iranchembook.

ir/edu
iranchembook.ir/edu

TECHNIQUES A N D INSTRUMENTATION IN ANALYTICAL CHEMISTRY - VOLUME 2

Handbook
of Laboratory Distillation
With an Introduction to Pilot Plant Distillation
iranchembook.ir/edu

TECHNIQUES AND INSTRUMENTATION I N ANALVTICAL CHEMISTRY

Volume 1 Evaluation and Optimization of Laboratory Methods

and Analytical Procedures. A Survey of Statistical
and Mathematical Techniques
by D. L. Massart, A. Dijkstra and L. Kaufman

Volume 2 Handbook of Laboratory Distillation

by E. Krell
iranchembook.ir/edu

TECHNIQUES AND INSTRUMENTATION I N ANALYTICAL CHEMISTRY - VOLUME 2

Handbook
of Laboratory Distillation
With an Introduction to Pilot Plant Distillation

Completely revised second edition

Erich Krell
Akademie der Wissenschaften der DDR
Zentralinstitut fur Isotopen- und Strahlenforschung, Leipzig

Translation, exclusive of the parts retained from the 1st

English edition as prepared by C. G. Verver, Amsterdam,
by Dr. phil. Manfred Hecker, Leipzig.

E LSEVl E R SCI E NTI F I C P U BLI S H IN G C O M P A N Y

Amsterdam -
Oxford -
NewYork 1982
iranchembook.ir/edu

Publish& in co-edition with

VEB Deutmher Verlag der Wissenechaftan, Berlin
Dietribntion of thie book is being handled by the following publishers
for the E.S.A. and Canada
Elsevim North-Holland, Inc.
62 Vanderbilt Avenue
NewYwk, NY 10017

for all remaining areas

Eleevier Scientific Publishing Company
1 Molenwerf
P.O. Box 211, loo0 AE Amsterdam, The Netherlands

Lisray ef Congress Cataloging in Publication Data

Krell, Erich.
Handbook of laboratory distillation.
(Techniques and instrumentation in analytical
chemistry ; v. 2)
Trsnslation of: 3., bearb. und em. A d . Handbuch
der Laboratoriumedestille,tion.
Bibliography: p.
Includes index.
1. D i ~ ~ t ~ ~ 8 t i o n - h b o r a tmanuale.
ory I. Title.
II. Series.
QD63.D8K713 1982 542l.4 82-9855
ISBN 0-444-99523-7 AACR2

@ VEB Deutscher Verlag der Wissenechaften, Berlin, 1982

All rlghts reserved. No part of this publication may be reproduced, stored in a
ret,rieval system, or transmitted in any form or by any means electronic, mechan-
ical, photocopying, recording, or otherwise, without the prior written permission of
the publisher.

P r i n t e d in the German Democratic Republic

iranchembook.ir/edu

Foreword to the third edition

The first edition of 1958 was sold out in a relatively short time. The second edi-
tion followed in 1960. The book has become a standard work and has been translated
into Russian, English and Hungarian.
This edition has again been written with the object of giving an account of the
subject of laboratory distillat)ion including recent views and developments. The
literature has been reviewed up to the year 1979. The author has adopted the course
of dealing only with generally accepted facts ;there are still numerous problems in sim-
ple and countercurrent distillation which have not yet been completely clarified and
in which there exist differences of opinion among various investigators. Distinct
trends in development have nevertheless been mentioned, in order to give an incentive
for further work. Owing to the large mass of material, a critical selection has been
necessary. An attempt has been made to introduce the mathematical deductions
and formulae required in laboratory work in a readily understandable form. Readers
with a mathematical turn of mind and those interested in particular problems will
find extensive references to the literature for further study.
The fundamental scheme of arrangement as applied in the first edition has
been retained. Section 5.1.3 has been extended to cover pilot plant distillation.
Section 4.2 now deals with fluidand interface dynamics. Chapter 8 could be drastically
shortened as there are a variety of components of distillation apparatus and the
pertaining measuring and control devices commercially available. The nomograms,
which were presented separately, have been inserted in the text. The references
for the various chapters have been rearranged and important new items added
to them. A great number of review articles serve to provide coinprehensive lists
of references for a longer period.
The book is intended primarily for physicists, chemists and engineers engaged in
chemical industry and in research or development centres, whose work includes
distillation on a laboratory or semi-technical scale. It will, however, also be useful
t o undergraduates, chemical technologists and laboratory assistants as a source
of answers to inany questions in the field of practical distillation and separating
processes. It is hoped that it will prove a guide to better and inore economi-
cal methods of operation for all these who have to carry out distillation in the
laboratory.
The author wishes to express his special gratitude to Prof. Dr. habil. K1. Wetzel
for his interest in the book, his valuable suggestions and constant support. He also
wishes to thank Dr. H. Stage and E. Giebeler for the numerous helpful comments
they have made. Further his thanks are due t o the various manufacturers of labor-
atory apparatus and glassware who have provided him with prospectuses and
iranchembook.ir/edu

6 Foreword to the third edition

technical data. Last but not least, he is grateful to the publishers for t,he generous
lay-out of the book, in particular to the staff of the chemistry department for their
thorough work on the manuscript.
It is to be hoped that this third edition will also contribute to the further develop-
ment of laboratory distillation and that, in laboratories, in industry, in technical
schools and universities, it will serve as a textbook and as a guide in the solution
of problenls of separation by distillation.

Dr.-Ing. Erich Krell

Akademie der Wissenschaften der DDR
fur Isotopen- und Strahlenforschung, Leigzig
Zentralir~et~itut
iranchembook.ir/edu

Contents

List of symbols . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

a . A review of the history of laboratory distillation . . . . . . . . . . . . 20

3 . Standardization and data on concentrations. . . . . . . . . . . . . . 32

3.1 Standardization of distillation a p p q a t u s . . . . . . . . . . . . . . . 32
3.2 Definition of concepts . . . . . . . . . . . . . . . . . . . . . . . . 37
3.3 Symbols and units . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.4 Definitions and conversion of concentrations . . . . . . . . . . . . . . 38

4 . Physical fundamentals of the separation process . . . . . . . . . . . . 43

4.1 Principles of simple and countercurrent distillation . . . . . . . . . . 43
4.2 Fluid and interface dynamics . . . . . . . . . . . . . . . . . . . . 46
4.2.1 Wetting columns . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.2.2 Film formation . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.2.3 Column dynamics . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.3 Miscibility of t h e components . . . . . . . . . . . . . . . . . . . . 65
4.4 Vapour pressure-temperature relationship. p - t . . . . . . . . . . . . 58
4.4.1 Measurement of saturated vapour pressures . . . . . . . . . . . . . . 59
4.4.2 Calculation and representation of saturated vapour pressures . . . . . . . 66
4.5 Equation of state and ps-diagrams . . . . . . . . . . . . . . . . . . 76
4.6 Boiling point diagram, B - 2, equilibrium curve. y* - x . . . . . . . 80
4.6.1 Calculation of vapour-liquid equilibria . . . . . . . . . . . . . . . . 82
4.6.2 Volatility, separation factor oc and activity coefficient y . . . . . . . . . 57
4.6.3 The experimental determination of equilibrium curves . . . . . . . . . 91
4.7 Number of theoretical plates (separat)ingstages) . . . . . . . . . . . . 102
4.7.1 Calculation of separating stages by the McCabe-Thiele method in batch
operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.7.2 Calculation of separating stages by the McCabe-Thiele method in contin-
uous distillation . . . . . . . . . . . . . . . . . . . . . . . . . . 111
4.7.3 Determination of separating stages by the McCabe-Thiele method for
equilibrium curves with an inflection point or a n azeotropic point . . . . 115
4.7.4 Determination of separating stages for flat equilibrium curves and for
equilibrium curves close t o operating line . . . . . . . . . . . . . . . 117
4.7.5 Methods for determining the plate number in batch distillation arithmet-
ically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
iranchembook.ir/edu

8 Contents

4.7.5.1 Determination of the plate number from difference in boiling point . . . . 120
4.7.5.2 Calculation of the plate number with the aid of the fractionating factor
and the Rose formulae . . . . . . . . . . . . . . . . . . . . . . . 121
4.7.5.3 Calculation of the minimum plate number by the Fenske equation for
ideal mixtures and w = ca . . . . . . . . . . . . . . . . . . . . . 125
4.7.5.4 Other methods and comparieon . . . . . . . . . . . . . . . . . . . 127
4.8 Theory of packed columns . . . . . . . . . . . . . . . . . . . . . . 128
4.8.1 Process of separation in a packed column . . . . . . . . . . . . . . . 128
4.8.2 Determination of the transfer unit. TU . . . . . . . . . . . . . . . . 132
4.8.3 Intensity of countercurrent exchange; time required . . . . . . . . . . 136
4.9 Determination of t.he number of plates and transfer units in the batch
and continuous separation of multicomponent mixtures . . . . . . . . . 140
4.10 Testing plate columns and packed columns . . . . . . . . . . . . . . 144
4.10.1 Column diameter . . . . . . . . . . . . . . . . . . . . . . . . . 146
4.10.2 Effective separating length; the introduction of column packing . . . . . 146
4.10.3 Test mixtures end the composition of the charge . . . . . . . . . . . . 150
4.10.4 Reflux ratio and quantity of reflux . . . . . . . . . . . . . . . . . . 162
4.10.5 Total. static and operating hold-up . . . . . . . . . . . . . . . . . . 167
4.10.6 Operating presaure . . . . . . . . . . . . . . . . . . . . . . . . . 159
4.10.7 Load and vapour velocity . . . . . . . . . . . . . . . . . . . . . . 160
4.10.8 Method of column calibration . . . . . . . . . . . . . . . . . . . . 162
4.10.9 Data for packed and plate columns . . . . . . . . . . . . . . . . . . 166
4.11 Pressure drop. limiting velocity and calculation of column dimensions . . . 167
4.13 Heat calculations . . . . . . . . . . . . . . . . . . . . . . . . . 182

4.13 Distillate properties and distillation diagrams . . . . . . . . . . . . . 187

4.14 h t r u c t i o n s for the oalculation of distillation conditions . . . . . . . . . 193
4.14.1 Batch distillation a t atmospheric pressure . . . . . . . . . . . . . . . 193
4.14.2 Continuous distillation . . . . . . . . . . . . . . . . . . . . . . . 197
4.14.3 Vacuum distillation . . . . . . . . . . . . . . . . . . . . . . . . 198
4.16 Diatillation calculations by computer . . . . . . . . . . . . . . . . . 198

(i . Separating processes . . . . . . . . . . . . . . . . . . . . . . . . 203

6.1 The scale of operation . . . . . . . . . . . . . . . . . . . . . . . . 203

6.1.1 Nicro- and semi-micro-distillation . . . . . . . . . . . . . . . . . . 203
5.1.2 Analytical distillation . . . . . . . . . . . . . . . . . . . . . . . 210
5.1.3 Preparative and production distillation . . . . . . . . . . . . . . . . 214
5.1.3.1 Semi-technical columns . . . . . . . . . . . . . . . . . . . . . . . 215
6.1.3.2 Pilotplants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
5.1.3.3 The preparation of distilled water . . . . . . . . . . . . . . . . . . 225
5.1.4 The separation of isotopes . . . . . . . . . . . . . . . . . . . . . . 228
5.1.4.1 Isotope separation by low-temperature countercurrent distillation . . . . 231
5.1.4.2 The preparation of D, and lSO, by countercurrent distillation of water 232
5.1.4.3 The preparation of various isotopes . . . . . . . . . . . . . . . . . 242
5.2 Methods of operation . . . . . . . . . . . . . . . . . . . . . . . . 244
5.2.1 Batch and semi-continuous distillation . . . . . . . . . . . . . . . . 245
5.2.2 Continuous distillation . . . . . . . . . . . . . . . . . . . . . . . 246
5.2.2.1 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
6.2.2.2 Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
iranchembook.ir/edu

Contents 9

5.2.2.3 Apparatus for continuous column distillation . . . . . . . . . . . . . . 250

5.2.2.4 Examples of application from laboratory practice . . . . . . . . . . . . 254
5.2.2.3 Starting up continuous distillations . . . . . . . . . . . . . . . . . . 255
5.2.3 Separation by partial condensation . . . . . . . . . . . . . . . . . . 256

5.3 Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . 260

5.3.1 Low-temperature distillation . . . . . . . . . . . . . . . . . . . . . 260
5.3.2 High-temperature and isothermal distillation . . . . . . . . . . . . . . 269
5.4 Distillation pressure . . . . . . . . . . . . . . . . . . . . . . . . 274
5.4.1 Simple and countercurrent distiHction under reduced pressure . . . . . . 274
5.4.2 Continuous equilibrium vaporisation (flash distillation) . . . . . . . . . 280
5.4.3 Thin-film distillation . . . . . . . . . . . . . . . . . . . . . . . . 283
5.4.4 Molecular distillation . . . . . . . . . . . . . . . . . . . . . . . . 292
5.4.5 Pressure distillation . . . . . . . . . . . . . . . . . . . . . . . . . 305

6 . Selective separating processes . . . . . . . . . . . . . . . . . . . . 307

6.1 Carrier vapour distillation . . . . . . . . . . . . . . . . . . . . . . 307

6.2 Szeotropic and extractive distillation . . . . . . . . . . . . . . . . . 312
6.2.1 dzeotropic distillation . . . . . . . . . . . . . . . . . . . . . . . . 317
6.2.2 Extractive distillation . . . . . . . . . . . . . . . . . . . . . . . . 327

6.3 Solution distillation and special methods . . . . . . . . . . . . . . . 334

7 . Constructional materials and apparatus . . . . . . . . . . . . . . . . 336

7.1 Const.ructional materials for distillation apparatus . . . . . . . . . . . 336
7.2 Standard apparatus and unit parts . . . . . . . . . . . . . . . . . . 337
7.2.1 Taps and valves . . . . . . . . . . . . . . . . . . . . . . . . . . 344

7.3 Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347

7.3.1 E m p t y columns . . . . . . . . . . . . . . . . . . . . . . . . . . 347
7.3.2 Packed columns . . . . . . . . . . . . . . . . . . . . . . . . . . 366
7.3.3 Plate columns . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
7.3.4 Columns with stationary elements . . . . . . . . . . . . . . . . . . 365
7.3.5 Columns with rotating elements . . . . . . . . . . . . . . . . . . . 373
7.4 Condensers and dephlegmators . . . . . . . . . . . . . . . . . . . . 382
\
7.5 Adapters; still and column heads . . . . . . . . . . . . . . . . . . 387
7.5.1 Adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
7.5.2 Still heads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
7.5.3 Column heads . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
7.6 Still pots. receivers and fraction collectors . . . . . . . . . . . . . . . 398
7.6.1 Still pots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
7.6.2 Receivers and fraction collectors . . . . . . . . . . . . . . . . . . . 401
7.7 Insulation and heating devices . . . . . . . . . . . . . . . . . . . . 405
7.7.1 The heating of still pots and flasks . . . . . . . . . . . . . . . . . . 405
7.7.2 The heating of feed-stock and bottom . . . . . . . . . . . . . . . . . 410
7.7.3 Insulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412

7.8 Packings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420

7.8.1 Shape of packing units . . . . . . . . . . . . . . . . . . . . . . . 420
7.8.2 Constructional material of pnckings . . . . . . . . . . . . . . . . . . 427
iranchembook.ir/edu

10 Contents

8 . A n t o d o devices. measuring and control equipment . . . . . . . . . . 430

8.1 Automatic equipment . . . . . . . . . . . . . . . . . . . . . . . . 430
8.1.1 Fully automatic equipment for standardized boiling point analysis . . . . 431
8.1.2 Fully automatic equipment for fractionation . . . . . . . . . . . . . -134
8.2 Temperature measurement and control . . . . . . . . . . . . . . . . 440
8.2.1 Temperature measurement . . . . . . . . . . . . . . . . . . . . . 440
8.2.2 Temperature control . . . . . . . . . . . . . . . . . . . . . . . . 446
8.3 Pressure measurement and control . . . . . . . . . . . . . . . . . . 448
8.3.1 Pressure measurement and control above 760 mm and from 760 to 1 mm Hg 449
8.3.1.1 The method of controlled evacuation . . . . . . . . . . . . . . . . . 453
8.3.2 Pressure measurement and control from 1 to Hg . . . . . . . . . . 457
8.3.2.1 The MeLeod compreaion manometer . . . . . . . . . . . . . . . . . 459
8.3.2.2 Vacuum control to pressures of mm Hg . . . . . . . . . . . . . . 162
8.4 Reflux and rate of evaporation . . . . . . . . . . . . . . . . . . . -163
8.4.1 Time-operated devices for reflux control . . . . . . . . . . . . . . . 463
8.4.2 Control of boil-up rate . . . . . . . . . . . . . . . . . . . . . . . 465
8.5 Measurement of physical data during distillation . . . . . . . . . . . . 467
8.6.1 Melting point . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
8.5.2 Refractive index . . . . . . . . . . . . . . . . . . . . . . . . . . 468
8.6.3 Dielectric constant . . . . . . . . . . . . . . . . . . . . . . . . . 470
8.5.4 Various determinations . . . . . . . . . . . . . . . . . . . . . . . 471
8.6 Measurement and metering of gases and liquids . . . . . . . . . . . . 472

9. Arrangement of a distillation laboratory. starting up dietillation apparatus 479

9.1 Lay-out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479

9.2 Frameworks and services . . . . . . . . . . . . . . . . . . . . . . 482

9.3 Building u p the apparatus . . . . . . . . . . . . . . . . . . . . . . 184
9.4 Sealing ground joints . . . . . . . . . . . . . . . . . . . . . . . . 488
9.5 Starting up distillation apparatus . . . . . . . . . . . . . . . . . . . 490
9.6 Safety measures . . . . . . . . . . . . . . . . . . . . . . . . . . 491

Relerenoes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492

Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . . 512

iranchembook.ir/edu

List of symbols

9 bottom product rate in continuous distillation

effective surface area
quantity of charge t o a batch distillation
B, degree of wetting
BP boiling point
c specific heat
CI,
specific heat at constant pressure
cv specific heat at constant volume
u quantity of vapour
n diffusion coefficient
Di degree of air-tightness
d diamet,er
E product rate
F volatility
F. melting point
FP solidification point
Fr fractionating factor
f cross-sectional area
9 acceleration due t o gravity
!7u standard value of y
H quantity of operating hold-up
column height (separating length)
height equivalent t o a theoretical plate
intensity
heat content
heat content of vaporized distillate
heat content of vaporized reflux
Kelvin
equilibrium constant
boiling point
heat transfer coefficient
LG velocity constant
L molar heat of vaporization
E mean liquid flow density over column cross-section
liquid flow density a t a given point
length (also, maldistribution = uneven distribution)
mean free path
inner diameter
molecular weight
mole fraction
number of moles
mass
mixing number
column efficiency
refractive index for the sodium D line
iranchembook.ir/edu

12 Liet of symbols

na number of transfer units

"J specific stage number
%h number of theoretical plates or stages
0 surface area
P total pressure
P pressure (partial preasure, distillation pressure)
P* vapour preesure of a pure substance
Q quantity of heat
VB he& supplied to still pot
QKO heat removed by condenser
Qv heat loss
QZ heat supplied to preheater in continuous distillation
R universal gas constant
R reflus rate
T heat of vaporization
s gas flow rate
8 stage efficiency
8, mean stage efficiency
T thermodynamic temperature
Tb equilibration time
t time
li voltage
V volume
Vb fraction by volume
VIllOi molecubr volume
V reflux ratio
'Jgew selected reflux ratio
"min minimum reflux ratio
W work
WS constructional material
2u velocity of flow
100 limiting velocity
ICk velocity constant for packiug unit
x quantity of liquid
t liquid composition referred to low-boiling component
2\ liquid concentration of bottoms
ZZ liquid concentration of feed
Xt3 mole fraction of low-boiling component in still pot
ZE mole fraction of low-boiling component in distillate
XO initial concentration
Xe final concentration
Y vapour composition referred to low-boiling component
?/* vapour composition in equilibrium
z feed rate in continuow distiiation
1 separation factor (vapour preseure ratio, relative volatility)
P evaporation coefficient (molecular distillation)
-I activity coefficient
I difference
6 Separation parameter (log, a)
& relative free volume of packing
17 dynamic viscosity
e characteristic temperature (reduced temperature)
6 temperature, "C
x ratio of speaific heats
1 thermal conductivity
iranchembook.ir/edu

List of symbols 13

i" reduced mass

V kinematic viscosity
5 composition, in fractions by weight or weight o/o
x reduced pressure
Q density
U surface tension
r time interval, time constnut
d, reduced volunie
cp contact angle

Subscripts and abbrerations

0 initial state1 ges total
1, 2, 3 pure components of a mixture 1 ideal
A bottoms K column
aequ equivalent korr corrected
az azeotropic krit critical state
D vapour R reflux
E product S side stream
e final state sa saturated
eff effective TGL Technisrhe Kormen, Gutevor-
eq equilibrium state schriften nnd Lieferbedingungen
F liquid (GDR)
Fk packing unit W water vopour
g gaseous Z feed
iranchembook.ir/edu

This Page Intentionally Left Blank

iranchembook.ir/edu

1. Introduction

Although simple and countercurrent (rectified) distillation are among the most
important physical separating methods employed in chemical industry, and hence
also in research and works laboratories, i t is often found that the apparatus used
for this purpose in the laboratory has a low efficiency by present-day standards.
Furthermore, calculations on the process are seldom made; instead, the work is
as a rule based entirely on experience and empirical data.
I n this field, nevertheless, a large amount of research work has been carried
out during the last thirty years. Today we have available modern components,
high-vacuum and completely automatic equipment ; methods of calculation have
been developed, whilst laboratory separations now range from rnicro-distillations
with less than 1 gram of charge to continuous operations with a throughput of up
to 5 litres per hour, from the distillation of liquefied gases a t low temperatures to
that of tars at high temperatures and from separations a t normal pressures to
so-called molecular distillations a t pressures below 10-4 mm of mercury. Selective
procedures have been perfected, and it is now possible to separate mixtures formerly
considered inseparable by appropriately influencing the vapour pressure relationship.
The classical textbooks of v. Rechenberg [l]and Young [ 2 ] provide an excellent
review of theory and practice, including that of industrial installations, but are
now out of date in many respects, particularly as regards methods of calculation
and apparatus. The works written by Thormann [3] and Badger and McCabe [4]
around 1930 already contain the graphical method of computation of McCabe and
Thiele [5] and excel in clarity of presentation. They are, however, mainly concerned
with large-scale operations and fall short in their treatment of the special problems
of laboratory distillation. A great impetus to research in the latter field was given
by the work of Jentzen and his students. In a Dechema monograph [6] published
in 1932 he gave a detailed description of the fundamental requirements for coltunns
(previously presented in 1923); these are still largely valid today. The books of
Kirschbauin [7], Gyula [8] and Jacobs [9] have a mainly industrial orientation ;
this also applies to that of Robinson and Gilliland [lo] which, apart from the theory,
deals with difficult separations of multi-component mixtures and with azeotropic
and extractive distillation. Perry’s Chemical Engineer’s Handbook [10a] contains
a chapter on distillation with numerous examples, tables and nomograms for
calculating industrial installations ; laboratory distillation is, however, but briefly
discussed. All these books presuppose a knowledge of the basic theory and a measure
of practical experience as do the 1944 work of Schultze and Stage [ll] on problems
of column distillation and the Dechema booklets published by Thormann [12] on
“Arbeitsmethoden und Gerate - Destillieren und Rektifizieren” and “DestiUieren -
iranchembook.ir/edu

16 1. Intmdnction

Betriebstechnii" [131 covering laboratory methods and engineering aspects, respec-

tively. The development of distillation techniques between 1920 and 1944 is reflected
by a bibliography of Stage-Schultze [14] containing about 2 300 publications which
m e baaic to the theory, apparatus and methods of distillation and rectification.
The entire distillation literature of the world for the years 1941 to 1946 and 1946
to 1952 was summarized in short references by A. and E. Rose [15]; the first group
contains lo00 investigations, the second 5000. In the publication ,,E'ortschritte der
Verfahrenstechnik" [16] biannual reviews are given in the sections on ,,Destillieren
und Rektiiieren", ,,Rektifikation bei tiefen Temperaturen", ,,Stoffubertragung",
,,Gleichgewichtsverhalten von Ein- und Mehrstoffsystemen" and ,,Vakuumtechnik
und Verfahren im Vakuum".
Walsh [17] annually gives a critical summary of literature in "Unit Operations
Reviews". As regards material constants and methods of calculation valuable
suggestions are found in the manual for petroleum engineers by Orlicek-Poll [18]
whioh includes an extensive chapter on boding points and phase equilibria. The
latest state of distillation techniques both in the laboratory and in industry is
discussed by Sigwart [19]in a very good survey of the field for Ullmann's Ency-
clopaedia of Engineering Chemistry. The theory of distillation and extraction is dealt
with on a strictly thermodynamical basis in the books of Kortiim and Buchholz-
Meieenheimer [20] and Bosnjakovii: [21]. The books of Gattermann-Wieland [22]
and Wittenbecher [23] which are well-known to every student of chemistry &re
intended to introduce the organic chemist into practical work. Distillation, however,
ia only represented by brief synopses. Although Weygand [24]and Bernhauer [ 2 4
devoted shorter sections to modern apparatus and methods in their books on labo-
ratory techniques an extensive discussion of laboratory distillation problems includ-
ing methods of calculation was lac- until the works of Carney 1261, Rosengart [27]
and Rose et al. [28] gave 8 systematic treatment of the special requirement6
of laboratory distillation. The contribution of von Weber [29] concentrates on the
preparation of essential oils by &tillation, especially by means of steam distilla-
tion. Siwart wrote the chapter on distillation and rectification for the 4th edi-
tion of the Houben-Weyl manual [30]. It offers an excellent survey together with
numerous examples of laboratory problems. The 13 problems set in the chapter are
all bawd on situations where some difficulty, expected or unexpected, has to be
overcome. Short books baaed on fundamental principles have been published by
Zuiderweg [31] and by Coulson and Herington [32]. Problems of fractional distilla-
tion in the laboratory are treated thoroughly in a monograph written by Bukala,
Majewski and Rodzibki [33].
Shoe 1960 the number of publications dealing especially with the theoretical
foundation of distillat ion has increased considerably. Hence documentatim has been
done largely by computer since the number of abstracts is steadily increasing.
Periodical reviews of the literature on distillation still appear in various journals [34].
Information is readily obtained through bibliography cards and abstracts [35,16]. It
b-me necessary to publish extensive laboratory handbooks for undergraduates
which include chapters on distillation methods [36], since most of the textbooks for
pmctiml comes do not give enough attention to separating techniques 1371. A
iranchembook.ir/edu

1. 1nt.roductioii 17

number of monographs deal with important special fields of distillation. Thus. the
book of Rock [38] gives a n excellent introduction to extractive and azeotropic
distillation and in their books Stage et al. [39] make clear the loading conditions in
packed columns and the experimental determination of vapour-liquid phase equilibria
as exemplified by the boiling behaviour of fatty acids. The 1960 standard work of
HBla et al. on the theory and the experimental deterniination of liquid-vapour
equilibria [40] has been complemented by a n extensive collection of equilibrium data
11411. X new edition of Kogan-Fridman’s compilation of tables contains equilibrium
data for over 2000 systems. Besides, these authors have published a large collection
of azeotropic data (21069 systems of 2 to 5 components) [42]. The data on vapour-
liquid phase equilibria edited by Hirata et al. [59] were prepared using computers.
The book contains lo00 systems for 800 of which experimental data and coiiiputed
curves are given. It should be noted that 200 high-pressure equilibria covering 133
systems are included. Systems with solubility gaps are not represented. By order of
DECHEMA, Gmehling and Onken [66] compiled a collection of equilibrium data.
Valuable suggestions for obtaining material data required for distillation calculations
are made by Bittrich et al. [60]. Billet has written 3 very instructive monographs on
the fundamentals of the thermal decomposition of liquids [43] and on optimization in
rectification (with special consideration of vacuum rectification) [44]. These books
refer particularly to technical distillation but they provide a number of calculating
methods for laboratory and even more for semitechnical distillation. The two inore
comprehensive books of Oliver [45] on “Diffusional Separation Processes” and of
Pratt [46] on “Countercurrent Separation Processes” as well as the books of Sattler
[62] on thermal separation processes and of Kafarov [62a] on the fundamentals of
material transfer illustrate the general trend of jointly presenting separation techni-
ques from a particular aspect. From a theoretical point of view this offers great
advantages. The introduction to separation processes written by Krell et al. [17] also
attempts to give a general survey of the various separation processes and possible
combinations for students of chemistry and process chemistry. The state of duneu-
sioning and precalculating processes of the thermal separation of material is reported
by Wunsch e t el. [63].
Apart from this the last few years have brought more books on special problems
of distillation, such as those of Hoppe and Mittelstrass [48] and of Stichelmair [6I] on
the fundamentals of plate and plate column dimensioning, resp., and of Jungnickel
and Otto [49] on the use of low temperatures in process engineering. While the book
of Frank and Kutsche [50] deals with the distillation of sensitive substances i n the
laboratory that of Olewskij and Rutschinskij [51] refers to the same problem on a
semitechnical scale. The applications of molecular distillation in the laboratory as
well as in pilot and industrial plants are described in a monograph written by Holl6
et al. [52] which includes an extensive list of references and a large number of tables
and illustrations. For their book on “Recent Developments in Boiling and Conden-
sation” Winter, Merte and Herz [65] collected over 500 references, 250 of which were
evaluated. The book is a n excellent description of the state of the art which also
includes interface phenomena.
Part I of Schuberth’s [53] ,,Thermodynamische Grundlagen der Destillation und
2 Krell, Handbook
iranchembook.ir/edu

18 l! Introdnation

Extmktion" treats the fundamental principles of the thermodynamics of mixed

phases, the classification of binary systems and the distillative separation of two
components. Part I1 will contain ternary systems, the extractive separation of two
componentfi by means of one or more solvents, special methods of selective distilla-
tion, and problems of multicomponent distillation and extraction. It is reasonable to
comider this exact theoretical presentation as a complement to the theoretical hints
concerning distillation problems given in the present book. The theoretical basis of
material exchange and of single- and multi-phase separation has been summarized by
Brauer [a]. He considers material transport in fluids both at rest and in motion. Of
particular interest for distillation are section 10 on material transfer in packing
layers at rest and with turbulence as well as part V on material transport through the
interfaces of simple two-phase systems and part VI on material transport in technical
n,ppamtus with a stream of two phaaes. Kaibel et al. I641 give an account of reaction
columns in which the column dynamics are complicated by the superposition of
reaction and distillation. Holland [55] presents a thorough discussion of questions of
multicomponent distillation while those of teohnical distillation are dealt with in a
more recent book of Billet [43a]. Scharov and Serafimov 1581have devoted their book
especially to questions of the countercurrent distillation of azeotropic multiwmpo-
nent mixtures.
The international symposium on distillation a t Brighton, England, in 1969 [56]
clearly showed the necessity to enlarge the theoretical basis of separation by distilla-
tion using coniputers, especially for the precalculation of separation processes. The
sS.7nposium covered the whole field of research and development and reviewed the
latest developments. It became evident, however, that even now there are a host of
problems waiting to he solved in the complex area of distillation. I n comparing the
present state, the developments and prospects of simple and countercurrent distilla-
tion with other thermal separation methods Billet [57] comes to the same assessment.
He points out that our knowledge of material exchange processes and of column
loading limits is far from sufficient for optiniizations so that further extensive and
detailed experimental and theoretical work will have to be done. This Situation waR
again brought out at the second international symposium on distillation in London,
1979. The 55 papers held in 4 sections were published in the proceedings of the sym-
posium [56a]. As was seen from the literature of the years before, the symposium
demonstrated considerable progress for laboratory distillation in the area of the
precalculation of vapour-liquid equilibria of non-ideal systems and of multicompo-
nent mixtures. The calculation of the azeotropic composition of binary mixtures by
means of the LJNJFAC method is also gaining ground. Moreover i t wafi pointed out
that increasing attention is paid to experimental results from laboratory and pilot
plants, o.g., In the examination of prohlems involved in the combined process of
distillation and chemical reaction.
The present state and trends in the field of material transfer in distillation co-
lumns have been discussed by Weisfi [67]. In the future, emphasis will be placed o n
the optimization and intensification of techniques, particularly from the point of
v i t w of saving energy.
In view of the enormous development in laboratory distillation during the last
iranchembook.ir/edu

1 . Introduction 19

thirty years and the extensive specialization in this field it seemed desirable that
an introduction to laboratory and pilot-plant distillation technique should appear,
not assuming any previous knowledge in this field, but nevertheless containing the
methods for determining vapour pressures and equilibriuni curves and giving a
detailed description of continuous and selective separating processes, together with a
vhapter on measuring and control devices. This review has the object of removing
iiiany erroneous notions about the factors affecting the process of separation and of
giving a comprehensive account of the performance of simple and difficult distilla-
t ions.
The chapter entitled “A review of the history of laboratory distillation” at once
introduces the general ideas, whilst chapter 3 clarifies concepts and defines the units
of measurement and the symbols. Special attention is given to standardization since
it brings about conventions concerning instrument and process parameters and thus
makes data comparable, which is a prerequisite for successful research. After the
physical theory of separation and the properties of the mixtures to be separated have
been discussed, the general and selective separating processes used in various cases
are dealt with from several points of view (chaps. 4 to 6). The apparatus required,
together with auxiliary equipment such as measuring and control apparatus, is
described in chapters 7 and 8. Finally chapter 9 details points to be considered in
fitting out a distillation laboratory and preparing the apparatus.
It was considered particularly important to deal with the procedures followed in
laboratory and pilot-plant distillation from the aspect of large-scale distillations,
since the former are often the forerunners of the latter. Formerly it happened all too
often that distillation methods were developed in the laboratory without any con-
sideration of development ;the result was that serious difficulties frequently occurred
in scaling up the laboratory experiments to the full dimensions. If, however, the
experiments are properly designed a t the outset for technical interpretation, much
expense and time can often be saved and the data obtained can be employed in
technical calculations without appreciable correction. This does not, of course,
exclude the need for using conditions in certain cases, say in analytical distillation,
that wonld be entirely uneconomic in industrial operation. Only with a knowledge of
the fundamentals of the separating process it is possible to decide on the optimum
conditions for every problem. It is the purpose of the present book t o provide this
knowledge.
The symbols for physico-chemical quantities and units have for the most part been
adopted unchanged since distillation technique requires some special notation. It has
heen attempted, however, to use internationally adopted symbols wherever possible.
The references have been grouped according to the chapters. Figures, formulae and
tables have been given running numbers throughout the book.
iranchembook.ir/edu

2. A review of the history of laboratory distillation

Not only i t is interesting to study the development of a chemical operation through

the centuries from a historical point of view, but it is often found that valuable
pointers for research are obtained from parallels between old and present-day methods.
As an introduction to a special field of chemical technology, such as distillation, a
historical review gives the reader an appreciation of how the human mind has always
sought for new ways of achieving better processes and more efficient apparatus with
the facilities available at the time.
Schelenz [I], Gildemeister [2] and von Rechenberg [3] have considered much
historical material concerning the development of distillation for their discussions of
the preparation of essential oils. Underwood [a] presente a brief outline of the evo-
lution of distillation technique up to the 19th century, and in his excellent book
Forbes [5] tells the history of the art of distillation from its beginnings to the death
of Cellier-Blumenthal.The latter, a French engineer who died in Brussels in 1840, was
one of the most talented column designers of the early 19th century. Further histori-
cal details can be found scattered in the various accounts of the history of alchemy
and chemistry. Of the more recent publications, the history of science and technology
of Forbes and Dijksterhuis [6] may be mentioned. In two volumes, the evolution of
technology, astronomy, mathematics, chemistry, and, in particular detail, of physics
is presented in an easily comprehensibleway. The development of distillation methods
is described in some detail in Bittel’s [16] paper on the history of multiplicative
separating niethods. W. Schneider [7] has published a lexicon of alchemistic and
pharmaceutical symbols which helps to understand ancient writings.
In the book of Strube [15] the main strands of development are described.
Distillation is dealt with according to its significance and its history from the 3rd
mntury A. D. is illustrated by numerous pictures.
The present review has been written, not with the object of providing complete
historical details or etymological derivations, but rather of giving a clear account
of the steps by which laboratory methods and apparatus (not omitting those for
semi-technical use) have developed to their present forms.
Distillation is an art that m a practised long before the Christian era by the
ancient Egyptians and was cultivated and protected as secret science by temple
priests. It also appears to have been carried out in early times in India, Persia and
China. Schelenz [I] is of the opinion that the discovery of distillation must be ascribed
to the Persians, who employed the art for the preparation of rose water. Another
view Rhared by v. Lippmann [8-lo] is that the principle of distillation found its
origin in the carbonization of wood, since descending distillation is referred to in the
“Ebers ppy-ru~~’’ of about 1500 B.C., so that the process would be almost 3500 vears
iranchembook.ir/edu

2. A review of t,he history of laboratory distillation 21

old. It should he noted in this connection that the word distillation was a t that time
a collective tern1 for all separating processes then known; the word may be trans-
lated as “separation drop by drop’’ and in alchemistic speech denoted the separation
of more or less subtle (or “fine’)) elements from each other. The concept of distillation
also covered operations such as filtration, crystallization, extraction and the ex-
pression of oil.
We shall here deal only with the history of distillation in the present meaning of
the term : the separation of bodies by evaporation and condensation of the vapour.
I n this connection it should be observed that by the foregoing definition descending
distillation is a true distillative operation.

Fig. 1
Alembic on furnace, surrounded by magic
signs. From a treatise on t h e making of
gold by t h e Egyptian alchemist Cleopatra
(2nd century A.D.)

The earliest uses of distillation were the preparation of rose oil and other ethereal
oils, distilled water for sailors (Aristoteles mentions how fresh water can he matie
from salt water) and a large number of alchemical mixtures and draughts. Fig. I
shows a so-called alembic (helmet) on a furnace, surrounded by the magic inscriptions
which in early ages played such a large part in the process of distillation. The illu-
stration is taken from a treatise on the preparation of gold dating from the second
century A.D., by the Egyptian woman alcheniist Cleopatra. A typical apparatus of
this period is shown in the next figure (2a). It is a glass still on a sand or water bath,
an arrangement still in use today, as demonstrated by the adjoining illustration of a
mercury still (2b). The four separate parts - the heating bath, the still (curcurbita),
the head (alembic) and the receiver (receptacula) - have remained in use as compo-
nents to this day. It is interesting to note that the collar for collecting the distillate
is also still found in a number of modern forms of equipment. The material employed
for the apparatus in antiquity was chiefly glass, a ceramic compound or copper.
From about 1300 A.D. onwards the methods of distillation may be divided into
two basic types:
per ascensum = “rising distillation”, per descenszm = “descending distillation”.
iranchembook.ir/edu

22 2. A review of the &tory of laboratory dietilation

The deanding procedure (l9g. 38) sank into oblivion after about 1800, although
it is the best method for certain types of separation. We find the pciuciple again in a
water still of thr year 1952, where it is chosen to enswe an economical use of heat
(we Fig. 3b).

Fig. 2a)
Glass dist,illation apparatus with sand or water bath (2nd century -4.D.)
Fig. 2 b)
Variiiim d i l l for merriiry with collecting collar for the distillate (20th century)

Fig. 3a)
Dry distillation of bark and herbs “per
deacensrni” (1 300 A.D.)
Fig. 3 b)
Water still of Blome according to the
“per desc,ernaacni” principle (20th
b) century)
iranchembook.ir/edu

2. A review of the history of laboratory distillation 23

Soon after the invention of printing a series of verygraphic descriptions of distilla-

tion were published, showing the state of development at the end of the middle ages.
The most important of these were the following:
1483, Schrick: Verzeichnis der ausgeprannten Wasser (Account of the hurnt -out
watefs) ;
1500, Brunswig: Das Buch der rechten Kunst zu destillieren die eintzige Ding (Book
of the true art of distilling the sole things);
1507, Brunswig: Das Buch der wahren Kunst zu destillieren (Book of the true art of
distilling) ;
1528, Ulstad : Coelum philosophorum (The philosopher’s heaven) ;
1536, Rpff: Neu gross Destillierbuch wohl begrundeter kiinstlicher Uestillation (Kew
large distilling book of well-founded artificial distillation).
After the 16th century a large variety of methods of heating the stills is observed.
Heating is carried out by air bath, water bath, sand or ash bath and also with the a d
of wax candles. The furnaces are provided with fuel hoppers, in order to permit of
working without interruption. Very strange systems of heating are also encountered.

Fig. 4a)
Separate heating furnace for wooden
distilling apparatus, with condensing coil
(17th century)
Fig. 4 b)
Separate boiler for continuous industrial
plant (20th century) b)
iranchembook.ir/edu

24 2. -4review of the history of laboratory distillation

such aa those utilizing the heat of fermentation of bread dough or pressed-out fruit.
In hot climates heat was occasionally obtained from burning mirrors and it is of
inbrmt to find that the same method of heating is covered by a patent in the year
1943. In industrial installations it is often necessary to separate the heating from the
actual still (Fig,4b). This arrangement had already been described by Ghuber
(lSOa-l668), aa shown in Fig. 4a. For obtaining larger amounts of distillate it' was
customaq-, even in the 16th century, to operate several &ills simultaneously on a
furnace, which was often arranged in terraces (Fig. 5 ) . The first attempts at heat

Fig. 5
Distilling furnace in terraces. The
fire is in the rentre of the cone; at
the sides there are draught
channels for temperatureregulation.
The earthenware fans support the
alembics (air bath heating) (16th
century)

Fig. 6a)
Semi-terhnical or labolatory distilling apparatus with serpentine air or water
condenser (16th century)
Pig. Bb)
Coil condenser (Dimroth) with standard joints (20th century)
iranchembook.ir/edu

2. A review of the history of laboratory distillation 25

insulation, using clay mixed with animal hair, are due to Lully (1415). I n the 16th
century the advantages of heating in stages were recognized and steam distillation
was already practised. The use of steam as heat carrier first became general, how-
ever, around the year 1800.
For condensing the vapour, air-cooling was the only method available up to
about 1300. Long vapour tubes were necessary t o obtain the required effect. Later
these were led through a barrel or trough filled with water. Finally it was found that
the best form of condenser was a coil (Fig. 6a), a type stillvery common today (Fig. 6b).

Fig. 7a)
Distilling apparatus ot thcb
Pharmacopoeia Medico-
Chylmica. Large Turk’s-he,itl
still and still with helm.
using air cooling, on f u i -
naces. Zig-zag ascending
tube. Receiver with a d j n s t -
able support (1709)
Fig. 7 b)
Distilling apparatus with coil
(Jantzen; 20th centurj )

In the 16th century continuous condensation by water followed. The concentrating

effect of a long vapour riser and of partial condensation (Fig. 7a) were also recognized
(1648). The modern coiinterpart is illustrated in Fig. 7h. The countercurrent principle
in condensation was introduced by Pissonnier in 1770, the same principle fonnrl
today in the well-known Liebig condensers (Fig. 8b), which dat,e back to Dariot
(1533-1594; see Fig. 8a).
The development of distillation technique from the 16th to the 19th centiirj-
kept pace with the general improvement in equipment construction. By the middle
of the 16th century it was known that metal stills are subject to corrosion, whereupon
glass and ceramics came to be the preferred materials. Kunkel (1638-1703) was the
first to use glass as a general laboratory material, as is done t o this day. Fig. 9 showb
the large variety of still heads emploj-eda t that time. Micro-forms were a1read;v in use.
The appearance of an alchemical laboratory of about 1700 is illustrated by Fig. 10,
representing the apothecary’s laboratory of the Capuchin monastery in Paris. The
large nuinher of distilling devices proves that they were among the most frequentlj-
used laboratory apparatus. Up to the 18th century laboratory stills were almost
iranchembook.ir/edu

26 2. A review of the history of laboratory distillation

). b)
Fig. 8a)
Dsriot’s condenser with continuous c o u n t e r c m n t cooling (16th century)
Pig. 8b)
Liebig condenser using countercurrent principle (19th century)

Fig. 9
Still heads after Andreas Libao, about 1600
Alembic with long beak m) Tin head with beak-shaped
Alembic with short beak end
Blind alembics n) Head with extension tube
Alembic of tin 0 ) Alembic with cooler
Sublimation heads p) Dwarftypes
Blind alembics with tube q) Blind beaked alembic
Bell-shaped alembic r) Triple blind alembic
Tiara a) Triple beaked a,lembic
iranchembook.ir/edu

Fig. 10
Apothecary’s laboratory of the Capuchin monastery in Paris (about 1700)

Fig. 11
Boyle’s apparatus for vaciiiim distillation (1627- 1691)

a) C)

Fig. 12
Semi-technical distillation equipment in the 19th century
a) Germany, b) China, c) Bulgaria
iranchembook.ir/edu

Fig. 13
Continuous distillation appa-
ratus of Cellier-Blnmenthal
(1813)

Fig. 1Pa)
Bubble-cap colamn of Chtiui-
ponnois (1854)
Fig. 14b)
Glass plate column of Briiun
with 20 actiial t,rays and
vacuum jacket (1931)
iranchembook.ir/edu

2. A review of the history of laboratory distillation 39

exact copies of those utilized in early antiquity, but towards the middle of the 18th
century chemists like Baum6 and Woulff began to create new forms. Systematic
experiments on distillation were first performed by the physicist Boyle in England
(1627-1691)) who even carried out experiments in vacuum and under pressure
(Fig. 11).By the middle of the 18th century a few standard types of equipment had
evolved, which were used in all countries (Fig. 12). The diameter was about 45 t o
75 cni and the height 90 to 120 cm. I n the 19th century there were considerable
developments,particularly as a result of the activity of French constructors concerned
with the alcohol industry. After several stages of development (Adam, Berard, Perrier)
a continuous still was patented in 1813 by Cellier-Blumenthal; in its basic features
it corresponded to modern apparatus (Fig. 13). Stills for vacuum distillation were
built in 1828 by Tritton, and the sieve-plate column of Coffey (1830) constituted

el fl 4 hl i)

Fig. 15
Stages of evolution of the bubble-cap column
a) Distillation flask with vapour side-tube g) Spray-plate column of Le Bel-Henniger
b ) Claisen flask (1875)
c) Still heads without and with bulbs h) Rectifier of Young and Thomas (1889)
d) Ball head according t o Wurtz (1854) i) Baum’s plate column (1910)
e) Sieve-plate head of Linnemann (1871) k) Bubble-cap column of Bruun (1951)
f ) Sieve-plate head of Glinsky (1875)
iranchembook.ir/edu

30 2. A review of the history of laboratory distillation

anot,her advance. The principle of bubble-cap plates was introduced in 1864 by

Champonnois for industrial installations (Fig.14a). This process is basically that used
in large-scale tinits up to the present day, together with packed columns. The lat'ter
were introduced by Ilges, who first utilized spheres as packing material in 1873.
Reviewing what has heensaid above, we realize that the hasic principles of distilla-
tion were already known in antiquity and in the middle ages, in spite of the simple
equipment then availahle. Vntil the middle of the 19th century, laboratory and semi-

Fig. 16e)
Von Rechenberg's V U C U I I ~distillation apparatus (1920) with Hempel column and Bertrend
receiver
Fig. 16h)
Elsner's rectifier for normal pressurg and vacuum (1920)

technical apparatus differ in their dimensions only. It was not until the tempestuous
development of organic chemistry around t,he middle of the 19th cent,ury that distilla-
t ion equipment evolved entirely designed for experiments in the laboratory. Noted
chemists, between this time and 1900, such as Claisen, Dimroth, Glinsky, Hempel,
1,t. BPI, IAehig, Mitscherlich, Mohr and Wurt,z, evolved apparatus for laboratory
(lintillation practice. Retorts were wed as stills and as receivers; the distillation flask
with air-cooled side tube developed, by way of the Claisen flask and the Wurtz ball
head, into t,he spray-plate column. The still heads of Linnemann, Glinsky and Le Bel-
Henniger are precursors of the hubble-cap column [ l l ] (Figs. 14b and 15). The
cdunin packed with glass beads was introduced into t.he laborat,orv in 1881 hy
Herlipel.
iranchembook.ir/edu

2. A review of the history of laboratory distillation 31

Toward3 the end of the 19th century came the need to compare the many still heads
thus available as regards efficiency. Kreis, Young and Friedrichs [12] carried out
measurements in this field, and also studied condensers. Besides the Liebig condenser,
that of Diniroth (Fig. 6b) became of importance; a specially effective condenser for
low-boiling substances, designed by Mitscherlich, led to the construction of similar
types.
During the period between 1900 and 1920 numerous pieces of apparatus at
present still in use were developed: for instance the Jantxen column (Fig. 7 b), and
the Raschig and Prym rings for column packing (1916 and 1919, respectively). Heating
systems wer? improved, whilst pressure controllers for work in vacuum were evolved.
Fig. 16a shows an apparatus used for vacuum distillation by the well-known specialist
von Rechenberg [3] between 1900 and 1920, which, like Elsner’s rectifier (Fig. 16b),
illustrates the stage of development a t the end of this period.
It was not, however, until Jantzen [13] and his pupils had systeniatically in-
vestigated the physical fundamentals of the distillation process that the numerous
clevelopinents after 1920 could take place. This later phase is still too recent to allow
11s to consider it historically. Ever and again we find, however, that multiple threads
lead us hack from our present complicated apparatus and methods to long-vanished
times when t,he fundament a1 principles of our modern knowledge were first recognized
and elaborated [ 141.
iranchembook.ir/edu

3. Standardization and data on concentrations

3.1 Standardization of distillation apparatus

The significance of standardization froin both a scientific and an economic point of

view is generally accepted. No detailed arguments supporting it need, therefore, he
given here. I t may suffice to point out that it always pays to use standardized appa-
ratus. In all countries there is a tendency to provide standard apparatus for the
various separations to be carried out: it can then be mass-produced and hence
is relatively cheap. Further, it is readily available and gives reproducible results. The

Fig. 17
Apparatus for simple distillation
in vacuum assembled from stan-
dard components
a = three-neck round-bottomed
flask NS 29 with 2 side tubes
NS 14.5; b = thermometer with
ground joint NS 14.5; e = boil-
ing capillary; d = still h a d
with ground joints NS 29 and
14.5; e = Dimcot,hcondenser;
f = vacuum adapter; g =short-
W necked round-bottomed flask

first step in this direction were taken when the well-known Engler and ASTM
devices were designed. Even for vacuum work, for example, similar apparatus can be
assembled from standardized components (m. 17). Components and apparatus as
I
well as special methods for distillation are listed in [l, 21.
The main advantages of standard components are that the apparatus required
can be rtssembled safely and quickly and that in many cases expensive special appa-
r a t u s can be dispensed with, since the components can be combined in many ways.
Noreover, it is possible to copy the set-up of technical plants to a considerable
extent. In the case of a fracture only the delnaged component needs to be replaced.
Today, apparat.us having standard ground-glass joints is used almost exclusively
in every organic laboratory and hence in every distillation laboratory (Tab. 1). An
encyclopaedia of chemical laboratory apparatus including the G.D.R. standards has
been edited by Telle [33.
iranchembook.ir/edu

3.1 Standardization of distillation apparatus 33

Table 1

Sizes of standard ground-glass joints for interchangeable connections

a) Conical joints (KS), cone l:lO, according to TGL 14972 (Nov. 1972)
and DIN 12242 with cone length K 6 (IS0 recommendation)

Symbol Largest diameter Length

of cone (mm) in mm

58 5/13 0 13
7/16 7.5 16
10/19' 10 19
12/21 12.5 21
s 14/23 14.5 23
x 19/26 18.8 26
24/29 24 29
2: 29/32 29.2 32
34/35 34.5 35
x 45/40 45 40
60146 60 46
x 71/51 71 51
85/55 85 55
100/60 100 60

x = preferred sizer,

b) Spherical joints (S), according to TGL 20678 (Nov. 1971)

and DIN 12244 (July 1963, Spherical joints)
~ ~ ~~

Symbol
Diameter of sphere
TGL DIN') in mm

R c
I 7.144
13.2 1312 12.700
13 1315 12
x 19 1919 19.050
29 29/15 28.575
35 35/20 34.925
S 38 40125 38.100
41 41/25 41.275
51 51/30 60.800
S 64 64/40 63.500
76 76.200
102 101.600

x = preferred sizes
~~

1) The second number refers to the inner diameter of the glass pipe

3 Krell, Handbook
iranchembook.ir/edu

34 3. Standardization and data on concentrations

Conical joints (Fig. 18) are usually employed: they have been in use as inter-
changeable joints since the beginning of this century. Spherical joints (Fig. 19) have so
far been restricted to those devices that would be too rigid if conical joints were used.
Spherical joints are made according to standards TGL 20678 and DIN 12244. The
precision spherical joints of nominal widths NW 25 to 150 miu (manufacturer,
W. Buchi, Flawil/Switzerland)ensure high vacuum tightness without the use of any
packing.
The standard DIN 12242 specifies types of ground-glass joints and their appli-
cationu in laboratoq apparatus. The sizes of the interchangeable standard joints
(conical, 1 : 10) are covered by TGL 14972, sheet No. 2, and DIN 12242. Laboratory
apparatus is equipped with joints of serieR 1 and 2, of which Specimens 14.5/23,29/32

Fig. 18 Fig. 19
Standard conical ground Standard spherical ground
joint. 1 :10 joint with clip

and 4-5/40 are preferably employed. In micro and semi-micro apparatus, NS 7.5116,
lO/l9 and 19/26 are also used. I t is desirable to construct general laboratory apparatus
using exclusively size NS 14.5/23, which is now commonly employed for thermo-
meters ( E L 40-339, DIN 12784), and NS 29/32 since this strongly enhances the
interchangeability and combining-power of components. In the author's Destinorrn
series t,hisprinciple has been followed and has proved to be very successful. According
to standards TGL 1497213 and DIE 12243,ground joints with a 1:5 cone are required
for special purposes where the joint must loosen more easily, as in the distillation of
high-boiling mixtures, and particularly in high-vacuum work. Moreover, lenticular
and plane joints are in use, mostly for larger apparatus. Even some draft standards
c*oncerningglass components and piping for use on a technical scale have been pre-
p a d ~71.
The testing and handlimg of standard ground joints are dealt with by Friedrichs
[.I1 and Fliedner [5].
.k for distillation procedures, numerous conventional methods have become
known which are concerned with particular niixtures, such as the determination of
the boiling behavionr of phenol crude acids. ,4 rapid analysis for petrols up to 180 "C
iranchembook.ir/edu

3.1 Standardization of distillation apparatus 35

has been elaborated by Kuehnhanss et al. [6] which covers the paraffins, 5- and 6-ring
cycloparaffins and aromatics present. The Destinorm column head shown in Fig. 312
is used. So far the boiling analyses listed in Table 2 have been standardized, with the
dimensions of the apparatus exactly determined.
Since ground-glass joints have to be greased attempts have been made to design
new joints for glass apparatus. A conical joint where vacuum tightness is provided by
labyrinth packing instead of by ground glass has been developed by Wissenschaft-
lich-Technische Glasgeriite GmbH of Wertheim. Two teflon gaskets render greasing
unnecessary. No packing a t all is required for Biichi spherical ground joints. Precision
grinding ensures high vacuum tightness and pressure resistance. These spherical
joints are available for nominal widths 25, 50, 80, 100 and 150 mm. The French

threaded pipe
hulf-flange with
individual screwing
sikone rubber goskel
with PTFE coating

intermediat? ring

half-flange with
individual screwing

threaded pipe Fig. 20

Sovirel connection (manu-
facturer, Sovirel, Levallois)

laboratory glassware manufacturers Sovirel have developed a connection which

employs novel connecting pieces. The principle makes use of a chemically inert gasket
placed between t>wogrease-free glass pieces (Fig. 20). The joints are also pressure
resistant and vacuum tight. Their temperature resistance corresponds t o that of the
packing material.
According to Kramer [16] who described modern systems of joints for glass
laboratory apparatus these can he classified as follows :
a) Ground-glass joints
conical joints
spherical joints
cylindrical joints
plane joints
spherical flanges
combination joints
b) Polished joints
precision clear-glass joints
precision calibrated pipes

3*
iranchembook.ir/edu

36 3. Standardiecltion and data on concentfations

Table 2
Standard distillation methods
~ ~ ~~

Standard Distillation method

TGL 21120 Analyeis of mineral oils, liquid fuels and related products;
determination of boiling behaviour
21 125 Analysis of liquid fuels; vapour pressure determination
after Reid
0-51761 Analysis of technical benzenes ; determination of boiling
behaviour after Kraemer-Spilker
0-51786 Analysis of liquid fuels; determination of the content of
non-settling water in fuel oils by distillation with the
rylene method
DIN 51751 Determination of the boiling behaviour of Otto fuels and
pet,rols
51 752 Analysis of the boiling behaviour of Diesel fuels and similar
substances
51 761 Determination of boiling behaviour after Kraemer-Spilker
51 567 Analysis of mineral oils, higher boiling mineral oil fractions
and mineral oil distillation residues - E’ractional distilla-
tion - after Grosse-Oetringhaus
51611 Lom-temperature distillation of gases

c) Joints with paclung (except screw joints)

flexible ball joint
conid joint with gasket
plane flange with gasket
spherical flange with packing insert
d) Screw joints
a t u f f i i boxes
@crewcaps
sorew couplings
e) Screw flanges
screwed plastic flange coupling
flanged stuffing boxes
flanged screw caps
f) Accessories for joints
Plugs
flexible tubes
sleeves
bellows
safety devices for ground joints
joint gremes, and the like
With the screw capsystems, Kramer [16] distinguishes between the designsoffered
by the manufacturers Quickfit, Witeg and Sovirel.
iranchembook.ir/edu

3.3 Symbols and units 37

3.2 Definition of concepts

When comparing various publications on distillation technique it is repeatedly

found that fundamental concepts are defined in different ways, a fact often leading
to misunderstanding. The word “distillation” itself is employed for the most diverse
operations in this field, so that a distinction between “simple” and “countercurrent”
distillation (the latter also known as rectification) seems useful. I n this way the word
distillation becomes a collective term for processes in which liquid mixtures are
separated by evaporation and condensation of the outgoing vapour. Hampel [7a]
has dealt with the difficult question of the “purity” of solvents and discussed the
concept of “ultrapurity”. By this he means the highest degree of purity a t present
achievable. I n manv cases the impurity concentration is required not to exceed a
few ppb. Besides extraction, distillation is a method of achieving this aim. As early
as 1943 German standards were elaborated on the “decompositicin of liquid mix-
turcs by distillation and rectification” (DIN 7052) which are no longer up to date and,
in addition, do not apply to the specific conditions of laboratory distillation. There-
fore, new definitions based on a proposal of the author’s [S] have been worked out
h y the Working Cornmittee on Apparatus for Distillation and Rectificatoin of the
sect ion of the German standardization committee on Laboratory Apparatus. They
have been included in the respective passages of the text.

3.3 Symbols and units

The recent alterations are due to a resolution of the X. General Conference of

Weights and Measures in 1954. I n the GDR the change was enacted by the first
regulation on physical and technological units issued on 14th August, 1958, which
has been replaced by that of 31st May, 1967.
I n the PRG the regulations pertaining to the law on units of measurement ap-
peared on 28th June, 1970. Inforniation on the practical use and the area of applica-
tion of the new international system of units (SI)as compared to the previous systems
is given in the books of Padelt and Laporte [ S ] , Forster [lo] and Haeder and Giirt-
ner [ll]. I n the meantime further standards have been elaborated (Table 3).

Table 3
Standards on symbols, quantities and units

TGL 0-1304 General signs and symbols; signs and symbols for
generally nsed physical quantities
TGL 18-762 Sheet 1: Quantities and units; names, symbols and
abbreviations
Sheet 2: Explanations
DIN 1301 Units; symbols and abbreviations
DIN 1313 Notation of physical equations in science and techno-
logy
iranchembook.ir/edu

38 3. Standardization and data on concentrations

The SI unit of force now is the newton = N = m x kg x s-,. It replaces the

kilogramnie force and k defined as follows:
1 kgf = 1 kg 9.80665 in XS-, = 9.80665 N.
Forliquidandgas pressures theSI unitpascal = Pa = N x 111-3 has to be usednow.
I n addition, the larger unit bar = 105 Pa is provided. Conversion is done according to
atm. = kg f x c n r 2 = 0.980665 bar = 735.56 tow,
1 torr = 1.333 x bar.
The thermal unit used 80 far, the calorie = cal, is replaced with the joule:
1 cal = 4.1868 J.
For thermodynamic temperatures the unit Kelvin (no longer degree Kelvin) is
-
now in use. The degree Celsius for Celsius tetnperatures T To where T o is the
therniodynamic temperature of the triple point of water is a special name for the
Kelvin. Grad = grd is no longer used to denote temperature differences.
When using the list of symbols, as is done in the following pages, the use of the
same symbol for more than one concept is t o be avoided. The Latin alphabet proves
to be insufficient for this purpose, SO that Greek letters have had to be added, whilst
further variations have been made possible by the addition of suffixes [12].

3.4 Definitions and conversion of concentrations

In the technique of distillation it is usual to calculate with mole fractions and
mole percentages, as this greatly facilitates the computation of vapour volumes,
vapour velocities, limiting velocities and 80 on. Since as a rule it is further the practice
to consider the separability of components in sequence, the calculation may be
based on that of binary mixtures, in which case of course, the second component may
represent a mixture of several constituents. The a.verage molecular weight is then deter-
mined by the fortnrda :
M -
+ + w,
w-1 w, - WI
- + +
u', w3
Wl/M, + WdM2 + Ws/Mis M ~+ + Mns. (1)
'" - I Mn2
number of mole8 M, = W/M
Example: Determination of the average molecular weight of a mixture:
300
W , = 300 g of benzene ; W J M , = M , , = -= 3.84
78.11
400
W , = 400 g of t.oluene; W2/M2= M,, = - = 4.35
92.13
500
W 3= 500 g of xyiene'; W3/M3= Mn3 = -= 4.71
106.16
WI + W , + R', = 1200 g of mixture; M,1 + M,, + Mn3 = 12.90
1200
MI = 78.11; M, = 92.13; 1113 = 106.16; M,,, = -- = 93.0.
12.90
iranchembook.ir/edu

3.4 Definitions and conversion of concentrations 39

If it is not otherwise stated, concentrations refer t o the low-boiling component.

The concentration can be reported as a fraction or percentage by volume or weight or
a‘ a molecular fraction or percentage. The fractional concentration is the ratio of the
coniponent to the sum of all components.

Fraction by volume; component 1 el, = I7,/(lTl + 17,) (3

component 2 I!, = V,/(171+ 1;) (3)
Fraction by weight; coiiiponent 1 w1 = lt’l/(M’l n-,) + (4)
component 2 w 2= n’,/(w,+ W,) (5,
Mole fraction; component 1 x1 = Mnl/(Mnl Mn2) + (6)
component 2 x2 +
= Mnz/(Mn1 L W , ~ ~ ) (7)

To convert these numbers into percentages inultiply by 100.

Ex:crniple:Determination of the molecular fractions and percentages in a tiiirture:

Component 1 300 g of benzene; WJM, = M , , = 3.84

Coniponent 2 400 g of toluene; W J M , = M,, = 4.35

700 g of mixture ; M,, + M,, = 8.19

3.84
Mole fraction xl = - = 0.47; molecular percentage: 47.0
8.19
4.35
Mole fraction x, = - = 0.53 ; molecular percentage : 53.0.
8.19
I n the case of multi-component mixtures the denominator must be extended to
contain the additional components :

Mole fraction x, = Mnl/(M,81 M,,+ + Mn3)

Mole fraction x, = Mn2/(Mnl M,, + + Mn3)

+
Mole fraction x3 = Mn3/(MnI M , , + Mn3)
For the conversion of weight yoto mole o/o in ternary mixtures Lessels r l 3 ] gives
a useful nomogram.
I n the case of binary mixtures the various conversions are performed with the
following formulae, which all refer to the lon~-boilingcomponent 1 .

Weight 7; t,o Molyo:

Volume yo to Weight% :
Weight% = . 100 (12)
where p = density VICI +
Vl@l
1 1 ~ 2 ~ 2

Mole/, to Weighto/, :
iranchembook.ir/edu

40 3. Standardization a d data on concentration8

Weight% to Volume% : Volume yo = WJel . (14)

Wlkl + WZ/&
M O I ? ~to Volumeyo: Volumeo/ - M,de, 100 (16)
'O +
- J f P l / @ l Jfzxzlez
4

As these calculations are often time-consuming, various nomograms have been

developed for carrying out the conversions rapidly. In converting from molecular or
weight percentages to percentages by volume it should be remembered that the for-
mulae m d i d only if no change in volume (contraction) occurs on mixing the
components. A nomogram suitable for many purposes has been designed by Orlicek,
Poll and Walenda [14] (Fig. 21).
I n using it the following directions should be followed:

Tor converting Value to be For component 1 the following

from used for Q values are read off

on scale 2 on scale X

3101. fraction to
wt. fraction JfIlJf, mt. fraction mol. fraction
Vol. fraction to
mol. fraction Vmol~l~/V,,,,,,~,~
rol. fraction niol. fraction
Vol. fraction to
mt. fraction &?ZIP1 vol. fraction wt. fraction

Jf = molecular weight; = molecular volume; e = density

The choice of the indices should always be so made that Q becomes greater than
unity. The u0c of the nomogram is illustrated by the emmple. What weight percen-
tage corresponds to 88 mol yo if the two components have molecular weights of 150
and 60, respectively?
MI = 150 Y = 0.88 mol. fraction
MJMZ = 2.5
M, = 60 Z = 0.948 wt. fract'ion = 94.8 wt.%
The opposite conversions, weight per cent to mol. per cent, mol. per cent to
volume per cent, weight per cent fo volume per cent, are of course also possible.
Baehr's circular diagram [16] for converting niolecular into weight percentages or
vice versa is very convenient, though its accuracy is not so grmt as that obtained by
calculation. It,s use will be seen from the example indicated in the nomogram (Fig. 22).
What, molecular percentage corresponds to 300/o by weight of benzene in a benzene-
tolucwr mixture:
MI for benzene = 78.11
.l!ll/1W* 0.85
M , for toluene = 92.13
iranchembook.ir/edu

3.4 Definitions and conversion of concentrations 41

Z
0.03 @
0.04 @
0.05 @)

0.65
c
0.10 @ 10.60

0.20 @
0.25 @
0.30 @
0.35 @

-
0.80 @ -0.04
@ -
iranchembook.ir/edu

42 3. Standardization and data on concentrations

A straight line is drawn through the point for 30 wt.% on the lower semicircle
and 0.85 on the horizontal scale for M J M , ; it intersects the upper semicircle at
33.6 molo,,.
50

Fig. 22
Nomogram for convorsion
of moI?/, to wt..% and
rice verm (Baehr)

When dealing with systems of isotopes it is customary to express the con-

wntrations in atomic percentages. Nat lira1 water, for instance, is composed as
follows (see Table 38):
99.9844 atoniic yo 1H 99.757 atomic 0 4 1 6 0
0.0156 atomic yo D 0.039 atomic 04 170
0.340 atomic 180
1OO.OOO atomic yoof hydrogen 1OO.OOO atomic "4 of oxygen
I n a paper as well a8 in a book Spath [171 has dealt with possibilities of denoting
and representing binary systems.
iranchembook.ir/edu

4. Physical fundamentals of the separation process

4.1 Principles of simple and countercurrent distillation

The reader is reminded that the word distillation can be translated as “separation
drop by drop” (see Chap. 2). It can therefore he used as a collective term for processes
in which mixtures of ilrntually soluble liquids can be separated by evaporation and
condensation of the liquid, the condensed part becoming richer in the niost volatile
component. The word gives no indication of the technique adopted in the separating
process. The ternis “simple distillation” and “countercurrent distillat ion”. tiowevw.
define the inode of operation (Fig. 23). I n a strictly physical wnse distillation need
not produce any separation; we also speak of distillation when a pure liquid is w i t -
porated, the vapour is condensed and the condensate is removed.
The exchange of material can be described by the following basic equation.
m =K x ueffx h,, (15Ri

(in words : transition current = transition coefficient times effective area of phaw
interface times driving force).

