional informa-
phase 1 clinical trial.
tion may be needed to demonstrate their
3. “[T]he general method of preparation of the
safety.
drug substance.” The sponsor should provide a
2. “[W]here applicable, the quantitative compo-
brief description of the manufacturing process.
sition of the investigational new drug product,
A detailed flowchart is also recommended.
including any reasonable variations that may
4. “[T]he acceptance limits and analytical
be expected during the investigational stage.”
methods used to assure the identity, strength,
A brief summary of the composition of the
quality, and purity of the drug substance.” A
investigational new drug product should be
set of scientifically acceptable specifications
submitted.
(tests, acceptance criteria, and analytical
3. “[T]he name and address of the drug product
procedures) should be provided. There is no
manufacturer.”
need to submit any validation data at this stage
4. “[A] brief, general description of the method
of investigation.
of manufacturing and packaging procedures as
5. “[I]nformation sufficient to support stability of
appropriate for the product.” Usually, a
the drug substance during the toxicological
manufacturing flow diagram is submitted
studies and the planned clinical studies.” Pre-
together with a brief description of the
liminary stability data to support the proposed
manufacturing process.
clinical study(ies), along with the tests,
5. “[T]he acceptable limits and analytical
1
methods used to assure the identity, strength,
Drug substance means an active ingredient that is
quality, and purity of the drug product.” A
intended to furnish pharmacological activity or other direct
effect in the diagnosis, cure, mitigation, treatment, or pre- brief description of the acceptance limits and
vention of disease or to affect the structure or any function their justifications and the scientifically sound
of the human body, but does not include intermediates analytical methods should be submitted.
used in the synthesis of such ingredient (21 CFR 314.3(b)).
Method validation data are not needed at this
2
Drug product means a finished dosage form, for exam-
ple, tablet, capsule, or solution that contains a drug sub-
early stage of the investigation.
stance, generally, but not necessarily, in association with
one or more other ingredients (21 CFR 314.3(b)).
6. “[A]nd information sufficient to assure available information. However, if sufficient

product’s stability during the planned clinical information and data are already generated, it is
studies.” The sponsor should provide a brief desirable to include preliminary acceptance
description of the stability study and stability criteria for these items to ensure the quality of
specifications (tests, acceptance criteria, and the test drug product. Similarly, the proposed
analytical procedures). Preliminary stability specifications for the drug product are usually
data of a representative batch in the proposed based on the information from a very limited
container closure should be provided (FDA number of batches of drug product manufactured
1995). and any short-term stability study results. Prior
knowledge of similar drug products developed
At phase I stage, preliminary information on
may be a useful source of reference. Therefore,
the physicochemical properties of a drug sub-
at this stage of the investigation, clear establish-
stance is usually available, such as pKa, logP,
ment of the acceptance criteria for certain specifi-
solubility in various solvents, pH-solubility pro-
cation items may not be realistic due to limited
file, solution stability, early reading of solid-state
experience and lack of data accumulation for the
stability, etc. Depending on the development
drug substance and the drug product. The limits or
strategies among different sponsors, certain infor-
acceptance ranges proposed for these specification
mation may not be available at this stage of
items are usually tentative, and some parameters
development, such as polymorphism/solid-form
may be monitored simply for the purpose of infor-
selection, or drug–excipient compatibility for
mation collection. In addition, during the devel-
supporting formulation development. In the case
opment, it is very likely that changes will be made
of poorly water-soluble compounds, based on the
to the drug substance (such as synthetic routes,
characterization data (chemistry and nonclinical
reagents, organic solvents, reaction conditions,
studies) from early development, a number of
batch size, crystalline form, etc.) or drug product
approaches may be tested during formulation
(formulation composition, manufacturing pro-
development for phase I studies; however, the
cess/equipment, etc.) for improvement. Under
most commonly used approaches are solid-form
these circumstances, the specifications of the
selection, particle-size reduction, lipid formula-
drug substance and drug product may be revised
tion, solubilization using cosolvent, or higher
accordingly post Phase 1 trial, provided that these
energy form (e.g., solid dispersion) of the drug
changes can be justified scientifically with
substance with the final dosage forms usually
supporting data.
suspension, tablet, or capsule.
In the early stage of clinical trials, from the
regulatory perspective, the main focus is to ensure
14.2.2 CMC Submissions After Initial
the safety of the subjects. The initial proposed
Phase 1 Study
drug substance specifications are usually based
on the limited information from the batch analysis
Any new CMC information generated after the
results and the stability data of the drug substance
initial IND submission is required to be submitted
used in the preclinical toxicology studies for the
as an information amendment as per 21 CFR part
IND submission. Generally speaking, toxicology
312.31 to ensure the identity, strength, potency,
thresholds as recommended by ICH Q3A (2006a)
quality, and purity of the drug used in the clinical
and Q3B (2006b) are reasonable approaches.
trials. As per FDA (2003) guidelines for INDs for
Known genotoxic and other noxious impurities
phase 2 and phase 3 studies, any CMC modi-
may require additional controls. It may be prema-
fications that are critical to ensure the safe use of
ture to include certain specification items in an
the test article should be submitted through infor-
initial IND, such as crystalline form, particle
mation amendments before any clinical studies
size, or surface area, due to the limited body of
start.
Regarding the manufacture of drug specification items, relaxing acceptance criteria,

substances, generally the sponsor should provide should be reported with justification and support-
more detailed information on process description, ive data. At this stage, the dissolution method
establishment of critical steps, control of (FDA 1997a) should usually be developed,
intermediates, and final drug substance (FDA finalized, and validated, as appropriate. Ideally,
2003). Changes in the manufacturing process the dissolution method should be discriminatory
that may affect safety should be highlighted. As to detect any significant changes in the critical
the drug-development program progresses, phys- product quality attributes caused by any major
ical properties linked to the performance of the manufacturing deviations. A well-designed disso-
drug product should be well characterized and lution method may help establish in vivo–in vitro
should be included in the drug substance correlation (IVIVC) if the sponsor plans for that
specifications and submitted to the IND with the route (FDA 1997b). With successful validation,
data. As more knowledge is gained and batch the dissolution method may also possibly be used
analysis, toxicology studies results, and stability as a key parameter to link between the drug prod-
data become available, the control of drug sub- uct batches used for bioequivalence (BE) studies,
stance (specifications) may be refined. The spon- clinical trials, and to-be-marketed formulations
sor should report any changes in the (FDA 2014). Similar to drug substance, a stability
specifications to the IND in an amendment. Vali- program for the drug product development should
dation data for the analytical methods that are not be submitted and any changes be reported. Stress
from the official references (e.g., United States tests on both the drug substance and drug product
Pharmacopeia (USP)) should be available and be should be carried out to investigate the potential
submitted upon request. Because of major safety degradation pathway as well as validate the ana-
concerns, new impurities should be qualified and lytical procedures (FDA 2003).
reported in the information amendment. In addi- For the development of any new drug product,
tion, stability data from an earlier phase study the regulatory pathway is similar. However, each
should be submitted. Description of the stability new drug product may face its own particular
program (and any changes thereafter) supporting technical challenges on the path to a successful
the clinical studies should be submitted, including commercialization. During the development of a
tests, acceptance criteria, analytical procedures, poorly water-soluble drug substance for oral
time points, storage conditions, and the duration administration, in order to improve the bioavail-
of the study. Generally, as the development pro- ability, it almost becomes the routine strategy for
gram moves to a later stage and larger-scale clini- the pharmaceutical industry to take one of the
cal trials, more detailed information on the drug following approaches, which have been covered
substance will be needed, including, but not lim- from the technology perspective in the previous
ited to, the manufacturing process, characteriza- chapters of this book.
tion, control of starting materials, critical steps
(a) Selection of an appropriate solid form of the
and intermediates, and the final drug substance
drug substance for formulation develop-
and its stability.
ment. This route usually involves a compre-
Regarding the manufacture of the drug prod-
hensive salt/cocrystal/polymorph screening
uct, information amendments supporting phase
(Brittain 2009; Flinn et al. 2008; Stahl and
2 and phase 3 studies should update any changes
Sutter 2006). Based on the screening results,
from the previous submission (FDA 2003). This
the solid forms that have satisfactory physi-
includes components and composition of the for-
cochemical properties (such as aqueous sol-
mulation, manufacturers, manufacturing process
ubility, hygroscopicity, stability, etc.) are
and controls, control of excipients and drug prod-
further investigated and evaluated for formu-
uct, container closure system (packaging), and
lation development and tested for improved
stability. Any changes in the drug product
bioavailability. After the final decision is
specifications, such as adding or deleting
made regarding which solid form is to be (c) Lipid formulation (Cuine 2009; Porter et al.
used for product development to support 2008; Pouton 2006). This is being used as a
clinical studies, a full characterization and broad definition in the context of covering
continuous understanding of this form will the use of any lipid, surfactant, or cosolvent
carry through the whole drug development. to solubilize the drug substance for oral for-
Properties of the solid form that are critical mulation development, for example, Self-
to the product quality, safety, and efficacy Emulsifying Drug Delivery System
should be built into the control strategies. (SEDDS) (Tang et al. 2008). In this case,
For example, if a metastable form of a drug the drug substance is completely dissolved
substance is used in a solid oral dosage form in the formulation vehicle, and the solution
rather than a stable crystalline form, it is is then filled in capsules. Due to the solution
appropriate to characterize the drug product state of the poorly soluble drug substance in
to ensure that there is no phase transition the vehicle, it is desirable to characterize the
induced during manufacturing and/or stor- formulation system, which includes, but not
age (Gift et al. 2009). The stability of the limited to, the solubility of the drug sub-
metastable polymorph in the drug product stance in the vehicle at a relevant tem-
can be assured by the long-term stability perature range, potential precipitation/
studies, and appropriate regulatory control crystallization because of supersaturation,
on the solid form should be established chemical stability, etc. Under ambient tem-
based on these results (ICH Q6A 1999). perature conditions, the formulation solution
(b) Control of the drug substance particle size. could be in a liquid, semi-solid, or solid
Due to the poor aqueous solubility of the state, depending on the composition of the
drug substance, its absorption/bioavailabil- vehicle as well as drug loading.
ity may be far from expectation to achieve (d) Amorphous solid dispersion prepared
the desired therapeutic effect. However, the through various technologies. This includes
bioavailability of a poorly soluble drug sub- most of the amorphous formulation develop-
stance may be greatly enhanced by reducing ment involving pharmaceutical technologies
the particle size to micron- or nanolevel to such as hot-melt extrusion, spray drying,
increase the surface area and therefore faster drug-carrier coprecipitation, and bead
dissolution (Hu et al. 2004; Merisko- layering (Lakshman et al. 2008; Serajuddin
Liversidge et al. 2003). If particle-size 1999). The physical stability (and possibly,
reduction is the selected route for formula- the chemical stability depending on the tech-
tion development for clinical trials and nology being used) of the drug product is the
future commercial drug product, the chal- primary concern for this approach. Again,
lenge would be to define the particle-size the desired stability profiles of the drug
range and link it to the desired exposure. In product can be supported by characterization
addition, experimental data can be collected and understanding of the amorphous system.
to demonstrate that the particle size of the It is recommended to use a scientifically
drug substance does not change significantly sound strategy for the screening of the dis-
during routine manufacturing and during persion carrier, selection of a plasticizer, and
long-term storage within the proposed retest technology for the manufacturing process.
period. For this purpose, appropriate regu- With the data collected during formulation
latory specification(s) (e.g., particle-size dis- development using this approach, the critical
tribution and specific surface area) should be process parameters and their operation
developed and established based on suffi- conditions can be finalized and justified.
cient data (ICH Q6A 1999).
The above four approaches are the most com- 14.3.1 Drug Substance
monly used strategies for the formulation devel-
opment of poorly water-soluble drugs for oral Information required for the drug substance in the
administration. These four approaches are far NDA is more detailed and comprehensive as
from a complete list of approaches used for this compared to that required for an IND (21 CFR
type of drug. Throughout the clinical studies, 314.50(d)(1)(i)). The following general informa-
sufficient controls should always be in place to tion shown below, as stated in the CFR, should be
ensure the safety of the subjects. In the late-stage submitted in the NDA.
pivotal clinical trials to support the efficacy claim
(a) “A full description of the physical and chem-
of the drug product, additional or more stringent
ical characteristics and stability.” The physi-
controls are strongly recommended to ensure the
cal and chemical characteristics include
consistent product quality as well as satisfactory
parameters for the description (appearance,
performance of the drug product. For instance, it
color, odor, etc.) of a compound, chemical
is a good idea to monitor the solid-form stability
name (including established name), Chemi-
of the drug substance in the drug product by
cal Abstracts Service (CAS) number, molec-
powder X-ray diffraction/FT-Raman and by dis-
ular formula and molecular weight,
solution when a meta-stable form is used in the
stereochemistry, solubility profile (pH--
drug product so that the exposure of the drug is
solubility profiles as well as solubility in
not reduced due to the potential phase transition,
various solvents), logP, pKa, melting point,
and therefore, the efficacy of the drug product is
polymorphism, etc. The chemical structure
not affected.
of the drug substance should be elucidated
and confirmed using appropriate
methodologies. Examples include UV spec-
14.3 New Drug Application (NDA) troscopy, FTIR, single-crystal X-ray crystal-
Stage lography, NMR (proton and 13C solid state),
mass spectrometry, and elemental analysis.
21 CFR part 314 regulates NDAs for commer- The stability characterization for the drug
cialization. Specifically, 21 CFR 314.50(d) substance is usually carried out as per ICH
(1) covers the requirements for CMC submission Q1A(R2) (2003a) and Q1B (1996). Gener-
in an NDA application. Compared with the ally, stability data for three primary batches
requirements in an IND, the CMC section in at a minimum of at least pilot scale by the
an NDA application should be a complete pack- same synthetic route and manufacturing
age including every aspect of the CMC of the procedures as the final commercial process
drug substance and drug product containing data should be submitted for the purpose of NDA
and information in sufficient detail to ensure the submission, with the drug substance tested at
identification, strength, quality, and purity of both long-term (minimum 12 months) and
both the drug substance and the drug product, accelerated storage conditions (6 months)
as well as potency and bioavailability of the (ICH Q1A(R2) 2003a). Depending on the
drug product. The following contents of this proposed storage conditions for the drug
chapter focus on the general requirements for substance, the long-term and accelerated
the CMC section of the NDA as well as a few conditions for stability testing are adjusted
examples from the currently approved drug accordingly. For drug substance intended for
products of poorly water-soluble compounds storage under ambient conditions,
for oral administration. 25 C 2 C/60% 5% RH is generally
used as the long-term storage condition and of the submission should include definition
30 C 2 C/65% 5% RH may be and control procedures of the regulatory
included as the intermediate condition in starting material; reagents, solvents, and
case any significant change occurs within auxiliary materials controls; description of
6 months under accelerated conditions the synthesis by flowchart and written state-
(40 C 2 C/75% 5% RH). If the drug ment of steps; and purification. For a drug
substance is proposed to be stored in a substance produced by chemical synthesis,
refrigerator, the long-term and accelerated both ICH Q7 (2000) and Q11 (2012) state
storage conditions are 5 C 3 C and that a starting material should be a signifi-
25 C 2 C/60% 5%, respectively. For cant structural fragment of the drug sub-
a drug substance proposed to be stored in a stance, with defined chemical properties
freezer, 20 C 5 C should be used as and structure. The starting materials should
the long-term conditions. Under this situa- have adequate controls, which include tests,
tion, ICH Q1A(R2) (2003a) does not specify analytical procedures, and acceptance
the accelerated storage condition; however, criteria. Similarly, the reagents, solvents,
it recommends testing a single batch of drug and auxiliary materials should have
substance at an elevated temperature (e.g., specifications. The description of the syn-
5 C 3 C or 25 C 2 C) to address the thetic process should include the details,
short-term temperature excursions outside such as typical equipment, solvents, catalyst
the proposed long-term storage condition and reagents, reaction conditions, in-process
under 20 C 5 C. The evaluation and tests for reaction completion, isolation and
analysis of the stability data generated under purification procedures, etc. The general
both long-term and accelerated conditions to principles for the selection of starting
establish the retest period of the drug sub- materials for synthetic drug substances and
stance usually follow ICH Q1E (2003b). the scientific justification are described in
Depending on the stability of the drug sub- ICH Q11 (2012) and FDA guidelines (2012
stance under the storage conditions tested, and 2018).
different scenarios for predicting its retest (d) “[T]he process controls used during manu-
period are used, which are summarized in facture and packaging.” With the experience
the decision tree (Fig. 14.1) in ICH Q1E and knowledge gained during the IND stage,
(2003b). the applicant should establish in-process
(b) “[T]he name and address of its controls to monitor the critical reaction
manufacturers.” The applicant should pro- steps and intermediates. These controls are
vide the FDA Establishment Identifier generally developed and justified to ensure
(FEI) number and the Data Universal Num- the conditions and completion of the reac-
bering System (DUNS) number tion, to confirm identification and purity of
(if available) (FDA 2019), the address of the isolated intermediate(s), or to determine
the drug substance manufacturer(s), applica- a critical physicochemical parameter. The
ble Drug Master File (DMF) number tests, acceptance criteria, and analytical
(if any), and contact and telephone number. procedures should be submitted with justifi-
The address should be that of the cation for the acceptance criteria.
manufacturing, packaging, and control (e) “[T]he specifications necessary to ensure the
(release as well as stability testing) facilities, identity, strength, quality, and purity of the
including contract facility(ies) used for this drug substance and the bioavailability of the
purpose. drug products made from the drug sub-
(c) “[T]he method of synthesis (or isolation) and stance, including, for example, tests, analyt-
purification of the drug substance.” This part ical procedures, and acceptance criteria
Fig. 14.1 Decision tree for stability data evaluation for establishing retest period of a drug substance or expiry of a drug
product (frozen products not included)
relating to stability, sterility, particle size, stability testing to support the retest period
and crystalline form.” The applicant should (also referred to as shelf life), detailed
provide specifications for the release and description of analytical procedures, and
their validation reports where needed (FDA proposed and justified based on batch analy-
2000). ICH Q6A (1999) provides guidelines sis data, qualification through toxicological
for establishing and justifying specifications studies, and drug products used in pivotal
for a new drug substance. Briefly, the fol- clinical studies.
lowing items are generally included in the
drug substance specifications: description,
identification, assay, and impurities.
14.3.2 Drug Product
Description is usually a qualitative statement
on the color, odor, and state of the drug
Similar to drug substance, the information
substance. Identification tests should be spe-
required for the drug product in the NDA is
cific to the drug substance so that the drug
described in 21 CFR 314.50(d)(1)(ii), which is a
substance can be differentiated from struc-
complete compilation of information and data
turally similar and related compounds. ICH
collected for the to-be-marketed drug product,
Q6A (1999) recommends two identification
more detailed and comprehensive as compared
tests with different principles or a combina-
to that required for an IND. The following infor-
tion of tests into one procedure. The assay
mation is required to be submitted in the NDA.
test is used to determine the content of a drug
substance, with the specification typically (a) “A list of all components used in the manu-
set within 98–102%, or otherwise justified facture of the drug product (regardless of
with sufficient supporting data. The impurity whether they appear in the drug product)
profile of a drug substance includes organic and a statement of the composition of the
(starting materials, intermediates, drug product.” All the components used in
by-products, and degradants), inorganic the manufacturing of the drug product
(elemental impurities, catalysts), and resid- should be included and their grades be
ual solvents. ICH Q3A(R2) (2006a) clearly stated. In addition, the quantitative
provides detailed information for reporting, composition of a unit dose of the drug prod-
identification, and qualification of organic uct should be provided by the weight of the
impurities in the drug substance. Also, ICH drug substance and the excipients (usually,
Q3C(R6) (2016) provides detailed informa- the percentage is also provided). The repre-
tion on reporting and control strategies for sentative batch formulation for each dose
residual solvents. Official compendial ana- strength should be provided in the
lytical procedures (such as USP) may be submission.
adopted if applicable. If nonofficial test (b) “[T]he specifications for each component.”
methods are used, which may be the most The specifications for the excipients, includ-
likely case for a new drug substance for ing those removed during the manufacturing
assay and organic impurities testing, the ana- process, should be provided in the submis-
lytical procedures should be validated sion. For novel excipients that are used for
against specified criteria, as described in the first time in the drug product, full details
the FDA (2000) and ICH Q2(R1) (2005) of their manufacture, characterization, and
guidelines. As certain physicochemical controls, with cross-references to supporting
properties of the drug substance may signifi- safety data (nonclinical and clinical), should
cantly affect the bioavailability of the drug be provided (FDA 2001). This information
product, particularly for poorly water- may be included in the NDA submission or
soluble drugs, specification for items such in a DMF (FDA 1989).
as crystal form, particle size distribution, or (c) “[T]he name and address of each manufac-
surface area may be needed (ICH Q6A turer of the drug product.” The name and
1999). The acceptance limits should be address (including building number if appli-
cable) of all the facilities (including contract
manufacturing/packaging/testing facilities) impurities, which are similar to that for the

