View Article Online / Journal Homepage / Table of Contents for this issue

REVIEW www.rsc.org/softmatter | Soft Matter

Cellulose nanowhiskers: promising materials for advanced applications

Stephen J. Eichhorn*
Received 23rd March 2010, Accepted 29th July 2010
DOI: 10.1039/c0sm00142b

This review covers the production, structure, properties and applications of nanowhiskers of cellulose.
It is shown that these nanowhiskers can be generated, from various plant sources, with transverse
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

dimensions as small as 3–30 nm, giving a high surface to volume ratio. Since the nanowhiskers are
rod-like, it is shown how they can be self-assembled into chiral nematic liquid crystalline structures, not
only in solution, but also in the dry state. The production of thin films of cellulose nanowhiskers, by
spin coating and in combination with polymer electrolytes, is also covered. A wide range of chemical
modification of cellulose nanowhiskers are reviewed; including 2,2,6,6-tetramethylpiperidine-1-oxyl
(TEMPO)-mediated oxidation, polymerisation from the surface using Reversible Addition–
Fragmentation chain Transfer (RAFT) and Atom Transfer Radical Polymerisation (ATRP), and the
rendering of the surface with cationic and anionic charge. The mechanical properties of cellulose
nanowhiskers will also be covered, including the measurement of their stiffness using both Raman
spectroscopy and Atomic Force Microscopy (AFM) measurements. The final part of the review will
cover the applications and potential industrial use of cellulose nanowhiskers; namely for
nanocomposite materials, thin films and other applications. Finally, some conclusions, including
perspectives and future developments will be presented.

1. Introduction candidate for replacing oil-based feedstocks. This is of course by

no-means a new concept. Indeed, the first ever plastics were
In recent times there has been a drive to utilise non-oil-based derived from cellulose (cellulose nitrate) in the mid-19th century.2
polymers due to a predicted diminishing supply of this feedstock During the depression in the 1930s, and also the Second World
for polymeric materials.1 Cellulose, being the most abundant War, supplies of raw materials were limited, and so the use of
material on the planet and also a renewable resource, is a prime natural materials, such as cellulose, increased.
Cellulose is produced by plants, trees, bacteria and some
School of Materials and the Northwest Composites Centre, University of animals (tunicates) via the condensation polymerisation of
Manchester, Sackville Street, Manchester, M60 4QD, UK. E-mail: s.j. glucose, which is a product of the photosynthesis process in
[email protected]; Fax: +44 (0)161 306 3586; Tel: +44
(0)161 5982
plants and trees. Long chains of anhydroglucose units, joined via
b-1,4-glycosidic linkages (C–O–C), are formed during this
process.3 Cellulose, in its native state (termed cellulose-I) in the
Dr Steve Eichhorn is a Reader in cell walls of plants and trees, forms a crystalline structure,
Polymer Physics and Biomate- whereby chains aggregate via hydrogen bonding.3 Two forms of
rials within the School of Mate- crystalline cellulose are known to occur in plants; namely cellu-
rials at the University of lose Ia and Ib.4 The crystal structures of these native forms of
Manchester and also the Director cellulose have recently been elucidated.5 These structures are
of Research for the Composites slight modifications of previously reported structures of cellu-
Centre. He has a first degree in lose-I,6 differing particularly in the hydrogen bonding network.
Physics (Leeds) and a Masters The crystalline state of cellulose can be therefore thought of as a
in Forestry and Paper Industries composite material, comprising two phases. All cellulose sources,
Technology (UMIST/Bangor). however, are not completely crystalline. The presence of para-
He conducted his PhD on the crystalline or amorphous cellulose is often significant, although it
‘‘Deformation Micromechanics varies from species to species. The presence of this amorphous
of Cellulose Fibres’’ and has portion to cellulose makes the structure susceptible to
Stephen J: Eichhorn published widely in the area of acid hydrolysis, and the eventual breakdown into individual
cellulosic and renewable mate- crystallites.
rials. Dr Eichhorn has particular The aggregated chains of cellulose form what are termed
expertise in natural fibre composites, particularly the use of elementary crystallites or fibrils.2 The measurement of the size of
spectroscopic and X-ray diffraction techniques to analysing these these elementary crystallites, or fibrils, has been extensively
materials in terms of their interfacial properties. He also has interests researched.7 What constitutes an elementary crystallite or fibril in
in biomimetic and other biological materials. He is a member of the terms of the number of chains has been a subject of debate for
Royal Society of Chemistry and also the Institute of Physics and the some time. Even their very existence has been disputed with
American Chemical Society. a number of protagonists,8 but also some antagonists to the

This journal is ª The Royal Society of Chemistry 2011 Soft Matter, 2011, 7, 303–315 | 303
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM. View Article Online

Fig. 2 Atomic Force Microscope image of cellulose nanowhiskers

produced using acid hydrolysis of cotton. The left hand scale gives height
information (units: nm Div1). Image courtesy of Mrs Rafeadah Rusli,
University of Manchester.

dimensions ranging from 3–30 nm.10 A distinction between what

is typically called Microcrystalline Cellulose (MCC) should also be
made. Whilst MCC is generated using acid hydrolysis of plant-
based cellulose,11 a much wider reaction window is used. Both the
terms ‘‘microcrystalline cellulose’’ and ‘‘cellulose microcrystals’’
were applied in cellulose nanowhiskers in the 1990s. For instance
a seminal paper by Gray et al. on this very subject refers to
cellulose nanowhiskers as ‘‘cellulose microcrystallites’’.12 It should
Fig. 1 The hierarchical structure of wood from the tree (top left) to the
trunk, cells or fibres, cell walls, fibrils and cellulose molecules (bottom
also be pointed out that cellulose nanowhiskers are not the same as
right). Artwork by Mark Harrington. Copyright University of Canterbury, nanofibrillar cellulose (also referred to as ‘‘microfibrillar cellulose’’
1996. or ‘‘microfibrillated cellulose’’). This form of cellulose, originally
developed by Turbak et al.,13 is generated by mechanical/chemical
means,14 or purely by chemical means,15 and produces long
idea.9 The concept of breaking down the cell wall into structural nanosized individualised fibrils and should not be confused with
components that represent these elementary crystallites or fibrils nanowhisker materials. A typical image (taken using Atomic
has in recent times become a topic of great interest. The hierar- Force Microscopy) of some cellulose nanowhiskers (produced by
chical structure of cellulose is best illustrated by considering the acid hydrolysis of cotton) is shown in Fig. 2.
full range of length scales. For wood this goes from the tree- It has been well known, for about 50 years or more, that stable
trunk, right down to the cellulose molecules. A schematic illus- suspensions of cellulose nanowhiskers can be produced by a harsh
trating this length-scale structural evolution is shown in Fig. 1. acid hydrolysis (usually sulfuric acid, although hydrochloric acid
It is possible to extract both the fibrils and nanocrystals from will also work) of plant material followed by ultrasonication. The
the cell walls of plants. This article will cover only the latter, with first report of such a process was by R anby in 1951.16 In 1953
sections on the production of cellulose nanowhiskers (also called a similar study was published, this time using X-ray diffraction to
nanocrystals), their self-assembly and chiral nematic nature, the ascribe lateral dimensions to the crystallites.17 Marchessault et al.
production of thin films of nanowhiskers, chemical modification showed that, owing to their rod-like nature, these nanowhiskers
and surface polymerisation, their mechanical properties and can form birefringent gels and liquid crystalline structures, remi-
finally applications in nanocomposites. niscent of spherulitic structures in polymers.18
Following the methods of Mukherjee and Woods,17 Revol
et al. showed that ‘‘cellulose microfibrils’’ (nanowhiskers) were
2. Production of cellulose nanowhiskers
shown to exhibit helicoidal self-ordering.19
Before describing their production, it is perhaps best to define Cellulose nanowhiskers can be produced, using the acid
what is meant by the term ‘‘cellulose nanowhisker’’. There are hydrolysis approach, from a range of initial cellulosic materials,
many names used to describe what will be defined, in this review, namely tunicate cellulose,20 bacterial cellulose,21 Kraft pulp,22
as a ‘‘cellulose nanowhisker’’; for example ‘‘cellulose nano- MCC,23 sugar beet (primary cell wall),24 softwood pulp,25
crystals’’, ‘‘cellulose nanofibres’’ and ‘‘cellulose whiskers’’. ramie,26 sisal,27 straw28 and cotton.12 The dimensions of cellulose
Perhaps, as a definition, ‘‘cellulose nanowhiskers’’ ought to be nanowhiskers do depend on the initial source.
defined in terms of the method by which they are produced, and The effect of processing conditions on the physical properties
also their lateral dimensions. In this respect, these materials are of nanowhiskers has been investigated.12 Beck-Candanedo et al.
a fibrous form of cellulose produced by the acid hydrolysis of showed that the length of nanowhiskers produced from wood
plant (or animal or bacterial) based cellulose, with lateral cellulose (black spruce acid sulfite pulp) was reduced by increasing