Fig. 23
a) Principle of simple distillation
b) Principle of countercurrent distillation (;S = Vapour, $ = Liquid)
iranchembook.ir/edu

44 4. Physicel fundamentals of the separation procese

The exchange of material takes place by diffusion through the phase interface. It
depends on the diffusion constant, D, the diffusion paths, b, concentration, x, and
t,he.phase interface area, aeff,per unit of length.
The driving force results from the differences of the concentrat,ions in the two
phases of the system (y - z).A t t,hermodynamicequilibrium they are saturated, the
driving force becomes zero and the exchange of material taking place at non-equili-
briwn ceases.
In eimple distillation the molecules emerging from the evaporating surface move
uniformly until they reach the condensing surface. In countercurrent distillation part
of the condensed vapour, termed “reflux”, returns to the boiler, meeting in its passage
the rising vapour. Provision is made for intimate contact between the liquid and

to column head
lower temperature higher
concentration of low-boilina mnwonents

I 1 vapour enriched

region in whid
approach to
equilibrium
takes Dlace
-~

Ir r$’ref,ux enriched
in high-boiling
components

to still pot
higher temperature lower
concentration of low-boiling components
Fig. 24
Separating proceas on a plate in countercurrent distillation

vapour in a tube, or *icolumn”, between the boiler and condenser. Thus, while simple
distillation consiRts merely of evaporation and condensation, in countercurrent
distillation there is an exchange of material and heat between the two phases in the
column. This exchange tends towards an equalization of temperature between the
phases and to an alteration in the composition of the phases so that equilibrium is
approached (Fig. 24).
The nature of this equilibrium ib invariably such that as the concentration of a
component increases in the liquid, it also increases in the vapour (or in a certain
limiting case, remains constant; it never decreases). As the first vapour reaches the
condenser at the beginning of a distillation it condenses completely to a liquid of the
same composition and starts to return through the column. Now this vapour, at the
start of the distillation, is in equilibrium with the contents of the boiler, or stillpot,
and contains a greater proportion of the lighter components; thus it cannot also be
iranchembook.ir/edu

4.1 Principles of simple and countercurrent distillation 45

in equilibrium with its own condensate. The vapour in equilibrium with this conden-
sate would contain still more of the lighter components [l]. The exchange between
the vapour and its condensate will therefore be in the direction that brings the vapour
into equilibrium with a lighter liquid than the contents of the still-pot, :.,nd this mill
enrich the vapour in the lighter components. Correspondingly the liquid will be
enriched in the heaviel' components.
This exchange continues as the liquid travels down the column and as the distilla-
tion proceeds, until a steady state is reached. There is then a gradient in the concen-
tration of liquid and vapour in the column so that each contains snore of the lower-
boiling material at the top than a t the bottom of the column. The concentration
gradient is accompanied by a corresponding temperature gradient, with the lowest
temperature at the top of the column. This is the principle of the fractionating colunln.
The exchange of material and heat is a physical process taking place a t the inter-
face between the two phases, and the surface area for exchange should therefore be as
large as possible. This surface may be supplied by the empty column, by packing or
by elements in the column such as plates, wire gauze or rotating components (chaps.
4.3 and 7 . 3 ) .
The separation is dependent on numerous factors, in the first place on the proper-
ties of the components of the mixture, and further on the characteristics of the
column and its contents and on factors related to the method of operation,
As a rule it can be assumed that two components of a liquid mixture having a
difference in boiling point of more than 50°C can be separated to a fair extent by
simple distillation. For this reason simple distillation is chiefly used on liquid mix-
tures containing high-boiling or even non-volatile constituents in small amountq.
Examples that can be quoted are the removal of dissolved, non-volatile substances in
the distillation of water, and the purification of solvents from high-boiling conta-
minants. An exception is formed by the so-called Engler distillation [ 2 ] , used for
determining the boiling range of mixtures (such as gasolines) having boiling points u p
to 2OO0C, where successive components usually differ but little in volatility. Simple
distillation is used here with a view to obtaining easily reproducible conditions.
Mixtures with a narrow boiling range cannot be separated by simple distillation.
As regards the mode of operation, batch distillation and continuous distillation are
distinguished :
- batch distillation ; simple or rectified (countercurrent) distillation in which a given
charge is partly or completely distilled
- continuous distillation; simple distillation or rectification in which the feedstock
is uninterruptedly passed into the apparatus and the separated different products
are continuously removed from the process.
Countercurrent distillation enables components to be separated having differen-
ces in boiling point of about 0.5 deg C, whilst this figure can be as low as 0.05"C if
extremely efficient columns are employed, as in the separation of isotopes. By the use
of selective methods and, in difficult cases, by combination with other methods of
separation such as extraction, countercurrent distribution and gas chromatography,
separations have been performed with mixtures previously regarded as inseparable.
iranchembook.ir/edu

46 4. Physical fundamentals of the separation process

I n the following sections the complex processes taking place in countercurrent

distillation will be discussed further (cf. Kuhn and Kuhn and Ryffel [3]).
Before every distillation, whether simple or countercurrent, a series of points
must be considered prior to starting the calculations. A plan of work for judging a
separation problem is given in Table 4.
More detailed instructions for carrying out distillations are given in section 4.14.
Inforriiation about the assembly of apparatus and the starting of the process is given
in sections 9.3 and 9.5, respectively.

Table 4
Plan of work for judging a separation problem

Criterion Point to be decided or calculated See

Section

1. Difference in boiling point Simple or countercurrent distillation 4.1

4.7.5.1
2. Vapour preesure curves Optimum pressure in distillat,ion 4.4
4.6.2
5.3
5.4
3. Equilibrium curve Theoretical plate number and conditions 4.6 -4.7.4
of distillation
4. Required separating effect Choice of separating method and oalcu- 4.7.5
lation of conditions in distillation 4.8
4.9
5.
6.
5. Required throughput Choice of apparatus, calculation of 4.11
dimensions and heat balance 4.12
5.1
5.2
I.
6. Separating ability of t h e column Testing the column 4.10
7. Automation 3Ieasuring and regulat.ing devices 8.

4.2 Fluid and interface dynamics

4.2.1 Wetting columns

The trherinalseparation of material makes use chiefly of four types of column:

plate columns (chap. 7.3.3)
c o l u m for falling film distillation (chap. 5.4.3)
colunins with various elements (chap. 7.3.4)
packed columns (chap. 7.3.2)
iranchembook.ir/edu

4.2 Flnid and interface dynamics 47

All these types have as a common characteristic the countercurrent of the t n o

fluid phases. Due to the geometries of the various coluinns, however, marked differen-
ces are found in the way the flows pass the column. The last three types can t - ~
classified as “wetting columns” since thin-film formation is their outstanding feat ur(1
and the two closed phases move against each other without mutual penetration. It i v
characteristic of the plate column, on the other hand, that more or less dispersed gas
bubbles penetrate through the liquid and combine again on the next plate. (Fig. 2 5 ) .
Suitable model liquids for a study of the inaterial exchange in countercurrent coluinns
are freons and inorganic chlorides with a dilute iodine solution [3a]. Radionuclides
have also been used to study liquid currents, e.g. szBr in a column for the distillation
of butadiene [3b].
The basic feature of a column with stationary or rotating elements is that the
elenients are regularly arranged. Sulzer’s packing is a wire gauze insert for falling film
distillation which gives rise to a prolonged zigzag path for exchange while in the
spray column (Spraypak) the liquid phase is dispersed due to the kinetic energy of
the vapour, the latter moving through the packing as a closed stream. This conipari-

Fig. 25
Phase flow diagrams for va.rious column types (Stage)
a) Bubble-c8.p column d ) Spraypak column
b) Sieve-plate column wit’h guided liquid e) Packed column
r) Sieve-pl2.te column with non-guided liquid f ) Falling-film column
iranchembook.ir/edu

48 4. Phpical fandementale of the separation process

son already gives an idea of the problems involved in paoked columns. The random
arrangement of the packing leads to non-uniform, changeable flow of the fluid
phsses, whereas in the other types of columns the currents move along regularly
arranged channels.
In packed columns a non-uniform distribution (maldistribution) is to be expected
both in the vapour and in the liquid phase. In the liquid it occurs for these reasom
r41:
a) The liquid flows more readil3- from the packing units toward the column wall than
in the opposite direction. Vapour condensation on the wall due to heat loss may
enhance this process. This form of maldistribution is called wall flow (chap. 4.8.1).
b) The distributing effect of the packing units themselves is insufficient so that a
channel formed by chance persists (channelling) (chap. 4.8.1).
c) The packed column contains packing layers which tend to give the liquid stream
some preferred directions. This is also called channelling. Such layers may be due
to the method of filling or the shape of the packing units.
d) The column is not in an exactl:. vertical position.
e) The reflux is unevenly distributed initially.
Hence we have:
maldistribution = channelling + wall flow.
Wall flow depends on both separating length and column diameter. With a ratio
of dK/dpKR= 20 a fixed wall flow of 10 to 20% is established, which in turn is in-
fluenced by the ratio of separating length to column diameter.
The relative effect of maldistribution is the greater the more separating stages a
column has. A maldistribution of, say, 10% reduces the plate number of a column
with 100 plates to 30 while a column of 10 plates is only reduced to 9. Since also in
industry the trend is toward more and more efficient columns the question of the
maldistribution of the fluid phases is gaining importance.
While it is relativelj- easy to cope with the problems of vertical column position-
ing and the distribution of the liquid research is concentrating on getting better in-
sight into the distribution of the fluid phases during the separation process. The
increase of both column diameter and separating length results in a relative decrease
of column efficiency. Thus, e.g., the radial extension and the distribution of a liquid
from a point source In passing through a layer of Raschig rings were investigated
experimentally by Bemer and Zuiderweg [4a]. The parameters varied were wettability
and ring size, effective separating length and liquid load. Measurements bg Farid and
Gann [4h] of radial and axial dispersion coefficients in packed columns of 0.1 and
0.3 m diameter for packings of spheres and Raschig rings of sizes 1.27 and 3.80 were
niade such that the influence of wall flow could be eliminated.
In all cases, the maldistribution in a column causes a reduction of the plate
iiumber since the local molar ratio of vapour to liquid is disturbed. This influence
was best estimated theoretically by- Huber and Hiltenbrunner [5]. They envisaged a
quadrangular model coliimn partitioned along its long axis with the two compartments
iranchembook.ir/edu

4.2 Fluid and interface dynamics 49

loaded differently. As a measure of maldistributionl they defined the relat,ivedeviation

of the local flow density from the mean value taken over the whole cross-section:
T
-
T

where L, = liquid flow density a t a given point, L = mean liquid flow density over
column cross-section. The partition is interspersed with several mixing points where
the two partial streams are niixed to a degree m, the mixing being associated with
radial transport of inatter (Fig. 26).

Fig. 26
Model of a column without lateral mixing Huber and
Hiltenbrunner [ 5 ] )
The column is divided into two compartments by a par-
tition. Within a compartment, the flow densities of the
gas G and the liquid L are constant but they may
differ from compartment to compartment. In the diagram
the gas flow is equally shared by the two compartments
whereas the reflux is not. I is a measure of maldistriba-
tion. Any given maldistribution can be approximated by
a division into a greater number of compartments.
Concentration for number of separating stages n = w .

The efficiency nm/n,of a packed column then is a function of maldistrihution 1

and degree of mixing m per separating stage :

where in = degree of mixing per separating stage, which indicates what proportion
of t,he column cross-section is involved in the mixing as the material stream traverses
a separating stage.
According to a model of Huher the degree of mixing m per separating stage can be
estimated to be
m = 27 rgp.
4 Krell. Handbook
iranchembook.ir/edu

50 4. Physical fundamentale of the separation proceee

where dm =- packing unit diameter and dK = c o f u m diameter. For example, for

&/dm = 10, m = 0.27; for &Idm = 30, tt? becomes 0.025.
In the former case the degree of mixing ni is so large that even a strong nddistri-
bution reduces the efficiency only slightly. In the latter case this does not hold any
more. The example is in accordance with the empirical rule that the dK/dFKratio
should lie hetween 10 and 30. The true value for the plate number of a column can
thiis be obtained only with a &IdFK ratio of 10: 1, hence at I RZ 0.
The relationships between the quantities of flowing matter in the two phases and
the various hydrodynamic parameters such as dynamic and static hold-up (chap.
4.10.5), longitudinal mixing and pressure drop (chap. 4.11) were studied in detail by
Kafarov, Dorokhov and Shestopavlov [6]. They found a quantitative connection
between dynamic hold-up and pressure drop and a dependence of the static hold-up
on regions of flow. A model for the non-stationary liquid flow in the packing was
elaborated on the basis of effective and dead space liquid volumes and was used for
the calculation of the liquid parameters. Moreover the dependence of the longitudinal
mixing coefficient on the vapour and liquid loads as well as the physical properties
of the liquid were investigated. The main causes of axial mixing in packed columns
were reported by Jonas [7]. The measurements of Timofeev and Aerov (chap. 7 [65])
were also made with a view to the question of the influence of column diameter on
efficiency. Liquid flow and liquid-phase exchange of material in falling film distilla-
tion with various packings were studied by Zech and Mersmann [270] since it is
generally i inpossible in physical exchange of material to measure the interface
which is effective in the exchange of material between a liquid and a gas.
The separation process in a packed column can already he influenced markedly
by the different surface tensions of the components of a mixture [8]. An improved
met,hod of calculating the surface tensions of vapour-liquid mixtures was developed
by Bauer [8b]. It is especially suited for inclusion in programmes for the calculation
of material data. If in separating a binary mixture the surface tension u of the reflux
increases (positive system) the exchange of material can proceed much faster than in
Rpstems with decreasing reflux surface tension (negative systeius). In negative
systems the reflux is dispersed into narrow streams and drops. However, this effect
becomes appreciable only when Jc > 3 dynelcni-l. Systems with interface tension
differences < 3 dyne/cni-' are termed neutral.
The influence of column diameter and surface tension on the HTU in packed
colunins in the countercurrent distillation of binary iiiixtiires was studied by Gomez
and Strumillo @a]. They found the relation
- gO.86 x F,O*O8 (18%)
with D =column diameter (62 - 200mni), F, = quotient of ~,low-boilingcomp. to
%lgh-bolllng u = surface tension.
camp.,
Values for F, of < 1, e 1 and > 1 correspond t o positive, neutral and negativp
systems, respectively.
Ponter et d.[ 8 c ] investigated the effect of adding an interface-active substance
on the efficiency of a packed colunin. They observed an increase in packing efficiency
corresponding t o a change of the wetting behavioiir as determined by contact angle
iranchembook.ir/edu

4.2 Fluid and interface dynamics 51

measiirenients. The additive had virtually no influence on the vapour-liquid equili-

brium. These investigations were compleinented by an assessment of the influence
of surface tension on binary liquid systems in distillation with total reflux [271] and
of the separating efficiency of packings on the basis of wetting data obtained under
conditions other than material exchange [272].
Jn process engineering the ratio of the effective surface area of the packing to the
given total surface area is called the degree of wetting B,:

B, = A x loo (%I,
Aeff
(19)
where Arrf = wetted surface area, A = total surface area.

a) 'p - 0 0

Fig. 27
\Vetting of it solid by a liquid
phase (p, = contact angle)
a ) total wetting
h) partial wetting
c) no wettahility
C) 900 * cp 6 180°

Niinierous relations have been worked out for the determination of the degree of
wetting of packing units b u t they differ very strongly [9]. Questions of the effective
surface area of packings and practical possibilities of determining it independent of
the kind of process have been thoroughly discussed by Kolev [gal. He considered,
above all, the influence of viscosity.
It seeins essential that the energies of adhesion of the systems involved, the
possibility of a contraction of the liquid film in negative systems and the contact
angle (Fig. 27) are not taken into account in these relations. This may account for
the great differences between calculated and experimental data. Combining the
rquations after DuprC. and after Young yields the following simple relation for the
work of adhesion:
ws-l = o(1 + cos pl). (20)
It is obvious from this equation that a high value of A,-* requires a high surface
tension and as small a contact angle as possible.

4*
iranchembook.ir/edu

52 4. Physical fundamentals of the separation process

Since a system to be separated has a fixed surface tension there remains only to
achieve a small contact angle by choosing the appropriate constructional material
for the packing. Table 5 gives the energies of adhesion according to eq. (20) for
various vapour-liquid systems [ 111.

Table 5
Energies of adhesion of solid-liquid systems (liquid phase: water, 20’C)

Group Constructional 9 (deg) Nean ws-1

material depth of (erg. em-?)
roughness
Rrn (Pm)

Pure tungsten 62 109.0

metals platinum 71 96.5
copper 84 0.1. * ~0.5 80.4
aluminium 85 79.2
nickel 86 77.9
Alloys “St. 37” steel 79 86.7
X-10-Cr Xi Ti-18.9
steel 83 0.1...0.5 81.7
phosphor bronze 84 80.4
Plastics polyamide 72 96.3
polyester 72 96.3
plymethyl metha crylate 72 96.3
polyvinyl chloride 73 94.1
polycarbonate 77 < 0.1 89.2
polystyrene 80 85.2
polypropylene 90 72.8
polyethylene 90 72.8
teflon 106 52.5
Ceramics porcelain 50 5...15 119.0
glass 20 0.1 140.8

Obviously, the grest’er the energy of adhesion the greater the stability of the
liquid film formed on the packing although, on the other hand, it becomes increasingly
difficult to make a film of liquid form on the whole of a solid surface [ll].The wetting
capacity can be increased considerably by preflooding (chap. 4.10.8) and by choosing
an optimum geometry of the solid siirface of the packing [9]. Titov and Zelvenskiy [lo]
have reported 3 methods of calculating active phase interfaces in packed columns.
Graphical representations are given for the dependences of the activated surface area,
the HTU value and the material exchange coefficient on the liquid rate.
Krell and Heinrich [ 11m] assumed that the degree of wetting can only be exactly
precalculated if the contact angle of the liquid on the packing material is taken into
account. They investigated Raschig rings of ceramic and PVC material of sizes be-
tween 10 and 50 m.They varied the liquid phase so as to give contact angles y in
iranchembook.ir/edu

4.2 Fluid and interface dynamics 53

the range of 27 to 90". As a result of extensive series of experiments [ l l n ] the follow-

ing relations for the calculation of the degree of wetting B, have been formulated:
1. contact angle 0 5 pl 5 58":

_I
3. contact angle 58 5 93- 90":

q~= 0.489e (T
0.0192
--
0.39
d + - ) d2
6
1.245

x (1 - e-0.210 WL), (20b)

where d = diameter of packing units (mm), WL = liquid rate (m3/m2x h).
Relations (20a) and (20b) can be employed
- for all organic and aqueous solutions
- for all operating conditions below the point of flooding onset
- for any packing inaterial
- for any temperature
- for all column diameters and packing heights if these recomniendations are con-
sidered :
packing height < 1.000 mnm: 1 redistribution of liquid for each 14 emS of column
cross-section
packing height 1.000 nim : at least 1 redistribution of liquid for each 180 cin2
of colunm cross-section.
The successful use of the relations has been reported e.g. by Weia and Schmidt
[110] and Schmidt [llp]. They differ from those suggested by other authors in that
they allow for the constructional material of the packing and the surface ronghness.
The critical surface tension is not sufficient for an exact calculation of the degree of
wetting.
Kwasniak [llaJ studied condensation and vaporization effects in packed columns
on the assumption that there is a temperature difference between vapour and liquid
in each cross-section of the colunin. Thus the iinwetted portions of the packing sur-
faces can be regarded as heat exchange surfaces. The packing units were of identicaI
4hapes and consisted of copper or copper-coated plastic platelets so that there resulted
greatly differing heat conductivities. The platelets were arranged such that their
downward surfaces remained unwetted. The two packings differed very much in
separating efficiency which was due to the condensation and vaporization effects
occurring in the copper packing. This demonstrates that the effect of such processes
should by all nieans be taken into account. These considerations led Kwasniak to
develop a novel regular packing consisting of zigzag strips of sheet metal which are
oriented differently. Thus turbulences in the liquid and vapour phases and high
degrees of wetting are achieved.
iranchembook.ir/edu

64 4. Physical fundamentals of the separation process

4.2.2 Film formation

The model concept of accelerated wavefree flow on flat plates was elaborated by
Yilmaz and Brauer [I1h] for a theoretical description of the fluid dynamic behaviour
of liquid films in packing layers. Their considerations have been confirmed by mea-
surements. Investigations of the onset of wavelike flow on acceleration yielded a
diagram which is valid for the different flow states. From this it follows that, for
example, film flow in packing layers is practically always laminar and wavefree.
On these theoretical grounds equations have been developed for the calculation of
the operating liquid contents of Raschig ring layers (see chap. 4.10.5).
A detailed study of the hydraulics of countercurrent, columns for falling film
distillation with an insert of corrugated sheet metal was made by Kiinne [ l l c ] with
air-water as test system at room temperature and normal pressure. He gives in-
formation about column dimensioning and suggests models for calculating the
hydraulic resistance and the upper limiting load of the vrtpour phase. A method for
the experimental determination of the vapom-liquid interfaces in falling filni columns
is reported by Antonov et al. [ 11h].
Marschall [ 11kl describes methods which allow the undisturbed measurement of
the local film thickness, the wave amplitude, wave frequency add wave length as
well as the inclination of the film surface as a function of time. It is interesting to
note that these methods make use of the scattering of laser beams. Vorontsov [ l l q ]
made a systematic study of the effect of the t3-pes and dimensions of regular surface
featurea providing roughness on walls with vertical film flow.
Starting from the hydrodynamic model of the liquid film generated mechanicall>-
in coluinns wit,h rotating elements and using simplifying assumptions Dietz et al.
[ l l d ] deduced an equation for the calculation of film thicknesses in the range 0 < film
thicknem < gap width :

where

0 = mean film thickness, x = 0.304, Ti = inner radius, e = density, n = number of

r.p.m., r = mass flow related to circumference, p = dynamic viscosity, rA= radius
u p to outer edge of brush.
The special flow processes in thin-film stills were studied by Godau [lle].
Representing the relevant parameters and mathematical relations Billet [11f ] also
deals witahthe continuous distillation from a thin film as exemplified by a Lipotherni
thin-film still. Arithmetical work with the relations obtained indicates that operating
with reflux can yield maxiinum separating results under certain loading conditions.
The possibilities of intensifying the transport of material in falling films were
studied systematically by Wiinsch et al. [ llc]. Special consideration was given to
effects due to the ripple of the film, the curvature of t.he phase interface, the roughness
of the solid surface and the intentional distnrhances of the film flow.
iranchembook.ir/edu

4.3 Miscibility of the components 55

4.2.3 Column dynamics

On the basis of mathematical models attempts are made to precalculate the

dynamic hehaviour of distilling columns. According to Kohler and Schober [364]
column dynamics involves these problems :
- the study of the column hehaviour in time due to process disturbances (single or
combined disturbances)
~ the simulation of starting and stopping processes and of changes of operation
(intentional or unintentional)
~ the recalculation of plants during periods of nonstationary operation
~ the calculation of stationary states as limits of dynamic conditions
~ model calculations concerning the control of plants
- recommendations for improved designs taking into account essential effects
occurring within nonstationary periods (changed vapour-liquid loads)
- the theoretical explanation of “pulsating effects” in separating plants and the
calculation of variants t o eliminate such effects.
Adolphi [11g] evolved a graphical method for calculating the dynamic beharioitr
of hinary distillations which is based on the McCabe-Thiele diagram. The behaviour
of packed columns with countercurrent distillation was examined by Wagner et
al. [ 162a, h]. The behaviour of ideal and non-ideal binary and of ternary systems is
reported by them. The behaviour of coupled distilling columns has also heen in-
vestigated. Of course, electronic computers are employed on a large scale for coluinn
tl) namic calcnlations (see chap. 4.15). The limiting states in the separation of binary
inixtures as affected by disturbances and possibilities of removing or diminishing the
latter have been examined by Wunsch [lll]. He concludes that the enthalpy coin-
position diagram may provide important information about the disturbing behaviour
of separating units.

4.3 Miscibility of the components

Experience tells us that the solubility of liquids is the higher the more closely the
snbstancea are related chemically (homologous series). Regularities occurring in the
iiiixing of organic compounds have been tabulated by Staudinger [ 121. I n most
cases the miscibility increases with increasing temperature until we find complete
iniscibility above the critical mixing temperature Tmkrlt. I n Fig. 28 the solubility
diagram of water-phenol mixtures is given as an example.
I n general distillation uses completely niiscible liquids since in case of phase
formation a separation is carried out a t first by decanting. I t should be noted,
however, that there is no complete insolubility. Since dissolved substances cannot be
separated mechanically it is distillation which results in a separation.
Carrier vapour distillation - chiefly steam distillation - and azeotropic distilla-
iranchembook.ir/edu

56 4. Physical fundamentals of the separation process

tion deliberately use mixtures with no or with partial solubility. Carrier vapour
distillation is employed in order to have lower vapour temperatures in the mixture
while in azeotropic distillation the addition of a selected substance serves to produce
an azeotrope of the added substance and one component so that the latter can be
separated from the other components. It is necessary that the azeotropic distillate
should be easy to separate into its components (see chap. 6.2.1). Thus in these cases
it is very important to know the solubility diagram of t,he azeotropic mixture to
achieve R phase separation by proper cooling. From Fig. 28 it can be seen, for
example, that above 68.8“C the solution is homogeneous for all concentrations where-
below this teniperature demixing is dependent on the concentration.

While the solubility of’two coinponents may range from vi~tuallyinsoluble to

uoiupletely miscible the boiling behaviour displays greater variety. I n ideal mixtures
the interiuolecular forces ect8ingamong like and unlike molecules are equal. If thtl
forces between unlike nlolecules are smaller than those between like molecules a
minimum-boiling azeotrope can be assumed, while in the converse case a maximum-
boiling amtrope will he found. Intermediate are those non-ideal mixtures the equilib-
iranchembook.ir/edu

4.3 Miscibility of the components 57

rium curves of which will approach the diagonal asynlptotically from above or helow.
Stage [13] has given a snmniary of the relationships involved (Table 6) and Fig. 29
shows the corresponding types of equilibrium curves. These questions will he ~11s-
cussed in greater detail in chaps. 4.6 and 6.2.
Hildebrand and Rotariu [14] have considered differences in heat content, entro])?
and activity and classified solutions as ideal, regular, athermal, associated and sol-
vated. Despite niuch fundamental work the theory of binary liquid mixtures i h still
essentially unsatisfactory as can be seen from the systematic treatment of hinary
mixtures by Mauser-Kort urn [ 151. The thermodynamics of mixtures is presented nioqt
instr~ictivelyin the books of Mannchen [16] and Schuberth [17]. Bitkrich et al. [17a]
give an account of model calculations concerning thermophysical properties of piiro

loo~71~F~looo-q
a n d mixed fluids.

50
b) C)

0 50 100 0 50 100 0 50 100

Fig. 29
A in liquid, mol% -
Vapoq-liquid equilibrium curves of binary mixtures
a ) Benzene-water e ) Ethanol-water
b) Water-farfural f ) Methanol -muter
(solubility of A in €3 < B in A) g) Benzene-toluene
c) Water-n-butanol 11) Ac&one-acetic acid
(solubility of A in B > B in A) i ) Acet’one-chloroforni
d) Sec.-but.anoI-wa.t,er Kit vie acid- r\at.er
iranchembook.ir/edu

58 4. Physical fundamentals of the separation process

Table 6
Boiling behaviour of binary liquid mixtures

Type in Solubility of the Boiling Comparative magnitudes

Fig. 29 two components behaviour of forces between identical
(a,/,, a,la) and non-identical
(allo) molecules

a virtually minimum a,/, 4 a,/1 and/or a,/2

insoluble boiling point
b partially soluble minimum
C
d
(solubility gap) boiling point a1i2<a,/, and/or a,I2
e completely minimum all2 .<a,/, and/or agi2
miscible boiling point
f completely non-ideal, a, ,: and aZi2
miscible without ezeotrope
g completely ideal a1 * and a?
miscible
I1 completely non-ideal, al,, > alil and
miscible without azeotrom ~ ~ ~~

1 completely maximum a, > alll and a,!,

ni iscible boiling point
completely formation a],*>>a111 m d we
m i ~ible
c of compound

4.4 Vapour pressure-temperature relationship, p -t

In order to assess a separation problem we need to know the vapour pressure-

temperatiire relationship. By plotting the pressure of the saturated vapours against
the temperature we may gain some idea of the pressure at which the separation
appeam most promising; the abscissa is best drawn on a logarithmic or related scale
(Fig. 39). The choice of the pressure depends on whether me wish to perform
a ) an analytical distillation,
h) a preparative distillation,
c) a production distillation or
(1) a distillation for comparison with a large-scale operation.
In general we should choose a pressure for the distillation such that, at the one
extreme, the temperature in the column head is adequate for complete condensation
with the cooling medium available and, at the other extreme, that the temperatures
are not high enough to cause decomposition.
In analytical distillation it should he the aim to achieve a niaximuni separating
effect : the relative volatility should therefore he as large as possibie, and in an ideal
mixtiire thiR is so at the pressure at which the ratio of the vapour pressures is a maxi-
iranchembook.ir/edu

4.4 Vapour pressure-temperature relationship, p - t 59

mum (see chap. 4.6.2). The same applies to preparative distillation, whilst in the case
of production distillation account should be taken of the increase of the volume of
vapour with decreasing pressure, which may limit the throughput. In comparative
distillation the pressure must be adjusted to that maintained in the larger installation.
Extensive collections of saturated vapour pressures have been given by Dreishach
[l8], who also worked out the physical constants for 38 “Cox chart” families. I n these
tables [19] the boiling points for about 500 organic substancesare quoted at 1,10, 100
and 760 min Hg. Stull[20] has collected the data for about 1200 organic compounds
at 1, 5 , 10, 20, 40, 60, 100, 200, 400 and 760 inm Hg. This collection furthermore
contains figures for about 300 inorganic compounds and gives about 100 values for
the region from 1to 60 atm. Stage has published corrected saturated vapour pressures
for various classes of compounds, viz. for the saturated straight-chain C, to C,, [21]
fatty alcohols, for the methyl esters of the saturated straight-chain C, to C!,* [ 2 2 ]
fatty acids, for phenols [25]and the volatile components of alcoholic malts and crude
spirits 11241. Extensive tables have heen given by Jordan [25]. Riddick and Toops jr.
[26] have provided vapour pressures and other physical data for 254 solvrnts, whilst
Sage and Lacey [27] have tabulated vapoiir pressures and equilibrluni curves for
low-boiling hydrocarbons, H,S and CO,. Maser and Rueland [28] determined thr
vapour pressure ciirves in the region of 1 to 60 atm for 33 substances, previously not
investigated. The literature on vapour pressures for medium- and high-vacuum worli
has been reviewed by Jaeckel [29]. More recent cornpilations are t o he found in the
publication ,,Fortschritte der Verfahrenstechnik“ and in abstract-type publications
(cf. ref. chap. 1 [16]).
Although numerous data are hence already available, it may be necessary for tlir
laboratory worker to determine vapoiir pressures himself. The more reliable course
is to measure data experimentally, though several methods of calculation, descrihed
in section 4.4.2, can be followed.

4.4.1 Measurement of saturated vapour pressures

Fundamentally there are two methods by which saturated vapour pressures of a
pure substance can be determined, namely
a) the dynamic method : the measurement of the boiling point at various pressures :
h) the static metshod: the measurement of the vapour pressure a t various tempera-
tures.
I n the Houben-Weyl manual, Kienitz [30] gives a detailed description of how the
measuring should be done. Methods and apparatus for the determination of low
vapour pressures are reported by Milazzo [31]. He shows that the McLeod manometer
yields very satisfactory results with pressures in the range of 10 to 10-5 tnrr.
As the experimental execution of the static method is rather difficult, parti-
cularly a t high temperatures, whilst pressure control presents no difficulty with the
accurate regulators now available, the dynamic method will generally be chosen. I t
is important that the substance to be examined should be checked for purity.
For an accurate measurement of the boiling point it is essential for the boiling
iranchembook.ir/edu

60 4. Physical fundamentale of the sepaxation process

liquid to be mixed thoroughly by the bubbles of vapour. The classical aqparatm for
measuring boiling points is Swietoslawski’s ebulliometer [32]. The semi-micro form
of the differential ebulliometer is depicted in Fig. 30. Besides boiling point deter-
mination,it is employed for emmining the purity of substanms and for studying the
azeotropy of multi-component systems.

Fig. 30 Fig. 31 Fig. 32

Differential ebulliometer Swi@oelamki’s ebullio- Ebullioscope with heat.
(semi-micro form for meter, as modified by insulating jacket,, after
10- 12 mi) (Swigtoslawski) Olszewski, for hetero- von Weber
azeotropic mixtures

The substance (10-12 ml) is placed in the arms A and B. In arm A there is a
fritted glas plug and an electric heating element. The resulting vapour entrains tJhe
liquid upwards and enters the tube T,,where the temperature is measured. The
vapour flows further along the tube T2,where the condensation temperature is
observed, and is completely condensed in condenser D. The liquid then returns to the
hoiling arin vis the drop counter C and tube E. Drop counter C serves to measlire the
rate of evaporation.
A t hree-stage ebulliometer is used for determining the azeotropic coniposition of
hygroscopic substances, such as pyridine bases and phenol. In this case the boiling
temperature and two condensation temperatures are measured. If the charge is a
pure substance or an azeotrope, all three temperatures should be the same. A further
iranchembook.ir/edu

4.4 Vapour pressure-temperature relationship, p -t 61

iiiodification for boiling point determination in the case of hetero-azeotropic mixtures

has been developed by Olszewski (Fig. 31).
The ebullioscope of von Weber, shown in Pig. 32, operates in a similar manner.
With this apparatus vapour pressure curves in the range of 10 to 760 mm Hg can be
determined, whilst it can also be employed for other purposes, for instance for the
calibration of thermometers, ebullioscopic measurements, the determination of
vapour-liquid equilibria and the characterization of distillate fractions, such as those
from petroleum and tar [33]. The apparatus consists of a vertical boiling tube A of
34 inn1 I.D. and 500 nim length. At its lower end the liquid is heated by an electric
soldering iron element B of variable output, which is contained in a well C , the outer

short
manometer
“I
vacuum

Fig. 33
L oil pump

General arrangement for the measurement of saturated vapour pressures witah the apparatus
of von Weber

wall of which is coated with a layer of fritted glass to prevent bumping. A collar D,
reaching almost to the bottom, collects the bubbles of vapour, which are led together
with liquid to the spray tube R (4 mm I.D.),where they are whirled together in two
bulbs and then sprayed against the thermometer by the jet F . The spray tube and
bulbs are surrounded by an evacuated jacket G ; the liquid separating from the vapour
a t the thermometer enclosure flows back into circulation through a narrow annular
space between the vacuum jacket and the boiling tube. The thermometer is immersed
in a little mercury in the protecting tube H , which has a fused-on dropping cap
diverting the downflowing condensate from the thermometer enclosure. The mercury
thread is surrounded by condensing vapour. The thermometer scale is illuminated
from behind and is read through the layer of condensate, estimating to l/lOth of a
scale division; no difficulty will be experienced in doing this if the upper part of the
ebullioscope is etched smooth before use with 1% hydrofluoric acid for 2 minutes,
and is occasionally boiled out with soap solution. The apparatus, up to the condenser
iranchembook.ir/edu

62 4. Physical fandamentala of the separation process

K, is surrounded by an insulating jacket J of glass fibre provided with a narrow

viewing slit ; inside this jacket the tube carries a winding of thin resistance wire E for
thermal compensation. It can be regarded aa an infinitely long cylinder surrounding
the measuring point, in which the source of heat is uniformly distributed. With the
aid of a voltmeter and a calibration chart this heater is so regulated, with no current
in the boding element, that the predicted temperature of measurement is approxi-
mately maintained in the appsmtus. Even the vapour of a high-boiling liquid will
then rise up to the condenser, 250 mm above the thermometer bulb. The fourfold
protection provided by the evacuated jacket, the liquid flowing back at boiling
temperature through the annular space, the thermal compensation and the fibre
jacket ensures an adiabatic condition in the spray tube. The liquid charge is introduced
at the top through the side tube I;, which also serves to connect the suction tube for
measurements in vacuum.
The whole arrangement, provided with a manifold to which pumps, ebullioscope
and manometer are connected, is shown in Fig. 33. The pressure is regulated to
exactly 760 nun Hg or to any desired pressure by the methods mentioned in section 8.3.
A similar device has been described by Eblin [34].
Measurements with an accuracy of f0.003deg. C can be carried out with an
arrangement due to Herington and Martin [35], who used it for determining the
vapour pressures of pyridine and its homologues in the range of 50" to 157°C. The
method of operation is shown in Fig. 34. The high degree of accuracy is attained by
evaporatingspeciallypurifiedwater invessel B under thesame conditionsas in the boi-
ling vessel A, and by calculating the pressure from the vapour temperature of the water.
9 very elegant method, which, however, yields results accurate only to f0.5 nini
Hg, is described by- Natelson and Zuckerman [36]. It consists in causing a drop of the
liquid hanging at the end of a vertical capillary (thermometer tubing) to faU by
gradually evacuating the surrounding space, the prevailing pressure at this moment
being read off from a mercury manometer. For carrying out measurements on a
micro-scale, with only a few milligrams of material, a procedure due to Emisch has
been so modified [37], that observations can be made at pressures below 100 mm Hg.
Boiling points can also be determined withminute amounts of substance by the use of
Kofler'a micro heating plate [38].
Radioactive isotopes may be employed in the measurement of vapour pressures at
low temperatures [38]. This technique is described in detail by Nesmeyanov [39],
with special reference to metals and alloys. For substances of low volatility such as
iodine, naphthalene and phenol, the effusion method is also suited [40].
We will now describe some static procedures, in which the temperature is kept
constant and the pressure is observed.
Vapour pressures can be determined in the range of -20" to +20°C by a method
described by Henning and Stock [41]. The apparatus is illustrated in Fig. 35.
The substance is introduced into the flask through the opening at B. A thernio-
meter divided into tenths of a degree C is attached to the hook on B and the flask is
placed in a cooling mixture. A vacuutn pump is connected to. C; cock D serves to
shut off the vacuum. The riser tube E is placed in the mercury bath F of the barometer
G. The apparatiis is now evacuated and cock D (which has previously been tested for
iranchembook.ir/edu

4.4 Vapour pressure-temperature relationship, p - t ci 3

tightness) is closed. Flask A is then heated a little, so that the vapour pressure in-
creases; heating is interrupted and cock D opened. This operation is repeated until
all the air has been removed from the apparatus. The actual measurenients can now he
5tarted. A constant teniperature is established in flask A , and by the use of a plate
glass scale - or better still a cathetometer - the difference in height of the niercur;\-
menisci in the limbs E and G is measured. This difference is the vapour pressure, i n
inn1 Hg, of the liquid tested. It is necessary to carry out a series of observations a t
each temperature and to take the mean figure. It is best to test the method first on
a piire substance having known vapour pressure values.

A B
-
Fig. 34
Arrangement of apparatus for boiling point determination (Herington and Martin)
A = Boiling vessel for pyridine bases, B = Boiling vessel for pure water, C = Oil viicuum primp.
D = Gas holder (30 l), E = Pressure regulator, F = Gas holder (30 1) in water bath of 400 1.
C: = High-vacuum oil pump, II = Vessel with P,O,,I = Mercury vapour pump, .I = McLeocl
manometer, K = Nitrogen bottle, L = Solution of ammonia +
ammonium carbonateon
metallic copper. M = as L, N = Washing bottle with dilute sulphnric acid, 0 = 10% caustic.
soda, P = Conc. sulphnric acid, (3 = Anhydrous calcium chloride, R = Silicagel, S = Phos-
phorus pentoxide, T = Glass wool, U = Cold t r a p
iranchembook.ir/edu

64 4. Physical fundamentals of the separation process

thermostat
Fig. 35
Apparatus for measuring vapour pres-
sures of volatile substances (Henningand
Stock)
A = Flask for liquid with thermometm,
B = Ground joint.with hook for thermo-
meter, C = Vacuum connection,
D = Cock, E = Riser tube, F = Dish
of mercury, G = Barometer

t-

=
Fig. 36
Isoteniscope as modified by Schuberth
iranchembook.ir/edu

4.4 Vapour pressure-temperature relationship, p - f 65

With the isoteniscope of Smith and Menzies [42] it is possible, by the use of an
auxiliary manometer and a n artificial atmosphere, to determine the vapour pressure
curves of higher-boiling substances. The method gives fairly reliable results. Fig. 36
shows a modification of the apparatus after Schuberth. One of its main features is
that the substance to be investigated can be distilled into the auxiliary manometer B
and that the vacuum from G in the buffer flask can be so regulated by the cock that
the auxiliary manomet,er B becomes adjusted to zero difference in level. The main
manometer then indicates the vapour pressure of the substance exerted on the left-

0'
Pt

Apparatus for static vapour pressure determination of Kortiini et al.

a = Conical ground joint, b = Choke, c/d = Platinum contacts, e = Ground
joint, B 19, f = Ground joint, B 12, g = Leads to platinum contacts, h = To the
main manometer HM, h,/h, = Cocks, i = Vessel for liquid, k = Condensation
vessel, E = Vapour space, ml/m, = Floats, n = Storage vessel, ZM = Inter-
mediate manometer

hand limb of the auxiliary manometer B. The isoteniscope is placed in a thermostat

jacket G, so as to maintain the liquid and the auxiliary manometer a t the same tempe-
rature.
The determination is carried out as follows [43].Bulb A of the isoteniscope tube
is filled about three-quarters full with the substance in question and the isoteniscope
is connected by a piece of vacuum tubing to the artificial atmosphere C. The thermo-
stat bath is warmed, whereupon the whole apparatus is slowly evacuated. As soon as
the pressure has fallen sufficiently the substance in A will begin to boil, but will again
condense in bulb D. The downflowing condensate collects in the auxiliary manometer
3 Krell, Handbook
iranchembook.ir/edu

66 4. Physical fiindamentals of the separation proces8

B and thus shiits off the substance boiling in A from the air in the artificial atmo-
sphere. When it is believed that the air has been entirely eliminated from the left-
hand part of the isoteniscope, a constant temperature is established in the thernio-
stat, and the vacuuni is regulated such that the liquid in both limbs of the auxiliary
manometer is a t the saine level. When this is the case, the thermostat teniperature
tlnd the presslire indicated by t,he mercury manometer are read off. In order to niake
sure that there is no longer any air in the closed part of the isoteniscope, the evacuation
and pressure adjustment are repeated. If the same result is found, the temperature
i s gradually raised and the pressure is changed so as to maintain the liquid levels in
the I'-tube a t approximately the same height. An excessive outer pressure must at
all times he carefully avoided, so as to prevent air from being forced back into bulb A.
The final adjustment is best done by making small changes in the bath temperature
at a constant pressure of the artificial atmosphere.
For more accurate measurements, with an uncertainty of f0.2 mni Hg, Kortdrn
et al. [a] designed the apparatus shown in Fig. 37, which was employed for deter-
mining the vapour pressure diagram (i.e. the total vapoiir pressure of a mixture) of
the s\-steni water-dioxane at 35°C.
A straightforward static method for vapour pressures of aqueous salt solutions
(LiQ, MgCl, and Mg(NO,), solutions at 20 t o 90°C) is reported by Schliinder [46].
The mass-spectroniet ric met hod of nieasiiring low vapour pressures enables the
partial pressure of a particular substance to be determined independent of any
i m pri t ies.
Three different static methods were studied by Rock [16],who measured very low
aniline vapour pressures (0.056 to 17.9 tom). Measurements using the Hg manometer
and the aniline nianonieter proved to be niost convenient. An autoclave of V4A steel
was used by Glaser and Riiland [28] for the determination of vapour pressures u p to
60 atni. and 500°C.

4.4.2 Calculation and representation of saturated vapour pressures

If there is no opportunity to determine vapour pressures, it is possible to obtain

the required data from figures published in the literature by the method of increments.
This method, whether carried out graphically or by calculation, gives approximate
resul t e.
Pbter [47] investigated the relationship between vapour pressure and molecular
structure on the basis of the van der Waals pressure equation:

where I = the reduced pressure PIP, and 0 = the reduced temperature ?'/T,.
By calculating the constant a additivelg for atomic and structural increments a
good agreement with experimental results is obtained. In about half the cases in-
vest igated the difference hetween calculated and measured values of a amounted to
less than 1 per cent, and the variations in a due to temperature were small. By
iranchembook.ir/edu

4.4 Vapour pressure-temperature relationship, p - t 67

evaluating 73 experimental results, and by a process of averaging, the following

values for the increments in a were determined.
C -2.154 N -0.765
H 1.139 C1 1.356
CH, 0.124 COOH 2.247
0 0.266 COO (ester) 0.507
OH (alcoholic) 2.327 Double bond 2.295
CoH, 1.708 Triple bond 4.605
Branch in C-C chain 0.056
Knowing the structural -.xniula of a su-stance, its critical temperature and critical
pressure, its vapour pressure a t any temperature can be calculated according to (21).
The additivity of the constant (I is comparable to that of the rnolecular refraction a n d
the parachor. Kolljr and Nagy [47a] report a method which is based on Kinney's
method and enables vapour pressure curves to be calculated from structural data in
a range of boiling points between 0 and 400°C. However, the niolecules should not be
too polar. The vapour pressures of high-boiling complex hydrocarbons may be deter-
iiiined according to Smith et al. [47b] if a boiling temperature (10 or 760 torr) and
the rough structure (proportion of side chains, aroniatic and naphthenic groups)
are known. The calculation is based on the fact that vapour pressure curves of coniplex
hydrocarbons are similar to those of n-paraffins with the same numbers of carbons.
4, similar method for estimating the boiling point of organic substances has been
developed by Pearson [48], who employed a factor for dealing with association and
justified his method on the basis of the volunie and shape of the molecules. Similarly,
Eigloff's boiling point equation [49] allows boiling points to be calculated. If the
boiling points of a compound at two pressures are known, as well as the total vapour
pressure curve of a reference substance, the vaporir pressure curve of the compound
in question can be calculated by an application of Diihring's rule 1501:

(k14 - tPA)/(tlB - & B ) = q (22

in which: tI4 = boiling point of the substance A in question a t a pressure P,;
t,, = boiling point of the substance a t a pressure P,;
t l L c= boiling point of the reference substance B at a pressure P,;
tZn = boiling point of the reference substance B a t a pressure P,;
q = a constant
This rule hence states that the differences in boiling points a t two pressures of two
substances A and B bear a constant ratio.
Taking water as a reference conipound, as has been done by von Rechenberg [51]
for a nuniber of compounds (water being a substance with very accurately known
vapour pressure data), we might arrive a t the following calculation:
I . Known values :
P, = 760mni P , = 100 inin Hg
Acetic acid tl, = 118.1"C t,, = 63.0"C
Water LJj = 1oo.o"c t2p, = 51.5"C
iranchembook.ir/edu

68 4. Physical fundamentals of the separation process

From these figures we first compute the value of q, which is used in the subsequent
calculation :
118.1 - 63.0
4 = 100.0 - 51.5 = 1.14;

2. Calculation of the boiling point tan of acetic acid at 200 mm:

PI = 760 111111 P, = 200 mm Hg
Acetic acid t l A = 118.1 "C t2* = 1 "C
Water tl, = 1oo.o"c t,, = 66.4"C
t 2 A = t l q(t,B
~ - t Z B ) = 118.1 - 1.14(100.0- 66.4) = 8o.o"c.
This value agrees with that found experimentally, and further figures can be
calculated in the same way. It should be observed, however, that measured data of
other reference compounds in the lower pressure range are frequently very inaccurate,
in which case calculation by Duhring's rule will lead to erroneous figures. An accuracy
of at lea& k0.05 deg. C in the basic data should be a requirement. It is advisable to
choose reference substances that are closely related in structure to the compound
in question, for instance by belonging to the same homologous series. Jantzen and
Erdmann [52] employed myrist'ic acid as reference substance for calculating the
vapour pressure curves of the higher saturated straight-chain fatty acids. In practice
it,will be found that water can be used for polar liquids and n-hexane for non-polar or
weakly polar liquids. A similar method has been given by Kirejew [53].
Frequent,ly the Antoine equation [54] is used as a vapour pressure-temperat~ire
relationship:
b
logP=a--
t+c'
This formula contains three constants depending on the nature of the substance, a, b
and c, of which c is a group constant that can be taken to be 230. The use of this
formula therefore entails a knowledge of the constants a and b for the compound in
question, as found by experiment. Vapour pressure data are quite often quoted in
this form in the literature, for instance in the tables of Dreisbach 1191.
Wichterle and Linek [58] have prepared lists of Antoine constant<sof pure com-
pounds to be used in electronic computers for process control. The individual para-
meters were calculated using an iterative procedure; standard deviations are given.
According to extensive investigations carried out by Thomson [55], the Antoine
equation is the equation giving values closest to the experimental results. Constantsa
and b of the Antoine equation can be calculated if the boiling point of the substance
atrtwo pressures P and P , is known. Constant b is given by:
log P - log P ,
b=
t - tW
(t + 230) (t, + 230) (24)

whereupon constant a can be obtained by int,roducingb and c (= 230) into equation

(23) as
b
a = log P - +
t 4.230'
iranchembook.ir/edu

4.4 Vapour pressure-temperature relationship, p - t 69

The values for P, and t, can be taken from Table 7 for the individual “COX chart
families”.
If the normal boiling point and the critical point of a substance are known the
whole vapour pressure curve can be calculated with a universal vapour pressure
formula after Riedel [56]. If data obtained a t any two temperatures are known thci
critical point can be estimated. The formula is characterized by particular exactness.
It indicates whether or not the substances associate. The method of determining
boiling points as developed by Reckhard [57] is particularly suited for the vacuum
distillation of multicomponent mixtures such as tars, tar oils and similar product,s.
Two nomograms facilitate computational work. A comprehensive presentation of
the methods of determining vapour pressures is contained in [59]. Othmer and
Yu [60] give a survey of the application of relations with reference substances. 9
generalized vapour pressure equation for hydrocarbons has been established by Zia
and Thodos [6Oa]. The mean deviation of calculated oapour pressures from those mea-
sured is about 0.38%. A critical coniparison of various vapour pressure equations for
a large number of substances was made by Gomez-Nieto and Thodos [60b] on the
basis of the correspondence principle. A new equation using particularly the tempera-
ture of the normal boiling point besides the critical values for a given substance
yielded a mean deviation of 1.43% for 7633 data from 139 substances. Employing
Riedel’s parameter even lowers this deviation to 1.05%.
Further results of these workers [ ~ O C60d], concern hydrocarbons and few in-
organic substances. Westmeier and Picht [60e] made a comparative study of general-
ized methods for the calculation of vapour pressure. They place particular emphasis
on those methods which describe the vapour pressure behaviour of polar substances
with adequate accuracy.
Data on vapour pressure calculations for liquid metals and alloys (Li, Na, K, CS,
Mg, A l , Pb) are found in a paper by Waseda et al. [6Of].
The graphical methods for determining vapour pressure generally have the ad-
vantage of being simpler and more rapid. For instance, Diihring’s rule can he followed
graphically by plotting the boiling points a t equal pressures on linear scales along
two co-ordinate axes, those of the substance A as ordinates and those of the reference
compound B as abscissae. These points then lie on a straight line. Fig. 38 shows a
plot of the data for acetic acid and water, mentioned above. The vapour pressure of
acetic acid at a required temperature can then be found by reading off the correspond-
ing temperature for water from this graph and then finding, on a vapour pressure
curve for water (Fig. W),what the vapour pressure pertaining to the latter tempera-
ture is. It is also possible to draw a scale for pressures in the Diihring graph (as has
been done in Fig. 38)) but it, is then not easy to read off intermediate values.
Fig. 39 shows the normal course of vapour pressure curves, and i t is evident that
a fairly large number of points are needed t o draw the curves accurately. Methods
have therefore been sought for obtaining a linear representation. It follows from the
Clausius-Clapeyron equation

A,H 1
logP= --.-
R Tfcy
iranchembook.ir/edu

70 4. Physical fundamentals of the' separat,ion process

Table 7
Cox chart families and their focal points 9, and p ,

Family Formula am 'C p, Factor

+
'.x

rnm Hg (8- 230)

x log
(PcOl760)

Aromatics with Ratureted

side chains
2.320 6621
Halogenated aromatics with
saturated Ride chains
Unsaturated aromatics
U nsatnrated halogenated 0.9lG 3 985
aromatics
Naphthalenes 0.530 3 274
Halogenated naphthalenes
Aromatic amines
Halogenated aromatic 3.800 7 324
amines
Phenols 11.170 10 697
Halogenated alkylphenols
Nitrobenzenes 3.701 8 304
Halogenated nitrobenzenes
Aromatic aldehydes
Halogenated aromatic 0.150 1905
aldehydes

I
Phenylethyl alcohols
Halogenated phenylethyl 3 300 20.00 15550
alcohol8

1
Aromatic ketones
Halogenated aromatic
ketones
5.186 9 263
Aromatic-aliphn tic k e t ~ n e s
Halogenated aromatic-
aliph~ticketones
i
1
2186

Aromatic ethers
Aromatic-aliphatic ethers
Cyolohexyl-aliphatic ethers
with saturated and
unsaturated side-chains
} 1850 3.400 7 595

A41kylcyclopentanes 1450 1.950 5 560

Alkylcyclohexanefi 2 076 2.076 8 104
Paraffins 1240 1.300 4 753
Halogenated paraffins I450 1A60 5 O9R
Mono-olefins 1725 1.698 F 648
Primary 7) -dkylamines 1125 1.441 4 424
Kp,,, below 190°C
iranchembook.ir/edu

4.4 Vapour pressure-t,emperaturerelationship, p - t 71

i
Table 7 (Continuatio?)

Family Formula 6, OC p, x Factor

mni Hg (6, f 230)
x log
(PCu/7W

Primary n-alkglamines 1270 1.709 5w29

K z ) 190"-350"C
~ ~ ~
Aliphatic ethers 1579 2.525 G 370
Aliphatic ketones 3293 G.929 13 950
Nitro paraffins 1350 2.000 5404
Aliphatic nitriles 2025 3.461 8 250
Aliphatic esters 1639 3.800 G913
Aliphatic alcohols 12150 128.200 64718
Aliphatic acids 364 0.060 1 12.5
Kp,,, below 142°C
Aliphatic acids
Z<p760above 145°C
Halogenated aliphatic ac I 1458 9.060 6881

10 20 304050
pltorr
100
-
200 300 500 760
tI I !

Kp. of reference compound water, B

Fig. 38
Plot of boiling points of acetic acid and water following Diihring's role
iranchembook.ir/edu

72 4. Phpeical fundament.& of the separation process

(in which R is the gas constant, T the absolute temperatyre, A,H the heat of
evaporation and c a temporarily unknown integration constant), that a plot of log P
against 1/T should give a straight line,
Fig. 40 shows such a plot for a number of substances. The slope of these various
lines is obviously a measure of the heat of evaporation. Rewriting t,he Clausius-
Clapyron equation as

the,logarithm of the vapour pressure of a substance 2 may be plotted against, the

logarithm of the vapour pressure of a substance 1. This also yields a straight line. The
slope of this line is now determined by the ratio of the two heats of evaporation.
Hence it is appropriate to use n-heptane as reference substance for hydrocarbons and
water for associating substances, such as acids and alcohols, since these substances
have large heats of evaporation. On the bask of eq. (27) Ohmer et al. [63] have pre-
pared a notilograni and corresponding tables hy means of which vapour pressuredata,
latent heats and other physical parameters can be determined.
If the boiling point and heat of evaporation of a substance are known the diagram
of Bergholm-Fischer [66] which is given by Wittenberger [64] can be used. Besides the
1/T vs. log p plot i t has, in its upper part, a kcal scale.
A method in which the logarithm of the vapour pressure is plotted along a “sub-
stance axis”, thereby yielding radiating lines has been given by Hoffmann and

Fig. 39
Vepour pressure curves
1 = water, 2 = benzene, 3 = sum-

t/t - 0
mated curve for water and benzene
iranchembook.ir/edu

4.4 Vapour pressure-temperature relationship, p - t 73

Florin [66] and is based on the well-known Cox charts [67]. I n these diagrams, which
can be drawn for classes of substances - the so-called “Cox chart families” - all
straight lines for the vapour pressures lead to a point (P,, T,) having a definite
position for each class of substance. I n this case it is therefore sufficient to know tlir
boiling point a t a certain pressure, t o plot, this point and connect it to the abovt.

700 -
600 -
500 -
400 -
300 -
200 -

100
-
28 I
I
5 30-
-& 20
-
I01
-
7-
L
4 -
3 -
2-

* ’ ’
-
1 I I f I I I I 1 I
127120110100 90 80 70 60 50 40 30 20 10 0 -10°C -20 3
t/”C
Fig. 40
Linear vapour pressure representation in the co-ordinate system log p agaimt
1/T
1 =propionic acid, 2 = water, 3 = ethanol, 4 = n-hexane; 5 = acetone, 6 = ether

inentioned “focusyyin order to obtain the whole vapour pressure line. I n Table 7 tlir
21 Cox chart families indicated by Dreisbach have been collected, together with their
focal points T m and Pm.Fig. 41 shows a Cox chart for the alkylbenzenes.
At first Cox plotted logp ngainst 118 and obtained a st,raight line a t a n angle
with the ordinate. According to the p-values of the ordinate the experiniental tempera-
ture values for water were plotted on the abscissa. Davis [68] and Calingart and Davis
1691 developed the Cox chart further, plotting logp vs. l/(S 230). Such a Cox +
chart may he considered to be the graphical representation of the Antoine equation
(18) with c = 230. However, this group constant applies only to aliphatic hydrocar-
bons [70].
I t is iiiiportant that a COXchart should be drawn on a sufficiently large scale. 111
order to represent the temperature with an accuracy of 1’ up to 200°C, the dimensions
iranchembook.ir/edu

74 4. Phvaical fundamentals of the separation process

~

of t)hediagram have to be about 0.96 x 1.5 m. Dreisbach [19] gives detailed directions
for constructing a Cox chart, in order that the values deternuned [nay have an accura-
cy sufficient for laboratory practice:
Oldinetes (pressures) - Divided logarithniically ; the eight divisions from 0.1 to
107 mm Hg should have a total length of 56 inches.

000

Fig. 41
Cox diagram for the family of the alkplbenzenes

Abscissae (temperatures) - The region froiii -100" to 10.0oO"C should occupy a

width of 38.75 inches. The temperature scale may he calculated by means of the
formulae
830L
I = or
33.17 - L
L = - 33.171
330 1 +
( 1 = teruperature, "C, and L = length in inches).

The zero point in the middle lies about 17.25 inches from the left-hand point for
-lOO"C, and the value of L is the positive or negative distance from this zero point.
It is advisable to subdivide the temperature scale according to the directions given in
Table 8.
iranchembook.ir/edu

4.4 Vapour pressure-temperature relationship, p - t 75

Table 8
Temperature subdivisions for a Cox chart
~

Temperature Subdivision
range I ("C) a t every

-100 to 200 10°C

200 t o 600 50 "C
600 tlo 2000 100°C
2000 to 4000 500°C
4000 to 10000 1000 "C

Further subdivision:

Temperature - 100/O°C 0/200"C 20O/60O0C 600/1 OOOOC > 1000°C

range I1

Subdivision' 1 10°C 10°C 50 "C 100°C subdivision

Subdivision 2 10 x 1 ° C 5 Y 2°C 5 x 10°C 2 Y 50cC yields no
useful values

190
180 -

170 -
160 -
150 -

140 -
130 -

1 120-

,u 110 -
100 -

90 -

80 -

6olv
70 -

60 -
Fig. 42
60 - Isobars for the homologous series
40, I I I I I I I I I of n-paraffins. C,,-C,,
20
straight-chain paraffins --+
iranchembook.ir/edu

76 4. Physical fundamentals of the separation proceas

Finally a method that is specially suited for homologous series will be described.
It consists in plotting the boiling points as ordinates on a linear scale and the carbon-
numbers of the components as abscissae on a logarithmic scale. The isobars a t various
pressures are then drawn in the diagram, as has been done in Fig. 42 for the straight-
chain paraffins C,,-C,, a t pressures of 5, 10 and 30 mm Hg. This method allows
values t o he interpolated, and data for members below C!, and above C,, to be found
bp extrapolation. It can also be used with advantage for giving an indication, during
8 laboratory distillation, of the molecular weight range in which distillate is situated.
There are many more methods of determining vapour pressure data which aim a t
obtaining higher accuracy. They are to be found in the publications of Hoffinann and
Florin [SS], Faldix and Stage [70], Stage and JuiLfs [71] and in [591. It is up to the
practical worker to choose the one method most suited to the given conditions. 16
should be pointed out that all these iiiethods are approximations and that high
accuracy is obtained only by exact measurement.

4.5 Equation of state and px-diagrams

liquid cnntains molecules moving at different velocities. For those moving

fastest it is easiest to penetrate the &quid surface, thus escaping into the space above
the liquid as rapour (evaporation). For the conversion of the liquid particles into
r a p u r heat of conversion is required, which is removed from the liquid, the latter
cooling down. If heat of evaporation is supplied from outside evaporation is achieved
more quickly and the liquid begins t o boil.
Hence, vapour is a gas near its boiling or dew point. I n distillation iisuczlly only
saturated vapours occur which are associated with their liquid phase, with equal
numbers of molecules going over froin the liquid to the gaseoi!s phase and vice versa
in the eqiiilibriuni state. An unsaturated vrqpour is forined only when the saturated
vapour is no longer associated with the liquid and is heated beyond the saturation
temperature or is expanded. Then it cannot take up more rnolecules from the liquid
phase according to the increased teniperature.
The general equcrticm of state (31) can only be regarded as an approxiination and a
limiting caRe for gases near the dew point a t high pressures w-ith the molecules asso-
ciated, i.e., it applies only to ideal gases and, strictly speaking, not t o saturated
vnpours which normally occur in distillation.
The van der Wads equation seeks t o account for this:

(50)

Here and b are constants, b and a/Vhol considering the covolunie of the mole-
cules and the intermolecular forces of attraction, respectively. p is the pressure
eserted by the gas on the walls of the vessel, and Vmolis the molecular volume. But
rrrn the van der Waals equation of state does not apply exactly t o any particular
fiiitxhnce since the intermolecular forces do not obey so simple a law as asmmed in
iranchembook.ir/edu

-r
4.5 Equation of state and px-diagrams i l

this latter equation. A fifth-order equation formulated by Planckprovides an equation

of state which approaches the real behaviour of gases and vapours most closely.
Obviously, calculations with third and fifth-order equations are difficult to carry
out and thus the simple gas equation based on the laws of Gay-Lussac and Bople-
Mariotte :
m
pv = povo 1
273.16
or
pV = RT (31)
is generally used in the practice of distillation, it being assumed to represent con-
ditions with sufficient accuracy. A gas approaches the more closely to its ideal state,
the lower its pressure and the higher its temperature.
If into the formula R = poVo/273.16 we introduce the values for normal con-
ditions, namely p , = 760 mm Hg (1 atm) and 8, = 22.414 1 (volume of one iiiole of
gas at O T ) , R becomes
1 x 22.4141273.16 = 0.0821 atm/'C.
For an arbitrary number M z of moles of a gas or vapour we hence obtain the equation
pJ- = M Z 0.082T. (32)
(In large-scale work one frequently uses a corresponding formula for kg-moles instead
of g-moles.)
To calculate the volume of vapour from a substance a t a definite pressure and
temperat Lire, the following formula can be employed:
M z x 0.082 x T x 760
V= (33)
P
At normal pressure (760 nim = 1 atm) and for one mole of substance formula (33)
simplifies to

T7 = 0.082T (34)
(T' in litres, p in mm Hg, Mz referred to g-moles).

The density of the vapour under normal conditions is given bj-

?j-J = M/22.4g/l, (35)

where M is t,he molecular weight.
At another pressure and temperature this becomes
M p 273.16
(JD = -.-. (36)
22.4 760 273.16 6 +
or
iranchembook.ir/edu

78 4. Physical fundamentals of the separation process

The weight of a certain volume of vapour is then simply

As we have seen it is the intermolecular forces that influence the solribility of the
components. This, in turn, determines the other prvperties of binury mixtures which
are summarized in Fig. 43 after Thorinann [72].
With practically no .mhbility the two components and the vapoiirs arising from
them behave as if either of them were there alone a t the same temperature (Fig. 43,
column 1). This applies to carrier vapour distillation (chap. 6.1).
-4ccording to Dalton's law the total pressure of a gas mixture is equal to the sum
of the pressures that each gas would exert at the same temperature in the same vapour

(39)
This leads t o the rompawlion. 01 the vnpour of insoluble niixtures, referred to the lowest-
boiling component, in mole per cent :

I n order that one niay obtain the partial pressures (E'ig. 43, series I) for such
calculations of vapour compositions, both vapour pressure curves are plotted and
the siinimat,ed curve for the two partial pressures is drawn, as has been done for water
and benzene in Fig. 39. S t 760 mill a boiling point of 69.0"Cis found for the mixture.
-4ccording t o fonriiila (40)the composition of the vapour is
5.36
y* = -x 100 = 70.2 inolo/o,
760
a figure that can he converted by the noniograiri of Fig. 22 to 91.Ogb by weight.
If we comider rows I and I1 of Fig. 43 where the diagrams show the composition
of the mixtiire in molo/o referred to the low-boiling component (see also Fig. 29) we
find these relationships:
Row I
Partial pressures of the low boiling component L nnd th high-boiling cornpunent W
( p x d i a g m m ) nt ccniatunt temperatwe
-At constant teniperature the vapour pressure of a mixture depends on the com-
position only. If the intermolecular binding forces between like molecules are greater
than those between unlike ones there is only a weak tendency for the components
to mix. The inking is accompanied by heat consumption, which results in a decrease
of the heat of evaporation required. Thus the volatility increases and the vapour
presswe is higher than what would correspond to a n ideal mixture (Fig. 43, row I,
columns 1 to 3 ) . The deviations from the ideal behaviour can be fornially expressed
in terms of the activity coefficient P so that Raonlt's law takes the form p l = p* x q
;< *; (see also (70)).
Colrirnn 4 presents those ideal mixtures with complete solubility (equal binding
forces het ween like und unlike rnolecdes) which are mixed without heat consumption
iranchembook.ir/edu

4.5 Equation of state and pz-diagrams 79

Type o f m i x t u r e
I 2
insoluble or partially
slight soluble with maximum
solubility

p--~ J+&l ! &

0 100: 0 100 0 -.- 100% 0 1009 0 100

0
I3 'M Ul 1009
0 1007 0 100% 0 1009 0 100%

El 7Zl
fp71fR 0 100% 0 100%
t 100%

Ipj
10 4

0 100% 0 100%
'M M
-
0 100% 0 100%
0 100%

0 100%
benzene -water
water- st ear ic
0
water-
isobutanol
100%
N'r-1:r;,
y0
/

100% 0 100% 0

acetone-
chloroform
100%

acid water- phenol acetone- am monia-water n i t r i c acid-

:arbon disulphide water- iso- methanol water
water amylic alcohol water-form ic
acid

Fig. 43
The properties of binary mixtures as a function of the concentration of light component
Row I: Partial pressures at constant temperature for light and heavy components
Row 11: Total pressure a t constant temperature
Row 111: Equilibriuni curve
Row I V : Temperature a t constant pressure (boiling point curve)
Row V : Temperature of constant pressure (dew point curve)
iranchembook.ir/edu

80 4. Physical fundamentals of the separation process

and thus have their partial pressures proportional to t.he liquid concentrations. The
partial pressures are t,hen calculated by Raoult’s law :

where p*] and p*, are the vapour pressures of the pure components at the given
temperature.
If the binding forces between unlike molecules are stronger than those between
like onea the mixing is an exothermic process and due to increased solubility the heat
of evaporation of one component in the mixture is greater than that of the pure
component. This hinders the formation of vapour and causes a lowering of the vapour
pressure. Row I, column 5 in Fig. 43 shows mixtures with minimum pressure as
examples of this.

Row 11
Tdal p e a s u r e of the naizture at canstant temperature (px-diagram)
The total vapour pressure is simply the sum of the partial pressures. In insoluble
mixtures the total pressure is ihdependent of the composition; in partially soluble
mixtures this holds for the concentration range in which insolubility occurs. Thus in
columns 1 to 3 we always find a maximum vapour pressure lying above that of an
ideal mixture, the total pressure of which yields a straight line (column4). Con-
versely, the example of column 5 displays a minimum pressure lying below the
values for ideal behaviour. The vapour pressure curves which are convex upwards
and concave upwards are termed positive (maximum pressure) and negative (mini-
mum pressure), respectively [73, 171. Adding foriuulae (41) and (42) together gives
the total pressure for an ideal mixture :

For a mathematical interpretation of the mixing behaviour further methods of

calculation are described in Matz [73], Kortiim and Buchholz-Meisenheiiner [74],
Bohjalcovik [75] and, above all, in Schuherth [17].