that are involved in the manufacturing, pack- drug substance. Also similar to the drug
aging, testing (release and stability) of the substance, ICH Q3B(R2) (2006a) covers
finished drug product should be provided. the rationale for reporting organic impurities
The FEI and DUNS (if available) numbers (e.g., degradation products) in the drug prod-
for the manufacturing facilities should be uct. The specifications for degradation
provided (FDA 2019). The name and products are usually established based on
address of the suppliers of the novel or criti- the available stability data, the proposed
cal excipients may be needed in the shelf life, and storage conditions, and the
application. daily dose of the drug product. However,
(d) “[A] description of the manufacturing and the acceptance limits should always be
packaging procedures and in-process within the levels qualified in toxicology
controls for the drug product.” A copy of studies. ICH Q1D(R1) (2019) and FDA
the proposed or actual master production (2020) recommend that an elemental impu-
batch record for the manufacture of a com- rity risk assessment be conducted on a new
mercial lot of drug product or a comparably finished drug product and establish appropri-
detailed description of the production pro- ate control as necessary. Additional
cess for a representative batch of the drug references for the limits of elemental
product is required (21 CFR 314.50(d)(1)(ii) impurities and analytical procedures can be
(c)). A flowchart and a step-by-step descrip- made to USP General Chapters <232> and
tion of the manufacturing process should be <233>, respectively (USP 43-NF 38 2020).
provided in the submission. In both For a potential genotoxic impurity with no
situations, the equipment, operating available carcinogenicity data, ICH M7
conditions, and sampling points for (R1) (2017) recommends a maximum daily
in-process controls are usually included. intake of 1.5 μg for long-term (>10 years)
In-process controls are usually established treatment. FDA (2021a) published
and justified based on appropriate data gen- guidelines on the control of nitrosamine
eration and analysis during development and impurities in human drugs, which outlines
process validation. The tests, acceptance the regulatory recommendations for drug
criteria/ranges, and the description of the substances and/or drug products that have
in-process analytical method should be potential contamination concerns with the
provided for the in-process controls for com- nitrosamine class of impurities.
mercial drug product manufacture. In addition to the universal tests, additional
(e) “[T]he specifications necessary to ensure the specific tests may also be established
identity, strength, quality, purity, potency, depending on the dosage form and adminis-
and bioavailability of the drug product, tration route. For the solid oral dosage forms
including, for example, tests, analytical (e.g., tablets and capsules), additional spe-
procedures, and acceptance criteria relating cific tests may include dissolution, disinte-
to sterility, dissolution rate, container clo- gration, hardness, content uniformity, water
sure systems.” Specifications should be content, microbial limits, etc. (ICH Q6A
established to ensure the batch-to-batch 1999). Similarly, for the liquid oral formula-
quality and hence the safety and efficacy of tion (e.g., solution and suspension), addi-
the drug product. ICH Q6A (1999) provides tional specific tests may include uniformity
detailed guidelines on the establishment of of dosage unit, pH, microbial limits, assay of
specifications. The universal tests that are preservative, dissolution, particle-size distri-
generally applicable to all drug products bution, redispersibility, etc. (ICH Q6A
are description, identification, assay, and 1999). The dissolution method development
report is usually provided in the NDA appli- accelerated storage condition is

cation, which supports and justifies the dis- 40 C 2 C/75% 5% RH. It is strongly
solution conditions selected. Dissolution recommended that aqueous-based products
specification is usually established based on packaged in semi-permeable containers be
the dissolution characteristics of the batches evaluated for potential water loss during stor-
used in pivotal clinical trials and/or in con- age (ICH Q1A(R2) 2003a). The stability data
firmatory bioavailability studies (FDA of the drug product from both long-term and
1997a, b). Depending on the solubility and accelerated storage conditions are evaluated to
rate of dissolution, one-, two-, or multi-point predict the shelf-life of the drug product (ICH
dissolution test specification may be Q1E 2003b), as summarized in Fig. 14.1.
established using USP Type 1 or Type
Reference to the current edition of the US
2 Apparatus. Standard tests, such as disinte-
Pharmacopeia and National Formulary may sat-
gration, content uniformity, and microbial
isfy relevant requirements for drug substance
limits, can be referenced to the respective
(21 CFR 314.50(d)(1)(i)) and drug product
official USP methods and acceptance criteria
(21 CFR 314.50(d)(1)(ii)(a))
unless modification is necessary as justified
The above information on the general filing
scientifically (USP 43-NF 38 2020).
requirements of the CMC section in an NDA
If an analytical method for the drug product is submission is far from complete since there are
not referenced to official sources (e.g., USP), the additional filing requirements such as environ-
submission is required to include a description of mental assessment (21 CFR 314.50(d)(1)(iii)),
the analytical procedures and method validation CMC relevant labeling information (package
report. insert, carton and immediate container labels),
and container closure information, which are not
(f) “Stability data with proposed expiration dat-
discussed in this chapter. The brief summary of
ing.” ICH Q1A(R2) (2003a) recommends sta-
the NDA filing is based on the authors’ under-
bility data from at least three primary batches
standing of the current FDA regulations and
of the drug product, with two batches mini-
available guidelines to the pharmaceutical
mally at pilot scale and the third batch maybe
scientists. If detailed guidelines are needed for
smaller with sufficient justification. These pri-
certain specific aspects of an NDA application,
mary batches should be representative of the
it is strongly recommended to refer to the relevant
commercial manufacturing process. The sta-
ICH (2021) guidelines and/or FDA (2021b, c)
bility study should be conducted in the com-
guidelines.
mercial container closure system under the
proposed long-term and accelerated storage
conditions. Similar to the drug substance,
ICH Q1A(R2) (2003a) recommends that sta- 14.3.3 Case Studies of Approved Drug
bility data of at least 12 months under long- Products for the Poorly Soluble
term and 6 months under accelerated storage Drugs
conditions be submitted at NDA filing. The
long-term and accelerated storage conditions The formulation development and commerciali-
used for stability studies also depend on the zation of poorly water-soluble compounds for
proposed storage conditions of the drug prod- oral delivery involves various approaches and
uct as well as the container closure system. In strategies to improve their in vivo exposure, as
most cases, where the drug product is pro- discussed in Sect. 14.2.2. Poorly water-soluble
posed to be stored under ambient conditions, drugs that have been approved and
the long-term storage condition is commercialized have their own characteristics
25 C 2 C/60% 5% RH or and technical difficulties arising from the poor
30 C 2 C/65% 5% RH, and the solubility, which have been successfully
addressed. However, they all may have common should be explored and investigated to avoid
challenges from the regulatory perspective. The any future surprises (Bauer et al. 2001; Dunitz
following are a few examples of the approved and Bernstein 1995; ICH Q6A 1999). Thermody-
drug products. namic characterization is usually carried out on
the solid crystalline forms that can be produced in
Solid-Form Selection: Carbamazepine bulk and are relatively stable under ambient
Carbamazepine ([298-46-4], 5 H-dibenz[b,f] conditions, as in the case of carbamazepine, the
azepine-5-carboxamide, Fig. 14.2) is currently anhydrous polymorphs, and the dihydrate. Ide-
approved for the indication of epilepsy and tri- ally, the phase transition kinetics from the meta-
geminal neuralgia. It is practically insoluble in stable to the stable form can also be characterized
water but soluble in organic solvents such as using the pure phases with the parameters such as
alcohol, acetone, and propylene glycol (O’Neil temperature, water activity (relative humidity),
2006). Four anhydrous polymorphs (Forms I– particle size, etc. In addition, if the metastable
IV) and a dihydrate of carbamazepine have been form is chosen for formulation development, the
reported and investigated widely (Lang et al. manufacturing process (e.g., wet granulation and
2002; Rustichelli et al. 2000). It has been shown tablet compaction) that may induce phase change
that the anhydrous polymorphs convert to the requires characterization as well to ensure mini-
dihydrate in an aqueous medium (Lehto et al. mal or absence of phase conversion to the unde-
2009; Tian et al. 2006). In vivo pharmacokinetic sired solid form (Gift et al. 2009). Thereafter, the
data of carbamazepine in dogs (Kobayashi et al. stability of the selected crystal form can be further
2000) have shown that the solid form is a critical monitored and confirmed during the drug product
factor for bioavailability. The commercial carba- stability studies in the proposed commercial
mazepine drug substance is anhydrous Form III packaging configuration by testing against the
(or the β form) as specified in the USP mono- established specifications (e.g., dissolution) and
graph. However, due to the phase conversion of examining for any potential trend of change. Fail-
Form III to the dihydrate, bioinequivalence was ure to meet these specifications or detection of a
observed in the 200 mg carbamazepine tablets significant trend of change due to phase transition
almost two decades ago (Meyer et al. 1992). would alert the formulation scientists to modify
Therefore, from a scientific and regulatory per- the formulation composition, manufacturing pro-
spective (ICH Q6A 1999), it is desirable to con- cess, or the container closure system.
trol the phase purity and establish specifications ICH Q6A (1999) decision tree#4 (Fig. 14.3)
for both the drug substance (e.g., powder X-ray provides the criteria for the establishment of a
diffraction, particle size, FTIR) and the drug specification for polymorphism in drug
product (e.g., dissolution test, powder X-ray dif- substances and drug products. Briefly, if the
fraction) to ensure the in vivo performance of the polymorphs of a drug substance have different
drug product. physicochemical properties that affect the safety
For a new drug substance, the potential poly- and efficacy of the drug, it is recommended to
morphism (including hydrates and solvates) establish specification on the polymorphic form.
If adequate control and test (e.g., dissolution)
have been established to detect any changes in
the content of the specified polymorphic form in
the drug product, typically no further action is
needed. If a phase conversion occurs, which
affects the safety and efficacy of the drug product,
and there is no established control in place to
monitor the change, then a specification for the
polymorphic form should be established.
Fig. 14.2 Molecular structure of carbamazepine
Drug Substance
Can NO
Conduct polymorphism
1. screen on drug substance.
different polymorphs No further action
be formed?
YES
Characterize the forms:

e.g., - X-ray Power Diffraction
- DSC / Thermoanalysis GO TO 2.
- Microscopy
- Spectroscopy
2.
Do the
forms have NO
different properties?
(solubility, stability,
melting point)
No further test or
YES acceptance criterion
for drug substance
Is drug
product safety, NO
performance or
efficacy affected?
YES
Sol acceptance criterion

for polymorph content GO TO 3.
in drug substance
Fig. 14.3 ICH Q6A Decision Tree#4 for setting up specifications on polymorphism
Does
drug product
performance testing YES Establish acceptance criteria
provide adequate control if for the relevant performance
polymorph ratio changes test(s).
(e.g., dissolution)?
NO
Monitor polymorph form during

stability of drug product.
Does a
change occur NO No need to set acceptance criteria
which could affect for polymorph change in drug
safety or efficacy? product.
YES
Establish acceptance criteria

which are consistent with
safety and/or efficacy.
Particle-Size Reduction: Fenofibrate Fenofibrate is currently marketed or approved

and Sirolimus under several brand names in tablets (Triglide,
Fenofibrate ([49562-28-9], isopropyl 2-[p-( p- Fenoglide, and Tricor) and capsules (Antara and
chlorobenzoyl)phenoxy]-2-methylpropanoate, Lipofen), and also as generic drug products
Fig. 14.4) is approved for the adjunctive therapy (Orange Book 2021).
to diet for hypercholesterolemia and hypertrigly- Sirolimus ([53123-88-9], also known as
ceridemia. It is practically insoluble in water, rapamycin
slightly soluble in methanol and ethanol, soluble (3S,6R,7E,9R,10R,12R,14S,15E,17E,19E,21-
in acetone, ether, benzene, and chloroform, and S,23S,26R,27R,34aS)
very soluble in methylene chloride (O’Neil 2006). 9,10,12,13,14,21,22,23,24,25,26,27,32,33,34,
OH
H
H OCH3
H
O
N CH3
O O O
O
OH
H3C
O H3C
H
O CH3 OH
H3C
Cl O CH3 O
O CH3 H3CO H3CO
H3C CH3
CH3
O CH3
Fenofibrate Sirolimus
Fig. 14.4 Molecular structures of fenofibrate (left) and sirolimus (right)
34a-hexadecahydro-9,27-dihydroxy-3-[(1R)-2- particle size of less than 2000 nm. Similarly, US