304 | Soft Matter, 2011, 7, 303–315 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

the hydrolysis time.29 It was also found, by the same group, that
increasing the acid to fibre ratio reduced the dimensions of the
nanowhiskers.29 Another group has also shown that the drying of
pulp, before acid hydrolysis, has an effect on the dimensions of the
nanowhiskers.30 An attempt to optimize the processing of nano-
whiskers from MCC derived from Norway spruce has been
reported.23 In this study the concentration of MCC and sulfuric
acid, the hydrolysis time, temperature and the ultrasonic treatment
time were all varied in order to arrive at optimal conditions in terms
of surface charge, dimensions, yield and birefringence.23 Rod-like
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

cellulose nanowhiskers, with a range of dimensions can be gener-

ated using this processing route.
A recent study by Elazzouzi-Hafraoui et al. measured the
dimensions of cellulose nanowhiskers produced from cotton, Fig. 3 Schematic of the chiral nematic structure of concentrated
microcrystalline cellulose and tunicate sources.10 A range of suspensions of cellulose nanowhiskers where P is the pitch of the chiral
techniques were used in order to do this, including transmission twist. Left-handed chirality is observed for these nanowhiskers, whereas
electron microscopy, wide angle X-ray diffraction and atomic right-handed is precluded. Image courtesy of Prof. D. G. Gray (McGill
force microscopy.10 An in-depth analysis of the polydispersity, University, Canada).
and the relationship with the size of the elementary crystallites in
these materials, was given.10 Tunicin was found to yield the
lowest polydispersity of lateral size (width), but the highest in thermodynamically. This is the cause of the formation of
terms of length. A low lateral association of elementary fibrils or a spontaneous nematic phase of rod-like particles. The effect was
crystallites was proposed to be the reason for this, whereas in the first observed much earlier than Onsager’s theoretical explana-
plant material aggregation was presumed to occur more readily, tion in 1937, for the tobacco-mosaic virus, by Bawden and Pirie33
leading to a large distribution of crystallites.10 The longer lengths where an isotropic low density phase and a nematic higher
observed, but also the highest polydispersity of this parameter density phase were found to co-exist in solution. Liquid crystal
for tunicate nanowhiskers, were thought to be due to the fact that forms were also observed, with similar structures to synthetic
plant celluloses are found to have periodic domains susceptible polymers. This phenomenon was later explained by Oster in 1950
to acid hydrolysis (presumably disordered or amorphous in terms of Onsager’s theory.34 As mentioned earlier in this
domains) that leads to a narrower distribution of crystallite article, Marchessault et al. subsequently showed that rod-like
lengths.10 cellulose crystallites would also form liquid crystal like gels.18 In
One result of the acid hydrolysis of cellulose into nanowhiskers exactly the same way as the tobacco mosaic virus, at low
is the presence of charge on their surface after processing. In the concentrations, cellulose nanowhiskers will form an isotropic
case of sulfuric acid hydrolysis, negatively charged (anionic) phase. However, at high enough concentrations they form what
sulfonate ester groups remain, leading to nanowhiskers that are is known as a chiral nematic phase, typical of a liquid crystalline
easy to disperse in aqueous solvents due to electrostatic repul- polymer. This structure is pictorially depicted in Fig. 3. The
sion. Nanowhiskers can be generated using hydrochloric acid, nanowhiskers, when forming such structures, occupy ordered
although a weaker charge density results from this processing domains in layers, with a twist in the mean orientation of the
route, leading to poorer dispersion in organic solvents. Recent nanowhiskers in each layer (or pitch) perpendicular to the
developments have enabled cationic surface charges to be cholesteric axis. This form gives rise to interesting optical prop-
generated on cellulose nanowhiskers, albeit by an additional erties.35 The appearance of a chiral nematic phase yields
reaction with epoxypropyltrimethylammonium chloride, and not a characteristic fingerprint when suspensions are viewed under
as a result of hydrolysis.31 cross-polarised light, an example of which is reported in Fig. 4.
These optical properties, also seen for cholesterol36 and deriva-
tives of cellulose,37 are due to the optical rotary power of the
3. Self-assembly and chiral nematic nature
cellulose nanowhiskers and their helicoidal packing. Interestingly
It has been well-known for some time that rod-like particles will, enough, the chiral nature of the nanowhiskers only exhibits itself
at a critical concentration, spontaneously form ordered struc- in left-handed chirality, and not right-handed (see Fig. 3). It is
tures. Onsager was the first to show that this spontaneous not only possible to achieve biphasic suspensions (isotropic/
ordering, or nematic phase transition, will occur with rod-like nematic), but also triphasic, when in the presence of dextrans.38
particles.32 The critical finding of his study was that rod-like Triphasic suspensions were shown to be isotropic–isotropic–
particles possess two components to their entropy; from their nematic, and were distinguished using blue dextrans.38 Unusual
translation and also from their orientation.32 Both of these liquid crystalline properties can be obtained from surfactant
components are coupled, and the parallel configuration of the stabilised cellulose nanowhisker apolar solvent suspensions (in
rods is favoured since the effective excluded volume is zero. cyclohexane), with pitches much lower than that seen in water.39
When there is a low density (or concentration) of rods, there is Nematic structures of cellulose nanowhiskers can also be taken
little to be gained from the free volume component, and so the from solution to the dry state, provided that careful and
translational entropy component is dominant. At high densities controlled drying/evaporation of solvent takes place. This allows
of rods, the entropy of parallel rods is lower, and so favoured the generation of iridescent films,40 whose reflected colour can be

This journal is ª The Royal Society of Chemistry 2011 Soft Matter, 2011, 7, 303–315 | 305
 View Article Online

of model films of cellulose nanowhiskers has been where inter-

actions between the model surface and other polymers have been
studied, using AFM (Atomic Force Microscopy) in colloid probe
mode.50 Cellulose nanowhiskers have been spin coated onto
a silica surface (oxidised silicon) and a colloidal probe, consisting
of an amorphous cellulose sphere attached to an AFM probe,
was brought increasingly close (until touching) to the cellulose
whisker surface.51 Most previous studies, using this technique,
used spin coated cellulose-II films,52–54 and so did not investigate
cellulose in its native state. It is important to investigate cellulose
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

interactions, in the native state, since this fundamental interac-

tion governs aggregation of cellulose fibres within the cell wall of
plants and the bonding between cellulose fibres within paper. The
basic principle of the measurement is that as the cellulosic sphere
Fig. 4 An example of a ‘‘fingerprint’’ texture of a chiral nematic phase
within the anisotropic region of a suspension of cellulose nanowhiskers
viewed under cross-polarised light. Image taken from Dong et al.46 Image
courtesy of Prof. Derek G. Gray (McGill University, Canada). Copyright
ACS Publishing (1996).

controlled using the ionic strength of the whisker–solvent

suspension,19,40 or by applying an electric field.41 The effect of an
electric field on the colours of these films has also been investi-
gated.42 Strong magnetic fields can also affect the chiral axis.43 A
comprehensive study on the effect of many parameters on the
chiral nematic properties of cellulose nanowhisker films from
suspension has been published by Pan et al.44
The wavelength of the reflected circular polarised light l from
films of cellulose nanowhiskers can be readily calculated using
a simple relationship suggested by de Vries, where

l ¼ nP (1)

and n is the refractive index of the material and P the pitch (see
Fig. 3). Eqn (1) is true when the films or suspensions are viewed
along the cholesteric axis, but if viewed obliquely then l is
reduced by a factor of cos q; where q is the viewing angle from the
cholesteric axis which is typically coincident with the normal to
the surface. Typically an electrolyte, such as NaCl, is added
before evaporation takes place in order to generate coloured
films, and applications of these materials will be discussed later.
A more recent development has been the discovery of parabolic
focal conics, liquid crystalline defect structures, in thin films of
dried-down cellulose nanowhiskers.45

4. Thin films of cellulose nanowhiskers

Model smooth surfaces of polymers are extremely useful for
understanding the interactions between materials since one can
isolate the effects of surface roughness and bulk properties. This Fig. 5 (a) A schematic of a typical experimental set-up for measuring the
is especially true for cellulose, as in the native state it is often non-contact and contact forces in an AFM (Atomic Force Microscope)
found in close proximity to, and with complex interactions with, in colloid probe mode and (b) normalized force profile between two
cellulose spheres immersed in 0.1 mM NaCl solution. The lines are fits of
other components (lignin, hemicellulose, extractives, etc.). The
DLVO theory to the data with Debye length 30 nm (calculated value, not
first thin films (30–40 nm) of cellulose to be made were deriva-
a fitting parameter) and Hamaker constant 8  1021 J. The solid line
tives, and were spin coated onto mica wafers.47 Thin films of represents the constant charge boundary condition, and the dashed line
cellulose nanowhiskers can also be produced using a similar the boundary of constant surface potential. The fitted surface potentials
technique.48 A large review of the structure, properties and are: 8/8 mV. Image (b) adapted from ref. 51 with permission and
applications of model cellulosic films has been recently published courtesy of Prof. M. W. Rutland (KTH, Sweden). Copyright Elsevier
by Kontturi et al.49 Perhaps one of the most useful applications (2006).