4.6 Boiling point diagram, 6 - x , equilibrium curve, y* - x

Whiie we have studied the properties of binary mixtures at constant temperature
60 far we shall now examine the behaviour of these mixtures at constant pressure.
The conditions are those found in distillation, which is normally an isobaric process
tending to establish equilihriuiii between the liquid and vapour phases. A boiling point
diagram shows the boiling points and the equilibrium compositions of binary mis-
iranchembook.ir/edu

4.6 Boiling point diagram, 8 - 2, equilibrium curve, y* - 2' 81

tnres at constant pressure (Fig. 43, rows I V and V). In a coordinate system with 6
as the ordinate and xB as the abscissa are plotted the boiling point curve a and the
dew point curve b whose ends coincide. Fig. 44 represents the boiling point diagrani
for benzene and toluene. In conditions represented by points above the dew point
curve b the mixtures are entirely unsaturated vapour, while those on curve b are
saturated vapour. Between the two curves, points represent liquid in equilibriuni
with vapour, while below the boiling curve a only liquid is present. If a tnixture with
the coniposition xB is heated, it begins to boil at the temperature 6, and the vapour
has the coniposition y*; the liquid xB and the vapour y* are in equilibrium at the
t ernperature 8. The boiling point diagram is determined experimentally in the mine
way as the equilibrium curve (see Chap. 4.6.3) and may be similarly used to deter-
mine the nuniber of separating stages required. Fig. 59 shows the equilibrium curve
for benzene and toluene as obtained from the values of the boiling point diagrarn.

120

110

> 90
2

80 P1g. 44
I lli I i I I I Boiling point di,igrani for benzene-toluene
solutions a t 760 torr
0 20 40 60 80 mol%
concentration ---+

Here points A and B coincide. The boiling point diagraui ha.i the advantage that in
distillation processes the concentration of the top product can be read off through the
top temperature. When plate columns are used thc boiling point diagram allows the
eontinuous checking of the bottom concentrations via the colunin temperature gra-
dient. Bottom temperatures may also serve to determine the most suitable feed plate
or draw-off plate for side streams.
Florin [76] gives a method of calculating the boiling point diagram of a n ideal
niulticomponent mixture. Only few points are determined but in addition the slopes
of the curve at these points need to be calculated.
In rows Tc' and V of Fig. 43 we find the boiling point diagram for various kinds
of mixtures. I n coluuins 1, 2 and 3 the boiling temperature is below the bubble point
of the pure components (niininium-boiling mixtures) over a wide range of concen-
trations whereas, on the other hand, colunin 5 displays a range with boiling tempe-
ratures above those of the pure components (maximum-boiling mixtures). The dew
point and boiling point curves coincide a t the azeotropic compositions.
The equilibriuiii curve reveals the relationship between the concentration of the
liquid, xB, and the corresponding vapour concentration, y*, in equilibrium. Thus
the equilibrium curve provides the basis for the calculation of separating stageq by
the McCahe-Thiele method [77] which IS no doubt most coininonly used due to its

tj Krell. Handbook
iranchembook.ir/edu

82 4. Physical fundamentals of the separation process

simplicity. Row I11 of Fig. 43 gives the equilibrium curves for the various mixtures.
For insoluble mixtures the equilibrium curve is E straight line (column 1)intersecting
the diagonal at a point where the vapour and liquid concentrations are equal. It is
impossible then to enrich the vapour in low-boiling component at higher liquid con-
centrations x g ; on the contrary, the vapour will then contain less of the low-boiling
component than the liquid. Hence, in insoluble or partially soluble mixtures the
distrite vaporizes with the same composition over a wide range and only near the
points 0 and loOq, do transition concentrations occur (cohimns 1 and 2). The con-
centration of the vapour is lower than that of the liquid, xB, above and below the
azeotropic point for mixtures with maximum (column 3) and minimum (column 5 )
vapour pressure, respectively. Column 4 shows the curve of an, ideal mixture; y* is
alwajs greater than I : ~ .

4.6.1 Calculation of vapour-l iq uid eq uiIibria

It is a simple matter to obtain the equilibrium curve of an ideal or an almost com-
pletely insoluble mikture by calculation, and for partially soluble mixtures the cal-
ciilat ion is fair1)- straightforward. For non-ideal but completely miscible systems
calculation is less accurate and takes rather longer, and where some degree of preci-
sion is required equilibrium data should be obtained experimentally. However, con-
siderable experience is needed if reliable equilibria are to be measured, and the
procedures are time-consuming, particularly if materials have to be purified.
In the last few ;rears many contributions dealing with the calculation of phase
equilibria have been published. It would lead us too far to mention all of these, parti-
cularly since this subject is exhaustively treated by Schuberth [17] and H d a et al.
[78]. Their books include extensive bibliographies. Of particular interest in this con-
text is a remark made by Spath (cf. chap. 3, ref. [17]) about a simplification in
representing and evaluating vapour-liquid equilibria.
The trend is toward precalculating the data for multicomponent mixtures with
digital computers on the basis of equilibrium measurements on binary mixtures [79].
The reason is that enough such data are available [105, 1061 whereas measurements
on inulticoniponent mixtures are exacting and time-consuming. At the international
symposium “Distillation 1969” at Brighton [6]no less than 7 papers were devoted to
the precalculation of phase equilibria on the basis of the material properties of the
pure components. There were treated equilibrium data for electrolytes and non-
electrolytes, consistency tests for non-electrolytes, and methods of precalculating
activity coefficients. In many cases the Wilson equation [80] has proved to be useful.
According to Donth and Weichmann [Sl] iterative methods can be used to obtain
equilibrium data of binary solutions from p - I: (or T - x)data. The most important
relation8 as formulated by Margules, Redlich-Kister, van Ness, van Laar (Scatchard-
Hauler), Wilson. Heil and Renon-Prausnitz (NRTL = non-random two-liquid) are
discussed by Krng, Haberland and Bittrich [82]. Schuberth [171 considers the Renon-
Prausnitz assuinption to be quite promising. It avoids the disadvantage of a three-
parameter Wilson equation whose application, as a niatter of principle, is restricted
to non-deniixing liquid systems. The NRTL relationship provides a connection be-
iranchembook.ir/edu

4.6 Boiling point diagram, 6 - x , equilibrium curve, g* - z 83

tween Wilson’s model and Scott’s theory. I n an excellent survey article Onkeii and
Gniehling [83a] suggest ways of precalculating vapour-liquid equilibria. I t is no
problem any longer t o include even complex non-idealities in column calciilations. It1
most, cases the resiilts could he more accurate if the data input were more coulplete.
The survey therefore explains the UNIPAC method and mentions the Dortmund
data bank (DDB). The efficiency of the method is illustrated by some exaniples.
Onkeii [83c] reports recent developments in equilibrinni calculations and Bittrich
[83f] gives an excellent account of the proceedings of the section on “Vapour-liquid
equilibria” of the Sixth CHISA Congress, 1975. The state of the field of modelling
phase equilibria of binary and multicomponent s-stems is presented by Haut ha1
[83g] in a survey of improved inethods of approxiination and precalculation.
A purely j’henomenological description of the effect occurring in high-pressure
vapour-liquid equilibria was given by Wichterle [83b] in the first part of a series of
papers.
The work was continued by Wichtwle et al. [ S d ] and resulted in a generalized
method of calculating and predicting vapour-liquid equilibria at high pressures.
The p a c t ical application of molecular therniodpnanrics to the calculation of
phase eqii ilibria is ilhstrated by Prausnitz [83c].
I n f urt8her papers the calculation of high-pressure vapour-liquid equilibria and
the use of the UNIQUAC equation for the calculation of ruulticoinponent mixture
equilibria are reported [83c].
The equilibrium curves of ideal mixtures, which arc exemplified by chemically
related substances arid particularly bg; near homologues, may be calculated bv
various procedures. Some of these will be outlined to explain the basic method.
I n order to obtain the equilibrium composition of the vapour from the vapour
pressure data for the pure components, we must apply Raoult’s law. This law states
that the partial pressure of a component is equal t o the product, of the mole fraction
of this component and its vapour pressure at the existing temperature.
Raonlt’s law:

2’1 = P*, * x1 (zl as a mole fraction)

or
1’*1 * x1
PI = (xl in molo/b).
100
~

For the second component we have, similarly

pz = ~ * ~- xl)
( l (2,as a mole fraction)
100 - x1
Pz = P*. (q in in01 ‘4;).
100
The total pressure pgesis obtained from the partial pressures pl and p 2 by

I)ges ==PI + ~z P,, ’ XI

1 + ~ * 2 ( 1-

(xl in mol 76).

iranchembook.ir/edu

84 4. Physical fundamentals of the separation process

8ince, now, the equilibrium composition of the vapour, y*, is equal t,o the ratio of the
partid pressure of component 1 t o the total pressure pgeswe find that

Thie will be illustrated by an example. We wish to calculate the eqriilibriiiin curve a t

780 m m pressure for the mixture benzene-toluene, by Raoult's law. We first plot
the vapour pressure values on as large a scale as possible and then derive the vapour
pressures of both components at definite teiuperatures (Table 9).

Table 9
Vapour pressures of benzene aud toluene a t various
temperatures

Temp. "c' p , benzene p , tolncne p,,/p,, = i\

80 748 294 2.54

90 1013 408 2.46
i(ni 1335 555 2.41
110 1739 747 2.33

By a rearrangement of formula (48), the composition of the liquid is obtained as

.rl = 1)seJ - p*.2 . 100 111010,. (51)

P*1 - P*,
With the values for Y0"C this formnla gives 11s

760 - 408
x1 = . 100 = 58.18 mole,.
1013 - 408
By int.roducing this value of zl into fortnula (19)we find the corresponding equilibrium
vapour coinposition
1013 x 58.18
Y* = = 77.55 111010". (54)
760
Further figures can be calculated in the same way. The more data that are avail-
able, the hetter can the equilibrium curve he drawn. It is evident that this calculation
can be carried out not only for normal pressure, but also for any pressure in question,
uay for 20 mni Hg. In that case the figure 20 must be introduced for pgea, and the
temperatures have to range between the boiling points of component9 1 and 2 at this
pressure. It is usefril to arrange the data as shown in Table 10.
Mention should also be made of the graphical determination of eqliilibrium curves.
which, however, is far more cumbersonie than the arithmetic method. Details i m , v
he found in Thormann [72] and Rosengart [841.
iranchembook.ir/edu

4.6 Boiling point diagram, 8 - x, equilibrium curve, y* - 5 85

Table 10
Calculation of the equilibrium curve for the normal fatty acids
C,, and C,, a t pgeJ= 20 torr

8 P*1 P*2 21 Y* P*,lP*r = a-

("C) (torr) (torr) (mol%) (molY0)
by by
formula (50) formula (49)

161.5 20 11.0 100 100 1.819

163 21 11.9 83.9 93.4 1.765
164 22 12.5 78.9 86.8 1.760
165 23 13.2 69.4 79.8 1.742
166 24.1 13.9 59.7 71.9 1.733
167 25.2 14.5 51.3 64.7 1.736
168 26.4 15.2 42.8 56.5 1.735
169 '7.5 16.0 34.7 47.15 1.717
170 28.7 16.9 26.3 37.7 1.696
171 30.0 17.4 20.6 30.9 1.723
172 31.4 18.0 14.95 23.45 1.743
173 32.6 18.8 8.70 14.3 1.733
174.j 35.0 20.0 0 0 1.750

For ideal binary mixtures the course of the equilibrium curve can be cotriputed b ~ -
rniploying a coinhination of Raoult's and Dalton's laws :

which reduces to
NX
!J* (55)
1 + x(o( - 1)
and again yields the vapour composition as a molecular fraction. In order to give
the composition in molyo , formula (55)can be modified (with x in moly') to
lOOa
Y* =
LOO/X + rx - 1
Jn these formulae we have employed the fraction 01 = p*1/p*2,which we denote
as the separation fuetor (section 4.2.6) or the relative volatility. In the case of ideal
mixtures LY can be taken as being approximately constant between the tlwo boiling
points (see Table 10). (At constant temperature it is absolutely constant in an ideal
rnisture.) The figures for cy calculated from p*, and p,, for benzene-toluene mix-
tures in Table 9 are seen to lie between 2.54 and 2.33, while the arithmetic average is
2.43. One may also calculate a geometric mean from the two terminal values of DL
(;. e. those corresponding to the boiling points of the components) by
iranchembook.ir/edu

88 4. Physical fundamentals of the separation process

or
1
log Ly, = - (log a1
2
+ log n z )
the figure obtained being almost the same as the arithmetic average, provided the
t,erminal values of OL are not too different. The geonietric mean is preferred. This figure
for a is now introduced into equation (57) or (58), and the vapour concentration y*
calculated for particular values of x. In the case of the benzene-toluene example we
thus obtain the figures shown in Table 11 (colnmn I).

Table 1 1
Calculation of equilibrium concentration y* for benzene-toluene with a = 2.43

I1 Determination
y* determined from by
Kirschbaum
Formula Difference Ideal equi- Difference [1031
(53) or (54) (molyo) librinm (moly")
(mol?/o) from curve from
deter- a = 2.4 deter-
mination mination

10 21.2 -0.4 20.0 +0.8 20.8

20 37.8 -0.6 35.2 & 0.0 37.2
30 51.1 -0.4 51.0 -0.3 50.7
20 61.8 +0.1 61.0 + 0.9 61.9
50 70.8 +0.5 70.8 +0.5 71.3
60 78.7 +0.4 78.4 -0.5 79.1
70 85.0 +0.i 84.8 +0.9 85.7
80 90.7 to.5 91.0 +0.2 91.2
90 96.6 +0.3 95.6 +0.3 95.9

The nomograni of Orlicek and YOU (Fig. 21) may also be used for the determination
of equilibria of ideal mixtures [85]. The left-hand vertical scale gives the molecular
fraction in the liquid phase (zB), the right-hand scale the equilibrium composition of
the vapour phase (y*). Let us again use the system benzene-toluene as example.
For z = 50 mo1oOand ~y = 2.4 a figure of 70.8 mo1°/6 is read off for y* on the right-
hand scale, as shown by the line drawn in the nomogram ; this figure agrees reasonably
well with that found by calculation. The nomogram is valid for molecular fractions.
When it' is used, it should be noted that if 5 or Q lies above nnitg, the unenclosed
figures on scales X and 2 should be employed: if the figure for oc or Q (central scale)
is less than 1, the figures enclosed in circles against scales X and 2 should be used.
The circular diagram after Foorwijk [SS], which is based on formula (53),also enables
equilibrium data to be determined very quickly.
We have so far confined ourselves to the treatment of binary mixtures, partly in
view of the fact that miilticoniponent mixturt~sma)- sometimes be regarded as
iranchembook.ir/edu

4.6 Boiling point diagram. 8 - 2, equilibrium curve, y* -x 8i

composed of a number of binary mixtures. Methods of calculation for mixtures of

three substances, with complete solubility and with a miscibility gap, have been
described in detail by Thormann [72j and Matz [73]. Hollo and Hollo and Lengyel
[ 871 are concerned with problems of vapour-liquid equilibria of ternary and quater-
nary systems. They examined three non-ideal quaternary systems using new appa-
ratus and a new computing method which shortens experimental work considerably.
Quaternary equilibrium data can be obtained by means of a double orthogonal
projection.
To meet the particular requirements of the petroleum industry Umstatter and
Flaschka [88] have developed a method of obtaining equilibrium diagrams froin the
vapour pressure curves of the components which, in turn, may be mixtitres of various
substances. By considering the heats of evaporation and dilatation the accuraq- of
the calculation is greatly increased. The method is reported to be applicable to
azeotropic mixtures as well.
A Fortran programme has been elaborated by Williams and Henley [891 for the
computation of multicomponent vapour-liquid equilibria. To take into account real
behaviours a number of subprogrammes are available which enable fugacities to he
calculated by means of the virial equation, the Redlich-Kwong relation or according
t o Chao-Seader. Activity coefficients may be considered following Wilson, van 1,aar
or Hildebrand. The state of the art of precalculating vapour-liquid equilibria i n
ainlticomponent mixtures was surveyed by H&la [89al. Lu and Polak [89h] dis-
cuwed the special requirements for the calculation of phase equilihria at low tenipera-
tiires (20 K to room temperature).

4.6.2 Volatility, separation factor a and activity coefficient y

By the volatility F we denote the tendency of the liqiiid to pass into the vapoiir
phase. It is expressed as the ratio of the partial pressure to the mole fraction in the
liquid at a given temperature:

The relative volatility of a binary niixture is defined a‘:

PI
-
p, 4 -
Pl x *% . Fi > F2. (60)
F2 1’2
- $32 xx1
x2

For ideal mixtures the volatility is independent of composition and equal to the
vapour pressure of the pure material. A4ccordingto Raoult’s law we have, for the
low-boiling componeiit 1 of an ideal mixture,
PI = P*1 x x1
and for the high-boiling component 2,
iranchembook.ir/edu

88 4. Physical fundamentals of the separation process

That is, the partial pressure of a component is equal to the product of the vapour
pressure of the pure component and the concentration of the same component in the
liquid, expressed as a mole fraction.
According to eq. (39) the equilibrium concentrations in the vapour are
1 1
y,* = fi and y2* =&.
I’nr. Pees

Substituting eqs. (61) and (62) into these latter equations we have

y,* = P*l
- x x1 and y2* - P*2
- -. x $2
(63)
Pew Pees

The quotient of eqs. (59) is found to be

?/I*
-
P*l x J-1 or Y,* X $2 - P*l - 2.
(64)
!I,* P*r XJZ YZ* X X I P*2

This expression for the separation factor iy, however, is identical with (60)BO that
the relative volatility of an ideal mixture niay be expressed much inore easily by the
ratio of the vapour pressures of the pure coniponents, where p,, > p*2.
The relative volatility can he determined from the boiling points of the compo-
nents. By starting froiii Trouton’s rule, with K = 20.5 (= &H/T cal/”C) and the
Clapeyron equation, Rose [91] obtained a relationship that can be used for purposes
of orientation, between the boiling points and the separation factor of two normal
liquids:

Melpolder and Headington [90]developed the following inore accurate formula,

giving errors of 0 to 6.2%, tjhe vast majority of which, however, lie below 1%.
For normal pressure :

loge = -
iE2 - (3.99
T
+ 0.001 193911).
For pressiires above or below at niospheric (10- 1500 nim Hg)

logs =
8 2 - 61
-
T
7-30 - 1.15 log p + 179 log p
(6, - I?,) deg. C = boiling point difference of components;
T K = boiling point of mixture;
p torr = distillation pressure.
Eyiiation (66) is represented graphically in Fig. 45, which covers the range from
-100” to +:3OO‘C for the boiling points of the mixtures.
In order to check the validity of the separation factor Rose and Biles [90]tested
a column, arranged as i n Fig. 88, with finite reflux ratios between 9.2 and 15.8. In
thr case of the mixture n-heptane-niethylcyclohexanethey found a good agreement
iranchembook.ir/edu

4.6 Boiling point diagram, 8 - 5 , equilibrium curve, y* - 5 89

with phase equilibrium measurements for an average value of n = 1.074. The cal-
culation from the experimental figures was perforiiied by means of the forniula

This method is very suitable for calculating the value of n for ideal and non-
ideal mixtures froni test data obtained with various colunin packings. Jt is also appli-
cable to inulticomponent mixtures.
It is only in the case of ideal mixtures that the difference in boiling point of two
components is a consistent nieasure of the ease of separation. If the mixture is not
ideal the dependence of LX on the pressure gives a valuable pointer for the solution of
a prohlem in separation.

Pig. 45
1 HS a function of the boiling point difference J K p . , with the boiling point R p .
of the mixture a t 760 t o n

Tf becomes unity (log (Y = 0) a separation of the two coniponents hy distillation

I
IS riot possible. Only selective methods of separation (section 6.2) can then lead to a
solution of the probleiii. The larger the value of L X ,the higher does the hyperbola of
the ideal equilibrium curve lie and the easier is the separation.
-1s a rule, particularly in homologous series, the value of ,x rises with decreasing
teiiiperatnre, so that the separation of a mixture should becoiiie easier in vaciiiiiii.
This is a h o the conclusion at which Hawkins and Brent [92] arrived after extensivr
distillation tests; the columns were just as efficient under vacuuni as at 760 n i t 1 1
pressure and it is onlv the increase in relative volatility in vacuuni that facilitates the
separation. As an example, according to Tsypkina [93] the number of theoretical
plates required for the separation of ant hracene-carbazole and of pyrene-fluor-
anthrene dropped by more than 50% if the distillation was performed at 60-100 iiini
Hg. There are mixtures for which LX remains constant over a wide range of pressures,
for instance chloroform-carbon tetrachloride and n-heptane-niethylcyclohexane,
and such mixtures are the most suitable for testing colunins (see section 4.10.3). In
iranchembook.ir/edu

90 4. Physical fundamentals of the eepsrrrtion process

some cases, however, oc increases with increesing temperature, an example being

the mixture 2.4-dimethglpentane-2.2.3-trimethylpntane ; such mixtures mn be
separated more easily under pressure. In theorg it would sometimes be separatd
more easily under pressure. In theoF it would sometimes be preferable to employ iso-
thermal distillation, in other words to keep the still temperature constant’and pro-
gressively lower the pressure.
In the case of non-ideal mixtures eq. (64) has to be corrected by int,roducingthe
activity coefficient ;J :

a = - P*1x :’I
P*, x Y2
Since, according to (a),
y* = (Dalton’s law) or p, x y*
= pges we have
Ygrs

For an azeotropic composition this becomes, since y,* = x l r

A plot of VR. x1 gives curves which are an immediate and clear quantitative
measure of the degree of deviation from Raoult’s law (Fig. 46).
For ideal mixtures the activit,p coefficient is unity. The deviations of y from 1are
a memure of non-ideal behaviour. If the vapour pressures of both components are
higher than what would correspond to Raoult’s law (pl = p*, XX,), y1 and y2 will
be greater than 1(log y > 0). This situation is referred to as a positive deviation from
Raoult’s law. If the deviation is sufficiently great the mixtiire is a minimum-boiling
azeotrope (with minimum vapour pressure). Conversely, a great negative deviation
(logy < 0) implies a maximum-bbhg azeotrope (with minimum vapour pressure).
Very great deviations from Raoult’s law result in the breaking up of the mixtiire
into two separate liquid phases and thus in the formation of a heterogeneous azeo-
trope. The latter displays the same behaviour as a homogeneous azeotrope (see
chap. 6.2)[94].
A knowledge of activitj- coefficients and their dependence on the composition of
a mixture is necessary, above all, to identify liquid mixtures as belonging to certain
groups. Further, for example, the reliability of measured equilibrium data may be
checked see p. 100. The determination of activit>+coefficients is also gaining impor-
tance for calculations concerning azeotropic and extractive distillations involving
ternary systems. The methods of calculating the activity coefficients of such systems
were thoroughly dealt with by Kortum and Buchholz-Meisenheimer [74] and Schu-
berth [17].
The value of the activity coefficient y of a mixture is determined experimentally
by isothermal partial pressure measurements (Fig. 37). Substituting the measured
value of p into eq. (70) yields the y values as functions of 1:for a constant tempera-
ture (see chap. 4.4.1). The activity coefficients of systems forming azeotropes should
iranchembook.ir/edu

4.6 Boiling point diagram, 8 - x, equilibrium curve, y* - x 91

be determined according to the method of Carlson and Colburn [95] as illustrated by

Orlicek-Po11[85]. I n a graphic method reported by Orlicek [96] exact values inay br
derived from the course of the curve representing the total pressure.
Measuring data for vapour pressures and phase equilibria can be checked for
therinodynaniic consistency by means of the activity coefficient. The data for x and 11
or x and y* are used to calculate the activity coefficients after eqs. (70) and (71) and
log y is plotted as a function of x. The curves thns obtained are compared with thaw

0.3

\
0.2
1‘

ix
9

Fig. 46
Activity coefficients in the
system methanol-water
w9 1
0 0.2 0.4 0.6 0.8 1.0
X ’ 4

derived from the van Laar [97] or the Margules [98] equation [17, 781. Fig. 46 gives a n
example based on measurements made by Gelbin [1091. If the vapour-liquid equilibria
have associations in the vapour [98a] these have to be taken into consideration s o
that activity coefficients inay be obtained which fulfil the requirement of thernio-
dynamic consistency.

4.6.3 The experimental determination of equilibrium curves

The phase equilibria of binary and multicomponent mixtures are essential for the
determination of the plate number required for a separation and of ot’her distillation
requirements.
iranchembook.ir/edu

92 4. Physical fundamentals of the separation process

The vapour-liquid equilibria published up to 1933 have been collected in the

book of Landolt and Bornstein [99]. Equilibrium data are also to be found in books
bv Perry [ 1001, Yu Chin Chu [ 1011 and m u , Wang, Levy and Paul [ 1021.
Stage and coworkers published a number of tables containing phase equilibrium
data. Figures for the lower hydrocarbons, H,S and CO, are t o be found in the book by
Sage and Lacey [27]. Jacobs [lM] gives the plots of equilibrium curves for 50 mix-
tures. Kogan and Fridman present a n ample amount of tabulated data in their
handbook [105]. The 1966 edition contains the equilibriuni data of 1765 binary
mixtures, 359 ternary mixtures and 32 midticomponent mixtures as well as an int,ro-
duction into the experimental investigation and the checking of phase equilibria.
The whole subject of vapour-liquid equilibria has been discussed extensively by
H&la et. al. [78]. Part I of this work deals with the thermodynamics of solutions of
non-electrolytes, part TI with the laboratory technique for the measurement of phase
equilibria, part I11 contains literature references for 1232 investigated systems.
One of the most extensive compilations of phase equilibrium data a t normal
pressure was published by Hhla et al. [lo61 in 1969. PPDS, which is intended to be
a long-term project, will be able to provide distillation data on a large scale [106a]. I t
is noteworthy that Renon has edited a n international journal entitled “Fluid-Phase
Equilibria” since April 1977, which is published by the Elsevier Publishing Company.
In laboratory practice, nevertheless, we frequently encounter mixtures for which
the phase equilibria have not been measured so that. the question arises what appara-
tus will be suitable to obtain the desired data. We distinguish between isot(herma1
methods and circulating methods with constant pressure. The principle of the latter
method consists in evaporating a binary mixture and allowing phase equilibrium to
become established a t a certain pressure, whereupon the temperature and the com-
positions of the contents of the still pot and of the liquefied vapour are determined.
.4n excellent introduction into the technique of measuring as exeniplified by the
hoiling behaviour of fatty acids is given by Miiier and Stage (Refs. chap. 1 [39]).
Tt is also possible to carry out the determination isothermally, the pressure that is
established a t a constant temperature being measured. Static methods of deterniina-
tion have the advantage of requiring small amounts of substance. Kortiirn arid
Freyer [ 1071 report a straightforward method of determining isothermal vapour
pressure diagrams of binary and ternary mixtures of liquids. A versatile arrangement
for passing the mixture through the thermostat is described by von Weber [108].
Gelbin [lo91 compared various methods of determination.
I t is particularly important to know the course 01 the eqdibriunr curwe accurately
near its beginning and end; for this reason as many points a s possible shoiild be deter-
mined between 0 and 10molyo and between 90 and 100niol%. In general the
equilibriuni vapour concentration y* is measured at the following compositions .zB of
the liquid:
1, 3, 5, 10 and every 5O/, up t o 95, 97, 99 moly/,.

The number of points depends, of course, on the nature of the mixture and the use to
which the data are to be put; in the case of ideal mixtures, intermediate points in the
central region may be omitted.
iranchembook.ir/edu

4.6 Boiling point diagram, B - 2, equilibrium curve, y* - x 5 :3

To determine these values in a fairlysimple apparatus about 250 to 500 nil of the
mixtures are placed in flask a (Fig. 47) and the liquid is heated to boiling. The vaporir
liquefies in condenser b, and the condensate returns to flask a as long as tap c is kept
closed. The temperature established is read off on thermometer d. After equilibrium
has hecoiiie established a sample of about 0.1 to 0.2 in1 is taken at tap c. At the sanie
time a sample of the flask contents is taken through tap e. The deterininations are
continued until three successive concentrations are the saiiie. When ineasureiuents
arp earned out in vaciiuni a sampling device (which is of course also snitable for

Fig. 47
Othmer's phase equilibrium apparatus
A = Sampling point for liquid
B = Sampling point for condensed vapour

normal pressure) may be used to ensure that the sample is not contaminated with tap
grease. The compositions are deterniined by measurement of the refractive index,
with the aid of a calibration curve (nD0"O - per cent wt.), which has previously been

prepared for iiiixtures of the pure components. Another criterion, such as densit\ or
freezing point may be adopted instead of the refractive index if required. The series
of determinations is usually started with the lowest content, xB, of low-hoiling coinpo-
nent. After each measurement enough of the low-boiling component is added to p v ~ "
the next intended concentration; the amount added need not be very exact, since
the position of the point is not critical.
Care is necessary to guard against the following sources of error in equilihririiii
determinations :
1. superheating of the liquid :
2. the occurrence of a concentration gradient in the flask contents:
3. partial condensation above the flask;
iranchembook.ir/edu

94 4. Physical fundamentals of the separation process

4. entrainment of droplets by the vapour;

5. total evaporation of sprayed liquid;
6. the disturbance of equilibrium by the withdrawal of excessively large samples;
7. contamination of the samples during withdrawal.

Attempts have been made to reduce these sources of error by modifications in

apparatausdesign and by standardizing the conditions of operation. The ideal method
would he to determine compositions of the flask contents and condensate without
removing samples. Recently, flow refractometers have been used for this purpose
(see section 8.5).According to Stage et al. [I101 the time required to attain equilibrium
with their circulation-type apparatus amounted td less than 10 minutes, as against
several hours for the usual Othmer apparatus [lll]. It is at all events advisable to
emplo>Tas large a charge as possible, in order that the samples taken shall not affect
the equilibrium.
If one of the components liars a veq- high vapour pressure (as in the system
iuethanol-uiethyl formate), or if the refractive indices differ but little, a satisfactory
method for determining compositions is by gas chromatography. This method will be
particularly useful for measuring the equilibrium concentrations of multicomponent
mixtures and requires very small samples [ 1121. Gillespie [113] has investigated the
possibility of eliminating entrainment and superheating by the use of a CottreU
pump. This principle seems promising, since apparatus so equipped (for instance that
of Rock and Sieg [ 1141, Figs. 48 and 49) gives thermodynamically sound results.
A heating coil b, directly immersed in the liquid in the still pot produces a regular
stream of vapour bubbles. The rising bubbles stir the contents thoroughly and entrain
an appreciable amount of liquid along the Cottrell tube c, so that a mixture of vapour
and liquid is ejected onto the temperature measuring point f , which is provided with
R sprayguard e (Fig.48). Thevapour and liquid phases there separate; the liquid flows
downwards, passing the sampling chamber g, and returns to the flask, whilst the
vapour passes through a settling chamber, is totally liquefied in condenser i, flowrs
through the receiver k and finally also returns to the still pot. In the improved appara-
tus shown in Fig. 19 the vapour space h and the Cottrell pump d are surrounded by a
vacuum jacket for the prevention of partial condensation. The outer wall of this
jacket is maintained at the temperature of ebullition by means of a circulating
thermostat. By this means an error in the equilibrium composition of the vapour
resnlting from partial condensation or the evaporation of sprayed liquid is avoided.
The hydrostatic pressure head due to the Cottrell pump and still pot has no in-
fluence on the temperatures and pressures measured, provided the experiment is
properly performed. The mixture nowhere comes in contact with valves or ground
joints, so that contamination is excluded. The ratio between the volumes of the distil-
late receiver and the still pot is smaller than l :20.
Appamtus equipped with a Cottrell pump has also proved suitable for systeim
u-ith relatively high volatilities (AKp. > 1 0 0 O C ) [115]. The improved Labodest
circulating apparatus makes use of a Cottrell pump and electromagnetic valves with
Teflon cones for sample withdrawal [112]. Its microscale construction is designed
along the xaiiie lines and requires only 100 m l charge (Fig. 50). The sources of error
iranchembook.ir/edu

4.6 Boiling point diagram, 8 - x, equilibrium curve, y* - x 95

Fig. 48
Clrculation-type apparatus of Rock and Sieg (dimensions in rnm ; Internal di<t-
meters given).
u = Boiler, 6 = Heating coil, c = Cottrell tube, d = Cottrell pmnp, e SpraF
L

guard, f = Temperature measuring point, g = Sampling reszel, h = Vaponr

space, i = Condenser, li = Receiver

Fig. 49
Heated part of Rock and Yieg’s apparat.us with
vacuum jacket (designations as in Fig. 48)
2 Vacuum jacket, m Tlieriiiostnt, jacket
iranchembook.ir/edu

-I
Fig. 60
Labodcst circulation-type apparatus with Cottrell pump
Micro-acale version with 1.iboratory recorder (52). device for gas chromatography ($54,Plcctronic intrgrator (53), control panel
(50) and vacuum controller ( 4 8 )
(Manufacturers: Fischer, Labor- iind Verfali~ctistrchnik. Bonn -Bad Godesberg)
iranchembook.ir/edu

4.6 Boiling point diagram, 29 - x, equilibrium curve, y* - z 97

mentioned on p. 93 have been practically eliminated in this type of apparatus. The

smile holds true for the equilibrium apparatus after Schuberth [ 1161 the latest design
of which, named Cr 15, is shown in Fig. 51. The apparatus of Raffenbeul and Hart-
niann [ 116al which is also of the Gillespie type allows quite accurate routine measure-
ments. It is practically free from the disadvantages mentioned above and enables
hot h isothermal and isobaric measurements to be made.
The connection between the still pot and the vapour space is provided by the
Cottrell tube. The tube mounted on top of the still pot is essentially that of the G 13

cket w i t h ground
e
nt 14.5 K bent to rear

;oft ir
:ore
displaced t o front
i
a l l ground joints of reflux ( t o condenser
size NS 14.5, w i t h hooks i in 2 m m capillary tube)

Fig. 51
Constructional sketch of the C: 15 equilibrium apparat,ns (Schuberth)

7 KreU, Handbook
iranchembook.ir/edu

98 4. Physical fundamentals of the separation process

apparatus. The heat required for the evaporation of the liquid is supplied by a heater
whioh consists of a curved glass tube carrying a heating coil. The tube surface is
sintered in order to avoid boiling delays. The tube is placed at a height above the
bottom which allows the operation of a Teflon-coated magnetic stirrer under it
ensuring better mixing of the liquid. Due to the shape of the curved tube the produc-
tion of vapour bubbles is most intense at the centre of the equilibrium space since the
heat f b w density is highest where the tube is most curved. Samples are with-
drawn by ineans of suspended rods which are operated electromagnetically. The
totally condensed vapour or the cooled liquid flows down these rods. The use of the
roda in contrast to the suspended funnels of the older apparatus is advantageous
in that two ground glass joints are saved in each withdrawal unit. Another
charact4eristicof the G 15 apparatus is the joining of the reflux capillary tubes
(3inm) for condensate and liquid outside the central part. The mixing of the two
reflux streams thus brought about prevents too great concentration fluctuations
being set up at the inlet of the equilibrium space, which might disturb the equi-
librium. A11 the other details of the apparatus are identical with those of its
forerunners.
A microscale apparatus for isothermal phase equilibrium determinations in multi-
component systems with sample volumes as small as 2 rid is described by Wichterle
and HBla [ 1171. The equilibrium apparatus developed by Neumann and Walch [ 1181
for low-temperature nieasurements was used to study the system C02-CH,. The
development of new low-temperature phase equilibrium apparatus and work on
binary and ternary systems were reported by Lu and Pobk [89b]. Details of the
improved Labodest circulating apparatus with a view to measurements on mixtures
with a miscibility gap were given by Stage and Fischer [118a]. Schmidt [118b]
suuggesta that equilibrium measurements for rapid determinations should be combined
with a gas-chromatographic vapour space analysis.
Kmarkrishna R m et al. [119] used the equipment shown in Fig. 52 for measure-
mente on the systems benzene-n-hexane and benzene-cyclohexane at pressures up
to 18atm. It is also possible, however, to place an ordinary glassapparatus inside a
pressure vessel. Jost et a]. [120] built a circulating apparatus of glass with an adia-
batic jacket into a 20-litre autoclave.
A glass apparatus described by Slocum [120a] is reported to be applicable to
vapour pressure and vapour-liquid equilibrium mwurements at temperatures from
-50 to 250°C and pressures from 0 to 70 bar. An extensive description of methods
and apparatus for the measurement of isothermal vapour-liquid equilibrium data at
high pressures (4.5.. .110.4 bar) is given by Grans~,and Fredenslund [120b]. Thus,
these authors together with Mollerup [ 1 2 0 ~ measured
1 the vapour-liquid equilibrium
data of the systems C2HJN2, C,H,/N,, C3H,/N2 and C,&/N, isothermally in the
temperature range from 200 K to 290 K. The pressures involved extended to
316.32 atm.
For measuring equilibria down to 1 mm Hg pressure Lydersen arid Hammer [ 1211
developed an apparatus that is rotated, so as to produce evaporation from a con-
stantly renewed, downflowing film and to ensure good mixing of the boiling liquid
and the condensate (Fig. 53). This arrangement was employed, for instance, for
iranchembook.ir/edu

3.6 Boiling point diagram, 6 - x, equilibrium curve, y* -I 99

determining phase equilibria of the systeiii di-n-hutyl phthalate-di-n-bntyl seba-

cate at 1111111 pressure.
A method of equilibrium determination involving the entrainment of saturated
vapour by a current of an inert gas was described by Junghans and von Weher [ 1221.
The method was used in measurements on the binary mixture ethyl benzene-styrene.
Polymerization was avoided by adding an inhibitor. The entrainment method has the
great advantage of proceeding isoharically, the varying partial pressures p , and pz
being made up to the exterior pressure by the partial pressure ps of the carrier gas

Fig. 52
Equilibrium apparatus of Kumarkrishna Rao e t al. for measurements up to
15 a t m
V , = Valve for liquid samples. V , = Valve for vapour samples, V , = Drain
valve, V4- 17, = Valves for admitting air, R = Evaporator, D = Perforated
copper tube, T = Thermo-wells, C = Condenser, P = Manometer, A' = Safety
valve

7"
iranchembook.ir/edu

100 4. Ph.ysical fundamentals of the separation process

(nitrogen), and the total mixture then passing isothermally through the thermostat
(Fig. 54).
The use of radioactive substances for equilibrium determinations may offer
appreciable advantages [123].Thus. e-g. SS has been used [124].

P
e-

Fig. 53
Equilibrium apparatus of Lydersen and Hammer for pressures up to 1 mm Rg
A = Flask (6 cm dia.), B = Heating jacket, C = Connecting tube, D = Air
condenser, E = Water condenser, P = Receiver, G = Capillary connecting tube,
HJH2= Electrically heated aluminium cylinders, K l / K 2= Dewar vessels con-
taining solid CO,, M = McLeod manometer, P = Vacuum pump, N = Valve,
T = Buffer vessel, V = Pressure regulator, RJR, = Sampliag vessels, '3 =
Shaft, 2 = Rotating table

Though very careful work has been done in the past, it will be noted that there is
often a considerable spread in the results found by investigators. A large number of
comparative measurements have therefore been performed recently, with the object
of finding the sources of error and of providing exactly reproducible data.
The equilibrium between liquid and vapour in binary systems with the association
of the component8 in both phases was discussed by Trybula and Bandrowski [12aa].
They reported measurements on the mixtures acetic acid-acrylic acid and formic
acid-acrylic acid aa illustrations of their method (see also Refs. chap. 6 [29a]).
Experimentally determined equilibrium data may be checked for thermo-
dpnamic consistency by methods described by Kortiini et al. [44],Carlson and Col-
iranchembook.ir/edu

4.6 Boiling point diagram, 19- I, equilibrium curve, y” - z 101

burn [95], Redlich and Kister [ 1251 and others [ 1261.Redlich and Kister showed that
the criterion of thermodynamic consistency for a set of isothermal measurements
could be expressed by
1
J’ log (yl/yZ) . d.c1 = 0 , (72a)
0

in which ; ~ ~is/ calculated

y ~ from equilibrium data as
-
log y J y 2 = log Y1 -
* xz - log P*l

Yz . x1 P*z
where y = the activity coefficient;
y = the concentration in the vapour phase ;
n = the concentration in the liquid phase and
p* = the vapour pressure of the pure component.

If, in accordance with equation (72a), log y I / y 2 is plotted against xl, the areas
above and below the x-axis must be equal if the isothermal data are thermodynarni-
cally reliable. As an example, Fig. 55 shows this relationship for the system chloro-
form-ethanol a t 45°C. The difference between the area? amounts to only 1.6%.
Herington [ 1271 extended this method to ternary systems and isobaric data.
Bittrich [ 1281 studied the possibilities of checking the consistency of equilibrium
nieasurements. He assessed the more important methods auggested in the literature.
Herington’s graphic method of “symmetric areas” [6] allows individual data to be

Fig. 54
Equilibrium apparatus employing the
thermornete entpainment method
(Junghans-Weber)
iranchembook.ir/edu

102 4. Physical fundamentals of the separation process

checked in limited concentration ranges of binary and ternary systems. Starting

from the Gibbs-Duhem equation Tao [129] established a relation for checking the
consistency of niulticomponent equilibria. Chapter 6 of Schuberth's book [171 offers
a thorough discussion of the techniques of measuring activity coefficient,s and the
methods of checking binary systems for consistency.

Ridich and Bister's criterion of

thermodynamic consistency
applied to data for the system
-1,o chloroform-ethanol ( t = 45OC)
0,OO 0,25 0,50 0,75 1,OO
xl-

4.7 Number of theoretical plates (separating stages)

It hrts already been pointed out in the historical summary in chapter 2 that,
together with the development of the various t>-pes of column, there arose the need
to compare their separating efficiencies. B;v sepuruting efficiency we imply the ainount
of distillate, having a definite concentration, obtained under predetermined conditions
of distillation in unit tiiiie from a mixture of a certain composition. At first the tests
were carried out on a charge consisting of 50°& w. of benzene and 50% m. of toluene,
under definite distillation conditions, a plot being subsequently made of the distilla-
tion teniperature against the amount of distillate. The course of such ciirves mas
compared, and it was thus possible to show trends in a qualitative manner, for in-
stance that the separating efficiency is dependent on the length of the column and
on the load. More accurate inforniation is obtained if a physical characteristic of the
distillate such as the refractive index is determined and is also plotted against the
amoimt of the overhead; the boiling point alone is frequently not a sufficient criterion
of the purity of a substance. These methods are very suitable for comparing various
columns, but do not allow the column anddistillation conditions required for a certain
separation to be calculated beforehand. For this reason the following criteria for
separating efficiency were proposed :
1. The yield of distillate of a certain purit- [130].
2. The amount of the transition fraction between two concentrations of distillate.
3. The slope of the distillation curve in the transition region [131].
4. The amount of residue [132].
The last two criteria, in particular, are suitable for coniparing separating rfficien-
cies in the batch distillation of binary mixtures. The concept of the pole height was then
iranchembook.ir/edu

4.7 Number of theoreticdl plates (separating stages) 103

developed [131]; this is obtained froin the slope of the distillation curve at the point
where the distillate contains exactly 50 mol% of the low-boiling coiiiponent (Fig. 56).
I t was not, however, until the concepts of the theoretical stage and the transfer
unit had been introduced that it became possible to express in numerical terms the
efficiency of a column and the difficulty involved in a separation. A theoreticctl (or
iclcal) plate (or stage) is defined to be such a section of a column that the vdporir
leaving it upwards (towards the next higher theoretical plate) and the liquid leaving
it downwards (towards the next lower theoretical plate) are in thermodyiainic
equilibriuni (Fig. 24).

Y distiliation curve

dY for y=0.5
tan 0 = -
dW
I

Fig. 56
Definition of pole height in batch
distillation (after Bowman and Cichelli)

To make this matter clear we will consider the development of the plate column.
The simplest apparatus for distillation consists of a still pot for evaporating the liqiiid
together with a still head for condensing the vapour and collecting the distillatc
(Fig. 67a). We have here a single theoreticalstage, since the vapour rising froin the still
pot attains thermodynamic equilibrium with the still pot contents. In order t o
obtain a higher separating effect, Adam (see chapter 2 ) applied the niethod of placing
several stills next to one another, the following still in each case being heated h v
passing into it the vapour froin the preceding one. 4, ccrtain amount of refliiv IS
produced by condensation in the connecting tubes. We can, now, place the conseciitive
stilly above each other and then obtain the well-known plate colrunn (Fig. 5711).
For a further explanation a diagraininatic plate colunin has been shown in Fig. 58:
together with a n equilihriurn curve representing the processes taking placc in the
colnmn graphically. We shall postdate that the operation is performed at total
reflnx, as in the case of equilibrium determinations, that is to say, no distillate is
taken off. In the still pot we have a mixture xl, containing 10 ino19; of the low-boiling
coinponent 1. The vapour y,* formed from the still pot contents at equilihriurn has a
concentration of 25 ruolo/o. If this vapour is condensed we obtain the liquid sp011 tht
second plate. (The still pot is regarded as the first plate.) The vapoiirs rising froin the
still pot evaporate part of the low-hoiling constituents from the second plate liquid,
the vapours yz* leaving this plate having a concentration of 50.4niol0,,. These
vapours condense on plate 3, so that the liquid on this platc again has the wine cotn-
iranchembook.ir/edu

104 4. Physical fundament& of the separation process

Fig. 57
Development of the plate column
a) Simple appwatus with one separating
stage ;
b) Plate column wit,h 3 actual plates
a

Fig. 58
A column with ideal plates and its behaviour as shown by the equilibrium curve
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) 1 05

position x3 = 50.4 molyo. The vapours rising from plate 3 then have the composition
y3 = 76 molyo, and so on. If the vapour y,* rising from plate 1is condensed, we
obtain a distillate with a concentration of 91 inol%.
Hence, in order to enrich the charge having 10 mol yo of t>helow-boiling coiiipo-
nerit to a distillate with 91 moly/, concentration, we must employ a total of 4 theo-
retical stages, constituted by the still pot (one stage) and a column of 3 ideal plates.
l n the following we shall speak of theoretical stages instead of theoretical plates.
a4 the former term best express the progressive enrichment taking place both in
plate columns and packed columns.
The concept of theoretical stage has been adopted for packed columns and several
other separating procedures. I t allows the efficiency of apparatus t o be characterized.
The distance required to achieve a n equilibrium step y* - x is called HETY (Iwight
rqiiivalent to a theoretical plate). Mathematically, we have

where :is the coordinate of length. For apparatus of length 1, the relation will hr
1
HETP = -.
%h

The nuiiiber of theoretical plates is a function of numerous apparatus and operat -

ing parameters. Thus, it has not been possible j-et to precalculate the HETP valw
for a packed column on the basis of mathematical considerations alone. So far, oiilj-
empirical relations have been established resulting from a large number of experinieri-
tal tests (chap. 4.2, 4.8).
l n practice plates do not have an efficiency of lo%, like the ideal plate, hut of
about 50 t o 90y(, only. This is partly because the mixing of vapour and liquid does not
produce complete equilibrium in the time available, and partly because the vapour
entrains droplets of liquid t o the next plate, particularly at high vapour velocities.
The term reflux ratio denotes the ratio between the amount of liquid R flowing
back to the column after condensation, and the quantity E drawn off as distillate:
Reflux ratio v = R/E
If no distillate is taken off, we obtain an infinite ratio:
1' = liin RIE = 00
E+O

We may express the reflux ratio graphically in the following way. The amount of
vapour D removed from the still pot is the ~ i i m
of reflux and distill~t~e:
D=R+E (75)
iranchembook.ir/edu

106 4. Physical fundamentals of the separation process

This applies t'o each component, as is expressed by t,he formula

DXY=RXXB+EXX,~.
By introducing (77) into (78) we obtain

If t.he right-hand member of (77) is divided h,v E we find

and hp introducing r for RJE this becomes

2' *.re x,
y=- (operating line). (79)
? L l + Z
This linear equation relates the composition of the vapour to the composition of
the liqiiid at any level in the colunin (on the assumption of adiabatic operation) and
represents the operating line (I - c (Fig. 59). If the molar heats of vaporization differ
markedly this has to he taken into account when using the McCabe-Thiele method.
Differences in the molar heats of vaporization may appreciably influence the
theoretical plate number, particrilnrly near the minimum reflux ratio or with low
volatilities of the components and high column efficiency. The graphical method
iroiild he greatly complicated by the calculation and plotting of the resulting curved
operating lines. However, a modification bg Fischer [134] makes the McCabe-Thiele
method easy to handle. Billet [135] derived equations for the determination of
operating lines for binary mixtures with different molar heats of vaporization of the
components. ,4 method which allows the inimediate calculation of the number of
theorrt tcal stages for ideal binary mixtures at a finite reflux ratio and considers the
diferent heats of vaporization was evolved by Thum [ 1.361. Xeretnicks et al. 11371
introduced a factor -U for each substance such that all substances will have the same
hmt of vaporization in the new system of units. Thus, a simple transformation
enablers qtraight operating lines to be plotted in McCahe-Thiele diagrams for systems
with greatly differing heats of vaporization. In cases of constant relative volatility
the transformation needs to he carried out at a few points only.
Tf plotted on the same diagram as an equilibrium curve, that is on scales of vapour
c+onqwsition iw.ms liquid composition, eqnation (79) becomes a straight line with
+
gradient r / ( r l ) , and an intercept on the y-axis, when x = 0, of J$(V + 1).
\ t an infinite reflux ratio, or total reflrrx, the gradient is unity and the intercept
Zt'ro :
lini y = 0;
r-lj

t h a t the operating line is then the 45" line, as employed in Fig. 59.
S ~ J
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) 107

We now have all the niaterial available for the graphical calculation of distillation
conditions by the McCabe-Thiele method. This is one of the most used and simplest
niethods for the calculation of batch and continuous distillations of binary inistiires.
It involves the siniplifying assuniptions that the niolar heats of evaporation of the
components and their mixtures are identical, and there are no heat losses from the
coliinin; the consequence ist that the vapour and liquid flow rates, in moles per ilnit
time, are constant throughout any section of the coliinin. provided there is no acldi-
tion or withdrawal of niaterial.

x-
Pig. 59
Equilibriiiin curve for the system benzene-toluene a t 1 atm. showing the con-
struction of the operating line and the method of stepping off plates in t h e 3IcCaI)e-
Thiele method

So far all our corisidrrations have applied to plate colunins. In these, eschaiipt.
occiirs during the passage of the vapour through the layer of liquid on the plate,
where the two attain something falling short of equilibrium. The changes in concen-
tration and temperature are repeated from plate to plate, so that the process actually
takes place in steps. Matters arc different in a packed column. In this case there is a
cotit iniioiis exchange between the liquid filni on the packing and the vapour flowing
past it. The vapoiir nowhere remains so long that it can establish equilibrium with
the liquid. Therefore we have a progressive enrichment caused by many elementary
st.Tiarating stages. For this reason thrre has been introduced a concept known as the
iranchembook.ir/edu

108 4. Physical fundamentala of the mpration process

height, of a transfer unit (IFTU) for packed columns; this will be discussed in sec-
tion 4.8.
In laboratory practice, nevertheless, we can apply to packed columns the simpler
methods of calculating separating stages for plate columns, as long as the relationship
between the two is clear and we realize that such calculations are merely good approxi-
mations wit.hout theoretical justification. An analogy between the two typed of
cdumn can be pict,iired if we imagine a plate column and a packed column of the
same dimensions, in which separations are carried out under identical conditions.
If both columns give the same separating effect it is reasonable to say that they
contain the same niimber of separating stages.

4.7.1 Calculation of separating stages by the McCabe-Thiele method

in batch operation

Let us keep to the mixture benzene-toluene as a practical example. It is our

purpose to enrich a mixture of 40 mol% benzene to 98 mol%.
We construct the equilibrium curve (Fig.59) and through the point xB = 40 molyo
draw a parallel to the ordinate, intersecting the equilibrium curve at the point b.
The line parallel to the ordinate through the point 2 , = 98 molyo intersects the
diagond at u. By joining a and b we obtain an operating line, which when continued
intersects the vertical axis at c (y = 37).
According to eq. (79) we have

98
wmin = -
37
- 1 = 1.65.
This figure, 1.65, is the minimum kflm ratio, at which an infinite number of
phtes would be required for separation. Since in this case points xE, xB and y* lie
on the operating line the minimum reflux ratio may also be expressed as [lo31

where y* refers to point b, at which the opmting line u - c int,ersectsthe equilibrium

curve,
98 -62
Qmi II = -M 1.65.
62 -44
If the selected reflux ratio is less than w = 1.65 a distillate compositionof 98 molyo
cannot, be reached. Increasing the reflux ratio beyond this value leads to finite plate
nilinhers which are the smaller the higher the selected ratio. In our example we choose
a refliix ratio 77 = 1.85. The operating line intersects the ordinate at
98
= 1.85 + 1 = 34.1.
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) 109

The intersection, c’ (Fig. 59), is connected with a, yielding the operating line which
corresponds to the distillation conditions chosen. Now we construct steps between b
and a, moving alternately from the operating line to the equilibrium curve and back.
We find I 0 theoretical plates.
We can show how the number of separating stages varies when the reflux ratio
is increased. Pig. 60 represents the equilibrium curve with a number of different
operating lines and Table 12 gives the number of stages found for each reflux ratio.

90

10 20 30 40 50 60 70 80 90mol%100
x-
Fig. 60
Operating lines for various reflux ratios with xE = 98 moIo4,

It will be seen that in this case there is not much object in increasing the reflux
ratio above, say, 5; the saving in separating stages becomes so small that it is scarcely
worth while.
I n a batch separation the concentration of the still pot contents gradually chang-
es owing t o the removal of low-boiling component with the distillate; the value
of xB shifts towards zero. Let us assume that we have distilled down t<oa concen-
tration xB of 10 molyo and let us again carry out the above calculation. We then
obtain the data shown in Table 13 where they are compared with those found for the
beginning of the distillation.
The number of separating stages has purposely been kept constant in order that
we may find how much the reflux ratio must be increased to produce the same separa-
iranchembook.ir/edu

110 4. Physical fundamentals of the separation process

tion. If we had kept the reflux ratio constant at 1.85, the concentration xE would have
fallen with the change in composition of the boiler cont,ents.
In the case of a batch distillation it is therefore not sufficient to determine the
conditions of distillation at the start; it is necessary to consider what the final com-
position of the contents of the still pot is to he, or what is the highest reflux ratio
that one is prepared to use, considering the total time required. Suppoue that we
adopt z- = 25 as the limiting reflux ratio, that xE = 80mol% and that we keep
x, and the number of theoretical stages constant at 98% and 10 respectively. The data
found for a reduction of the still pot concentration from 80 to 5 mol76 are then as
shown in Tahle 14.

Table 12
Separating stages required for the separation of the mixture
benzene-toluenein batch operation
(ZE = 40 moloh; xs = 98 molyo; cmin= 1.65)

v chosen Interaept on ordinate Number of theoretical stages

Y = xz/(v + 1) %h

1.85 34.4 10.0

3.00 24.5 7.0
5.00 16.4 6.0
10.00 8.9 5.5
20.00 4.7 5.0
50.00 1.9 5.0
.x 0.0 4.5

Table 13

58 2, v (minimum) u (chosen) ‘bth

At start of distillation 40 98 1.66 1.85 to

At the later stage 10 98 7.10 8.80 10

Table 14
Reflur ratios for the mixture benzene-toluene
with Z B = 80 to 5 molyo, nth= 10
and XE = 98 mol%

.L‘B Minimum reflux Reflux ratio

ratio chosen

80 0.58 0.65
60 0.94 1.09
40 1.65 1.85
20 3.46 3.90
10 7.10 8.80
> 13.00 25.00
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) “1

If we plot a graph of the selected reflux ratio against xB (Fig. 6 l ) , it will h~ qtlc11
that when the concentration falls below 10% every reduction in xB entails a very
large increase in the reflux ratio. In practice we might start with a reflux ratio of 2
and increase i t to 5 as the top temperature rises, and later to 10. I t must be decidd
in each case whether a further increase in the reflux ratio is justified, bearing in rnirltl
that the amount of the transition fraction becomes greater the less closely om’ ad-
heres to the optimum conditions.
On tlie basis of the Raleigh equation uon Weber [ 1381 developed a diagraru frottl
which the concentration of the liquid in the still pot after distilling a charge B, down
t o a residue B,can be read off for ideal mixtures if o( is known. A method of calciilat-
ing the stepwise countercurrent distillation of ternary systems was evolved by Vogel-
pohl [139]. A mathematical model was constrncted by Doinenech et al. [ l N a ] . -4

Fig. 61
Dependence of the reflus ratio on the still
pot concentration xswith ZE = 98 nml‘’~,
and 10 separating stages

Irocednre, partly mathematical and partly graphical, has been developed b?- Billet
[ 1401 for batch countercurrent distillation of binary mixtures, with which one w t i
calculate the reflux ratio, the yield, the still pot concentration and the quantit? of
mixed vapour flowing through the colurnn as a function of the time of distillation.
With a view to the high purification of liquids by means of batch distillation Wi1-
cox [140a] derived special equations for distillations which aim a t achieving ver!
high distillate purities.
Just as the concentration of the liquid in the still pot decreases during a hatch
distillation, so also does the proportion of the low-boiling cotnponent in the o p e ) d ; ~ t c /
hdd-up of the colnmn (see section 4.10.5).

4.7.2 Calculation of separating stages by the McCabe-Thiele method

in continuous distillation
Although distillations in a laboratory are for the greater part done batchwise,
cases occur in which a continuous procedure is preferable or even essential (see sect ion
5.2.2). -4s opposed to hatch distillation, in which tlie concentrations of the still liot
iranchembook.ir/edu

112 4. Physical fundamentals of the separation proceea

contents, the distillate and the hold-up are changing dl the time, continuous distilla-
tion is chamcterized by constant conditions throughout. A mixture to be distilled is
preheated to a certain temperature and introduced into the column a--d at some
point along its length; thiR column yields an overhead product E and a bottom
product A (Fig. 62). The apparatus operates in such a way that

Z = E + A

Fig. 62
Destinorm apparatus,
Model IIIv, for continuous
fractional distillation in
vacuum
iranchembook.ir/edu

4.7 Suniber of theoretical plates (separating stages) 113

and hence
X, X = XE X E + XA xA (88)

where Z = the amount of feed in unit time,

E = the amount of top product in unit time,
A = the amount of bottom product in unit time.

concentration

Fig. 63
Equilibrium curve for a binary mixture of the normal fatty acids C, and C, a t
20 mm pressure (calculated), determination of the number of separating stages
in continuous distillat,ion

Let us assume that we wish to separate a mixture of the C, and Ci normal fatty
acids, the mixture containing 81 molyo of the C, component, and that we have to
detertnine the number of separating stages necessary to obtain a top product of
96 molyo and a bottom product of 0.5 molyo concentration. The distillation pressure
is to be 20 mni Hg. The equilibrium curve for this mixture has not pet been measured;
it was calculated by formula (58) with 01 = 2.10 (Fig. 63).
Fir& we draw a n operating line through the points a (xE= 96 n1oIo6) and
h (xz = 81 molyo);this int,ersects the y-axis a t 59.8 (point c). The minimum reflux
ratio is hence, according to formula (82)

vm,n = x E / Y - 1 = 96/59.8 - 1 = 0.6. (89)

I n order that, the number of separating stages needed shall not be too high, a
reflux ratio of 1 is chosen, which gives en intercept on the y-axis equal to
96
y= -= 48.
I t 1

8 Krell, Handbook
iranchembook.ir/edu

114 4. Physical fundamentals of %he eeparation process

The chosen operating line a - c' hence starts from the point 48 on the y-axis; it
intersects the vertical through xz = 81 moloh in the point d. From xA = 0.5 mol?;
a vertical line is drawn, meeting the diagonal at e. We join points e and d ; this
straight line is the operating line for the lower column section. Starting at d, we now
draw in the steps for the separating stages on the upper column section between the
equilibrium curve and the operating line d - a; this gives us slightly less than
4 stages. Going downwards from d, we draw further steps between the equilibrium
curve and the lower operating line d - e, which gives us 13 stages. If the equilibrium
curve has been constructed on a small scale (say 25 x 26 CH), or if the curve is very
flat, it is useful to draw the part between 10 and 0 mol% again on a larger scale, so
as to obtain greater accuracy.
The separation requires a total of 17 stages, of which 13 have t o be in the lower
column section and 4 in the upper. The way in which these 17 theoretical stages may
be realized is described in sections 4.10.9 and 7.3.
The above-mentioned determination of the minimum reflux ratio for a continuous
distillation assumes that the mixture to be separated is fed in at its boiling (bubble)
point. In this case the amount of reflux in the lower column section is increased by
the amount of the feed, i.e.
R' = R + 2. (91)
If the teniperature of the feed is lower, condensat.ion of a quantity of vapour will
take place, this amount being determined by the quantity of heat necessary to b h g
it to its boiling point. This may be expressed by a correction factor e, which modifies
equation (91) to
R' =R + eZ (92)
in which e has the value
Q g p . - QZ
e = l +
&H .
Qgp.here denotes the heat content. of the feed a t its boiling point, Qz that at the
actual feed temperature and AvH the molar heat of vaporisation. Consequently the
abscissa of the point g (Fig. 64)has the value

The straight line g - d then intersects the upper operating line in the point b.
The minimuni reflux ratio thus becomes lower. If the mixture is fed in at? its boiling
point, e becomes 1 and the points b and f have the same abscissa xz [103]. If tthe
feed consists partly of vapour a construction by Cavers [la]allows the McCabe-
Thiele diagram to he used also.
On the basis of the McCabe-Thiele graphical method Pohl [143] has derived a
numerical procedure for calculating the conditions of the lower column section for
mixtures containing a very low concentration of the light component. The eqiiilibriuin
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) I15

ciirve can then be approximated by a straight line. Horvath and Schuberth [I143
developed a method for dealing with high concentrations in the top product (say
99.990/,)and low concentrations of the bottom product. A high degree of accuracy is
obtained by plotting the McCabe-Thiele diagram on logarithmic co-ordinates. The
arithmetic determination of the theoretical plate number and the parameters for the
continuous separation of ideal mixtures will be treated in detail in chap. 4.15.

100.
mol%
90 -

il
(

10
I

20
,

30
,

40
1

50
I
XZpJ
60 70
,
,if , I
80 90mc
b-&----4
concentration
Fig. 64
Influence of the feed temperature in coiitiniious distillation

4.7.3 Determination of separating stages by the McCabe-Thiele method

for equilibrium curves with an inflection point or an azeotropic point

If an inflection point occurs in the equilibrium curve of a mixture, as in the case

of the system methanol-water, the number of separating stages is found as described
in sections 4.7.1 and 4.7.2, but in determining the minimum reflux ratio the ripper
operating line may have to be drawn as a tangent t o the equilibrium curve (Fig. 65).
If the upper operating line a--c, were drawn as described previously from the point
a through the point bp,three points of intersection with the equilibrium curve would
result, at bl, b, and b3. Then if the reflux ratio were increased by a small amount, an
infinite number of separating stages would still be required even to reach a residue
coniposition a t about b,. The true minimum reflux ratio is derived by drawing the
tangent from a to the equilibrium curve and extending it to c, so that a slight increase
in the reflux ratio leaves a clear path for the steps from u to bp,with the operating line
between the equilibrium curve and the 45" line. Because the operating line for niini-
muin reflux may be determined by the distillate composition and the tangent from it

8*
iranchembook.ir/edu

116 4. Physical fundamentals of the separation process

to the equilibrium curve, and not by the feed composition, we then find that, the
minimum reflux ratio is independent of feed composition over a wide range. Thus
Fig.66 shows the dependence of vmln on the feed concentration xz for the system
ethanol-water, with a constant distillate composition of 81.6 moloh (92.0% w t . ) of
ethanol; vmlnremains constant at 1.3 between x, = 7.50/6 and 66.3 ii~ol%.
If the equilibrium curve lies below the diagonal from the origin, and above it
after the intersection, we have a mixture with a maximum boiling point, that is

concentratlon +
Fig. 65
Determination of minimum reflus ratio for an equilibrium
curve with an inflection point

1-
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) 117

to say, the boiling point of the azeotrope lies above those of the pure compo-
nents. An example is the mixture nitric acid-water, for which the data are as follows
(Fig. 29i):
Kp.760nitric acid 86.0 "C
K P . , ~water
~ 100.0"C
KP.,~,,azeotrope with 37.8 molq/, 122.0"C
of nitric acid
-4 graphical method for the determination of the distillation conditions for this
system is given by Flatt [145].

4.7.4 Determination of separating stages for flat equilibrium curves

and for equilibrium curves close t o operating line

If the mixture in question has a very flat equilibrium curve, as may occur when
the boiling points are close together (an example being the system m-cresol-p-cresol,
Pig. 67), it becomes very difficult t o construct the separating Stages accurately in the
regions of low xB and high x8 values. As a, makeshift one can draw the equilibrium
curve on a large scale, say 2 by 2 metres, and enlarge the regions from 0 to 10 and
from 90 to 100 mol% still further, the portions of the curve then being practically
straight lines (compare Fig. 63).
,4 similar situation occurs when a relatively high equilibrium curve has a shape
such that the space between the curve and the upper operating line, where the steps
are to be constructed, is narrow. This kind of curve denotes a close approach to
azeotxope formation as in Fig. 68. Owing to the nearly asymptotic approach of the
curve to the diagonal (and still more so to the operating line) the number of steps to
be drawn is so that accuracy becomes poor.
For an infinite reflux ratio Stage and Schultze 11461 describe a method involving
t h e construction of a differential curve. The difference IJ*- xB is plotted against .c,

Fig. 67
Equilibriuni curve forcthe mixture
m-cresol-p-cresol at 30 torr

concentration ----)
iranchembook.ir/edu

118 4. Physical fundamentals of the separation process

on as large a scale as possible. This gives us a differential curve, as shown in Fig. 69

for the system benzene-n-heptane. From this diagram a curve for separating stages
v e r w concentration is constructed. Starting with a very low concentration of the
still pot contents - say 0.16 mole/, - we read off the enrichment y* - xB from the
differential curve to be 0.28. This value is added to 0.16 and we thus obtain the figure
of 0.44 for the concentration of the liquid in the second stage. Against this value we
again read off the enrichment (0.70),add it to 0.44 and so find the concentration in
tjhe next stage to he 1.14. This procedure is followed up to, say, 99.98 mol% and
gives us the curve for separating stages against concentration (Fig. 70). The number
of stages with infinite reflux ratio is now obtained by drawing lines vertically from the
points xB and xg to the last-mentioned curve, reading off the corresponding numbers of

concentration -+
Fig. 68
Equilibrium curve for the system toluene-n-octane at
760 tom, having a close approach to azeotrope formation

0.3 I0 2C X 40 5C 6p 7C 80 rnnio/,,
- .300
xs concentration of benzene in Ilquld-
Fig. 69
Differential curve according to Stage and Schultzc
for t h e system benzene-n-heptane
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) 119

separating stages and finding the difference between them. Fig. 70 shows that, for
tmriching a charge of 25 molyo benzene to 85 mole/, 11.2-6.6 = 4.6 separating
. t a p are needed with an infinite reflux ratio.
Instead of xB, other characteristic quantities may be plotted as abscissae, for
inqtance the density of therefractive index, the figure for the charge and the distillate
h i n g entered and the difference of the corresponding ordinate values determined.
Pig. 71 gives the curve for separating stages against refractive index for the tmt
1 nixtiire n-heptane-methylcyclohexane [ 1471.
For finite reflux ratios, Schafer [148] evolved a method involving two nomograms
<ind two diagrams.
The method is particularly suited to determining series of plate numbers. It makes
iiw of the approximate hyperbolic relation between plate number n and reflux ratio
7 % as employed by Fischer :

( n - ?tinin) X ( ~ 1 - t’rnin) = (n - n m i n ) X (v - ~rnin). (95)

Fig. 70
Curve for separating stages
against concentration after
Stage and Schultze for the
system benzene-
n-heptitne
concentration of benzene +

14200

%
t 1.4100
F?
14000
Pig. 71
Curve of separating stages against
1.3900 refractive index for t,he system
n-heptane-methylcyclohesane
0 I0 20 30 40 50 60 70 80 90
number of theoretical plates +
iranchembook.ir/edu

120 4. Physiral fundamentale of the separation process

Fischer himself describes a method for which, unlike Schafer’s, one nomogram
suffices [ 1491. Further graphical and semi-graphical methods are found in a compila-
tion by Stage-Juilfs [71].
By the “method of the separating factor” according to Matz [ 1501 plate numbers
can be calculated for ideal mixtures with batch and continuous operation even if the
ze-values are in the range from 99 to 99.W,,.