[(1S,3R,4R)-4-hydroxy-3-methoxycyclohexyl]-1- Patent 5,145,684 (Liversidge et al. 1991) is
methylethyl]-10,21-dimethoxy-6,8,12,14,20,26- identified to cover the nanotechnology for
hexamethyl-23,27-epoxy-3H-pyrido[2,1-c][1,4] Rapamune® tablets, where an average effective
oxaazacyclohentriacontine-1,5,11,28,29 particle size of less than 400 nm is claimed and
(4H,6H,31H )-pentone, Fig. 14.4), is a triene preparation method, grinding media, and disper-
macrolide antibiotic isolated from streptomyces sion media are included.
hygroscopicus, and it is currently approved for As recommended by ICH Q6A (1999), ade-
the indication of prophylaxis of organ rejection in quate control on particle-size distribution of the
patients aged 13 years receiving renal drug substance should be established for the drug
transplants. The currently approved drug products product developed via particle-size reduction
are Rapamune® oral solution (1 mg/mL) and when particle size is a critical factor for bioavail-
tablets (0.5, 1, and 2 mg) (Orange Book 2021). ability, and particle-size distribution should be
Sirolimus is insoluble in water but freely soluble included as part of the drug substance
in benzyl alcohol, chloroform, acetone, and ace- specifications. From a scientific and regulatory
tonitrile (O’Neil 2006). perspective (ICH Q1A(R2) 2003a), when particle
The literature (Bawa 2009) indicates that the size reduction is carried out in the drug substance,
Tricor® tablets (fenofibric acid eq. 45 and 135 mg through either dry milling (e.g., jet mill) or wet
strengths) and Rapamune® tablets (sirolimus 0.5, milling, sufficient stability data are desirable to
1, and 2 mg strengths) utilized nanotechnology demonstrate that there is no significant change in
during formulation development for improved the particle size of the milled drug substance
bioavailability. Orange Book (2011) listed three throughout the proposed retest period (as a final
patents relating to nanotechnology for the Tricor® drug substance) or storage period (as an interme-
tablets: US Patent 6,375,986 (Ryde and Ruddy diate during drug product manufacturing for a
2000), US Patent 7,276,249 (Ryde et al. 2003a), wet-milled suspension). In addition, it is desirable
and US Patent 7,320,802 (Ryde et al. 2003b). to have proper controls (e.g., dissolution) in place
Both patents 7,276,249 and 7,320,802 cover the for the final drug product to monitor any potential
nanoparticle technology of fenofibrate, which is change in the particle size of the drug substance
stated to eliminate the food effect upon oral throughout the proposed expiry so that the safety
administration and claims for an effective average
No drug substance particle

Is the drug product a solid NO size acceptance criterion
dosage form or liquid
required for solution dosage
containing undissolved
forms.
drug substance?
YES
1. Is the particle size critical to dissolution,

solubility, or bioavailability?
2. Is the particle size critical to drug product
processability? If NO to all
3. Is the particle size critical to drug product stability?
4. Is the particle size critical to drug product
content uniformity?
5. Is particle size critical for maintaining
product appaerance?
If YES to any
No Acceptance Criterion
Required
Set Acceptance Criterion
Fig. 14.5 ICH Q6A Decision Tree#3 for setting up specifications on drug substance particle-size distribution
and efficacy of the drug product are not affected dosage forms (e.g., tablet and capsule) as well as
(ICH Q6A 1999). suspension for oral administration.
ICH Q6A (1999) decision tree#3 (Fig. 14.5)
provides guidelines on setting acceptance criteria
Lipid Formulation: Ritonavir and Sirolimus
for drug substance particle-size distribution. As
Ritonavir ([155213-67-5], 5S-
per the decision tree, a specification for particle
(5R*,8R*,10R*,11R*))10-hydroxy-2-methyl-5-
size distribution is necessary for essentially all
(1-methylethyl)-1-(2-(1-methylethyl)-4-
poorly water-soluble drugs and for all solid
thiazolyl)-3,6-dioxo-8,11-bis(phenylmethyl)-
Fig. 14.6 Molecular structure of ritonavir
2,4,7,12-tetraazatridecan-13-oic acid, indicated in the Orange Book (2011), the formu-