306 | Soft Matter, 2011, 7, 303–315 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

is brought close to the crystalline cellulose surface there will be an polyelectrolytes.56 In this case an AFM tip, without the presence
interaction between the two materials. This interaction is of a cellulosic material, was used. It was found that there was
measured as a force, calculated from the deflection of the AFM little difference between the force-deflection curves of cellulose
tip, and knowing the compliance of the cantilever beam. A surfaces with and without polyelectrolyte, although differences
schematic of a typical experimental set-up is shown in Fig. 5a. between single and multilayers were noted.56 However, the
Data collected during these experiments are similar, for a wide addition of a salt solution swelled the latter, giving rise to
range of studies. A typical example of a dataset is reported in a change in the force-deflection curve, which was attributed to
Fig. 5b. When the probe is at some distance from the surface then the transfer of polyelectrolyte to the AFM tip.56 Adhesion
the data suggest a double-layer interaction is occurring, which measurements between polydimethylsiloxane (PDMS) and a thin
further suggests that the surfaces are charged. The data in this film of cellulose nanowhiskers have also been recently carried
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

region can be fitted (as shown in Fig. 5b) using DLVO theory.55 out.57 It was shown that the forces were mostly dispersive, and
At small distances between the surfaces, short range repulsions that the work of adhesion decreased with increasing degree of
are observed, which are generally thought to be electrosteric order (or crystallinity) of the substrate, being the lowest for
repulsions,51 where overlapping charged polymeric chains extend cellulose nanowhiskers; cellulose nanowhiskers were compared
from the two surfaces. This argument has been disputed, with other cellulosic thin films.57 The effect of substrate on the
however, with some people not observing the repulsion,54 or even deposition of cellulose nanowhiskers during spin coating,
associating it with a swollen layer on the surface or surface whether anionic, cationic or amorphous, has also been investi-
roughness.53 One of the rationales for using cellulose nano- gated.58 A layer of titania was formed on a silicon wafer
whiskers for these experiments is, however, their low surface substrate, which is cationic, and yields well-dispersed cellulose
roughness,51 and so since these repulsive forces are still seen, it is nanowhiskers, yet with high concentrations.58 Anionic surfaces
likely that surface charges are the dominant effect. caused aggregation of the nanowhiskers, and an electroneutral
Some other studies of note, using cellulose-I thin films, have cellulose substrate with a low concentration homogenous
been the measurement of force-deflection curves of adsorbed dispersion.58 Other approaches to generating smooth monolayer
cellulose nanowhisker surfaces have been reported, such as by
the Langmuir–Blodgett method.59
Some other more recent applications of cellulose nanowhisker
films have been where layer-by-layer films have been produced
with antireflective properties.60 In order to generate an anti-
reflective coating it should have a thickness of l/4, where l is the
wavelength of the incident radiation. The refractive index of the
coating is simply the average of the refractive indices of the air
contained within the coating (if it is porous) na and the substrate
material nm i.e. nc ¼ (nanm)1/2. If a coating is to be effective, say on
a material such as glass (nm ¼ 1.5), then nc should be about 1.22.
However, most homogenous dieletric materials rarely have
a refractive index less than about 1.34. In order to obtain a crit-
ical value of nc nanoporous materials can be used, and for this
purpose cellulose nanowhiskers are ideal.60 Podsiadlo et al. have
shown that 12 layers of cellulose nanowhiskers are sufficient to
achieve high levels of transmission of light on a coated glass
surface. An image of the effect of a 12 layer cellulose whisker–
polyethyleneimine coating is shown in Fig. 6a.60 In Fig. 6b
transmittance data are also shown indicating the increase with
each successive layer of cellulose nanowhiskers. This method
holds potential for a wide variety of applications, including
transparent display devices and more efficient solar cells.

5. Mechanical and interfacial properties of cellulose

nanowhiskers
Cellulose nanowhiskers, as already described, are produced by
the acid hydrolysis of cellulosic material which is generally
semicrystalline, leaving what is ostensibly a single crystal of
material. The absolute value of the crystal modulus of cellulose
Fig. 6 (a) Photograph of a 100  300 microscope glass slide coated with 12 has long been debated. The first determination of the crystal
bilayers of cellulose whisker–polyethyleneimine as indicated; (b) light modulus of cellulose was by Meyer and Lotmar in 1936 who used
transmittance at 400 nm as a function of number of bilayers. Images a pyranose structure for cellulose, which was later proved to be
modified from ref. 60 with permission from Prof N. A. Kotov (University incorrect.61 Lyons then reported a value of 180 GPa for the chain
of Michigan, USA). Copyright ACS Publishing (2007). stiffness of cellulose, but he also used an incorrect term in his

This journal is ª The Royal Society of Chemistry 2011 Soft Matter, 2011, 7, 303–315 | 307
 View Article Online

derivation of the bending stiffness contribution.62 Treloar then molecular dynamics/mechanics methods, have typically given
corrected Lyons’ calculation and obtained a very low value for values in the range 100–160 GPa,67,68 and so very high values,
the chain modulus of a theoretical single chain of 56 GPa.63 This such as the one suggested by Diddens et al.,66 seem unlikely. It is
value was arrived at by considering the contributions to the chain conceivable therefore, that after removing the vast majority of
modulus from individual bond stretching and angle bending the amorphous fraction of cellulose, which is what happens
stiffnesses, but not, most notably, hydrogen bonding.63 Sakurada during hydrolysis of plant material into nanowhiskers, that the
et al. were the first to experimentally determine the crystal remaining material should have a modulus close to that of the
modulus of cellulose using X-ray diffraction, and a series of crystal. Measurements of the mechanical properties of small
weights to load fibre bundles in tension, arriving at a value of 138 objects, with dimensions of the size of cellulose nanowhiskers,
GPa for cellulose-I.64 This value has subsequently been are difficult. An estimation of the modulus of tunicate cellulose
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

confirmed via similar experiments on plant cellulose fibres.65 A nanowhiskers was published by Sturcova et al. in 2005.68 This
recent determination of the crystal modulus of cellulose using estimation was obtained by following the molecular deformation
sound velocities of X-rays in plant fibres yielded a very high value of the nanowhiskers, when embedded in an epoxy resin beam
of 220 GPa, much higher than any other experimental value.66 (deformed in 4-point bending), using a Raman spectroscopic
No consideration of the water content of the fibres was given in technique. It was first shown that this technique could be used to
the paper, which could affect the values obtained, but it is worth follow molecular deformation of cellulose in 1997 by Hamad
noting that there is some disagreement on the true value of the and Eichhorn.69 They deformed cellulose-II fibres in tension and
crystal modulus of plant cellulose. Theoretical estimates, using found that a characteristic peak, located at approximately
1095 cm1, shifted towards a lower wavenumber position, which
is an indication of molecular deformation.70 Subsequently many
studies have shown the same effect in a large range of cellulose
fibres and composites thereof.71 A value of 143 GPa was obtained
for the modulus of the tunicate cellulose nanowhiskers68 and
subsequently values in the range 50–100 GPa for acid hydrolysed
material.72 These values, however, were not derived from a direct
mechanical measurement of the stiffness of the nanowhiskers.
However, later measurements of the stiffnesses of tunicate
cellulose nanowhiskers (both unmodified and TEMPO modified
nanofibrils) have been reported73 yielding values of 150.7  28.8
GPa and 145.2  31.3 GPa respectively, in agreement with those
of Sturcova et al.68 The method used to obtain these values was
an AFM bending stiffness approach73 and involves measuring
the deflection of the probe as it is pressed to bend a suspended
affine object. A schematic of the technique is shown in Fig. 7a.
An AFM probe is brought into contact with a fibre, which is
suspended between two supports. In order to generate the
substrate so that fibres were suspended between such supports,
a photolithographic technique was applied to generate grooves
on a silicon wafer.73 An image of a suspended nanofibril lying on
top of this grooved surface is shown in Fig. 7b.73 The deflection
of a beam d(a) when measured at a single point (which is the case
in an AFM experiment) is given by the equation74
 3
F aðL  aÞ
dðaÞ ¼ (2)
3EI L
where I is the moment of inertia of the beam and E is the elastic
modulus of the beam. In the case where the AFM tip is applied in
the centre of the beam then eqn (2) becomes
F
E¼ (3)
192dðaÞI
which can be used to work out the modulus, given the force and
Fig. 7 (a) Schematic of an AFM bending stiffness measurement showing
measured deflection are known. In order to calculate the moment
the force applied by the tip (F), the distance from the support to the point
of inertia of a cellulose nanowhisker (or nanofibril) their cross-
of contact (a) and the separation of the supporting points for the object
subjected to a bending moment (L); (b) an atomic force micrograph of
sectional shape has to be taken into account.73 This was deter-
a single cellulose fibril prepared by TEMPO-oxidation of tunicate mined for cellulose nanowhiskers based on a parallelogram
cellulose on the microfabricated silicon wafer.73 Image (a) Copyright cross-sectional shape, as proposed by Helbert et al.75 In this case
ACS (2009) and (b) courtesy of Prof. T. Iwata (University of Tokyo, the second moment of inertia I can be determined using the
Japan), Copyright ACS(2009). equation