4.7.5 Methods for determining the plate number

in batch distillation arithmetically

So far we have chiefly considered graphical and semi-graphical methods, as they

are best suited for clarifying the idea of a separating stage. Numerous procedures
have, however, also been developed for determining the conditions during distillation
arithmetically.
The graphical methods have the advantage of being based on the equilibriuiii
curve of the mixture, so that they are applicable to non-ideal systemti. The methods
based purely on calculation as a rule postulate an ideal system.

4.7.5.1 Determination of the plate number from difference in boiling point

In order to gain a rough impression of the relation between differences in boiling

point and the required number of separating stages, the values for binary hydro-
carbon mixbures at normal pressure, given by Hilberath 11511, are quoted in Table 15.
A better indication is obtained if we consider the desired purity of the low-boiling
component as well. From a diagram published by Bragg and Lewis [l62] the mini-
mum number of plates may be read off directly against the difference in boiling point
of equimolar, ideal binary mixtures (Fig. 72).
For given differences in boiling point the number of plates needed can also be
calculated by the following formula, which, however, does not consider the other ron-

Fig. 72
Minimum plates Venus boiling
point difference and distillate pu rity
(Bragg and Lewis)
boilivg point difference
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) 121

ditions of dist,illation such as reflux ratio, pressure and hold-up :

where T , and T , = boihng points of the low- and high-boiling components,

K = a constant. (By putting K = 2.5, 3 and 4 we obtain the maxi-
mum, optimum and minimum number of separating stage,.
respectively. )
nth = number of theoretical stages.

Table 15

A XP.7," Number of separating

"C stages required

1.5 100
:3.0 55
.i.O 30
7.0 20

With equations derived by Strangio and Treybal[152a] the distillation condition.

(v and nth)for multicomponent mixtures may also be estimated if constant) a-values
and gaseous or liquid feed are assumed.

4.7.5.2 Calculati8n of the plate number w i t h the aid of the

fractionating factor and the Rose formulae
The use of forinula (96) is equivalent to the calciilatjionof the number of s t a p h
for ideal mixtures from the relative volatility a and the fractionating factor Pr. The
frcictiomting fuctor is a value found enipirically by Cook as the result of niimeroiis
experimental distillations :

(97)

in which xBl and xB2 are the concentrations of low- and high-boiling componcnti.
respectively, in the liquid in the still pot;
xE, and xE2 are the concentrations in the distillate of low- and high-boiling
component8s,respectively ;
r&th = the number of theoretical plates.

Fr has a value of 10.000 for a sharp separation, in which the low-boiling compo-
nent present in the distillate amounts to 99 mole/, and in the still pot to 1 inolo,.
For a normal separation, in which these figures are 90 moly, and 1 i n o l 0 ~respw-
,
iranchembook.ir/edu

122 4. Physical fundamentals of t.he separation process

tively, Pr works out at 1OOO.

Fr = w
lox
xg
1
= 891 10oO. (99)

From this we conclude that.

4.0
for a simrp separation nth= -separating stages (loo)
log a
3.0
for a w m n a l separation nth = - separating stages. (101)
log a
Obviously the fractionating factor may also be calculated for other condit'ions,
such as a distillate concentration of 95 in01 % and a residue concentration of 1 inol yo:

= 1850

?hth log n = log 1850

3.3
nt,, = -separating stages.
log (Y
The figures resulting from equations (100) and (101) have been listed in Table 16
and plott'ed in Fig. 73.
Rose [154] developed this method further. He assumed t,hat in most separations
a purity of 95 molob map be regarded as sufficient for the first 40o/, of distillate and

Table 16
Number of separating stages from the empirical fractionating factor Fr

1 Difference in boiling Required number of stages nth

points of componentg
deg C Il'ormal separation, I Sharp separation, I1
(formula (101)) (formula (loo))

3.0 30 5 8
2.0 20 9 13
1 .R 10 15 22
1.3 7 25 3.5
1.25 6 30 42
1.20 5 38 50
1.15 4 50 65
1.12 3 70 80
1.05 2 100 150
iranchembook.ir/edu

4.7 Number of theoretical plates (separating stages) 143

that the iiiixturc in question is an ideal systeril containing 50 mole',, of each conlpo-
nent. He further postulated that the hold-up is negligible and that the relationshil)
n = w can he taken as valid for a laboratory distillation. From the average vapour
pressure ratio a he then derived the maximum, optirnnm and minimum number of
stages required (Fig. 74).
3.6
%nax = - (106)
log n
2.85
noat = - (107)
log a
2.3
nmin= -. (108)
log o(
These formulae give the following numerical valws (Table 17), which have heen
plott,ed in Pig. 74.

Fig. 73
Dependence of the value of J on the
difference in boiling point,s of the
components and dependence of the
required plate number on the frac-
tionating factor Fr
n(I) by formula (101)
Fr = 1000
n(I1) by formula (100)

0 50
At
n-
-100 150
Fr = 10000

''180
160
140
$120
; 100
80
5 60
i *20
t
Fig. 74
Separating stages required for the ,standard
distillation of Rose
10 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 . 0
vapour pressure rstio, a -+
iranchembook.ir/edu

124 4. Physical fundamentals of the separation process

Table 17
Number of separating stages acmrding to Rose

Ls log a %in nopt %ax

2.i I 0.3010 8 9 12
1.3 0.1761 13 16 20
1.25 0.0696 24 29 37
1.10 0.0414 36 69 87
1.05 0.0212 108 134 170
1.03 0.0140 203
1.02 0.0086 335

.As regards the reflux ratio, we have previously seen that a low value of n can be
compensated by a n increase in v, and that higher values of n allow lower values of v.
It is, however, important t o remain within certain limits, since excessively low
values of n cannot be offset, by increasing v and very high values of n do not allow a
proportional reduction in v. Rose expressed this fact by adopting the following
standard values for the reflux ratio (see Fig. 75).

3
Cm,, = - n,,,
2

Fig. 75
reflux ratio, v -
Dependence of the required plate number on the reflux
ratio for Rose's standard distillat,ion
iranchembook.ir/edu

4.7 Number of theoretical plates (separatingstages) 1%

4.7.5.3 Calculation of the minimum plate number by the Fenske equation

for ideal mixtures and v =m
The relative volat,ility of an ideal mixture can be expressed by:

(11)

in which yl* and y2* = concentrations of the low- and high-boiling components,
respectively, in the vapour ;
x1 and x2 = concentrations of the low- and high-boiling components,
respectively, in the equilibrium liquid.
If the elementary separating effect of a theoretical stage takes place in each of
)it,, theoretical stages a t total reflux, we find

(113)

By taking logarithms of both sides of this formula, we obtain the Fenske equation
[ 1551 for determining the minimum number of separating stages (see section 5.1.2.2)
in a batch distillation of an ideal mixture:

or

(see chap. 5.1.4),where 6 = cx - 1.

This formula has the advantage with respect to those of Rose that it can he
used with arbitrary values of XB and XE , though only with v = 00. The Fenske
formula has been converted by Melpolder and Headington [YO] into a nomogram, by
ineans of which the data required can be found quickly and easily (Fig. 76). In the
nomogram the following example has been shown :

xB = 15 molyo, ZE = 90 molyo, cx = 1.2; result): nth = 22.

On the upper horizontal scale for XB one first finds the point for 15 niolo,,
whereupon one proceeds vertically downwards to the xE curve for 90 mole,;. Froni
this point a horizontal line leads to the slanting straight line for oc = 1.2. One then
again proceeds vertically downwards to the lower horizontal scale and finds nt,,= 22.
Richter [1561 rewrote the Fenske equation, setting
iranchembook.ir/edu

Fig. 76
nth -
Nomogram for the determination of minimum plate numbers for ideal mixtures and v = CYI (Melpolder and Headington)
iranchembook.ir/edu

4.7 Kumber of theoretical plates (separating stages) 127

which yielded
K
nt1, = - - 1 (116)
log 111
or
K = (nth + 1)x log a. (117)

The formulae mentioned in sections 4.7.5.2 and 4.7.5.3 can all be redaced to t lit.
fnndamcntal type:
G
nth = - (118)
log Lx’

in which the value of c depends on the conditions of distillation. The use of thih
formula is limited to ideal mixtures.
The minimum refluxrutio vminrequired for an ideal mixture may be calculated 1 ) ~
ineans of the following formula:

in which t~,,,,,
= minimum reflux ratio,

xEl and xE2 = mole fraction of the low- and high-boiling components in the distillate,
xBl and xB2 = mole fraction of the low- and high-boiling components in the residue.
-1relevant noinograin after Smoker is included in the book of Kharabanda [2381.

4.7.5.4 Other methods and comparison

Of course, the above presentation of arithmetic methods is not exhaustive. Pol11

[ 1571 pnhlished numerical methods for ideal mixtures and batch as well as continuous
operation a t infinite and finite reflixx ratios which niake possible a rapid and relativel!
simple determination of the plate number. The contributions of Stage and Juilfs [ i11
should also be mentioned in which further accurate and approximate methods arc.
rninniarized. The same applies to the book of Rose et al. 13531. Ziridcrweg [158]
reports a procedure which considers the operating hold-up (see chap. 4.10.5) and thc
niagnitude of the transition fraction in batch distillation.
Finally, the calculating procedures described in section 4.7.5 have been s ~ n m i a r -
ized below and compared with the graphical McCabe-Thiele method for a particular
example of a separation (see the summary).
It can be seen that the truly comparable methods, that is to say I1 and IV, I and
VTI, as well as VJ with v = 00, agree well.
Methods TI and I V merely give a rough indication of the required number of
theoretlcal stages. The maximum value of nth would he chosen, the refliis ratio being
increased with rising top temperature.
iranchembook.ir/edu

128 4. Physical fundamentals of the separation process

Summary of methods of calculation for batch distillation and

comparison with the graphical McCabe-Thiele procedure
(zB = 40 molo/o; z8 = 98 moloh; LX = 2.4; A K p . w 30°C)

Nethod U %h Remarks

I. YcCabe-Thiele 1.85 10.0 ' Final state with

graphical met,hod 3.00 7.0 xE = 98 mol% and
(Table 12) 5.00 6.0 n = 10
10.00 5.5 Initial state V
XB
20.00 5.0
50.00 5.0 80 molyo 0.65
00 4.5, 60 mol% 1.09
40 moly" 1.85
20 molyo 3.90
10 molyo 8.80
5 molyo 25.00
11. Diagram ~ 99%)
5.0 ( z = xB = 60 moloo
Fig. 72 8.0 (ZZ = 99.9%)
111. Formula (96) 9.7 (maximum) Purity not con-
8.0 (optimum) sidered; zB not
6.0 (minimum) considered
IV. Formulae (100) and (101) 10.0 (zR=99mol%) Residue= lmol%
7.5 (zZ=90mol%)
5'. Formula (106) 8.8 9.0 xB = 50 molyo
Formula (107) 7.1 7.1 First 40% of distil-
Formula (108) 6.0 5.8 late > 95 molyo
~~

VI. Formula (114) r

x. 3.7
VII. Method of 26 10.0 Corresponds to
Zuiderweg [158], Table 14 with final
section 4.10.5 condition
ZB = 5 moly"

4.8 Theory of packed columns

4.8.1 Process of separation in a packed column
By considering a plate inacolumn we have been led to theconcept of the theoreti-
cal separating stage. What are the conditions in a packed column, which contains a
regular or irregular aggregate of packing for promoting the exchange between vapour
and liquid? As before, there is a countercurrent flow of the two phases between the
still pot and the condenser, so that, on the one hand, there exists a tendency for
equilibrium to become established by diffusion in a horizontal sense (whereby an
enrichment of low-boiling fractions in the vapour takes place), and on the other hand
the flow in a vertical sense tends to disturb this equilibrium. The different paths
travelled by the vapour and the liquid in a packed column and in a plate column are
iranchembook.ir/edu

4.8 Theory of packed columns 129

shown diagrammatically in Pig. 77. I n chap. 4.2 it was pointed out that randoiu
packing strongly influences the aerodynamic and hydrodpaniic properties of a
column.
I n addition, the separating behaviour of a packed column cannot be completely
described from a physical point of view by only one parameter, such as the HETP or
the HTU value. Sizmann [159] has shown that material transfer in a packed column
is the more intense the easier it is for the components to get from inside one phase to
the liquid-vapour interface and froni there t o the interior of the other phase. Hence,
two resistances to transfer, W , and W,, for material transport in the vapour and
liquid phases, respectively, have always to be considered. The magnitudes of these
resistances depend on the mean path travelled in the respective phase, the frequency
of mixing a t certain points along the flow path, the formation of turbulence and other
factors which have been discussed in chap. 4.2. The distribution of the resistances to
mass transfer for 7 different packings as determined by Sizmann is given in Table 18.
It is seen that the proportion of the resistance in the vapour phase may var- froin
9 to 96% of the total resistance.
By a proper choice of the packing an attempt, is made to provide the optimum
surface for phase contact. Channelling may be restricted by ensuring that the ratio
between the diameters of the column and the packing units is more than 10:1, by
the use of aerodynamically efficient packing units (such as those of saddle shape)
and by repeatedly sectioning the column, so as to recombine the reflux frequently
and distribute it again over the column contents. The influence of channelling on
column efficiency is demonstrated by Big. 78, which refers to measurements carricd

0 1 2 3 4705
a) b) reflux with channels-
Fig. 77 Fig. 78
Exchange of material and heat Effect of channelling on column
a) a packed column efficiency
b) a plate column with 2 P = Deterioration factor for
actual trays (TJ, paths of number of stages, N = Number
vapour and liquid; of theoretical stages, y = Liquid
- - - direction of exchange) composition in head

9 Krell, Handbook
iranchembook.ir/edu

130 4. Physical fundamentals of the separation process

Table 18
Packing data and distribution of resistance to material
exchange according to Sizman [159]
Chlorobenzene-ethylbenzene was used as test mixture in a column of 1 m height and 30 mm
diameter. Lower concentration of mixture, about 20 mol% of chlorobenzene ; separation
factor, 1.11;distillation tempersturn, about 110OC. Turbulent vapour flow

Packing type Size Material Free Load HETP H‘D/ H’D/ Flooded
(mm) column (cmY (cm) II’, (WD + FITF)
volume/ om2 x h)
packing
volume’)

Raschig rings 6 <: 6 glass 0.72 170 13.4 24 0.96

6 x 6 glass 0.72 170 13.8 3 0.75
Raschig rings 4 x 4 glass 0.70 170 10.0 0.75 0.43
4 7 4 glaas 0.70 130 11.4 0.80 0.45
1 4 glass 0.70 280 10.0 0.5 0.33
Berl saddles 4 ’ 4 ceramic 0.54 170 8.5 1.2 0.55
4 A 4 ceramic 0.54 140 7.5 0.5 0.33
4 x 4 ceramic 0.54 125 ti. 1 T.l 0.52
Beads 4 er glass, matt 0.38 170 7.1 0.20 0.17
Wire-gauze 5 < 5 V 4 9 s t e a l 0.90 210 3.2 17 0.95
rings with 5 , 5 V4Asteel 0.90 185 3.3 7 0.87
one turn
Wire-gauze 4 :c 4 V4A steel 0.88 195 9.8 0.1 0.09
rings with
one turn
Wire-gauze 0.66 170 3.3 0.05 0.05
rings with
two turns

1) measured in n tube of 30 mm diameter

out with the mixture n-heptane-methylcyclohexane. I t will be seen that a small

degree of channelling has no important effect in columns with 10 to 20 theoretical
stages, but is of great influence in colunins with 100 stages. In the case of mixtures
with a value of above 1.07 the influence of channelling is a niinor one, but when a
is below 1.07 it causes the separating effect to deteriorate appreciably [160].
In spite of a frequent redistribution of the liquid it inay occur that the packing is
not u.niformly wetted; the film itself may also have different forins. Stuke [161] has
investigated this subject in detail, and has examined the influ.ence of the nature of
the packing, the properties of the liquid, the column load and the distillation pressure
(with its associated temperature) on irregularities of film formation. His conclusions
ran he summarized as follows (see also chap. 1.2).
1 . Channelling on the packing will occur principally at low loads ; a large amount of
liquid is distributed better on the surface of the packing.
iranchembook.ir/edu

1.8 Theory of p c k e d columns 131

2. The higher the density of the liquid and the lower its viscosity, the sooner will
channelling take place by flow under gravity, even if the packing is fully wetted.
3. The lower the surface tension, the better the wetting of a packing unit a t low
loads.
4. The high separating effect to be expected with small packing can be realized only
if the wetting is extremely uniform. A t higher loads, irregularities in the film
have a progressively unfavourable influence.
5. An unambiguous influence of thedistillation pressure has not as yet been establish-
ed. Some investigators have reported a n optimum pressure in countercurrent,
distillation, others have stated that no such influence exists.
Another point of interest is the dynamic behaviour of packed columns. Heise,
Hiller and Wagner [162] investigated the response of a packed column to marked
disturbances of the liquid-vapour ratio in the countercurrent distillation of the
ternary, almost ideal mixture ethyl acetate-benzene-trichloroethylene(see chap. 4.2,
4.15).
What has been said so far, however, by no means mentions all the complicated
processes taking place in a packed column. Krell (see chap. 4.2) had previously
pointed out the effects of surface energies and had sunimarized the factors operative
in countercurrent distillation with packed columns as follows.

Factors in countercurrent distillation

1. Properties of the mixture to be distilled
Average relative volatility
Phase equilibrium
Molecular proportions
Solubility of the liquid components
Density of the vapour and reflux
Viscosity of the vapour and reflux
Surface tension of the vapour and reflux
Specific heat of the components
Heat of evaporation of the components
Wetting properties
2. Chccructeristics of the colunm
Diaineter
Length of the packed portion
Method of packing
3. Characteristics of the packing
Height, diameter and wall thickness
Nature of surface
Solid volume
Material, shape

9*
iranchembook.ir/edu

132 4. Physical fundamentals of the separation process

4. Fa&m m d with the n&ad of operation

Admission and distribution of the reflux
Hold-up or ratio of hold-upfcharge
Operating pressure
Pressure differential
Surface velocity
It is apparent from this summary that countercurrent distillation is such a
complex process that very great difficulties are often involved in determining the
influence of a certain factor on the separation by experiment. Owing to this fact,
contradictory statements are frequently encountered in the literature.

4.8.2 Determination of the transfer unit, TU

A quantitative description of the processes taking place in a packed column is
therefore possible only along semi-empirical, comparative lines. Chilton and Col-
burn [163] introduced the concept of the transfer unit T U for packed columns, in
which they took account of the fact that in a packed column, as opposed to a plate
column, the exchange of material and heat occurs in infinitely small elementary
stages. The factor causing the exchange of heat is the difference in temperature, and
the factor producing transfer of material is the gradient in partial pressure and con-
centration of the substances. The dimensionless magnitude “number of transfer units”
is derived from the concentration gradient represented by the space between the
equilibrium curve and the operating line (731.
Let us consider an enlarged section of the McCabe-Thiele diagram for d e t e r m i n i
the number of separating stages (Fig.79). T h e upper line is the equilibrium curve, the
lower one is the operating line. In a plate c o l u h the liquid present on an arbitrary
plate, having concentration y,, reaches thermodynamic equilibrium with the wend-
ing vapour, concentration yl*, and falls from the plate in countercurrent to a vapour

I II
Fig. 79
Relation between number of theoreticel stages 7bth and transfer units Tu
(-- --;- %h)
I Operating line steeper than equilibrium curve
I1 Operating line parallel to equilibrium curve
iranchembook.ir/edu

4.8 Theory of packed columns 133

of composition, determined from the operating line, of yl. The vapour yl* rises in
countercurrent t o the liquid of composition yz. Between the plat,es, that is, between
y, and y2,no interchange takes place. Matters are different in a packed column. The
rate a t which material is transferred from one phase to the other is proportional to
the extent to which the concentration of vapour, y*, in contact with liquid of con-
centration yl differs from that of the equilibrium vapour, y*, i.e. is proportional
to !J* - y. Then over any arbitrary height of the column the rate of transfer will be
proportional to the mean value of y* - y over that height, (y* - Y ) ~ and
, a trarisfer

Titble 19
Comparison between the theoret,ical number of separating stages i ~ ~and
, , the
transfer unit TU

lx Y1 Y2 TU %h
mole/, moly" by formula by formnla
(correspondingto xB) (corresponding to ZE) (126) (114)

1.08 10 99 119.5 119.5

1.08 50 99 61.0 59.8
2.50 10 99 10.7 10.0
2.50 50 99 6.9 5.0

unit is so defined that the step height is this quantity (y* - y),,,, in place of the step
gl* - y, used in theoretical plate calculations [71]. This mean quantity is shown as
y,,' - y, in Fig. 791. In general (y* - y)nLis equal t o y,* - y, only if the equilibrium
curve IS parallel to the operating line (Fig. 7911); if the value of dy = y* - y in-
creases as the vapour moves up the column, the number of transfer units is smaller
than the nuniber of theoretical plates; if dy decreases the reverse is true (Fig. 791).
J y usually increases as vapour rises up the column below the feed in a continuous
column, and decreases above the feed, and therefore for an ideal mixture a t total
rvfhix, with a distillation range equally divided about tJhe 0.5 mole fraction, the
iiuniher of theoretical plates will equal the number of transfer units; this is approxi-
mately true in the conditions of column testing. With highly non-ideal mixtures, and
systems with high relative volatility a t low reflux ratio the difference between the
numbers of transfer units and theoretical separating stages may be large. Table 19
shows some comparisons where the difference becomes appreciable [ 1641.
On the assumption that the resistance to transfer is associated with the vapour
phase only the number of transfer units is defined by

111

as compared with
X.
iranchembook.ir/edu

134 4. Physical fundamentals of the separation process

Corresponding to the height of a separating stage,

column length
(em) = height equivalent t o a theoretical plate, (HETP), (122)
'It11

we can formulate the effective unit of height as

coluiiin length
(cin) = height of a transfer iinit (HTU). (123)
TU
The height of a transfer unit is a measure of the resistance to mass transfer. It
varies but little with variation in vapoiir velocity [165].
Matz [73], taking account of the fact that the processes of inaterial and heat
exchange producing thermodynainic equilihrium do not proceed infinitely slowly,
hilt in R measurable time at a finite speed, replaced the integral (120)by the expression

-
(aH) ( K / w ) . (124)
l n this expression
( a l l ) = a characteristic nuiither for the interfacial area, with
u = the interfacial area relative to the volume (ni2/m3),
H = the column height ;
(h'lw) = a characteristic number for the velocity, with
K = the velocity of material transfer (m/sec) and
to = the vapour velocity (nilsec).

Starting from this expression, Matz developed equations with which the number
of trctnafer units may easily he calculated for a n infinite reflux ratio:

R T = '
-ln-+- 1 - ?I1 771
In -.
Y2
(125)
-1
'7?/ 1 - y2 ??Z - 1
in which ?ti = the tangent of the angle between a section of the equilibrium curve and
the abscissa of the co-ordinate system, it being assumed that this section may be
regarded with sufficient accuracy as a straight line.
On page 230 of Matz's book [73] detailed examples of the calculation are giveu
both for 7' = 00 and for a finite reflm ratio greater than the minimum value.
In the case of ideal mixtures and an infinite reflux ratio, eq. (120) can be sub-
jected to a closed integration, which gives
2.303 1 - xs
TU = - log %(l - +a.s0:3 log-. (126)
a-1 .L'B(1 - x E ) 1 - 2,
The calculation becomes much more coniplex if the reflux ratio Is finite [71].
A rapid graphical method, related to the McCabe-Thiele procedure, has been
described by Chilton and Colhurn [163]. I t is applicable to all mixtures for which the
equilibrium curve is known. The operating line should, however, not lie too close to
the equilibrium curve in the range considered. The operating line is constructed in
the usual way in the equilibriritii diagraiit, and through the point zBa vertical line is
iranchembook.ir/edu

. of packed
4.8 Theory - columns 135

drawn, intersecting the operating line a t y1 and the equilibrimn curve at yl*. This
has been illustrated diagrammatically in Pig. 79 on a section of the equilibrium
diagram; Fig. 791 deals with the case in which the equilibrium curve is not so steep
as the operating line, Fig. 7911 with the case in which the two lines are parallel.
According to Fig. 79 I a transfer unit gives an enrichment in the vapour correspond-
ing to ybt - yl, the latter being the mean of yb* - yb and yl* - yl. A theoretical
separating stage gives an enrichment corresponding to yl* - yl. The transfer units
may be drawn as steps in the McCahe-Thiele diagram if care is taken that the magni-
tilde of the enrichment, Y b ’ - yl, is equal to the average of y,* - y, and yb* - Y b .
Point yb on the operating line iniist hence he so chosen that

(127)

This involves a certain amount of trial and error, but does not require much time.
If the equilibrium curve and the operating line were both straight (hut not
parallel) lines, the method would be exact if the logarithmic mean of yl* - y1 and
!/b* - Yb were employed. In most problems, however, the inexactitude caused by a
curvature of one or both of the lines and by the use of the arithmetic instead of the
logarithmic mean will be of little significance.
We have chosen the simplest graphical method to illustrate the relation between
theoretical plate number and number of transfer unita Further exact and approximate
methods are mentioned in the papers of Stage-Juilfs [71]. I n a very interesting paper
Arkenbout and Smit [ 1661 present the inathenlatical foundation of nth and TU and
go on to suggest that, according to their results, in several cases HETP is more suited
to I:ohinin length calculations than HTU.

700 I I I I

8ig. 80
Heat content versus concentra-
1
’ tion for the system ethanol-
n
Y
l I I 1 I water at 1.033 ntm
0.2 0.4 0.6 0.8 1.0
t-
iranchembook.ir/edu

136 4. Physics1 fundsmentele of the separation proceea

Two more methods should be noticed here, which are independent of the column
type and hence may be used for plate columns as well a for packed columns and
columns provided with other elements. Mekel’s I1671 method is based on the heat
content-concentration diagram representing the countercurrent process, from which
the changes of the liquid and vapour compositions, the amounts of liquid and vapour
and the amounts of heat to be supplied and removed are seen (Fig. 80).Unfortunate-
ly, only few such diagrams are available at present so that t’he method has been
applied to a small number of mixtures only. Therefore we shdl only give the pertain-
ing references [73,75, 1031. The various ways of determining the theoretical plate
number for the separation of binary mixtures have been unified by Bitter [261]. He
gives equations for the operating lines. These provide the basis also for the graphical
solution by means of the enthalpy-concentration diagram.

4.8.3 intensity of countercurrent exchange: time required

The hold-up is an essential factor in a countercurrent distillation and has an

important influence on the time required for carrying it out, (see chap. 4.10.5). Whilst,
the theoretical plate number and the reflux ratio increase in proportion to l/log a,
the time expended increases as the square of l/log u. It is hence the time consumed
that best characterizes the difficulty of a separation. If we calculate t,he quotient
theoretical plate number x amount of reflux
I=
operating hold-up of the column
for two columns I and I1 from t,he operating data, the times t I and t I 1 required for
effecting the same separation are related as 11:111, as long as the distillation con-
ditions are otherwise identical. The characteristic value I is termed the intensity of
countercurrent exchange [68]:

(129)

Its physical diiiiension is a reciprocal time, which is that of a frequency or - if the

distillation is regarded from the aspect of labour economy - that of an investment
return. I is independent of the scale and can be forniulated for a volume element
of arbitrary size in a column, having an arbitrary height and diameter.
In t.he total time required we must include the period of equilibration tA during
which the lowest-boiling component in the vapour is concent,rating at the column
head and distillate should not be taken off. According to von Weber [1701 this time
t.k can be calculated by the formiila

( 130)

The time for running-in is therefore roughly proportional to the square of the number
of theoretical plates or to the column length [152], [169]. The equilibration time will
be discussed more fully in connection with the separation of stable isotopes (section
iranchembook.ir/edu

4.8 Theory of packed columns 137

5.1.4) as it requires special attention when the number of theoretical separating

stages exceeds 100.
Yon Weber [170], taking as basis Kuhn's model of a countercurrent systeni [ 3 ] -
which consists of two vertical, parallel, plane surfaces - calculated intensity values
for the test mixture n-heptane-methylcyclohexane, starting with the niaxiniunl
number of separating stages (3.54 per cm) and the optimum vapour velocity
(0.1.52.5cm/sec). The figures so obtained are collected in Table 20.

Table 20
Intensity values for the test mixture n-heptane-niethylcyclohexanein Kuhn's
model of a countercurrent system, as a function of the number of separating
stages and the vapour velocity

!I' = rapour velocity (cmlsec) '0.1525 0.305 0.4575 0.610 1.525 7.62
nth number of theoretical
=
stages per cm 3.54 2.83 2.12 1.66 0.70 0.14
Z = intensity valne (103/sec) 260 367 381 375 322 215

Table 21
Distillation conditions of columns packed with gauze discs a t a n optimum load
of 2/3 times the maximum load

Internal R nth H I . 103

diameter Smount Number of Operating Intensity
of reflux theoretical stages hold-up
mm g/sec cm-l g/cm sec-l

17.0 0.139 0.460 0.239 268

25.4 0.333 0.370 0.228 537
30.8 1.530 0.144 0.328 670

It will be seen from Table 20 that thr maximum nuinber of theoretical stages.
3.54 per cni, does not correspond to the optimum intensity, which is 381 with a stage
number of 2.12 per cni and a vapour velocity of 0.4575 cmlsec. Further i t appears
that the intensity falls again after its maximum of 381, but much more slowly than
one would expect from the greatly diminished number of separating stages. If a high
intensity is the object it does not therefore pay t o work with very lowvaponr velocities.
A t high vapour velocities the flow of vapour becomes turbulent, a factor having a
favourable effect, similar to that of an increase in the diffusion coefficient. If the
throughput is raised above the optimum value, the amount of reflux will increase to
such an extent that its return will be obstructed by the ascending vapour; the
operating hold-up H then increases more than twofold and the intensity is consequent-
ly lowered.
If, following von Weber's [170] experiments, we compare the separating effect of
coluinns of various diameters packed with gauze discs, the columns all being loaded
to 213 of their niaximuni throughput, we find the figures given in Table 21.
iranchembook.ir/edu

138 4. Physical fundamentals of the eeparation process

If these figures are converted to those for a constant number of theoretioal stages
and a constant ratio of the operating hold-up H to the charge, the following values
are obtained (Table 22).
The last and largest column operates the most intensively, though all three distil
at about, the same rate per unit cross-sectional area.
In order to demonstrate the order of magnitude of the intensity, three more
figures, determined by von Weber [170], will be quoted (Table 23).
The intensity is consequently a characteristic that is easily calculated and can
he employed for comparing the most diverse types of columns; it is of value because it
is a factor compounded of several parameters.
.A similar conclusion was reached by Kolling [l’il], who compared various forms
of packing w regards efficiency and defined the separating power as the amount of
a certain purity collected in unit time. Fig. 81 shows the relationship between the

Table 22
Distillation conditions for a separation with a constant number of theoretical
stages and a constant ratio of H to the charge

Charge Column Length of Duration of the

diamet,er column distillation
kg mm m h

1.0 17.0 1.0 100

1.18 25.4 1.24 49
4.40 50.8 3.20 40

Table 23
.Uaximum intensities of various columns

Type of column Max. intensity at

213 of the max. load

Column packed with glass helices, diam. 0.3 cm 165;. l W 3

Column packed with Raschig rings, 2.4 x 2.4 mm, m a t t 370 x
Bruun plate column 140 1:

Table 24
Distillation conditions for a distillate with 90 mol% of n-heptanr

Reflux ratio Number of theoretical Column load Distillate

stages per metre take-off rate
2‘ nth rni/cm2 . 11 ml/h

25: 1 61 50 14
50: 1 36 230 30
iranchembook.ir/edu

4.8 Theory of packed columns 139

degree of purity, the nutuber of theoretical stages and the reflux ratio for a mixture
of 50 molyo of n-heptane and 50 mol% of methylcyclohexane. Thus to obtain a
distillate containing, say, 90 molyo of n-heptane, either of the following conditions
may be chosen (Table 24).
One can therefore take off double the ainount of distillate of the same purity by
increasing the load by a factor of 4.6 and doubling the refhix ratio.
Collecting the distillate at a constant rate, thp following value.: (as shown in
Fig. 8 2 ) arc fonnd (Table 25).

0 20 40 3
r e f l u x ratio, v -----)

Fig. 81
Dependence of the p i ~ r i t yof the distillate on the number of tlieoretical stages
and the reflux ratio

0 50 100 150 d;/h 200

distillation rate ---*
Fig. 82
Dependence of t h e degree of purity of the distillate (with various reflux ratios t i )
on the rate of distillate removal
zB = 50 molyo n-heptane, 50 mol% methylcyclohexane; figures against the
curves are the load in ml/cm2 . h
iranchembook.ir/edu

140 4. Physical fundamentals of the separation process

Table 25
Distillation condition8 for a constant rate of distillate take-off (40 ml/h)
Reflux ratio Column load Concentration Distillate
of distillate take-off rate
U ml/cma h - mol% ml/h

1O:l 70 71.5 40
15:l 100 78.5 40
25:1 160 84.8 40
5O:i 300 88.0 40

143 7
, m,
18.7 $6
j 2 0 j 5
._
5 2 & 4
dI 3 J3 53
8
Fig. 83
SO p2
Dependence of the HETP of
100 8
25 mm Raschig rings and saddle
20 packing on the vapour velocity
C.2 0.4 0.6 0.& 1
-
1.2 1.4 1.6 1.8m/s
vapour velocity W,

The view, often held in the past, that the vapour velocity (and consequently
the load on the column) should be kept as low as possible in order to obtain the
greatest separating effect must therefore be revised, as already shown by the inten-
sity figures. The higher the column load, the better are the results obtained at a
particular rate of distillate take-off; in this connection it may be remembered that
large quantities of liquid are, relatively speaking, distributed better over the packing,
SO that the wetting becomes more complete. This statement is, of course, only true up
to a certain limit, above which the separating efficiency again drops. The relationship
is clearly seen if the H E T P of a packed column is plotted as a function of the vapour
velocity C10.31. In the case of 26 nim (1 inch) Raschig rings, for instance, the same
H E T P is found with loads of 0.2 and 1.0 m/sec (see Fig. 83),i.e. under conditions
corresponding to a throughput ratio of 1 :5. 25 mm saddle packing gives values of the
WETP of 19cm at 0.08 and 1.6mlsec. -s\ can be seen from the curves it is not
advisable to work with vapour velocities hetween these figures, as the separating
effect will then be lower.

4.9 Determination of the number of plates and transfer units

in the batch and continuous separation
of multicomponent mixtures
The term niultieoniponent mixture will be used to denote liquid mixtures contain-
ing more than two components. We shall further make a distinction between rnix-
tiires coniprising a few definite constituents and complex mixtures containing a very
iranchembook.ir/edu

4.9 Determination of the number of plates and transfer units of mixtures 141

large number of components, whose boiling points may lie very close together.
Examples of the latter type are crude mineral oils and gasolines; in general these
mixtures are characterized only by their distillation curves. This is determined,
according to circumstances, either with an Engler apparatus (see section 7 . 2 )or with
an efficient column capable of yielding narrow fractions.
If the mixture is a ternary one, the conditions of distillation may be deterinined
with the aid of a graph on triangular coordinates [72-74, 781. The minimum reflus
ratio in the continuous distillation of ternary and quaternary ideal and axeotroprc
iiiixt,ures can be calculated with a method evolved by Kohrt [172].
Hausen succeeded in solving the differential equation for ideal ternary mixt ures
[173] and van Wijk and Thijssen [174] developed a graphical procedure based on the
approximation of considering niulticoinponent mixtures as binary systems. This
procedure had already been suggested by Lvov [175], in whose fundamental study
the process of the countercurrent distillation of an arbitrary multicomponent mix-
tiire is regarded as consisting of the sum of nuineroiis separations of binary nustures
proceeding in parallel. The problem then reduces t o that of finding the binar) mixture
containing the “key components’) which are the most difficult to separate (i.e.
having the lowest value of a)and of basing the calculation on the equilibriuni curve
for this mixture.
In the usual practice of laboratory distillation the “key component method” is
the most suitable, particularly a s the approximations involved may be checked by
relatively simple trial distillations.
Mostafa [175a] divides a multicomponent mixture into equivalent binary niix-
tures of different volatilities and determines the plate number by using the McCahe-
Thiele diagram. For each group of several components a key component is selected
the &-valueof which corresponds to the average of the component a’s. Wagner and
Blafi [ 175dl have reviewed approximate calculations pertaining to inulticoniponent
separations. Their paper includes also methods of determining relative volatilities,
key components and the distribution of the components in the product flows.
A method for the exact calculation of the minimum plate number in a multi-
component distillation is reported by Chien [ 175b]. An approximate method was
developed by Serov e t al. [175a].
Continuous operation is simpler to calculate because the conditions of distillation
remain constant. Furthermore it is possible to obtain single components or several
components together as top and bottom products. If we wish to produce IIL fractions,
m - 1 columns are necessary. Let us take as a n example a mixture of the lower nornial
fatty acids C,-C,, the distillation curve for which [176] is shown in Fig. 83. To
separate it continuously into individual components we need four columns. Experience
proves that it is best to collect the fractions as far as possible as distillates if clear and
colourless products are required. We shall then perform the operation according to the
diagram given in Fig. 85. It is, however, also possible to work according to a, scheme
Such as that shown in Fig. 86.
Let us first consider column I V of Fig. 85. A binary mixture of C, and C, IS
separated here, and the calculation of the conditions of distillation is therefore
siniple. I n column I11the feed is a ternary mixture; the distillate is a single compound
iranchembook.ir/edu

142 4. Physical fundamentals of the aepuation process

and the bottom product is a mixture of two oornponents. I n this case we compare the
equilibrium curves of the mixtures C&, and C7/C,, and for the calculat,ion use the
equilibrium curve corresponding to the lower U-value of the two curves. The feed to
column II is a mixture of four components. We again determine u for the individual
binary mixtures, c5/c6, Ce/c7, C,/C8, and bme the calculation on the equilibrium
curve for the lowest value of cy. We must, however, consider the fact that the com-
pounds C, and CBare almost completely absent from the vapour ; they should rather
be looked upon as a diluent. It is therefore necessary to correct the value of OL,taking
into account the expected bottom temperatures, the volatility of the bottom compo-

.5 90 1.4100 x -580
80 d -560
=
t
=" 80 60 .E -540
--
-6 70
40 -520
20 - 500
n 1.4000 0 -LBO a
60 80k-4603
60 91 -&LO 9
50 40 -420.G
20 -400 0
40 1.3900 -380 '
- 360
30 - 340
-

nents and their respective amounts. If the bottom temperature is relatively- high, so
that a n appreciable proportion of each component is vaporized, the lowest a of thr
three mixtures should be chosen. If this temperature is lower, i t is sufficient t o choose
t8helowest U-value of the first two mixtures. If we take account of the hold-up and
of the fact that the C, acid is present in a relatively small amount (Fig. 84) it may
even be necessary t o base the calculation on the cS/c, mixture and to regard the re-
mainder aa diluent. Matters are similar in column 1. The bottom temperature will be
high with respect to the top temperature and we can hence choose the lowest o( of
t h e mixtures c,/c, and C5/c6.
In horndogous .geries, as occur in the present instance, the values of ,X diniinish
with rising C-number, and the separation consequently becomes more difficult as the
molecular weight increaaes (Table 26).
We nia-, however have to deal with a case in which the top and bottom products
to be taken off are mixtures of various components. An example is seen in column I
of Fig. 86. The top product is here CJC, and the bottoms are C,/C,; the cut, inust
iranchembook.ir/edu

1.9 Determination of the number of plates and transfer units of mixtures 143

therefore be made betweenC, and C,. We ma? follow the procedure of first using the
equilibrium curves for the two extreme pairs of components, i.e. C,/C5 and C,/C,, t o
calculate the number of stages and distillation conditions required to enrich the mis-
ture to 95 mol% (= zE).The average of the two figures so obtained is taken.
dlternatively we iiiay base the calculation on the equilibrium curve for C,/C7 since
C&, will preponderate in amount and C, scarcely takes part in the exchange bet ween
vapour and liquid.

c4 c5 c6 c7

Z4-C8- -L -m

T T
Fig. 85 Fig. 86
Scheme for the continuous distilla- Scheme for the continuous distilla-
tion of the C,-C, n-fatty acids tion of t h e C,-C, n-fatty acids
C, to C , collected a s distillates, C, as C,, C, and C, collected as distillates.
bottom product C, and C, as bottom products

Table 26
Average values of 01 for mixtures of the
normal C, to C, fatty acids a t 20 mm pressure

Nixture (x

Studies of heat economy made b>r Faldix and Stage [176a] have resulted in opti-
nium coliiinn connexions in two-colunin systems. Flow diagram variants for the
multicomponent separation in columns with intermediate take-off were developed
bv Muller [176b].
Billet [176c] shows that deviations from optinium column lap-out may lead to a
drastic increase in cost. The connexion of main and side columns and the use of
partial condensation in direct and countercurrent operation, where appropriate are
also discussed. These principles should be considered in labcratory and seinitechnical
work. Compara,tive studies of the applicability of heuristic rules and approximate
methods in the selection of appropriate coluniii connexions for niulticomponent
coiintercurrent distillation were made by Jobst et al. [ 176dl.
For ideal multicomponent mixtures there have been developed niinierous methods
of calculation, which have been reviewed by Thormann [177] and Ellis and Fresli-
water [178]. Special attention may he drawn to the method of approxi~~lation evolved
iranchembook.ir/edu

144 4. Physical fundamentals of the separation process

by Colburn [179] and the exact procedure of Underwood [180] for determining the
niinimum reflux ratio. A simple approximate method for finding the minimum
nnmber of stages with v = 00 is due to Fenske [ M I , who also adopts the simpli-
fication of considering a multicomponent mixture as a binary system. The method
involves the restrictions that, in the series of boiling points in the mixture, the succes-
sive key components must predominate, and that the differences in the boiling points
of the successive components shah be not too widely different. If xN denotes the con-
tent of the lowest-boiling key component, the amount of which in the bottom product
is t(Jbe limited, and xo the content of the higher-boiling key component, the amount of
which is ta be limited in the top product,, and if the compositions of overhead and
bottoms are denoted by E and W ,the Underwood-Fenske formula for the minimum
number of stages is
nmin = log ((zwelzos) (~owlznrw)I
(131)
1% &NO
in which uNOis the vapour pressure'ratio of the two key components.
An exact method for calculating the minimum number of separating stages has
been developed by Harbert [182], who refers the volatilities of the individual com-
ponents to the vaporir pressure of the most volatile constituent. An excellent article
by Bruijn [183] deals with the theory of multicomponent distillation at a minimum
reflux ratio. The t hermodgnainic foundation of multicomponent rectifications has
been treated systematically by Wagner in a series of publications [183a]. Starting
from the physical properties of mixtures the various steps in dist,illationpractice are
described in an exemplary manner. The description is supplemented by a great
number of references.
The exact calculation of column and operating parameters for the separation of
multicomponent mixtures requires electronic computers. Problems of modelling and
optimization can then be solved comparatively quickly. Applications of digital and
analog computers in the calculation of separations will be discussed in chap. 4.15.

4.1 0 Testing plate columns and packed columns

So far we have considered only the methods available for determining the number
of separating stages or transfer units necessary to effect a given separation. The
procedure of testing, or calibrating, a column with a known mixture has the object of
establishing the number of stages or transfer units that the column provides. Such
tests are essential because there is no satisfmtory method for findmg the efficiency
of a laboratory colrimn by calculation. A test mixture of known composition is put
into the still and brought to the boil. In general the test is carried out a t an infinite
reflux ratio, that is to say without withdrawing distillate, aa the conditions are then
easy to reproduce. Samples, of such small size that they do not affect the stationary
condition, are taken at the column head and from the still pot during the distillation.
By analyzing these samples one can find the enrichment obtained. The theory and
practice of column testing were discussed in detail by Underwood [I841some tinir ago.
iranchembook.ir/edu

4.10 Testing plate columns and packed columns 145

The number of theoretical stages present in a plate column is in general lower

than the number of actual plates that it contains. An actual plate thus does not func-
tion ideally, afact which weexpress byreporting the ratio of the number of theoretical
stages to the number of plates present (Kirschbaum’s average stage efficiency) [ 1031:
nth.
s=
, -, (132)
n
in which s, = the average stage efficiency, nth= the number of theoretical plates
and n = the number of actual plates.
I n the case of a packed column, the height of a theoretical stage or plate is used:
column length
la= cm . (133)
%h

Investigations of Hiibner and Schliinder [103a] show that the Murphree efficiency
can be calculated sufficiently exactly with the turbulent diffusion model if the
material transfer and the intensity of turbulent mixing in the vapour and liquid
phases are known. The efficiencies in the coiintercurrent distillation of ternary
mixtures are discussed by Vogelyohl and Ceretto [103b].
If the height of a separating staqe is identical with the height of an actual plate
the plate operates ideallv. I n a packed column the height of a separating stage
corresponds t o the HETP.
The number of separating stages found with a test mixture depends on its equili-
hrium curve and the conditions of distillation maintained, and it is obviously also
dependent on the dimensions of the column. If the manufacturer of a column merely
quotes a number of theoretical stages, without further information, the statenlent is
of little value. Only if the data in Table 27 are available can the separating efficiency
of a column be fully assessed and the reported figure be reproduced.

4.1 0.1 Column diameter

According to experiments performed by Murch [185] the height of a separating
stage increases with the column diameter if the size of the packing units remains un-
changed. This influence is, however, less pronounced for the wire mesh type of
packing (such as that of Stedinan and McMahon) than for Raschig rings or saddle
packing, for instance. It is shown by Mullin’s systematic experiments on wall flow
[186] that i t is advisable to keep t,he ratio between the diameter of the column and
the size of the packing units within the range of 1O:l to 12:l. Otherwise, as many
points of flow redistribution as possible should be provided for (see chap. 4.2, 4.8.1).
Groenhof and Steinerding [186a] examined the influence of column diameter on
material transfer, remixing and liquid distribution. They employed columns of 14,
29 and 44 cm diameter packed with Raschig rings of 8.4 intn diameter to a height of
2.07 m. The experimental conditions were the same for the three columns. The
most important conclusion is that the distribution is not influenced by the variation
of the column diameter provided that the liquid introduced into the column does not
change in distribution uniformity with increasing column diameter.
10 Krell, Handbook
iranchembook.ir/edu

146 4. Physical fundamentals of the separation process

Table 27
Data required for judging the number of separating stages

Section (a)Packed columns (b)Plate columns (0) Columns with

various elements

1. 4.10.1 Column diameter 1 and 3-7 as for 1 and 3-7 as for packed
packed columns columns
2. 4.1U.2 Length of packing Active column length Active column length
3. 4.10.3 Test mixture
(composition, mol yo)
4. 4.10.4 Reflux ratio and
amount of reflux
referred to the free
cross-section
(ml/cma h).
5. 4.1U.5 Operating hold-up
6. 4.10.0 Operating pressure
7. 4.10.7 Load (ml/h) or
vapour velocity
(mb4
8. 7.8 Shape, material Plste distance Description of operation
(nature of surface), with essential data (e.g.
dimensions (height, speed of rotation of
diameter, wall thick- revolving parts)
ness), active surface
area of packing
9. Dimensions of an
actual plate

4.10.2 Effective separating length; the introduction of column packing

By effective separation length we denote the effective height of any type of

column. I n packed columns the effective length is that of the packing. Owing to
channelling and wall flow, it is desirable to subdivide this length into sections, since
the countercurrent exchange in the lower part of the column will otherwise be defi-
cient (Refs. chap. 1 : [39]). Thus it, has been shown by Kirschbaum [187] that in in-
dustrial columns the number of theoretical stages does not increase in proportion to
the length. For laboratory columns, Kazanskii (1881 demonstrated that a column of
149 cm length, not subdivided, corresponded in certain conditions to 18 theoretical
plates, and t>hatthis number increased to 24 by subdivision into three sections. More
recent work by Bushmakin and Lydova [1891 has confiied these results. As column
packing they employed constantan helices of 1.8mm diameter. They found that
subdivision at every 25om of the length produced an optimuni efficiency of thi.
column. On increasing the number of subdivisions from zero to the optimum ever)
subdivision resulted in an increase of efficiency corresponding to 1.5 theoretical
iranchembook.ir/edu

4.10 Testing plate coluinns and packed columns 147

plates. The question of the sectionipg of columns for the better distribution of rcflns
has also been investigated by the writer, who used 4 x 4 nim saddle packing and an
infinite reflux ratio.
The data are suniniarized in Table 28.

Table 28
Influence of the effective length s n d of subdivisions on the H E T P

Effective Effective length 90 em

length 45 cm
1 section 2 sections

HETP (cm) a t a load of 200 ml/h 4.73 6.93 6.93

HETP (cm) a t a load of 600 ml/h 6.00 11.25 7.83

Test mixture: benzene-carbon tetrachloride; charge 300 ml ; 40 molyo carbon

tetrachloride
Pressure: atmospheric

is Fig. 87
Device for the random int,roduction of small
packing units (Allenby and L’Heureux)

These results demonstrate the following facts, which so far have been given little
attention in laboratory work.
1. The HETP, a t a constant load, is dependent on the effective column length.
2. Subdivision of the column has no influence at a low load, but, improves the effi-
ciency a t a high load.
The total surface of the packing in a lengt>hof cohimn depends on the way I’)L
which the packing is introduced; this also affects the thickness of the liquid film [190],
the pressure drop and the separating efficiency [191]. In order to obtain a random
distribution of the packing Myles et al. [192] fined the column with mineral oil and
dropped in the packing units one by one. It is simpler to fill the column by putting in
three or four packing units at a time and tapping the column repeatedly with a length
of wood. By the use of a device described by Allenby and 1’Heureux [193] (Fig. 87),
small packing elements may be rapidly introduced singly into the colunin, so that a
random arrangement of the packing is assured. The packing is put into the flask in a
layer about 1 cm thick. Air is hlown in through the tube a ; the packing is whirled

1 o*
iranchembook.ir/edu

Table 29
Properties of test mixtures
a,, @B = boiling point8 of substtlnces A and B a t 760 torr (T),
At9 = OH - B A = boiling point, difference

I. Normal Binary mixt8ure Boiling point .n' " A a For test of

pressure nt I1
@,, ("C) 8" ("C) A @ ("C) A R >I:;

760 torr benzene-toluene 80.1 110.6 30.6 1.50122 1.49647 0.0048') 2.36/2.61 2-7
benzene-n -heDtane 80.1 98.4 18.3 1.501 1 1.3876 0.1185 2.12/1.11 5 - 20
methylcyclohexanc- 100.98 110.6 0.62 1.4232 1.4968 0.073fi .tOb/l.:+28 5 --25
t,oluene
benzene- no.1 83.5 3.4 1.501 1 1.4448 0.0563 .ltL2/1.107 15 -20
1.2-dichloroel hanr
n-hoptnne- 9u.4 fOO.Hii 2.4:; 1.3R 7S 1.4232 0.0:153 .076/ t 474 20 - 8 0
methy lcyclohoxnne
p-xylene-in-xylenc 139.10 138.35 0.75 1.4960 1.4973 0.00132) 1.O2O3/1.0204 A 0 - -250

11. Vacuum a a a ng A Application For teRt of

5 torr 10 torr 100 torr "th
A B ?LE
to t torr 2-methylnaphthalene- 1.1400/ 1.131 2/ 1.1035/ ') d n g = 0.0060 max. pressure 15-45
1-methylnaphthalene 1.1417 1.1328 1.1047 100 torr
to 5 t,orr trans-decalin- 1.483/ 1.440/ 1.314/ 1.46932 1.48098 0.011661) min. pressure 4-20
cis-decalin 1.460 1.420 1.300 5 torr

a a a
60torr 300torr 760torr

to 20 torr benzene- 1.188/ 1.140/ 1.130/ 1.5246 1.49.58 0.0288 min. pressure 10-30
ethylbenzene 1.194 1.143 1.130 20-30 torr
iranchembook.ir/edu

111. N n fi Bpplic~tioii For test of

Pressure 8 ictiii. 12 atin. 16 :\tin. nth

t o 8 atm. isobutane-n-butnne 1.308/ l.%il/ 1.220/ inin. pressure ti -25

1.241 1.217 1.196 8 atm.

L+ N a a
1 atm 3 atin. 5 atin. 19 atm.

t o 1 atin. methanol-ethanol 1.6441 1.4741 1.3951 1.289/ ’) 4-20

1.866 1.559 1.447 1.319

l) ilnalysis preferably by gas chroinntography

2, Analysis preferably by infrared spectroscopy or gas chromatography
iranchembook.ir/edu

150 4. Physical fundamentals of the separation process

round in the flask. The projection b acts as a deflector; the hole in the cork c allows
only one packing unit to pass at a time through the connecting tube into the column.
In plate columnfi and columns provided with other elements the separating length
is equal to the effective column height, i.e. to the distance between the lower edge
of the lowest plate and the lower edge of the topmost plate plus 1 plate distanoe in
the case of plate columns. For columns carrying certain elements the separating
length is defined as the total height of the effective element ; in the case of rotating
elements this will be the height of the rotor.

4.10.3 Test mixtures and the composition of the charge

Strictly speaking, every colunin should be tested with the mixture that is to be
separated in it. However, for purposes of comparison, and because the pure compo-
nents of the mixture in question may not be available, it is usual to employ certain
I%& midurea having particularly suitable properties. They should primarily be SO
chosen that the separability of the components is suited to the expected efficiency
of the oolumn. A test mixture with a low value of u may possibly be used both for
oolumns with a large and a small number of separating stages; mixtures with a high
value of a can be used only for columns having a small number of stages. According to
Haldenwanger [194] the propertiesof a test mixture map be summarized as follows:
1. The properties of the mixture must approximate to those of an ideal system, i.e.
it must conform closely to Raoult’s law and have a separation factor that is
nearly constant a t all concentrations. ,
2. The vapour-liquid equilibria must be known or be obtainable by calculation.
3. The mixture shall consist of two components only, so as not to complicate the
ineasurenlents and calculation.
4. The separation factor niust have such R value that an adequate, but not too
thorough separation is attained in the apparatus to be tested.
5. The boiling points should lie in such a range that no difficulties are experienced
with the thermal insulation of the colunin.
6. The substances must’ be stable under the conditions of the distillation.
7. The substances and their mixtures must not attack the materials of construction.
8. The substance^ must he readily obtainable.
9. The Substances must be free from contaninants; it shonld be possible to test
their piirity by anal*wis.
10. 9 simple analysis for determining the composition of the mixtures should be
available.
On the basis of cxt ensive experimental work international co-?peration within
the European Federation of Chemical Engineering has resulted in the standardization
of test mixtures. The booklet published b r Zuiderweg [ 1951contains the equilibrium
data as well as all iinportant physical properties and data concerning the chemical
stability of the components for 11 s-=terns. I t is suggested to use mainly the test mix-
tures listed in Table 29 so that it should be possible to compare the efficiency of
packiiigs and coliimns.
iranchembook.ir/edu

4.10 Testing plate coliimns and packed columns 151

It has become international practice to einploy the mixtures n-heptane-tnethyl-

cvclohexane, 1.2-dichloroethane-benxene,benzene-carbon tetrachloride und benzene-
rtkylene dichloride (for plate numbers up to 50 or 60) and benzene-carbon tetra-
chloride (for plate numbers up to 30) for tests a t atmospheric pressure. Brandt and
Riick [ 1961 examined the system n-heptane-methylcyclohexaneas regards its ideality
of hehaviour. It was found that the mixture must be classed as a separate type,
termed “pseudo-ideal,’, because the heats of mixing are not negligible in spite of the
constancy of the relative volatility. The system is also very suitable for testing under
reduced pressure. A further advantage is that the molar heats of evaporation of the
voinponents are almost the same (7.575 kcal/mole). For plate numbers from 100 to
300 heavy wat,er is suitable as a test substance [214]. -The system H,lsO-H,lsO

Table 30
Constants k,, k, and k3 in Murch’s formula [185]

Type of packing Size k, k, k3

(mm)

Ravchig rings 9.5 13.58 -0.37 1.24

12.7 44.95 -0.24 1.24
Saddle packing 12.7 25.61 -0.45 1.11
?IIcRlahonpacking 6.35 0.0114 0.5 1.0
9.5 0.199 0.25 1.0
12.7 0.356 0.20 1.0

allo\vs tests to be performed up t o a few thousand theoretical plates although the

equilibration times are of the order of months (chap. 5.1.4). Exact column tests map
also he perfornied using substances labelled with the radionuclides 14C, 35Sand 36Cl.
I n the pressure range from 20 to 400 mm the mixture cyclopentylcyclohexane-n-
rlodecane has been found to be nearly ideal, whilst dicyclohexyl-n-tridecaneproved
to be approximately ideal only in the concentration range of 30 to 90% dicpclohexyl
[ 1971.
Murch [1851 succeeded in establishing the influence of certain physical p qitrties
of the mixture on the separating effect. He found that the height of a separating stage
is directly proportional to the expression 111 . q/p, in which 111 is the relative volatility.
i , the dynamic viscosity in centipoises and e the density in glnil. By an evaluation of

experimental data Murch arrived a t the following empirical formula :

Height of a separating stage (HETP)
1
- k,Mka dk“ hT
- - ar/e (134)
in which M = mass velocity of the vapour in kg/m2h, d = column diameter in m,
7~ = height of the packed part of the column in in.
The constants, kl,k, and k3 are dependent on the nature and size of the packing
units and are listed in.Table 30.
iranchembook.ir/edu

162 4. Physical fundamentals of the eeparation process

The validity of Murch’s formula is subject to the following restrictions:

1. The distillation must be at atmospheric pressure.
2. The value of M must lie between 25 and 80% of the limiting load.
3 . The ratio of the column diameter to that. of the packing units must be larger than
8 -10.
4. The HEW is that valid for an infinite or very high reflux ratio.
The accuracy of the forniula is said to be &lo%.Its originator, however, sug-
gested that 20% should be added to the calculated figure as a safety factor. Formula
(134)hence gives a rough approximation to the true value and its results should be
checked by experiment.
The HETP is dependent on the co~zcmtmtionof the still pot liquid [92, 1991 and
experience shows that this concentration is best kept above a value of 40mol%,
whilst the amount of the charge should preferablj- be about 8 times the operating
hold-up of the column, The test mixture should he so chosen that the distillate does not
contain more than 90% of low-boiling component. We can consequently divide the
test mistiires, according to the differences in boiling points of the components, into
three groups :
Test mixture
with 4 Kp.
for a large number of separating stages (> 40) 0.5 to 3.0
for a iuedium number of separating stages (10-40) 3.0 to 10.0
for a small number of separating stages (< 10) > 10
For the most important test mixtures tables have been drawn up [92, 146, 147
1951 froin which it is possible to construct a diagram corresponding to Fig. 71,
showing the number of separating stages against the refractive index of the mixture.
The refractive indices of the top and bottom products are marked along the abscissa
and the corresponding difference in the number of separating stages is read off; the
latter figure is then the number of stages for the column tested (section 4.7.4).

4.10.4 Reflux ratio and quantity of reflux

As a rude the number of separatingstages is determined a t total reflux. In practice,
however, it may be necessary to measure how the number of stages changes with finite
rejlux ratios, or what factor must be taken into account for a finite reflux ratio if the
test. has been perfornied at v = 60. This may be done by carrying out an experiment
at a given load and a finite reflux ratio, the distillate not being actually removed,
however, hut returned to the still pot (Fig. 88). Periodically, simultaneous samples
are taken of the top distillate and the still pot contents and their composition is
determined [164]. After the samples have been taken, distillate should be removed for
a time in the normal manner, whereupon the determination is again carried out at8
the new coniposition of the boiler contents. Bxperinients performed in this way by
the author, with a colunin of 30 1nm diameter and 1.16 m packed length containing
2 x 2 x 0.2 ulni helices, resulted in the factors P shown in Fig. 89. Investigations by
Seuinann and Leihnitz [201] confirmed the results.
iranchembook.ir/edu

4.10 Testing plate columns and packed columns 153

P is the factor by which the number of separating stages, determined at u = 00,

should be.multiplied in order that the actual nuinher of stages at a certain finite
reflux ratio may be obtained:
xF = ~W,",,, * (135)

Big. 88 Pig. 89
Arrangement for testing Factor F for 2 :' 2 mni helices and reflns
a colarnn with N finite ratios of 5: 1 to 70: 1
reflux ratio

Prom Fig. 89 it will be seen that with a colunin c;uch as that tested, containing
60-70 theoretical plates, there is little point in increasing the refhix ratio a b o V C ,
say 70: 1. This agrees with formula (110)developed by Rose, which iniplies that there
is no advantage in laboratory distillation in using a reflux ratio greater than the
theoretical plate number of the column.
Another experimental approach is to carry out a batch countmurrent distillation
with the test, mixture a t the selected load and reflux ratio; siiiall samples of t i i t s
distillate (about 1%of the still charge) are regularly taken and also - together with
every second sample - a few drops of the liquid from the still pot ; the coinpo-ition*
are then determined. By plotting the concentrations of the distillate and still pot
iranchembook.ir/edu

154 4. Physical fundament& of the eeparmtion process

contents against the amount of distillate removed, two curves are obtained, as ill-
ustrated in Fig. 90for the system chlorobenzene-ethylbenzenea t 760 and 20 mmpres-
sure a d 1; = 8 : 1 [92]. Theae results are compared with tests a t infinite reflux to
obtain the plate equivalent and the useful efficiency.
The plate equivalent is the minimum number of plates required at infinite reflux
ratio to attain the mme enrichment (xB --f x E ) rn in a countercurrent distillation
with a finite reflux ratio. All distillation conditions except the reflux ratio remain the
same. Thus, in the McCabe-Thiele diagram the separating stages are drawn between
the diagonal and the equilibrium curve (v = ce).
As an example (Fq.W),after the removal of 30 volo/b of distillate (at 20 mm
pressure and 2' = 8: l), the composition of the still pot contents, 17 mol%, (ZB)
corresponds to a distillate concentration (ze) of 60 mol yo chlorobenzene. If these

Pig. 90
0 20 40
distilled -
60 vol% 90

Composition of the distillate and boiler contents as a function of the distillate taken
off (chlorobenzene-ethylbenzene)
(Offtake: 44 ml/h, = 8: 1)
xB at start of distillation: 37 mol?,
A - 760 torr pressure; B - 20 torr pressure

two values, 17 and 60 mole/,, are now entered into a McCahe-Thiele construction or
the Fenake formula (lla), it is found that this enrichment, with v = 00, corresponds
to 17.5 theoretical st,ages. This procedure can be repeated for a number of points and
then yields a curve as shown in Fig. 91. It is seen from this figure that the plate
equivalent - as would be expected - becomes higher at a larger reflux ratio, and
that with a low reflux ratio its value drops appreciably in the early stages of dist<illa-
tion after equilibrationat total reflux. Further it will be observed that pressure has
no appreciable influence on the plate equivalent. The relation between the number of
theoretical stages a t v = do and 2' = 27: 1 and the plate equivalent is demonstrated
by Fig. 92, constructed from values determined by Collins and Lantz [200]. The
column employed, containing 30 actual plates, gave the following results when tested
with the mixture n-heptane-methylcyclohexaneat a load of 30oO ml/h:
with c = do, 19 theoretical stages;
with 2' = 27: 1, 17.5 theoretical stages;
with 1, = 27: 1, 12.5 plate equivalents.
iranchembook.ir/edu

4.10 Testing plate columns and packed columns 155

This result signifies that the separation effected at, v = 27: 1 could also have been
attained a t v = 00 with 12.5 theoretical stages. Conversely, one would have to choose
a column that, at a load of 3000 ml/h, produces an enrichment equivalent to 19.0
stages a t D = 00.
Relating the plate equivalent nmUto the plate nuniber nc0at D = 00 for the sainr
load we obtain, according to Obolonzev and Frost [202], the useful efficiency of a
coluinn :
N = n8qu (at the same load). (136)
ncc

This value may a t most, be unity and approaches unity as the selected reflux ratio
increases.
The method involving plate equivalents and the useful cfficienc3-has the distinct

I
+ 28
C

3 24
.-
$ 20
a,
3 16
d
a
12 1 I I I I I
0 10 20 30 40~0l%50
distilled
Fig. 91
Plate equivalent R S n functionof the amount of distillate taken off,
from 0 -50 vole/, (chlorobenzene-ethylbenzene)
A:v=24:1 (20torr).B:v=8:1 (20torr),C:u=24:1 (760torr),D:u-8:1
(760 tom)

Fig. 92
load -
Theoretical stages -( ) and plate equivalents (- - -) as a function of load
A = Number of theoretical stages, u = co, B = Number of theoretical stages,
= 27: 1, C = Plate equivalents , v = 27: 1
iranchembook.ir/edu

156 4. Physical fundamentals of the separation process

advantage of allowing columns to be compared under varied conditions, as all plate

n u m b ? are referred to v = do.
The determination of the amount of rcfliix is in principle performed at the
column head. If the colunin were to operate adiabatically, that is to say without gain
or loss of heat, and the molar heats of evaporation of the components were identicd,
the amount of reflux measured just above the still pot would be equal to that mea-
sured a t the top. I n practice, however, the quantity of “bottom” reflux is generally

Pig. 93
Destinorni unit for measuring the amount
of bottom reflux and the hold-up

n
mark

Fig. 94 Fig. 95
Reflux measuring device with Labodest reflux meter with
syphon (Sigwsrt and Stage) magnetic valve (Stage)
n = connection to still pot,
b = connection to column,
c = reflux collectringvessel,
a = syphon
iranchembook.ir/edu

4.10 Testing plate columns and packed columns 157

higher than that of top reflux, owing to loss of heat. I n order to find out.whether the
column is operating adiabatically i t is hence useful also to determine the amount of
bottom reflux. This can be done by employing the apparatus shown in Fig. 93.
During a measurement the tap is so regulated that the reflux does not flow into the
flask, but is collected in the graduated funnel. The amount flowing into i t in unit
time is observed with the aid of a stop watch. The form of apparatus illustrated has
the advantage that the quantity measured can be varied in accordance with the
conditions of the distillation, but a disadvantage is that tap grease may be dissolved.
This is avoided in modifications developed by Sigwart. Fig. 94 shows a model with a
syphon and Fig. 95 the preferred form with a magnetic valve. An excellent review of
reflux measuring devices including numerous constructional details is given hy
Gennneker and Stage [245].