5-thiazolylmethyl ester, (Fig. 14.6) is a human lation composition appears to be claimed in US
immunodeficiency virus (HIV) protease inhibitor Patent 6,232,333 (Lipari et al. 1997).
indicated in combination with other antiretroviral As mentioned previously, in addition to the
agents for the treatment of HIV infection. Crys- tablet dosage form, sirolimus is also marketed as
talline ritonavir is practically insoluble in aqueous Rapamune® Oral Solution (1 mg/mL). The cur-
media with a solubility of 400 μg/mL in 0.1 N rently approved prescribing information for
HCl and 1 μg/mL in pH 6.8 phosphate buffer at Rapamune® Oral Solution (2020) indicates that
37 C (Law et al. 2004). As per the Orange Book the drug product contains the following inactive
(2021), the currently approved innovator drug ingredients: Phosal® 50 PG (phosphatidylcholine,
products containing ritonavir are Norvir® Oral propylene glycol, mono- and di-glycerides, etha-
Solution (80 mg/mL), Norvir® Capsule nol, soy fatty acids, and ascorbyl palmitate) and
(100 mg), Norvir® Tablet (100 mg), Norvir® polysorbate 80. It appears that sirolimus is
Powder (100 mg/packet), Kaletra® Oral Solution solubilized as a solution in the lipid for oral
(80 mg/mL lopinavir and 20 mg/mL ritonavir), administration, which is covered by US Patent
Kaletra® Capsule (133.3 mg lopinavir and 5,536,729 (Waranis and Leonard 1994; Orange
33.3 mg ritonavir), and Kaletra® Tablet (200 mg Book 2021).
lopinavir and 50 mg ritonavir; 100 mg lopinavir The scientific and regulatory perspective of the
and 25 mg ritonavir). A well-known case is the lipid type of formulation may involve
currently approved Norvir® 100 mg strength cap- establishing appropriate specifications to ensure
sule, which has to be reformulated due to the both physical and chemical stability of the drug
precipitation of ritonavir polymorph II in the orig- product. Physical stability concerns such as crys-
inally approved soft gelatin capsule (Chemburkar tallization of the drug substance from drug prod-
et al. 2000). uct solution may be addressed by the inspection
The currently approved labeling for Norvir® of appearance of the oral solution or the contents
(2020) indicates that the reformulated soft gelatin in the capsule. The phase transformation of the
capsule has the following inactive ingredients: drug substance and lipid excipients may be moni-
butylated hydroxytoluene, ethanol, gelatin, iron tored by dissolution testing. Concerns on the
oxide, oleic acid, polyoxyl 35 castor oil, and chemical stability of the drug product can be
titanium dioxide. Based on the formulation addressed by establishing appropriate stability
components, it appears that the drug product specifications of the drug product. In addition to
uses an SEDDS formulation with oleic acid for the concern related to the drug substance stability
the oil phase, polyoxyl 35 castor oil as an emulsi- in the formulation, lipid excipients may also have
fier, and butylated hydroxytoluene as an antioxi- potential stability issues, such as peroxidation and
dant to prevent the peroxidation of oleic acid. As hydrolysis (Canon 2008; Radomska-Soukharev
and Mueller 2006; Radomska-Soukharev 2007). comprising a solid dispersion comprising” (a) a
To prevent peroxidation of certain lipid defined group of compounds of interest “and
excipients, antioxidant(s) can be used in the for- (b) one or more pharmaceutically acceptable
mulation. In that case, in the authors’ opinion, it water-soluble polymers.” The manufacturing pro-
may be desirable to establish a specification for cesses, as covered in the patent, include spray
the antioxidant content. In addition, depending on drying, film casting, or hot melt extrusion of
the type of lipid used, it may be desirable to store etravirine with polymers. As described in the
the drug product with protection from light. Addi- currently approved labeling (Intelence® Tablets
tional specifications may also be desirable to con- Full Prescribing Information 2019), hypromellose
trol the hydrolysis of some lipid excipients, based is the only water-soluble polymer used in the
on the investigation results of the individual case. formulation; therefore, it is the primary polymeric
The necessity of these controls is a scientific and carrier in the solid dispersion.
regulatory decision based on available literature Ritonavir 100 mg tablet (Norvir® Tablet) is
information and/or research conducted by the another example of solid dispersion formulation
applicant. of a poorly soluble drug, which was approved in
2010 in addition to the previously approved oral
Solid Dispersion: Etravirine and Ritonavir solution and softgel capsule formulations. Orange
Etravirine ([269055-15-4], 4-[[6-amino-5-bromo- Book (2011) listed US Patent 7,364,752 (Fort
2-[(4-cyanophenyl)amino]-4-pyrimidinyl]oxy]- et al. 2000) for the tablet formulation, which
3,5-dimethylbenzonitrile, Fig. 14.7) is a second- covers the hot-melt extrusion process to produce
generation non-nucleoside reverse transcriptase the solid dispersion of ritonavir in a water-soluble
inhibitor indicated for HIV infection. It is practi- polymer.
cally insoluble in water with less than 1 μg/mL These approved drug products demonstrate
solubility in both water and 0.1 N HCl with a logP solid dispersion to be a successful approach for
>5 (Weuts et al. 2011) and appears to be the formulation development of poorly water-
lipophilic. Intelence® tablets with three strengths, soluble drugs. The dispersion may be produced
25 mg, 100 mg, and 200 mg, are currently by hot-melt extrusion, which is used in the manu-
approved (Orange Book 2021). Orange Book facture of Norvir® tablets, or by alternative
(2011) also listed US Patent 7,887,845 (Verreck technologies such as spraying, drying, beads
and Baert 2006) for the Intelence tablet formula- layering, etc. In most cases, the technical chal-
tion, which claims various processes for lenge for the solid dispersion approach is the
manufacturing the solid dispersion of etravirine selection of suitable excipient(s) as well as appro-
with water-soluble polymers. Specifically, Claim priate drug loading for the dispersion system to
1 of Patent 7,887,845 covers “a particle provide satisfactory physical stability to the drug
product. To tackle this challenge, the drug–poly-
NH2 mer dispersion system should be thoroughly
N
characterized and understood. In the authors’
opinion, to help ensure consistent quality of the
NC NH Br drug product, it would be desirable that the drug
N product specifications include a specification to
O
control the crystallinity of the drug substance in
the finished drug product, such as powder X-ray
diffraction. To serve the same purpose, a dissolu-
tion test may also be desirable as an indicator of
any change in the crystallinity of the drug prod-
CN
uct. Due to the plasticizing effect of water in an
amorphous system, which facilitates the crystalli-
Fig. 14.7 Molecular structure of etravirine zation of the amorphous drug substance in the
product (Tong and Zografi 2004), the water con- form in the development stage, preformulation
tent should also be tightly controlled and the characterization of the drug substance provides
acceptance range be justified based on stability more reliable and more accurate aqueous solubil-
data (ICH Q6A 1999). ity results. Therefore, with the solubility informa-
tion generated from preformulation studies and
the permeability/absorption data derived from
14.3.4 Concept of Biopharmaceutics human pharmacokinetic studies, in vivo animal
Classification System (BCS) models, or in vitro methods, it is relatively simple
in Developing Poorly to determine whether a compound is highly solu-
Water-Soluble Drugs ble and highly permeable as defined in the FDA
(2017b) guidelines. The reality is, however, in
The concept of BCS classification was introduced many cases that the drug substances selected for
by Amidon et al. (1995) in order to understand the clinical development are very poorly water solu-
importance of solubility and permeability on drug ble. Regardless of their BCS classifications (2 or
absorption. Four BCS categories are defined 4), if the pharmaceutical technologies described
based on the solubility and permeability of earlier are used for bioavailability enhancement
a drug: case 1 with high solubility and high per- and final commercial drug product marketing,
meability, case 2 with low solubility and high regulatory approaches similar to those described
permeability, case 3 with high solubility and low in the examples may be taken.
permeability, and case 4 with low solubility and
low permeability. The original article provides
the theoretical basis for BCS classification using 14.4 Concluding Remarks
a transport model and analysis of human perme-
ability results for a number of representative Discovery and development of new drug
drugs. products, as well as regulatory assessment of
FDA (2017b) issued guidelines on the waiver their safety and efficacy before approval for mar-
of in vivo bioavailability and bioequivalence keting, have always been based on the advance-
studies for immediate-release solid oral dosage ment of science and technology. From the very
forms based on BCS classification, which defines first steps in drug discovery with disease target
the two terms high solubility and high permeabil- identification and validation, identification of the
ity from a regulatory perspective. For an pharmacophore and leads selection, to the later
immediate-release (IR) drug product, if the stages in pharmacology/toxicology investigation,
highest dose strength is soluble in the aqueous preformulation characterization and formulation
media of 250 mL or less in the pH range of 1–7.5, development, and safety and efficacy demonstra-
the drug substance is considered highly soluble. tion in clinical trials, the success of a new drug
High permeability is defined by the extent of product is the outcome of the close collaboration
absorption of a drug determined in humans at among the scientists from various backgrounds
85% or higher. As pointed out in the guidelines, and expertise. Regulatory assessment of a new
the extent of absorption is the fraction of dose drug product prior to commercialization is carried
absorbed rather than systemic bioavailability. out with the scientific criteria outlined in the FDA
The initial characterization of the aqueous sol- and ICH guidelines, which are accessible to the
ubility and permeability of a new chemical com- public and serve as useful resources to the indus-
pound is generally carried out in the discovery try. As for the formulation development and com-
stage, where preliminary information is generated mercialization of poorly water-soluble
to provide an early reading on the BCS classifica- compounds, if a new technology is used in the
tion. With the availability of the compound of drug product manufacturing, the finished drug
better purity as well as final selection on solid product should meet the CFR requirements on
identity, strength, quality, purity, potency, and FDA. Guidance for industry extended release oral dos-
bioavailability, as supported by the underlying age forms: development, evaluation, and application of
in vitro/in vivo correlation. Rockville: FDA; 1997b.
science. FDA. Guidance for industry analytical procedures and
methods validation. Rockville: FDA; 2000.
Acknowledgments The authors wish to thank FDA. Guidance for industry M4Q: the CTD – quality.
Dr. Richard Lostritto, Dr. Christine Moore, and Rockville: FDA; 2001.
Dr. Stephen Moore for critically reviewing the manuscript FDA. Guidance for industry INDs for phase 2 and phase
and insightful discussions during the preparation of this 3 studies chemistry, manufacturing, and controls infor-
book chapter for the first edition and Dr. Ramesh mation. Rockville: FDA; 2003.
Raghavachari for the valuable comments during the revi- FDA. Guidance for industry – Q11 development and man-
sion of this book chapter for the third edition. ufacture of drug substances. Rockville: FDA; 2012.
FDA. Guidance for industry bioavailability and bio-
equivalence studies submitted in NDAs or INDs –
general considerations (Draft Guidance). Rockville:
FDA; 2014.
References FDA. The FDA’s drug review process: ensuring drugs are
safe and effective. 2017a. http://www.fda.gov/Drugs/
Amidon GL, Lennernaes H, Shah VP, et al. A theoretical ResourcesForYou/Consumers/ucm143534.htm.
basis for a biopharmaceutic drug classification: the Accessed 10 Feb 2021.
correlation of in vitro drug product dissolution and FDA. Guidance for industry waiver of in vivo bioavail-
in vivo bioavailability. Pharm Res. 1995;12:413–20. ability and bioequivalence studies for immediate-
Bauer J, Spanton S, Henry R, et al. Ritonavir: an extraor- release solid oral dosage forms based on a
dinary example of conformational polymorphism. biopharmaceutics classification system. Rockville:
Pharm Res. 2001;18:859–66. FDA; 2017b.
Bawa R. Nanopharmaceuticals for drug delivery – a FDA. Guidance for Industry – Q11 development and
review. Drug Deliv. 2009;3:122–7. manufacture of drug substances (chemical entities and
Borchardt RT, Kerns EH, Lipinski CA, et al., editors. biotechnological/biological entities): questions and
Pharmaceutical profiling in drug discovery for lead answers. Rockville: FDA; 2018.
selection. Arlington: American Association of Pharma- FDA. Guidance for industry identification of
ceutical Scientists; 2004. manufacturing establishments in applications submit-
Brittain HG, editor. Polymorphism in pharmaceutical ted to CBER and CDER: questions and answers.
solids. 2nd ed. New York: Informa Healthcare; 2009. Rockville: FDA; 2019.
Canon JB. Chemical and physical stability considerations FDA. Guidance for industry – Q3D(R1) elemental
for lipid-based drug formulations. Am Pharm Rev. impurities. Rockville: FDA; 2020.
2008;11(132):134–8. FDA. Guidance for industry – control of nitrosamine
Chemburkar SR, Bauer J, Deming K, et al. Dealing with impurities in human drugs. Rockville: FDA; 2021a.
the impact of ritonavir polymorphs on the late stages of FDA. Guidances. 2021b. https://www.fda.gov/regulatory-
bulk drug process development. Org Process Res Dev. information/search-fda-guidance-documents.
2000;4:413–7. Accessed 4 Mar 2021.
Cuine JF. Lipid-based oral drug delivery systems to FDA. Manual of policies and procedures (CDER). 2021c.
enhance solubility and absorption of poorly water- https://www.fda.gov/about-fda/center-drug-evalua
soluble drugs. Am Pharm Rev. 2009;12:74–83. tion-and-research-cder/cder-manual-policies-
Dubin CH. Formulation strategies for poorly soluble procedures-mapp. Accessed 4 Mar 2021.
drugs. Drug Del Technol. 2006;6:34–8. Flinn T, Northen J, Fernandes P. New drug development:
Dunitz JD, Bernstein J. Disappearing polymorphs. getting to the optimal physical form. Pharm Chem.
Acounts Chem Res. 1995;28:193–200. 2008;7:20–3.
FDA. Guideline for drug master files. 1989. https://www. Florence AJ. Approaches to high-throughput physical
fda.gov/drugs/drug-master-files-dmfs/guideline-drug- form screening and discovery. Drugs Pharm Sci.
master-files-dmf. Accessed 11 Feb 2021. 2009;192:139–84.
FDA. Guidance for industry content and format of Fort JJ, Krill SL, Law D, et al. Solid dispersion pharma-
investigational new drug applications (INDs) for ceutical formulations. US Patent 7,364,752. 10 Nov
Phase 1 studies of drugs, including well-characterized, 2011; 2000.
therapeutic, biotechnology-derived products. Gift AD, Luner PE, Luedeman L, et al. Manipulating
Rockville: FDA; 1995. hydrate formation during high shear wet granulation
FDA. Guidance for industry dissolution testing of using polymeric excipients. J Pharm Sci. 2009;98:
immediate release solid oral dosage forms. Rockville: 4670–83.
FDA; 1997a. Guarino RA, editor. New drug approval process. 4th
ed. Marcel Dekker: New York; 2004.
Hou T, Xu X. Recent development and application of Lipari J, Al-Razzak LA, Ghosh S, et al. Pharmaceutical
virtual screening in drug discovery: an overview. composition. US Patent 6,232,333, 7 Nov 1997; 1997.
Curr Pharm Design. 2004;10:1011–33. Lipinski CA. Drug-like properties and the causes of poor
Hu J, Johnston KP, Williams RO III. Nanoparticle engi- solubility and poor permeability. J Pharmacol Toxicol
neering processes for enhancing the dissolution rates of Methods. 2000;44:235–49.
poorly water soluble drugs. Drug Dev Ind Pharm. Liversidge GG, Cundy KC, Bishop JF, et al. Surface
2004;30:233–45. modified drug nanoparticles. US Patent 5,145,684,
ICH. Stability testing: photostability testing of new drug 25 Jan 1991; 1991.
substances and products Q1B. Geneva: ICH; 1996. Lowe JA, Jones P, Wilson DM. The importance of target
ICH. Specifications: test procedures and acceptance validation in drug discovery and development. Curr
criteria for new drug substances and new drug Opin Drug Discov Devel. 2009;12:581–4.
products: chemical substances Q6A. Geneva: Merisko-Liversidge EM, Liversidge GG. Drug
ICH; 1999. nanoparticles: formulating poorly water-soluble
ICH. Good manufacturing practice guide for active phar- compounds. Toxicol Pathol. 2008;36:43–8.
maceutical ingredients Q7. Geneva: ICH; 2000. Merisko-Liversidge E, Liversidge GG, Cooper
ICH. Stability testing of new drug substances and products ER. Nanosizing: a formulation approach for poorly-
Q1A(R2). Geneva: ICH; 2003a. water-soluble compounds. Eur J Pharm Sci. 2003;18:
ICH. Evaluation for stability data Q1E. Geneva: ICH; 113–20.
2003b. Meyer MC, Straughn AB, Jarvi EJ, et al. The
ICH. Validation of analytical procedures: text and meth- bioinequivalence of carbamazepine tablets with a his-
odology Q2(R1). Geneva: ICH; 2005. tory of clinical failures. Pharm Res. 1992;9:1612–6.
ICH. Impurities in new drug substances Q3A(R2). Norvir® Soft Gelatin Capsule prescribing information.
Geneva: ICH; 2006a. 2020. https://www.accessdata.fda.gov/drugsatfda_
ICH. Impurities in new drug products Q3B(R2). Geneva: docs/label/2020/020945s048lbl.pdf. Accessed 15
ICH; 2006b. December 2021.
ICH. Development and manufacture of drug substances O’Neil MJ, editor. The Merck index: an encyclopedia of
(chemical entities and biotechnological/biological chemicals, drugs, and biologicals. 13th ed. Whitehouse
entities) Q11. Geneva: ICH; 2012. Station: Merck; 2006.
ICH. Impurities: guideline for residual solvents Q3C(R6). Orange Book: approved drug products with therapeutic
Geneva: ICH; 2016. equivalence evaluations. Silver Spring: U.S. Food
ICH. Assessment and control of DNA reactive (muta- and Drug Administration; 2011 and 2021. http://
genic) impurities in pharmaceuticals to limit potential www.accessdata.fda.gov/scripts/cder/ob/default.cfm.
carcinogenic risk M7(R1). Geneva: ICH; 2017. Accessed 31 Mar 2011 and 11 Feb 2021.
ICH. Guideline for elemental impurities Q3D(R1). Porter CJH, Wasan KM, Constantinides P. Lipid-based
Geneva: ICH; 2019. systems for the enhanced delivery of poorly water
ICH. Quality guidelines. 2021. https://www.ich.org/page/ soluble drugs. Adv Drug Deliv Rev. 2008;60:615–6.
quality-guidelines. Accessed 4 Mar 2021. Pouton CW. Formulation of poorly water-soluble drugs
Intelence® (etravirine) Tablets full prescribing informa- for oral administration: physicochemical and physio-
tion. 2019. https://www.janssenlabels.com/package- logical issues and the lipid formulation classification
insert/product-monograph/prescribing-information/ system. Eur J Pharm Sci. 2006;29:278–87.
INTELENCE-pi.pdf. Accessed 29 Mar 2021. Radomska-Soukharev A. Stability of lipid excipients in
Kobayashi Y, Ito S, Itai S, et al. Physicochemical solid lipid nanoparticles. Adv Drug Deliv Rev.
properties and bioavailability of carbamazepine 2007;59:411–8.
polymorphs and dihydrate. Int J Pharm. 2000;193: Radomska-Soukharev A, Mueller RH. Chemical stability
137–46. of lipid excipients in SLN-production of test
Lakshman JP, Cao Y, Kowalski J, et al. Application of formulations, characterization and short-term stability.
melt extrusion in the development of a physically and Pharmazie. 2006;61:425–30.
chemically stable high-energy amorphous solid disper- Rapamune® Oral Solution prescribing information. 2020.
sion of a poorly water-soluble drug. Mol Pharm. https://www.accessdata.fda.gov/drugsatfda_docs/
2008;5:994–1002. label/2020/021083s067,021110s085lbl.pdf. Accessed
Lang M, Kampf JW, Matzger AJ. Form IV of carbamaze- 29 Mar 2021.
pine. J Pharm Sci. 2002;91:1186–90. Rogge MC, Taft DR, editors. Preclinical drug develop-
Law D, Schmitt EA, Marsh KC, et al. Ritonavir-PEG 8000 ment. 2nd ed. Taylor & Francis: Boca Raton; 2010.
amorphous solid dispersions: in vitro and in vivo eval- Rustichelli C, Gamberini G, Ferioli V, et al. Solid-state
uation. J Pharm Sci. 2004;93:563–70. study of polymorphic drugs: carbamazepine. J Pharm
Lehto P, Aaltonen J, Tenho M, et al. Solvent-mediated Biomed Anal. 2000;23:41–54.
solid phase transformations of carbamazepine: effects Ryde NP, Ruddy SB. Solid dose nanoparticulate
of simulated intestinal fluid and fasted state simulated compositions comprising a synergistic combination of
intestinal fluid. J Pharm Sci. 2009;98:985–96. a polymeric surface stabilizer and dioctyl sodium
sulfosuccinate. US Patent 6,375,986, 21 Sept the dihydrate in aqueous suspension using Raman
2000; 2000. spectroscopy. J Pharm Biomed Anal. 2006;40:271–80.
Ryde T, Gustow EE, Ruddy SB, et al. Nanoparticulate Tong P, Zografi G. Effects of water vapor absorption on
fibrate formulations. US Patent 7,276,249, 23 May the physical and chemical stability of amorphous
2003; 2003a. sodium indomethacin. AAPS PharmSciTech. 2004;5:
Ryde T, Gustow EE, Ruddy SB, et al. Methods of treat- 2.
ment using nanoparticulate fenofibrate compositions. USP 43-NF 38. United States pharmacopeia 43/national
US Patent 7,320,802, 27 Oct 2003; 2003b. formulary 38. Rockville: United States Pharmacopeial
Salyer KL. Preclinical pharmacokinetic models for drug Convention; 2020.
discovery and development. Drugs Pharm Sci. Verreck G, Baert L. Antiviral compositions. US Patent
2009;186:659–73. 7,887,845, 3 Feb 2006; 2006.
Serajuddin ATM. Solid dispersion of poorly water-soluble Waranis, RP, Leonard TW. Rapamycin formulations for
drugs: early promises, subsequent problems, and recent oral administration. US Patent 5,536,729, 9 Sept
breakthroughs. J Pharm Sci. 1999;88:1058–66. 1994; 1994.
Stahl PH, Sutter B. Salt selection. In: Hilfiker R, editor. Weuts I, Van Dycke F, Voorspoels J, et al. Physicochemi-
Polymorphism: in the pharmaceutical industry. cal properties of the amorphous drug, cast films, and
Weinheim: Wiley-VCH; 2006. spray dried powders to predict formulation probability
Tang B, Cheng G, Gu JC, et al. Development of solid self- of success for solid dispersions: etravirine. J Pharm
emulsifying drug delivery systems: preparation Sci. 2011;100:260–74.
techniques and dosage forms. Drug Discov Today. Yang SY. Pharmacophore modeling and application in
2008;13:606–12. drug discovery: challenges and recent advances. Drug
Tian F, Zeitler JA, Strachan CJ, et al. Characterizing the Discov Today. 2010;15:444–50.
conversion kinetics of carbamazepien polymorphs to
Index
A Aldehydes, 202
AbbVie, 359 Aluminum salts, 460
Acetyl-11-keto-b-boswellic acid (AKBA), 603 Alzheimer’s disease, 197
Active loading method, 234 Amino acids, 636
Active pharmaceutical ingredients (API), 33, 611 Aminobenzoic acid, 557
API–excipient incompatibility, 50 Amorphous cyclosporine nanodispersions, 23, 24
development, 142 Amorphous forms, 111
dispersion, 147 Amorphous material, 476
experimental aqueous solubility determination, 35–36 Amorphous ritonavir, 186
explosive dust, 155 Amorphous solid dispersions (ASD), 5, 404, 600, 601,
hardness and friability, 148 603, 605–607, 609, 614, 622, 632
intrinsic dissolution rate, 39 analytical characterization, 408, 409
production, 142 carrier selection, 405
solubility prediction using machine characterization of
and deep learning, 37–39 differential scanning calorimetry, 314
stability testing, potential formulations, 69 IR and Raman spectroscopy, 313
substantial particle-size reduction, 60 microscopic technique, 314
thermal decomposition, 50 X-ray powder diffraction (XRPD), 313
wet milling, 156 clinical considerations, 414, 415
with pH-dependent solubility, 74 combined ASD screening approach, 305
Acute hypersensitivity reactions, 10 copovidone-based ASDs, 311
Additive manufacturing (AM) coprecipitation, 310
compositions and polymeric components, 612 dissolution method for, 316, 317
definition, 611 drug molecules, 430
drug release doravirine (MK-1439), 431
multilayered tablets, 615 elbasvir, 435
tablet-in-tablet, 615 etravirine (TMC125), 431, 432
FDM, 612, 613 everolimus (EVE), 431
fusion-(heat/laser)based processes, 612 grazoprevir, 434
material jetting, 611 ivacaftor (VX-770), 433, 434
PBF, 614, 616, 617 ledipasvir (LED), 430
sintering process, ASD formation, 616 telaprevir, 432, 433
3D printed tablet, 611 torcetrapib, 434
3D printing, 611 tucatinib, 435
Administration routes velpatasvir (VEL), 430
IA, 220 drug–polymer miscibility, 405
IM, 219 final dosage form, 415
IT, 219, 220 bottle formulation, 417
IV, 218, 219 bulk spray-dried dispersion, 416
SC, 219 bulk spray-dried powder, 415
Aerodynamic diameter, 142 capsule formulation, 417
Aerosol solvent extraction system (ASES), 533 dry granulation, 416
Affinisol™, 346 tablet formulation, 418
Albendazole solubilization, 241, 243 wet granulation, 417
# American Association of Pharmaceutical Scientists 2022 675

Pharmaceutical Sciences Series 50, https://doi.org/10.1007/978-3-030-88719-3
676 Index
Amorphous solid dispersions (ASD) (cont.) spray drying, 306