308 | Soft Matter, 2011, 7, 303–315 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

bh3 and any functionality on this surface will have a strong interac-
I¼ (4) tion with its surroundings.
12
Various chemical modifications of cellulose nanowhiskers
where b and h are the base width and height of the parallelogram. have been reported. Perhaps the area where chemical modification
Using literature values for whisker dimensions the value of E was and polymerisation from the surface of cellulose nanowhiskers
therefore determined.73 The cross-sectional area is thought to be have been most useful has been for nanocomposite materials. One
modified (from a parallelogram shape to an ellipse) during issue with cellulose nanocomposites, based on whisker materials,
TEMPO oxidation,75 and this was taken into account for the is their dispersibility and interaction with matrix materials. The
calculations of the elastic modulus of this form of nanofibre.73 literature and methods involved in obtaining dispersion and
Another paper on the determination of the transverse elastic strong matrix interactions are remarkably similar to that available
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

modulus of cellulose nanowhiskers from dissolving pulp, using for carbon nanotubes. This is not surprising since the issues are the
a similar method to Iwamoto et al. has been published by Lahiji same for nanocomposites comprising these materials. However, it
et al.76 They reported values in the range 18–50 GPa.76 is worth noting that cellulosic surfaces are more highly reactive
(through the –OH valency) than a carbon nanotube. Nevertheless,
they are presupposed to have a weak interaction with hydrophobic
6. Chemical modification and polymerisation from matrix materials given their inherent hydrophilicity. Cellulose
the surface of cellulose nanowhiskers nanowhiskers have been effectively dispersed in organic solvents
using adsorbed surfactants.77 Stabilised suspensions of cellulose
Cellulose nanowhiskers possess high surface to volume ratios, nanowhiskers have also been obtained by covalently binding
which mean that they have a highly reactive and readily func- poly(ethylene glycol) (PEG) with a terminal amine group (NH2) to
tionalisable surface. A simple calculation, based on treating the their surface.78 It should be pointed out that these nanowhiskers
whisker like a cylinder, shows that the surface to volume ratio is were modified using TEMPO oxidation, and so were heavily
2/r where r is the radius of the cylinder (neglecting the ends of decorated with carboxylic groups.78 Surface silanes (silylation)
cylinder as available surface). Therefore a dramatic decrease in r, have also been grafted onto cellulose nanowhiskers,79 which is
which one obtains when going from a typical micron-sized a method used for micron-sized cellulose fibres to enhance the
natural fibre (20 mm for cotton) to a cellulose nanowhisker interface with polymer matrices.80 This treatment alters the
(20 nm for a cotton produced nanowhisker), gives rise to morphology of the nanowhiskers, as can be seen from the trans-
a thousand-fold increase in the surface-to-volume ratio. This mission electron microscope images reported in Fig. 8.79 These
means that the material is represented almost entirely by surface, images show that swelling of the nanowhiskers occurred, which
was suggested to be due to some derivatization of the bulk of the
material.79 The dispersion of the nanowhiskers in organic solvents
was also shown to be affected by the modification.79 Flocculation
occurred for unmodified material, with increasing levels of
dispersion depending on the degree of substitution of silane
groups (see Fig. 9).79 Grafting of maleated polypropylene to the
surface of tunicin cellulose nanowhiskers has also been reported.
Modification of the surface of cellulose nanowhiskers using

Fig. 8 Typical transmission electron micrographs of silylated and

unmodified tunicin cellulose nanowhiskers: (a) unmodified nano- Fig. 9 Typical suspensions of silylated tunicin cellulose nanowhiskers in
whiskers, (b) silylated using isopropyldimethylchlorosilane (IPDM-SiCl) tetrahydrofuran (THF): (a) phase separated suspension of unmodified
for 16 hours with a molar ratio of SiCl/AGU of 8/1 and (c) silylated with nanowhiskers, (b) flocculated suspension corresponding to a degree of
n-octyldimethylchlorosilane (ODMSiCl) for 16 hours with a molar ratio substitution of 0.4 and (c) non-flocculated suspension corresponding to
of SiCl/AGU of 6/1. Images courtesy of Dr E. Fleury (INSA, Lyon, a degree of substitution of 0.6. Image courtesy of Dr E. Fleury (INSA,
France). Copyright Elsevier (2002). Lyon). Copyright Elsevier (2002).

This journal is ª The Royal Society of Chemistry 2011 Soft Matter, 2011, 7, 303–315 | 309
 View Article Online

alkenyl succinic anhydride (a common sizing agent used in

papermaking) has been reported.81 Significant hydrophobicity was
reported and dispersion in organic solvents and polystyrene was
achieved, although no mechanical properties of the nano-
composites formed with this matrix were reported.81
Long chain modification of cellulose nanowhiskers has also
been reported in the literature. One of the first such modifications
(not on cellulose, but starch nanocrystals) was the grafting of
stearic acid chloride and poly(ethylene glycol) methyl ether
(PEGME).82 In order to graft the latter, toluene 2,4-diisocyanate
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

was used.82 Both modifications resulted in crystalline hydro-

phobic shells around the hydrophilic starch particle, although the
stearic anhydride exhibited a clearer crystalline form than
PEGME.82 Ring opening polymerisations have also been Fig. 10 Schematic of the build-up in fibre stress sf(x) as a function of the
performed on cellulose nanowhiskers to form poly- position along its length (x) governed by Cox’s shear lag theory.95 lc is the
caprolactone,26,83,84 the interface with which has also been critical length of the fibre.
modelled using molecular dynamics.85 Long chain isocyanates
(n-octadecyl isocyanate, C18H37NCO) have also been grafted to up in fibre stress sf(x) over the ends of the fibres (or stress-
cellulose nanowhiskers using a novel solvent exchange process.86 transfer) to the maximum value (the fibre stress s0) as a function
In fact most of the chemical treatments so-far mentioned involve of the length along the fibre x is schematically depicted in Fig. 10
solvent exchanges which negate the environmental benefit of and can be simplified (if bl $ 10) to
using cellulose. To address this problem, gas phase esterification
of cellulose nanowhiskers has been recently reported.87 TEMPO sf(x) ¼ s0(1  exp (bx)) (7)
mediated oxidation of cellulose nanowhiskers has also been
reported, showing birefringence in a water suspension, indicative
of a liquid crystal.88 A number of other modifications of cellulose Half the critical length (lc/2) is marked on this schematic, and
nanowhiskers have been reported, namely single-step esterifica- represents the distance along the fibre that has to be traversed in
tion,89 click-chemistry modification with terminal amine order for the stress to reach a maximum, which plateaus gener-
groups,90 fluorescently labelled whiskers for bioimaging appli- ally around the centre of the fibre. If this maximum stress is
cations,91 grafted polystyrene chains by ATRP (Atom Transfer greater than the breaking stress of the fibre then it will break. If
Radical Polymerisation)92 and other means93 and triblock co- a short fibre is deformed within a composite, it will ultimately
polymers by reverse catalysis.94 fragment into smaller and smaller lengths, until a critical length is
reached, at which no further breaking of the fibre can occur, since
the stress never reaches the breaking stress of the fibre. This is
7. Nanocomposites using cellulose nanowhiskers a process known as fragmentation, the theory of which is the
Cellulose nanowhiskers lend themselves to composite reinforce- basis of our understanding of composites, and was developed by
ment due to their potential high aspect ratio. This is certainly true Kelly and Tyson in 1965.97 The critical length for a fibre is
of tunicate cellulose nanowhiskers, but perhaps less true of acid therefore a minimum length that a fibre must have in order for
hydrolysed plant material. It is worth spending some time to effective stress-transfer to take place. If the stress does not reach
discuss the effect of aspect ratio on composite reinforcement in the maximum value, for a given modulus of the reinforcement,
general. Stress, in a fibre composite, is known to transfer across then we do not effectively utilise the mechanical properties of
fibre ends by a process known as shear-lag, a theory that was that reinforcing phase. It can be shown that the critical length is
originally developed by Cox in 1952 for fibrous networks such as governed by the equation
paper.95 This theory essentially yields an efficiency factor hl sf d
which can be applied to the simple ‘‘rule-of-mixtures’’ law for lc ¼ (8)
2si
stress-transfer in a composite,96 namely
where sf is the strength of the fibre reinforcement, d is the
Ec ¼ hlEfVf + Em(1  Vf) (5) diameter of the fibre (so lc/d is effectively the aspect ratio, or
critical aspect ratio) and si is the interfacial shear stress between
where Ec is the modulus of the composite, Ef is the modulus of the matrix and the fibre. If we take an example of a cellulose
a reinforcing fibre (in this case the nanowhisker), Em is the nanowhisker in an epoxy resin matrix we can calculate the
modulus of the matrix material and Vf is the volume fraction of expected critical length for this particular resin system. Theo-
the reinforcing phase. The efficiency factor hl is governed by the retically sf z Ef/10 (typically) and since we know that Ef z 143
equation GPa from measurements using Raman spectroscopy and AFM
bending68,73 we can estimate their strength (no measurements of
tan hðbl =2Þ
hl ¼ 1  (6) the strength of cellulose nanowhiskers have been made to the
ðbl =2Þ
authors knowledge). Cellulose nanowhiskers have been reported
where l is the distance along the fibre from its end, and b is to have various lengths and diameters, but some typical values
a constant and a function of the composite geometry. The build- are given in Table 1.10

310 | Soft Matter, 2011, 7, 303–315 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

Table 1 Typical values of the lengths and diameters of some cellulose

nanowhiskers10 and their estimated aspect ratios and critical lengths using
eqn (8). Standard deviations are shown in brackets