4.10.5 Total, static and operating hold-up

The term total hold-up is used to denote the amount of substance that is present
in operating a distillation apparatus, as liquid and vapour, between the surface of
the liquid in the still pot and the condenser. The total hold-up is the sum of the
static hold-up, remaining when the column is a t rest, and the dynamic column content
or Operating hold-up. A knowledge of the total hold-up is of importance, since - as
we have already seen in section 4.7.1 - its influence increases with the number of
separating stages and is unfavourable for separation a t a high reflux ratio, but favour-
able a t a low reflux ratio, whilst with a very high hold-up the reflux ratio loses almost
all its influence on the separation. For Raschig ring packings the theoretical cal-
culation of the liquid operating hold-up after Yilmaz and Brauer [ 1161 is possible
(see chap. 4.2.2). BlaB and Kurtz [268] express the liquid static hold-up of Raschig
ring packings as a function of the Bond number:
BOL pL X g/eL X a'. (1%)
A t small widths, B,, < 1, the static hold-up assumes a n almost constant value.
For the total hold-up of liquid, VLo, below the flooding point an empirical corre-
lation is given.
The rise of liquid hold-up with increasing vapour velocity WGis proportional to
(WG/WGF)lo,where WGFis the vapour velocity a t the flooding point.
Hold-up profiles in various packings were determined by Achwal and Stepanek
[269] in a column with a diameter of 5 cni and a packing of 6 mm ceramic cylinders.
The quantity of liquid remaining in a column after flooding or after terminating
a distillation and cooling is known as the static hold-up. To determine it, a n amount of
liquid equal to about five times the expected value is put into the still pot and distilla-
tion is allowed to proceed for an hour. After the column has cooled the liquid that
has collected in the pot is measured. The difference with respect to the charge is the
static hold-up. In packed columns this quantity is made up of liquid clinging to the
packing, to the walls of the columns, the still head and the condenser. I n plate columns
the major part consists of that remaining on the individual plates. A simplified
procedure that can be adopted is to pour a measured amount of the liquid to he
iranchembook.ir/edu

158 4. Physical fundamentals of the separation process

distilled through the upper end of the condenser and subsequently to determine how
much is retained by the column. "his mectsurement, (which does not refer to the
distillation temperature) should he repeated several times, so as to ensure that
wetting is complete and that a good average is obtained.
If we add to the static hold-up the amount of Liquid temporarily present during a
dist>dlationbetween the still pot and the condenser (which we call the operacing
huZd-up),we get t.he total hold-up. The operating hold-up is found by rapidly removing
the still pot heater during distillation (and if possible also disconnecting the still pot),
and collecting the downflowing liquid in the piece of apparatus shown in a.
95.
For an indirect determination of the total hold-up, both with packed and phte
colums, a small quantity of a high-boiling substance, such as stearic acid, or a-
bromonaphthalene $ .( = 1.6680) is added to the test substance (say n-heptane).
During distillation some of the low-boiling substance passes into the column, so that
the concentration of the additive in the liquid remaining in the still pot increases.
Samples of the still pot contents are periodically taken during distillation and are
analyzed for their additive content. If stearic acid is used, analysis may be performed
bp titration or by evaporating off the test liquid. The content of a-bromonaphthalene
is determined refractornetrically. The total hold-up H is then calculated by the
fortiiula :
H = tB(1 - 1). (138)

in which V , = the volume of the still pot charge; x, = the concentration of additive
before distillation ;x2 = the concentrationof additive in the still pot during distillation.
For purposes of comparison it is convenient to calculate the static hold-up,
operating hold-up and total hold-up per theoretical or practical plate. The total
hold-up varies with the load, but the literature gives little definite information on
this relationship. For an Oldershaw sieve-plate column with a diameter of 28 mm,
contaming 30 actual plates, Collins and Lantz [200] published the values shown in
Fig. 92. The total hold-up lies between 43 and 60 ml, depending on the load, so that
it can be taken to he 1.4-2.00 ml for each actual plate and 2.5-3.5 ml for each
theoretical plate. According to measurements of the author's, the total hold-up of
packed columns is about the same, as will also be seen from Fig. 96, which indicates

60
t, rnL
Fig. 96
250
c
a Dependence of the total
f
I
hold-up on the load with
40 various types of packing
c
0 at 730 torr
a
3 I = Perforated half
' 30
E cylinders of metel sheet,
0
L 2 = Gauze saddles,
20 3 = Raschig rings
0 , 400 800 1200 1600 200omuh2400
iranchembook.ir/edu

4.10 Testing plate colnmns and packed columns 159

the dependence of the total hold-up in ml of n-heptane a t 97 “Con the load a t 730 i m i
pressure [203]. The column diameter was 19 mm, its effective length 812 mm.
I n the calculation of the plate number for batch distillation Zuiderweg [158]
takes into account the influence of the total hold-up and the magnitude of the transi-
tion fraction. Mixtures having a-values from 1.07 to 2.42 were used t o study the
tlependence of “separation sharpness” on relative total hold-up, reflux ratio and plate
number. By nieans of the pole height S the optimum reflux ratio can be determined.
The method corresponds to the determination after McCabe-Thiele with a final state
xB = 5 inol~’.

4.10.6 Operating pressure

The operating pressure is understood to be that observed at the head of a coluliili

during distillation, measured in the gas space just beyond the condenser. Strict1;r-
speaking the pressure should be measured in the vapour at the point of condensation,
i.e. at the same place as the temperature measurement. As this would be difficult.
the tubes of the condenser should be wide enough for the difference in the pressure
between the vapour space before the zone of condensation and the gas space be!-ond
it to be negligible. The difference is more important when high vacuum is employed.
in which case it is advisable to make comparative ineasurements before and beyond
the condenser. The pressure in the still pot is equal to the sum of the pressure in the
vapour space a t the point of condensation and the pressure drop occurring in the
cohinin. I n a vacuum distillation there is a greater volume of vapour to be handled :
it is then often the practice to use packing larger than 1/10 t h of the column diameter,
at least in the upper part of the column, where the lowest pressure prevails and flood-
ing is most !ikely to take place.
The question of how far a reduction in pressure affects the number of separating
stages has so far not been definitely answered. Von Weber [204] found that the number
of theoretical stages changes little; the throughput is lower owing to the larger
volume of vapoui, but the relative volatility generally becomes more favourable (see
section 4.6.2). Myles et al. [192] observed with various forms of packing that tlic
HETP has a minimum a t about 200 mm pressure, whilst Struck and Kinney [20:3]
concluded that the HETP passes through a minimum between 50 and 100 nun, the
value a t about 100 nim being equal to that a t atmospheric pressure (see also secLtion.
4.8 and 4.10.4).
On the basis of extensive experimental data - some of them collected hy Myles [ 1921
- David [191] and Zuiderweg [205] showed conclusively that the number of stage>
becomes smaller the lower the pressure. If the specific number of stages ?is (theoretical
stages per metre) is plotted against the pressure drop, a good comparison of tlir.
efficiency of separation is obtained. Pig. 97 shows such curves for the niixture hen-
zene-ethylenedichloride, as determined by David [206]. A similar conclusion was
reached by Kirschbaum et al. [191], who drew attention to the fact, that, owing to
the pressure dependence of the relative volatility, the change in the specific stage
number will not follow the alteration in the di~tillateconiposition.
Zelvenskij et al. [209] studied the influence of pressure on material transfer in a
iranchembook.ir/edu

160 4. Physical fundamentals of the separation process

packed column using the radionuclides 14C, and aC1. They found that reducing
t,he operating pressure decreases the efficiency of t,he column.
Szapiro [207] derived an equation which can be used to calculate the composition
of the liquid in the flask and of the reflux at the n-th plate as a function of pressure
for ideal binary sptems at full reflux.

11

tg
Fig. 97
c :- Specific stage number n8in relation
.u !il 3 to the pressure drop (David). Test

0 80 ‘I60 240 320 400 480 mrn water

chloride, 2
,-
mixture: benzene-ethylenedi-
14 mole/,
pressure drop, dp ---+

4.10.7 Load and vapour velocity

The term “load” is used to denote the amount of substance which reaches the
column head a-9 condensate in unit time, and is composed of the quantity of reflux
plus the amount of distillate removed. I n industrial usage it is customary to express
the load on a column in terms of the vapour velocity, referred to the free column
cross-section. I n laboratory practice this figure is calculated only if a comparison
with a semi-technical or industrial installation is required. It may be usefully ex-
pressed ats ml/cm2 h.
The separating efficiency of a column is dependent on the load and this influence
is greater, the higher the reflux ratio (Fig. 98). In reporting tests it is therefore neces-
sary to date the load. According to a proposal of von Weber tests should be performed
at 21.7 of the maximum load, in order that results may be comparable. The niaxi-
mim load is understood to be just below the load at which the ascending vapour
prevents the reflux from flowing down. I n this case the column “floods” and counter-
current exchange becomes impossible.
Further important observations are that the efficiency of small packing units
(2-4mm diameter) is more dependent on the load than that of larger packing
(5-10 mm), as shown in Fig. 98, and - as has already been stated - that the
optimum load with respect to intensity does not lie at a low figure, but rather just
helow the point of flooding (Fig. 83) [214]. The highest efficiency of laboratory
columns, however, is not situated at the upper limiting velocity, as indicated by
David [ 1911for industrial columns, but at a very low load, approaching the minimum
load.
If it is desired to compare laboratory columns with semi-technical or industrial
columns, account should be taken of the fact that the vapour velocities in the former
Itre generally very much lower, and lie in the range of 0.02 to 0.5 m/sec, whilst in
iranchembook.ir/edu

4.10 Testing plate columns and packed columns 161

load --m
Fig. 98
Specific stage number n, (nth per m) as a function of the load for various packings according to
Stage (Refs. chap. 1: [39])
(:urves 1 - 1 2 : Test mixture benzene-1.2-dichloroethane
Curves 13-23: Test mixture n-heptane-methylcyclohexane
Measurements of Schultze and Stage Measurements of H. Kolling
I = V2A helices, 2 x 2 x 0.2 mm 13 = V2A helices, 2 mm o
2 = V2A helices, 4 x 4 x 0.4mm 0 14 = V2A helices, 3 mm @
3 = glass Raschig rings, 6.5 x 6.5 mm 0 15 = V2A helices, 4 mm 0
4 = glass Raschig rings, 4.5 x 4.5 mm o 26 = V2A helices, 6 mm 0
5 = small glass spheres, 3 mm 0 I7 = V2A helices, 8 mm 0
6 = large glass spheres, 7.5 mm 0 , matt Measurements of H. Brauer
7 = small glass spheres, 4 mm 0,matt
= porcelain Raschig rings, mm 18 = V2A spaced helices, 3 x 3 V.4 mrn z
= earthenware Raschig rings, 10 10 mm la
19 = V2A spaced helices, 4 x 4 i0.45 iiim a
10 = glass spirals, 3 mm @, 1-5 mm long 20 = V2A spaced helices, 3 x 3 x 0.45 mm 3
21 = iron single-turn helices, 1-3 mm 0, 21 = V2A helices, 2 x 2 x 0.2 mm 0
ca. 2 mm thick 22 = V2A helices, 3 x 3 x 0.45 mm 0
13 = porcelain Raschig rings, 5.5 x 5.5 mm 0 23 = V2A "" mm

11 hrell, Handbooh
iranchembook.ir/edu

162 4. Phvsical funhment.de of the separation process

industrial installations they attain values of 0.1 to 2.0 m/sec and more [184]. With
packed columns it is important to keep this velocity constant; in the case of plate
columns the efficiency is not so greatly affected by variations in the vapour velo-
city.
Otherfactorsdependent on theloadare thetotalandoperating hold-up, thepressure
drop and the limiting velocity. The latter varies with the shape and size of the packing
(or the size and arrangement of theplates) and with the properties of the substances tfo
he separated. These aspects will be dealt with in detail in chap. 4.11. An exhauetive
trea.tment of the loading conditions in packed colunins was given by Stage and Bose
(Refs. chap. 1, [39]).

4.10.8 Method of column calibration

Preyctmtim of the column
In calibration it is necessary to work with painstaking care. The whole apparatus
must be rigorously clean and dry, and RO it is desimble to leave the jacket heating
switched on overniqht before the test. Column packing must also be thoroughly
cleaned before use. A good method i s to carry out a preliminary cleaning with carbon
tetrarhloridt- and trichioroethylene, then to w d i with hot benzene and finally to
rinse again with trichloroethylene. When filling the column care should be taken not
to touch the packing with the fingers.
The method employed for filling a packed column is important. The best proce-
dure is to put in 3 or 4 packing units at a time, the column being constantly tapped
with a piece of wood. With small-sized packing a rapid method of obtaining a random
arrangement is to use the apparatus Bhown in Fig. 87.It is also useful to remove the
packing after a test, to re-introduce it after cleaning and drying end then to repeat
the test. This will show whether the method of filling exerts an influence on the
result.
The cleaning of plate columns is more difficult because traces of liquid tend to
remain on the plates. The niost satisfactory procedure is to distil a volatile solvent
through the column and then to dry it with a current of hot air, the column being
surrounded meanwhile by a heated jacket.
The ccnwunt o j test mixture should be 8 times the total coluinn hold-up, and this
figure should previously be determined, as described in section 4.10.5. The sirte of the
s t i pot should be such that it is initially about two-thirds full. The composition of
the test mixture should be adjusted to 30--40molo/, of low-boiling component.
The influence of the composition on the separating effect has as yet not been fully
established [208].
The purity of the components used for the test mixture requires close attention.
It is not safe to employ substances sold as “pure” or “analytical grade” without a
previous examination. A meaaurement of the refractive index alone is not enough.
C‘nambiguousanalyses can be carried out only with infrared spectroscopic and, above
all, gas-chromatographic measurements [1961. If the properties agree with published
figures and if the distillation range (measured with a thermometer graduated in
tenths of a degree C and with correction for barometric pressure) shows variations of
iranchembook.ir/edu

4.t0 Testing plate columns and packed columns 163

less than 0.1 rleg C, a further purification is not necessary. Most substances, however,
require purification by chemical means [210, 211, 2121 and a subsequent sharp
fractionation a t a high reflux ratio. If insufficiently pure components are used two
dangers exist: firstly that the refractive index calibration curve will be displaced,
secondly that the impurities will be concentrated during the test a t the top or in the
still pot and will vitiate the results.
With the carefully purified components mixtures are prepared in the range of
5 to 95% (wt),their refractive indices are measured and a calibration curve (nE-76 wt)
is plotted. If desired, the number of stages required with v = 03 for a series of enrich-
ments may be calculated by means of the Fenske equation (114)) and a chart for
these values (Fig. 71) against the refractive index constructed. From this chart the
number of stages may be read off directly after taking samples and determining the
refractive indexes.

Fig. 99
Destinorm sampling device for normal
and reduced pressures

It is necessary to ensure that the small samples (0.1-0.2 nil) taken from the
column head or the still pot are not contaminated by tap grease. The large errors
that can result have generally been overlooked. For the removal of samples from
the still the arrangement illustrated in Fig. 99 has proved satisfactory. The still head
employed must be so designed that it not only has a small dead space but also allows
single-drop samples to be taken without contact with a tap. An example is illustrated
in Fig. 100. During the equilibration period the small funnel a is turned downwards.
When a sample is to be taken it is turned LIP and liquid is drawn off by connecting b
to vacuuiii via a rubber tube and pinchcock and opening the latter carefully. The
liquid falls through the wide passage of tap d ; when 0.1-0.2 ml has been collected,
tap d and the pinchcock are closed and air is admitted for an instant through valve c,
forcing the liquid in the tube back into the column. Atmospheric pressure is restored
in the receiver space i through b. If normal pressure is used air is blown in a t the
vacuuni connection, in order to empty the funnel tube completely. Fiirther examples
will be given in chap. 7.5.3 (see also Refs. chap. 1, [39]).
If reliable data are to be obtained it is necessary to operate the column adiahati-
cally. The still pot and all components up to the condenser should therefore be
carefully insulated, for example with asbestos cord and glass wool, so as to minimize
loss of heat (cf. section 7 . 7 ) .Since, even a t temperatures below 80°C, a certain amount
of “false reflux” is formed, it is essential to compensate for the loss of heat from the

11*
iranchembook.ir/edu

164 4. Physical fundamentals of the separation process

column by providing a heating element inside the jacket. Automatic control of this
heading element may be effected by means of gas thermometers (cf. section 8.2.2).
Heating jackets that can be opened on hinges (section 7.7.3) may be used with ad-
vantage, since they can be fitted after building up the glassware and can be removed
at any time during a distillation in the event of trouble. The sections between the
still pot and the column jacket, between the latter and the still head and the section
up to the condenser are first encased in a layer of about 20 mm of asbestos cord and
this is again surrounded by a winding of gkss tape. The total thickness of the insula-
tion should be about 60 mrn. The safest course is to apply heating tapes to the in-
sulation.

Fig. 100
Destinorm column head with
device, having almost no dead
space, for ta.king single-drop
samples
(see accompany text)

Particular care rnust be taken that the point at which the vapour temperature is
measured in the column head is not wetted by undercooled liquid reflux. Furthermore,
when working at reduced pressures, there should be no loss in presslire between the
points where the temperature and the pressure are measured. These conditions are
sat,isfiedin the Destinorm column head (Fig. 100). The temperature is measured in the
insulated vapour tube e by means of a thermometer f or a thermocouple. Since the
reflux is returned to the column through g and the reflux measuring pipette h at a
level below the thermometer, the latter cannot be wetted. The pressure is measured
directly beyond the wide cold trap, so that no appreciable reduction in cross section
occurs between the thermometer and manometer.
Adjwtment of test colditions
The separating effect of the column is dependent in the first place on the load and
this must therefore be kept constant during a test run. A satisfactory method is to
control the rate of evaporation by the pressure drop, using a contact manometer
(section 8.4.2), and observing the heat input on an ammeter. Before the load is
iranchembook.ir/edu

4.10 Testing plate columns and packed columns 165

adjusted, however, the column (if of the packed type) should be flooded, so that the
packing is thoroughly wetted. To do this, the heating current to the still pot is
Increased, while watching the contact manometer, until flooding starts at the foot of
the coluinn and advances up to the colunin head. The column is flooded in this way
for about 15 minutes, whereupon the heat is reduced, so that the reflux can return.
This operation may be repeated several times, after which the desired load is ad-
justed by ineans of the contact manometer (see also Refs. chap. 1: [39]).
somewhat less effective procedure consists of “cold-flooding’’ the column by
filling it with the mixture to he distilled, drawing off the liquid and then carrying out
the distillation.
How strongly “warm flooding” influences the efficiency is clearly seen in Fig. 101.
0
The tests were done in a Podbielniak column of 25 mni diameter with Heli-grid
packing (Fig. 275) at a load of 1000iiil/h. It is seen that the stationary state was
reached only after 8 to 10 hours and that the flooding increased the plate number up
to *50?4 [213].

Fig. 101
Dependence of the separating
efficiency on flooding (after
Brandt et al.)

This flooding is particularly recommendable for wire packings of sizes between

1.6 and 4.0 mm whereas for Raschig rings of 4.5 x 6.0 inn1 the initial degree of
wetting does not influence the efficiency. For flooding under reduced pressure, see
chap. 5.4.1. Flooding points and points of optimum efficiency of a column packed
with Raschig rings (5.5 x 4.0mm) and wire gauze units (V4A steel, 2 mm) were deter-
mined by Lentz [214]. The test substances were H,O-D,O (27% DzO) and n-hep-
tane-methylcyclohexane, the column diameter was 37.5 mm. n, and the pressure
drop Ap were determined via the vaponr velocity. The point of optimum efficiency
lies in the flooding point for the Raschig rings and between 81 and 880/, of the maxi-
mum load for the wire gauze packing. Operating a packed column in the state of
“flooding onset” may increase the transfer of matter by a factor of 6 together with
an inreased throughput [214a]. Kafarov et al. [214b] call this state phase inversion,
in which the liquid fills the spaces in the packing almost completely and the vapour
in the form of threads and vortices disperses (“emulsifies”) the liquid. The liquid
hold-up as well as the pressure drop and the platle number rise drastically. Relations
and diagrams for estimating the particular parameters are given.
After the column has been adjusted and has run at total reflux for about 2 hours
a quantity of 1-2 ml of liquid is removed a t the top and is rejected. After a further
hour the first samples of 0.1-0.2 ml are taken from the top and the still pot for
examination. The same precautions are observed as in sampling for phase equilibrium
nieasurements (section 4.6.3). The compositions of the samples are determined;
iranchembook.ir/edu

186 4. Physical fundamentals of the separation process

samples are taken every hour until the compositions of the overhead distillate and
the still pot contents remain constant. The samples should 6e as small as possible,
so as not to upset the stationaq- condition of the colunin. Equilibrium re-cstablishes
itself more rapidly in a short column than in a long, multi-stage column; in the latter
case it may be necessary to make the intervals between the taking of samples much
greater than an hour. With the usual test mixtures the time required for reaching
a stationary state is approximately 10 to 20 hours. It is advisable to carry out the
tests in the order both of decreasing and increasing loads [208].
To find the separating efficiencyof a column under operating conditions, the proce-
dure described in section 4.10.4 for determining stage equivalents should hefollowed.

Ca.?-ctd&ion 01 the number of separdi+ stages

When 3 to 5 values of the top and bottom concentrabions remain constant the test
is stopped; it should, however, preferably be repeated another day under the same
conditions for corroboration.
The number of separating stages is now determined. I n order to do this bycalcu-
lation the figures for x, (dist,illate) and x, (still pot composition) are introduced into
the Fenske equation (114). For a graphical determination the method of McCabe-
Thiele is followed (section 4.7.1). If the run was carried out under operating conditions,
that is to say with a finite reflux ratio, the stage equivalent may be calculated as
described in section 4.10.4. The number of stages to which the separating efficiency
of the column corresponds under operating conditions may also be coniputed by
iiiyans of a formula due to Oholentsev Rnd Frost [202]:

(1'39)

in which u = mol. fraction of distillate removed from charge; xE = content of low-

hoilmg component in thedistillate (mole fraction) ;xF = content of lorn-boiling compo-
nent in the initial mixture (mole fraction).
Obolentsev and Frost proposed that the procedure for determining the number
of separating stages in the column should be as follows. After equilibrium has been
est.ablished a sample of the distillate is taken at a finite reflux ratio and at a point
where the fraction distilled corresponds approximately to the weight or mole
fraction 2,. in the original charge. In this way one obtains the number of plate
equivalents, ttmpqu/v(cf. section 4.10.4). According to this method the column is first
allowed to run for 3 hours with t i = 00 a t the selected load, and t'hen a t a finite
reflux ratio. As the forinitla is inconvenient to use, a corresponding nomogram has
been developed [84,202].

4.10.9 Data for packed and plate columns

It is practically iiupossible to obtain reliable comparative values for the HETP's
of packed columns and for the plate efficiencies (or average enrichnient ratios) of
plate columsfroni data published in theliteratnre, since thesedata refer to various test
mixtures and were obtained under different conditions. Only in rare cases are ell the
necessary data, mentioned in section 4.10, given. Some agreement on a standardized
iranchembook.ir/edu

4.11 Pressure drop, limitsingvelocity and calculation of column dimensions 167

testing procedure should be reached, so that published resiilts inay be comparable.

Other sources of uncertainty are that the purity of the substances used in testing
is not always certain, and that errors due t o dissolved grease from sampling taps inay
have arisen. With our present knowledge of such matters it is desirable to test the
more important types of packed and plate columns again, using accurate modern
analytical methods, such as infra-red spectroscopy, gas chromatography and mass
spectrometry.
However, with the object of providing a rough idea, the present data for the
most important types of packed and plate columns [215] have been collccted in
Table 31. The various forms of packing and plate columns will be described further
in sections 7.8 and 7.3.3.

4.1 1 Pressure drop, limiting velocity and calculation

of column dimensions

I n the laboratory it is usual to employ coluinns having diameters of 10-30 nim

for distillations a t normal pressures and of 20-50 mm for distillations a t rednced
pressiires. For very large throughputs, even at normal pressure, a diameter greater
than 50 rnm is necessary. The recent view that a column should be loaded to just
below the point of flooding makes it necessary to settle on its diameter beforehand,
qo that the desired throughput may be realized.
The load on a column is the amount of substance that passes through it and
condenses a t the coluinn head in unit time, i.e. the sum of distillate and reflux. .4h
the packing, or other obstructions present, offer a resistance both to the ascending
vapour and the descending liquid, t,heload cannot be increased beyond acertairi limit.
If we measure the pressures a t the top and bottom of the column we observe that
their difference (the loss in pressure or pressure differentid over the colunm) rises
with increasing speed of evaporation or coluinn load. This pressure loss is dependent
on the nature and dinlensions of the column, on the packing or plates present. on the
distillation pressure and the properties of the mixture distilled. The pressure drop
in annular columns may be calculated by formula (195) (section 7.3.1). Pressure drop
figures for columns with rotating elements have been ~ollect~ed in Tables 62 and 64.
Plate columns have the advantage that, they can he used at very low loads and
that the separating effect may then be a maximum, whilst packed columns require a
minimum load to ensure a proper countercurrent exchange. In large-scale operat ion
columns are often controlled by the pressure differential, so that flooding may he
avoided and an optimum separating effect maintained. The same method can he used
with advantage for laboratory columns (cf. section 8.4.2).
The term lower linriting velocity is used to denote the vapour velocity (referred
to unit cross section of the empty column) below which the effectiveness of the
column begins to fall off, the term upper limiting velocity for the vapour velocity at
which flooding is so intense that, in plate columns, the layer of liquid on a plate is
entrained upwards, and in packed columns a quantity of spraying liquid rises from
the foot of the column to the top. Since the fundamental pnblications by Mach [219],
iranchembook.ir/edu

168 4. Physical fundamentals of the separation process

Table 31
I. Test data for packed columns
Packing Column

Type of packing Material Diameter Height Wall Surface Column

thickness area of diameter
one litre
(mm) (mm) (mm) (m? (mm)

RaSchig M g s Glass, smooth 3.0 3.0 0.7 1.320

Glass, smooth 4.0 4.0 0.6 1.109
Glass, smooth 4.5 4.5 0.5 1.382 24
Glaas. smooth 5.0 5.0 35
Glass, smooth 6.0 6.0 20
Glass, smooth 6.5 6.5 0.5 1.109 24
Glass, smooth 10.0 10.0 50
Porcelain 5.5 5.5 1.1 24
Porcelain 8.0 8.0 1.1 24
Earthenware 10.0 10.0 1.7 24
Dixon rings Wire mesh 3.0 3.0 26
Wire mesh 5.0 5.0 20
Wire mesh 6.0 50
Prym riugs Metal 2.2 2.5
Wilson spirals Glass 3.0 1.5-3.0 0.4 24
Glass 5.0 5.0 0.65 1.66
Glass 6.3 25
Fenske helices Glass 8.1 10
Gl8SS 3.1 13
Glass 3.1 20
Glase 6.3 25
Helices Stainless steel 1.6 1.6 0.2 40
Stainless steel 2.0 2.0 0.2 4.55 24
Stainless steel -1.0 4.0 0.4 1.49 24
Beads Glass, smooth 3 24
Glass, smooth 4 0.9
Glass, matt 4 24
Glass, matt! 7.5 24
Saddles Porcelain 4 1.5 30
Porcelain 6 1.15 30
Porelain 8 30
Porcelain 10 0.72 30

Sigwart : No further data supplied.

"hormann : The lower HETP's are for smaller loads a.nd conversely.
iranchembook.ir/edu

4.11 Pressure drop, limiting velocity and calculation of column dimensions 169

Height equivalent to a theoretical plate, cm

Cross- Packed Schultze Sigwart Thormann Myles Other References

sectional height and Stage chap. 4. chap. 4. e t al.
area chap. 4. chap. 4.
(cm2) (cm) [216] [217] [a181 [192]

5-7
6-9
4.5 60 5.82-9.52
9.6 90 11.1 7- 12
3.1 155 15.5 8- 14
4.5 60 5.0-14.3
19.6 95 18.2
4.5 60 3.69-7.5
4.5 60 7.07-11.3
4.5 60 7.23-12.8
4.9 1u0 1.4-3.8 Ch. 7.5
3.1 60 4.33
19.6 100 1.4-3.5
3-3
4.5 liU 8.0-13.35
7-12
4.9 107 7-9.2
0.8 80 4.8
1.3 126 7.4 4-6
3.1 125 9.5
1.9 107 3.6-4.2
12.5 100 1.4 -2.0 Author’s measurements
4.5 60 1.25-3.32
4..i 60 1.82-5.0 2-5
4.5 60 6.00-6.06 3-6
6-7
1.5 60 5.82-7.22
4.5 60 3.31-7.50
7.0 45 4.7-6.3 Author’s nieasurements
7.0 4.3 6.0 -9.0 Author’s measurements
7.0 45 7.3- 11.4 Author’s measurements
7.0 45 9.0- 12.8 Author’s measureinents

Nyles e t al. : Test mixtures are n-heptane-methylcyclohexaneand n-dodecane-cyclopen-

tylcyclohexane; t h e HETP’s are for various pressures and a medium reflux flow.
iranchembook.ir/edu

170 4. Phyaical fundamentals of the separation process

Table 31 (continued)
11. Plate oolumns [215]

~lolumntype Distance Column Plate Plate

(doe to) between diameter capacity effioiency
plates
(mm) (mm) (mu (%)Y
Keesom 23 -2.5 % 33 1.o 50
Groll-Oldemhaw 25 25 ca. 1-2 80-90
Sigwart 65 50 ce. 7.5 loo- 110
Brunn 20 and 40 5 0.7 and 1.4 70-100
Greiner and Friedricha 40 46 ca. 2-2.5 64 -66
Stage (vapour-collision 00 70-90
plates)
Stajp (bubble-cap plates 50 80 -95
with shielded down-
comers)
Stage (vspour -collision 35-150 80-120 3-75 76-95
bubble-cap plates)
Yrhmticlder and Fritz 80 -90
Brand 40-90
Sorinschliffgerate 40 -95

l) Caution should be used in comparing these figures since they represent only rough estimates
of the plate efficiencies (cf. chap. 7.3.3)

Barth [220] and Leva, [221] appareed, a large number of contributions have been
written with the aim of elucidating the principles of pressure drop in packed columns.
I n an extensive invefltigatiorion the relation between pressure drop and the fractionat-
ing effect in packed columns Brauer [208] determined the pressure drop with 1 3 types
of packing (Table 32) as a function of the flow. The experiments were performed in a
column of 37 nim I. D. and 1 m length with the test niixtiire n-heptane-methyl-
cyclohexane. The results of these measurements are represented in Fig. 102 and
bring out the following facts :
1. curves have the same characteristic form. The limiting breaks in the curves lie
on straight lines a - ( I , b - b, c - c.
2. In all the ranges of load limited by the breaks in the curves the course of the
pressure drop lines can be represented by- a formula of the form
(14.0)
Type of vapour flow
11 = 1.0 in the range below a - (1 Laminar
n = 1.78 betweena - @.andb - b Turbulent
n = 2.7 between b - b and c - c Turbulent
n = 3.3 in the range above c - c Flooding, format,ionof a layer of spray
3. The value of c in formula (140) is a function of the properties of the vapour and
the liquid and of the chawcterist ics of the packing.
iranchembook.ir/edu

4.11 Pressure drop, limiting velocity and calculation of column dimensions 171

Pig. 103 shows the relation between pressure drop and the fractionating effect.
The number of theoretical plates per metre of the spiral packing listed in Table 32
was plotted against the pressure drop in the turbulent region. The linear form of the
curve shows that the same relationship exists between the flow and exchange of heat
and material in the whole of the turbulent region, in other words that the mechanism
of exchange is constant. Above the break line c-c the vapour passes as bubbles
through the accumulated liquid and the process of exchange follows other laws
[203, 206, 2081.
In his investigations Kast [222] considers the actual path of the two-phase flow

rnrn w

20 40 60 80 100 200 400 rnllcrnZh 1000

load -+
Fig. 102
Pressure drop caused by the packings listed in Table 32 as a frinrtion of the liquid
load (Brmer)
iranchembook.ir/edu

172 4. Physical fundamentale of the separation process

Table 32
Paokings used by Brauer for pressure drop measurements [208]

Xo. Dimensions Type Jiaterial Diameter Relative Specific Un-

ram ratio, free surfaceof wound
packing volume I packing o length2,
unit: EO (m*/m8) (mm)
column
~F&K

1. 3 ‘i3 x 0.45 spaced 18-8 stainless 1:12.3 0.777 1802 43

spiral steel
> 4
-. 4 Y 0.5 spaced 18-8 stainless 1:9.25 0.822 1480 61
spiral Steel
3. 3 3 x 0.40 spaced 18 -8 stainless 1: 12.3 0.771 1815 44
spiral steel
4. 2 2 .I‘ 0.3 close 18-8 stainless 1:18.8 0.661 2 680 32
spiral Steel
> :I :i 4 0.4.5 close 18 -8 st&dess 1 : 12.3 0.704 1717 60
spiral steel
6. + 1 ,’ 0.4 close 18-8-3 stain-’ 19.25 0.760 1543 100
spiral less steel
i. 2 , 2 0.25 Sheet-metal 18-8 stainless 1: 18.5 0.716 1965 10.5
rings with steel
cross piece
8. 4 4 0.3 Sheet-metal 18-8 stainless 1 :9.26 0.824 1176 18.7
rings with steel
cross piece
9. ti ti 0.4 Sheet-metal 18-8 stainless 1616 0.811 946 24.1
rings with stml
cross piece
10.
-
3 J
- Raschig Porcelain 1:7.1 0.478 977
rings
11. 3.5 7.4 Ellipsoids Aluminium 0.306
12. ? - Balls Glass 1:18.5 0.339 1786
13. I t i (16 Raschig Porcelain 19.3
rings

and replaces the length of packing H K with an effective length p H K , where p is a

physically justified path factor. In the region of transition from laminar to turbulent
flow in packed columns Kast. reports good agreement of measurements by himself
and others with the pressure drop formule for packing units up to a size of 50 nun.
Teutsch [223] derived a semi-empirical equation which allows a straightforward
calculation of pressure drop in terms of a dimensionless pressure loss. The range of
validity is mentioned, there being the limitation that Teutsch’s work covered only
packing units 2 25 mm. Gelbe [224] examined spaced helices of size 4 x 4 x 0.5 mni
with the mixture n-decane-transdecane in the range from 10 to 740 tom. On the
hasir of Gelbe’s measuring results Glaser [225] developed the following relation for a
iranchembook.ir/edu

4.11 Pressure drop, limiting velocity and calculation of column dimensions i73

where @D = vapour density (kg/m3), KK = length of packing (in), HTU = height of

a transfer unit (m), dp = pressure drop (kplmz).

2 4 6 8 10' 2 4 6 8 10' mm waterlm

pressure drop,dp/h-
Fig. 103
Relation between pressme drop and fractionating effect (Brauer) (cf. also Table 32)

2 50
mm water
200
4
:I50
P
U

5 100
Ln
Fig. 104
Pressure drop with porcelain
? saddle packing (4-8 mm) as a
a
50 function of the rate of gas flow
(air); liquid (water):
40 ml/cm2 x h
0 0.2 0.4 0.6 0.8 l.Om/sl.?
gas velocity +

The coefficient as a function of the relevant parameters can also be given ap a

dimensionless value. It varies with the load in a way which clearly indicates the close
connexion between flow and material transfer and the influence of pressure.
Reichelt and Reichelt and Blass used their own measuring data and those of
others to investigate the flow through packings consisting of Raschig rings and
spheres. They found equations for the calculation of the pressure drops in sphere and
cylinder packings with one-phase flow.
Heckmann and Krell [226] carried out measurements of the pressure drop across
porcelain saddle packing (4, 6 and 8 mm), with air as gas and water as liquid. The
effective column length was 1 m, the column diameter 22 mm. Fig. 104 gives t,he
results.
iranchembook.ir/edu

174 4. Physical fundamentals of the separation process

mm

Fig. 106
1.0 1.1 1.2
WD -
1.3 1.4 1.5 m/s 1.6

Pressure drop across Wilson spirals (3 mm) with an increa,singflow of water, as R

function of the gas velocity (air)

crn wa

Fig. 106
throughput - h

Pressure drop found with wire gauze rings (3 x 3 mm) as alfunctionof the through-
put a t atmospheric pressure. Column dismeter:12€iamm; average height of
parking: 1 m (measurementsby Thiirkauf)
iranchembook.ir/edu

4.11 Pressure drop, limiting velocity and calculation of colnmn dimensions 176

The effect of an increasing flow of liquid with 3 inni Wilson spirals is shown in
Fig. 10.5. A w r y favourable pressure drop relationship is foiind with wire gauze (e.y
Dixon) rings, as can be seen froni Fig. 106 and 107. The figures plotted are average
values obtained with a number of different test mixtures.
The “coniparable~’air velocity as a measure of the vapoiir or gas load was derived
by Neuniann [227] froin einpirical results.
Reed and Fenske [228] have evolved a forinula for calculating the pressure drop
in packed colinnns which is given by Rraiier [208] in this form:

(142)

tori

Pig. 107
throughput - Ti

Pressure drop per separating stage found with wire gauze rings, as a. function of
the throughput at various operating pressures. Column diameter: 50 mni
(2 inches); average height of packing: 1 m (measurements by Thiirkauf)

in which y = the absolute pressure (kg/ni2);- J p = the pressure differential (kg/ni2\:

M = the molecular weight; = (VK - VFk)/VK = the relative free volume of the
dry packing; VK = the volume of the effective length of cohiinn (m3); VFk = the
volume of the packing (m3; (measured hy displacement of liquid); BF = the
volnine of the liquid present in the coliinin (n13): R = the gas constant. (kg m/K)
T = the absolute teniperatiire (K); ElTE;= the effective length of the packing (ni).
g = acceleration due t o gravity (m/secz); 11 = dynamic viscosity of the vapour
(kg . sec/in2); g = density of the vapour (kg . sec2/m4); 0 = surface area of packing,
relative to colnmn volume (111~!m~); D = vepour rate (1n3/11i2 sec).
The agrceinent with experimental data is very good if the packing has been
charged in the optimum manner.
The fact that the properties of the substance undergoing distillation have an
af~precinhleinfluence on the pressure drop is apparent froni the curves shown in
Fig. 108 [22Y] and froni Table 33 [200]. The data listed in the latter table were
determined in an Oldershaw sieve-plate column (28 inin diainrter) with 30 actual
plates.
iranchembook.ir/edu

176 4. Physical fundamentals of the separation process

It should be noted that the substances mentioned in Table33 were so chosen

that they differed in density and surface tension at their boiling points.
Umholtz and van Winkle [230] also determined the pressure drop as a function of
load for sieve-plate c o l m s , which had a diamet'erof 25 mm (1inch), a plate spacing
of 50 m m (2 inches), a free area of 16.2% and contained a varying number of holes
uf different diameters.
Data showing the dependence of the pressure drop and the limits of load on the

mm

40

20

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35p 0.4

vapour velocity --+
Fig. 108
Pressure drop as a function of the vapour velocity for various substances, accord-
ing to Stage
1 = Ethanol, 2 = Chloroform, 3 = Benzene, 4 = Heptane, 5 = Di-isopropyl
ether, 6 = Amy1 alcohol. 7 = Toluene, 8 = Ethyl acetate
iranchembook.ir/edu

4.11 Pressure drop, limiting velocity and calculation of column dimensions 177

distillation pressure, for various types of packing, have been collected in Table 34
[192]. From these figures it can be seen that spheres have a relatively narrow load
range and cause a large pressure drop on account of their small free volume (269b).
On the other hand Heli-grid packing (section 7.3.4) gives a wide load range and a high
upper load limit (see also Refs. chap. 1: [39]).
For a calculation of the required column dianwter it is necessary t o know the
upper limiting velocity. Schurnacher [231], on the hasis of theoretical considerations,
has calculated velocity constants bG for various types of packing h p means of the
following formula :

(143)

where pR = eD/eF(= density of vaponrfdensity of liquid, in the same units at the

operating pressure), and further w G = the limiting velocity, m/sec; kG = the limiting
velocitjy constant; d = the diameter of the packing units in m; g = the acceleration
dne t o gravity (9.81 rn/sec2).

Table 33
Pressure drop in an Oldershaw sieve-plate column during the distillation of substances
having various densities and surface tensions [200]

n-Heptane Methylcyclohexane Benzene Formic acid

Load Pressure Load Pressure Load Pressure Load Pressure

drop per drop per drop per drop per
actual plate actual plate actual plate actual plate
(ml/h) (torr) (ml/h) (torr) (ml/h) (torr) (ml/h) (torr)

1000 1000 1000 0.76 200 0.88

1500 0.50 1500 0.62 1500 0.82 400 0.95
2 000 0.53 2000 0.63 2000 0.84 600 1.02
2 500 0.55 2500 0.65 2 500 0.87 800 1.09
3 000 0.57 3 000 ' 0.67 3 000 0.91 1000 1.18
3 500 0.60 3500 0.71 3 500 1.00 1125 1.23
4000 0.64 4000 0.77 4 000 1.11
4500 0.70 4500 0.85 4280 1.17
5 000 0.77 5000 0.93
5 230 0.80 5230 0.97

Column diameter: 28 mm; number of actual plates: 30.

Substance ICP.760 e20

Benzene 80.09 0.87895

n-Heptane 98.424 0.683 68
Formic acid 100.75 1.2206
Nethylcyclohexane 101.10 0.76944

The surface tension of these substances in contact with air lies in the range of 20-40 dynes/cm
at 20°C.
~

12 =ell, Handbook
iranchembook.ir/edu

Tuble 34
Presstire drop and load liniits for vctrioua types of packing ut n number of pressure's 11921

Pucking Size Column diameter: 26 mm 20 t,orr 50 torr 100 torr 160 torr 760 torr
(mm)
length: 1.07 m Min. Mux. Min. Max. Min. Max. Min. Max. Min. Max.

Heli-grid - Pressure loss (torr), - - 5 28 28 6 27 1 27

Load limits (g/h) - - 1200 1650 1600 660 1100 BOO 6500
Spheres, loose packing 3.6 Pressure loss (torr) 6 26 5 32 35 5 32 3 27
Load limits (g/h) 450 700 600 900 1150 1000 1400 126 2600
Spheres, clorte packing 3.6 Pressure lorts (torr) 3.5 31 3.5 30 30 4 35 5 26
Load limits (g/h) 380 720 620 900 1170 1000 1300 - -

Suddle packing 6.3 Pressure loss (torr) 2 8 1 8 - 2 8 2 8

Load limits (g/h) 450 800 (100 1100 .- 900 2200 2600 4800
Triple glass licliues 6.3 Pressure loss (torr) 1 ti 1 6 - 1 5 1 -
Load limits (g/h) 460 900 800 1400 _- 1000 2200 1200 1600
Singlc glass helices 6.3 Pressure loss (torr) 2 14 3 14 - 3 18 2 -
Load limits (g/h) 450 900 970 1200 - 1460 2000 2400 -
iranchembook.ir/edu

4.1 1 Pressure drop, limiting velocity and calculation of column dimensions 179

This formula deals with the influence on t h e limiting velocity of the size and type
of the packing and the density of the vapour and the liquid, but does not take into
account the viscosity of the vapour and liquid and the surface tension of the liqiiid.
The values found for the constant kG were as follows:
for Raschig rings 0.01
for Pryrn rings 0.021
for Rerl saddles 0.019
for solid packings 0.004
for helices 0.0126 (the author's measurements)
=\P an aid in determining the limiting velocity a noniogram was developed
(Fig. log), the use of which can be illustrated by the following example.
The limiting velocity is to be determined for the countercurrent distillation of
a Ci fatty acid at 20 nini Hg (boiling point 125.0"C at 20 mm), using 4 min saddle
packing.

Fig. 109
reduced vapour density, qR' -
Nomogram for the determination of limiting velocities for packings (Schumacher)

12"
iranchembook.ir/edu

180 4. Physical fundamentals of the separation process

Deternindon of the necessary &a

20
kc = 0.019; e D = -x 130 x 0.01606 = 0.105 (forniula (37));
398
@R = @D/@ = 0.106/900 (g/l at 200 m)= 0.1165 x w3.
Uae of the ?Lornogram
The examulr has been indicated by dashed lines in Fig. 109. Through the point
of intersection of the vertical line for the constant kG = 0.019 and the horizontd line
for the diameter of the packing used, dp = 4 mm, we draw a line parallel to the in-
clined network. Prom the point, where this parallel intersects the vertical line from
emr= 0.12 we then proceed horizontally to the vertical scale for the limiting velocity,
where the required figure, 2.5 m/sec, is read off.
Obviously the column cannot be loaded to this limit in practice. A load of 60 to
70% of the calculated limiting velocity, say.'60% = 1.5 m/sec, may be taken t w
suitable.
If we modify the previous example by adding the requirement that the C7fatty
acid shall be distilled at 20 mm pressure with a load of 440 g/h, we can calculate the
necessary column cross section from the limiting velocit,y just determined, by using
the formula

f=
D - 22400 - (273 6) 760+ - 1 (144)
W D * 3 6 0 * 273 * p

in which f = the column cross sectional area in cm2; D = the amount of vapour in
g moles/h; 6 = the distillation temperature in "C;wD = the selected vapour velocity
in cmlsec; p = the distillation pressure, mni Hg.
This formda reduces to

17.3 X D X T
f= (145)
WlI XP

A t a throughput of D = 440 g/h = 3.4 g moles/h and wD = 150 cmlsec we find

17.3 x 3.4 x 399
f= = 7.9 cni2
150 x 20

from which the radius works out at

r = = 1.6 C I ~ . (147)
When the distillation is carried out a t norma, pressure, formula (145) becomes
0.0228 * D*T
f= cma.
WIJ

For comparison we shall calculate the limiting load for a column of 30 mm dia-
meter, in which n-heptane is distilled at normal pressure; 4 mm saddle packing ia
iranchembook.ir/edu

4. I1 Pressure drop, limiting velocity and calculation of column dimensions 181

employed. The limiting velocit,y for this packing is 0.38 m/sec. Then, by formula
(148~
f * WD 7.9 x 38
D= - = 35.5 g mole/h
0.0228 . T 0.0228 X 371
= 3 550 g/h = 5 200 ml/h . (149)
That such high loads can actually be realized with a column of 30 inn1 diameter is
confirmed by measurements carried out by Kolling [ 1711. The second calculation
above was included to show to what extent the use of reduced pressure increasm the
vapour velocity and thereby reduces the possible load.
Schumacher [231a] has introduced a n equation for the determination of the wet
pressure drop and divided it into two dimensionless coefficients for pressure drop and
gas flow. This allows the points of flooding onset, spraying and flooding to be character-
ized consistently for all packings wetted by falling films. A more recent dimensionless
irrigation coefficient makes it possible to calculate the admissible gas velocities not
only for the regions of intense gas load hut also for those of intense irrigation.
Met hods for calculating the limiting velocity for plate columns of laboratory size
have, as far as the author is aware, not been published. The formulae valid for in-
dustrial columns prove to be inapplicable to small-scale apparatus. Experience shows
that plate columns can be submitted to only about one third of the load that can he
applied to packed columns of the same diameter; this is due to the resistance resulting
from (a) the liquid present on the plates and (b) the r6strictions in the vapour pass-
ages.
The aerodynamic aspect of vapour flow in packed columns has been dealt with
by Barth [220], who demonstrated that packing having sharp angles gives rise to two
or three times the pressure drop caused by rounded bodies.
By considering flow, David [191] derived a method for calculating the theoretical
plate number of a packed colunm.
He assumed that since the pressure drop, velocity and density of the vapour
greatly influence the efficiency of a packed column the calculation should be based
on these parameters. By analogy with a well-known equation for the presslire drop
in pipes, he formulated a flow factor as follows:

in which Ap = the pressure drop (nim water colunm), g = the acceleration due to
gravity (m/sec2),d = the diameter of the packing units (m), wD = the vapour velo-
city, relative to the empty column cross section (m/sec), Q~ = the density of the
vapour (kg/m3),I = the height of the packing (m).
The quotient of the specific stage number and the flow factor was regarded as a
fundamental value, independent of the Reynolds niirrtber Re. This quotient proved to
be 0.13 for all mixtures evaluated by David, a t any pressure. As he assumed that the
greatest separating effect is obtained a t the upper limiting velocity, the expression

)acpt = 0.13E (151)

iranchembook.ir/edu

i8a 4. Physical fundamentais of the mpamtion process

would give the optimum efficiency of a lahratmy c d w u with s paaked height of

1Ill.
All experiniental data taken together yield an enipirical formula for the height
of packmg :

HK = 1 + +10 d
( n t h - 0.13 X 6 ) X -b). (152)
120
The formula refers to a reflux ratio v = bo, i.e., for a liquid-vapour ratio FID = 1.
Beck [244] evaluated 115 experiments on the s-ptem water-air and derived new
equations for packing sizes from 8 to 801nm and column diameters from 150 to
1200 mm. SavkoviE-StevanoviE, Simonovii. and PopoviE [244a] examined the effi-
ciency of a laboratory column (diameter, 25 mm; separating length, 1.2 m) by deter-
mining t8he HETP values for various pressure drops and vapour velocities. The
packing units were of 0.35mni steel wire with a diameter of 4mm. The system
benzenefcarbon tetrachloride (30 mol~,,carbon tetmchloride) was used as test
mixture.
Eq. (152s) yields HETP values as a function of the reflux rate. The mean de-
viation froni experimental data is as small as 1.7%:
In H E W = 29.0028 + 11.0310 In M - 0.9774 ln2M, (152a)
M = reflux rate (g/cm2 h),
HETP = height equivalent to a theoretical plate (cm).
The dependence of the HETP values on the vapour velocity was determined using
eq. (152h):
HETP zz 0.0689 + 0.1310~D- 0.3823wDZ, (162b)
WD = vapour velocity Irnls).
A coniparison of experimental and calculated data for the HETP shows that the
equation of Murch [1%] and eq. (152a) give good residts. In contrast, the equation
of Hands-Whitt [.J. Appl. Cheni. (London) 1 (1951)pp. 135-1401 gives large devia-
tions.

4.1 2 Heat calculations

Heat calculations are seldom carried out in connection with laboratory distilla-
tion, since the amounts of energy consumed are relat.ively small, and heat economy
is usually of minor importance. With electricity a higher voltage than necessary
is generally used and the current is regulated with resistances or auto-transformers.
Gas is still utilized in the laboratory for certain forms of distillation, such as the
ASTM and Engler procedures, and for heating oil, sand or metal baths. As a rule,
however, gas flames are avoided because of the dangers of breakage and superheating,
and of explosion in t,he distillation of inflammable substances. Heating by means of
electric resistances in the form of embedded wires or hot plates is at present the
prrfrrred method. Recently, infrared heaters with bright or dark elements have also
iranchembook.ir/edu

4.12 Heat calculations 183

become popular, particularly for low temperatures; they have the advantage of
giving radiation in a concentrated form [232]. The use of high-frequency heating in
the laboratory is still in the experimental stage.
In batch operation the initial mixture is first heated in the still pot to its boiling
point. A further supply of heat, corresponding to the heat of evaporation, converts
the liquid into vapour. The rate of evaporation depends on the amount of heat
supplied in unit time. In continuous distillation part of the heat is usually introduced
by the preheater. An additional quantity of heat is supplied in the still pot, since an
exchange of heat takes place in the column in countercurrent distillation.
If the total heat required is Q, the heat balance for the various types of distillation
(divegarding the heat necessary for bringing the charge to its boiling point) can he
forninlated as follows.

I . Simple distilldim
Q QK = QD f Q! Q K ~ + Qv. (153)

11. Countercurrent distillation

a ) hatch
Q = Qr; = Qo + + Qv
QH + Qv = Quo (154)
h ) continuous
Q = QK + Qz = Qo + + = + Qv
QR QV Q K ~ (155)
where Q = the heat supplied to the still pot; QH = the heat supplied to the preheater
in continuous distillation; Q D = the heat content of the distillate as vapour: Q R = the
hwt content of the reflux as vapour; Qv = the heat loss; QKo= the heat, removed
by the condenser for condensation and cooling.
These equations are valid only if the distillate and reflux are cooled to the same
temperature as the liquid present in the still pot. Pig. 110 shows these quantities
diagrainniat ically.

QKO

,QV

QV '

. I
D
111

IIa
Fig. 110
Heat supplied or withdrawn in I. Simple distillation, Ila) batch countercurrent
distillation, I1 b) continuous countercurrent distillation
iranchembook.ir/edu

184 4. Phyeiml fundsmentale of the separation process

As an illustration of the quantities of heat required in a laboratory distillation the

following example is given.
A mixture of benzene and toluene (50-50% wt.) is to be separated at normal
pressure with a rpflux ratio of 2: 1 a t a load of 0.6 kg/h, the charge being 2 kg (temp.
20°C; Table 36).
In the calculation it will be assumed that only benzene evaporates, though
wtudy, of couwe, a certain amount of toluene will also vaporize. As the heat of
evaporation of toluene is lower than that of benzene, however, the small error so made
is on the safe side. The heat loss is difficult to estimate and can be largely avoided by
proper insulation (see chap. 7.7). The question of how heat losses may be avoided
will be dimussed below.

Table 35
Data for the components

Component KP.:~~ Spec. heat, c p Heat of vaporization, AvH

(kc4kg) (kcal/kg)

Benzene 80.1 0.43 (5OOC) 93.8 (B.p. 760 mm)

Toluene 110.6 0.43 (60°C) 86.5 (109.7 "C)

Heat required to bring the mixture to 80°C

QK = B(.B, - 61)Cp QV ;
&R -
= 2(80 - 20) 0.43 = 51.6 kcal
Heat loss (say 10%) = 5.2 kcal
56.8 kcal
(B = charge, kg; 6,and 8, = initial and final temperatures)

€?eat require&to evaporate the distilhte and reflux

QK = ( E + R ) - & H + Qv;
Qa = (0.5 + 1.0) . 93.8 = 140.7 kcal/h
Heat loss (say look) = 14.0 kcal/h
154.7 kcalfh
The heat equivalent is taken to be:
1 kwh = 86Okcal;

the heating source must therefore supply

164.7/860 = 0.18 kW
for the intended operation.
iranchembook.ir/edu

4.12 Heat calculations 185

The method of calculating the necessary condenser surface area will be described
in section 7.4. The formula in question (191), can also be used to estimate the heating
surface of the heater required in continuous distillation, Q then being the amount of
heat supplied to the liquid, F the surface area and 46, the difference in temperature
between the heating agent and the liquid. The mean temperature difference is calcu-
lated after (199). Jn the laboratory, glassware is generally employed. With glass
the heat transfer coefficient is only 2.5 x 10-2 kcalfcm2 hdeg. C, but if round-bottomed
flasks are used, as will normally be the case, a sufficient area for heat transfer is
available at the customary speeds of evaporation. The same applies to condensers,
which will become overloaded only a t extremely high distillation rates.

Table 36
Heat loss from a column insulated with magnesia (85%)

Temperature Thickness of insulation

differential
1 inch 2 inches

Heat loss Condensate Heat loss Condensate

-16 Substance
(”C) cal/h cal/cm2.h ml/h cal/h cal/cm2.h ml/h

55 Benzene 4800 20.0 54 3450 14.3 42

85 Toluene 7600 31.6 101 6400 22.4 73
151 p-Cymene 13800 57.2 238 9800 40.7 170

For insulating the column against heat loss, substances of low heat conductivity
are chosen (section 7.7), for instance magnesia, asbestos, mineral wool and glass wool,
or vacuum jackets [153]. They are, of course, never completely effective, as exempli-
fied by the data in Table 36. These data were obtained with a column of one inch
diameter and a length of one foot. The insulation consisted of magnesia, in layers one
and two inches thick. The temperature differential was the boiling point at 760 mm
minus 25°C (average room temperature). The fourth and seventh columns show the
heat loss per unit area of the glass columns.
Junge [233] has carried out similar measurements on a column provided with a
vacuum jacket and on a column insulated with glass wool 50mm (2 inches) thick.
The heat losses found, in calories per cm2 per hour, are shown in Fig. 11 1 against the
inner column temperature. This figure also includes the data of Rose [153]ahead>*
listed in Table 36. The heat lost by an insulated distillation column is by no means
negligible. Jt gives rise to a “false reflux” that can be avoided only by incorporating
a heating element in the insulating jacket. It is necessary for the column to operate
adiabatically if reliable and reproducible results are to be obtained.
The heat losses Q of an insulated column can also be calculated by nieans of the
following formula:

(162)
iranchembook.ir/edu

186 4. Physical fundamentals of the separation process

in which A = the thermal conductivity of the insulation in kcal/m h "C; EZK = the - -
mlumn length in m; A 9 = the difference in temperature (column interior ambient) ;
d, = the external diameter of the insulation (m); d , = the colun~ndiamet,er (m) =
inner diameter of insulation.

'4
/=-
e <,A
10

-
' A
1 , I I I I I I I L , 1 1 1

60 80 100 120 NO 160 "C 180

temperature

Fig. I l l
Heat 108s of insulated columns in cal/cm%h as a function of the inside column
temperature
(a ) = Column insulated with magnesia (850/,), thickness one inch (25.4 mrn)*;
( X ) = Column insulated with magnesia (85%), thickness two inches*;
($7) = Column insulated with glass fibre, thickness two inches**;
( EI ) = Column insulated with a vacuum jacket**
* Rose [ 1531; ** Junge [233]

Example. A colrinm 25 IUM in diameter and 5 m long has been covered with
channel shaped sections of glass wool insulation ( 2 = 0.055 kcal/m h ."C)with a wall -
thickness of 62.5 mi^. The internal teniperatiire is 76.6OC, room temperature 18°C.
Reqnired: the heat loss. By formula (162):

0.055 X 6.28 X 5 X 58.6 = 56.5 kcal,h.

Q= 2.3 x log 6

Thus, despite insulation a heat loss of 56.5 kcallh arise8 which has to be offset by
an additional heater in order to avoid wall flow.
iranchembook.ir/edu

4.13 Distillate properties 187

4.1 3 Distillate properties and distillation diagrams

I n the practice of laboratory distillation it is customary t o report not on15 the
boiling point curve, but also subsidiary data for the distillate, such as the density,
the refractive index, the solidification or melting point and occasionally also valiies
such as the molecular weight and the iodine value (Fig. 112). Of late, infrared spectro-
scopy, mass spectrometry and gas chromatography have been used to an increasing
extent for identifying the fractions. The present trend is towards determinations
with continuous flow apparatus, so as to obtain continuous readings. The collection
of comprehensive data is particularly desirable in the case of an analytical distilla-
tion, where the characterisation of the distillate by boiling point alone is usually not
sufficient [234].
Constants for chemical compounds are to be found in the usual books of reference,
e.y. [235], but, t,he published figures should be considered critically. Considerable

destiilate 4

Fig. 112
Distillation cnrre of a gasoliiie with curves for distillate properties
iranchembook.ir/edu

188 4. Ph.piersl fundamentals of the separation process

differences often occur between the values reported by various authors, even in the
cme of well-known substances such as benzene and 1.2-dichloroethane. In general,
preference should be given to recent data, which are more likely to have been deter-
mined with sufficient accuracy. By plotting the constants for pure compounds in a
distillation diagram (Fig. a), identification is greatly facilitated.
Unfortunately it will frequently occur that the desired constants are not to be
found in the literature in the requiredform, and that they will need conversion to other
conditions or recalculation. The fundamental work by Reid and Sherwood [2S] deals
with the properties of gases and liquids and their calculation. Another exoellent
reference hook has been written by Hecht et a]. [236]. It contains a systematic survey
of the methods for calculating thermodynamic data of gases and liquids.

Pig. 113
Relation between the refractive indices at
various temperatures of the normal fatty
acids and the number of carbon atoms
C-atoms in n-fatty acids-

The refractive index is generally given for various temperatures in books of

reference, so that interpolation is possible If the refractive indices of a homologous
series of compounds, at a series of temperatures, are plotted as a function of the
number of carbon atoms, one obtains, according to Stage [237], a family of hyper-
bolae. This property may be used for extrapolation (Fig. 113). It should be noted
that, on an average, a change in temperature of 1deg. C causes a change in the re-
fractive index of about 5 units in the 4th decimal place, and it is therefore necessary
t o niea.wre the temperature of observation accurately (if possible to within 0.1 deg. C).
Convenient noniograms for deriving various properties of materials are to be found
in publications of Kharhanda [238] and Orlicek and Poll [239].
When a distillation is perfornied at atmospheric pressure it is necessary to take
several readings of the barometric pressure. Variations of 20 mm in the reading can,
for instance in the case of benzene, give rise to differences in the boiling point of
1 deg. C. The barometer should be checked against a precision instrument and a
correction should be applied for the influence of temperature on the mercury column
and the scale. This can be done witshoutcalculation by means of a nomogram due to
Haiissler [ W ](Fig.114). It is based on the formula

bkorr = b . [I - ( h - 5,)’ - 61 torr (164)

iranchembook.ir/edu

4.13 Distillate properties 189

in which bko, = the corrected barometric pressure, mni Hg; b = the barometer
reading, mm Hg; a = the coefficient of expansion of mercury (0.OOO 182/deg. C) ;
/3 = the coefficient of expansion of the scale material (brass: O.OOOO19; glass:
0.000008/deg. C); 6 = tlhe air temperature in "C.
Glass thermometers (cf. section 8.2.1) are as a rule, but not always, calibrated
with the stem immersed in the constant temperature bath up to the top of the mer-
cury thread. In this case a correction must be applied if part of the thread is not a t

Pig. 114
Nomogram for barometer correction (Haussler)

the required temperature. For this purpose a second thermometer is suspended

alongside the projecting stem of the first, so as to indicate the average teniperature
of the emerging mercury thread. The observed temperature is then corrected by
means of the formula
79korr = 6 + k * a(8 - &)) (165)
in which Okorr = the corrected temperature; 6 = the observed temperature; k = the
apparent coefficient of expansion of mercury in the glass of which the thermometer is
made (generally about 0.00016); a = the length of the emerging mercury thread,
expressed in deg. C ; 6, = the average temperature of the emergent thread.
A nomogram due to Cole [241] for performing this correction is based on formula
(165) with k = 0.000156. It further enables the boiling point, in distillations at
atmospheric pressure, to be converted to a pressure of exactly 760 mix, whereby t,he
data reported become strictly comparable. The correction K,, of the boiling point, for
iranchembook.ir/edu

im 4. Physical fundamentals of the separation process

pressure is based on the formula

h'l = (b - 760) * d81d.p (166)
where dt?/dp is the pressure variation of the boiling point of the substance being
diatded. By a combination of formulae (166) and (166) the total correction, K , to
760 mni pressure becomes
-
K = [k X ~ ( 6 a,)] - [b - 7601 x d6/dp. (167)
Nomogram 9 (Fig. 115) is used as follows. The temperature difference 6 - 4,,is
marked on scale 9 and this point is joined to the scale reading for a. The point of
intersection of this line with scale C is then connected by a straight line to the baro-
metric pressure on scale D. The point where thia st.raight line intersects the line
corresponding to d8id.p for the substance in question gives the final correction in the
boiling point for 760 mm pressure and a completely immersed thermometer stem.
Example. In an atmospheric distillation, the top temperature suggests the
presence of phenol. The observed temperature PSI is 179.8"C.The emergent mercury
thread of the thermometer is 40 centgrade degrees long (a), and its average tempera-
ture, 99irn, is 30°C. The barometric pressure read off is 760 mm Hg;the value of dt?/dp
for phenol is 0.045. Point 150 (= 6 - 8,) on scale A is joined to point 40 on scale a ;

temperature
4.8 -c
--
correction, "C

4.6
4.4 --
4.2 --
4.0 --
3.8 --
3.6--
3.4--
3.2 --
3.0 --
2.8--
2.6
2.4
2.2
2.0
1.8--

emergent
stem correction barometer
reading

Fig. 115
Nomogram for thermometer correction (Cole)
iranchembook.ir/edu

4.13 Distillate properties 191

this line intersects scale C a t 0.95. This is the correction for the emergent thread onlj-.
If we now connect this point to 750 on scale D, we find that the line for d6fdp = 0.0$5
is intersected a t a point corresponding to +1.4deg. C. The boiling point, fully
corrected t o 760 mm pressure and for emergent stem is then 179.8 1.4 = 181.2"C. +
A similar procedure for determining boiling points at 760 torr was evolved hy
Reckhard [57]. He obtained group data for 400 organic substances. The procedure i\
also applicable to multicomponent mixtures.
I t is iiseful to employ printed fornis for entering the results of distillations. Such
f o r m make the sixhseqiient graphical representation much easier.

22

:I4
c
'p 12
U
Lc 10
E 8
n
€6
z4
2 1 1 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I
I I I I I I I I I I I I l i l l l l l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 41 46 48 50 52%
distillate +
Fig. 116
Step diagram of the distillation of a crude fatty acid mixture

A distillation curve is obtained by plotting the amount of the distillate against the
vapour temperature, with the temperature either as the horizontal scale (Fig. 118)
or the vertical scale (Figs. 84, 112). Whilst the first type is still employed to some
extent in industry, the second type is generally used. It is advisable to draw the
diagrams to standard dimensions. The temperature scale may with advantage he
inade larger than the quantity scale, in order that sniall differences in temperatiire
may show up. If there is no danger 'of loss of clarity, all observed values should be
included in the diagram, for instance the still pot temperature and the distillation
pressure. If certain components have been identified from the results, the fractions in
question should be marked off by heavy vertical lines. An example is shown in Fig. 84,
which illustrat,es the fractional distillation of normal fatty acids a t 20 nini pressure.
For binary mixtures a satisfactory form of diagram is also obtained by plotting the
coinposition of the distillate on one axis against the amount distilled on the other
(Fig. 90).
For evaluating the results of a distillation performed on a mixture of homologous
conipounds it is useful to construct a n idealized step diagram (Fig. 116) [242].
From the latter, a diagram of the distribution of the individual compounds
may he derived (Fig. 117). A somewhat similar type of figure is that constriicted
according to the method of Meier-Grolman and Wesolofsky [243], which provides a
iranchembook.ir/edu

192 4. Physical fundamentals of the separation process

Fig. 117
Distribution diagram showing
amount of constant-boilingfractions
against the number of C-atoms, for
R crude fatty acid mixture

number of C-atoms

Fig. 118
Boiling point diagram (a)and
surface diagram (b) for a

a)
temperature - I)
shale oil
iranchembook.ir/edu

4.14 Instructions for the calculation of distillation conditions 193

iiieatlii of visualizing distillations (particularly those carried out by the ASTM, Engler
ant1 other standard procedures) on multicomponent mixtures, such as fuels, tars or
+ale oils. I n this method the diagram is drawn by plotting the temperature along the
y-axis, and the percentages that have passed over between intervals of 10 deg. C are
plotted in both directions along the x-axis. I n this way there result “root” or “onion-
shaped” figures, having very characteristic and distinctive forms. On the left-hand side
of such a diagram a column of figures is included, indicating the total percentages
passing over up to various temperatures, on the righthand side a similar column
showing the percentages obtained between certain temperature limits. In the case of
crude oils, for instance, 180°C can be taken as the limit of the gasoline and 300°C for
kc~osine.Fig. 118 gives a comparison between this type of diagram and the boiling
point curve of a shale oil.

Fig. 119
O~S, Surface diagram for (a) diesel fuel and
(b) special diesel fuel

The advantages of such surface diagrams are :

1. the amount of residue is seen at a glance;
2. the onset of decomposition is shown by a swelling out of the surface at high
temperatures (320°C or more) ;
3 . various distillations may be compared rapidly by superimposing corresponding
(transparent) diagrams.
A s a further illustration of this form of graphical representation Fig. 119 gives the
diagrams for a diesel fuel (a) and a special diesel fuel (b).

4.1 4 Instructions for the calculation of distillation conditions

The plan of work indicated in Table 4 will be illustrated by means of an exaniple
showing the methods of calculation employed.

4.14.1 Batch distillation at atmospheric pressure

I n a certain laboratory a considerable quantity (about 10 litres per week) of a
benzene-n-heptane mixture is obtained as waste product, and this is to be separated
into its components. The distillate is to contain 99 molyo of benzene, the residue not
more than 2 molyo.
13 Krell, Handbook
iranchembook.ir/edu

194 4. Physical fundament& of the separation process

Data for component8

Component Amount in Mol. KP.760 Density Q Separation

inixt ure wt. at factor LY
(mol?/o) "C 20 "C

Benzene 60 78.11 80.2 0.87895

n-Heptane 10 100.20 98.42 0.68368 1.77

1. Eatintation of the nuniber of theoretical stages required

At atmospheric pressure the difference in boiling point is 18.3 deg. C. Arough
estimate, assuming ideal behaviour of the mixture, indicates that there will be
required
(by the diagram in Fig. 72) 9 theoretical stages,
(by formula (100)) 16 theoretical stages.
However, as will appear in paragraph 3, the mixture is not ideal and the number
of theoretical stages needed will therefore be higher. The following example shows
that only a strict calculation yields reliable values and that the difference in b o w
point provides no more than a rough approximation.

2. Vapour pressure curva

As i t has been decided, for convenience, to carry out the separation at atmospheric
pressure, the optimum pressure for distillation need not be gone into.
It should be noted, however, that the boiling point should be corrected to 760 mm
(see Figs. 114, 115) and that temperatures should be corrected for emergent stem.