first-generation ASDs, 288 stability prediction, 318
fourth-generation ASDs, 288 stability screening, 408, 410, 411
glass transition temperature, 297, 298 states, 199
hot-melt extrusion formulation and process structured development approach, 288, 291, 320
parameter, 312 structured development, 288
hot-melt extrusion, 308, 309 supersaturation and sustainment, 199
hydrogen bonding and/or ionic interactions, 299 supersaturation screening, 302
hygroscopicity and water activity, 300 temperature-composition phase diagram, 298
initial assessment third-generation ASDs, 288
of bioavailability, 292 Amorphous solids, 104, 112
of manufacturability, 293 Amorphous spray-dried dispersion (ASDD), 398
of stability, 292 Amorphous systems, 44, 45, 55, 71
in vitro dissolution testing, 407, 409 Amoxicillin, 542
in vitro screening excipient system, 406, 407 Amphiphilic lipids, 233
manufacturing technology, 300 Anaphylactic shock, 8
miniaturized methods for, 301 Anionic polymers, 606, 607
molecular arrangement in, 315, 316 Anticancer agents, 166
oral formulations of, 359–362 Antimicrobial effectiveness testing (AET), 224
pharmaceutical literature, 419 Antioxidant/chelating agent, 239
Albendazole with HPMCP and HPMC, 427–429 Antioxidants, 240
AMG 517 with HPMCAS-MF Antisolvent precipitation (AP) process
and HPMC E5, 426, 427 atomization, 565
B-Raf Inhibitor (G-F) with HPMCAS-M, 419 CP process, 575, 576, 585, 586
compounds with HPMCAS, 422, 423, 425 driving mechanism, 565
copovidone, 425 FN process, 570–575, 585
cyclosporine A (CsA) with nonpolymeric high-velocity jets, 567
carrier, 429 limits of processing parameters, 566
MK-0364 with HPMCAS and Copovidone, mixing energies, 567
421, 422 MSMPR model, 566
Protease Inhibitor with Eudragit L100-55, organic drug solution, 566
424, 426 P127-stabilized-Itz particles, 566
Simvastatin with Povidone, 423, 425 particle formation process, 566
Sirolimus with HPMC-TPGS, 420 poorly water-soluble drug, 564
surfactant, 425 process and formulation parameters, 565
Telmisartan with Povidone K30 and a pH process parameters, 566
Modifier, 420, 421 scalability, 563
traconazole (ITZ), 425 stabilizers, 564
preclinical considerations, 411 temperature, pressure and solvents, 567
drug loading, 413 Approximate dielectric requirement (ADR), 181
formulation technology, 414 Aqueous solubility, 1–3, 8, 11, 21
solid dispersions, 414 Aqueous suspension, 35–37, 85
suspension stability, 412 Artificial intelligences, 37
toxicology formulation, 412 Ascorbic acid–ethanol–CO2 system, 543
XRPD spectra, 413 Asthma, 17
polymer and drug loading, 288 ATIVAN® injection, 228
polymer screening, 293, 294 Atmospheric freeze drying (ATMFD), 462
polymeric formulation matrix, 198 Atomic force microscopy (AFM), 59, 60, 300, 350, 484
quality by design (QbD) approach, 288 and conventional light microscopy, 59
second-generation ASDs, 288 designs, 59
single-phase system, 198 instrument designs, 59
solid-state screening, 303, 304 and mDSC, 60
solid-state stability, 291 mechanical properties, 59
solubility parameters, 297 ATRIGEL® Delivery System, 236
solubility ratio, 199 Attenuated total reflectance (ATR), 52
solubility, 405 Attenuated total reflectance (ATR)-FTIR
solvent-based spray drying, 199 spectroscopy, 54, 194
solvent evaporation method, 404 Avogadro’s number, 531
solvent selection, 404–405 Azodicarbonamide, 164
Index 677
B Chlorofluorocarbons (CFCs), 18
Ball media mills, 147 Cholesteryl ester transfer protein (CETP) inhibitor, 362
Basic homogenization principles, 164 cimetidine (CIM), 633
Bead mill DISPERMAT, 162 CIM–indomethacin (IND) mixture, 633
Belsomra®, 362 CIM–NAP co-amorphous mixtures, 633, 634
Bile dilution-induced drug supersaturation, 273 Circular dichroism, 194
Bile-mediated supersaturation, 275 CLEVIPREX®, 231
Binary HPBCD ASD, 605 Clofazimine-loaded nanoparticles, 578
Binding/complexation stability, 189 Closed stirred media mills, 158
Bioavailability, 119, 141 Co-amorphous mixtures
Bioavailability enhancement of poorly ASDs, 632
water-soluble drugs CIM–NAP, 634, 635
emerging technologies, 600 definition, 633
formulation-based techniques (see Formulation-based drug–amino acid, 633, 636, 637
techniques, poorly water-soluble drugs) drug–drug, 633
process-based techniques (see Process-based FUR–TRP, 637
techniques, poorly water-soluble drugs) IND–NAP, 634, 635
Bioavailability-related outcomes, 142 IND–TRP, 637
Biocompatible stabilizers, 23 polymer-based solid dispersions, 633
Biodegradable poly (DL-lactide-co-glycolide) preparation, 633
(PLGH/PLG), 236 Coarse suspensions, 231
Biodegradable polymeric microspheres, 231 Coaxial electrospinning, 619
Biopharmaceutics classification system (BCS), 3, 4 Coconut oil, 256
Biopharmaceutics Drug Disposition Classification Cocrystal approach, 130
System (BDDCS), 5, 6 Cocrystal formation, 128
Bio-relevant dissolution media, 76 Cocrystallization mechanism, 128
Biphasic dissolution test, 78, 274 Collisional transport, 272
Blood–brain barrier, 16 Colloidal SiO2, 208
Bowman’s membrane, 12 Combination technology (CT), 167
Budesonide nanosuspension, 24 Combinational approach, 239
Buffer capacity, 16 Combinative particle-size reduction techniques
Buffered systems, 221, 222 CT, 167
Bulk density, 16 levels, 168
NANOEDGE technology, 165, 166
C smartCrystal technology, 166–167
Caffeine (CAF), 613 Commercial Itz solid dispersion (Sporanox), 581
Carbamazepine (CBZ), 17, 488, 489, 513–515, 539, Common-ion effect, 227
562, 564, 609, 663 Compatibility studies, 226
Carboxylated mesoporous carbon microparticles Complexation efficiency (CE), 190
(c-MCM), 630 Complexation techniques
Carboxymethyl chitosan (CMCS)-conjugated coprecipitation, 193
poloxamer P407 gels, 237 dynamic equilibrium, 191
Cardiac arrest, 8 extrusion, 193
Carfentanil, 197 freeze drying, 193
Carvedilol (CAR), 630 kneading, 193
Cavitation, 163 mechanical grinding, 193
CD applications solvent evaporation, 193
nasal drug delivery, 197 spray drying, 193
ophthalmic drug delivery, 197 Compressed fluids, 532, 540, 558
pulmonary drug delivery, 197, 198 Computer-aided design (CAD), 614
CD–drug inclusion complex, 191 Concentration-dependent, 20
CEB S-SEDDS formulation, 281 Confined impinging jet with a dilution mixer
Celecoxib (CEB), 281, 282, 456 (CIJ-D-M), 573, 574
Central nervous system (CNS), 14 Contact angle, 485, 486
Centrifugational forces, 206 Container closure integrity tests, 225
Cerebrospinal fluid (CSF), 219 Container closure selection, 226
Charging powders, 156 Contamination-free milling process, 161
Chemical derivatization, 20 Controlled drug-delivery systems, 625
Chemically-modified cyclodextrins, 234 Controlled precipitation (CP) process, 575, 576, 585, 586
678 Index
Conventional DSC, 44 pharmaceutical powders, properties of

Conventional lyophilization, 461, 462 brittle matrix powder, for pulmonary drug
Conventional ssNMR approaches, 64 delivery, 479–481
Coprecipitated hydrochlorothiazide nanoparticles, 545 engineered high surface area powder, for oral
Coprecipitation, 193, 306, 310 drug delivery, 477, 479
Cornea membrane, 13 poorly water-soluble drug compositions
Coronavirus disease 2019 (COVID-19), 198 aluminum salt-adjuvant vaccines, 515, 516
Cosolvent excipients, 183 carbamazepine (CBZ), 488, 489, 513–515
Cosolvent system, 10 cyclosporin, 487, 488
Cosolvents danazol, 490, 491
concentrated drug solutions preparation, 183 fenofibrate (FB), 502, 503
emulsification process, 183 fixed-dose combination drug formulation, 509, 510
ethanol and propylene glycol, 183 itraconazole, 498, 500, 502
higher solubility, 183 niclosamide, 520, 521
physical properties, 183 rapamycin, 507–509
preclinical studies, 182 remdesivir (REM), 511–513
self-emulsifying formulations, 183 tacrolimus compositions, 494–496, 498
usage, 183 voriconazole (VCZ), 504–507
water-miscible organic solvents, 182 rapid freezing, 472, 473
Cosolvents solubility theoretical modeling solid dispersion/solution, supersaturation, 454, 455
ADR, 181 solid-state characterization
crystalline nonelectrolyte, 180 spectroscopic methods, 483
dissociation process, 180 thermal analysis, 483
Jouyban–Acree model, 181 X-ray diffractometry (XRD), 482
log linear equation, 180 solubility advantage of nanoparticles, 455, 456
MPD, 181 solvent/cosolvent systems
parameter, 180 fluid dynamics, 474
volume fraction, 180 lyophilization, 475
water-soluble drug, 181 solubility, 472, 474
Counter-rotating intermeshing designs, 329 spray freeze drying (SFD) (see Spray freeze
COVID-19, 198 drying (SFD))
Cremophor, 10, 11, 198 storage of dried powders, 471
Critical micelle concentration (CMC), 228, 229 surface analyses
Critical process parameters (CPPs), 312 AFM, 484
Critical quality attributes (CQAs), 312 contact angle, 485, 486
Cryogenic milling energy dispersive X-ray spectroscopy (EDX), 484
apparatus, 163 scanning electron microscopy (SEM), 484
average particle size, 162 surface area measurement, 485
crystalline nanoparticles, 162 time-of-flight secondary-ion mass spectrometry
inherent limitation, 163 (ToF-SIMS), 484
liquid nitrogen, 162 TFF (see Thin film freezing (TFF) process)
nanosuspension, 163 Crystal packing diagrams, 132
performance, 162 Crystalline lens, 12
polymeric excipients, 163 Crystallization, 186
process conditions, 162 Curcumin, 17
Cryogenic spray processes, 456 Cyclodextrin, 8
Cryogenic technologies Cyclodextrin-based formulations, 11
atmospheric freeze drying (ATMFD), 457 Cyclodextrin complexation, 8
bottom-up particle engineering, 456 Cyclodextrins (CD), 11, 13, 16, 19, 20, 229, 230, 639
density measurement, 486 chemical structure, 190
dissolution, 475, 486 commercial products, 194
drug composition, 477 complexes, 189
dynamic vapor sorption (DVS), 487 concentration, 192
external factors, 457 cyclic oligosaccharides, 189
inverse gas chromatography (IGC), 487 derivatives, 189
micro-/nanoparticle precipitation technologies, 456 drug’s intrinsic solubility, 189
morphological instability and poor wettability, 476 equilibrium constant, 189
nonpolymeric natural products, 476 glucopyranose units, 190
particle-size measurement, 486 hydrophobic and hydrophilic, 191
Index 679
individual drugs, 189 Dissolution improvement, 529

internal hydrogen bonding network, 191 Dissolution testing
phase solubility profile drug, 192 alternative dissolution studies, 76–78
physical and chemical characteristics, 190 dipsolution performance, machine and deep
thermodynamic driving forces, 191 learning, 78
Cyclone efficiency, 388 excipient screening for supersaturation maintenance
Cyclosporin A (CsA), 14, 19, 487, 488, 554, 555, ability, 74–75
560–562, 564, 573, 576, 584, 585 in vivo testing (see In vivo dose administration)
Cytochrome P 450 (CYP 450), 5 sample handling, 72–74
Cytomegalovirus (CMV), 230 supersaturation dissolution studies, 75–76, 92
Dissolving drug molecules, 7
D Distributive mixing, 329
Damkohler number (Da), 565, 568 Docetaxel, 10, 229
Danazol, 262, 263, 279, 280, 490, 491, 562–564, 575 Docosahexaenoic acid (DHA), 230
Danazol desorption, 277 Dosage form manufacturing techniques
DEPO-Testosterone® injection, 230 carrier excipients, 199
Diethylene glycol monomethyl ether (DEGEE), 183 fluid bed melt granulation, 199
Differential scanning calorimetry (DSC), 300, 338 fluidized bed melt granulation, 207
analysis of drug–excipient interactions, 86–87 gelatin encapsulation, 199
Diffuse reflectance IR (DRIFT), 54 hard gelatin capsule (see Hard gelatin capsule)
Diffusion and stranding mechanism, 183, 198 molten low-viscosity liquids, 199
Diffusion-limited colloid aggregation, 579 powered solution technology, 207–208
Digestion-induced supersaturation, 254 soft gelatin capsules (see Soft gelatin capsules)
Diluents, 225 spray congealing, 199, 206–207
Dispersed systems Dose-independent clearance, 10
coarse suspensions, 231 Dose–response proportionality, 141
emulsions, 230, 231 Doxorubicin, 10
hydrogels, 237 Doxycycline, 227
ISFIs, 238 Drug candidates, 168
liposomes and colloidal systems, 233 Drug compounds, 117
Dispersion and digestion test, LBFs Drug concentration, 9
acidic and neutral drugs, 267 Drug crystallinity, 188
bile, 275 Drug degradation, 187, 188
bile-salt concentration, 266 Drug delivery, 623, 627, 631
drug loading, 266 OMS materials, 628
drug precipitation, 267 Drug discovery
drug separation, 266 high-throughput screening techniques, 599
drug solvent capacity, 265 Drug dissolution, 4, 233
fatty acids, 264, 265 Drug loading, 624
gastric lipases, lipid digestion, 264 Drug metabolism and pharmacokinetics (DMPK), 652
HPMC, 270 Drug migration, 202
in vitro digestion test, 265 Drug molecules, 2
in vitro evaluations, 269 drug particle size, 142
in vitro paclitaxel concentration, 270 Drug product stability, 225
ionized and unionized fatty acids, 264 Drug release, 232
LFCS consortium, 264 Drug substance, 64, 103, 104, 339
LFCS consortium fasted-state, 266, 269 Drug’s solubility, 4, 9, 263
lipolysis profiles, drug-loaded type IIIB-MC, 268 in aqueous media, 530
lipolysis-triggered precipitation, 267 in CO2, 533, 534
pH-stat apparatus, 264 in scCO2, 532
PPIs, 268–271 Drug–water interface, 562, 564
stressed digestion conditions, 268 Dry-milling techniques
stressed digestion testing, 269 correlation, 147
titrated fatty acid concentration, 265 environmental limitations, 154–156
type IIIA formulations, 271 equipment, 147
type of drug, 266 fluidized bed jet milling, 148
Dispersive mixing, 329 particle–equipment interactions, 147
Dissipation, 52 pin mill, 152–154
DissoCubes, 163 scale-up process, 147
680 Index
Dry-milling techniques (cont.) carbamazepine and danazol, 563