Nanowhisker source Length/nm Diameter/nm Aspect ratio lc/nm

Cotton 131 (39) 21 (52) 6 1313

MCC 105 (35) 12 (42) 9 750
Tunicate 1073 (67) 28 (46) 38 1750

It is clear that only the length of tunicate nanowhiskers

Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

appears to approach the required critical value, with other

materials produced by acid hydrolysis being almost an order of
magnitude too small. The aspect ratio of cellulose nanowhiskers
is also small, compared to other nanofillers such as carbon
nanotubes (typically 300–1000) for instance.98 This is, however,
not the complete picture. Reinforcement of cellulose nano-
whisker composites began with the work of Favier et al.20 They
showed that dramatic reinforcements could be achieved based on
the fact that the nanowhiskers form a percolating network,
bound together by hydrogen bonds. Percolation theory was
originally developed to describe what were then called two-phase
‘‘molecular composites’’ by Takayanagi et al.99 This was later
adapted by Ouali et al.100 to incorporate some idea of topology,
and the fact that the rigid phase must reach a critical concen-
tration before percolation can take place. According to the
theory, the tensile modulus (G0 ) of the composite is given by the
equation
0 0 0
0 ð1  2j þ jXr ÞGs Gr þ ð1  Xr ÞjGr2
Gc ¼
(9)
1  Xr Gr0 þ ðXr  jÞGs0

where
 0:4
Xr  Xc
j ¼ Xr (10)
1  Xc
where Xr is the volume fraction of the rigid (r, nanowhisker)
component, G0 s and G0 r are the shear moduli of the neat soft (s,
polymer) and rigid (r, reinforcement) constituents, and j is the
volume fraction of nanowhiskers. This equation showed excel-
lent agreement with Favier’s data, and has recently been shown
to fit data for both tunicate101 and microcrystalline cellulose102
derived whisker nanocomposites. It is hard in a nanocomposite
to discriminate between matrix–filler and filler–filler interactions.
A recent study, using Raman spectroscopy, has attempted to do
this, with limited success.103 It was shown in this study that under
polarised light, pressed nanocomposites of tunicate nano-
whiskers in poly(vinyl acetate) resin exhibited light and dark
regions (see Fig. 11a).103 By rotating the specimen whilst
recording Raman spectra of main chain vibrations (parallel to
the cellulose chains) it was shown that the local orientation of
nanowhiskers could be determined i.e. oriented and non-oriented
Fig. 11 (a) Polarised light micrograph of a poly(vinyl alcohol)/tunicate
regions.103 This was achieved by plotting the intensity of whisker nanocomposite showing dark and light regions corresponding to
a Raman band located at approximately 1095 cm1, which isotropic and oriented domains of nanowhiskers; the intensity of the
corresponds to the C–O stretch modes in the cellulose main Raman band located at approximately 1095 cm1 as a function of rota-
chain,104 as a function of rotation angle.103 Where nanowhiskers tion angle for (b) dark regions of the sample where isotropic domains
are oriented the peak intensity was found to be a maximum, of nanowhiskers are present and (c) light regions where oriented domains
commensurate with the orientation of the nanowhiskers, and of nanowhiskers are present.103Copyright American Chemical Society
invariant in regions where isotropic random dispersions exist (see (2010).
Fig. 11b).103 The local mechanical properties of regions of

This journal is ª The Royal Society of Chemistry 2011 Soft Matter, 2011, 7, 303–315 | 311
 View Article Online

randomly dispersed and oriented nanowhiskers were shown to be (DMF), can redisperse cellulose nanowhiskers from the freeze-
different.103 In the areas of isotropic nanowhisker dispersion, dried state,125 although only single phase films have been gener-
nanowhisker–nanowhisker interactions were thought to domi- ated, and not composites per se.
nate, whereas matrix–nanowhisker interactions were thought to A less conventional route of producing a cellulose nanowhisker
be dominant where they are oriented. Moisture was shown to composite material has recently been reported, whereby
interrupt the matrix–whisker and whisker–whisker interactions, a template of nanowhiskers is generated and then subsequently
although it was impossible to discriminate the two.103 This infused with a polymer resin.101,126 The process involves generating
perhaps highlights the fact that the interface between cellulose a homogenous aqueous dispersion of nanowhiskers followed by
nanowhiskers and between the nanowhiskers and the matrix a gelation via solvent exchange with a water miscible solvent. This
material is also affected by both the processing conditions, and gel is then immersed in a polymer solvent to form the composite.
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

the environment under which they are tested. Many different This facile approach has been shown to yield a wide-range of
approaches to the fabrication of cellulose nanowhisker composites with interesting stimuli-responsive properties (e.g. to
composites have been reported, which will be summarised respond to the presence of water).101 More conventionally, and
shortly. A wide range of traditional polymer matrices have also perhaps in-line with traditional composite processing routes,
been used for the fabrication of cellulose nanocomposites attempts to extrude cellulose nanowhiskers in thermoplastic
including epoxy resins,68,72,105 polypropylene,106 poly(vinyl polymers, such as starch have been reported.127 Grafting of the
acetate),101,103,107 poly(vinyl alcohol),108 poly(vinyl chloride),109 cellulose nanowhiskers can also aid extrusion. Long chain fatty
carboxymethyl cellulose,110 cellulose acetate butyrate,111 poly- acids have been used to successfully extrude cellulose nano-
(caprolactone),26,84,112 poly(oxyethylene),113 polyethylene114 and whiskers with low density polyethylene.114 Long chain poly-
polyurethane.115 Biopolymers such as starch-based plastics,84,116 caprolactone grafted cellulose nanowhiskers have also been
chitosan,117 other plant polysaccharides,118 regenerated cellu- reported to be readily thermoformed (by compression and injec-
lose,119 poly(lactic acid)120 and polyhydroxybutyrates121 have tion moulding).128
also been reported as suitable matrix materials for cellulose Perhaps the most exciting development in processing cellulose
nanowhiskers. nanowhisker nanocomposites has been the use of electrospinning
There are a range of processing routes available for the to generate fibres. Electrospinning is a technique which relies on
production of cellulose nanocomposites. It is worth pointing out the use of a highly charged polymer solution or melt, which
that whilst some of them, such as solution casting, can produce deforms due to the repulsion of like charges at the surface. This
good properties, they are unlikely to be adopted as a large charge repulsion eventually overcomes the surface tension of the
industrial scale process due to the presence of volatile solvents. It polymer and a fine jet is ejected towards either an earthed or
is more likely that commodity markets, where the properties charged target at high speed, generating fibres from the nanoscale
override the need for mass production, will be the place where (20 nm is readily achievable) to several microns in diameter. A
this processing route is adopted. More conventional routes of number of reviews of the range of polymers that can be electro-
composites processing, such as injection moulding, extrusion and spun, and their applications have been recently published129 but
pressing will ultimately have to be the methods for large-scale we will concentrate on fibres containing cellulose nanowhiskers.
production of cellulose nanocomposites. Also key to the manu- The first report of the successful production of a nano-
facture of composites by these routes is the large-scale produc- composite electrospun fibre was with poly(ethylene oxide) and
tion of cellulose nanowhiskers. This latter issue has not yet been bacterial cellulose derived nanowhiskers.130 Increases in the
addressed to the author’s knowledge. mechanical properties of fibres when loaded with cellulose
Cellulose nanowhiskers, produced by sulfuric acid hydrolysis, nanowhiskers were observed, and alignment and clustering of the
have negatively charged sulfate ester groups on them, which nanowhiskers were also noted from transmission electron
make them readily dispersible in most organic solvents. microscopy images (see Fig. 12). Cellulose nanowhiskers have
However, when no charge is present, for instance when hydro-
chloric acid is used as a hydrolysis medium, then dispersibility is
difficult as the nanowhiskers have a high affinity for each other.
This aggregation is exacerbated by the process of freeze-drying,
which is often used as a final stage in the production of cellulose
nanowhiskers. That said, some impressive properties can be
achieved from dispersing cellulose nanowhiskers in water soluble
polymers, such as latexes.20,28,122 It is possible, by solvent
exchange processes, to incorporate cellulose nanowhiskers into
organic solvents, for subsequent dispersion in polymeric resins.
This has been achieved to a certain extent, from the freeze-dried
state, although with some aggregation of the nanowhiskers, by
using toluene and then into polypropylene.123 Formic acid has
been shown to give much better dispersions of nanowhiskers, and Fig. 12 Transmission electron microscope images of bacterial cellulose
combinations, by solution casting, with fully conjugated polymers, nanowhiskers embedded in electrospun poly(ethylene oxide) fibres at
such as polyaniline (PANI) and poly(p-phenyleneethynylene) a weight fraction of 0.4 wt% where (a) orientation and (b) entanglement
(PPE) have been reported.124 It has also been shown that aprotic of the nanowhiskers have taken place. Image courtesy of Prof. H.-J. Jin
solvents, like dimethyl sulfoxide (DMSO) and dimethylfuran (Inha University, Korea). Copyright Wiley-VCH (2007).

312 | Soft Matter, 2011, 7, 303–315 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

also been spun with amorphous polymers, such as polystyrene.131 aspect ratio. It has been possible in recent years to generate
An improvement in the storage modulus of the fibres was noted, composite nanofibres from carbon nanotubes alone, due to their
although it was not possible to image, using TEM, the presence high aspect ratios and subsequent entanglement.136 This, to the
of the nanowhiskers due to the low diffraction contrast with author’s knowledge has not been achieved using cellulose
the polymer; this did not improve even with uranyl acetate nanowhiskers, although it could be a future development in the
stained nanowhiskers.131 Low crystallinity thermoplastics, such as area. Thin film coatings on materials, using cellulose nano-
poly(lactic acid) have also been electrospun with cellulose nano- whiskers, for modifying adhesion and optical properties (reflec-
whiskers.132 Relationships between the increase in the strength of tion and colour) could be an area that sees significant
the fibres and their reduced diameter, when nanowhiskers were development in the future. Facile routes to generating films of
added, were made, and the nanowhiskers themselves were found controlled thickness are now available, and many areas of
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