3. calculcction of the number of separating stages und the reflux ratio

The equilibrium curve of the system has been published (see Table 29, chap. 4.10.3)
and shows an asymptotic approach to the diagonal in the neighbourhood of 100 mole/,.
Taking the values of chapter 4.10.3 as basis, we now draw the equilibrium curve
to a R i z e of 50 x 60 cm on a mole fraction basis.
As described in section 4.7.1 we then proceed to determine the conditions for
distillation graphically. The operating line (xB= 60 molyo and xE = 99 mol%) cuts
the vertical axis at y = 32. From formula (82) we find that

We decide on a reflux ratio of v = 9, which by (79)g'ives us

99
y= - = 9.9.
10
The operating line connecting y = 9.9 and xE = 99.0 for v = 9 is employed for
drawing in the steps between this line and the equilibrium curve. We find that' 25
theoretical stages are needed in the initial conditions.
iranchembook.ir/edu

4.14 Instructions for the calculation of distillation conditions 195

The final conditions required are a distillate of xE = 99 molyo and a residue

concentration of 2 molyo. The operating line corresponding to these conditions cuts
the ordinate a t y = 3, so that now
99
2),in = - - 1 = 32.
3
We estimate that the reflux rat,io selected again requires 25 theoretical st>agesand
choose a value of v = 50, which gives iis

The points y = 1.94 and .c., = 99.0 are connected. Between this new operating
line and the equilibrium curve the steps are again drawn in. As expected, they
indicate ?z = 25 theoretical stages.
The conditions for distillation so determined are thus :
Required number of theoretical stages nth = 25
Reflux ratio, initial v= 9
Reflux, ratio, final v = 50

4. Load, distillate take-off and r u n n i n g - i n t i m e

First we must convert molecular percentages into percentages by volume, (see

section 3.4, formula (15)):
78.11 x 0.6
0.879
benzene, vol o//o -
- = 48 voly0.
78.11 x 0.6 100.3 x 0.4
0.879 + 0.684
dboiit half the still pot contents will therefore remain as residue. The operating
hold-up, estimated a t 100 ml, w ill be taken off as intermediate fraction and added
to the charge in the next distillation, since such a small quantity cannot be separated
adequately in the same still.
If we require a n hourly distillate take-off of 300 nil, the column must he loaded to
the extent of 2000 ml/h, with a reflux ratio of 9 in the initial conditions. The flow of
reflux is then 1800 ml/h.
At a constant column load of 2 litres/h, the distillate take-off a t the end, where
2 000
the reflixs ratio is to be 50, would ai~iount~ to only -= approx. 40 ml/h. Were
51
we to keep the the take-off rate at 200 nil/h, the column load would finally have to
be 200 x 51 = 10,200 ml/h. This would clearly be impossible in practice, since the
col~inni,with a fivefold increase in load, would no longer provide the 25 theoretical
stages needed for the separation. We therefore need to maintain the load constant
and to increase the reflux ratio from 9 t o 50 in accordance with the reduction in the
amount of volatile coniponent in the still pot.

13*
iranchembook.ir/edu

196 4. Physical fundamentale of the separation process

The complete calculation can be performed using an equation derived by Billet

[135]. Since this requires extensive preliminary calculations the reader is referred to
the original paper.
We will now estimate the equilibration time, i.e. the time that must elapse before
the distillate has attained a concentration of 99 mole/,. (See section 4.8.3). For this
purpose we have to know the operating hold-up. If this is not t o be determined
exactly (section 4.10.5), it may be taken to be about 3 1111 per theoretical stage of a
packed column. We shall msume the figure to be 75 ml.
2000
By formula (129) the intensity of countercurrent exchange (R= -= 0.55d/
sec) can be calculated: 3 600
26 x 0.55
I = = 0.18 sec-l.
75
If we introduce this figure int,o equation (130), we find the approximate equili-
bration time to be
Ma
t A= -= 1877 sec = c a 30 min.
0.18 x 2
The summary of rate8 and times is: Column load 2 litreslh; distillate take-off at
outset,, with v = 9.200 d / h ; equilibration time 30 min, assuming adequate flask
heating.
5. Calculation of column dimemions
I n order to calculate the column diameter needed at a load of 2 litres/h we must
first know the limiting velocity for the packing employed. We have decided to use
4 mm porcelain saddles. The method of calculation is fully described in section 4.11.
The reduced vapour density is

& = -@D
=- 27 3.07 x
pF 878.95
in which, by formula (37)
78.11 x 760
el) = x 0.01605 = 2.7 kg/ma.
353.35
With the aid of the nomogram of Egg.109, using a value of k, = 0.019 for saddle
packing, we find that the limiting velocity in the conditions concerned is 0.49 m/sec.

wo = /-
This value may also be obtained by means of formula (143):
0.019 x 0.004 x 9.81 = 0.49 inlsec.
0.00307
Taking the limiting velocity a t 6OyL of this figure, i.e. 30 cm/sec, we find the
required cross-sectional area of the column, by formula (148) to be
0.0228 x 22.5 x 353.35
f= = 6.0 cm2 at a value
30
of D = 2000 ml/h = 1758 g/h = 22.5 mol/h.
iranchembook.ir/edu

4.14 Instructions for the calculation of distillation conditions 197

We decide on a column diameter of 30 mm (f 7.06 em2) so that we obtain a

specific load of 285 ml/cm2 h. A t this load the HETP with 4 mm saddle packing can
he taken to be about 8 cni (see Table 31).
Hence we require a column having the following dimensions:
Inside diameter 30 mm; effective length 2000 mm.
Total length, including ground joints, 2150 nini.
For the heat calculations the reader is referred to section 4.12, where examples
have been given. The amount of heat necessary t o bring the charge to the boiling
point is found by formula (156), that for evaporation of the distillate and reflux by
formula (158). The required condenser surface area may be calculated by nieans of
formula (198).
The heat that iuust be supplied t o compensate for l o s ~ e smay be calculated with
thc ~ i t of
l formula (162). An example in chap. 4.12 illustrates its use.

6. Testing the coluniii

The coluinn cannot be tested (section 4.10.8) with the mixture in question. Even
if a charge having a concentration of only 2 mole/, were employed, only 18 theoretical
stages would be needed to enrich thc distillate to 99 niolyo with L, = 00. The test
might he performed with the mixture n-heptane-met hylcyclohexane at a load of
2 l/h. If, under these conditions, 25 theoretical stages were not attained, the column
would have to be lengthened. An HETP of 8 cm, as quoted above for 4 nini saddles at
a specific load of 285 ml/cm2 h, was only an extrapolation.

7 . Aritrrqement of the appuratus

The apparatus must be adequately insiilated and the coluniri jacket provided
with the requisite heat compensation (section 7.7.3). The iise of an automatic column
head (section 7.5.3) allows the distillation to he performed with a minimum of atten-
tion. I n the present case it would be useful to place a contact thermometer, connected
to a bell via a relay (section 8.2.2),in the column head. The therinometer would he set
to the boiling point of benzene, and when a signal was heard, the initial reflux ratio
of 9 would be increased, say, to 20. Jn the same way it would be progressively raised
to the final value of 50. The load may consenientl? bc regulated by means of a cnn-
tact manometer actuated by the pressure differential (section 8.4.2). I n accordanctk
with the amount of the charge the still pot may he a pilot-plant flask (section 7.6.1)
or a round-bottomed flask of 4 to 10 litres capacity, heated hp a flask heater (section
7.7.1).
For further automation the reader is referred to chapter 5, for the construction
of the apparatus to chapter 9.

4.1 4.2 Continuous d isti I lation

The specialmethodsof calculation nsedfor continuous distillation are dealt within :
Section 4.7.2 Exainple: Separation of a C6-C7 fatty acid mixture containing
81 mo1°6 of C,, a t a pressure of 20 mm Hg, xE = 96 ~nolO/~,
.?:, = 0.5 mol%.
iranchembook.ir/edu

198 4. PhysVSicsl fundamentah of the separation process

Section 4.9 Example: Separation of a multicomponent mixture of normal fatty

acids, C,-C,, at 20 mm pressure.
The apparatus for continuous distillation and its execution are described in
section 5.2.2.

4.14.3 Vacuum distillation

The calculations necessary for vacuuni distillation are carried out along lines
corresponding to those for distillations a t atmospheric pressure. The following addi-
tional points should he observed.
1. Choice of the optimum distillation pressure (section 4.6.2; 4.10.6).
3. Calculation of the equilibrium curve for ideal and non-ideal uiixtures a t a given
pressure (section 4.6.1, Table 10).
3. Dependence of the volume of vapour on the pressure (section 4.11).
4. Basic principles, calculation of tube diameters and testing for vacuuiii tightness
(section 5.4.1).
5. Measurement and control of pressure (section 8.3).

4.1 5 Distillation calculations by computer

If the use of electronic compiiters is advantageous on the laboratory and pilot-

plant scale it, is a must in industrial dktillation with its complex control problems.
Nowadays the whole separating process in a multistage plant is controlled by a
process coinputer. A comparison of the essential criteria of the various process coni-
puters used for laboratory autoniation is given by Blatt and Fleissner [259].
We distinguish between digital [246] and analog computers [247] and the combi-
nation of the two in hybrid computational systems [248]. The digital computer sup-
plies the data in tables of figures whereas the analog machine represents the results
as timedependent voltages which may be recorded in various forms. It would lead
U R too far to go into details here: the reader is referred to books on the subject t2.18.
2491.
All calculations occurring in distillation practice may be carried out on electronic
computers. The determination of vapour pressure and phase equilibrium data as well
as the calculation of optimum conditions for the plate number and the reflux ratio
tnay in certain cases he done even with desk calculators. To save time in the investi-
gation of the dynamic behavionr of packed columns, for example, the mole fractions
were derived from density and refractive index values using a n IBM 7040 computer
[ 1621. Kunierous programmes have been elahorated for the thermodynamics of the
phases [109]. An extensive Fortran progranime serves to calculate the phase equiiih-
rium of multicomponent niixtures [89]. According t o Roth [250] differential evapora-
tion (Engler and ASTM methods, see chap. 7.2), which is mostly used t o control
industrial processes, involves complex therniodpaiuic conditions. On the basis of
a model recalculations can he done with a ZRA 1 computer. A programme for an
iranchembook.ir/edu

4.15 Distillation calculations by computer 199

1BX 705 machine based on the McCabe-Thiele method was worked out by Schub-
ring [251] for binary mixtures. It may be used t o calculate plate numbers for ideal,
non-ideal and non-ideal azeotropic binary mixtures on the assumptions (which arc
frequently sufficient in practice) that the molar heat of vaporization is independent
of concentration and that no heat of solution occurs. The equilibrium data and the
concentrations of feed, top and bottom are fed into the programme through punched
cards. The computing times range from 15 seconds to 5 minutes for one problem. For
each problem the machine prints a series of possible plate numbers as a function of
reflux ratio or re-evaporation ratio. The block diagram of Pig. 120 illustrates the> corn-
puting process and Fig. I21 shows the diagram for the continuous separation of a binary
system in terms of plate number vs. reflux or re-evaporation ratio. The representation
of the relation between vR and vD in a nomogram enables the corresponding values
for the upper and the lower section of the column to be plotted in one diagram [251].
Chien [251a] presents a method for calculating the minimum reflux ratio on a digital
computer, with an exact description of the requirements for the calculation.
The calculation of batch distillations of multicomponent mixtures is particularly
difficult. Rose [251 b] developed a computer programme which simulates the dynamic
behaviour of discontinuous distillation processes.
Sealey [252] used an English Electric KDF 7 coinpnter to optimize the total
efficiency of an Oldershaw laboratory plate column (cf. Fig. 268) with a diameter of
31.8 mm and 12 and 10 actual plates in the lower and upper sections, respectively.
He analyzed 280 cases using methyl cyclohexane-toluene as test mixture. The opti-
miim plate number was defined as the one which minimizes the errors. These sources
of error were investigated:
- measurements of mixture composition
- position of operating lines
- slope of operating lines
- equilibriim data.

Fig. 122 shows the dependence of the total deviation in per cent on the plate
number over a range of interior reflux ratios from 1.0 to 1.3 in the lower column
section. The curves for the upper column section have similar shapes. The minima of
the curves increasing with rising v decrease from 1.0 to 0.7.
The mathematical model of a semitechnical rectification column with stationary
operation was elaborated and experimentally tested by GroBhennig [267]. The tests
were made in a n NW 310 bubble-cap column with 35 plates using the system chloro-
form-carbon tetrachloride. A comparison showed that even the simplified version of
the model gave satisfactory agreement.
The use of computers is imperative for the exact calculation of distillation
coluinns run for the separation of multicomponent mixtures [264]. Possibilities of
modelling a primary distillation of crude oil were discussed by Kirbach, Frenzel and
Strankmiiller [265]. The computing programme as prepared for a ZRA 1 computer
for ideal multicomponent mixtures was published by Leibnitz and Schuhler [253].
The calculating procedure for non-ideal multicomponent mixtures was dealt with
by Nagel et al. [254]. Morozova and Platonov [266] developed a method for esaniin-
iranchembook.ir/edu

200 4. Physical fnndamenfsla of the sepamtion process

read in problem

t 1
line o f points
of intersection computed all parameters ?

ref lux and

re-evaporation ratios

compute plate number

o f uppen coiumn pa of lower column part

add one to second

reached first reached second

limitation ? limitation ?

I I

Fig. 120
Block diagram after Schubring [251]
Process of determining the number of theoretical plate8 on an electronic com-
puter of the IBM 705 type

Fig. 121
Data for a cont,inuousdiatillation [251] nit,h
3 = 1.5; xz = 85 mol%; z~= 99.9 mol%; z.1 = 0.1 molY/u
iranchembook.ir/edu

4.15 DistiIla.t,ioncalculations by computer 20 1

ing the various possibilities of separating multicomponent mixtures with azeotropcs.

The method allows the influence of pressure changes and additives t o be determined.
-1programme for plate-to-plate calculations with columns for complex mist ures
(crude oil and petrol mixtures) without circulation and with the addition of stripping
steam was elaborated by Gohrke and Kleeniann [273].

I4

I
I

73 t v = 7.00
I

1.10

't.,
0 2
I
'I-
I
6 8
I I

10
I
'2
'
I&
qurnber of theoretical stages, nth 4
I
16 18
1
20

Fig. 122
Total error for calculated number of stages in the lower part a t various ii1nc.i
reflux ratios v (methylcyclohexane-toluene)

The progress made in modelling the distillation of coitiplex mixtures with pseudo-
cLoniponents has been reviewed by WeiB and Kirbach [374] who in their turn linl-c.
presented inore advanced methods of calculation.
.Analog computers may be employed in many ways in process engineering for t h v
calculation of material and heat transfer proceqses [255].In the last 10 years chivfli
those methods have been further developed which can only be used by nieaw of
electronic computers [6]. These are, above all, thc so-called matrix methods as evolvrd
by Wang and Henke [256],Sargent and Murtagh [257] and Stainthorp et al. [258].
iranchembook.ir/edu

202 4. Physical fundamenfsla'of the aepration process

Together with studies of flow resistance and material and heat transfer in packing
layers at rest Brauer and Mewea [260] workedout Fortran programmea for the calcu-
lation of presslire drop and packing layer height, taking into account the structure of
the packing.
['sing the system dichloroethane-toluene as an example Wagner and Blass [262]
describe a procedure for a hybrid amlog computer which simulates stepwise recti-
fication. This allows the determination of the still head or still pot concentration in
the separation of binary mixtures for given operating parameters and equilibriuiii
curve^ as a function of time.
The theory and practice of column dynamics for large-scale plants are discussed
hj. Kohler and Schober [26.1].
iranchembook.ir/edu

5. Separating processes

When the preliminary calculations for a separation have been carried out. the
next step is to consider the actnal procedure by which the separation is to he per-
formed. The method chosen will need to be adapted to the quantity of substance, to
the desired throughput and to the properties required of the distillate. Besides, thtb
properties of the mixtures to be separated as well as the apparatus and the energy
sources available have to be taken into account.

5.1 The scale of operation

The diniensions of the apparatus to be used for the distillation must obviously he
snited to the amount of the charge. It would, of course, be possible in an emergency
to distil large quantities in a small apparatus with a great expenditure of time, but on
the other hand too large a n apparatus would lead to excessive losses and erroneous
results.
I n the laboratory cases may occur in which the distillation of a few milligrauis of
inaterialis necessaq-andamicro-technique has tobe adopted, whilst at the other extreme
it may be essential to work with a throughput of 5-10 kg/h, so that the operation
falls into the serni-technical range. Pilot plant distillations involving scaling prohleins
have increasingly been carried out in recent, years.

5.1 .I Micro- and semi-micro-distillation

I n the first place a definition of the quantities involved will be given. I n preparative
micro-operations the amounts of substance handled often lie below 0.1 g. Since,
however, the physical properties of liquids give rise to considerahle difficulties in the
distillation of such small quantities, a distillation with a charge up to 5 in1 can be
considered as a microprocedure, and one with a charge between 5 and 50 nil as a semi-
micro process. Furthermore, we must distinguish between simple unfractionated
evaporation or distillation on the one hand and countercurrent distillation on the
other. Stage and Gemmelrer [l] point out that micro-distillation procedures have by
no means been replaced by gas chromatography, in particular by preparative gas
chromatography. On the contrary, it has been found that in most cases a preliminary
distillation of the mixtures is necessary in order to make full use of a gas-chromato-
graphic technique. An excellent survey of the appropriate types of colunin is given
and constructional details and devices used are described. A detailed description of
modern distillation equipment has been given by Stage [ 1 a].
For the simple distillation of a liquid in a n amount of a few niilligrains up to 1 in1
iranchembook.ir/edu

204 5. Separating processes

?

distillation capillaries, bulb tubes and small flasks are employed. Columns with a
rery small hold-up - empty columns, columns with concentric tubes (annular columns)
and spinning hand columns - are utilized for quantities of 1.0-25 ml, whilst charges
of 25-50 ml can be distilled with semi-microapparatus provided with packed
cornruns.
Bodenheinier [2] has described apparatus with which milligram quantities of
material can be distilled and boiling points can be determined with an accuracy of
1-2deg. C (Fig. 123). The end of tube a, which has been drawn out to an internal

Fig. 123
Bodenheimer’s mirro-distillat.ionapparatus
a = Condensation tube, b = Capillary containing
sample, c = Glass ponder, d = Heating block with
b c a two thermometers

Fig. 124
Horizontal micro-distillingdevice (Jantzen)

diameter of 1 mm, is inserted in a heating block d. Two thermometers are provided

for temperature measurement. The sample of liquid is placed in the capillary at b and
the reniaining volume of the capillary within the block is filled with glass powder. The
portion of tube a emerging from the block hae been etched internally so a s to give it a
matt sirface. The temperature of the heating block is raised at a steady rate of
4-6 deg. C per minute. When the boiling point8is approached a film of liquid is seen
to form on the cold part of tube a, and at the boiling point the glass becomes tmns-
parent. From the length of the transparent section an estimate can be made of the
amount of distillate, on the strength of preliminary experiments. In order to obtain
further fractions, one can again warm up the first deposit, so that another region of
condensation is formed further along tube a, which is cooled with wet filter paper.
By means of the horizontal distillation device of Jantzen (Fig. 124) it is possible
to carry out a countercurrent distillation with 1-100mg of substance in a tube of
120 ern length and 5.5-5.7 mm I.D. [3]. In the heating cylinder a temperature gra-
dient of about 30 deg. C is produced; a current of hydrogen assists the transport of
material in the sense of this gradient. As a result, the low-boilingcomponents migrate
fnrther than those with a high-boilingpoint. The device is intended for substances of
iranchembook.ir/edu

5.1 The scale of operation 905

low volatility. Fig. 125 shows the separation that it gave with a mixture of dodecanoic
and tetradecanoic acids. The most, recent, autoniatic version of the device is described
by Fischer [3a].
Quantities of 0.05 -0.5 nil are frequently distilled in a small bulb blown at the end
of a tube. The method of filling is to heat the bulb (similar t o Fig. 126) gentlj- and
allow it to suck up the liquid during cooling, or to employ a capillarl- pipette. The use
of bulb tubes has been extensively treated by Bertele and Humbel [3b]. Even

t;
._ Fig. 125
-
4-
Separation of a mixture of

.
;5-ll-
40
44
dodecanoic and tetradecanoic acids
in Jantzen's horizontal distillation
apparatus
5

Fig. 126
Babcock's micro-distillat,ion apparatus for
use a t reduced pressures

vacuum distillations may be performed on a micro-scale. Fig. 126 shows an apparatus

developed by Babcock [a],that gives three fractions a t a,b and c. To collect several
fractions without interrupting the distillation, micro-equipment is provided with
rotating vacuum receivers (Pig. 127). Any number of fractions can be taken with the
apparatus of Krell (Fig. 128). This small device may be employed both for atnio-
spheric and reduced pressures and can be adapted to the size of the charge. If the
latter is only 1-2 ml it is best to work with a n empty riser tube. If necessary, one or
two gauze cones may be placed as spray traps above the flask, and these also give a
certain rectifying effect (see Fig. 273). For distilling a charge of 3-5 d,glass
Raschig rings or 4 x 4 mm helices may be placed on a glass grid in the riser; if the
charge is 5- 10 i d , 2 x 2 min helices, supported by a gauze cone, can be used.
Distillation is started with the fine-control valve closed until a constant temperature
iranchembook.ir/edu

208 5. Separating processes

is established, after which arbitrary, measurable loads and reflux ratios can be
applied with the aid of thedropping capillariesa. These provisions make it possible to
carry out reproducible distillations with small amounts of material.
In contrast to an empty column (chap. 7.3.1), the Vigreux column has a con-
siderable surface area and leads the reflux to the centre by means of the indentationa
in the walls. Shrader and Ritzer [5) equipped their micro-distillation apparatus with
such a column (Fig. 129). Other special features of this equipment are the flat-bot-
tomed f h k a, which ensures a constant surface area for evaporation, and the receiver b
containing tubes c of 0.1 ml capacity. The distillate is led to the tube in question by
the glass thread d. The apparatus can, however, be employed only for high-boiliig

Fig. 127 Fig. 128

Micro-distillation apparatus, Micro-distillation apparatus of lire11
model 00100 for atmospheric and reduced
distilling flask: 5 ml, fraction collec- pressures; charge 1-10 ml
tors: 6 X 6 ml or 9 x 2 ml
iranchembook.ir/edu

6.1 The scale of operation 207

materials, as there is no condenser. The larger apparatus due to Klenk [6] is also
equipped for condensation by air cooling only. Its column, 8.5 or 13 cm long, which
contains a spiral of metal tape and is surrounded by a vacuum jacket, appears
capable of effecting sharp separations with a low hold-up (Fig. 130).
Both packed and plate columns can only partly be used for semi-micro-distillations
since their hold-ups are too large (see chap. 4.10.5). The point is to keep the column
dimensions as small as possible for a given amount of material. Only helices of I .A to
2 inn1 size can then be used as packing (chap. 7.3.2, 7.3.3 and 7.8). Concentric tube
(chap. 7.3.1) and spinning band columns (chap. 7.3.5), however, are particularly
suitable for sharp separations of very small amounts. Both types are also of great

Fig. 129 Fig.

" 130
1'
Xicro-distillation apparatus of Klenk's apparatus for micro-
Shrader and Ritzer with distillation, with spiral column,
Vigreux column for charges of 4 nil or more

importance in analytical and micro-scale distillations of high-boiling substances in

vacuum [7].
Winters and Dinerstein [8] have dcqcribed a n assenibl>*of 8 senii-micro c o l u u ~ ~ i ~ ,
which are used for analytical distillations and are provided with a central measuring
point. The assembly contains spinning band columns, Hyper-Cal columns bj- Pod-
bielniak and concentric tube columns with characteristics as shown in Table 37. The
separation of essential oils with spinning band columns a t 4 to 13 torr has been re-
ported by Spiegelberg 191. Examples of separations of high-molecular organic coin-
pounds with a concentric tube column are given by Jantzen and Witgert [31 and
Fischer [7].
Molecular distillation (section 5.4.4) on a micro-scale inav be perfornied by both
the cold-finger method and that of the falling film or thin layer. A practical example
of a cold-finger apparatus is illustrated in Fig. 131 [lo]. The mixture to be separated
is heated in A , and condenses on the cold finger B; the distillate huilds up a t C , froni
iranchembook.ir/edu

208 5. Separating processes

Table 37
Data of semi-micro columns used in analytical work [8]

Spinning Hyper-Cal Conceutric

band column tube
column column

inner diameter rn 111 5 8 8

length cm 90 60-90 60
capacity ml 10-50 20 -100 10-50
size of fractions ml 0.5 0.5-1.0 0.5
throughput ml/h 10-30 20-30 20-30
amount of take-off ml/h 0.5-1.5 0.5-1.5 0.2-0.5
reflux ratio 40: 1 40: 1 loo: 1
HETP cm 1.1. 0.7 0.9
hold-up per 90 cm ml 1.9 4.8 3.2

which it drips into the rotatable receiver D containing the tubes E . For quantities
smaller than 4 g Breger [ l l ] has described a useful arrangement (Pig. 132). The
evaporatmg surface is on the outside, the condensing surface on the inside of the
c>-lindricalspace. The apparatus, which has a length of only 18.5 cm, can be rotated
nhoiit its centre, so that the mixture inti?-first be degassed, and the original charge or
the distillate may if desired be repeatedly distilled as a falling film a t increasing
temperatures. With the apparatus of Paschke et al. [12] a molecular distillation of
abont 0.5 g can be performed. Fig. 133 shows the method of operation. The substance
to hr distilled is dripped onto a hall of glass wool a, which is suspended on a spring
balance in the tube. After h g h vacuuin has been established the heating jacket b is
put into operation; the quantity of distillate is read off on scale c.
The heating of small flasks presents difficulties in micro-distillations, and flooding
of the column often occurs. If possible the diameter of the latter should therefore he
not less than 6-8 mm. B micro-heating device, equipped with a dark infra-red lamp,

Fig. 131
bficro-form of molecular distillation apparatus according to Gould et al.
(cold-finger method)
(see accompany text)
iranchembook.ir/edu

5.1 The scale of operation 209

Fig. 132
Micro-form of molecular distillation apparatus after Breger
(falling film principle) for less than 4 g charge

Fig. 133
Apparatus for molecular distillation
on a micro-scale for 0.5 g charge
(Paschke e t al.)

14 Krell, Handbook
iranchembook.ir/edu

2 10 5. sepllratingprocesses

has given excellent results in practice. It also allows micro-distillations to be carried

out under reduced pressure, as bumping is avoided.
Preparative microprocedures including distillation apparatus have been surveyed
by Pfeil [13]. In this connexion mention should be made of Eigenberger’s micro-
density balance which allows densities of very small amounts of material to be deter-
mined relat,ivelg quickly. A simpliiied version has been built by Classen [15].

5.1.2 Analytical distillation

The object of carrying out an analytical distillation is to obtain the components
of a binary or multicomponent mixture in as pure a form as possible. Their purity is
determined by the measurement of suitable physical properties, such as the refractive
index, the density, the solidification or melting point end the molecular weight. As it
is frequently not known what constituents the mixture contains, or in what propor-
tions they are present, the distillation is usually carried out batchwise. A column with
an adequate number of separating stages, calculated as described in sections 4.7 to
4.12, is employed and it is provided with a column head that allows the reflux ratio
to be regulated precisely. For an accurate determination of the amounts of the com-
ponents it is necessary to keep the transition fractions as small as possible. A transition
fmtion is the distillate paasing over between two pure components and consisting of
a mixture of the two in changing proportion (Fig. 56). The size of the charge must be
chosen according to the smallest fraction that is to be isolated; for the transition
fractions to be small, the ratio of the hold-up to this smallest fraction must be suffi-
ciently low. The use of a high hold-up to improve the separation (cf. section 4.7.1)
need not be considered, since an analytical distillation is always performed with a
high reflux ratio. The book of Bock [16a] gives a good introduction intp analytical
distillation. The author points out that column distillation allows larger amounts
of material to be employed and thus appears to be especially suited for the enrich-
ment of trace impurities.
-4s a considerable number of separating stages will be required for an analysis and
as the hold-up increases with the number of stages, a narrow column should be
chosen for the purpose. Sccording to Stage [16] the smallest fraction to be iaolated
must be at least ten tinies as large as the hold-up. This means that if the hold-up is
say, 120 ml, and the smallest fraction to be obtained pure is present to the extent of
lo%, a charge of 12 1would be necessary. This probably applies only to very narrow-
boiling mistures, requiring a high reflux ratio, and where the purity to be attained is
nearly 100%. Thus, besides packed columns especially the efficient concentric tube
and spinniig band columns as well as columns with stationary inserts will be used for
analytical purposes on account of their very low hold-ups (chap. 7.3.1, 7.3.4, 7.3.5).
For mdticomponent mixturm, if a sufficient amount ia available, the best procc-
dure is first to separate the raw material into a number of fractions by flash distilla-
tiou (cf. section 5.4.2) or continuous countercurrent distillation (section 6.2.2). If
the cuts are made on the basis of a preliminary distillation the sharpness of this
continuous separation need not to be high. The fractions obtained are then further
submitted to analytical distillation. An advantage of this method is that the in-
fluence exerted by the hold-up is less nnfavourable.
iranchembook.ir/edu

5.1 The scale of operation 211

Notable examples of analytical distillations of multicomponent mixtures are the

separation of a lignite light oil by Kuhnhanss and co-workers [171, the investigations
of Profft and Buchmann [MI, who tried to obtain chinoline bases, and the distillative
separation of fatty amine mixtures by Furst and Schwetlick [19]. Destinorm apparatus
has been used, e.g., in the separation of natural amino acid mixtures by means of
the countercurrent distillation of the methyl esters of the respective acetyl amino
acids [20]and in the isolation of the components of lignite tar oils to establish an
element balance [21]. Gutwasser and M a e r [22] developed a special thm-film still for
the fractionation of spermaceti oil fatty acids a t 10-1 torr. Mair et al., for example,
have demonstrated that it, carefully planned combination of normal distillation,
azeotropic distillation and the methods of zone melting can lead to a thorough iso-
lation of the components of a petroleum fraction [23].
Unfortunately it often happens that insufficient mixture is available for the
isolation of minor components unless special precautions are taken. A column having
a very low hold-up, for instance one of the spinning band type (section 7.3.5) may then
be chosen. If a packed or plate column is employed, it will be necessary to increase the
reflux ratio when the transition fraction starts to pass over, so as to ensure that i t
will be as small as possible. Bubble cap columns (section 7.3.3) have SO far been used
relatively little for analytical distillations, though Oldershaw sieve plate columns
have been in use for some years. They have the advantage with respect to packed
columns of allowing samples to be taken from the upper plates, so that the concen-
tration gradient in the top part of the column may be determined. By doing this i t
is possible to start increasing the reflux ratio as soon as the concentration on, say,
the third plate from the top begins to change. Another method is to separate the
mixture as sharply as possible into a number of fractions in a large column, and
subsequently to redistil the cuts so made with a smaller column (having a lower hold-
up), to obtain the components in a purer state. For very difficult analytical recti-
fications, requiring 125 or more stages, concentric tube and parallel tube columns (cf,
section 7.3.1) or columns packed with a wire spiral or wire-gauze elements (section
7.3.4) are the most suitable. R E T P values of less than 5 mm are now obtained also
with concentric tube columns under vacuum conditions 171.
If a n effective type of packing, giving an HETP of a t most 2 cm, is not to hand,
the construction of a column with a large number of stages may present difficulties
of size. In a room 12 feet high the longest column that can be set up will be about 9feet
effective length. Good results have been achieved in such cases by placing two
columns in series, as shown in Fig. 134. This method has proved useful, e.g., in
analytical distillations of the bases of lignite tar [24]. The distillate from column I is
fed continuously a t a point above the boiler into column 11, which may be smaller in
size. The load and the reflux ratio in column I1 may be different from those in column
I. It should be checked that the quantity of liquid in the second boiler remains
approximately constant. This method of combining a batch and a continuous
procedure saves the time needed for carrying out a second distillation separately.
Whilst the contents of the first pot change in composition from beginning to end (say
from 50 to 1 molO,(o), the concentration of the liquid in still pot I1 is kept within as
close limits as possible (e.g. between 90 and 95 molyo of light component) by an

14*
iranchembook.ir/edu

212 6. Separating processes

increase in the refluxratio in column I, with the result that an almost stationary state
correapondmg to a continuous process prevails in colunin I1 (see chap. 5.2.2). The
higher-boiling component concentrates in the second still pot' and the transition
fraction may consequently be kept very small. After the supply from the first column
has been utopped, the liquid in still pot I1 can be worked up batchwise, or it may be
tested for its content of low-boiling component.
An important point in analytical distillation is to uae a still head having as low a
dead space possible. By dead space that volume of liquid is denoted which is
retained above the column in the still head instead of taking part further in t,he
oountemurrent process. The automatic column heads for the partition of vapour or

Fig. 134
Combined batch-continuous operation, by the arrangement in series of two Destinorm
columns
iranchembook.ir/edu

.5.1 The scale of operation dl3

liquid (section 7.5.3) have the important advantage of possessing almost no dead
space. The reader is reminded that the presence of sample taps may cause contamina-
tion of the distillate by grease, from which unexpectedly large errors may resiilt.
There is consequently a trend towards sampling devices in which the use of taps is
avoided (section 4.6.3). Of late automatic column heads (cf. section 7.5.3) h a w heerr
equipped with magnetic valves, which prove to give Satisfactory service.
In view of the precision denlanded of an analytical distillation it is neccssary to
examine which of the following two courses is the most favourable: to take off the
distillate as a constant fraction of the condensate, or to operate intermittently, i.e.
to return the condensate to the column entirely as reflux for a certain period and
nr,xt to remove it completely as distillate for another period. In the latter case th(x
wflnx ratio is the quotient of the periods of reflux return and of distillate with-
drawal.
150
OC

120

1
"s
90 Fig. 135
Distillation diagrams of a
solvent mixture
60
a = Engler distillation,
b = preliminary fractionat,ion,
30 c = sharp rectification
0 20 40 60 80 V O I % 100
m i x t u r e ---+

If the reflux ratio is low, the average sharpness of separation will always hc 1)oorer
with an interinittent than with a constant take-off. It has been found, however,
that with a reflux ratio of about 1OO:l the separation is better if the distillate is
taken off once an hour than if it is removed every 30 seconds [%I. Jf the reflux ratio
is still higher (e.y. 1000: 1 to 10,OOO:l), as in the fractionation of isotopes, inter-
mittent operation will in any case be necessary for technical reasons (chali. 5.1.1).
Boiling point analysis may be carried out with simple apparatus if the coiuponents
differ considerably in volatility or if a standard record of the boiling range is all that
is required. Apparatus for such purposes will be described in chapter 7.2. An exact
evaluation of the distillation curve, however, becomes difficult, as appears from
Fig. 135, in which the analysis of a solvent mixture is taken as an exaniple. It should
he noted in particular that the initial boiling point is very different in the three forms
of distillation chosen. This is a result of the differing degrees of separation in the
three distillations ; the sharp rectification shows initially the boiling point of the most
volatile component, 37 "C; the Engler distillation cwve starts with the boiling point
of a poorly fractionated mixture which naturally has a higher boiling point, 59". the
preliminary fractionation is intermediate in separating power and in initial boiling
point, 45". A s the distillations continue, the sharp rectification reflects closely the
composition of the mixture, and the distillation ciirve of this mixture is irregular.
iranchembook.ir/edu

214 6. Sepereting proceaees

The Ehgler distillation averagea out the boiling points of a sequence of fractions and
the o w e is therefore smoother.
The considerable expenditure of time for a sharp analytical distillation (e.g.,
120 hours for the separation of a petroleum fraction with boiling points in the range
from -30 to +26OoC) gave rise to the development of a simulating technique ueing
a special gas-chromatographic appamtus [26]. It gives data comparable to those of a
column with about 100 theoretical stages and can be employed for analyses of both
crude oils and petroleum fractions in the range of boding temperatures from C, to
C,,,. The Aerograph Simultaneous Still provides boiling point-mass per cent curves.
The areas under the gaa-chromatographic curves are continuously integrated by
computer and recorded as measuring points a t 10 second intervals. The time spent
on an analysis of the above-mentioned fraction (-30 to +26OoC) amounts to about
1 hour [27].

5.1.3 Preparative and production distillation

If an analytical distillation has shown that a mixture contains certain compo-
nents, the next step may be to isolate these components in larger amounts. If the
differences in their boiling points are large and the purity of the products need not be
high, simple stills that can be constructed from standard parts will be adequate.
F ig.136a) illustrates an apparatus of this type that can be equipped with a column
if desired (Fig.136b). Equipment of the kind shown in these two figures, however,
operates with an uncontrollable load and “fake reflux”. For reproducible work it is
necessary to use 8 column head permitting the reflux ratio to be adjusted accurately
(cf. section 7.5). The Anschutz-Thiele receiver, shown in Fig. 136, though intended for
work in vacuum, has also been found convenient for distillation at atmospheric

Fig. 138
Apperetns from standard parta for: a) simple distillation, b) rectified distillation
iranchembook.ir/edu

5.1 The scale of operation 315

pressure. Besides, high throughputs allow the techniques of partial condensation

(chap. 5.2.3) and flash distillation (chap. 5.4.2) to be employed.
Due to the use of standardized ground joints laboratory columns are comnionll-
inanufactured up to a ground joint size NS 45. Semi-technical glass columns provided
with ball, cone and plane joints have nominal widths up to about 150 min. In pilot
plants columns may be made of glass, ceramics or metal, questions of corrosion
possibly determining the choice of material. The nominal diameters are in the sangr
100 to 400 mm. A survey of apparatus used for preparative separations on a laliora-
tory and semi-technical scale is given by Stage [28].
The preparation of ultra-pure solvents has been thbroughly dealt with by Hanipel
[29]. He points out that besides extraction, distillation is still oneof the most effective
niethods of purifying solvents up t o > 99.9%. With alcohols, ether, benzene, pyridine
and a number of hydrocarbons extremely high degrees of purity can be attaincd. c q . .
c; l 0-4% absorbing impurities in carbon tetrachloride.

5.1.3.1 Semi-technical columns

The dimensions of the apparatus used should be in accordance with the reqnireri
rate of throughput and separating effect (section 4.11). Ground joint components are
conimercial1;v available in sizes NS 10, NS 14.5, NS 29 and NS 45. Arrangenients for
continuous operation, permitting high distillation rates, are dealt with in sections
5.2.2,5.4.2and 5.4.5. For distilling quantities up to 301/h the semitechnical Destinonn
apparatus, which is made with columns of 50, 70 and 90 mm diameter, mas developed.
It is designed for work both a t atmospheric pressure and reduced pressures down to
1 nim Hg and is largelp automatic. Fig. 137 shows a semi-technical apparatus for
continuous distdlation, put together from standard parts. I t allows various arrange-
ments of full-scale plant to be reproduced. The column is connected by means of
conical joints; th~rmometersand tubing carrying liquids are held by NS 14.5 joints,
whilst flexibility is obtained in critical places by the use of 35 min ball joints (chap.
3.1). The column head (Fig. 137) is provided with a multi-channel tap for accurate
control of the quantity of distillate; the reflux ratio may be measured and adjusted
with the aid of graduated burettes for the reflux and distillate. A condensing coil
built into the vapour tube above the column allows the operation to be performed with
dephlegmation (partial condensation). The amount of dephlegniate can be c*alciilated
from the rise in temperature of the cooling water or measured in a graduated receircr.
The semi-technical colunin is also obtainable with liquid-dividing automatic reflnu
control (see Fig. 312). In continuous distilla.tion the feed is heated by a preheater,
constructed as a U-tube (Fig. 138). When tap a is partially closed the liquid level in
the preheater rises to the eonnect,ion b and the liquid circulates. By this means the
feed can he heated to over 250°C. Fig. 137 shows only the basic arrangemrnt which
may be extended to any degree of automation by adding measuring and control
devices (see chap. 8.).
For batch distillation round-bottomed glass flasks were as a rule used only np to
a .ize of 10 1; a larger volurne was distilled in a suspension flask (Fig. 316). or a nietal
itill pot with a flat or spherical ground closure, to which the glass apparatus was
iranchembook.ir/edu

216 6. Separating processes

14
I ’

Fig. 137
Semi-technical Destinorm
apparatus for throughputs up
to 20 I/h
iranchembook.ir/edu

5.1 The scale of operation 217

connected. A suitable vessel of 25 1 nominal content is a stainless steel still pot, which
is equipped with a heating jacket of 3 k W maximnm rating, controlled by a %heat
switch (chap. 7.7.2). An advantage of continuous operation is that even a t a throiigh-
put of 20 l/h the reboiler need have a capacity of only 2-5 1 (cf. section 7.6.1).
The plate column (with diameters up to 90 mm) in the author’s Destinorm appa-
ratus (Fig. 139), on the other hand, has internal reflux like industrial colunms, so
that distillations with this form of apparatus are comparable in this respect. The
various trays may be provided as required with sampling valves and thermometers,
in order that concentration and temperature gradients may be followed during the

6-

Fig. 138
Preheater for continuous distillation
a = tap, b = connecting tube

Fig. 139
Krell’s plate coluinn with internal reflux
(50, 70 and 90 mm diameter)
iranchembook.ir/edu

Frg. 140
Distillation plant 111 d e of Rasotherm glass, Irith bubble-cap column XIV 400
iranchembook.ir/edu

5.1 The scale of operation 219

operation. The Labodest vapour collision plate after Stage (chap. 7.3.3) a.lso operates
with internal reflux and is manufactured in sizes up t o 100 mm diameber. The depen-
dence of the plate efficiency on the handling of t,he reflux w ill be dealt, wit,h in
chap. 7.3.3.

5.1.3.2 Pilot plants

Glass has proved to be a reliable constructional material in pilot plants for batch
and continuous countercurrent distillation as wpll. It offers the basic advantage that
during operation the hydrodynamic processes going on in the column and the other
parks can be watched. In addition, the glass manufacturers offer a variety of compo-
nents with ball or socket shaped tube ends or with plane ground joints [165] by means
of which virtually all distillation procedures can be automated to a high extent. The
material commonly used is borosilicate glass which is highly resistant to chemical%
and can withstand both high temperatures and extreme temperature changes
(chap. 7.1). Further details will be exemplified by the glass components made by
VEB Jenaer Glaswerk Schott & Gen. of Jena. The material employed is Rasotherni
glass of type 320, TGL 7209 [30].
A wall thickness of 5 to 8 mm gives the glass apparatus a sufficient mechanical
strength. Due to it8 low expansion coefficient, (a = 3.3 x Rasotherni glass
display^ a high resistance to thermal changes. Pilot plants are mostly made up of
standard parts. Examples of such parts are tubes with their fittinrs, heat exchangers,
round-bottomed flasks and cylindrkal vessels, bubble-cap plates for nominal widths
(NW) of 200,400 and 500 mm (Fig. 140), column sections up t o hW 500 mm, stop-
cocks and valves, metering pumps and other components serving measuring and
control functions. Besides the standardized plants, which are manufactured serially,
such as water processing plants (distillation and ion exchange), circulation stills
operating a t reduced pressures with throughputs from 5 to 100 l/h (water) (Fig. 141)
or plants for the recovery of solvents, special components, such as automatic reflus
controllers and pneumatic glass valves, are also made. If a batch distillation requires
larger flasks than can be manufactured (the present limit is 200 1) glass columns may
be mounted on metal or ceramic flasks (chap. 7.7.2).
Closed tubing systems can be assembled without difficulty on the basis of a
system of modules consisting of straight lengths of tubing, fittings, transition pieces,
stop-cocks and valves [165]. The tube diameters are graded in the range from N W 15
to NW 600. The individual lengths (at present up to 3 m), are connected by means of
loose flanges, the ball and socket-shaped tube ends allowing a deflection up to 3‘.
There are various other kinds of fittings available for special uses (screwed and flanged
fittings for various kinds of tubes). The necessary precautions t o achieve more safety
in the operation of semi-technical glass apparatus have been reviewed by W’inliel-
sesser [30a]. I n particular, the possibility of armouring endangered parts and the
prevention of electrostatic charging are discussed. A paper by Pietzsch and Rauten-
berg [30b] on the uses of glass in the chemical industry gives information about the
chemical resistance, surface quality, mechanical and thermal strength of Ragotherin
glass and also deals with the question of armouring.
iranchembook.ir/edu

Fig. 141
Circulating-t,ypeapparatus for work at reduced pressures, evaporating capccity 100 I/h H,O
iranchembook.ir/edu

5.1 The scale of operation 22 1

The distillation plants shown in Pigs. 140 and 141 (heights up to 14.5 ni) arc
equipped with steam heating elements. The evaporating capacity for each PJJV 200 or
NW 300 element (flattened spirals) a t a saturated steam pressure of 2 atin. orer-
pressure is 20 l/h of water. An advantage is that the condensing and heating snrfa;cs

vacuum
main connection

reflux d,ivider

acuum

Fig. 142
Apparatus for continuous distillation a t reduced pressures, with nominal width
of column of 150 mm
Height: 7.2 m, width: 1.9 m
iranchembook.ir/edu

222 5. Sepereting prooesaes

can easily be extended. For heat exohanger calculations the heat transfer coeffioient
k (kcal/max h x “C) may be taken to be 400 when liquids are to be heated by steam.
For the condensation of v a p o m by m a n s of condensers and for the cooling of
liquids the respective values are k = 300 and k = 150.
Quickfit glass components, tubing systems and pilot plants are equally well
suited to operating a t reduced pressures. They contain elements with conical tube
ends of nominal widths N W 16 to 600 with plane ground joints and of NW 16 t o 225
with ball ground joints. The wall thickness ranges from 2 to 10 mm. AU Quickfit
components may be protected by a polyester coating which is heat resistant up to
150°C [165]. For vacuum operation in pilot and small-scale technical plants, Quickfit-
Wiegand steam ejectors of glass [128a] (see chapter 5.4.1) are now available.
There is also a tendency to manufacture standardized distillation apparatus, be-
ginning with circulation stills, including ah0 combinations of non-glass flasks and
glass columns, and ending up with automatic continuous pilot plants with electro-
magnetically operated reflux heads, level control and flow control devices (Fig. 142).
The maintenance-free operation of such plants requires numerous glass valves.
These will be described in detail in chap. 7.2.1. Further, controlled bottom and
feed heating is necessary for the automation of glass-made pilot plants (chap. 7.7.2).
The reflux division (chap. 8.4.1 and Fig. 142) and the evaporation rate (chap. 8.4.2)
must,also be controlled. For the measurement and metering of gas and liquid quanti-
ties special glass components for pilot plants were developed (chap. 8.6).
The combination of the constructional materials glass and PTm is the optimum
solution for metering devices for corrosive liquids (temperatures up to 100”C, kme-
matic viscosity up to 30 cSt and maximum densities of 2 g/cms). VEB Jenaer Glas-
werk Schott & Gen. of Jena developed a series of metering pump heads without
stuffing boxes (Fig. 143) which are designed to be mounted on the standardized
drive of type PAE 32/50 (VEB Pumpenwerk Salzwedel). The nominal capacities
range from 63 to 650 l/h a t maximum pressures between 3.5 and 2.5 kp/cma. The gas
flow valve N W 15 mm of the same manufacturer is also highly resistant to corrosion
(F’ig. 144). The valve chamber is made of glass, the diapbragm and diaphragm support
are of PTFE. The valve is closed by spindle lit.
Distillation apparatus made of hard porcelain is used for corroding substances
which also attack glass (Fig. 145). Here the smallest still pot volume, however, is
50 1, yet column tube diameters begin at 50 mm nominal width.
Podhielniak Inc., of Chicago, manufacture automatic pilot, distillation plants
(“Practioneer” series, Fig. 146) in four types, suitable for batch or continuous opera-
tion in the following ranges:
Still charge 45-450 litrea
Column diameter 76-304 lll~ll
Column length 1.5-3 m
Distillation pressure 3 mm Hg to 135 atm. abs.
Distillation temperature -40” to +350”C
The column packmg employed is usually of the Heli-pak or Octa-pak type (section 7.8).
Heating can be by electricity, steam or Dowtherm liquid according to choice.
iranchembook.ir/edu

5.1 The scale of operation 223

Fig. 143
Metering pump head of glass with Teflon bellows for combiriation with the
standardized drive of type PAE 32/50

75C1 77+1

.
1 c
b u i l t - in i e n g t h 150
LI

Fig. 144
Gas straight-way valve, NW 15 mm
iranchembook.ir/edu

22 5. Separating processes

Snitable components for distillation pilot plants are described by Carpenter and
HcAlwig [31]. Jordan, in his book on chemical pilot plants [32] devotes special atten-
tion t o the scaling up of results from pilot plant to industrial scale.
*\gain and again the question arises whether the results of pilot plant experiments
c.ould not be replaced by purely arithmetical scaling up. Since, however, there is
SI111 no inethod of precalculating thernial separations available, such experiments will

Fig. 145
Semi-technicalapparatus for batch distillation made of hard porcelain

have to continue to provide the data required to scale up separation processes [33].
This is quite obvious in the case of packed columns where the aerodynamic and
hvdrodynamic paratneters are of special importance (chap. 4.2). In addition, as
Billet [34] points out, the evaluation of separating columns and hence the niini-.
iiiization of costs involves the exact determination of t,he dependence of efficiency
and pressure drop on the load. Even today this relation has to be determined experi-
mentally for most columns. The same experimental conditions and systems should be
chosen so that different column types may be compared (see chap. 4.10 and 4.11).
iranchembook.ir/edu

6.1 The scale of operation 225

Fig. 146
Podbielniak automatic pilot distillation plant

5.1.3.3 The preparation of distilled water

9 typical instance of production distillation occurring in the laboratory is the
process of making distilled water. This is generally performed continuously. As a rule
a constant level is maintained in a water boiler, preheated cooling water is led to it
and the surplus is syphoned off continuously. By using quartz glass a high purity of
the distillate is ensured (Fig. 147). Moreover, borosilicate glass is now the preferred
material. It is attempted to prevent the water froin getting into contact with the
metal parts.
The multitude of types that are made indicates that the ideal still, from the aspect
of economy, has not yet been realized. To achieve better utilization of heat internal
16 Krell, Handbook
iranchembook.ir/edu

-lntus heater

still head

heat exchanger
levelling tube
overflow Fig. 147
Water still “ P ~ r a t ~ r ”
made
, of
head with bell quartz glass, with throughput of
and dropping tip 2 I/h of distilled water

fi‘
Fig. 148
Automatic water distillation apparatus made of Duran 50
Throughput: 10 to 70 I/h
iranchembook.ir/edu

5.1 The scale of operation 227

electric heating is often adopted and the apparatus is provided with efficient insula-
tion. The throughput of laboratory units usually lies in the range of 0.5 to 2 l/h. For
throughputs above 2 l/h water stills are also often made of copper, nickel-plated or
tinned internally. Such apparatus has the advantage of being unbreakable, but has
only a moderate heat, efficiency unless the insulation is very thorough.
An electrically heated water still made of Duran 50 was developed by Schott &
(+en. of Mainz (Fig. 148). The various designs have throughputs from 10 to 70 l/h
of distilled water with a conductivity < 0.5 $3 when they are fed with fully desalted

Fig. 149
Apparatus for the simultaneous produc-
tion of distilled and doubly distilled
water (Kullmann)

water. Apparatus for the single distillation of water with throughputs from 8 to
10 l/h is manufactured by VEB Jenaer Glaswerk Schott & Gen. of Jena. The heating
is by electricity or steam. For the electrically heated bidistilling apparatus (8 and
2-1 l/h) the immersion heater TH Q6 (standard power, 6 kVA) was developed.
An interesting form of water still (Fig. 3b) is one employing the “descending”
principle of distillation for the more economical utilization of heat. A heat efficiency of
9 5 O I O is said to be attainable. Kullniann’s [35] device operates in a similar way. It)
vields a highly degassed distillate. An additional unit allows the preparation of
doubly distilled water (Fig. 149).
Zellner’s “Bidestillator 1600” [36], Fig. 150 which has a throughput of 1.3-1.6 l/h,
functions very economically. Heat is supplied only in the first evaporating stage for
the production of monodistillate from tap water. The re-evaporation of this distillate
takes place a t the reduced pressure produced by a water jet pump placed before the
condenser, a t a temperature of 30”-40°C, solely by the heat of condensation of the
primary vapour. The water used by the pump first functions as cooling medium, and

1.5”
iranchembook.ir/edu

228 5. separating processes

a part of it is finally utilized as feed for the first evaporating stage. Thus, a degassed
bidietihta of extreme purity is produced without a residue with the energy supply
required for a single distillation and a minimum coneumption of tap water.
It seems quite convenient to purify the feed water in an ion exchange filter as is
done in the water stills manufactured hy Heraeus-Quanschmelze GmbH of Hamu,

Fig. 160
Zellner's water still 1600

whioh are made of a completely water-resistant quartz. The path of the vapour
bubbles is short and yet the liquid surface is large so that spraying of the boiling
water and a.ems01 formation are largely prevented [36a]. The hi-tridistillation appara-
tus designed by Cammenga et al. [36b] is capable of continuously producing high-
purity water of intrinsic conductivity. I n routine operation the still yields water of
k w 0.06 $3 cm-' (20").

5.1.4 The separation of isotopes

The chemical elements occurring in nature consist for the greater part of mix-
tures of isotopes. They may be regarded as mixtures, the individual isotopes represent-
ing pure components. Isotopes of one element occupy the same position in the
periodic system and have the same nuclear charge, but different masses. The carbon
isotopes laC and 13C, for instance, merely differ by having 6 and 7 neutrons in the
iranchembook.ir/edu

5.1 The scale of operation 229

nucleus, respectively. The atomic weights found for naturally occurring elements are
the weighted averages of the atomic weights of the isotopes present.
Isotopes were first discovered in radioactive substances, but the majority of non-
radioactive elements also contain a number of (stable) isotopes. Table 38 lists a
variety of isotopes, many of which can be obtained by distillation processes [37]. In
organic chemistry considerable use is made of "tracer compounds", usually con-
taining heavy isotopes of the elements C, H, 0 and N. Except for carbon, none of these
has a suitable radioactive isotope possessing the requisitely long half-life period, but
the stable isotopes ZH, W,1 8 0 and l5N are of great importance for chemical research
and €or biological, medical and geological investigations.
Elements having an odd nuclear charge in general consist of only one or two
isotopes with a n odd mass number; exceptions are H, Li, B and N. The element Hg,
on the other hand, comprises seven stable isotopes with mass numbers of 196 to

Table 38
Some stable isotopes, their relative differences in mas8 and relative
natural abundance

Symbol Atomic Mass Relative difference Relative natural

number number in mass (%) abundance

1 1 100 99.9844
2 0.0156
2 3 33.3 1.3 x 10-4
4 99.9999
5 10 10 18.83
11 81.17
6 12 8.35 98.9
13 1.1
7 14 7.15 99.62
15 0.365
8 16 6.25 99.757
17 5.9 0.039
18 12.5 0.204
10 20 5.0 90.51
21 4.75 0.28
22 10 9.21
16 32 3.13 95.06
33 3.03 0.74
34 5.9 4.18
36 0.016
17 35 5.7 75.4
37 24.6
18 36 5.55 0.307
38 5.25 0.060
40 11.11 99.633
iranchembook.ir/edu

230 5. Separating processes

2434, the relative differences in mass being of the order of 0.5%. Since all separating
methods for enriching and isolating isotopes depend on difference in mass, which in
turn gives rise to some difference in properties, the relative differences in mass provide
an indication of the separability. An equally iniportsnt matter, however, is the rela-
tive abundance of the isotopes in the starting material. Aniong the elements listed
in Table 38 the relative natural distribution is particularly suitable in the case of
loB--llB, mNe-=Ne and 3Wl-37Cl.
For the concentration or isolation of isotopes the iiiethods of diffusion, thermal
diffusion, electrolysis and exchange reactions are employed. An enrichment, can also
be attained with the aid of the mass spectrometer and the centrifuge. Distillation
procedures are used for obtaining *He, BH (D), lo& W,l5N, lSO, 22Ne,37Cland 36Ar.
An extensive survey of the production and use of stable isotopes including the special
prohlems of isotope distillation has been given by Brodaky [38] and by Wetzel [39].

Table 39
Vspour preRsure ratios of isotopic systems

HI-D, 18.65 3.6 ~481

3He--PHe 2.19 5.4 c481
llBCl,-lOBCI, 260 1.003 [52, 531
"BF3 -'OBF3 170 1.01 C531
12Cl60-13C160 68.2 1.010 ~481
12C160 - 13C110
68.2 1.007 [481
- IaC13CH4
12C2H4 120 1.0014 [541
169.3 1.0019
1.0022 1541
1.0006
W H , -I3CH, 97.56 1.0099 ~541
111.8 1.0070
14N2-l*xLSN
63.3 1.006 [481
14N - 1 4 p J l S N
71.5 1.OM [551
1.03 [551
1.04 ~501
160160 -i s g i n 0
63.14 1.O 1286 "W
74.05 1.00912
2ONe --22Ne 24.6 1.043 ~481
s6Ar- 83.8 1.006 [481
12aXe --'zXe 161.4 1.000 [481

= vepour pressures of compounds with light and he'cavy isotope, respec-

tively
iranchembook.ir/edu

5.1 The scale of operation 231

Of late a conibination of cheniical exchange and distillation, known as “exchange

distillation”, has become of increasing importance [40]. The analysis of stable isotopes
is done chiefly by mass spectrometry [39]. There are, however, several nonniass-
bpectrometric techniques used which have been described by Mercea [41]and Miiller
a n d Rilauersberger [42].
The concentrating of stable isotopes by distillation is particularly suitable for
gases and water (the latter as a source of deuterium and l80).The vapour pressure
ratios of such isotopic systems are given in Table 39. Except for the compound BCl,,
the 53-stems require a sitbstantial consumption of cooling agents for separation.
Ftirtlierniore, t,he nuinber of separating stages necessary will be in excess of 500
(except in the case of hydrogen and helium) if a reasonable concentration is to be
attained. Kuhn et al. [43] determined the separation factors of isotopic coinpounds
with boiling points up to 80°C by employing an extremely large nuniher of staseq.

5.1.4.1 Isotope separation by low-temperature countercurrent distillation

Low-temperature distillation (see section 5.3.1) has been used to separate H-I),
lOB-llB, 1zC-13C and the isotopes of the noble gases helium, neon and argon.
Due to the great difference in mass and the high vapour pressure ratio tjheobtain-
ing of deuterinm from the gaseous mixture H,-D, by low-temperature distillation
appears to be relatively straightforward from a theoretical point of view. In a n
extensive systeinatic investigation Sellers and Augood [44] examined the probleins
involved in the countercurrent distillation of the systems HD-n-H,, 1 6 0 1 8 0 -1W160
and 0,-N,. They used a bubble-cap column of 27 m height. Further series of
vyminients on the low-temperature distillation of H-D were performed by Tinmer-
hails et al. [45] who used a sieve-plate column of 150inni diameter and deter-
mined a number of separating parameters of the system. A small laboratory apparatus
for the countercurrent distillation of H2-HD mixtures is described by Weisser [46].
By now, a number of large-scale plants for the industrial production of heavy water
have come into operation [47].The difficulties, which are considerable, are associa-
ted with the apparatus and gas purification. Thus, at present isotope separation a t
the temperatures of liquid nitrogen and liquid air seem to be too expensive on a
laboratory scale. If, however, a distilling plant were connected with an oxygen produc-
ing plant having apparatus for the technical decomposition of air, 36Ar, 1 8 0 and l5N
could br enriched very economically [48, 491.The low-temperature distillation of NO
with the simultaneous separation of I5N and 1 8 0 [50], as well as that of CO for the
enrichment of 13C [51] appears to be quite promising. Based on these techniques
remitechnical processes were developed for enriching the carbon isotopes “X and
13C, the nitrogen isotopes 14Nand l6N and the oxygen isotopes 160,170and lSO [167].
Clusiiis and Meyer [48]enriched 15 1 of argon per day t o 0.60/, (as compared to
0.307°!0 in naturally occnrring Ar) in a low-temperature distillation using 130 theo-
retical stages, They employed a packed brass column of 12 mini i. d. and 3 in height.
The packing consisted of stainless-steel spirals of 2 x 2 miii size. The specially shaped
still pot with a volunle of 250 ml and the condenser which was cooled nith liquid
nitrogen are shown i n Fig. 151. The separation of the isotopes 22Neand W e and the
iranchembook.ir/edu

23.a 5. Separating processes

anri&ment of 21Xe by countercurrent distillation at 28 K are reported by Bewilogua

et al. [164].
Miihlenpfordt and collaborators realized an annual production of 4.5 kg of BFa
containhg 83% "B, using a semi-technical column of 12mm diameter and 12m
length filled with metal fings of 1.2 mm diameter (Fig. 162). Because of the lower
volatility of l0BF3 this column has the feed vessel for BF3 a t its upper end and the
tae-off for the enriched BF, at its lower end. The constancy of the liquid BF3 flow
to the upper end of the separating tube, which iR a requirement for the 8uccess of the
separation, wae attained by means of a speciallydesigned measuring device.

Pyrex copper

f u s e d - i n Pt
31 b)
Fig. 151
a) Distilling flask
b) C!onderiser used in the low-temperatiwe distillation for 3sAr enrichment

5.1.4.2 The preparation of D, and l80,

by countercurrent distillation of water
The special circmmtances of an isotope separation will be illustrated further by
describing the preparation of D, and 180, by the distillation of natural water a t
n o d and reduced pressures.
Hydrogen has three isotopes: 'H, hydrogen, ?H or D, deuterium, and SH, tritium.
The last-mentioned, unstable isotope is almost completely absent from natural
compounds. Oxygen occurs in the form of the three isotopes leO, l7O and 1 8 0 , so that,
if we disregard tritium, we can formulate water in the nine ways shown in Table 40.
Of t,hese, we are interested only in the compound known a8 heavy water, DZ1eO
(aHzleO)and the compound lHz18O.
The physical properties of D,O and HzO are compared in Table 41 [37,66].
The difference in boiling point is 1.42 deg. C. Mass spectrometry and the difference
in density with respect to natural water are most frequently used for determining the
concentration.
iranchembook.ir/edu

5.1 The scale of operation 433

Fig. 152
Distillation column for'the:production of highly enriched lOBF, (Miihlenpfordt)
I = Evaporator with electric heater, IIi= Column for separating SiF, and HF,
I I I = Separation tube and packing, I V = Double condenser for BF,, V = Feed
vessel for liquid BF,, with electric heater. V I = Ethylene condenser, VKI =
Rubber bladder for pressure regulation, I'III = Storage vessel for liquid air.
IS = Auxiliary condenser
1 = To diffusion pump, 2 = To diffusion pump, 4 = Draw-off for impurities,
-1 = Take-off for highly enriched 'OBF,, 5 = Connection to manometer, 6' =
Connection t o manometer, 7 = Take-off for depleted RF,, S = Inlet for BF, feed.
9 = Feed line for d l y l e n e and hydrogen, 10 = Hydrogen gas buffer
iranchembook.ir/edu

2.34 5. Separating processes

Table 40
Theoretically possible isotope forms of water

Table 41
Physical properties of D,O and H,O

Property D,O H,O Difference

Density e4*" glcm? 1.1059 0.998 2 0.107 7

Melting point "C 3.82 0.00 3.82
Boiling point "C 101.42 100.00 1.42
Temp. of max. density "C 11.6 4.0 7.6
Viscosity CP 1.260 1.009 0.251
Surface tension dynesled 67.80 72.76 -4.95
Refractive index nDaO 1.32844 1.33300 -0.00456

The initial material for producing heavy water is natural water, which contains
0.0146atoniic % of D (ratio 1:6850). Rain water has a sbghtly higher content,
0.0200 at. (ratio 1:6000)[56]. From the point of view of distillation the main con-
stituents of natural water are H,O as low-boiling component and D,O as high-
hoiling component, but the compound H D O must also be taken into account, SO that
thv separation factor a* between D and H is given by

__
- XHDO + 2xDI0 2YHI0 f YHDO
(168)
ZXH,O f tHDO YHDO + 22/D,O

=\wording to Urey [57], a* can he expressed in terms of the vapour pressure ratio of
the two components as

= 1/1)*H,OIp*D,0. (169)
The latter formula aasumes that the solution is an ideal one, that the vapour pressure
of HDO is the geometric mean of the vapour pressures of H,O and D,O and that the
equilibrium constant of the reaction
H,O + D,O 2 2HDO (170)
in the x-apour phase has a value of 4.0. In Fig. 153 the values of
by Kirschenbauni [56] have been plotted against the temperature.
fxo given

I n theory, therefore, the smallest number of stages would he needed if the separa-
tion were carried out at as low a pressure as possible, and at a correspondingly low
ttwtperature. These conditions, however, would render condensation difficult and
iranchembook.ir/edu

5.1 The scale of operation 235

would also liniit the throughput owing to the large volume of the vapour. Large
pressure differentials and velocities would then result, the time required would be-
come extremely long and the procedure would be uneconomical. As is so often t he case
in counterciirrent distillation, a compromise must be made ; as a rule, therefore, the
operation is performed a t 50 to 125 mm Hg pressure. The value of a* then lies i n the
neighbourhood of 1.05-1.06 and the temperature between 40 and 60°C.

Fig. 153
Values of ip&,o/pg,o as a
l.00L""" '
10 20 30 40 50 60 70 80 90 100
temperature -
I " '
'c ' 0
function of temperature

In this case it is not possible to derive the required number of separating stilgtls
by the McCabe-Thiele method because, firstly, the equilibrium curve is too flat and,
secondly, extreme concentrations are involved, as shown by the following esaiirple :
zo = initial concentration = 0.000 146 D,O
(atomic fraction) :
re = final concentration = 0.998 D,O (atomic fraction)
= 1.069 (at 40°C and 55.3 mni Hg).
LY*

With these figures, the Fcnske equation gives the following (very approxinrnte)
value for the ininimunr number of stages required (i.e. at total reflnx):

nmin = log
xe(l - X o )
r o ( 1 - 5) I log a* = 308.

The number ?f theoretical stages required a t the optininni take-off rate will he
about 700 [58]. It should be noted that in this case the desired component is the high-
boiling compound D,O, which in batch operation concentrates in the still pot and in
continuous operation is drawn off as bottom product. The overhead product is
"stripped" water containing less than 0.0146 atomic 7; of D,O.
iranchembook.ir/edu

236 5. Separating proceases

According to Kuhn [59] it is convenient to express the separating effect for auch
ideal isotopic systems occurring with a single partial evaporation of a mixture of
liquids as
Y*
1 - y*
-- - e*.
20

1 - xo
The quantity b is characteristic of the binary mixture at a given temperature no
matter how large an initial concentration is chosen (e.g., a mole fraction of 0.99 or
0.5 or even etc.). I t is called separation parameter and can he calculated from
the vapour pressure values provided that the mixture in quest,ionis athermal, which
implies that the heat of mixing is zero:
p*1 - ed (173)
P*s
where p,. = vapour pressure of the pure low-boiling component,
y*s = vapour pressure of the pure high-boiling component,
or, if the relative difference of the saturation pressures is small:

b = P*l - P*s

l'*S
6=3-1. (175)
If the mixtures are athermal, as is mostly the case with isotopes, approximate values
of 6 can he calcula.ted from the formula
AT
S = 10.7 -,
TP
where AT = boiling point difference of the two components at normal presmre and
T p = boiling point of the low-boiling component a t distillation pressure p.
Another requirement is that "routon's rule applies to both components and that
the temperature dependence of t,he heat of vaporization is nearly the same for the
components. Table 12 shows values of 6 thus obtained.

Table 42
Separation parameters 6 of isotopic systems

Isotope mixture Temp. Difference in Separation Separation

boiling point, 4.8 parameter factor
"C "C I3 c(

H%O/D*O 70 1.6 0,05 1.05

&lnO/H,lsO 70 0.165 0.005 1.006
'*CC14/"CCI, 35 0.036 0.001 25 1.001 25
CaaCl,/C"Cl, 35 .- 0.003 o.Oo0 10 1.00010
iranchembook.ir/edu

5.1 The scale of operation 237

The multiplication of the elementary separating effect in countercurrent columns

with a given number of theoretical stages nthleads t o an enrichment according to the
relation
Xe

with xo = initial concentration, x, = final concentration. Hence, t,he number of

theoretical stages required for a particular separation which could bring about a
change in concentration from xo t o x, is

n - - In- Xe -1n-l. 5
(178)
0 - " 6 1-& 1 - xo
This expression corresponds t o the well-known Fenske equation (108b) for v = 00.
To obtain a particular separating effect a minimum reflux ratio is required which
to a first approximation is larger than or a t least as large as
2 1 f -zo
Omin = - =- (179)
t 6 xo(l - x,)
I n the preparation of a concentrate of heavy isotope as the bottom product, the mini-
mum reflux ratio (179) is replaced by the maximum bottom take-off ratio :

where Z = amount circulating (reflux) and t = amount taken off (distillate).

With a n amount of take-off calculated according to eq. (180) an infinit,e number
of stages would be necessary. The number of stages nBrequired wit,h a finite reflux or
take-off ratio may be calculated in the manner described below.
The number of stages n, necessary a t a given take-off ratio z1.Z is given by

2, - X" t
.- .L}]. (181)
Fig. 154 shows the number of theoretical stages as a function of the take-off
ratio in the preparation of the stable isotope lSO. The three curves represent the
following cases (all figures are mole fractions) :

Curve I xo = 0.002 6 = 0.0045

xe = 0.015
Curve I1 x, = 0.002 6 = 0.0045
n, = 0.020

Curve I11 X, = 0.002 6 = 0.0065.