spiral jet (see Spiral jet/pancake mills CBZ particles, 562
Dust explosion control strategies, 155 contact-angle measurements, 563
Dust explosions prevention and mitigation, 155 CsA particles, 562
Dynamic mechanical analysis (DMA), 51, 52 dichloromethane (DCM), 560
drug particle properties, 563, 564
E high-potency particles, 562
Effective solubility, 19 nozzles, 560
Eicosapentaenoic acid (EPA), 230 operating parameters, 560
Elan Drug Technologies, 142 parenteral applications, 563
Electroanalytical techniques, 194 rapid stabilization of particles, 563
Electrospinning, 311 schematic of EPAS, 561
Electrostatic spinning, 618 stabilizer, 560, 562
CAR drug release, 620 unadsorbed surfactants, 562
coaxial, 619 Explosive dust clouds, 155
drug loading, 617 Extended-release products, 219
electric field intensity, 616 Eye drops, 11–13
electrostatic and MES, 620 Ezetimib(EZB)–indapamid (IDP) co-amorphous
Eudragit®, 618 mixtures, 635
factors, 617
HME particles, 620 F
hydrophilic amorphous solid solutions, 617 FDA's inactive ingredient database, 184
in vitro dissolution rates, pH values, 619 Fenofibrate (FB), 207, 260, 502, 503, 665
manufacture polymer-free inclusion complexes, 620 Fenofibrate-HPβCD inclusion complex, 193
nanofibers, 618 Film dissolution method, 303
pharmaceutical applications, 617 Fines, 155
polymer industry, 616 Fixed-dose combination (FDC)
POLYOX®, 618 multilayered and tablet-in-tablet geometries, 613, 614
PVP matrix, 618 Flash nanoprecipitation (FN), 570–575, 585
solid dispersion Flory-Huggins equation, 300
itraconazole, 640 Flory-Huggins interaction parameter, 298, 344
preparation, 618 Flory Huggins theory, 47–49, 88, 89, 129
unmilled fibers, 617 Fluid bed granulator, 207
Electrosteric stabilization, 580 Fluidized bed jet milling
ELIGARD® (leuprolide acetate) suspension, 236 advantage, 148
Employing spray-drying, 20 air pressure effects, 149
Emulsion preparation techniques, 231 classifier, 148
Emulsion-freeze-drying technique, 455 disadvantage, 148
Emulsions, 230, 231 particle–particle collisions, 148
Emulsion templating, 563 scheme, 148
Encapsulation, 458 surface, 149
Energy dispersive X-ray spectroscopy (EDX), 484 working pressures, 148
Enhanced permeation and retention (EPR), 233 Fluidized bed melt granulation, 207
Ephedrine crystallization, 107 Fluidized spray drying (FSD), 402
Ephedrine salts, 107, 109 agglomerated particle, 402
Escherichia coli, 223 fluidizing beds, 403
Ethanol, 10, 19 Food and Drug Administration (FDA), 309
Ethoxylated surfactants, 560 Food effects
Ethylenediaminetetraacetic acid (EDTA), 239 likelihood, 21
Etoposide, 10 negative, 3
Etravirine, 669 occurrence, 3–5
Eudragit, 577, 578 positive, 3
Eudragit® L100-55, 602, 619, 639 Food intake, 4
European Medicines Agency (EMA), 198 Formulation-based techniques, poorly water-soluble drugs
Eutectic mixture formulation strategy, 186 co-amorphous mixtures (see Co-amorphous mixtures)
Evaporative precipitation into aqueous solution mesoporous materials (see Mesoporous materials)
(EPAS) process Fourier transform infrared spectroscopy (FTIR), 52, 194
analysis of particles, 561 ATR, 52
APIs, 560 ATR–FTIR, 54
Index 681
DRIFT, 52 H
elastic and inelastic interactions, 53 Hard gelatin capsules
excipient interactions, 54 antioxidant, 205
polymorph screening, 54 compatible excipients, 205
and Raman spectroscopy, 52 compositions, 203
sample preparation, 53–54 fill materials, 203, 204
Freeze drying, 193 fill-weight uniformity, 204
Freezing-induced nanoparticle precipitation formulation, 204
methods, 530 Gelucire 44/14, 203
Functionalized OMS systems, 627 hydrophilic cosolvents, 204
Fused deposition modeling (FDM), 612, 613 hydrophilic solubilized formulations, 205
Fused filament fabrication (FFF), 612 manufacturing, 203
Fused filament modeling (FFM), 612 opacifier, 203
Fusion-based technologies, 301 PEG, 204
PEG-based solubilized nifedipine formulation, 205
G polymers, 204
Gas antisolvent (GAS) precipitation relative humidity conditions, 204
API’s solubility, 535 water migration, 204
batch process, 535 Hemolysis, 9
CBZ, 539 Heterogeneous particle-size distribution, 142
CO2 addition rates, 538 Heterosynthons, 125
CO2 concentration, 536 High melting point APIs, 603, 604
CO2–organic solvent solution, 535 High-melting-point drug substances, 340
drawback, 535 High-molecular-weight PEG, 184, 204
effect of CO2 addition rate, 539 High-performance liquid chromatography (HPLC), 194
mass-transfer rate, 558 High-pressure homogenization, 142, 144
mathematical modeling, 539 disadvantage, 165
nanoparticles, s-nitrosoglutathione, 540 microfluidizer, 164, 165
paracetamol and lysozyme, 537 microprecipitation, 167
particle growth, 538 particle-size reduction, 171
PCA process, 540 piston-gap homogenizers, 163, 164
PLLA, 538 High-throughput screening technique, 107, 109, 599
poorly water-soluble drug, 536, 537 High-velocity jet techniques, 567, 568
pressurized CO2, 537 Hildebrand equation, 180
PSD, 537 Homogenization, 163, 164, 530, 532, 581
salicylic acid–propanol–CO2 system, 536 Homosynthon, 126
schematic of GAS process, 536 Hot-melt extrusion (HME), 50, 293, 297, 300, 308,
toluene–CO2 system, 536 600–607, 610, 613, 614, 620, 628
Gas antisolvent recrystallization (GASR), 188 Hot-stage microscope (HSM), 194
Gas-assisted injection, 333 HPMC-based stabilizers, 579
Gastric emptying, 3 Hydraulic packing, 160
Gastrointestinal (GI) tract, 2, 3, 6 Hydrochloric acid, 226
Gastrointestinal degradation, 8 Hydrochlorothiazide liquisolid tablets, 208
Gel and ointment-based formulations, 13 Hydro-fluoroalkanes (HFAs), 18
Gelatin banding process, 203 Hydrogel-based commercially marketed products, 239
General solubility equation (GSE), 34, 35 Hydrogels, 237
GI environment, 3 Hydrogen bonding, 198, 476
Gibbs energy difference (GED) method, 292 Hydrolytic degradation, 240
Gibbs free energy, 114, 115 Hydrophilic cosolvents, 204, 258
Glass-forming ability (GFA), 112, 292 Hydrophilic drugs, 233
classification, 113, 115 Hydrophilic-lipophilic balance (HLB), 257
concept, 112 Hydrophobic core, 228
determination, 113 Hydroxycarbonate apatite (HCA), 631
Glucose hexamer, 190 Hydroxypropyl B cyclodextrin (HPBCD), 605
Glycerin, 200 Hydroxypropyl cellulose (HPC), 609
Glycerol tripalmitate, 207 Hydroxypropyl methylcellulose phthalate
Grinding air pressure (GAP), 169 (HPMCP), 341, 577, 578
Griseofulvin, 272, 537, 543 Hydroxypropyl β-CD (HPβCD), 192
Guest molecules with less polarity, 191 Hydroxypropylmethyl cellulose phthalate (HP-55), 477
682 Index
Hydroxypropylmethylcellulose (HPMC), 568–571, 576, device and diluent compatibility, 225

579, 586, 587 packaging and manufacturing considerations, 225, 226
Hydroxypropyl-β-cyclodextrin (HP-β-CD), 11, particulate matter, 223
230, 245, 246 pH, 221
Hygroscopicity, 300 preservatives, 224, 225
Hypromellose acetate succinate (HPMCAS), 5, 300, sterility and endotoxin requirements, 223
571, 574, 577 tonicity and biological implications, 222, 223
Hypromellose phthalate, 5, 577 Inorganic particles, 20
In-situ forming implants (ISFIs), 238
I Insoluble Drug Delivery Microparticle technology
IA products, 220 (IDD-P®), 142
Ibuprofen, 625 Instrumental techniques, 319
ICH guidelines on residual organic solvents Intra-articular (IA) injections, 219, 220
class 1 solvents, 396 Intrathecal (IT) injection, 219, 220
class 2 solvents, 396 Intravenous (IV) administration, 218, 219
class 3 solvents, 396 Intravitreal injections, 12
ICH Q6A (1999) decision tree Intrinsic dissolution testing, 41
particle-size reduction, 667 Inverse gas chromatography (IGC), 487
solid-form selection, 663, 664 Investigational New Drug Application
Ignition, 155 (IND) stage, 652
In situ gelling microemulsion, 22, 23 CMC submissions, 654
In vitro deposition properties, 197 amorphous solid dispersion, 656
In vitro digestion tests, 268, 270–272 control of drug substance, 656
In vitro dissolution, 34, 37, 79, 83 lipid formulation, 656
In vitro hemolysis test, 223 solid form, 655
In vitro nasal deposition testing, 519 phase 1 clinical study
In vitro–in vivo correlations, 76, 77 drug substance, 653
In vitro–in vivo relationship (IVIVR), 3, 282 investigational drug product, 653
In vivo dose administration physicochemical properties, 654
administration via inhalation, 80–83 submission and maintenance, 652
oral route of administration, 83–85 Ionic surfactants, 161
rat models, 80 Ipriflavone, 549, 551
simulated intestinal fluids, 79 Isothermal titration calorimetry (ITC), 194
Inactive Ingredient Guide (IIG), 187 Isotonic formulation, 222
IND–ARG co-amorphous formulation, 637 Itraconazole (Itz), 129, 477, 498, 500, 502, 529, 562, 564,
IND–ARG co-amorphous tablet, 637, 638 566–569, 576, 579, 580, 585, 586, 601, 625,
Indomethacin (IND), 625, 634 626, 628, 640
Indomethacin-HPβCD inclusion complexes, 193 Itraconazole-loaded OMS, 641
IND–RAN mixtures, 634 ITZ supersaturation, 604
IND–RTV co-amorphous mixtures, 635 IV nanosuspension formulation preparation, 246
Injectable emulsions, 230
Injectable formulations J
administration routes (see Administration routes) Jouyban–Acree model, 181
buffer systems, 222
cosolvents, 228 K
drug product development, 218 Kelvin equation, 143
FDA, 221 KinetiSol®, 300
IA injections, 220 KinetiSol® dispersing (KSD)
inter/intra subject variability reduction, 217 anionic polymers, 606–609
materials, 225 ASDs homogenous production with low-drug
medications, 217 loading, 610
multidose, 229 cross-sectional view, 601
poorly water-soluble drugs, 218 fusion-based processing technique, 601
preservatives, 224 high drug loading reduces pill burden, 609, 610
tonicity, 222 high melting point APIs, 603, 604
toxicity, 229 high molecular weight, viscous polymers, 604, 605
Injectable product development workflow, 241, 242 HME, 601–603
Injectable products formulation considerations ITZ dissolution rates, 611
buffered systems, 221, 222 KC254B Compounder, batch-mode operation, 602
Index 683
KC254C Compounder, semi-continuous triglycerides, 256

operation, 602 type I formulations, 255
processing and equipment, 601 type II formulations, 255, 256
prolong mixing with thermal conductivity type III formulations, 255, 256
modification, 609, 610 type IV formulations, 255, 256
PXRD profiles dilution/digestion-induced supersaturation, 254
bulk ITZ, 607 drug solubility determination
GRIS crystalline drug, 606 danazol, 279, 280
ritonavir plasma concentration, 608 formulation composition, 255
short chain, low molecular weight oligomers, 605 lipid digestion, 255
solid dispersion solidification (see Solid lipid-based formulations)
cyclodextrin, 639 solubility-limited drugs, 253
Eudragit® L100-55, 639 solubilization, 254
temperature profiles, 604 stressed digestion test, 281
weakly basic APIs, 606–609 supersaturation effects, 254
Kneading, 193 supersaturation pathway, 254
Lipid Classification System Consortium (LFCS
L Consortium), 258–260, 263–266, 268, 269, 271
Large bronchi, 17 Lipid colloidal phases, 275
Large porous carriers of nanoparticles (LPNP), 24, 25 Lipid Formulation Classification System (LFCS), 258
Large-scale manufacturing method, 128 Lipophilic, 228
Laval-shaped nozzle, 151 Lipophilic excipients, 205
Layer by layer (LbL) polymer coating, 631 Lipophilic formulations, 262
LBFs characterization Liposomes
dispersion and digestion testing (see Dispersion carrier systems, 233
and digestion test, LBFs) categories, 233
drug solubilization/precipitation, 259 drug loading, 234
equilibrium solubility lipids, 234
danazol, 263 nanocarriers, 233
lipophilic drug, 263 poorly soluble compounds, 233
lumefantrine compounds, 262 preparation techniques, 234
equilibrium solubility values water-soluble drugs incorporation, 233
danazol, 261 Liquid and semisolid formulations, 209
fenofibrate, 261 Liquid argon, 457
tolfenamic acid, 261 Liquid nitrogen, 457
in vitro methods, 258 Lovastatin (LOV), 629
in vivo, 270–273 Low-melting-point drug substance, 340
in vivo bioperformance, 278 Lyophilization, 20, 167, 240, 475
LFCS Consortium, 259, 260
solubility, 259, 262, 264 M
LCT lipid, 271 Macrosuspension, 161, 167
Lead formulations, 75, 79 Marketed nanoparticle formulations, 237
Lecithin, 4 Marketed nutritional emulsions, 230
Leidenfrost effect, 457 Mass median aerodynamic diameters (MMAD), 17
LEMS™ sealing operation, 204 Material bed–based 3D printing, 611
Lidocaine-loaded solid lipid microparticles Material jetting, 611
(SLMPs), 558 Material Studio Blends module, 298
Light sensitivity, 226 Maxwell model, 51
Limiting oxygen concentration (LOC), 155 Maxwell unit, 51
Lipid-based drug solution, 202 MCT lipid, 270
Lipid-based formulations (LBFs) MD-FIS model, 78, 79
absorption, 253, 279 Mean percentage deviations (MPD), 181
characterization (see LBFs characterization) Mechanical grinding, 193
classification and components Mechanical particle-size reduction techniques
hydrophilic cosolvents, 258 aim, 142
mixed glycerides, 256 combinative strategies (see Combinative
pharmaceutical ingredients, 255 particle-size reduction techniques)
pharmaceutical nonionic surfactants, 259 dissolution, 168
surfactants, 257, 258 microprecipitation, 171, 172
684 Index
Mechanical particle-size reduction techniques (cont.) Merck, 361

need, 142 Mesoporous carbon nanospheres (MCN), 631
rationale (see Particle-size reduction rationale) Mesoporous materials
top-down techniques (see Milling) amorphous CAR, 632
Media (bead) wet milling definition, 623
alternative model, 160 drug delivery, 631
API manufacture, 160 HCA, 631
chamber and stirrer geometry, 158, 159 inorganic mesoporous carriers, 631
disadvantages, 160 mesoporous carbon, 628–631
grinding media, 159 OMS (see Ordered mesoporous silica
grinding process, 161 (OMS) materials)
intensity factor, 160 organic mesoporous carriers, 631
materials, 158 porous starch, 632
nanosuspension mean size, 161 Mesoporous silica, 311, 358, 359, 641
parameters, 159 Mesoporous silica microparticles (MSM), 626
particles obtained, 161 Mesylate salt, 109
particle-size reduction, 169 Metabolic extraction, 5
particle-size reduction control process, 158 Metabolism, 2, 3, 5, 6, 8, 14
physical drug properties, 161 Metal–chelator complex, 239
rapid drug dissolution properties, 161 Metastable solids, 114, 122
scaling up, 160, 161 Meter-dosed pump sprays, 15, 16
specific characteristics, 158 Metered dose inhalers (MDIs), 18, 19
steric and electrostatic stabilization, 161 Methoxy b-poly(L-lactide) 2000-poly (ethylene glycol)
stress energies, 160 2000 (mPLLA-PEG), 553
stress model, 158 Methylated cyclodextrins, 20
variables, 158 Micellar solubilization, 241
volumetric concentration, 160 Micelles, 236
vs. rotor–stator wet mill, 158 Microcalorimeter, 207
Media milling, 142 Microcrystalline cellulose, 208
MedinCell, 238 Microcrystalline wax, 206
Medium chain triglyceride (MCT), 230 Microemulsion, 207
Medium-chain fatty acids, 256 Microfluidic reactor, 459
Medium-chain triglycerides, 256 Microfluidization, 166, 168
Melt extrusion, 306 Microfluidizer, 164, 165
bioavailability, 351, 352 Microhydrodynamic model, 160
co-amorphous systems, 357 Micronization, 142, 455, 552, 557, 558
currently marketed and developed drug products, 367 Microprecipitated bulk powder
drug-polymer incompatibilities, 351 (MBP), 310, 577, 578, 587
equipment description, 329, 331, 332 Microprecipitation, 165, 293, 576
feeders, 332 Microprecipitation process solvents, 300
for improved oral bioavailability, 369 Microscopic technique, 314
manufacturability, 348 Microsphere preparation techniques, 233
mechanical energy in, 337, 338 Microspheres, 17
melt-extruded solid dispersions, 341–345 Microwave-induced dielectric-induced
non-oral sustained-release formulations, 363, 365 heating, 621, 622
non-polymer carriers Milling
mesoporous silica, 358, 359 categories, 145
pharmaceutical co-crystals, 357 collision events, 145
pharmaceutical extrusion, 338, 339 disadvantage, 146
process optimization, 353 dry-milling (see Dry-milling techniques)
QbD and PAT, 355, 356 interparticle collisions, 146
scale-up of extrusion operations, 354, 355 Mohs scale, 146
solid-state characterization of, 349 particle morphology, 146
stability, 350, 351 particle-size distribution, 146
steady-state processing and production particle-size reduction, 146
feedback, 333, 334, 336 research, 145
Melted solid dispersions, 188 slurry milling (see Wet milling techniques)
Melt-extruded solid dispersions, 341 solid density, 145
Menthol, 556, 557 surface area, 146
Index 685
Minimum ignition energy (MIE), 155 recovery methods, 579