to migrate to the surface of the electrospun fibres.132 Other poly- applications (security tagging, solar cells and biomedical devices)
mers that have been electrospun with incorporated cellulose are envisaged because of this. On the fundamental science of
nanowhiskers have been poly(acrylic acid),133 polycaprolactone134 cellulose nanowhiskers, the ability to now make thin films using
and regenerated cellulose.135 Although large quantities of material these materials opens up opportunities to understand the inter-
have not been made using electrospinning, this route to the faces between polymers and cellulose. For instance, complex
production of nanocomposite fibres is promising. The ability to interactions take place between plant cell-wall polysaccharides,
orient the nanowhiskers offers the opportunity to generate truly and our understanding of these interactions could be better
novel fibres, which could be incorporated into composite materials approached by generating biological mimics of the structure. The
in a hierarchical manner. development of AFM force probe microscopy for this purpose is
particularly exciting. In conclusion then it is clear that cellulose
nanowhiskers offer much promise for advanced applications.
8. Future perspectives and conclusions
Although much research needs to be carried out in order to
One thing that we must always remember, when it comes to realise those applications, the area is likely to expand due to
cellulose, is that it is cheap and abundant. Both of these are a focus on renewable materials. Another area that shows promise
primary considerations in any industrial application. The ability is the ability to generate other nanoparticles, aided by cellulose
to take a cheap and abundant material, and generate products nanowhiskers.137 Finally, the use of cellulose nanowhiskers for
with value-added properties, is desirable. The field of cellulose biomedical applications has large potential. Some work has been
nanowhiskers offers all of these properties, and therefore it is published on using ordered monolayers of cellulose nano-
likely, with the present emphasis on renewable materials, that we whiskers to direct tissue growth,138 but this remains a topic for
will see a large increase in research efforts in this area. The ability future work.
to generate highly functional nanomaterials within areas such as
smart coatings, pharmaceutical applications, nanocomposites,
electronic materials is clearly evident. Translation of current Acknowledgements
technology is now key to the development of natural nano- The author wishes to thank the Malaysian Government, EPSRC
materials like cellulose nanowhiskers. If the range of applications (EP/F036914/1; EP/F028946/1) for funding and to Prof. Derek
is to be truly exploited industrially then we need to be able to Gray for helpful suggestions on the manuscript.
generate material on an industrial scale, and to be able to make
materials with consistent properties. Cellulose nanowhiskers, like
carbon nanotubes, have encountered similar hurdles in their Notes and references
development as truly industrial materials (aggregation and 1 S. J. Eichhorn and A. Gandini, MRS Bull., 2010, 35, 187–193.
dispersion, ease of manufacture, etc.). However, it is true to say 2 R. J. Young and P. A. Lovell, Introduction to Polymers, CRC Press,
that carbon nanotubes are now at a stage where they are being 1991.
3 D. Klemm, T. Heinze, U. Heinze, K. J. Edgar, B. Philip and
sold commercially as a product; the same cannot be yet said of P. Zugenmaier, Comprehensive Cellulose Chemistry, Wiley-VCH,
cellulose nanowhiskers, although that will come in time. Carbon 2nd edn, 2009.
nanotubes are readily processable using melt extrusion and 4 R. H. Atalla and D. L. Vanderhart, Science, 1984, 223, 283–285;
pressing thermoplastic technology. Cellulose nanowhiskers on R. H. Atalla and D. L. VanderHart, Solid State Nucl. Magn.
Reson., 1999, 15, 1–19; J. Sugiyama, R. Vuong and H. Chanzy,
the other hand have significant problems in this respect, mainly Macromolecules, 1991, 24, 4168–4175.
associated with their inherent hydrophilicity. Efforts to over- 5 Y. Nishiyama, J. Sugiyama, H. Chanzy and P. Langan, J. Am.
come this problem need to be stepped-up if they are to make it Chem. Soc., 2003, 125, 14300–14306; Y. Nishiyama, P. Langan
and H. Chanzy, J. Am. Chem. Soc., 2002, 124, 9074–9082;
into the composites markets. Nevertheless, since cellulose
Y. Nishiyama, G. P. Johnson, A. D. French, V. T. Forsyth and
nanowhiskers are charged (using sulfuric acid production P. Langan, Biomacromolecules, 2008, 9, 3133–3140.
methods), they can be readily dispersed in aqueous solvents. This 6 S. Nishikawa and S. Ono, Proc. Tokyo Math.-Phys. Soc., 1913, 7,
ability to disperse has advantages over carbon nanotubes 131–138; K. H. Meyer and H. Mark, Ber. Dtsch. Chem. Ges.,
1928, 61, 593–614; K. H. Meyer and L. Misch, Helv. Chim. Acta,
(without modification at least). Crystalline cellulose has a density 1937, 20, 232–244; K. H. Gardner and J. Blackwel, Biopolymers,
of about 1.6 g cm3, and, as we have seen, nanowhiskers have 1974, 13, 1975–2001; A. Sarko and R. Muggli, Macromolecules,
a modulus of about 140 GPa. Carbon nanotubes have a modulus 1974, 7, 486–494.
7 J. Hengstenberg and H. Mark, Z. Kristallogr. Kristallgeom.
in excess of 1 TPa, with a slightly lower density of 1.3–1.4 g cm3,
Kristallphys. Kristallchem., 1928, 69, 271–284; H. Chanzy,
and so by far out-compete cellulose mechanically. Another area K. Imada, A. Mollard, R. Vuong and F. Barnoud, Protoplasma,
where carbon nanotubes out-compete cellulose nanowhiskers is 1979, 100, 303–316.

This journal is ª The Royal Society of Chemistry 2011 Soft Matter, 2011, 7, 303–315 | 313
 View Article Online

8 A. N. J. Heyn, J. Ultrastruct. Res., 1969, 26, 52–68; J. Blackwell and 47 R. D. Neuman, J. M. Berg and P. M. Claesson, Nord. Pulp Pap. Res.
F. J. Kolpak, Macromolecules, 1975, 8, 322–326. J., 1993, 8, 96–104.
9 I. Nieduszy and R. D. Preston, Nature, 1970, 225, 273; 48 C. D. Edgar and D. G. Gray, Cellulose, 2003, 10, 299–306.
D. F. Caulfield, Text. Res. J., 1971, 41, 267; R. D. Preston, 49 E. Kontturi, T. Tammelin and M. Osterberg, Chem. Soc. Rev., 2006,
J. Microsc. (Oxford, U. K.), 1971, 93, 7; W. W. Franke and 35, 1287–1304.
B. Ermen, Z. Naturforsch., B: Anorg. Chem. Org. Chem. Biochem. 50 W. A. Ducker, T. J. Senden and R. M. Pashley, Nature, 1991, 353,
Biophys. Biol., 1969, 24, 918. 239–241.
10 S. Elazzouzi-Hafraoui, Y. Nishiyama, J. L. Putaux, L. Heux, 51 J. Stiernstedt, N. Nordgren, L. Wagberg, H. Brumer, D. G. Gray
F. Dubreuil and C. Rochas, Biomacromolecules, 2008, 9, 57–65. and M. W. Rutland, J. Colloid Interface Sci., 2006, 303, 117–123.
11 O. A. Battista and P. A. Smith, Ind. Eng. Chem., 1962, 54, 20–29. 52 M. W. Rutland, A. Carambassis, G. A. Willing and R. D. Neuman,
12 X. M. Dong, J. F. Revol and D. G. Gray, Cellulose, 1998, 5, 19–32. Colloids Surf., A, 1997, 123, 369–374; A. Carambassis and
13 A. F. Turbak, F. W. Snyder and K. R. Sandberg, J. Appl. Polym. M. W. Rutland, Langmuir, 1999, 15, 5584–5590; S. Zauscher and
Sci.: Appl. Polym. Symp., 1983, 37, 815–827. D. J. Klingenberg, Colloids Surf., A, 2001, 178, 213–229;
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