5, = 0.020
iranchembook.ir/edu

238 5. Separating processes

Taking, for example, curve 111, we see that an enrichment from 0.20/, to 2% of
I8O is the possible under the following circnmstances:
at a take-off ratio t/Z = 10-5, with n = 388 theoretical stages;
at a take-off ratio t / Z = 10-4, with 71 = 409 theoretical stages;
at a takeoff ratio T / Z = 2 . with n = 437 theoretical stages;
at a take-off ratio r / z = 5 . with 71. = 567 theoretical stages.
Further calculating iiiethods based on nornograins are described by Huher [61] and
Hilthrunner e t al. [62].

the take-off ratio t / Z (K&)

P
140 20 %

initial concentration -
Fig. 155
Number of stages required
at v = mfor preparing D,O.
Pressure 760 torr (100°C)
and 100 torr (BOcC)
xo = 0.1-100 rnol%;
n, = 99.8 rnol?;
iranchembook.ir/edu

5.1 The scale of operation 2 39

In their preparation of D,O Kuhn and collaborators [60] began by electrolysing

natural water to a D,O content of l.Omol%, and then proceeded by distillation.
Jn industrial operation the opposite route is also followed. Fig. 155 shows the nniilher
of stages necessary with an infinite reflux ratio at 50°C (= lOOmmHg) and at
1OO'C (=760 mm Hg) for various initial concentrations and 2, = 99.8 mol9, D,O.
The material balance in Kuhn's two-stage operation is represented in Fig. 156, which
shows that the take-off ratio in the first stage is 1: 1270, in the second stage i:'i8.
The realization of these take-off ratios gives rise to some practical difficulty on a
laboratory scale. Using Kuhn's apparatus [59] (Fig. 157) as a guide, we shall now

2% qo .L\
0.1 '10 DZO

40llday
*
3m>I
-u.
m

lo/. DzO -
0,
column 1
effective length 530cm
H.E.T.P. 1.8 cm
head pressure 120mmHg
i
o
0
2
v o%

~
column 2
effective length 680cm
H.E.T.P. 1.7cm
pressure 6 0 m m Hg

Fig. 157
Diagram of Kuhn's multi-tube
column for prepararing D,O
iranchembook.ir/edu

240 5. Separating processes

discuss the operation more fd.lF. The initial mixture is present in the storage vessel
1; it is evaporated and passes through the superheated capillary tubes 2 at a constant
mte into the columns, which are filled with wire-gauze packing. The amount of
vapour is controlled by the difference in pressure between points 3 and 4 by means of
manometer 5. The lower part of the columns (zone 6 ) is heated by water vapour from
7 so that the total amount of descending liquid evaporates. The hase of the columns is
constricted by capillary tubes, through which only an amount of vapour correspond-
ing to the take-off ratio is drawn into chamber 8 by a pressure difference. Valve 9
controls the quantity of vapour thus removed; it is condensed and collected in 10.
The vapour ascending in the columns leaves them a t 4 and is totally condensed in 11.
The condensate collecting in 12 is pumped back into vessel 3, from which a controlled
amount, exactly corresponding to that introduced a t 2, leaves the column through

1.014

1012

4 1.010

1 1.008

1.006

1004

1.002
1.000
10 20 30 CO 50 60 70 80 90 100 ~ O ° C 1 2 0
temperature
Fig. 138
Vapour pressure ratio a of the system H,1@O/H280against temperature
according to Dostrovsky and Raviv

the barometric overflow 14. If reduced pressure is employed, vacuum is connected to

13 and to the bottoms receiver 10.
With this apparatus the throughput of starting material amounted to 1.5 kg/h.
The operation was performed in two stages: in the first, D,O was concentrated from
1.0 to SO molo;b, in the second from 50 to 99.8 mol%. The load was kept at about 10%
below the flooding point. With the tubes packed with wire-gauze Raschig (Dixon)
rings, the HETP amounted to 1.7 to 1.8 cm. The use of packing was found to be
necessary, because empty tubes proved to be wetted irregularly on account of the
high s u d m tension of water. The production of a uniform liquid film on the packing
IS the crucial problem in the preparation of heavy water (see chap. 4.2). The present
rtate of the large-sale production of heavy water has been reviewed by Lumb [59a].
The normal distribution of the isotopes of oxygen is approximately l 6 0 :170:1 8 0
= 2500: 1:6. During the distdation of natural water **Oconcentrates in the bottoms.
A s can be seen from Fig. 158, however, the value of B for H216O/H,l*Olies a factor of
iranchembook.ir/edu

5.1 The scale of operation 241

ten lower than for H,O/D,O. At 100°C (760mm) LX is about 1.004 and at 50°C
(100 mm) about 1.007. The data presented in Fig. 158 were collected by Dostrovskyv
and Raviv [53]. The relatively strong scattering of the values is due to the experi-
mental difficulties involved in the determination of a. It is a favourable point,
however, that natural water contains 0.204 at. yo of l80,a relatively high concentra-
tion.
Kuhn, with his column, reached a concentration of 6 at. o/o of 1 8 0 . Dostrovsky and
his collaborators a t the Weiszmann Institute in Israel [53], prepared ' 8 0 of about
95:,, purity and 170of about, 2y0. The preliminary concentration from 0.2 to 1.67; of
l*O was carried out in 10 columns connected in parallel (diam. 100 mm). The multi-
stage cascade (diam. 80-100 mm) was fed a t 800 ml/h and was operated with a take-
off ratio of 1.37 x The procedure is to be improved so as t o produce a concentra-
tion of 99.8%. The highest concentration of 1 7 0 is present a t the centre of the cascade
( a 10%). I n these separations the Dixon packing, made of 100 mesh phosphor bronze
wire gauze in the shape of a Raschig ring with an S-shaped bridge, proved its value
~31.
The writer, using a single-stage continuous column (Fig. 160)) attained a con-
centration of 5.25% lSO. The experiments performed with this unit at 300mm
pressure indicated that the most probable value of a a t this pressure (76°C)is 1.0068
[64]. Uvarov et al. [54, 651 effected an enrichment from 3 to 24.5% of l*O in a column
of 9.5 m effective length and 52 mm diameter, filled with triangular wire spirals
(2.0x 1.6 mni) and operated batchwise a t normal pressure. Their HETP amounted
to 1 cm. The high enrichment of the oxygen isotopes by countercurrent distillation
of heavy water was studied by Staschewski [70].
Fig. 159 shows a continuous apparatus made of copper. The packing consists of
1.6 mm ( l / l S inch) Dixon gauze rings. The columns are extensively automatized;
in an auxiliary unit the bottom product enriched with l80is electrolyzed. The 180z
gas is recombined with hydrogen, which is also obtained by electrolysis. The final
products are therefore H2lSOand D,W.
Whilst the time for the establishment of a stationary condition in normal distilla-
tions varies from a few hours to a t most 24, it may amount to weeks or months in the
fractionation of isotopes. Since the older formulae for calculating this time proved
unreliable, the problem was studied anew. Kuhn and collaborators [59] evolved the
following equation for the equilibration time Tbof parallel-tube and packed columns :

(182)

with H = hold-up, x, = initial concentration, Z = feed rate, xe = final concen-

t ration.
Further contributions to this subject have been made by Jackson and Pigford
[67], Rosen [68] and Brodsky [38]. The books of London and Cohen [69] deal with the
whole theory of isotope separation. This includes the optimization of cascade arrange-
ments which is of great importance in isotope separation as well as in the separating
plants of the chemical industry [71].
16 Krell, Handbook
iranchembook.ir/edu

242 5. Separating processes

Fig. 159
&fetal still for continuous preparation of & I 8 0
1 = Water supply, 2 = Distilled water apparatus, 3 = Flow meter, 4 = Con-
stant level device, 5 = Feed line, 6 = Overflow, 7 = Column, 8 = Boiler,
9 = Product line, 10 = 1-inch copper pipe between boiler and column, 11 =
Reflux cooler, 12 & 13 = Main condensers, 14 = Stock tank, 15 = Drain,
16 = Boiler manometer, 17 = Differential manometer, 18 = Manost.at, 19 =
Magnetic valve, 20 = To pump, 21-23 = Piping

5.1.4.3 The preparation of various isotopes

I n addition to the foregoing, some examples may be given of cases in which other
isotopes have been concentrated by distillation.
The relative difference in mass and the relative natural abundance are particularly
favourable for separating the isotopes of carbon and chlorine :

relative difference in mass 8.35%

3w1 75.4%
relative difference in mass 5.7%.
37ci 24.604
Both pairs of isotopes are separated in the rectification of carbon tetrachloride.
Using a parallel-tube column having about 250 effective stages, Kuhn [69] succeeded
in concentrating the lSC by about 1% of the original value. (Concentrations were
determined by mass spectrometry). An unexpected and interesting result was the
observation that heavy carbon (IT)accumulates in the low-boiling overhead fraction,
with a7Clon t,he other hand in the high-boiling fraction. A similar effect is found with
the boron isotopes loB-I1B. Baertschi and Kuhn [72] found that the infrared contri-
iranchembook.ir/edu

3.1 The scale of operation 243

Fig. 160
The author’s still for enriching H,lYO

16*
iranchembook.ir/edu

2.44 5. Separating proceases

bution to the dispersion interaction may be considered to be the cause of the greater
volatility of the heavy molecules.
In the distillation of cc1, a double isotope aepration thus takes place. The
separation factor of t,hesystem W--"C is about 1.00013,corresponding to a difference
in boiling point of approximately 0.036 deg. C (Table 42). The effect referred to above
- i. e. a concentration of W in the overhead produot - also occurs in the recti-
fication of chloroform, benzene and methanol. When chloroform ia distilled, 37Cl
accumulates in the bottoms, and lW concentrates in the bottoms when methanol is
fractionated. Data for such distillations [43]have been colleoted in Table 43.

Table 43
Single arid double isotope concentration by countercurrent distillation
nefl = effective number of stages
p i , p s = vapotu preesnre of the substance with the light (1) and the heavy (p) isotopes a t the di-
stillation temperature (34.6OC)
zB1 = concentration a t the foot of the column
zE1 = concentration a t the columo heed
T, = boiling point of the substance containing the light isotope
T, = boiling point of the substance containing the heavy isotope
Substance neff Isotope separation of
~~ ~~~~ ~ ~ ~ ~

W (overhead) $'Cl (in flask) lSO (in flask)

Chloroform 610 0.62 -8 x lo4 +0.02 1.06 +1W6 -0.003

Carbon-
tetrachloride 80 0.90 -1.3 x lo-' +0.033 1.003 +3 x lo4 -0.001
Methanol 140 0.97 -2 x lo-' +0.0055 1.50 3 x 10-3 -0.07
Benzene 80 0.98 -2.6 X lo4 +O.W

Trauser and co-workers investigated the application of molecular or short-path

distillation in the enrichment of the lithium isotopes. They developed single and
multi-stage apparatus and found separation factors between 1.052 and 1.064 for one
&age in the temperature range from 535 to 62'7°C. I n a similar way the mercury
isotopes were separated.

5.2 Methods of operation

Although simple and rectified batch distillation are the methods most used in the
laboratory, they are not the only ones. Procedures employed mainly in indust,ry have
also been added to the laboratory repertoire. In order to increase output, semi-
continuous and continuous proceases have been adopted, whilst in certain cases
partial condensation also preaents advantages.
iranchembook.ir/edu

5.2 Nethods of operation 245

5.2.1 Batch and semi-continuous distillation

The basic difference between batch and continuous distillation is that the coni-
position of the liquid in the still pot and that of the distillate change progressively
dnring the former operation, but are constant during the latter.
I n batch distillation (whether of the simple or countercurrent type) a definite
charge is put into the still and treated. A distinction should be made between fractio-
nal and imfractionated distillation. In the latter, the whole distillate is collected in
one lot, and the initial material is thus separated into two fractions: the distillate and

Fig. 161
Apparatus for semi-continuous va.cuuiu distillation
with column and heating of the feed

the residue. In fractional distillation, on the other hand, the distillate is collected in
it niiniber of successive portions. In a n interesting paper Block et al. [168] have in-
vestigated the influence of the operating variables on the result of a separation for
discontinuous operation. A small total hold-tip is particularly favonrable for sub-
stances a sinall proportion of which is contained in the flask.
In certain cases a semi-continuous procedure is to be reconiniended owing to the
higher throughput and the better heat econoniy that can be attained. The procedure
is to supplement the contents of the still pot with feedstock a t the same rate at
which distillate is removed. The feedstock may be admitted either directly into the
still pot or at some point in the column and should previously have been heated nearly
to its boiling point (Fig. 138). The flask must be large enough to contain the residue
froill the quantity of feedstock t o be introduced. This method is chiefly eiriployed for
separating a low-boiling preliminary fraction or for concentrating a mixture con-
taining a high proportion of solids. I n the latter case the continuous admission of
feedstock prevents the contents of the still from becoming too viscous or even solid.
iranchembook.ir/edu

248 6. separating processes

After a certain time the feed is stopped and the liquid in the boiler is further eva-
porated batohwise, or drawn off. clontinuous withdrawal is impossible since the feed
steadily introduces low-boiling material which has to be evaporated before the residue
is drawn off. The apparatus illustrated in Fig. 161 is built up from standard parts, is
provided wit,h a column and allows the feedstock to be preheated in a jacketed vessel.

5.2.2 Continuous distillation

In continuous distillation the feedstock - usually preheated to a certain degree -

ie uninterruptedly passed into the apparatus and is separated into an overhead
distillate and a bottom product, both of which are continuously removed from the
process. After the ratio and compositions of these two products, and other predeter-
minedconditions,have become established, they should not change during the further
course of the operation (see chap. 4.7.2). Continuous countercurrent distillation, like
the corresponding batch process, is carried out witsha column, and the feed is usually
fed into the column at some point between its base and top (Fig. 162).
The advantages of continuous, as opposed to batch and semi-continuous distilla-
tion [28], are the leRs severe thermal treatment to which the material is submitted
and the higher throughput that can often be realized. In this respect the results of
the process on the laboratory scale may frequently equal those of semi-technical
batch distillation. Such continuous distillation can, in fact, be used in many cases for
preparative pnrposw (with a throughput of, say, 10--20kg a day), e.g. for the
production of temperature-sensitive pharmaceutical substances or for the recovery
of solvents. Throughprits in the laboratory mostly range from 0.5 to 5 I/h. Corrosive
substances attacking the more common metals may, with a few exceptions, be
distilled continuously in glass; on an industrial scale these ~nhstanceswould require
the use of glass-lined plants or plants of porcelain or special metals. Further basic fea-
tures are that with uniform column operation top and bottom products of constant
composition may be obtained and that the heat consumption of a continuous distilla-
tion is appreciably lower than that of a batch process so that time and energy are
saved. An important requirement of a cont.inuous method - constant composition
of the raw material - can easily be realized in the laboratory and this composition
may readily be determined in advance by a preliminarJ- distillation (section 7.2).

Fig. 162
M i n o r m apparatus for continuous vacuum distillation
a = filling device, h = heated feed vessel, c = graduated pipette, d = heating
tube for feed, e = connecting tube with feed inlet, f = lower column part,
g = coil, h = column head with graduated pipette, a' = graduated pipette,
k = vacuum connection. I = contact manometer, m = flask heater, n = switch-
board, o = bottom, p = sampling point:g = bottoms reflux meter
iranchembook.ir/edu

5.2 Methods of operation 247

h,

i,
/Y
e,
iranchembook.ir/edu

248 5. Separating proceases

5.2.2.1 Applications
The principal applications of continuous laboratory distillation may be classified
as follows:
For cornpurism with industrial operatiam
a) The development of a process to be converted to a semi-technical or industrial
scale :
b) The establishment of a small-scale distillation comparable with one on an in-
dustrial scale ;
G ) Demonstrations for training purposes.

The preliminary lraetionatim of mixture8 in quantities of, say, SO to 100 1. f

a) The removal of a low-boiling fraction;

b) The separation of a medium-boiling cut;
c) The separation of high-boiling material.

Production distilldim for ainounk up to 20 kg a day (see chap. 5.1.3)

a) The separation of temperature-sensitive materials, e.g., essential oils, scents ;
h) The isolation of pure substances from crude mixtures;
c) The purification of starting materials and solvents;
d) The production of pharmaceutical substances.
There is a niarked tendency in industrial distillation practice to convert batch
operations into continuous ones, and it is important, therefore, that laboratory
distillation shonld be able to follow suit. It is hemming more and more necessary
for industrial processes to be developed first with corresponding apparatus on a small
scale. I n the case of multicomponent mixtures, in particular, information may be
obtained more rapidly by experiment than by calculation. Trials in the laboratory
are, of course, far more economical than experiments on an industrial scale, since the)-
save material, power and time, and by the introduction of standard laborator\-
components it has become easy to construct small distillation equipment in accordance
with the flow diagram of an industrial plant. The distillation tests described in
section 5.5.2.1 for example, were all performed with industrial application in mind.
In the same way it is now also possible to evolve laboratory tests for checking pro-
cesses that have already found application on a large scale, with the object of finding
the causes of malfunctioning or irregularities that may occur [73]. Thus, e.g., the
lay-out of coluuiiis for the sharp fractionation of isoprene on an industrial scale was
prepared by semi-technical test distillations. The concentration profiles obtained
from a test column were checked with a computer programme. The agreement was
found to be close. This enabled an optimum lay-out of the industrial columns (see
chap. 1.15; [73a]).
The advantages of continuous distillation are particularly clear for binary
mixtures of components having a large difference in boiling point. The apparatus can
then be relatively simple and can run for days on end without interruption; even in
gless equipment a considerable throughput can be obtained, both at normal and
reduced pressures. By the use of a second column, ternary mixtures can also be
iranchembook.ir/edu

.5.2 Methods of operation 249

separated continuously (chap. 4.9). I n theory a mixture of n constituents can be

separated by means of (n - 1) columns, but in practice a mixture containing more
than three components is unlikely to be dealt with in one operation, since the process
tends to become complicated. I n such a case it is usually preferable to separate the
mixture first into two or three fractions continuously, and then to fractionate each of
these separately in batch into its constituents.
If the mixture to be fractionated contains a large high-boiling “tail” it is advisablr
to remove the latter first by thin-film distillation (cf. section 5.4.3) since high-niolr-
cular-weight substances can often give rise t o decomposition and polpierization. A

crude C1-C3,, mixture

i
cont.thin-film distillation

4
ront flash distillation

c t

4
:ant. column distillation
+
:on:. column distillation
q3--7q q
c,c=
C?, 7-
7 =c=C1
cstch &st. tctch dist botch dist. batch dist.

Fig. 1 ti3
Distillat,ion scheme for i~ C,-C,, multicomponeat, mixture

preliminary removal of the high-boiling fraction by a thernially mild method allows

the yield of distillate to be increased appreciably. Low-boiling “tops” may be sejm-
rated from a product a t a high rate (up to 3 l/h) by flashdistillation (section 5.4.2). If
both of these procedures are applied, a residue-free main fraction is obtained which
can then be split up into fractions by continuous countercurrent distillation. ‘l’u
illnstrate this sequence of operations, Fig. 163 gives R scheme for the separation of a
multicoiiiyonent mixture ranging from C1 to C&.

5.2.2.2 Calculation
As with distillation in batch, it is necessary hefore undertaking a continuow
rectification to select the conditions, namely:
1. The pressure, atmospheric or reduced;
3. The number of theoretical stages required;
3. The niinimuni reflux ratio ;
4. The temperature of the feed;
5. The aniount of heat, necessary.
iranchembook.ir/edu

m 5. Separating processes

Prior to every continuous distillation, an ASW-type distillation - or, preferably,

a sharp rectification - in batch should be performed, and this will usually give the
information needed.
The question of the distillation pressure will be fully discussed in section 5.4. The
number of stages and the minimum reflux ratio are calculated FM described in 4.7.2;
the temperature of the feedstock is usually kept FM near as possible to its boiling
point. The dimensions of the column and the heat requirements can be found as
detailed in sections 4.11 and 4.12, respectively. When calculating the colulnndiaineter
one should remember that the lower half-column is loaded with an amount of reflux

Pig. 164
Plate efficiency as il function of
‘U 120 the bad (McWhirter and
*
U O Y ~ ) [761

“/P J
40
throughput 4
-4, B, C = cyclical operat,ion
wit.h three different divisions of
the cycle, D = continuous
operation

csceeding that in the upper half-coliiiiin by the high-boiling part of the feedstock. If
littlc distillate is to be taken off - saj- 10 or 20% - it is advisable to make the lower
coliiinn wider than indicated by the calculation described, which applies only to the
upper column half.
Here the pofisibilityof cyclic operation should be mehtioned, which was studied by
Gelbin [74]. Thus, according to Cannon [75], the vapour flow from the boiler t’othe
column may be controlled such that a cycle consists of a t>hree-secondflow followed
by (t one-second interruption. McWhirter and Lloyd [76] used a combined plate and
packed column in which each of the five plates carried a layer of packing. With the
mixture met,hj-lcpclohexane-toluene they increased the column efficiency consider-
,rbly by meansof cyclic operation, the optimum depending on the t h i n g of the cycle
and the shape of the quantity-time curve during metering. Rg. 164 illustrates the
astonishingly high plate efficiencies with cgclic operation.

5.2.2.3 Apparatus for continuous column distillation

For laboratory and semi-technical operation the packed column is the type chiefly
rmployed; in special cmes use is made of plate columns constructed of glms, which,
owirlg to the resistance of the material, are also suitable for the distillation of corrosive
substances.
iranchembook.ir/edu

5.2 Methods of operation 25 1

An all-metal laboratory apparatus with a tubular heater was described as early

as 1931 by Burstin and Winkler [77]. It was designed for distilling crude mineral oil
residues a t atmospheric and reduced pressures and has also proved convenient for
heating readily decomposable or foaming liquids, since it operates on the principle of
thin-film distillation. It may be noted that this unit (Fig. 165) corresponds to industrial
installations of the same period and could thus he used for solving problems encoun-
tered in large-scale practice. The control systems for continuous 1aborator;v coluniiis
described by Kolling and Tramm [78] are intended to provide automatic operation.
Fig. 166 shows the arrangement used for distillation a t at,mospheric pressure.

Fig. 163
Apparatus for continuous distillation with tubular heater (Burstin and Winliler)
I = Constant level vessel, 2 = Keedle valve, 3 = Observation bulb, 4 = Vacuum
connection, 5 = Condensers, 6' == Metal block, 7 = Thermometer wells, 8 =
Insulated jacket, 9 = Dephlegmator, 10 = Reeeivers, 11 = Vacuum connections.
12 = Burners

Further details of a Destinorm apparatus for continuous distillation at reduced

pressures (Fig. 162) will be discussed in chap. 5.2.2.5. A continuous plate column is
exemplified in Fig. 167 by a Labodest apparatus which can operate from atmospheric
pressure down to about 20 tom. It is provided with vapour-collision bubble-cap
plates of glass (chap. 7.3.3) having a n efficiency between 80 and 100%. Their mode
of operation is identical with that of industrial plates so that the test results can be
transferred directly. Besides, they are well suited to demonstrations since the gradual
exchange process can be observed. This type of column is built with lengths between
600 and 1300 mm and column diameters between 30 and 120 mm and has between
5 and 28 plates depending on the distance between the plates (35 to 150 nim).
iranchembook.ir/edu

262 5. Separating processes

A remarkable feature of the apparatus of Pig. 167 is the boiler with circulatory
heat,ing (4) which imitates the industrial construction. The switch cupboard (20)
contains all tJheineasuring and control devices.
This is a good illustration of the general trend to learn from industry and to
shape laboratory and pilot plant methods according to large-scale procesReFi [79].

-+

-- /

Fig. 166
Arkmgement of an automatic apparatus for continuous distillation a t atmo-
spheric pressure used by Kolling and Tramm
a = Contact manometer, b = Relay, c = Electfic heater, d = Insulating jacket,
e = Heating jacket. f / g = Contact thermometers, h/i = Relays, k = Valve,
1 = Storage vessel, m = Flowmeter, n/o = Electric contacta of flowmeter,
p = Relay, q = Electroma,gnetic valve, r = Resistance capillary, 8 = Total
condenser, t = Capillary tube, u = Needle valve, v = Receiving vessel, w =
Thermometer, z = Contact thermometer, y = Syphon-shaped tube, z = Draw-
off cock, a, = Capillary tube, b, = Electromagnetic valve, c1 = Needle valve,
d, = Relay, e, = Overflow
iranchembook.ir/edu

5.2 Methods of operation 25:3

70

Fig. 167
Labodest bubble-cap column with vapour collision plates made of glass for
continuous operation a t 760 to 20 torr (Stage)
1 = pump, 2 = feed vessel, 3 = feed heater, 4 = circulating evaporator. 5 =
bottoms condenser, 6 = bottoms receiver, 7 = reflux meter, 8 = safety device,
9 = differential pressure controller, 10/11 = plate columns with heating jacket,
1 2 = column head, 13 = condenser, 14 = expansion vessel, 15 = distillate
condenser, 26 = exchangeable receiver for distillate, I7 = buffer volume,
18 = tap distribution, 19 = cold trap, 20 = switch cupboard
iranchembook.ir/edu

2% 3. Separating processes

The importance of keeping the equipment as simple as possible was soon recog-
nized, and accordingly the Labodest apparatus after Stage [80]as well as the series of
components manufactured by Fischer Labor- und Verfahrenstechnik of Bonn-Bad
Godesbergand by several other firms were designed. This is true also for the Destinorm
equipment of VEB Kombinat Technisches Glas of Ilmenau, and the components made
hy VEB Jenaer Glaswerk Schott & Gen. of Jena. Standardized components enable
fractured parts to be replaced promptly. Further, all kinds of problems may be tackled
as regards equipment, and large-scale plant,s may be copied to a large extent.

5.2.2.4 Examples of application from laboratory practice

With the aid of the columns and other apparatus described it is possible to separate
liquid mixtures continuously at normal and reduced pressures as well as at small
overpressures (chap. 5.4.5) as sharply as in batch distillation. This is clear from the
distillation curves of the products obtained in the continuous distillation of a mixture
of crude fatty acids in the C, to C,* range (Fig. 168). A further example shows that by
the use of control devices it was possible to separate a mixture of phenols into its main
components with great constancy. Fig. 169 illustrates the results of the first separa-
tion in this sequence, which was made between the ortho- and rmtaJpara-cresol
fractions.
120 samples were taken during the run for determination of the density of the
bottom and the solidification point of the tops. As the diagram shows, there were
negligible variations in the temperatures of the feed, the head and the bottom, and

260

L
2220
ga200
E 180
I60
140
0 10 20’ 30 40 50 60 70 80 */.wt.
a) distillate -w

+‘.”c”
$
J ao Fig. 168
2190 ‘Distillation curves of products
2 prepared by continuous labo-
%I70 ratory distillationfrom B C, -C,,
d 150
+
mixture of fatty acids
a) C,.’.C,
130
b) c,,-*c,
0 10 20 30 40 50 50 70 80 ‘Iewt.
b) distillate -+
iranchembook.ir/edu

5.2 Methods of operation 255

in the physical properties in the course of 22 hours. The vacuum, also, could be kept
constant by the control mechanism. In a similar way a phenol fraction, solidifying
at 37.0°C, could be worked up continuously t o an overhead of pure phenol, solidifying
at 40.3 "C (purity 99.2%), by the use of a reflux ratio of 10: 1a t 60 mm pressure. The
bottom product was continuously withdrawn so that its residence time in the still
n as considerably shorter than with batch operation. Thus, decomposition of the
bottom product was largely prevented. Further continuous methods, including azeo-
tropic and extractive distillation, flash distillation, thin-film distillation and molecular
distillation will be discussed in sections 6.2, 5.4.2, 5.4.3 and 5.4.4 respectively.

a
t 200
U.

Q
3120
zc xc 1.08 >I
p
g 80
$ 4 E
-:50 1.06 E 40

-
30
10 1.00 0
0 6 16 2% 32 40 48 56 64 72 80 88 96 104 112 120 128
sample

Fig. 169
Continuous rectification of crude phenols. Pressure: 30 tom; duration of run:
22 hours
Distillate: phenol +
o-cresol;bottoms: m/p-cresol +
xylenols and higher

5.2.2.5 Starting up continuous distillations

The function of the various control devices has been discussed in section 5.2.2.3
we need thereforedeal only with a few points concerning the operation of the process
(Pig. 162).The mixture charged to the reboiler o should have the expected composition
of the bottom product, so that long equilibration times are avoided. I n order to wct
the packing properly a t the same time, this reboiler charge may be prepared hatch-
wise, by distilling off the required amount of material. Then the supply of feedstock
is started (e), after first, heating it to the required temperature in the storage vessel d.
The feed rate is regulated by means of the measuring burette c and the reflux ratio
is adjusted a t the column head h. The load on the column depends on the feed rate
and is in addition governed by the contact manometer 1. It is obvious that the ad-
justment must he such that the feed rate is the sum of the overhead and bottom
take-off rates. The valves on the distillate h and bottoms receivers i are therefore
so regulated that the proper ratio is obtained. As an example we can take the separa-
tion of a benzene-toluene mixture containing 20% of benzene. A t a feed rate of
500 ml/h we must take off 100 ml/h a t the head and 400 ml/h from the bottoms. If
the reflux ratio is to be adjusted to 2: 1, the vapour load will have to be regulated at
300 ml/h. The reflux will be 200 ml/h above the feed point and 700 nil/h below the
feed. Equilibration will probably require1/, to lhour and its progress can be judged by
iranchembook.ir/edu

256 5. Separating processes

variations in the top and bottom teinperatures (FQ.169). 4 s soon as the feed tempera-
ture has also become constant, the apparatus should operate without further adjust-
ment. The operator’s attention can then be confined to checking the quantities of feed
and take-off and to recording the experimental data.

5.2.3 Separation by partial condensation

The term partial condensation is used to denote the liquefaction of a fraction
from a flow of vapour. It can be brought about deliberately by interposing a special
condenser - a so-called &phJqmator - in the vapour line, and it also occurs as an
undesired effect by loss of heat from an insufficiently insulated column (Fig. 170a).

Fig. 170
Partial condensation
a) Dephlegmator (Labodest, type I11 of Stage) and lateral condensation
b) fractional partial condensation

The gradual dephlegmation of a vapour flow in a tube system consisting of sections

with progressively falling temperatures serves for preliminary separations which give
various fractions in the case of components with great differences in boiling point
(Fig. 170b). Examples of calculations applicable to such partial condensation (assum-
ing the condensate to be drawn off immediately) have been given by Sigwart [81].
The enrichment achieved by partial condensation can be increased appreciably by
employing the countercurrent principle [82].
The flows of vapour leaving the column have to be in a particular ratio depending
on the object of the distillation. In industry i t is common to make a dephlegmator
produce the reflux and cool the distillate to be drawn off in a production condenser.
The aim is to use the amplifying effect of the dephlegmator which is due to the partial
condensation of the high-boiling components. Partial condensation is employed
iranchembook.ir/edu

5.2 Methods of operation 257

especially in the low-temperature separation of gas mixtures. Otto [83] has reported
the partial condensation of binary mixtures in vertical tubes.
The performance depends on the nature of the mixture to be separated [84].
Another advantage is the low hold-up. I n Table 44 two examples concerning mix-
tures of technical importance are given.
The calculation of the enrichment caused by a dephlegmator can be carried out
with a formula due to Fabuss:

log (V, + 1) = -
ix-1
(.log ;
Y2 + log -
"1
1 - YZ
(183)

in which
amount of condensate
Vi = the internal reflux ratio = (both in nioles/h) :
amount of vapour feed
y, = mole fraction of vapour feed;
y2 = mole fraction of residual vapour.

Table 44
Separating action of a dephlegmator

Mixture XI Xg 2' 'lt h

Concentration before Concentration after Reflux Number of
the dephlegmator the dephlegmator ratio theoretical stages

Ethanol-water 67 molyo 78 molyo 3 3

Benzene-toluene 63 molo/o 80 molyo 5 1

A nomogram has been developed for the solution of this equation [85]. An
equation for the calculation of residual vapour and condensate concentrations of a
dephlegmator in multicomponent separation is given by Fischer [86]. On the basis
of extensive experiments Herrmann [87] has tried to disentangle some points of the
complicated problem of partial condensation. Although the study deals with in-
dustrial dephlegmators Herrmann's findings are of use for the laboratory and pilot
plant scales as well. Further methods of calculating the separating efficiency of a
partial condenser have been reported by Troster [88]. Wondrak [%a] is credited
with having pointed out that the optimum fields of application of distillation and
partial condensation are quite different. Consistent treatment of simple thermo-
dynamic model concepts reveals the functional relations between the two separating
processes for ideal binary mixtures.
I n analytical and preparative distillations in the laboratory total condensation is
generally used, and dephlegmation is employed only when the simulation of an
industrial plant is required. I n this case a dephlegmator (Fig. 170a) is inserted above
the column. Total condensation has the advantage that it is relatively easy to process
the liquid condensate in a certain ratio whereas the establishment of a definite reflux
ratio from a dephlegmator is difficult, since small variations in the amount and
temperature of the cooling water are sufficient to cause large changes in the quantity
and composition of the reflux and residual vapour. On an industrial scale it is the
1 7 Krell, Handbook
iranchembook.ir/edu

268 5. Separating processes

practice not to measure the quantity of reflux in a partial condenser, but to reg&& it
only by controlling the t,emperature a t the head. If the quantity and temperatures of
the cooling water entering and leaving the dephlegmator are measured, the quantit,y
of reflux can be calculated roughly from the heat of evaporation of the distillate. In

Fig. 171
Destinorm column head for partial conden-
sation with capillary take-off tubes
a = Dephlegmator, ZJ = Collecting funnel,
c = Measuring pipette,
d = Glasswool imulation,
e = Thermometers, f = Condenser,
g = Perforated funnel, h = Cock,
i = Capillary distillate take-off,
k = Receiver, I = Measuring pipette

cmes where the products are more or less constant, as generally occurs in manu-
facture, this procedure may be adequate, but when mixtures of varying composition
have to be dealt with the calculation becomes very uncertain.
I n laboratory experiments utilizing partial condensation to imitate a teclmicrtl
process, a special form of the author’s Destinorm column head, shown in Pi. 171,
may be used. The ascending vapour is partially condensed in dephlegmator a and
the resulting condensate is passed into measuring pipette c through the collecting
funnel b; pipette e isprovided with a capillaryinlet, so that theamount of reflux can be
observed a t any moment. The section between the funnel and the column is insulated
by the glass wool jacket d. The temperatures of the incoming and outgoing water are
read off on thermometers e. The uncondensed vapour flows upwards to condenser f ,
iranchembook.ir/edu

.5.2 Methods of operation 259

where it is completely liquefied. The condensate is collected by the perforated funnel

q and when cock h is opened it passes down through capillary tube i into receiver k.
-\ cooler may, if desired, be interposed between k and the receiving vessel 1.
As the separating effect of a dephlegmator is relatively low and its control on a
lahoratory scale presents difficulties it will, as has already been said, be einployed
mainly for comparative distillations. Partial condensation in stages may, however,
be applied for the preliminary separation of mixtures with large differences in boiling
point, especially for the removal of water or some other low-boiling fraction. The
column is then replaced by a system of tubes (Fig. 170h) containing condensers a t
progressively decreasing temperatures. I n this way it is, for instance, possible to split
up a mixture of crude fatty acids (up to Czo)fairly rapidly into the fractions C, -C4,
C4-C,, and C, t o Czo.
=\ccording to Junge [SS] a proper application of partial condensation within a
coliinin can increase its efficiency. The effect in question is a partial wall conden-
sation due to a heat loss in the column, i.e., to non-adiabatic operation. Trenne [go]
has reported a similar process. On the other hand the extensive calculations of
Kuhn [91] promote the view that the most effective procedure is to avoid all conden-
sation except a t the upper end of the column. Yon Weber [92] has pointed out that
partial condensation offers advantages if it is applied in connection with a coluinn
narrowing towards its top (see Fig. 172). Owing to the increase in concentration

Fig. 172
Partial condensation in countercurrent distillation [92]
iranchembook.ir/edu

260 5. Sepsrating proceases

in the upper part of the column there is an advantage in reducing its hold-up. On the
other hand the required reflux flow becomes smaller towards the top. On this account
the column cross-section may be reduced in proportion to the enrichment and a deph-
legmator incorporated in order to reduce the throughput at the upper end. Voigt [93]
hati shown on theoretical grounds that the separating effect may be improved appre-
cirtbly if heat is withdrawn, not only in certain sections, but from the whole of the
oolumn's surface. This produces an improvement only as regards the enrichment of
the low-boiling components. If, on the other hand, the concentration of the high-
boiling conatituents is to be increased (as,for instance, in the isolation of stable iso-
topes) it Will conversely be necessary to introduce heat inh the lower part of the
column,so as to reduce the throughput in this region (cf. 5.1.4; [93a, b]). Recent in-
vestigations by Blafi and Sauer [93d] of the partial condensation and nonadiabatic
rectification of binary mixtures in packed tubes point to ways of improving separation
effects for particular ratios of tube length to diameber.

5.3 Ternperature
In the most commonly occurring forms of distillation the temperatures lie within
the range of about 20 to 250°C. If the boiling points of the components to be separated
are below room temperature it becomes necessary to employ one of the special methods
of low-temperature distillation, in which additional coolig agents are required for
condensation. Distillations involving temperatures in the range of 250-400 "C may
be termed high-temperature distillations. A distillation performed isothermally by
keeping the temperature constant and varying the pressure is also conceivable.

5.3.1 Low-tern perature d ist iIlation

Mixtures of low-boiling substances, for example light hydrocarbons and gases
such as H,, Nay0,and CO can be fractionated either at atmospheric pressure, with
the aid of cooling agents, or under pressure. In the latter case the use of pressure
establishes overhead temperatures above those of the usual cooling media (sectmion
€1.4.6). Since pressure distilhtion involves difficulties with apparatus, low-tempera-
tare distillation is generally preferred in the laboratory and in pilot plants. The
technique of low-temperature distillation has evolved to a high degree of perfection,
and completely automatic apparatus suitable for the temperature range of -190"
to 20°C has been developed. Various types of columns - empty or packed - are
used.
For analytical purposes low-temperature distillation has now to some extent been
superseded by gas chromatography (chap. 5.1.2). If, however, the object is to prepare
a considerable quantity of a pure gas or to effect a comparison with a large-scale
installation, low-temperature distillation is still the best method. This applies espe-
cially to industrial waste gases without hydrogen concentrations and, more recently,
to natural gas. For example, the separation of helium and nitrogen from natural gas
is still a difficult problem. The separation of fission rare gases from the used air of
iranchembook.ir/edu

5.3 Temperature 26 1

nuclear plants is now gaining importance. After the removal of accompanying com-
ponents the task is to separate the mixture Kr/Xe/N, with radioactive krypton and
xenon isotopes by means of low-temperature distillation [93 c].
The basic work on the analytical distillation of gases (comprising the permanent
gases, the gaseous and lower liquid hydrocarbons) was carried out by Podbielniak
[94], who for this purpose developed an effective column containing a closely-wound
wire spiral known as Heli-grid packing (see chap. 7.3.4), which has also proved its
wort,h for normal dist.illations.

Fig. 173 Fig. 174

Low-temperature column with Gas density balance after Stock
fused-on dephlegmator
(Grosse-Oetringhaiw)

Grosse-Oetringhaus [95] gives detailed methods for performing low-temperature

distillations. To start with, the taking of samples requires great care. Semi-automatic
and automatic apparatns for drawing instantaneous and continuous samples have
been devised. The actual separation is carried out in the following manner. First,
those components not condensed by liquid nitrogen (-195.8"C) are removed and
analysed in an Orsat absorption apparatus. The condensable components are treat,ed
as gas in washing bottles for the removal of CO,, H,S and NH3 and then again
liquefied. The distillation is performed in a packed column insulated by a silvered
vacuiiin jacket (Fig. 173); the column has spiral sections to take up strains. The
iranchembook.ir/edu

262 6. Separating proocesses

.-

dephlegmator with its conical tube is either fused to the column or connected to it by
a ground joint. I t has a vessel f9r the cooling medium which again is surrounded by
a vtwuum jacket. The column and the dephlegmator are filled with 2 x 2 xO.2 mtn
stainless steel spirals. The apparatus has been rendered largely automatic by the

Fig. 175
Low-temperah r e column (Koch a.nd
Hi1berath)

incorporation of various control devices, including a manometer with automatic

pressure control and an arrangement for maintaining the temperature of the cooling
bath. An ingenious set-up for cooling the column head to a constant temperature has
also been described by Stokes and Hauptschein and Drawin 1971. The molecular
weights of gases may be determined by means of Stock’s balance (Fig. 174) [98].
Vapour pressure measurements (see chap. 4.4.1) are used to determine n-butane and
isobutane contents. A useful siimniary of gas-handling techniques has been given by
iranchembook.ir/edu

5.3 Temperature 263

Miller [99]. I n their book on high-purity gases Miiller and Gnauck [loo] deal with the
production and use of equipment for work on gases and with gas analysis, separation
and purification.
The low-temperature column of Koch and Hilberath [loll, which is very simple
to operate, will be taken as an example to describe the procedure. The column has
twen given the form of a spiral, like Jantzen’s column shown in Fig. 7 b, and it conse-
quently has a low hold-up (3--4ml), so that a charge of 15--25g is sufficient
(Fig. 175). It functions in the same manner as a distillation apparatus at nornial
temperature. By the production of total condensation any desired reflux ratio may he
ohtained. The distillate may be taken off as gas or as liquid,
Thermometer c is suspended from a thin wire and the capillary opening d is
closed with picein. The side-tube e of the ground-on cap is connected to a siiiall
drying tube.
The degree of vacuum in the insulating jacket should be checked at intervals by
means of a high-frequency vacuum tester, I n a darkened room no lighting up of the
gas space should be noticeable, at, most a wavering green fluorescence of the glass
walls. If the gas space lights up, the jacket must be exhausted anew. This is done, by
means of a three-stage mercury diffusion pump, with a cooled adsorption vessel
containing active charcoal or silica gel between the pump and the jacket in order to
freeze out mercury vapour. Cocks are greased with high-vacuum grease (section 9.4).
When the requisite vaciium (10-6 mm Hg or better) has been reached, the Schiff valve
f is closed.
The boiling tube y is now cooled with a solid-C02-methanol slurry or another
suitable cooling mixture to the required temperature. Condenser h is filled with its
cooling agent. Liquid air or liquid nitrogen may be used. Alternatively, a salt solution
iiiny be used and the cooling temperature be kept constant by means of a cryostat.
The dried gas sample (if necessary freed from CO,) is then condensed into the boiling
titbe g, which for this purpose can be removed at a ground joint i. Insteadof the
cooling bath a Dewar vessel k is now placed so far around the boiling tube that its
upper end is in contact with the bottom of the stand. The evaporation of the charge
in g is then brought about as usual by the heat from an electric resistance element 1.
The sapours pass up the spiral column m, which is surrounded by the silvered in-
sulating vessel and by a glass-wool jacket. The temperature is measured a t the
column head in a well that prevents the thermometer c from being cooled by the con-
densate. The distillate is drawn off below the condenser through the regulating
valve n.
I n order that the distillate may, if desired, be collected as liquid, a spiral tube
leads down to a fused-on, graduated receiver and the whole is surrounded by a
Dewar vesseI containing a cooling medium in which the distillate is liquefied. The
weight of the liquid condensing in the spiral yields sufficient,suction a t the regulating
valve to draw off the vapour from the column. If the distillate is to be collected as
gas, i t can be passed from the valve to a vessel b in which it displaces a suitable
liquid. The necessary suction is then produced bv a difference in level between the
gas entry and the liquid overflow.
If the boiling and solidifying points of the distillate lie close together the tempera-
iranchembook.ir/edu

ture in the condenser haa to be controlled exactly. Ebr preparative purposes a low-
temperature column was developed by Stage [lo21 in which the coolant temperatwe
is controlled automatically (Fig. 176). This refers to the liquid in both the condenser
and the still pot, which is cooled by a streald of evaporating liquid air. When the
temperature reaches the minimum value pre-set on a contact thermometer a magnetic

111) cut A-B

Fig. 176
Packed rolumn for preparative low-temperature distillation (Stage)

valve opens and admits atmospheric air. The distillate is also drawn off through
magnetic valves. The low-temperature still after Grosse-Oetringhaixs (Fig. 177) is
used for the distillation of low-boiling liquids or liquids containing dissolved gases.
The sapours or gases which cannot be condensed in the first condenser will undergo
condensation in the low-temperature condenser of t,he receiver. This is filled with a
suitable cooling agent.
The latest development. in the field of low-temperature distillation is illustrated
by the fully automatic apparat.us of Podbielniak. The “Thermocon” model, series
iranchembook.ir/edu

5.3 Temperature 265

8 700, is designed to operate in the temperature range of -200 to 20 “C (Fig. 178 [ 1041 ;
cf. section 8.1).
-4 modified Podbielniak apparatus, the “Ruhrgas” model, employs charges of
3-4 1 of gas (at normal pressure). A distillation takes 2 hours; the introduction of
the charge by condensation requires about the Same time. The built-in recorders
continuously register the temperature of the still head. Fig. 179 gives an example of
such a distillation curve [105].
Low-temperature distillations of very small amounts of liquefied gases (about
5 1111) inay be carried out with the apparatus of Simons [lW], which fits into a Dewar
vessel (Fig. 180).

F-

Fig. 177
Low-temperature,distillation
apparatus (Grosse-
Oetringhaus)
iranchembook.ir/edu

266 5. Seuaratinp: processes

Fig. 178
Fully automatic Ion--temperaturedistillation apparatus of Podbielniali
iranchembook.ir/edu

6.3 Temperature 267

+<O
C
?O
-M
t
I
-40
$ -60
z -80
E%-loo Fig. 179
d -120 Boiling curve obtained with the
-140 “Ruhrgas”low-temperature
-160 apparatus
7 00
distillate +

Fig. 180
Low-temperature micro-column (Simons)

The minimum reflux ratio required in a low-temperature distillation ma\- be

calculated approxiniately by a formula due to Podhielniak [107], which assumes the
concentration of the charge to be 50 mol% :

(x. = mol. fraction of low-boiling component in distillate;

xB = mol. fraction of low-boiling component in still pot).
For calculating the minimum number of stages in a low-temperature distillation
at infinite reflux ratio, the nomogram of Fig. 181 after Podbielniak [107], may he
used. The diagram again assumes that the concentration of the charge is 50 niol%,
but takea account of the relative volatility of the mixture and the desired concen-
tration of the distillate. Firstly, the relative volatility of the gas mixture to be sepa-
iranchembook.ir/edu

268 5. Separating processes

rated is determined from vapour pressure values, as described in section 4.4.4. Prom
this value of OL one proceeds vertically t,o the curve corresponding to the required
purity of the distillate (80-99.9 molyo). From the intersection with this curve a
horizontal line is followed to the scale for the number of theoretical stages. For a
rough guide Table 45 [lo81 gives the minimum stage number for a few commonly
occurring mixtures of low-boiling hydrocarbons at infinite and minimum reflux
ratios.
It can be seen from the table that the separation of low-boUng mixtures is
relatively straightforward. Thus, it is usually sufficient to have columns with low
efficiency. It has to be borne in mind, of course, that the data given refer to v = 00.
For finite reflux ratios the requirements have to be altered appropriately.

3, _ _ f

Fig. I81
Nomogram for the detmminat.ion of the minimum plate number a,,,, for low-temperature
distillation at v = s
iranchembook.ir/edu

5.3 Temperature 369

Table 45
Minimum numbers of stages and reflux ratios required for the separation of
various low-boiling binary mixtures
~~

Mixture OL A B C D

Methane-ethylene 61.2 1.32 1.82 0.03 0.161

Methane-ethane 107.0 1.14 1.60 0.018 0.093
Ethylene-ethane 2.3 6.5 9.2 1.60 7.95
E thane-propylene 8.1 2.52 3.57 0.27 1.1
Ethane-propane 9.3 2.38 3.36 0.23 1.2
Propylene-propane 1.3 20.2 28.6 6.65 33.0
Propane-isobutane 3.3 4.6 6.5 0.89 4.5
Isobutane-n-butane 1.5 13.0 18.4 3.99 20.0
Isobutane-isobutylene 1.2 28.8 40.8 10.0 30.0
Isobutylene- 1-butylene 1.03 25.5 36.0 8.98 43.5
n-Butane-isopentane 2.8 5.3 7.5 1.16 5.8
Isopentane-n-pentane 1.33 18.6 26.3 ti.25 30.1
n-Pentane-n-hexane 3.0 4.9 6.95 1.04 5.25
A = number of stages required with xS = 50 molyo, zB = 99.5 mole/, and c = iu
B = number of stages required with ZB = 10 molyo, ZE = 99.5 mole/, and 2' = ca
C = minimum reflux ratio for a mixture with ZB = 50 molyo and zE = 100 niolqo
D = minimum reflux ratio for a mixture with zg = 10 molo/b and zE = 100 mole/,

5.3.2 High-temperature and isothermal distillation

Temperatures above 250 "C may cause some decomposition in organic compounds
and the range of 250 to 400°C is therefore avoided if possible in the laboratory. The
necessity may, however, arise in determinations of boiling range of mixtures of low
volatility at atmospheric pressure, as with tars or high-molecular-weight waxes.
Crude waxes are often purposely distilled destructively to obtain fractions of low
molecular weight [log]. On the other hand inhibitors are sometimes added in an
attempt to reduce decomposition and polymerization [1101.
I n the distillation of substances of low volatility a t atmospheric and reduced
pressures a short column is all that is employed and its function is mainly to act as
spray trap. Coarse packing, placed on a gauze support, is used; plate columns are
unsuited for such purposes. It is obvious that the column should be provided with a
heating jacket and that no part of the apparatus, from the still pot to the condenser,
should be unheated, so that there may be no partial condensation. The receivers
should also be capable of being warmed. A useful measure is to heat the take-off
valve by radiation from an infrared source or by a hot air blower (hair dryer), so as
to prevent it from sticking. Better still is the use of magnetic valves.
I n laboratory high-temperature distillation the boiler is as a rule heated indirectly
by a n oil, Wood's metal or molten salt bath [lll], controlled by a contact therxno-
meter. (The maintenance of a constant high temperature in such a bath is, however,
usually not a n easy matter.) Table 46 shows the temperatures that are attainable
with various baths (cf. secttion 7.7).
iranchembook.ir/edu

270 5. Separating processes

A molten salt bath has the advantage that it can be used over a wide temperature
range and that it does not give rise to objectionable vapours. However, the flask must
be removed from the bath when heating is stopped and the adhering salt washed
off with wat>er.
In high-temperature distillation cooling sometimes poses serious problems in that
the condenser cross-section must be prevented from being narrowed or even blocked
by solidifying (or sublirmng) matter. By m a n s of circulatory thermostats and with
water or glycol as warming agent the required condensation temperatures can be set
and kept. Another simple possibility is offered by the boiling condenser which con-
sists of two concentric tubes. The inner tube contains the liquid whose boiling point

Table 46
Heating baths for high-temperature distillation

Medium Temperature range ("C)

C;lycerin up t o about 160

Sulphuric acid upto about250
Poly-glycols up to about 300
Paraffin wax, melting point 30"- 60 "C up to about 300
Mineral oil uptoabout330
Mixture of 40% wt. SaNO,, 53% wt. KNO, and 150-500
70; wt. SaXO,
Wood's metal 70 -300

lies somewhat above the solidifying point of the distillate but below the boiling point
of the latter. The distillate vapours heat the liquid and cause it to evaporate. The
distillate condenses on the outer wall of the cooling tube. Fig. 182 shows a distillation
apparatus with a simple boiling condenser after Stage as presented in [112]. In that
paper, further arrangements for the distillation of high-melting, subliming materials,
partly with the addition of solvents, are described.
Of late the purification of metals by distillation has become important, for
example with the mixtures Al-Zn, Al-Mg, Pb-Zn, Ag-Zn and Ag-Pb [113].
Vapour pressure data for metals can be found in Leybold's handbook [lla]. The
inveatigation of such problems on a laboratory scale, with small amounts of material,
may become common in the near future. Fig. 183 shows two forms of apparatus
suitable for this purpose, having the flow of vapour directed sideways in the one
cafie and downwards in the other [115]. An extensive account of the vacuum distilla-
tion of non-ferrous metals and alloys on a laboratory scale has been published by
Spendlove [ 1161. Horsley [ 1171 has described an apparatus for the distillation of alkali
metals. The metal is melted in vacuum, filtered and distilled a t mm pressure.
The rubber gaskets are protected from excessive heat by cooling coils. The metal
vapour is condensed on a surface cooled by circulating oil. The chloride TaCl,
( K p . = 210°C) was used by Parker and Wilson [118] for the purification of
tantalum. Bezobrazov et al. [118a] developed an apparatus made of quartz for the
iranchembook.ir/edu

5.3 Temperature -77 1

LA
to rough
vacuum

otled vacuum

L switch board
Fig. 182
Apparatus for high-temp?-
rature distillation with boiling
condenser (Stage)
iranchembook.ir/edu

272 6. Separating processes

continuous distillation of high-boiling substances up to 1OOO"C (e.g., sulphur, sele-

nium, tellurium, zinc, cadmium, arsenic sulphide).
An operation frequently occurring in the laboratory is the (unrectified)distillation
of mercury to remove tin, cadmium and noble metals. Generally, a preliminary chemi-
cal purification is carried out so that there remain, in addition, traces of the purifying
agents in the mercury. In general the evaporation process exhibits nonstationary and
strongly pulsating states, which have been analyzed by Schmucker and Grigull
[118b].
The apparatus, which is constructed of silica or high-melting glass, is often de-
signed so as to operate continuously on the principle of the Sprengel pump: after a
preliminary pumping-out the apparatus evacuates itself further continuously during

al
Fig. 183
Unit for metal distillation
a) with horizontal vapour flow
b) with downward vspour flow
1 = furnace, 2= melt, 3 = receiver, 4 = condenser, 5 = condensate

its operation [119]. Fig. 184 illustrates the apparatus diagrammatically. For the
self-pumping action to be effective, the condensed mercury must flow down through
a capillary tube of about 1.5 mm bore. A better procedure, howaver, is to evacuate
the apparatus first with a mercury diffusion pump and to close valve H after 1 to 2
hours' distillation, reopening it only when evacuation is again necessary.
Sincevery pure mercury is required in increasing quantities a mercury bidistillation
apparatus as shown in Fig. 186 was evolved by VEB Jenaer Glaswerk Sch0t.t & Gttn.
of Jena. After vacuum evaporation in the first stage the condensate is passed through
a tube bridge into the second stage where it is again subjected to evaporation. The
distillate then flows through a tube working on the barometric principle to the recei-
ver. At a pressure of 1torr about 2 kgJh distillate may he obtained with a heating
power of 300 VA for each distillation stage. According to spectral analysis results, the
mercury thurj obtained has a very high degree of purity. This is partly due to the
use of precision ball ground joints which need no vwnum grease.
In isobaric distillation - the form normally practised - the pressure is kept con-
stant and the di&illate paspes over with rising temperature. In isdhcrmal distillation
the still pot temperature is kept constant by a thermostat and the pressure is pro-
iranchembook.ir/edu

5.3 Temperature 273

Fig. I84
Apparatus for the distillation of mercury (dimensions in mm)
A = Storage vessel for intake, B = Riser tube, C = Distillation vessel, D =
Electrical heater, E = Aluminium block, F = Descending capillary tube,
G = Receiver with overflow, H = Mercury float valve with connection to
diffusion pump

Fig. 185
L Mercury bidistilling unit

18 Erell, Handbook
iranchembook.ir/edu

274 5. Separating processes

greseively reduced. One can then construct a diagram for the teinperature in question
by plot.ting the pressure along one axifi and the amount of distillate along the other.
This method is applied when it is necessary to know the pressure at which a definite
percentage will be evaporated from-a multicomponent niixture a t a particular
temperature, for instance in the ewe where steam at a certain preseure (corresponding
to a particular temperature) is available as heating medium.By nieans of isothermal
distillation Echols and Gelus [120] determined the properties of the residue of a
niixture which at B constant temperature corresponded to a particular pressure.
Equilibrium curves may also be determined isothermally. The theory of iijothermal
distillation as compared to isobaric distillation has been discussed by Ulusoy and
Sakaloz [121]. Accodng to their findings isothermal distillation m y be t.he more
favourable method in some cases.

5.4 DistiIlation pressure

An important point to be considered prior to every distillation is the pressure atj
which the separation is best performed. If there are no objections to its use, normal
atmospheric pressure will generally be chmen, since a reduced or incremed pressure
involves complications in the apparatus and a vacuum distillation gives a lower
throughput.
As a rule gss mixtures containing low-boiling hydrocarbons are distilled a t
atmospheric pressure (i.e. at low temperatmes) or at an increased pressure, whilst
easily decomposable or high-boiling organia substances are distilled at a reduced
pressure in order to prevent overheating. High throughputs may be realized with
flash distillation at 20 to 1mm Hg pressure, whilst temperature-sensitive compounds,
which cannot he distilled from a flask, may be separated by the mild process of thin-
film distillation a t 20 t o 10-1 mm. For the distillation of substances of low vapour
pressure and high molecular weight (20-1200) the method of molecular distillation
was developed, in which - by the use of pressures of 10-8 to 10-6 mm - the mean
free paths of the molecules are of the same magnitude as the distances travelled by
the vapour in the apparatus. Experience gathered in the simple and countercurrent
distillations of sensitive substances under reduced pressure and methods evolved for
such procedures have been reviewed extensively by Frank and Kutsche in their book
on mild distillation [ 1221.

5.4.1 Simple and countercurrent distillation under reduced pressure

Rectified distillation, with a coIum, may be carried out at pressures down t,o
about 0.5 mm Hg. At lower pressures special forms of apparat,us must be employed.
The hasic advantage associated with the use of reduced pressure in distillation is
the lowering of the boiling point and the consequent possibility of separating com-
pounds below the temperatmureat which they decompose or undergo chemical altera-
tion, such as polymerization. Examples from industry that may be quoted are the
distillation of lubricating oils, the fractionation of crude phenols (which is perforniecl
iranchembook.ir/edu

5.4 Distillation pressure 2i5

a t 20-60 mni pressure) and that of synthetic fatty acids (carried out a t 1-20 aim).
Reduced pressures are also employed for distillation where, although no direct danger
of chemical change is present, the boiling points lie so high a t normal pressure that,
it is profitable to reduce them for reasons of heat economy, as with certain ethereal
oils. A fiirther advantage of a reduced pressure in distillation is the fact that the
equilibria are often more favourable for the operation than at atmospheric pressure.
An uzeotrcvpe frequently becomes richer in the low-boiling component as the
pressure is reduced. By continuing the reduction, a pressure is reached a t which the
azeotropic point vanishes. As an example, the mixture ethanol-water may be quoted ;
at 70 mni Hg this system no longer has an azeotropic point (cf. section 6.2.1). Hence
it is possible to prepare absolute alcohol by the distillation of dilute spirits at a
pressure below 70 mm, without the addition of auxiliary substances. The boiling
point, however, is then rather low (28°C). I n this connection it is obvious that the
choice of distillation pressure may also depend on the temperature of the cooling
medium.
Another motive for using reduced pressure in distillation - especially on an
industrial scale - may be that low-pressure steam is to be used as heating agent and
that its temperature would be too low for distillation a t normal pressure. In practice,
furthermore, the corrosion of a metal still can be a n important consideration, and on
this account it may be necessary not to exceed a certain temperature.
It thus appears that there are numerous technical and econoniic factors to be
considered in choosing the distillation pressure.
Billet and Raichle [ 1231 described a method of calculating column dimensions
which optiinizes a vacuum distillation in terms of minimum total pressure drop. The
reader will recall that in chap. 4.6.2 and 4.10.6 it was attempted to clarify, from
various points of view, to what extent reduced pressure influences the efficiency of
a column. Gelbe [124] has pointed out that the findings of the various authors are
still contradictory. Using a column of 45.7 mm diameter and 500 mni length (4 mm
spaced helices) he found that with constant throughput the efficiency is pra+icalIy
independent of pressure between 10 and 100 torr. As the pressure is raised to 740 torr
the number of transfer units gradually increases due t o the increase of the relativc
velocity. I n experiments with n-decane-trans-decaline he observed an increase by
about 15% (cf. Fig. 97). According to Gelhe the contradictions found in the litera-
ture are due to column flooding being carried out differently or in some cases even
being omitted, I n chap. 4.10.8 this was already mentioned and a method of flooding
described. In his distillations with operating pressures below 100 torr Gelbe eniployed
a higher initial pressure to obtain a higher reflux rate. What was important was that
the bubbling layer moved steadily upward from the column base through the pack-
ing. The layer of liquid formed a t the head was pressed through the packing several
times, then the pressure was lowered with the column remaining in the state of
bubbling. The optimum wetting of the packing thus achieved is clearly shown in
Fig. 186. The straight line f corresponds t o the optimum flooding conditionsde-
scribed.
The calculation of the column dimensions has been dealt within section 4.11. It
is important to ensure that the vacuum lines are sufficiently wide. The drop in

18*
iranchembook.ir/edu

276 5. Separating processes

pressure in a vacuum line can be calculated by Poiseuille's law if the pressure is

not too low and t,he tube has a diameter of less t,han 200 mni:

where p = the pressure drop in kg/cm2 (1atm. = 1.03 kg/cma);

V = the gas flow rate, cmg/sec;
-
77 = the dynamic viscosity of the gas in kg seclcm2
-
(1kg sec/cmz = 1.02 poises);
1 = the length of the tube, cm;
T = its radius, cm;

t = time, sec.

Fig. 186
Influence of flooding at 20 torr
(Gelbe [124])
2 3.10-'4 6 8 10' 2 d s3
mean vapour velocity, wD *

A useful nomogram (Fig. 187) interrelating these variables hrts been published
by Earries [125]. I t assumes that the suction is not choked more than 30%. The
followhg examples explain its use.
?i
1. Calculation of t h e maximum p u m p c a p a c i t y
a) Connect p i n t s on scale d end 1 to intersect A at A,.
b) Connect A, with p, and produce the line to meet scale 8,.
c) The reading on scale a2 is t.he maximum pump capacity.

2. Calculation of t h e smallest allowable d i a m e t e r of t u b i n g

a) Connect the p i n t p,, on scale p,, representing the maximum allowable pressure
drop, and the point on scale a,, representing the gas flow mte, by a line inter-
secting scale A at A1. Scale A shows a function of the resistance of the tubing to
gas flow.
b) Connect A, and the point on scale 1 corresponding to the length of the tube, and
produce it to meet' scale d. The reading on scale d is the smallest allowable dia-
meter.
iranchembook.ir/edu

5.4 Distillation pressure 277

3. C a l c u l a t i o n of t h e m a x i m u m l e n g t h of v a c u u m l i n e
a) Connect points on scale p1 and s2 to intersect scale A, as in 2a.
b) Connect the point A with the point on scale d showing the diameter of the tubing,
and produce the line to intersect scale 2. The reading on 1 shows the maximum
allowable length of vacuum line.

Fig. 187
Nomogram for the sizing of vacuum lines (Harries)

Example 1 Example 2
p 1 = pressure in vacuum vessel, mm Hg 1.0 1.0
d = diameter of line Cl3l 1.0 3.0
9, = gas flow rate m3/h 2.0 50.0
1 = length of line om 3 50 450
I n example 2 the intersection of pl-S, with A is 6.5 scale divisions above that
of d-Z with A. The vacuum line therefore has some capacity in reserve

The intersection of the two straight lines in example 1 corresponds to an efficiency

of 41%:
N = v
m - POIPI, (186)
where p o represents the minimum pressure a t which a vacuuni apparatus can be run
with sufficient efficiency.
For checking an existing vacuurn lay-out it Ruffices to connect points p , - sp
and d - I, respectively, by straight lines and to find thepointsof intersectionof these
lines with scale A. If tzhepoint of intersection of p , - s2 is situated above that of
d - 1 the vacuum installation is adequate. For more detailed calculations the reader
is referred to the original paper [125]. A vacuum system constructed of 20-30 mni
I.D. glass tubing is in generalsatisfactory in distillationlaboratories ; it may be built up
iranchembook.ir/edu

278 6. Separating proceeaes

of 1-2 m lengths of such pipe, connected by short pieces of rubber vacuum tube or,
better, by ball and socket joints.
The apparatus! for simple or countercurrent distillation at reduced pressures
differs from that used a t atmospheric pressure only by its wider dimensions (cf.
section 4.11). A few additional components are necessary, such as a vacuum oonnec-
tion, which is generally provided with a cold trap (Fig. 188).Besides packed columns,
empty columns (chap. 7.3.1) and columns with stationary (chap. 7.3.4) and rotating

Fig. 188
Vacuum connection with cold trap
a) For apparatus without control by
preasure drop
b) for apparatus with control by
preasure drop

Fig. 189
Vacuum tap having body closed on one side

Vacuum tap witrhmercury pocket

elements (chap. 7.3.5) are especially suited for distillation under reduced pressure.
The optimum ratio between the diameters of the inner and the outer tube of a cold
trap is 1.6. Fnrthemore, special vacuum receivers are required (section 7.6). It is
iniportant to measure pressure directly after the vacuum connection and to ensure
that there is no loss in pressure between this point, and that. where vapour temperature
is observed. To be on the safe side, a differential manometer can be connected be-
tween these two points (cf. section 8.3).
The standard ground joints normally employed for vacuum work are those of
series 1 ; in cases where the grease is rapidly dissolved away or high vacuum is re-
quired series 0 is also employed, which might be called high-vacuum ground joints
(chap. 3.1). Because normal glass taps are subject to leakage, special types have been
evolved for vacuum work. Fig. 189 shows a tap with the body closed on the one side
and Fig. 190 R similar type with a mercury pocket. Further special taps and valves
for vacuum work will he described in chap. 7.2.1.
iranchembook.ir/edu

5.4 Distillation pressure 279

Reduced pressure is produced by means of water-jet (filter) pumps, ejectors, and

various types of electrically driven mechanical pumps ; for high vacuum, mercury
and oil diffusion pumps are used. A description of these adjuncts would take us
outside the scope of the book. The reader is therefore referred to relevant books [113$
114, 119, 122, 126, 1271. Leybold's book [124] discusses the principle underlying the
choice of pumps, and von Ardenne [ 1281 reviews the operating ranges and the charac-
teristics of the more important pump types. All-glass Quickfit-Wiegand steam
ejectors [128a] have now become available. The main advantages of these pumps
are corrosion resistance and transparence. Two types are provided for operation
with or without water circulation. They require a saturation vapour and water
jressure of 3 bar. Two steam jets in series effect compression from 1 torr and 2.5 torr
to about 120 torr with types G. 01.1 and G. 01.1-K (0.1 kg/h air) and G. 05.2 and
(2.052-K (0.5 kg/h air), resp. The water jet condenser condenses the working vapour
and compresses air, gases and noncondensable vapours to atmospheric pressure.