Mixed glycerides, 256 sodium sulfate (Na2SO4), 579
Mixed-product-removal crystallization techniques for solvent removal, 578
(MSMPR) model, 544, 566, 567 Nanoparticles, 8, 20, 236
Mixing energies, 567, 571 bottom-up approaches, 530
Modified CDs, 192 high-energy methods, 530
Modified HPMC (Affinisol™), 368 top-down approaches, 530
Modified-RESS process, 555 Nanosuspensions, 20, 160, 161, 166, 167, 235, 455
experimental parameters, 554 Naproxen (NAP), 534, 539, 551, 554,
PGSS process, 557, 558 575, 576, 580, 633
phospholipid stabilizers, 555 Narrower plume geometries, 16
RESAS/RESOLV process, 553, 555 Nasacort® AQ, 17
RESS-SC process, 556, 557 Nasal administration
scCO2, 554 advantageous, 14
schematic of RESAS process, 554 airflow, 14
US-RESAS process, 555 bioavailability, 14
US-RESOLV, 556 challenges, 16, 17
water-insoluble materials, 556 CNS, 14
Modulated differential scanning calorimetry disadvantages, 15
(mDSC), 42–44, 47, 51, 60, 349 drug delivery, 14
Mohs scale, 146 meter-dosed pump sprays, 15
Molecular dynamic (MD) simulations, 297 nasal cavity, 14, 15
Molecular-level solid dispersion, 288 olfactory neuroepithelium, 14
Molten mass, 187 respiratory area, 14
Monodisperse droplet generators (MDGs), 382 Nasal cavity, 14–17
Mononuclear phagocyte system (MPS), 233 Nasal drug delivery, 197
mRNA-based COVID-19 vaccines, 236 Nasal mucosa, 16
Mucoadhesive tablets, 7 Nasopharynx, 17
Mucociliary clearance, 197 Nebulized itraconazole nanoparticle dispersions, 23
Muhrer’s model, 538, 539 Nebulizers, 18, 19
Multi-inlet-vortex mixer (MIVM), 573–575 Negative food effect, 3
Multilamellar vesicles (MLVs), 233 Nernst–Brunner/Noyes–Whitney equation, 2
Multilayered tablets, 613, 615 Neuroepithelium, 16
Multiparticulate-modified release dosage, 206 New chemical entities (NCEs), 1–3, 21
Multivesicular liposome (MVL), 244 New drug application (NDA), 652, 657
Multivesicular liposome formulation preparation, 244 drug product, 660–662
Multivesicular vesicles (MVVs), 233 drug substance, 657, 659, 660
Murine model poorly water-soluble drugs, 662
by dry powder insufflation, 93 BCS classification, 670
liquid formulation, 668, 669
N particle-size reduction, 665–667
NanoCrystal®, 168 solid dispersion, 669
NanoCrystal® technology, 142 solid-form selection, 663
NANOEDGE technology, 165, 166 Newly discovered drugs, 141
Nanofibers, 618 Niclosamide, 520, 521
Nanoparticle recovery Nicotinamide, 539, 552
amorphous drug–polyelectrolyte nanoparticle Nifedipine (NIF), 63, 187, 205, 226, 341, 609
complex, 581 Nimeodipine (NIM), 631
clofazimine-loaded nanoparticles, 578 Nimodipine, 188
crystallization of Itz, 580 N-methyl-2-pyrrolidone (NMP), 236, 238
curcumin and chitosan, 580 N-octyl-O-sulfate chitosan (NOSC), 237
drug loadings, 581 Nonaqueous solvents, 18, 19
electrosteric stabilization, 580 Nonaqueous/oily vehicles, 230
flocculate formation, 580 Nonfreezing-based precipitation techniques, 530
flocculation of primary particles, 579 Nonionic surfactants, 228, 258, 554
flocculation/filtration process, 581 Nonspecific formulation approaches, 142
Ostwald ripening, 578 Norvir®, 359, 607
polymer solvation and particle volume fraction, 579 Nose-to-brain delivery, 22, 23
process conditions, 578 Noyes–Whitney equation, 143, 530, 576
686 Index
O targeted drug delivery, 631

Ocular administration Ordered mesoporous silica (OMS) materials, 641
aqueous humor, 12 applications, 623
challenges, 12–14 controlled drug-delivery systems, 625
characterization, 11 drug delivery, 623, 627
eye drops, 12 drug loading, 624
ophthalmic dosage, 11 HME process, 628
periorbital routes, 12 ibuprofen, 625
Ocular surface, 13 IND, 625
Oil-in-water emulsion, 231 ITZ, 625, 626, 628
Olanzapine, 16, 123, 124 MSN, 626
Olfactory neuroepithelium, 14 nanosized pores, 626
Online sizing system, 149 pore diameter ranges, 623
Ophthalmic cyclosporine a-loaded nanosuspension, 22 release behavior, 625
Ophthalmic dosage, 11 SIM, 625
Ophthalmic drug delivery, 197 sol–gel or ion reaction method, 624
Oral administration synthesis, 627
BCS, 3 TEL, 625
challenges, 3–7 Organic drug solution, 535, 540, 541, 566, 567, 584
food and metabolism, 3 Organic solvents, 9
GI tract, 2, 3 Organic water-miscible cosolvents, 9
intestinal absorption, 3 Ostwald ripening, 560, 567, 570, 575, 578
NCE, 3 OVA-absorbed aluminum hydroxide, 517
solid dosage, 2 Oxidative degradation, 239
systemic administration, 2
Oral bioavailability, 5 P
Oral drug absorption, GI tract Paclitaxel, 10, 11
active transport, 182 Pair distribution function (PDF), 57–59
blood/lymphatic capillaries, 182 Parenteral administration, 240
constraints, 182 aqueous-based solution, 8
dissolution rate, 182 challenges, 8–11
Fick’s law of diffusion, 181 formulations, 8
membrane transport, 182 organic solvent mixtures/mixed aqueous/organic
passive paracellular diffusion, 182 cosolvents, 8
pathways, 181 pH and tonicity, 8
surfactants, 182 pharmacological effect, 8
Oral mucosa, 7 side effects, 8
Oral transmucosal delivery, 7 stabilizers, 8
Oral transmucosal dosage, 7 Parenteral emulsions, 230
Oral transmucosal drug delivery, 7 Particle size distribution, 496
Ordered mesoporous carbon (OMC) materials Particles from gas saturated solutions
adsorption isotherms, IBU solution concentration, 629 (PGSS) process, 557, 558
advantages, 630 Particle-size reduction rationale
Caco-2 cells, 628 crystalline surface, 145
c-MCM, 630 diffusion boundary layer thickness, 144
drug delivery, 628 direct effect, 143
glucose based MCN, 632 dissolution rate, 143
hard-templating, 628 high-energy processes, 144
IBU, 629 higher saturation solubility, 145
in vitro drug release, LOV, 630 high-pressure homogenization, 145
KOH, 631 Kelvin equation, 143
LbL polymer coating, 631 micro- to nanorange, 144
magnetic field, 631 micron range, 143
MSM, 630 nanosuspension formulations, 145
NIM, 631 saturation solubility, 144
poorly soluble drug, 629 surface area, 143
porous car, 628 top-down techniques (see High-pressure
ROS, 628 homogenization)
soft-templating, 628, 629 vapor/dissolution pressure, 143
Index 687
Particle–wall collisions, 145 hygroscopicity, 106

Particulate matter, 223, 224 of pharmaceutical drug, 104
PEG-based drug delivery systems, 187 precipitation process, 106
PEG-based solid dispersion screening process, 104
classification, 184, 185 Pharmacodynamic response, 16
fusion method, 187, 188 Pharmacokinetics, 9, 11
hydrophilic solubilized formulation, 184 Phase solubility diagram, 192
molecular weight, 184 Phenytoin, 8, 163
parallel helices, 185 Phlebitis, 8
PVP K-30, 184 Phospholipids, 233
solubilization, 184 Phospholipid stabilizers, 555
solubilized formulations (see PEG-based solubilized pH-solubility, 75, 85
formulations development) pH-solubility profile
solvent method, 188, 189 haloperidol, 85
water-soluble drugs, 184 PWS drug, 36–37
PEG-based solubilization formulations, 208 sildenafil, 36
PEG-based solubilized formulations development Physicochemical properties of drug, 3
crystalline structure, 186 Pin mill
drug/polymer/PEG dispersion, 187 dynamic selector, 153
drug–PEG eutectic crystallization, 186 equipment, 152, 153
fenofibrate dissolution, 186 mechanical energy impact mill, 152
formulation and process approaches, 187 particle–mill collisions, 152
fusion enthalpy, 186 rotor–stator mill, 152
hydrogen bonding/hydrophobic interactions, 187 solid material physical properties, 153
IIG, 187 SSA, 153, 154
moisture sensitive, 186 variable effects, 153
molecular/ionic crystals, 185 Piroxicam, 188
nifedipine, 187 Piston-gap high-pressure homogenization, 170
PEG melting point, 186 Piston-gap homogenizers, 163, 164
semicrystalline phase, 186 Pluronic F127 (P127), 554, 560–562, 564, 566, 586
solid dispersion, 187 Polarized light microscopy (PLM), 350
solubilization capacity, 186 Poloxamer, 205, 207
two-phase system, 186 Poly (Lactide-Co-Glycolide) microspheres preparation,
PEG-based solubilized nifedipine formulation, 205 243, 244, 247
PEG-induced oxidation, 187 Poly(L-lactide) acid (PLLA), 538, 542
PEGylation, 233 Poly(vinyl pyrrolidone) (PVP), 618
Penn-Century Model DP-4 M dry powder insufflator, 82 Polydispersity index values, 161
Periorbital routes, 12 Polyethylene glycol (PEG), 183, 184
Personnel protection, 156 Polymer carrier phase, 338
P-glycoprotein (P-gp), 5 Polymer-free electrospinning
pH-adjusted formulations, 8 poorly water-soluble drugs, 621
pH modification, 8 Polymeric biodegradable microspheres, 231
Pharmaceutical “quality by design” approach, 218 Polymeric excipients, 163
Pharmaceutical cocrystals, 124 Polymeric micelles, 237
compatibility, 125 Polymeric nanoparticles, 14
dissolution enhancement, 124 Polymeric precipitation inhibitors (PPIs), 254, 268–271
intermolecular interactions, 125 Polymers, 236, 343, 576, 612
preparation, 124 Polymethacrylate, 577
Pharmaceutical extrusion, 338, 339 Polymethylmethacrylate, 577
Pharmaceutical industry, 599 Polymorphic forms, 111
Pharmaceutical ingredients, 255 Polymorphic transformation, 206
Pharmaceutical melt extruders, 328 Polymorphism, 111
Pharmaceutical nonionic surfactants, 259 Polymorphs, 110, 111, 117
Pharmaceutical processes, 142 amorphous solids, 112
Pharmaceutical salt formation, 104 compounds, 111
Pharmaceutical salts, 104, 107, 109, 110 high-throughput screening, 111
counterions, 106 polymorphic forms, 111
drug substance, 104 solvates, 111
high-throughput techniques, 106 Polyoxyethylene groups, 202
688 Index
Polythiazide liquisolid tablets, 208 HPMC/PVP, 545

Polyvinyl alcohol (PVA), 148, 605 manufacturing technique, 545
Polyvinylphthalate, 577 mass-transfer efficiency, 540
Polyvinylpyrolidone (PVP), 161, 574, 621 mass-transfer model, 547
Polyvinylpyrrolidone (PVP)-stabilized CsA particles, 560 organic solution, 540
Poorly soluble drug solubilization, 245 particle formation and growth, 546
Poorly water-soluble (PWS) drugs PCA processing, 546
aqueous solubility determination (see Solubility precipitated microparticles, 541
studies) rifampicin and ethyl cellulose, 544
formulation design, 2 SAS, 535
Nasal Route of Administration, 14–17 scalability, 542
NCEs, 1, 2 schematic of PCA process, 540
Nernst–Brunner/Noyes–Whitney equation, 2 SEDS, 535
ocular route of administration, 11–14 semicontinuous technique, 540
oral drug delivery, 94 solute concentrations, 544
oral route of administration, 2–7 solvent–CO2 system/solute–solvent–CO2 system, 546
parenteral route of administration, 7–11 subcritical and supercritical conditions, 547
precipitation, 2 swelling test, 546
pulmonary route of administration, 17–21 two-component jet nozzles, 542
route-specific challenges (see Route-specific zein and diclofenac mixtures, 544
challenges) Precipitation-based particle formation techniques, 565
solubility, 1 Predicted solubility ratios, 116
Poorly water-soluble antiretroviral API ritonavir, 614 Prehomogenized liquid, 164
Poorly water-soluble drug, 529, 530 Preservatives, 224, 225
Porous hydrophilic polymer (PVA) scaffold, 455 Pressurized CO2, 369
Porous starch, 632 Presystemic metabolism, 5
Posaconazole–CD inclusion complexes, 209 Presystemic/first-pass metabolism, 8
Positive food effect, 3 PREVYMIS™ (letermovir) injection, 230
Posterior eye, 14 Probe rotor–stator mills, 157
Postproduction sterilization methods, 231 Process analytical tools (PAT), 309
Potentiometric titration, 194 Process-based techniques, poorly water-soluble drugs
Powder bed fusion (PBF), 614, 616, 617, 640 AM (see Additive manufacturing (AM))
Powder dissolution profiles, 120 amorphous fraction, celecoxib, 622
Powered solution technology, 207–208 ASDs, 600
Precipitation electrostatic spinning (see Electrostatic spinning)
into acid, 576–578 glass-transition temperature limitations, 600
mechanism of precipitation, 531 HME process, 600
solution-based techniques, 530 KSD (see KinetiSol® dispersing (KSD))
Precipitation processes microwave-induced dielectric-induced
by EPAS, 584, 585 heating, 621, 622
by GAS, 582, 583 solid dispersions, 600
by PCA, 583 spray-drying process, 600
by RESS, 584 thermal exposure, 600
Precipitation technologies, 532 vitro drug release profiles, IND, 623
Precipitation with a compressed antisolvent Processing techniques, 188
(PCA), 533 Product packaging, 225
acetaminophen–ethanol solution, 543 Propylene glycol, 19, 200
amoxicillin, 542 Pseudopolymorphs, 111, 117
ascorbic acid–ethanol–CO2 system, 543 Pulmonary administration, 24
ASES, 533 asthma/pulmonary infections, 17
atomization intensity, 541 challenges, 18–21
coprecipitated anthocyanin and PVP, 547 epithelial barrier, 17
coprecipitated particles, 544 formulations, 17
coprecipitation, substrates and polymers, 541 MMAD, 17
crystal formation, 547 poorly soluble drugs, 18
curcumin and PVP, 541 respiratory system, 17
dissolution rate of drugs, 540 trans-epithelial transport, 17
drug nanoparticles, 543 Pulmonary deposition and dissolution, 23, 24
higher-intensity atomization, 541 Pulmonary drug administration, 197, 198
Index 689
Pulmonary drug delivery, 142 Resuspendability, 231

Pulmonary infections, 17 Resveratrol nanosuspension, 570
Reticuloendothelial system, 9
Q Rheometers, 348
Quality by design (QbD) approach, 148 Ritonavir (RTV), 635, 640, 667–669
Quality by design (QbD) methodology, 328 RNA-based therapies, 236
Quality by design phase (QbD), 294 Rotor–stator media mills, 147
Quality target product profile (QTPP), 311 Rotor–stator wet milling
Quantitative preformulation techniques, 366 equipment, 157
QVAR®, 18 geometries, 157
high-shear devices, 157
R multimodal distributions, 157
Raman spectroscopy, 52, 621 parameters, 158
Rapamycin, 507–509 particle-size reduction, 157
Rapid expansion from supercritical solutions (RESS), 533 scaling-up purposes, 157
benzoic acid, griseofulvin and β-sitosterol, 552 Route-specific challenges
capillary/laser-drilled orifice nozzle, 549 nasal administration, 14–17
cholesterol particles, 551 ocular administration, 11–14
controlling-expansion conditions, 552 oral administration, 2–7
depressurization of CO2, 559 parenteral administration, 7–11
ipriflavone, 549 pulmonary administration, 17–21
micronization and co-crystallization, 552 Ryanodex® (dantrolene sodium), 11
modified processes (see Modified RESS processes)
nozzle design, 548 S
nozzle types and diameters, 549 Salt flocculation, 580, 581, 586, 587
operating conditions, 548 Salt formation, 104, 227
population balance model, 552 Salt-induced flocculation, 581
process parameters, 548 Sample analysis, 106
P–T diagram for CO2, 548 Scanning electron microscopy (SEM), 194, 350, 484
salicylic acid and griseofulvin particles, 550 SCF antisolvent
SCF as a solvent, 547 GAS precipitation, 535
schematic of RESS process, 548 into a liquid solvent, 532
spray distance, 549 precipitation processes, 532
supersaturation, 552 SCF precipitation techniques
theoretical models, 551 average size of particles, 559
trends in processing conditions, 550, 551 characteristic mass-transfer times, 558
Rapid expansion from supercritical to aqueous solution comparison of micronization techniques, 558
(RESAS), 553–556, 559–561, 588 comparison of precipitation processes, 558–560
Rapid expansion of a supercritical solution into a liquid EPAS process (see Evaporative precipitation into
solvent (RESOLV) processes, 555, 556, 559 aqueous solution (EPAS) process)
Rapid expansion of supercritical fluid solution GAS (see Gas antisolvent (GAS) precipitation)
(RESS), 188 PCA process (see Precipitation with a compressed
Rapid expansion of supercritical solutions with solid antisolvent (PCA))
cosolvents (RESS-SC), 556, 557, 589, 590 RESS (see Rapid expansion from supercritical
Rapid freezing, 472, 473 solutions (RESS))
Reactive oxygen species (ROS), 628 schematics of different nozzles, 542
Recrystallization, 45–48, 53, 57, 63, 67, 68, 71, 87, 88 SCF precipitation technologies, 559, 582
REDITREX (methotrexate) injection, 219 Schizophrenia, 197
Refined natural oils, 256 Seal integrity testing, 225
Refining process, 256 SEDDS formulations (S-SEDDS), 270
Regulatory acceptability, 239 Selective laser sintering (SLS), 612, 614, 640
Regulatory guidelines, 223 Self-emulsifying, 5
Remdesivir (REM), 511–513 Self-emulsifying drug delivery systems
Remdesivir-Captisol®-based dry powder formulation, 211 (SEDDS), 256, 276
RESAS/RESOLV process, 553, 554 Self-emulsifying formulations, 21, 22
Respiratory area, 14 aqueous environment, 198
Respiratory epithelium, 16 components, 198
Respiratory system, 17 Cremophor, 198
Restasis®, 14 drug product, 198
690 Index
Self-emulsifying formulations (cont.) S-SMEDDS formulations, 277