14 M. Paakko, M. Ankerfors, H. Kosonen, A. Nykanen, S. Ahola, K. Theander, R. J. Pugh and M. W. Rutland, J. Colloid Interface
M. Osterberg, J. Ruokolainen, J. Laine, P. T. Larsson, O. Ikkala Sci., 2005, 291, 361–368; J. Stiernstedt, H. Brumer, Q. Zhou,
and T. Lindstrom, Biomacromolecules, 2007, 8, 1934–1941; T. T. Teeri and M. W. Rutland, Biomacromolecules, 2006, 7, 2147–
K. Abe, S. Iwamoto and H. Yano, Biomacromolecules, 2007, 8, 2153.
3276–3278. 53 S. Zauscher and D. J. Klingenberg, J. Colloid Interface Sci., 2000,
15 T. Saito, Y. Nishiyama, J. L. Putaux, M. Vignon and A. Isogai, 229, 497–510.
Biomacromolecules, 2006, 7, 1687–1691. 54 S. M. Notley, B. Pettersson and L. Wagberg, J. Am. Chem. Soc.,
16 B. G. R anby, Discuss. Faraday Soc., 1951, 11, 158–164, discussion 2004, 126, 13930–13931.
208–113. 55 E. J. W. Verwey and J. T. G. Overbeek, Theory of the Stability of
17 S. M. Mukherjee and H. J. Woods, Biochim. Biophys. Acta, 1953, 10, Lyophobic Colloids: The Interaction of Sol Particles having an
499–511. Electric Double Layer, Elsevier, New York, 1948; B. V. Derjaguin
18 R. H. Marchessault, F. F. Morehead and N. M. Walter, Nature, and L. Landau, Acta Physicochim. URSS, 1941, 14, 633.
1959, 184, 632–633. 56 J. Lefebvre and D. G. Gray, Cellulose, 2005, 12, 127–134.
19 J. F. Revol, L. Godbout, X. M. Dong, D. G. Gray, H. Chanzy and 57 M. Eriksson, S. M. Notley and L. Wagberg, Biomacromolecules,
G. Maret, Liq. Cryst., 1994, 16, 127–134. 2007, 8, 912–919.
20 V. Favier, H. Chanzy and J. Y. Cavaille, Macromolecules, 1995, 28, 58 E. Kontturi, L. S. Johansson, K. S. Kontturi, P. Ahonen,
6365–6367. P. C. Thune and J. Laine, Langmuir, 2007, 23, 9674–9680.
21 J. Araki and S. Kuga, Langmuir, 2001, 17, 4493–4496; M. Roman 59 Y. Habibi, L. Foulon, V. Aguie-Beghin, M. Molinari and
and W. T. Winter, Biomacromolecules, 2004, 5, 1671–1677. R. Douillard, J. Colloid Interface Sci., 2007, 316, 388–397.
22 J. Araki, M. Wada, S. Kuga and T. Okana, J. Wood Sci., 1999, 45, 60 P. Podsiadlo, L. Sui, Y. Elkasabi, P. Burgardt, J. Lee, A. Miryala,
258–261. W. Kusumaatmaja, M. R. Carman, M. Shtein, J. Kieffer,
23 D. Bondeson, A. Mathew and K. Oksman, Cellulose, 2006, 13, 171– J. Lahann and N. A. Kotov, Langmuir, 2007, 23, 7901–7906.
180. 61 K. H. Meyer and W. Lotmar, Helv. Chim. Acta, 1936, 19, 68–86.
24 E. Dinand, H. Chanzy and M. R. Vignon, Food Hydrocolloids, 1999, 62 W. J. Lyons, J. Appl. Phys., 1959, 30, 796–797.
13, 275–283. 63 L. R. G. Treloar, Polymer, 1960, 1, 290–303.
25 J. Araki, M. Wada, S. Kuga and T. Okano, Colloids Surf., A, 1998, 64 I. Sakurada, Y. Nukushina and T. Ito, J. Polym. Sci., 1962, 57, 651–
142, 75–82. 660.
26 Y. Habibi, A. L. Goffin, N. Schiltz, E. Duquesne, P. Dubois and 65 T. Nishino, K. Takano and K. Nakamae, J. Polym. Sci., Part B:
A. Dufresne, J. Mater. Chem., 2008, 18, 5002–5010. Polym. Phys., 1995, 33, 1647–1651.
27 G. Siqueira, J. Bras and A. Dufresne, Biomacromolecules, 2009, 10, 66 I. Diddens, B. Murphy, M. Krisch and M. Muller, Macromolecules,
425–432. 2008, 41, 9755–9759.
28 W. Helbert, J. Y. Cavaille and A. Dufresne, Polym. Compos., 1996, 67 K. Tashiro and M. Kobayashi, Polymer, 1991, 32, 1516–1530;
17, 604–611. R. J. Marhofer, S. Reiling and J. Brickmann, Ber. Bunsen-Ges.
29 S. Beck-Candanedo, M. Roman and D. G. Gray, Phys. Chem., 1996, 100, 1350–1354; S. J. Eichhorn and
Biomacromolecules, 2005, 6, 1048–1054. G. R. Davies, Cellulose, 2006, 13, 291–307; M. Bergenstrahle,
30 E. Kontturi and T. Vuorinen, Cellulose, 2009, 16, 65–74. L. A. Berglund and K. Mazeau, J. Phys. Chem. B, 2007, 111,
31 M. Hasani, E. D. Cranston, G. Westman and D. G. Gray, Soft 9138–9145.
Matter, 2008, 4, 2238–2244. 68 A. Sturcova, G. R. Davies and S. J. Eichhorn, Biomacromolecules,
32 L. Onsager, Ann. N. Y. Acad. Sci., 1949, 51, 627–659. 2005, 6, 1055–1061.
33 F. C. Bawden and N. W. Pirie, Proc. R. Soc. London, Ser. B, 1937, 69 W. Y. Hamad and S. Eichhorn, J. Eng. Mater. Technol., 1997, 119,
123, 274–320. 309–313.
34 G. Oster, J. Gen. Physiol., 1950, 33, 445–473. 70 R. J. Young and S. J. Eichhorn, Polymer, 2007, 48, 2–18.
35 J. F. Revol, H. Bradford, J. Giasson, R. H. Marchessault and 71 S. J. Eichhorn, M. Hughes, R. Snell and L. Mott, J. Mater. Sci. Lett.,
D. G. Gray, Int. J. Biol. Macromol., 1992, 14, 170–172. 2000, 19, 721–723; S. J. Eichhorn, J. Sirichaisit and R. J. Young, J.
36 H. Devries, Acta Crystallogr., 1951, 4, 219–226. Mater. Sci., 2001, 36, 3129–3135; S. J. Eichhorn, R. J. Young and
37 R. S. Werbowyj and D. G. Gray, Mol. Cryst. Liq. Cryst., 1976, 34, W. Y. Yeh, Text. Res. J., 2001, 71, 121–129; S. J. Eichhorn and
97–103. R. J. Young, Compos. Sci. Technol., 2004, 64, 767–772; K. Kong
38 S. Beck-Candanedo, D. Viet and D. G. Gray, Macromolecules, 2007, and S. J. Eichhorn, Polymer, 2005, 46, 6380–6390; B. Mottershead
40, 3429–3436. and S. J. Eichhorn, Compos. Sci. Technol., 2007, 67, 2150–2159;
39 S. Elazzouzi-Hafraoui, J. L. Putaux and L. Heux, J. Phys. Chem. B, K. Kong and S. J. Eichhorn, J. Macromol. Sci., Part B: Phys.,
2009, 113, 11069–11075. 2005, 44, 1123–1136; W. T. Y. Tze, S. C. O’Neill, C. P. Tripp,
40 J. F. Revol, L. Godbout and D. G. Gray, J. Pulp Pap. Sci., 1998, 24, D. J. Gardner and S. M. Shaler, Wood Fiber Sci., 2007, 39, 184–
146–149. 195; W. T. Y. Tze, D. J. Gardner, C. P. Tripp and S. C. O’Neill, J.
41 Y. Nishio, T. Kai, N. Kimura, K. Oshima and H. Suzuki, Adhes. Sci. Technol., 2006, 20, 1649–1668; P. Peetla,
Macromolecules, 1998, 31, 2384–2386. K. C. Schenzel and W. Diepenbrock, Appl. Spectrosc., 2006, 60,
42 D. Bordel, J. L. Putaux and L. Heux, Langmuir, 2006, 22, 4899–4901. 682–691; N. Gierlinger, M. Schwanninger, A. Reinecke and
43 X. M. Dong and D. G. Gray, Langmuir, 1997, 13, 3029–3034. I. Burgert, Biomacromolecules, 2006, 7, 2077–2081.
44 J. Pan, W. Hamad and S. K. Straus, Macromolecules, 2010, 43, 72 R. Rusli and S. J. Eichhorn, Appl. Phys. Lett., 2008, 93, 033111.
3851–3858. 73 S. Iwamoto, W. H. Kai, A. Isogai and T. Iwata, Biomacromolecules,
45 M. Roman and D. G. Gray, Langmuir, 2005, 21, 5555–5561. 2009, 10, 2571–2576.
46 X. M. Dong, T. Kimura, J. F. Revol and D. G. Gray, Langmuir, 74 S. Timoshenko and D. H. Young, Elements of the Strength of
1996, 12, 2076–2082. Materials, D Van Nostrand, London, 1968.

314 | Soft Matter, 2011, 7, 303–315 This journal is ª The Royal Society of Chemistry 2011
 View Article Online

75 W. Helbert, Y. Nishiyama, T. Okano and J. Sugiyama, J. Struct. 3903; M. Samir, A. M. Mateos, F. Alloin, J. Y. Sanchez and
Biol., 1998, 124, 42–50. A. Dufresne, Electrochim. Acta, 2004, 49, 4667–4677.
76 R. R. Lahiji, X. Xu, R. Reifenberger, A. Raman, A. Rudie and 114 A. J. de Menezes, G. Siqueira, A. A. S. Curvelo and A. Dufresne,
R. J. Moon, Langmuir, 2010, 26, 4480–4488. Polymer, 2009, 50, 4552–4563.
77 L. Heux, G. Chauve and C. Bonini, Langmuir, 2000, 16, 8210–8212. 115 N. E. Marcovich, M. L. Auad, N. E. Bellesi, S. R. Nutt and
78 J. Araki, M. Wada and S. Kuga, Langmuir, 2001, 17, 21–27. M. I. Aranguren, J. Mater. Res., 2006, 21, 870–881; X. D. Cao,
79 C. Gousse, H. Chanzy, G. Excoffier, L. Soubeyrand and E. Fleury, Y. Habibi and L. A. Lucia, J. Mater. Chem., 2009, 19, 7137–7145.
Polymer, 2002, 43, 2645–2651. 116 D. G. Liu, T. H. Zhong, P. R. Chang, K. F. Li and Q. L. Wu,
80 A. K. Bledzki and J. Gassan, Prog. Polym. Sci., 1999, 24, 221–274; Bioresour. Technol., 2010, 101, 2529–2536; Y. Chen, C. H. Liu,
A. K. Bledzki, S. Reihmane and J. Gassan, J. Appl. Polym. Sci., P. R. Chang, X. D. Cao and D. P. Anderson, Carbohydr. Polym.,
1996, 59, 1329–1336. 2009, 76, 607–615; M. N. Angles and A. Dufresne,
81 H. H. Yuan, Y. Nishiyama, M. Wada and S. Kuga, Macromolecules, 2000, 33, 8344–8353; M. N. Angles and
Biomacromolecules, 2006, 7, 696–700. A. Dufresne, Macromolecules, 2001, 34, 2921–2931; X. D. Cao,
Published on 31 August 2010. Downloaded by Mahidol University on 11/9/2018 1:42:44 AM.