P-glycoprotein-mediated efflux, 198 structure, 278
Self-emulsifying soft gelatin capsules, 183 synthetic poloxamers, 273
Self-microemulsifying drug delivery systems syringe-based extrusion 3-dimensional
(SMEDDS), 256 printing, 276
Semicrystalline material, 184 techniques, 273, 275
Semicrystalline particles, 165 Solidification, 273
SFL Spray freezing into liquid (SFL) Solid-in-oil-in-water (S/O/W) emulsion-based
process of, 466, 467 method, 233
Shark skinning, 332 Solid-liquid equilibrium (SLE), 298
Short chain, low molecular weight oligomers, 605 Solid-state characterization, 34
Silybum marianum, 184 Solid-state characterization, APIs, 41
Simvastatin (SIM), 625 AFM (see Atomic force microscopy (AFM))
Single particle analysis (SPA), 36 differential scanning calorimetry (DSC)
Single-pass rat jejunum perfusion model, 272 advantage, 41
Sirolimus, 665 drug–excipient interactions, 46, 47
SLS 3D printing process, 614 Flory–Huggins theory (see Flory–Huggins theory)
Small unilamellar vesicles (SUVs), 233 parameter selection method, 42–43
SmartCrystal technology polymorphic transformations, 46
H 42 technology, 167 thermal events, 43–45
H 69 technology, 166, 167 DMA, 51, 52
H 96 technology, 167 FTIR, 52 (see also Fourier transform infrared
SmartPill®, 4 spectroscopy (FTIR))
Sodium glycocholate, 161 Raman spectroscopy, 52
Sodium lauryl sulfate (SLS), 412 residual solvent analysis
Soft and hard gelatin capsules cross-linking, 210, 211 analytical determination, 66–67
Soft gelatin capsules residual solvent guidelines, 66, 70
aldehyde, 202 specific surface area (SA)
bloom strengths, 200 BET surface area analysis, 61–63
challenges, 202 for high-surface-area heat-liable material, 89
commercial, 200 ssNMR (see Solid-state nuclear magnetic resonance
common dosage form, 200 (ssNMR))
drug migration, 202 thermogravimetric analysis
glycerin and sorbitol, 200 excipient interactions, 50
hydrophilic excipients, 202 parameter selection, 49
lipid-based drug solution, 202 secondary technique, 49
manufacturing, 200 thermal decomposition, 50
plasticizer, 200, 202 XRD (see X-ray diffraction (XRD))
primary drying, 201 Solid-state modifications, 133
rotary die machine, 200 Solid-state nuclear magnetic resonance (solid-state NMR),
secondary drying, 201 63, 64, 89, 90, 194, 349
water migration, 201, 202 Solid-state properties, 227
Soft gelatin compatible vehicle, 202 Solid-state technologies
Soft vs. hard gelatin capsules, 203 and storage conditions, 114
Sol–gel/ion reaction method, 624 concepts and applications, 104
Solid dispersion-based drug delivery systems, 5 Solubility, 1, 259, 262, 264, 472, 474
Solid dispersions, 305, 345, 366, 600 Solubility determination, 35
Solid dosage, 2 Solubility parameter, 180, 344
Solid lipid-based formulations Solubility studies, 68, 69, 74
advantages, 273, 274 aqueous solubility determination, 34
bioavailability, 274 BCS class II compounds, 34
danazol dose, 276 solubility prediction, 34–35
high hydrophilic surfactant, 276 Solubilization, 9, 189, 254
material properties, 273 Solubilization effect, 254
model drugs, 276 Solubilization of drugs, 530
nonionic surfactants, 277 Solubilized formulations
poorly water-soluble drugs, 278 application, 180
self-emulsifying LBFs, 274 complexation technologies, 180
SLNs, 277–279 cosolvent-based, 183, 184
Index 691
dosage form manufacturing (see Dosage form pneumatic two-fluid nozzles, 380
manufacturing techniques) pressure nozzles, 381
drugs, 182 rotary nozzles, 380
excipients function, 180 three-fluid nozzles, 382
preformulation support, 208, 209 ultrasonic nozzles, 382
Soluble drugs, 12 brick-dust compounds, 438
Soluplus®, 298, 603 closed-loop configuration, 390
Solution crystallization method, 128 collection systems, 388
Solution-enhanced dispersion by supercritical fluids cyclone, 388
(SEDS), 535 electrostatic precipitators, 389
Solvent crystallization process, 130 filter bags, 389
Solvent evaporation, 184, 193, 404 drying chamber, 378
Sorafenib-loaded MBP, 578 feed systems, 392
Sparging, 240 fluidized spray drying (FSD), 402
Specific surface area (SSA), 152 agglomerated particle, 402
Spectrophotometric analyses, 243 fluidizing beds, 403
Spectroscopic methods, 483 gas droplet contact and particle formation, 383
Spectroscopic techniques, 194 atomization process, 383
Spiral jet/pancake mills coalescence, 386, 387
advantages, 152 cocurrent configuration, 383
critical grinding pressure, 151 countercurrent mode, 383
gas velocities, 149 drying process, 385
geometry dependence, 150 gas dispersers, 387, 388
internal orientation, 149 product properties, 385
nozzle dependence, 151 innovative approach, 437
operating variables, 149, 151 laboratorial-scale equipment, 390
particle-size distributions, 152 microencapsulation, 440
particle-size reduction, 149, 168, 169 abrasion, 441
scheme, 150 atomization process, 441
solid particle feed rate, 152 carbohydrates, 440
specific energy consumption, 151 gums and gelatins, 441
SSA, 152 lycopene, 441
Spironolactone nanoparticles, 570 PLGA/PLA, 441
Sporanox®, 129 overview, 377
Spray-congealed clarithromycin wax matrix, 206 particle engineering of large molecules, 442
Spray-congealing pilot-scale spray driers, 391
acetazolamide, 207 polymers, 406
equipment, 206 product quality, 400
excipients, 207 product residual solvent, 390
food and chemical industry, 205 production-scale equipment, 392
hydrophobic matrix, 206 pulmonary drug delivery, 435
inlet temperature, 207 free particle fractions (FPF), 436
microencapsulation, 206 inhalable powder, 437
molten formulation, 207 PulmoSphere™ solid foam particles, 437
operation principles, 205 residual solvents, 398
processing parameters, 205 scale-up methodologies, 400
processing steps, 206 atomization, 401
residence time, 206 droplet drying, 401, 402
thermodynamic changes, 207 lab scale to large production, 400
verapamil hydrochloride, 206 schematic diagram, 379
Spray distances, 549, 557 secondary drying, 398
Spray drying, 193, 293, 306 agitated bed driers, 399
amorphous solid dispersions (ASDD) (see Amorphous fluid bed drying, 399
solid dispersions (ASDD)) rotary driers, 400
atomization, 379 tray drying, 398
atomizers, 383, 384 solution preparation, 378
flash nozzles, 383 solvents, 393
four-fluid nozzle, 382 alcohols, 394
monodisperse droplet generators (MDGs), 382 aqueous systems, 395
692 Index
Spray drying (cont.) feature, 532

dichloromethane (DCM), 394 precipitation techniques (see SCF precipitation
ICH guidelines (see ICH guidelines on residual techniques)
organic solvents) Supersaturable SEDDS (S-SEDDS), 268
ketones, 394 Supersaturation, 192, 266–268, 270, 271, 279
mixture of organic solvents, 395 definition, 530
solid dispersions, 393 primary nucleation rate, 531
tetrahydrofuran (THF), 394 reduction in solvent powder, 531
vaporization enthalpies, 394 Supramolecular hydrogels preparation, 244, 245
water, 393 Supramolecular structures, 236
spray congealing, 438, 439 Surface area measurement, 485
Spray freeze drying (SFD) Surfactants, 9, 10, 182, 228, 229, 257, 258
advantages of, 465 Suspension-based systems, 19
amorphous high-potency danazol compositions, Sweetening agents, 183
491, 493 Synthetic poloxamers, 273
amorphous solid dispersions, 458 Syringe-based extrusion 3-dimensional printing, 276
ATMFD, 462 Systemic administration, 2
BCS Class IV compound, 458
continuous SFD processes, 462, 463 T
conventional lyophilization, 461, 462 Tablet-in-tablet, 615
disadvantages of, 465 Tacrolimus, 493
freeze-drying process, 459 Target product profile (TPP), 218
liposomal dry powder inhalation, 459 Targeted drug delivery, 631
process of, 460, 461 Taxol®, 11
proteins and peptides, 458 Taxotere®, 10, 229
spray drying (SD), 459 Tegretol®, 130
Spray freezing into liquid (SFL), 456 Telmisartan (TEL), 625
Spray freezing techniques, 442 Ternary HPBCD ASD, 605
Spritam®, 611 Thermal analysis methods, 298
S-SEDDS formulation, celecoxib, 281, 282 Thermal analytical analysis, 194
S-SMEDDS formulations, 276 Thermal behavior, 300
Stability improving strategies Thermal conductivity, 609
excipients addition, 239 Thermal decomposition, 50
physicochemical stability, 238 Thermal energy, 614
processing techniques adjustment, 240, 241 Thermodynamic models, 298
Stability testing Thermogravimetric analysis (TGA), 194, 348
chemical stability, 68–69 Thermolabile drugs, 167
conditions for products, 69, 71 Thermoplastic polymer, 613, 614
predicting stability, 71–72 Thin film freezing (TFF), 198, 211–212
stability monitoring, 67–68 advantages of, 470, 471
Stabilizers, 531 disadvantages of, 471
Steric stabilization, 161 parameters, 469
Sterilization methods, 223 TFF-template emulsion technology, 470
Stress conditions, 443 Time-of-flight secondary-ion mass spectrometry
Stressed digestion test, 268, 281 (ToF-SIMS), 484
Stroma, 12 T-junction microfluidic chip, 459
Subcutaneous (SC) administration, 219 Tolfenamic acid, 281
Sublingual mucosa, 7 Tonicity, 222
Succinylated gelatin, 202 Tonicity-modifying agents, 222
Sulfobutylether-β-cyclodextrin, 11, 230 Torcetrapib, 21, 22
Sumatriptan, 15 Trachea, 17
Supercritical antisolvent (SAS) precipitation, 589 Traditional polymers, 605
organic solvents, 545 Traditional solvent-based process, 188
Supercritical CO2 (scCO2), 532, 540, 541, 548, Transcorneal absorption, 12
553–556, 584, 588 Trans-epithelial transport, 17
Supercritical fluids (SCFs), 532 Transmembrane pH gradient, 234
commonly used SCFs, 532, 533 Transmission electron microscopy (TEM)
density of CO2, 532 measurements, 300
diffusivities, 532 Triethyl citrate (TEC), 602
Index 693
Triglycerides, 256 W
Trojan micron-sized particles, 20 Water/dehydrated alcohol cosolvent
Trojan particles, 24, 25 system, 229
Troup classification system (TCS), 346 Water–cosolvent system, 181
Tryptophan (TRP), 636 Water-soluble compounds, 119
Tween 80, 554, 560, 562, 576, 584 Weakly acidic/basic drugs, 4
Twin-screw extruders, 328 Wet and dry extrusion, 193
Twin-screw melt extrusion, 328 Wet milling techniques
Tyrosine (TYR), 636 categories, 156
cryomilling (see Cryogenic milling)
U definition, 156
Ultra cryomilling, 162, 163, 169, 170 heating/cooling cycles, 156
Ultrasonic dispersion devices, 541 issue, 156
Unadsorbed surfactants, 562, 563 liquid shear forces, 145
Uncontrolled crystallization/precipitation, 142 liquid thermal properties, 156
Ursodeoxycholic acid (UDCA), 578 mechanical sealing, 156
media milling (see Media (bead) wet milling)
V particle size, 156
Van’t Hoff equation method, 222 rotor–stator milling, 157
Vehicle selection and solubilization White–Vincent method, 222
cosolvents, 228
cyclodextrins, 229, 230 X
nonaqueous and oily vehicles, 230 X-ray diffraction (XRD), 194, 482
pH adjustment, 226 excipient interactions, 57
physical/chemical stability, 226 limits, 55
salt formation, 227 nondestructive test, 55
surfactants, 228, 229 parameter optimization, 87
Vemurafenib, 578 parameter selection, 55–56
Vemurafenib amorphous MBP, 578 PDF, 57–59
Venclexta (venetoclax), 361 polymorph screening, 56–57
Vent zones, 331
Verapamil hydrochloride wax microparticles, 206 Y
Verteporfin, 233 Young’s modulus, 161
Viscoelasticity, 51
Viscous dissipation, 335 Z
Viscous polymers, 604, 605 Zwitterionic, 257
Voriconazole (VCZ), 504–507

2022 - Williams - Formulating Poorly Water Soluble Drugs

Uploaded by

Copyright:

Available Formats

2022 - Williams - Formulating Poorly Water Soluble Drugs

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2022 - Williams - Formulating Poorly Water Soluble Drugs

Uploaded by

Copyright:

Available Formats

AAPS Advances in the Pharmaceutical Sciences Series 50

Robert O. Williams III

More information about this series at http://www.springer.com/series/8825

ISSN 2210-7371 ISSN 2210-738X (electronic)

# American Association of Pharmaceutical Scientists 2022

To Allison, Westley, Gwendolyn, and Lyla for your constant love

High-throughput screening (HTS) methodologies for lead identiﬁcation in

considering that knowledge in most of these areas is relatively new and

poorly water-soluble compounds entering development, the role of the for-

Austin, TX, USA Robert O. Williams III

1 Route-Speciﬁc Challenges in the Delivery

10 Spray-Drying Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

Zachary Warnken, Hugh D. C. Smyth, and Robert O. Williams III

Abstract site, and the excipients commonly used in

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 1

Fig. 1.1 Occurrence of

Fig. 1.2 The

Fig. 1.3 Transporter

Table 1.1 Detection of hemolysis by in vivo and in vitro methods

Fig. 1.5 Illustration of 2500

1000 dilution curve

Fig. 1.6 Illustration of the

Fig. 1.7 Anatomy of the

or inﬂammation (Tolman and Williams 2009; with non-ozone-depleting propellants such as

Fig. 1.8 “Effective 1.4 60

Respirable Deposition (%recovered)

differently. It is assumed that inorganic Method Capsule 1

• Glass vessel Equipment and Reagents

Objective ultimate physical dimension of the spray-

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 33

formulation process, it is important to perform the 1:11ΔS f ðT m 25Þ

Fig. 2.1 Calculated

Solubility Measurement on Amorphous 2019a, b) and naproxen sodium (Stukelj et al.

Dissolution Vessel Rotating

Drug Drug Pellet

Plastic Cap Plunger

Table 2.2 Advantages and disadvantages of mDSC and conventional DSC

assessment of drug–polymer or polymer–polymer occurring upon heating without sufﬁcient poly-

Polymorphic Transformations product for tableting led to the inclusion of stearic

Flory–Huggins regions C and D are metastable, and regions E

Fig. 2.5 Phase diagram for

Fig. 2.6 API thermal 110.0

ROA : Eudragit®L100-55 1:2

ROA–HPMCAS blends (Hughey et al. 2010). In dσ dγ

Fig. 2.8 Illustration of

mannitol. Samples of the R-isomer compact Parameter Selection

Fig. 2.9 XRD patterns of

Fig. 2.10 XRD patterns of

80:20 API:PVP Dispersion

Fig. 2.11 Representative

Fig. 2.14 Surface area Surface enlargement factor

dW DAðC s CÞ detailed summary of the theory behind BET iso-

Fig. 2.15 Artiﬁcial neural

withdrawn samples must be treated immediately Alternative Dissolution Studies

Fig. 2.16 Diagram of a Media USP II with dual paddle

Fig. 2.18 Illustration of MD-FIS structure

2.4.2 In Vivo Testing However, these studies are only indicative of

Fig. 2.20 Itraconazole 20

Notably, this proposed classiﬁcation of crystalli- ΔGIII ¼ ΔH III TΔSIII , ð3:3Þ