82 W. Thielemans, M. N. Belgacem and A. Dufresne, Langmuir, 2006, Y. Chen, P. R. Chang, M. Stumborg and M. A. Huneault,
22, 4804–4810. J. Appl. Polym. Sci., 2008, 109, 3804–3810; X. Cao, Y. Chen,
83 M. Krouit, J. Bras and M. N. Belgacem, Eur. Polym. J., 2008, 44, P. R. Chang, A. D. Muir and G. Falk, eXPRESS Polym. Lett.,
4074–4081. 2008, 2, 502–510; I. Kvien, J. Sugiyama, M. Votrubec and
84 N. Lin, G. J. Chen, J. Huang, A. Dufresne and P. R. Chang, J. Appl. K. Oksman, J. Mater. Sci., 2007, 42, 8163–8171; Y. S. Lu,
Polym. Sci., 2009, 113, 3417–3425. L. H. Weng and X. D. Cao, Macromol. Biosci., 2005, 5, 1101–
85 M. Bergenstrahle, K. Mazeau and L. A. Berglund, Eur. Polym. J., 1107; Y. S. Lu, L. H. Weng and X. D. Cao, Carbohydr. Polym.,
2008, 44, 3662–3669. 2006, 63, 198–204.
86 G. Siqueira, J. Bras and A. Dufresne, Langmuir, 2010, 26, 402–411. 117 Q. Li, J. P. Zhou and L. N. Zhang, J. Polym. Sci., Part B: Polym.
87 S. Berlioz, S. Molina-Boisseau, Y. Nishiyama and L. Heux, Phys., 2009, 47, 1069–1077.
Biomacromolecules, 2009, 10, 2144–2151. 118 K. S. Mikkonen, A. P. Mathew, K. Pirkkalainen, R. Serimaa,
88 Y. Habibi, H. Chanzy and M. R. Vignon, Cellulose, 2006, 13, 679– C. L. Xu, S. Willfor, K. Oksman and M. Tenkanen, Cellulose,
687. 2010, 17, 69–81.
89 B. Braun and J. R. Dorgan, Biomacromolecules, 2009, 10, 334–341. 119 H. S. Qi, J. Cai, L. N. Zhang and S. Kuga, Biomacromolecules, 2009,
90 I. Filpponen and D. S. Argyropoulos, Biomacromolecules, 2010, 11, 10, 1597–1602.
1060–1066. 120 D. Bondeson and K. Oksman, Compos. Interfaces, 2007, 14, 617–
91 S. P. Dong and M. Roman, J. Am. Chem. Soc., 2007, 129, 13810– 630; D. Bondeson and K. Oksman, Composites, Part A, 2007, 38,
13811. 2486–2492; K. Oksman, A. P. Mathew, D. Bondeson and
92 G. Morandi, L. Heath and W. Thielemans, Langmuir, 2009, 25, I. Kvien, Compos. Sci. Technol., 2006, 66, 2776–2784; L. Petersson,
8280–8286. I. Kvien and K. Oksman, Compos. Sci. Technol., 2007, 67, 2535–
93 J. Yi, Q. Xu, X. Zhang and H. Zhang, Polymer, 2008, 49, 4406–4412. 2544.
94 Q. Zhou, H. Brumer and T. T. Teeri, Macromolecules, 2009, 42, 121 L. Jiang, E. Morelius, J. W. Zhang, M. Wolcott and J. Holbery,
5430–5432. J. Compos. Mater., 2008, 42, 2629–2645.
95 H. L. Cox, Br. J. Appl. Phys., 1952, 3, 72–79. 122 V. Favier, G. R. Canova, J. Y. Cavaille, H. Chanzy, A. Dufresne and
96 B. Harris, Engineering Composite Materials, IOM Communications C. Gauthier, Polym. Adv. Technol., 1995, 6, 351–355.
Ltd., London, 1999. 123 N. Ljungberg, C. Bonini, F. Bortolussi, C. Boisson, L. Heux and
97 A. Kelly and W. R. Tyson, J. Mech. Phys. Solids, 1965, 13, 329–350. J. Y. Cavaille, Biomacromolecules, 2005, 6, 2732–2739.
98 M. Moniruzzaman and K. I. Winey, Macromolecules, 2006, 39, 124 O. van den Berg, M. Schroeter, J. R. Capadona and C. Weder,
5194–5205. J. Mater. Chem., 2007, 17, 2746–2753.
99 M. Takayanagi, S. Minami and S. Uemura, J. Polym. Sci., Part C: 125 D. Viet, S. Beck-Candanedo and D. G. Gray, Cellulose, 2007, 14,
Polym. Symp., 1964, 113–122. 109–113.
100 N. Ouali, J. Y. Cavaille and J. Perez, Plast., Rubber Compos. Process. 126 J. R. Capadona, O. Van Den Berg, L. A. Capadona, M. Schroeter,
Appl., 1991, 16, 55–60. S. J. Rowan, D. J. Tyler and C. Weder, Nat. Nanotechnol., 2007, 2,
101 J. R. Capadona, K. Shanmuganathan, D. J. Tyler, S. J. Rowan and 765–769.
C. Weder, Science, 2008, 319, 1370–1374. 127 W. J. Orts, J. Shey, S. H. Imam, G. M. Glenn, M. E. Guttman and
102 J. R. Capadona, K. Shanmuganathan, S. Triftschuh, S. Seidel, J. F. Revol, J. Polym. Environ., 2005, 13, 301–306.
S. J. Rowan and C. Weder, Biomacromolecules, 2009, 10, 712–716. 128 G. J. Chen, A. Dufresne, J. Huang and P. R. Chang, Macromol.
103 R. Rusli, K. Shanmuganathan, S. J. Rowan, C. Weder and Mater. Eng., 2009, 294, 59–67.
S. J. Eichhorn, Biomacromolecules, 2010, 11, 762–768. 129 Z. M. Huang, Y. Z. Zhang, M. Kotaki and S. Ramakrishna,
104 J. H. Wiley and R. H. Atalla, Carbohydr. Res., 1987, 160, 113–129. Compos. Sci. Technol., 2003, 63, 2223–2253; D. Li and Y. N. Xia,
105 M. M. Ruiz, J. Y. Cavaille, A. Dufresne, J. F. Gerard and Adv. Mater., 2004, 16, 1151–1170; A. Greiner and J. H. Wendorff,
C. Graillat, Compos. Interfaces, 2000, 7, 117–131. Angew. Chem., Int. Ed., 2007, 46, 5670–5703.
106 D. G. Gray, Cellulose, 2008, 15, 297–301. 130 W. I. Park, M. Kang, H. S. Kim and H. J. Jin, Macromol. Symp.,
107 N. L. Garcia de Rodriguez, W. Thielemans and A. Dufresne, 2007, 249, 289–294.
Cellulose, 2006, 13, 261–270; M. Roohani, Y. Habibi, 131 O. J. Rojas, G. A. Montero and Y. Habibi, J. Appl. Polym. Sci.,
N. M. Belgacem, G. Ebrahim, A. N. Karimi and A. Dufresne, 2009, 113, 927–935.
Eur. Polym. J., 2008, 44, 2489–2498. 132 C. H. Xiang, Y. L. Joo and M. W. Frey, J. Biobased Mater.
108 S. A. Paralikara, J. Simonsen and J. Lombardi, J. Membr. Sci., 2008, Bioenergy, 2009, 3, 147–155.
320, 248–258; M. S. Peresin, Y. Habibi, J. O. Zoppe, J. J. Pawlak and 133 P. Lu and Y. L. Hsieh, Nanotechnology, 2009, 20, 9.
O. J. Rojas, Biomacromolecules, 2010, 11, 674–681. 134 J. O. Zoppe, M. S. Peresin, Y. Habibi, R. A. Venditti and
109 L. Chazeau, M. Paillet and J. Y. Cavaille, J. Polym. Sci., Part B: O. J. Rojas, ACS Appl. Mater. Interfaces, 2009, 1, 1996–2004.
Polym. Phys., 1999, 37, 2151–2164; L. Chazeau, J. Y. Cavaille and 135 W. L. E. Magalhaes, X. D. Cao and L. A. Lucia, Langmuir, 2009, 25,
P. Terech, Polymer, 1999, 40, 5333–5344; L. Chazeau, 13250–13257.
J. Y. Cavaille and J. Perez, J. Polym. Sci., Part B: Polym. Phys., 136 Y. L. Li, I. A. Kinloch and A. H. Windle, Science, 2004, 304, 276–
2000, 38, 383–392. 278; M. Zhang, K. R. Atkinson and R. H. Baughman, Science,
110 Y. J. Choi and J. Simonsen, J. Nanosci. Nanotechnol., 2006, 6, 633–639. 2004, 306, 1358–1361.
111 M. Grunert and W. T. Winter, J. Polym. Environ., 2002, 10, 27–30. 137 K. Benaissi, L. Johnson, D. A. Walsh and W. Thielemans, Green
112 Y. Habibi and A. Dufresne, Biomacromolecules, 2008, 9, 1974–1980. Chem., 2010, 12, 220–222; Y. Shin, I.-T. Bae, B. W. Arey and
113 M. Samir, F. Alloin, J. Y. Sanchez and A. Dufresne, Polymer, 2004, G. J. Exarhos, J. Phys. Chem. C, 2008, 112, 4844–4848.
45, 4149–4157; M. Samir, L. Chazeau, F. Alloin, J. Y. Cavaille, 138 J. M. Dugan, J. E. Gough and S. J. Eichhorn, Biomacromolecules,
A. Dufresne and J. Y. Sanchez, Electrochim. Acta, 2005, 50, 3897– 2010, DOI: 10.1021/bm100684k.

This journal is ª The Royal Society of Chemistry 2011 Soft Matter, 2011, 7, 303–315 | 315