Fig. 191
High-frequency leak detector with brush electrode
and ratio shielding

After the vacuum system has been assembled it is necessary to test it for efficienq-
and leaks. Testing should be carried out systematically and should start with the
pump, the capacity of which may be checked by connecting it to a buffer vessel of
5 -10 1volume. Valves and ground joints are then tested, passing on to the individual
coniponents, where fused connections often prove to be the cause of leakage. It is
tiseful - by including taps at suitable points - to construct the apparatus in such a
way that various sections may be tested separately for leakage. Testing for leaks can
be carried out with the aid of a high-frequency tester with a brush electrode working
according to the Tesla principle (Pig. 191). A t points where air is drawn in a luminous
spark breaks through. Leaks can also he detected aurally with a stethoscope: another
method is to apply a pressure of ahout half an atmosphere and dab soap solution
with a brush on suspected spots. An elegant method is to paint a weakly alkaline,
inethanolic solution of fluorescein or eosin onto the apparatus while it is under
vacuum; if i t is then irradiated with ultra-violet light in a darkened room, the leakages
hecome apparent by fluorescence.
For special high-vacuum testing methods see the books of Laporte [ 1191 and
Monch [126].
Devices for hunting leaks in vacuum and pressure equipment may be purchased.
Halogen leak detectors operate with Refrigerant 12 (CF,Cl,) as test gas. The principle
of operation is that incandescent platinum, in the presence of halogens, emits ions. If
an apparatus under vaouiiin is to be tested it is connected to the detector tube and
iranchembook.ir/edu

280 5. Separating procssaes

the apparatus is sprayed on the outside with the test gas. If the gm penetrates a t
any point this is indicated by a deflection on an indicator or by the emission of
a signal. For testing pressure apparatus the test gas is introduced into the equipment
and it is examined externally with the detector. The smallest leak that can be detec-
ted is about mm Hg l/sec (see eq. (187) below). Other devices employ hydrogen
or coal gas as testing medium and have the same sensitivity. The helium leak detec-
tom function according to the principle of the mass spectrometer and indicate leaks
down to 10-lo mm Hg l/sec [129].
The degree of tightness Di of an evacuated apparatus with the punip shut off can
be expressed by the formula

Di = mm Hg . llsec.
t
in which A p = the change in pressure in mm Hg;
V = the volume of the apparatus in litres;
t = the t,ime of observation (insec).
The value of Di should not, be greater than l O P to

5.4.2 Continuous equilibrium vaporisation (flash distillation)

One of the continuous processes employed on an industrial scale is continuous
equilibrium vaporisation - generally known as flash distillation. In spite of its
advantages this procedure has not been widely- used in the laboratory and in pilot
plants. A glass apparatus developed for this yiirpose by the author is illustrated in
Fig. 192 [ 1301. The most important componenK is the flask a of 250-600 ml capacity,
provided with a ground-in thermometer. It is placed in a thickly-lagged vessel and
is heated to a constant temperature, regulated by means of a contact thermometer.
Into the centre of the flask there projects an interchangeable injection capillary b,
which may be observed through a small window and is illuminated behind by a lamp.
The mixture to be distilled is forced from bottle c into the storage vessel d (operating
on the principle of the Mariotte bottle) and i6 brought to the required temperature
by the thermostatically controlled jacket. The feed rate may be checked in the
measuring burette e. The feedstock is heated further in heat exchanger f by a liquid
at constant temperature or by steam. The temperature of the heating agent is con-
trolled by contact thermometer g.
The preheated feedstock passes at a definite rate through the fine control valve h,
is sprayed into flask a (in which a reduced pressure is maintained), and forms a thin
f i h i on its walls. The release of pressure causes the low-boihg fractions to vaporiee
preferentially at once. Furthermore the walls of a (which can be heated to a higher
teinperature if desired) have a large surface, so that more of the light components
evaporate after a very brief interval of heating. The vapour passes through the
insulated tube i to condenser k, then as liquid, via the Anschutz-Thiele vaciiuiu
receiver ni to the collecting bottle 1. The vacuum punip is connected to n ; the pressure
is kept constant by a controller (cf. section 8.3). The high-boiling fractlionflowing
down from the flask is distributed by a perforated funnel in a short column 0. equipped
iranchembook.ir/edu

5.4 Distillation pressure 28 1

with a heating jacket. Any low-boiling constituents remaining in the residue are
removed in this column, which is supplied with heat for this purpose. The heating
jacket of the column is regulated by contact thermometer p. The bottom product
rims continuously into bottle q through the measuring device r.
Flash distillation has proved particularly convenient for distilling off low-boiling
“tops”; a t a pressure of 1 to 20 mm Hg a throughput of 800 to 1500 g/h niay he
attained. I n the separation, for instance, of a crude fatty acid mixture, a C,-C,,
fraction could be taken off with a feed rate of 1200 g/h a t 15 nim pressure.

Fig. 192
Apparatus for flash distillation (Krell)
iranchembook.ir/edu

282 6. Separating proceeses

Fig. 193 shows the distillation analyses of the distillate and residue obtained in
this sepamtion. A satisfactory cut proves to have been effected at Clo. The preli-
m h q fractionation of large amounts of a phenol oil total distillate was also carried
out by flash distillation. The cut was intended to lie between 210 and 230°C (normal
pressure). The result is given in Fig. 194. The separation performed in three stages
under different conditions yielded an overlap of about 15"C, as shown by the top and

Fig. 193
Distillation analyses of the top and
bottom products from a flash distillation
of crude fatty acids at 16 torr

Fig. lW
Flash distillation of a phenol oil total distillate. Analyses of the top
and bottom products

bottoni products analyzed in a test distillation at normal pressure. If the mixture to

he separated cont,ains fractions having large differences in boiling point, even higher
rates of throughput, up to about 3 l/h, may be realized.
A mixtiire of phenol containing 9% of water could be dehydrated in the ~ m i e
apparatus, with the omission of the column below the flash vessel. At a preesiire of
23 mi11 and a feed temperature of 80°C the optimum rate of input proved to be 4 1/h.
Such a large throughput could scarcely be attained in the laboratorj- by- the normal
method of distillation from a flask, if only on accountcof foaming.
Flash distillation has proved valuable in preparative work as a preliminary
oprration in the separation of midticomponent mixtures into their constituents.
iranchembook.ir/edu

5.4 Distillation pressure 283

Quantities of feedstock up to 100 1 may be split up in this way, a t a high rate and
without thermal ill-treatment, into fractions suitable for subsequent batch rectJi-
f ication.

5.4.3 Thin-film distillation

I n the method to be described now the crude material is not subniitted to distilla-
tion in the form of a thick layer of liquid (as is done in a boiler) but as a thin flowing
film ; the thermal treatment i t undergoes is consequently very mild, both as regards
time and temperature. The procedure may be regarded as a one-stage simple distilla-
tion with a maximum efficiencv of one theoretical stage. Arranging several stills in
series increases the efficiency. According to the author's experiments the procedure
can frequentJy he utilized with advantage a t any pressure in the range froin atmos-

1torr

I42 't, 5 torr

Fig. 195
Pressures in a stiil pot and in thin-film distillation

pheric down to 0.1-0.2 min Hg [130]. I t s favourable features can be deliionstrated

by means of an example. Suppose that a mixture of sat,urated straight-chain alco-
hols, Clo-C18, is t o be distilled a t 1 mm pressure. If the mixture were put into a
1-litre still pot and the d.epih of the charge were 54 mm, the pressures would be as
shown in Fig. 195. There would he a pressure of 1 mm Hg only a t the surface of the
liquid, whilst at the bottom of the still pot this would be increased hydrostatically
h y an amount of about 54/13.5= 4 nim, i.e. t o a total of 5 mm. At the beginning of
distillation C , , would evaporate under 1 mm pressure at 76"C, but the teniperatiire
in the bottom of the flask would be 142°C. In thin-film distillation no appreciable
hydrostatic pressure exists. A heated surface having a temperature of about 78°C is
hence sufficient to vaporise the first component. This example shows that thin-film
distillation reduces thermal hazards, so that the method is suitable for effecting a n
initial separation of temperature-sensitive substances.
When a high-boiling mixture contains a small amount of a volatile conyonent it
is often impossible, when a conventional still is used, to remove all the light fraction,
even by an increase in t e q e r a t u r e . If the pressure is reduced to prevent decompo-
sition there is the danger that, the volatile constituent will escape uncondensed. This
danger can be avoided by the use of thin-film distillation a t atmospheric pressnre.
iranchembook.ir/edu

284 5. Seperating processea

The applications of thin-film distillation may be summarized as follows (cf. also

[ 1221, chap. 5 and 0).
1. For a “mild” continuous distillation of temperature-sensitive materials.
3. For the distillation of high-boiling mixtures that cannot be separated in a normal
still.
3. For the continuous sepantion of mixtures containing a low-boiling fraction and a.
very high-boiling main fraction.
4. For the continuous degassing of liquids.
5. For the continuous evaporation of liquids.
6. For the continuous distillation of easily foaming substances.
The following types of evaporators for thin-film distillation or rather, evaporation
have to be distingnished.
1. Falling-film evaporators
- with vertical heated tubes, straight or spiral-shaped, down the outer walls of
which the liquid flows (Figs. 196, 198, 199, 212);
- with rotating elements for film circulation in the form of fractionating brushes
(Fig. 201), glass spirals (Fig. 210) or wiper systems consisting of brushes,
lamellae or rolls (Figs. 201, 202, 211).
2. Evaporating dishes, horizontal or inclined as used particularly for molecular
distillation (Figs. 205, 209).
3. Spraying stills, such as. the one described in the section on flash distillation
(Fig. 192).
5. Rotating stills
- apparatus with rotating still pots (Fig. 203);
- mixed-film drums (Q. 200);
- rotating-disk apparatus with centrifugal distribution (Fig.213).
An exhaustive discussion of all the problems associated with thin-film and flash
evaporation as well as falling-film columns illustrated wit.h numerous constructional
examples has been presented by Gemmeker and Stige [132]. Furthermore, the
d e r is referred to chap. 2 (secs. 1 to 8) of [122]. A review of the present state of
thin-film stills for laboratory and experimental use by Stage and Fischer [166] is
accompanied by impressive illustrations. The paper indicates the large variety of
applications of thin-film distillation.
Thin-film distillation is used for both the evaporation of solid solutions and the
partial separation of liquid mixtures. As shown by W a e l m and Mil3 [132a] the two
methods can be conibined in special cases. The possible combinations are compared
concerning the separating efficiency and heat requirement on the basis of an ideal
phase equilibrium of the volatile components.
The author’s thin-film apparatus, as developed further by VEB Glaswerk Stiitzer-
bach and illustrated in Fig. 196, was adapted from the early types of molecular stills
operating with a falling film. It contains two concentric tubes. The inner tube, which
ISinterchangeable by the provision of a ground joint, can be heated by an element a
and functions as a surface for evaporation. The outer tube acts as condensing surface.
iranchembook.ir/edu

5.4 Distillation pressure 285

As pointed out by Utzinger [133] the production of a uniformly thin, downflowing

film is a matter of considerable difficulty (see Chap. 4.2). The aut'hor has avoided
channeUing on the evaporator by introducing the feed into a weir, which distributes
i t evenly over the periphery of the cylinder. The end of the feed tube can be submerged
below the surface in this weir, so that the liquid is not broken iip into drops. The

Fig. 196
Krell's thin-film distilhtion apparatus

surface of the evaporating tube is roughened by sand-blasting, a condition also

favouring a n even film. Moveable metal rings 6 round the heating cylinder serve to
mix up the downflowing film; a protective funnel c over the distributing weir prevents
the liquid from spraying onto the condensing surface.
The apparatus is equipped with photo-electric control of the feed rate; the
temperature of evaporation is regulated by a contact thermometer. The distillation
area is surrounded by a cooled jacket, so that the procedure may also be employed
for substances of relatively low boiling point. The whole apparatus is operated from
a central panel. The material to be distilled is contained in bottle m of about 101
capacity. It is forced by air pressure through tubing to the supply vessel d, where it
iranchembook.ir/edu

986 5. Separating proceeses

is preheated. The feed rate is controlled by a float e which, by means of a photo-

electric cell, switches the pump off and on as necessary. A fine-control valve g on the
supply vessel allows the feed rate to be adjusted accurately. The distillate and residue
may be withdrawn from their respective receivers without interrupting the operation.
The apparatus can be used at normal and reduced pressures.
Experiments have shown that the rate a t which the mixture is introduced, at a
given temperature and pressure, has a great influence on the result. It the feed rate
is increased, the percentage of bottoms becomes larger, whilst if it is reduced the
amount of distillate rises and the separation becomes sharper. The best conditions
may be found in each case by keeping the pressure constant and varying the tempera-
ture of distillation.

240
“c
220
200

t 180
$160
+,
ga140 Fig. 197
Fractionation of products
E, 120 obtained in e continuous
* thin-film distillation of a C, -..C,
100
fatty acid fraction
80
0 10 20 30 40 50 60 70 80
-
fatty acid fraction C8.. . CZ5
%wt.
x = distillate, 0 = residue

If the composition of the mixture is known, the feed rate may be so adjusted that
the yields of distillate and residue correspond to the desired division. It is found,
however, to be preferable to work in two stages. I n the first thin-film distillation the
cut is purposely placed somewhat too high, so that the residue is free from the desired
distillate fractions. The tops are distilled in the same way a second time; the high-
b o i h g components carried over in the first operation then remain in the bottoms
and the distillate obtained is very pure. By the use of the graduated receivers the
separation ratios can be watched at frequent intervals (Fig. 197).
By continuous thin-film distillation fatty acid fractions up to an acid value of
90 could be separated off, so that the distillates were higher than CJo.The thm-film
procedure has also proved to be very suitable for fractionating silicone oils (which are
distributed with great uniformity on the heater) and for distilling waxes. Gutwasser
and Miiller [22] developed a thin-film evaporator which haa been successfully used
for the distillation of spermaceti-oil fatty acids. Fig. 198 shows the whole apparatus.
The principle of thin-film distillation has been developed further by Messrs.
Leybold-Heraeus QmbH, Cologne, to a so-called “mixed-film distillation procedure”.
This process takes place at pressures of 1mm or less on a large evaporating surface,
with special precautions to ensure that the film is energetically mixed, so as to renew
iranchembook.ir/edu

5.4 Distillation pressure 287

/1
r
I
1
I
I
I
%
77,
I r--

?
- _ _ _ cooling water
I - vacuum
r/
mi
.GI
4
1 '
heat carrier
I ,
$1
t o vacuum t o vacuum
pump pump
Fig. 198a)
Apparatus for thin-film distillation (Gutwasser and Moller [ 2 2 ] )
1 = still pot, 2 = thin-film evaporator, 3 = heating jacket, 4 = vapour-liquid
distributor, 5 = overhead condenser, 6 = safety condenser, 7 , 8 = therino-
meters, 9 = thermostat, 10 = vacuum gauge, 11, 12 = connections to thermo-
stat, 13, 14 = cold traps, 15 = ionization gauge, 16 = mercury diffusion pump.
17 = oil manometer

Fig. 198b)
Thin-film evaporator with standard ground joint
NS 70
I = main evaporator tube, 2 = condenserswith
additional evaporators, 3 = distributor for top
product, 4 = glass disc for distribution, 5, G = seals,
7, 8 = temperature measuring points,
9 = ball joint for receiver, 10 = condenser for
distillate, 22, 22 = connections to thermostat.
13 = vacuum connection
iranchembook.ir/edu

288 6. Separating prowsaes

the “active” surface. The following are the main forms of apparatus that have been
evolved.
1. A “mixed-film” column (Fig. 199),
intended for liquids giving a residue that is still capable of flowing at the dist.dstion
temperature e. g. high-molecular-weight esters and mineral oil fractions, scents,
monoglycerides, plant extracts etc. It should be noted that the vapour is taken off
at right angles to the direction of liquid flow. The trickle evaporator 11is employed
in cases where the product tends to cause caking of the column packing.

Fig. 199
Mixed-film columns

............ 4
4
Fig. 200
Mixed-filmdrum

2. A “mixed-film drum’’, (Fig. ZOO),

which can be used for substances with viscous to solid residues, such as coal tar
pitch, natural and synthetic waxes, shale oils etc.
3. A “fractionating brush” (Fig. ZOI),
intended for suhstances giving residues with a viscosity up to 800cP at the
temperature of distillation, or distillates with the same viscosity at the tempera-
ture of condensation, for instance tallow oil, chlorinated di- and terphenyls, wool
gwase [l46].
For these main forms components were designed which may be assembled to
give midtistage thin-film distillation apparatus for degassing and distillation a t
various temperatures. Further development hm led to short-path distdation units
with internal condensation and rotating wipers, which may also be combined t o
operate as multistage apparatus (see chap. 5.4.4).
iranchembook.ir/edu

5.4 Distillation pressure 289

Falling-film evaporators with rotating elements resemble columns with rotating

elcments (chap. 7.3.5). The basic difference is that with colunins evaporation takes
place in a normal manner in a flask and there exists a countercurrent of vapour and
liquid for material transfer so that a considerably higher separating efficiency can be
obtained with these columns. In both cases, however, the liquid film is made to cir-
culate by the rotational motions. Thus, the low-boiling component is prevented from
being depleted a t the film surface. The problem of mechanically induced areal
turbulence in thin films of liquid was dealt with by Janosfia [ 1341.
Palling-film evaporators may he used for simple and countercurrent distillat ion
a< well as for short-path evaporation (see chap. 5.4.4). Heating is mostly done via
thc onter wall (Figs. 198, 200, 201, 202, 203, 211) but the opposite design is also
employed (Figs. 196, 199, 210, 212).

00 heaters
condensation
material

Fig. 201
Fractionating brush

Thin-film evaporators have proved particularly useful on a semi-technical scale

[IS5]. The performance of various types of evaporators with rotors and the causes
and degrees of resistance to mass transfer were studied by Dieter [136]. Billet has
reported methods for the mathematical treatment of the distillation process in thin-
film evaporators with rotating elements [ 1371.
Fig. 202 shows the Sanibay glass-made evaporator of QVF-Glastechnik, Wies-
baden-Schierstein, as a n example of a falling-film apparatus with rotating wipers.
The feed is supplied to the evaporator from a jacketed dropping fnnnel ( A ) .
The feed rate can be controlled exactly by means of a needle valve ( B )and the funnel
itself which works on the principle of the Mai-iotte bottle. Before entering the eva-
poration zone the feed is passed through a heated coil ( C )and heated to about boiling
temperature. Thus, full use can be made of the evaporation zone for actual evapora-
tion. Besides, the product is already degassed in the coil and spraying in the eva-
porator tube is avoided.
The liquid film is formed on the calibrated inner glass tube (D)by the action of
the rotating wiper system having movable lamdlae ( E ) . (For corrosion resistance
these metal parts are made of tantalum or special steel.) The rotor is operated ria a
magnetic coupling ( F )by a variably controllable drive (G). Thus the disadvantages of
a stuffing box are avoided. The lower bearing is of the pendulum type with a Teflon
ball placed in a glass bearing. It is lubricated by the bottom product. Heating is
19 h i e l l , Handbook
iranchembook.ir/edu

290 5. Separatingprocesses

done by an electrically heated circulatory thermostat (1.5 or 2 k W ) (H),using paraffin

oil (up to 20OOC) or silicone oil (ahove 200°C) as heating agents. Both oils are com-
pletely transparent.
The apparatus is completed by a receiver for the bottoms ( J ) and a transitian
piece (K) with a connection to the vacuum gauge. For deodorization work or for the
processing of mixtures with a small proportion of low-boding material the transition
piece ( K ) is replaced by the deodorization piece (0). Through the lateral inlet the
vapour of a low-boiling product csn be passed counter to the f&ng liquid film in
order to enhance the flow toward the condenser and to prevent recondeneation of
the distillate. A siphon prevents condensation in the hottoms receiver.

I
f 7 \

Fig. 202
Sambay distillation apparatus made of glass
iranchembook.ir/edu

5.4 Distillation pressure 39 1

The low-boiling phase is condensed in a large-dimensioned condenser (L)and

collected in a receiver ( M ) . A transition piece ( N ) provides the vacuum connection.
The heating surface area of the apparatus is 0.016 1 3 , the condensing surface area
of the condenser is 0.2 ni2. The maximum working teiiiperature lies around 220°C
so that a t 25 torr the maximum rate of evaporation is 480 g/h.
On account of its flexibility, apparatus with rotating evaporators has become
standard equipment for laboratory and semi-technical work, particularly because its

1 = rotating still pot, 2 = heating bath, 3 = distillate receiving flask,

I = vacuum connection, 5 = condensers, 6 = switchboard

throughput actually covers the range from micro to pilot-plant scale. Thus there
are rotating hiilb tubes (chap. 5.1.1) for < 1 nil and 100 1 still pots for operation on a
seiiii-technical scale. Besides the degassing of oils and resins, units with rotat inp eva-
porators are preferably used for the mild separation of solvents and are especially
suited for foaming substances. Fig. 203 shows the operating principle, and Egli's
[ 1381 paper offers a discussion of constructional details and applications. The eva-
porator 1 is variably controlled in the range of about 10 to 220 r.p.ni. It is provided
with mechanical rapid siphons and automatic siphons. Kramer [ 1%a] has discussed
the passible applications of a system of rotating evaporator components. Sorbe [138h]
describes the IKA-DEST system which opens up a wide field of applications through
nianifold combinations of various condensers.
The LRV2 laboratory rotating evaporator of VEB Carl Zeiss, Jena, can be
provided with still pots of 500 to 2000 nil capacity. The heating bath can reach 90°C.
iranchembook.ir/edu

292 5. Separtlting procews

5.4.4 Molecular distillation

In molecular distillation, which is applied to high-boiling substances (mainly of
a temperature-sensitive nature), the material is distilled at a pressure < torr
in apparatus constructed in such a way that the distance travelled by the molecules
between the evaporating and condensing surfaces is shorter than their mean free
path. The purpose of this arrangement is that the majority of the evaporated mole-
cules reach the condensing surface without being deflected on collision with foreign
gm molecules. The mean free path is the theoretical concept of the distance a mole-
cule can travel without colliding with another molecule. For the normal t,riglyceride
fats with a molecular weight of 800, e.g., it assumes these values [141]:
distillation pressure mean free path
8x t’orr 7 min
3x torr 25 mm
1 x 10-3 born 50 m.
The process going on in molecular distillation is not the normal ebullition; it mlght be
called “molar evaporation”. The equilibrium between evaporated molecules and the
liquid is continually disturbed by condensation so that, in accordance with physical
lams, equilibrium has to be re-established. This means, however, that more molecule8
will evaporate from the liquid surface. Thus, we have a true example of a simple
distillation which is also termed “one-way evaporation” [lal]. The whole field of
molecular distillation is covered by the books of Burrows and Ho16 et al. [139].
Ridgway-Watt [ 1401 presents an introductory survey of apparatus from the micro
to the technical scale. I n addition, the reader is referred to chap. 1.5 of [122] and
some more review articles [108, 131, 145, 156, 1571.
The rate of evaporation depends on the vapour pressure p* of the substance at
the temperature of the evaporating surface T,and on the molecular weight ilf of the
substance to be distilled.
This relationship is expressed in Langmuir’s equation for the rate of evaporation
[1421 :
D = 0.0583p, W T (188)
in which
-
D = the rate of evaporation (g - cm-2 sec-1) ;
T = the temperature of the surface (K);
p , = the vapour pressure of the substance (mm Hg) at 1’;
64 = the molecular weight.
This equation aasumes that evaporation is not impeded by foreign gas niolecdes.
Sine it is inevitable that some evaporated molecules w i l l collide with those of residual
gm before arriving at the condensing surface, the value of D given by this formula is
not normally attained. It is therefore necessary to COReCt D by multiplying it by a
factor a, which approaches unity the more closely, the lower the pressure of the
residual gas. Thii factor a can amount to 0.9 in modern indristrial apparatus.
The validity of equation (188) has been examined by Burrows [143], who has
iranchembook.ir/edu

5.4 Distillation pressure 293

derived a number of semi-empirical formulae giving better agreement in various

practical conditions.
The amounts of distillate theoretically obtainable are quite miall, as will be
apparent from the figures given as examples [ 1081 in Table 47.
Tf we take the correction factor a, to be 0.8 and multiply the figures of the third
colicnin by 3.6 x 105 (to give the amount of distillate in one hour from an evaporating
Yurface of 100 em2),we obtain quantities in the range of 15 to 22 grams. To realize a
higher throughput, Utzinger’s method of “short-path distillation” [ 1331 or, Jaeckel’s
“free-path distillation” [144] are suitable, (the latter of which is chiefly utilized on a
or industrial scale) with pressures higher than
-c-~iii-technical mni. These processes
rely on the increase in distillation rate brought about by a higher pressure, which is
5hon t i by Langmriir’s equation, and compensate for the shortened mean free path hy
q1,roprrate design of the apparatus.

Tnhle 47
Theoretically obtainable amounts of distillate, at a saturation pressure of torr, from
hubstances n i t h molecular weights of 284 t o 891

S.;rthstnnce Xol. D D P, at VWT P,/@

wt by formuld (188) by formula (188) 120°C at a t 120°C
( g . m - 2 . sec-l) (mol em-?. sec-’) (torr) (torr)

Stearic acid 284 0.52 X lo-* 0.21 x 10-0 36.0 0.90 2.07
Cholesterin 387 0.56 > lo-* 0.14 x lo-’ 0.5 0.97 0.025
Tristenrin 891 0.76 ; 10V 0.09 x lo-” 10-4 1.32 -

It should he noted that a pressure of less than quoted as being necessary for
inolecnlar distillation, applies only to the residual gas. The vapour pressure of the
substance distilling may be considerably higher, up to about 1 mm. Only the mole-
cules of the residual gas rebound from the condensing surface; the niolecules of the
vapour are retained hy this surface [145].
The applications of molecular distillation are nuinerous and lie chiefly in the
field of temperature-sensitive substances having molecular weights of 250 to 1200.
The following examples have been chosen from the many that might have been given
(cf. [122], chap, 5 and 6 and [158]:
the preparation of vacuum pump oils and viscous lubricants with a flat viscosity
curve ;
the investigation of triglycerides (oils and fats) and high-molecular-weight fatty
acids, fatty alcohols, waxes and residues;
the separation of vitamin and hormone concentrates;
the piirification of plasticizers and other substances of low volatility ;
the pnrification of essential oils and scents;
the deodorization of materials of high molecular weight.
Of late molecular distillation has been employed for the investigation of the
high-molecular-weight components present in the residues of crude oils and similar
iranchembook.ir/edu

294 6. Separating processes

materiala. Recycling is employed to improve distillate purities. A survey of apparatus

and methods wed in the molecular distillation of fatty acids and lipids has been
given by Perry [145]. Frank [146] deals with special problems and, in particular,
describes multistage stills.
An essential feature of molecular distillation is that it is capable of separating
substances having the same vapour pressure, but differing in molecular weight. If we
examine Langmuir’s original equation

P
D= @zEF
in which D = the maximum amount that can be evaporated, in
moles . -
sec-l;
-
p = the vapour pressure, dynes cmr2;
34 = the molecular weight ;
R = the gas constant, 8.3 x lo7ergs/K. mole;
T = the temperature, K,
we observe that at constant temperature the amount evaporated is dependent only
on p / m . By analogy to the relative volatility we can therefore write the relative
quantities that can be evaporated as

where p , and p , represent the partial presmres of the components [108]. Substances
for whichp, and p, are identical can therefore be separated if MIand M , have different
values.
The apparatus employed for molecular distillation should coniply witfhthe follow-
ing requirements:

1. the vacuum system should be wide in bore, so that pressure differentials are
avoided (see chap. 5.4.1) ;
2. the liquid should be evenly distributed as a thin film and its residence time (cf.
chap. 6.4.3)should be short;
3. the distance between the evaporating and condensing surfaces should not be
greater than the mean free path (1 -2 cm); the condensation temperature should
be about 50-100°C below that of evaporation;
4. the substance to be distilled should undergo a preliminary degassing to minimize
the amount of uncondensable gas present.
The hypes of apparatus that have been developed so far may be divided into the
following groups according to their principle of operation :
a) flat-bottomed stills containing a thin film;
b) apparatus having horizontal or inclined trays as evaporating surfaces:
c) apparatus with a film descending vertically;
t i ) centrifugal apparatiis.
iranchembook.ir/edu

5.4 Distillation pressure 295

It should beobserved that’ recycling can be practised in all these forms of appara-
tus in order to separate low-boiling components effectively; this is often necessary
in view of the fact that niolecular distillation effectively corresponds to a single
equilibrium stage. We must not forget, however, that in a distillation from a flask
mtrainment may constitute a complication ; in molecular distillation, on the other
hand, evaporation takes place a t the surface only, so that the molecules can leave it,
selectively without mechanical disturbance [ 1471.

Fig. 204
High-vacuum niolecular still
for charges of 2, 5 or l o ml,
temperature range 20 to 200uC,
vacuum to about 10.‘ torr, with
support and temperatore
control device

The oldest fornis of apparatus are those having a flat-bottomed still pot con-
taining a thin film of liquid. Fig. 204 shows a modern construction developed by
Fischer. This type of still is well suited for dealing with substances having molecular
weights up to 300, and are mainly used for obtaining preliminary data on the boiling
range and the tendency to decomposition of the material in question. Methods
employed in micromolecular distillation based on the cold-finger and the falling
film principle for throughputs of 0.5 to 5.0 g were described in chap. 5.1.1 (Figs. 131
to 136).
A good example of an apparatus in which the evaporation takes place from a
tray is the arrangement of Utzinger [133, 1471, who eniploys the term “short-path
distillation” for one-way distillation at pressures above nini Hg. The apparatns,
iranchembook.ir/edu

2
s S. Separ~tingprooeeses

dating from 1943, has been developed further for continuous fractionation. Fig. 205
shows it in a recent form.
The previously degassed feedstock emerges from the degassing flask through a
capihry tube into the “still” and flows as a thin film over a tray of adjustable slope,
heabed by the circulation of high-boiling mineral oil. A temperature gradient is here
established, the temperature increasing in the direction of flow. The vapours are
condensed by a cooler surrounding the tray, inclined a t the same angle as thelatter.

Pig. 205
Short-path distillation apparatus
for three distillate fractions
sect ion CaO (Utzinger)

The condenser is subdivided into three sections and there are take-off tubes, three for
distillate fractions and one for the residue.
A small difference in pressure between the degassing flask and the distillation
space - both of which are initially connected to the same pump - allows the tray
to be charged with feedstock and the rate of admission to be adjusted as required.
Furthermore, the rate of flow of the liquid on the tray can be controlled, even
during distillation, by altering the slope; to do this, i t is rotated around a conical
ground joint. These two variables are necessary for establishing the desired fractionat-
ing ratio.
The heating medium h brought into circulation and warmed by heating the tube
leading to the entry jet with a Bunsen burner or electric element. The temperature
iranchembook.ir/edu

5.4 Distillation pressure 297

gradient in the tray depends on the rate of circulation of the heating oil, on the feed
rate. the downflow rate on the tray, the composition of the feedstock and the pump-
ing system. I n the glass apparatus described here temperature differences of 10 to
30 deg. C occur between the beginning and end of the tray. The distillation tempera-
ture is measured indirectly by submitting small samples of the feedstock to the nsual
boiling point determination a t the same pressure. The temperature of the heat iiig
niedium usually lies about 60 to 80 deg. C above this distillation temperature; the
evaporation in the film is thus accoinpanied by a marked cooling effect. The average
tliroiighput iq approsimat~ly100 ml/h if 50-70 ml/h of distillate is taken off.

Fig. 206
Micro-cascade apparatus for short-path iind molecular distillation
iranchembook.ir/edu

298 6. Separating proceases

Utzinger’s device has been developed further by the VEB Jenaer Glaawerk
Schott 8c Gen., Jena, Germany, who have constructed a macro- and micro-<tlbectlde
apparatus (Fig. 206).Since this device is built up of sepamte pasts it can be set up
for any number of fractions. Every step of the cascade may be adjusted to a different
temperature by means of its circulatoryheater. Degsssing is carried out independently
in the apparatus shown in Fg.207. In this way account is taken of the fact that
degassing and the actual distillation frequently have to be performed at different
rates.

Fig. 207
Utzinger’s degasser

ti90
0 80
c 70
6 60
5 50
2 40
30
8 20
g0 100
0 10 20 30 40 5060 70 80 90100
distilled,% --+

Fig. 208
Comparison of separating efficiencies according to Frank [146]
Molecular distillation with the test mixture:
di-(iso-octyl-)sebacata (DIOS), molecular weight 426
di-(iso-octyl-)phthalate(DIOP), molecular weight 391
Single-stage apparatus: - - - - -
Ten-atage apparatus:
iranchembook.ir/edu

5.4 Distillation pressure 299

Nelpolder et al. [148] have evolved a 20-stage apparatus for molecular connter-
ciirrent distillation, intended for the separation of high-boiling components of
petroleum. A 10-stage still described by Ridgway-Watt [140] rotat,es slowly in an
inclined position. Thus the product is distributed over the whole evaporator surface.
After condensation in the first condenser stage it is fed into the second, and so forth
until the product is withdrawn. Fig. 208 [146] demonstrates the separating efficiency
of a single-stage still (Fig. 204) as compared with that of a 10-stage apparatus after
[ 1401. Methods for calculating the number of theoretical stages in multi-stage devices
have been worked out by Malyusov e t al. [lag] and Shavoronkov et al. [150].
Fig. 209 shows the apparatus used by those authors.

u u
Fig. 209
Multi-stage apparatus for molecular distillation (Malyusov and Shavoronkov
ciw)
1 = tube vacuum, 2 = cooling water, 3 = electric heater, 4 = product feed,
3 = distillate removal

The principle of the falling film is that which has been most generally utilized
(cl. section 5.4.3).The very short residence time of such a falling film - amounting to
3-4 seconds in laboratory apparatus and as little as 1/1000th of a second in soiiie
industrial installations - ensures a very “mild” thermal treatment of the substaim
being distilled. I n laboratory molecular stills the thickness of a falling film may he of
the order of 0.1 t o 0.2 mm, corresponding to 50,000 molecular layers, whilst film
thicknesses down to 0.001 -0.005 mm (around 400 inolecdar layers) can be estab-
lished in large-scale units. The ideal solution would be mono-layers.
liiiportant factors in molecular distillation are the character of the evaporating
surface and, of course, the distance between this latter and the condensing surface.
For falling-film apparatus a good wetting of the evaporator surface and effective
circulation of the liquid film are achieved by means of the same constriictional
nieasures as those described 011 p. 286 [122].
As an example of a spiral still Fig. 210 represents apparatus for molecular distilla-
tion manufactured hy VEB .Jenaer Glaswerk Schott & Gen., Jena. The glass spiral 3
rotates around the heater 2. Thus a film thickness of about 0.1 mm and a good circii-
iranchembook.ir/edu

300 5. Separating processes

Fig. 210
Xolecular distillation apparatus MDR 600
1 = still with condensing jacket, 2 = evaporator, 3 = glass spiral, 4 = motor for spiral,
5 = thermocouple, 6 = electromagnetic valve for change of receivers, 7'a = receiving flask
for rmidue, 7 b = receiving flaek for distillste, 8 = cold trap for still, 9 = high-vacuum oil
diffusion pump, 10 = degasser, 11 = control valve for feed, 12 = storage vessel, 13 = de-
gasser cold trap, I4 = two-stage rotary-valve vacuum pump, 15 = one-stage rotary-valve
vacuum pump, 16 = vacuum stop valve, 1 7 = universal laboratory thermostat,
18 = switchboard
iranchembook.ir/edu

5.4 Distillation pressure 30 1

lation of the film are obtained. The residence times are as short as a few seconds.
These technical data are given by the manufacturer:

nominal throughput 1000 g/h

throughput1) 250. * -2000 g/h
rate of evaporation1) 18,000g/h (maximum)
spiral velocity approx. 40.. -90 r.p.m.
evaporating surface area approx. 600 cm2
distillation temperature 300 "C (maximum)
working voltage 3801220 V
power consumption 2 kW
condensing water consumption approx. 350 l/h
cooling agent for freeze traps liquid air (nitrogen) :
in some cases: C0,-acetone
mixture

The KDL 4 laboratorj- short-path distillation apparatus of Leybold-Heraeus

GmbH & Co KG, Cologne, was designed on the assumption that in general the first
step in a molecular distillation is the degassing of the product which is followed by
several distilling stages at different temperatures. The wipers consisting of Teflon or
Teflon-coated ceramic rollers were placed between the inner condensing cylinder and
the heated outer wall. The K D L 4 is a multistage apparatus which combinPs a
number of short-path stills made of glass or special steel. The continuous operation
is illustrated by Fig. 211. Pressiires of torr can be reached. The mininiuin through-
put is about 100 g/h. A more modern apparatus (KDT 6) of the same manufacturer
is designed to handle extremely corrosive substances. Hence glass has been used as
a constructional material alniost exclusively. The throughputs lie around 10 kg/h.
Constructional and operational aspects of short-path distillation equipment on a
semi-technical and technical scale were discussed by Habendorff [140al.
Some of the devices based on a falling film are equipped with an arrangement for
recycling, as shown in Fig. 212. The circulation is brought about by an electromagne-
tic pump and the liquid is generally injected onto ,z ring of wire gauze to produce an
even distribution [152].
The main problem in designing an apparatus of the falling-film type is to prevent
channelliig on the evaporating surfaces. The following measures have been adopted
for this purpose:
1. roughing the surface of the evaporator by sand-blasting, etching, or sintering glass
powder on (Pigs. 196, 205, 206) ;
2. the provision of a wire-gauze ring, or a closely-fitting concentric cylinder of wire
gauze or glass cloth at the point where the feedstock is admitted (Figs. 196, 212) :
3. the use of an evaporating surface having wave-shaped or spiral extrusions (Fig. 210) ;
4. the introduction of the feedstock by a rotor with or without wipers or the rotation
of the whole evaporator (Figs. 200, 201, 202, 203, 211).

1) depending on the substance processed

iranchembook.ir/edu

302 5. Separating processes

The principle of producing very thin films by centrificgal distribution has been
developed mainly by Hickman [163] for apparat,us of various sizes up to large
industrial unit.s [ 1511. Fig. 213 illustrates a laboratory apparatus arranged for circu-
lation. The mixt.ure to be distilled is introduced by pump a from the storage vessel e

Fig. 21 1
Laboratory apparatus for short-path distillation KDL 4 with roller wipers
1 = metering vessel, 2 = drive for wipers, 3 = cold trap, 4 = glass-made part
of still, 5 = metal part of still, 6 = circulating-type thermostat, 7 = graduated
vessel (section), 8 = sockets

onto the rot-ating,heated disk b, and is there distributed by centrifugal action into a
thin film. The condensing surface c is a short distance from the disk. The distillate
may be removed directly or returned to the storage vessel e according to choice. The
residue is collected by the device f and led to the receiver d, from which it can also be
pasxed back to the storage vessel for further treatment.
In their conxiderations on molecular distillation Hickman [152] and Einbree [I541
introduced the concept of the “distillability”. This term is used to denote the ratio
iranchembook.ir/edu

5.4 Distillation pressure 305

Fig. 212
Molecular distillation apparatus of Hickmann
with arrangement for recycling

Fig. 213
Apparatus for molecular distillation with centrifogal distribution (Hickrntinn)
iranchembook.ir/edu

304 6. Separating processes

between the number of molecules of a substance A leaving the evaporating surface in

unit time, and the number of molecules of A remaining in the fiLm under identical
conditions. By circulating repeatedly one can finally obtain A almost completely in
the distillate. The time of distillation can be shottened by increasing the temperature
of evaporation. An “elimination curve” may be obtained by distilling the mixture,
under the same vacuum, a t successively increasing temperatures (say a t intervals of
10°C) and determining the composition of the distillate. Fig. 214 shows the typical
shape of a n elimination curve. The concentration rises to a maximum and subsequent-
ly falls back to zero. The course of the curve naturally depends on the properties of
the components - especially the heat of evaporation - and on the residence time,
which for purposes of comparison must also be kept constant. The niaximiini in the
elimination curve corresponds approximately to the boiling p i n t in normal distilla-
tion. If the residence time increases, the elimination curve shifts in such a way [169]

tl I

i Fig. 214
Elimination curve

-
9 = short residence time
B = long residence time
temDerature

9%
-<G 65
-
P4
$ 3
4-

z2
180 160 140 120 100 80 OC 60
temperature 4

Fig. 215
Relative volatility of the test mixture EHP/EHS
EHF’: Kp. 0.5 torr 183-1Y4OC; EHS: Ep. 0.5 torr 199-202°C
I: From an equilibrium apparatus of flask type
11: from an apparatus operating on the falling-film principle

that its maximum lies a t a lower temperature (curve B in Fig. 214).To obtain the
optimum throughput in molecular distillation the influence of various parameters
should be studied, the results being expressed in the form of such elimination curves.
Theoretical investigations of this problem were carried out b y Gorriz et al. [155].
The method has been developed further with the aid of standard suhdances,
mostly volatile colouring matters capable of being determined colorirnetrically [1081.
Carriers are also frequently employed with the object of achieving a good distribution
of the film when the atnounts of the substances become small.
iranchembook.ir/edu

5.4 Distillation pressure 805

Recent investigations of Hickman and Trevoy [la11 on evaporation in a high

vacuuin and molecular distillation have yielded noteworthy results. The evaporation
coefficient of a pure substance may be changed appreciably by the presence of n i i -
purities in very sinall quantites. I n studying a test mixture, di-2-ethylhexyl-phtha-
late-di-2-ethylhexyl-sebacate (EHP-EHS), the investigators found that the
relative volatility is highl? dependent, on the operating conditions, as demonstrated in
Fig. 215.
Therefore, Malyusov et al. [160], in their measurements a t a pressure of torr,
distinguished between non-equilibrium (aneq) and equilibrium separation factors (e,,).
Aisthe temperature rises the ratio approachesunity, as is confirmed by Fig. 215.
The test mixture di-2-ethylhexyl phthalate-in-tricresyl phosphate (EHP-iiiTCP)
was studied a t < 1.5 x ~ O -torr ~ (2 x mbar) in the range between 130 and
I6O'C by Uyeha and Hagihara [163]. I n the lower part of this range the niixture
rlisplayed an ideal hehaviour for niole fractions of 20 to SO:/,.

5.4.5 Pressure distillation

Mixtures that are gaseous a t room teniperature and higher may be distilled a t
pressures above atmospheric with water as cooling medium. The expense of a low-
teniperature cooling agent is thereby saved: on the other hand, special apparatus
capable of withstanding and maintaining the increased pressure is necessary. Pressure
distillation inay also be useful in cases where the relative volatility increases with
riring temperature, for example in that of the mixture 2.4-dimethylpentane-2.2.3-
triinethylpentane, which a t atmospheric pressure exhibits a difference in boiling
1)oint of only 0.2"C, but has a value of a increasing from 1.006 at 80°C to 1.055 at
200°C: (cf. section 4.6.2).
Relatively few descriptions of laboratory apparatus for pressure distillation have
been published. Glass cannot normally be employed for pressures higher than 5 atni.
Schneider has described metal equipment that can withstand up to 45 atni; it is
provided with a 3-litre boiler and a column of 25 mni diameter [161]. The apparatus
of Sinions is made of glass and is designed for pressures up to 5 atni [ 1621. Its iiiani-
pulation is as follows (see Fig. 216). Conipressed air is admitted at A . An arrangement
for keeping the pressure constant consists of valve Bywhich is fdled with mercury and
contains a float C, terminating in a cone D that can close the opening E . By an
adjustment of the mercury level with the aid of the levelling bulb F , float C main-
tains the pressure a t the required value. If this value is exceeded by the conipressed
air admitted, float C is forced down and air can escape through S to the atmosphere
until equilibririm is re-established. The mixture to be separated is put in still pot f1
and heated by the electric element J . Column K may be filled with any type of
packing; the vapoiirs are liquefied in condenser L. The condensate flows into funnel
X , which has a rinall hole through which the reflux returns to the column. Thc
distillate passes through the capillary tube N , where it is evaporated by a heater.
The vapour is again condensed in P , the condensate flows down through R and
collcTc+s above valve S , which is of the same type as that previously described. This
valve is opened or closed by lowering or raising the levelling bulb T. Part of the

20 Krell, Handbook
iranchembook.ir/edu

306 5. Separating processes

Fig. 216 Fig. 217

Apparatus for distillations under Pressure tap with retaining
preseures up to 5 atm (Simons) spring

distillate remains floating on the mercury in S, another part flows directly to V . If

valve 8 is closed and valve W opened the pressure below S is reduced and the fraction
evaporates. It is re-condensed in receiver X,which is cooled by liquid air. In this
way mmples may be taken without any alteration of the pressure in the appamtus.
Normal glass laboratory apparatus can as a ride be used for distillation at a
pressure up to 2 atm if precautions are taken. All ground joints should be secured by
strong spring clips. Fig. 217 shows a pressure tap, the cone of which is kept in its
seating by a strong spring.
iranchembook.ir/edu

6. Selective separating processes

By this term we denote processes in which, as the result of the addition of a

substance or substances in the gaseous, liquid or solid state, a distillative separation
is favourably affected (or sometimes even rendered possible). The term also includes
procedures involving a reaction that gives rise to new substances, which are removed
by the distillation in the same operation. Furthermore it can be taken to cover
methods which by a combination with another procedure, for instance chromato-
graphy, supplement the process of distillation in specific cases.
According to Gibbs’ phase rulz a completely soluble binary mixture is enriched
in both phases, whilst an inimiscible binary mixture, with its three phases, cannot be
enriched (see Fig. 29, a-d). It w il
l be recognized, on the other hand, that three-
component systems having a miscibility gap, i.e. showing two liquid phases and one
vapour phase, are separable by countercurrent distillation [l]. A typical example is
the preparation of absolute alcohol by azeotropic distillation with benzene.

6.1 Carrier vapour distillation

Th? most familiar example of carrier vapour distillation is the distillation of

high-boiling materials wtth stenrrr. The use of a carrier vapour has the double object
of separating the volatile from the non-volatile components with a reduction in
boiling point, and of thereby avoiding thermal decomposition. The process takes place
under conditions comparable with those existing in vacuum distillation. The volatile
substance passes over a t a n effective pressure p , that is lower than the total pressure
pgesof the system. The difference pges-pl is the partial pressure p2 of the carrier
vapour. I n normal distillation a t a reduced pressure it is necessary either to increase
the temperature or to reduce the pressure during the course of the distillation. I n
steam distillation the same effect is obtained by an (automatic) increase in the pro-
portion of water vapour [2].
Steam distillation is particularly valuable for purifying and separating substances
partially or totally insoluble in water, such as ethereal oils, fatty acids, fatty alcohols,
aniline, tallow oil, waxes, fractions of petroleum and tar etc. A further advantage of
this process is that the steam displaces atmospheric oxygen and hence protects thc
material from oxidation. Steam distillation is much used in preparative work for
dealing with gummy or alkaline reaction products.
As has been pointed out in section 4.3 and 4.5, the two components in steam
distillation behave as though each of them were present alone a t the existing tempera-
ture, provided they are immiscible. The total pressure acting on the boiling mixture

20+
iranchembook.ir/edu

308 6. Selective separahg processes

is the sum of the partial pressures of the components, one of which is water. The
partial pressures and the boiling point of the mixture may be determined, as described
in section 4.5 by means of the additive formula. A more convenient procedure i8 to
draw a diagram according to the method of Badger and McCabe [2]. In this diagram
one first plots the vapour pressure curves of the substances to be distilled (Fig. 218).
Next, one draws curves for the values of the total pressure minus the vapour pressure
of water. For instance, water has a vapour pressure of 1 5 0 m a t 60.1"C; thus for
the curve for atiiiospheric pressure one plots 760 - 150 = 610 nini against 60.1 "C.

Fig. 218
Diagram for determining the boiling point and the partial pressures in steam
distillation (Badger and McCabe)

The point of intersection of the latter curve with that for the vapour pressure of the
substance gives the tanperatwe of the mixed system and the partial pressures of the
components. (The diagram is sometinies drawn on a logarithmic scale.)
From the partial pressures, the vapour composition can be calculated by formula
(40).The proportion h p weight of the component to carrier steam is determined by

_
GI - P , M ,
-
0, P&2'

For the mixture toluene-water we thus find

G,
- -- 3 3 7 x 9 2 = 4.1
G, 433 x 18
which signifies that to distil 4.1 kg of toluene we require 1 kg of water vapour. By
formula (40) we find that the toluene content of the mixture is 80.3% wt. and from
Fig. 218 it, is seen that the boiling point of the mixture is 84.4"C.
iranchembook.ir/edu

6.1 Carrier vapour distillation 30'3

Readily decoinpoeable compounds may be distilled with steain a t a reduced

pressure. From Fig. 218 we see, for example, that the mixed boiling point of aniline-
water, which is 98°C a t a total pressure of 760 irtm, falls to 75" if this pressiire i s
300 niin; it is assumed, of course, that only enough steam for saturation is hlown 111.
S t e q distillation at a reduced pressiire is desirable for distilling substances with low
vapoiir pressures (such a? fatty acids of high nioleciilar weight) and for dealing with
mixtures containing a coniponent in low concentration (for instance oils containing
traces of a solvent).

Fig. 219
t/oc -
Boiling points of the straight-chain, saturated, even-numbered fatty acids, with
,Ind without steam, a s i~ function of the distillation pressure

The results of rising .superhecrted steat/i - which is widely riiiployed in industry for
t h e distillation of taw, mineral oils, fatty acids, glycerin etc. -- have been critically
diccnssed by Stage. Taking the saturated C, to C,, straight-chain fatty acid.: as an
example, he showed that when using saturated steani hoiling points are depressed by
about 160-240deg. C, but that at the sanic time the differences in hoiling point,
from niein ber to nieinber, became so small that a separation into individual coinpoiinds
waq then iinpossible (Fig. 219). If, however, the separation was perfornied with lo"/,
of wperheated steam at a pressure of 10 mni, the differences in boiling point remained
iinaffected and the inherent advantages of steam distillation (good mixing and ab-
sence of local superheating) were still present ; the average reductions in the tenipera-
ture, on thf, other harid, amounted to only 20 deg. c' 131. I t may he pointed out that
mixtures of fatty acids and fatty alcohols can also he separated reasonably well in
stages by steani distillation coupled with partial condensation [4].
Althongh steani is the carrier vapour most used, on account of its low inolecular
weight, its high heat of evaporation and the ease with which it is condensed, other
gases also find application in industry for effective pressure reduction. I n the produc-
tion of Iiard coal tar pitch, for instance, waste gases containing COz, N2 and water
iranchembook.ir/edu

310 6. Selective separating processes

vapour are employed as carriers [6]. The presence of inert gas does not cause any
appreciable reduction in column efficiency [7].A graphical method for calculating the
condensation of stesni-gas mixt.rwes was evolved by Algermissen [8].
Apparatus for carrying out steam distillation can be put together without diffi-
culty from ordinary laborator>- components. Fig. 220 shows a set-up for distilliig
with saturated steam at atmospheric and reduced pressures. The flask a is thoroughly
insulated with glass wool, or slag wool; it is advisable to heat it, as well, in order to
prevent the condensattionof water. The steam inlet b is provided with a cock for

Fig. 220
Arrangement for distillation with eaturat,edsteam

drawing off condensed water and can also be employed for passing in other carrier
gases. Fig.221 illustrates the method of distilling with superheated steam in counter-
current operation. The steam is produced in a nietd boiler a,equipped with a sight
glass. Superheating takes place in the c6nical metal spiral tube b, which is connected
to a water trap and a thermometer. It is advisable to include a safety valve in the
steam line. An arrangement developed by Tropsch [7] has also proved suitable for
superhmting. For comparative experiments it is necessary to supply the steam in
constant, measumble amounts. A simple method of regulating the amount is shown
in Fig. 220 where water is admitted dropwise from the graduated cylinder d into the
steam boiler, care being taken to maintain a constant level. A more accurate method
is that described by Merkel [9], who regulates the amount of steam by the pressure
produced in one l i b of a manometer. A steam production unit has been developed
by Stage et al. [lo] to an arrangement in which the steam can be accurately mpasured.
iranchembook.ir/edu

6.1 Carrier vapour distillation 311

Water is admitted continuously from a measuring burette into the heated apparatus,
which is filled half full with sand to promote heat transfer. Alternatively the water
may be evaporated in a coil immersed in a metal bath.
For the distillation of sinall amounts of material the apparatus of Pozzi and
Ewot [11] is very convenient, as the steam-boiler simultaneously serves as heater
for the distillation vessel (Pip. 222). I’arnass and Wagner [I21 offer an arrangement
for inicroscalr work.

Fig. 221
Arrangement for distilling with superheated steam in countercurrent, operat,ion

Fig. 222
Device for the distillation with st.enm of small amounts
of material (Pozzi and Escot)
iranchembook.ir/edu

312 6. Selective separating processea

The methods of steam distillation have been summarized by Bernhauer [13] and
Thorniann [ 141. A detailed discussion of practical and theoretical aspects of steam
distillation a~ illustrated by the distillation of essential oils is given by von Weber
[15]. Rigamonti [l6] developed a nomogram which can be used to calculate the steam
requirements for various enrichments. Prenosil [ 16a] compared theoretical anti
e-xperimental steam distillation data for niulticomponent mixtures. He modified the
calculating niet hod by introducing a value for evaporation efficiency. Steam distilla-
tion can also be carried out in thin-film apparatus. Berkes et al. [ 16h] give a descrip-
t,ion of the material transfer conditions of a steam distillation performed in such
apparatus in ternis of the balance eqnations.

6.2 Azeotropic an'd extractive distillation

Whilst azeotropic and extractive distillation are now eiiiployed extensively for
difficult separations on an industrial scale [6], it has been usual in the laboratory to
resort to other processes, siich as extraction and chromatography, for separating
narrow-boiling and azeotropic mixtures. It will be shown below that under unfavour-
able conditions selective processes, such as azeotropic and extractive distillation,
offer considerable advantages. The common characteristic of the two is that the
ratio of the activity coefficients of the components is influenced by adding another
substance [ 171. ; Iconihination of the two processes termed azeotropic-extractive
rectification was proved to he feasihle by Kiiniierle [HI. Gerster 1191 conyared
these selective processes with ordinary distillation from thc point of view of economy.
A comparison between extraction and extractive distillation for the purpose of
separating aromatic hydrocarbons from petrol produced by pyrolysis and reformate
wits made by Miiller [19a]. He shows in which cases extraction and in which estritc-
tive dhtillation is advantageous, including the economical aspect. Helnis and John
[ 19bl have described the extractive distillation of aromatic hydrocarbons by the
"Lurgi-Mstapeu" method. They used n-methyl pyrolidone (LUMP).The purities
obtained were 99.95, 99.7 and 99.8% for benzene, toluene and xylene, respectively.
The book of Hoffniann [5] which contains niimerous calculations for binary, ternary
and niulticomponent systems offers a thorough treatment of the problems associated
with azeotropic and extractive distillation. Results of laboratory experiiuents on the
separation of strongly non-ideal mixtures by means of azeotropic and extractive
distillation as exemplified by the distillation of acryl nitrile are reported by Schober
et al. [19c]. In addition, the authors have made a theoretical study of mixtures of
HCN, acryl nitrile, acetonitrile, oxazole, H,O.
In the case of nowideul mixtures without a special point the equilibrium curve
approaches the diagonal asymptotically at the upper or lower end (for examples see
Fig. 29f and h). Even with relatively great, differences in boiling point between the
components a separation of such mixtures requires a considerable number of separat-
ing stages. Mixtures having a special point (azeotropes) give the following results
when submitted to countercurrent distillation with sufficient separating power.
iranchembook.ir/edu

6.2 Azeotropic and extractive distillation 313

Nixtures forming an azeotrope with minimum boiling point: distillate: azeotropc

niixture of the two coinponents; bottom product : the component in excess, pure.
itlixtures forming an azeotrope with nbaximum boding point: distillate : the conil)o-
neiit in excess, pure; bottom product : azeotropic niixture of the two components.
Which of the two coniponents can be obtained pure from a binary mixtiirtl of
siibstances A and B, forming an azeotrope, depends on the coniposition of the initial
iinxtnre. If this is between the azeotropic composition and 100% A, distillation can
yield compound A and the azeotrope, but not the second component, in a pur
Azeotropes with a iiiinirnuni boiling point (for examples, see Fig. 43,cohmin 3/III)
are far more numerous than those with a maximuni boiling point (Fig. 43,colunui
-5011). According to the tables of Lecat [20], who lists A287 azeotropes and 700:;
non-azeotropes, the ratio is about 9 to 1.
The books of Horsley [51] present azeotropic data up to 1962, and the handbook
on azeotropic mixtures published by Kogan et al. [21] contains 21,069 systems of
which 19735 are binary, 1274 are ternary and 60 are inulticoniponent mixtures.
The tables of the handbook are preceded by an introduction into the theoretical and
experimental aspects of azeotropy written by Kogan, who edited the book. Furthcr,
he discusses the influence of temperature, the coniposition of azeotropic niixtureh. t h v
predictionof the azeotropicpoint and the study of the properties of azeotropic mixturt.,
For the thermodynamic and kinetic theory of azeotropic mixtures the reader. I?
referred to investigations by Stuke [22], C’oulson and Herrington [23], Kuhn and
Massini [24], Enixstiin [25], Litvinov [26] and to the extensive publications of Swiyto-
slawski [27] and Lecat [28], who also deal with ternary and quaternary azeotroiws.
The theor\- and applications of azeotropic distillation were thoroughly dealt wit 11 1):
Othnier [%a]. I n the first volume of this fundamental book the last-named anthor
discusses the theory of azeotropes, the problems of experimental technique and t hc
classification of the various types of azeotropes, and considers the applicationh of
azeotropy in indiistry. The second volume of the book is to deal with azeotropy froni
a thermodynamic point of view. An excellent introduction into the probleiii* of
azeotropic and extractive distillation in the laboratory from a theoretical as well a+
from a practical point of view was given by Rock [17]. In Schuberth’s [29] boolis o i i
the thermodynamics of mixtures the azeotropy of binary systems is discussed in
Volume I.
-4model concept for the description of the vapour-liquid eqiiilibriiiiri of azeot ro1)iv
mixtures in the case of associations in the gas phase was elaborated by Sc rafinior rt
al. [29a]. Examples are the binary mixtures acetic acid-water, formic acid-water ntid
acetic acid-formic acid and the ternary azeotrope acetic azid-formic acid-water
(see Fig. 225). The real behaviour of the liquid phase is in all cases described
Wilson’s equation. Svohoda et al. [19b] measured the evaporation enthalpies oi Y
binary azeotropic mixtures in the temperature range frmi 30 to 80°C. On siniplif>ing
assumptions the additional enthalpies of the azeotropic niixtures can be calculated
froin the measuring data. With this method the temperature dependence of the
additional enthalpy up to the boiling point can be determined in a simple manner.
A mathematical description of phase equilibria in polyszeotropic mixtiires ha--
been given hy Serafiinov et al. [31a]. Tamir and Wisniak [31b] have reported the
iranchembook.ir/edu

314 6. Selective separating processes

correlation and precalculation of boiling temperatures and azeotropes for multi-

component systems.
Narrow-boiling mixtures can generally be expected to be azeotropic. Azeotropy
does not only occur in cases where rather specific interactions are present but also in
mixtures of non-polar substances, such as benzene-cyclohexaneor benzene-carbon
tetrachloride. According to W. Kuhn and H.-J. Kuhn [24]the frequent occiwrence of
azeotropy in narrow-boiling mixtures is due to the compensation of contributions to
the vapour pressure difference and the summation of contributions to the energy
required for misig. Zieborak [30] ehowed that the addition of hydrocarbons for the
&stillative dehydration of ethanol-water mixtures results in a number of quater-
nary heteroazeotropes and ternary homoazeotropes. (A heteroazeotrope is an
azeotropic mixture which separates into two liquid phases on condensation (example:
benzene-water).) The various types of polyazeotropic mixtures of liquids and the
influence of these azeotropes on the course of the distillation have been investigated
by Orszagh [31]. Malesinski 1323 and Stecki [33] have contributed a classification of
homo- and heteroazeotropic ternary s-wtems. In a survey of possible additives for
azeotropic and extractive distillation Berg [34] has dealt with the classification of
liqiiids with regard to hydrogen bonds.
Columns having about 100 theoretical stages, suitable for dealing with numerous
narrow-boiling mixtures and non-ideal mixtures without a special point, can now be
constructed, since an HETP of 1 to 2 cm is attainable with modern forms of packing.
If, however, 200 or 300 stages are required (for values of a equal to 1.03-1.02), it is
preferable to attempt, to increase the value of a. As an example we may mention the
ext mctive distdlation of the narrow-boiling mixture n-heptane-methylcyclohexane,
where the difference in boiling point is only 2.7 deg. C (a= 1.076).Normal counter-
current distillation requires 48 stages with v = 00 for enriching a mixture from
15.3 molyo to 96.4 molyo. If 70% wt. of aniline is added, the same enriohment is
obtained with 12.4 stages and a reflux ratio of 36.The separating factor a increases
from 1.075 to 1.30 [35]. With mixtures forming an azeotrope, a selective separating
process is indispensable if both compounds are to be obtained in a pure state.
The phenonienon of azeotropy can on the one hand be very troublesome, as in the
distillation of dilute alcohol and similar mixtures; on the other hand it offers a means
of “breaking” existing aieotropes and of separating very narrow-boiling mixtures.
By the addition of a foreign compound (known as an entrainer) the formation of a
new azeotcope, between the entrainer and one of the components of the binary mix-
ture, is purposely brought about; as a result of the larger difference in boiling point the
latter azeotrope can then be separated from the other component. An essential
requirement in this method is, however, that. the new azeotrope shall be readily
separable into its components. Methods by which this may be achieved are:
by cooling
by salting out
by chemical removal of the entrainer
by extraction
by a second azeotropic distillation.
iranchembook.ir/edu

6.2 Azeotropic and extractive distillation 315

In extractive distillation a solvent boiling about 50 to 100°C higher than the mix-
ture t o be treated is introdiiced; its preferential affinity for one component change4
the relative volatility [17]. The solvent chosen must not form an azeotrope with any
of the components to be separated, must be readily reniovable from the mixture and
must act so as to produce an increase in the relative volatility The compoiinds
employed are the same as, or similar to, those utilized in liquid-liquid extraction.
Examples of the apparatus used for azeotropic and extractive distillation a IT
shown in Pigs. 223 and 224.
The selectivity of distillation with a n entrainer does not in general depend only
on the relative volatilities but alqo on the diffusion rates of the components present

entrainer

charqe + entraine-

a) bl
Fig. 223
Apparatus for azeotropic distillation
a) batch b) continuous
iranchembook.ir/edu

3 16 ~~ ~
6. Selective separating processes
~

in the liquid and gaseous phases. Schliinder [35b] describes distillation with an
entrainer with an excesfi of entrainer where an equilibrium cannot be established and
selectivity is governed by diffusion. This kinetic effect could be used for the separation
of azeotropic mixtures. Moreover, distillation with an additive allows the examina-
tion of the hydrodpainic and kinetic causes of incomplete equilibration.

distillate A
distillate,

-
a followed by 8
A l v e n t feed

solvent feed

tacuum

1
;
start :‘charge
end: solvent

P
1 \
solvent
Pump

Fig. 224
Apparatus for extractive distillation

A) batch b) continuous
iranchembook.ir/edu

6.2 Azeotropic and extractive distillation 317

6.2.1 Azeot ropic d i s t i I lat ion

Azeotropic distillation is employed in the following cases:
1 . For the separation of narrow-boiling mixtures, which have usual]? been prepared
by a previous countercurrent distillation;
2 . For the separation of mixtures forming an azeotrope, generally having a conipo-
sition close to the azeotropic point.
As a rule the compound to be added is so chosen that it fornia an azeotrope of
rniniinuiii boiling point with one of the components. But it is also possible to select an
entrainer forming a binary or ternary minimum azeotrope with both of the components
to he separated; in the latter case it is necessary for the proportion of the coiiq10-
iients in the new azeotropes to be different from their initial proportions. Disciissinp
extensive investigations of various types of phase diagrams and of the elaboration of
cohiinn schemes Sharov and Serafiniov [35a] have treated the problems specific to
the countercurrent distillation of azeotropic niulticornponent mixtures.
A particularly striking example of the treatment of narrow-boiling mixtures is
the mixture indole-diphenyl, which can be separated with diethyleneglycol as en-
trainer. At atmospheric prtmure indole and diphenyl differ by only 0.6 deg. c' in
boiling point ; by the addition of diethyleneglycol this difference can be increased to
12.2 deg. C. The azeotropes produced: diphenyl-diethylene-glycol (boiling point
230.4 "C) and indole-diethylene-glycol (B.P. 242.6"C), both of wliich contain jiist
under GOo/, of diethglenz-glycol, can be separated with relatively inefficient columns
and low reflux ratios. A s diphenyl, unlike indole, is scarcely soluble in liquid diethvl-
eneglycol, the amount of the glycol needed is small: the azeotrope of diphenyl and
diethyleneglycol, which passes over first, deinixes in the receiver and the glycol can
be recycled eontinuousk?- to the distillation. From thP azeotrope of indole and di-
ethylcneglycol passing over a t a higher temperature the indole inap be precipitated
with water [36].
Coinniercial P-picoline is R mixture of a-picoline, P-picoline and 2.6-lut idine. The
difference between the boiling points of p-picoline and 2.6-lutidine is 0.15 deg. C at
atniospheric pressure. By means of azeotropic distillation, using acetic or propionic
acid as additive, one can separate the individual bases in a purity of 95-980/o [W].
Further processes eniployed in industry were reported by Dummett [38].
Two processes which are of interest in industry may be mentioned here. If the
ternary system water-formic acid-acetic acid is to be separated the following azeo-
tropes are to be expected according to Hiinsinann and Sinimrock [39] :
B : 107.65 "C binary high-boiling azeotrope 56.7 ~ i i o ofl ~formic
~ acid
43.3 mol*~,of water
T: 107.1"C ternary azeotrope 39.3 mol~/, of water
48.2 nioloi, of formic acid
12.5 niolq/O of acet,ic acid
Thiis, tlhe whole concentration region of t,he ternary system is divided into four
separate distillation areas (Pig. 225). By means of azeotropic dist'illation with a
iranchembook.ir/edu

318 6. Selective separating processes

higher ether the mixture can be dehydrated. The interdependences involved in the
countercurrent distillation of ternary mixtures with special points were studied by
Petlik and Avetyan [39a]. Corresponding to the azeotropic points of binary systems,
ternary systems haw limiting lines which cannot be exceeded in ordinary distillation.
The systems acetone-chloroform-methanoland acetone-chlorofonn-isopropylethene
were studied with a view to this phenomenon by Naka et al. [39b]. Kiidryavtseva et
al. [39d] discuss various methods of calculating multicomponent azeotropic mixtures
and suggest a calculating procedure on the basis of constants obtained for binary
mixtures which does not include a formulation for the concentration dependence of

tormic acid

water acetic acid

Fig. 225
Diagram of the four distillation area8 of the ternary system water-formic acid-
itcet,ic acid, having a binary and a ternary azmtrope, and position of composition
R WI

the activity coefficients. Results for ternary azeotropes are represented graphically'
The calculation using data from binary mixtures can be carried out with a precision
sufficient for practical purposes. Lino et al. [40] succeeded in reducing the proportion
of water in acetone to below 1,400 ppm by azeotropic distillation using ethyl bromide
as additive. Susa,rev and Toikka [39c] have reported a method of estimating the
compositions of ternary azeotropic mixtures if the vapour pressures of these mix-
tures have been measured or the vapour pressure of one of the binary azeotropic
mixtures is known.
It would lead us beyond the scope of this book to mention all the processes of
azeotropic distillation known at the present time, since the publications and patents
on this subject are numerous; the two examples selected show the wide possibilities
that exist. Another field of azeotropic distillation is the dehydration of organic
compounds, such as formic acid, acetic aoid and pyridine. Puther, mention may be
made of the separation of hydrocarbons from alcohols, the purification of aromatic
iranchembook.ir/edu

6.2 Azeotropic and ext'ractive distillation 319

hydrocarbons and the separation of mono- from di-olefins. Mair, Glasgow and
Rossini [41, 421 and Berg [34] have carried out a systematic investigation on thc
vparation of hydrocarbons by azeotropic distillation. The separation of the isoitirrs
of methyl naphthalene in the presence of undecanol shows that the distillation
pressure chosen is of importance also in azeotropic distillation. The optimuin is
300 t o 300 torr. Pure a-methyl naphthalene is obtained as residue while the distillate
is a mixture with 80% of $-methyl naphthalene [43].
Since the position of an azeotropic point is not stable, besides the azeotropic and
extractive distillations a change of the external thermal conditions (temperature or
pressure) in the form of a vacuum or pressure distillation may be effected t o make the
special point disappear. Schuberth [a] has reported relations by means of which the

I 184

00 01 02 0 3 0 4 0 5 07 1 2 3 4 5 6789at
pressure +
Fig. 226
Pressure dependence of the azeotropic points of the mixtures ethanol-water (t)
and water-phenol (11)

amounts of pressure or temperature change or (in azeeotropic or extractive distillation)

the mininium concentration of the additive required to remove azeotropy can he
approximated for hoinoazeotropic binary systems.
The result of a diminution in pressure is frequently that the azeotropic composition
becomes richer in the low-boiling constituent. A reduced pressure may thus be
reached a t which the special point vanishes. As examples, the mixtures ethanol-
water and water-phenol may be taken (see Fig. 226). By distilling dilute alcohol at
TO mm, absolute alcohol may be prepared without the use of anentrainer. The azeotro-
pic point water-phenol is eliminated a t 32 mm. The shift in the azeotropic point
has been determined by Shemker and Peresleni [45] for two other systems:
Formic acid-water : a t 50 mm Hg,66% wt. of formic acid;
a t 200 mm, 72% wt. of formic acid.
Butyl alcohol-butyl acetate: a t 50 mm, 37 molO/, of butyl alcohol;
a t 760 nim, 79 nioloh of hiityl alcohol.
iranchembook.ir/edu

320 6. Selective separating processes

Ruether and Lu [a] elaborated numerical methods which were tested on the
systems ethyl acetate-ethanol, ethanol-water and methanol-benzene. An azeotrope
can in some cases also be made to disappear by increasing the pressure.
Nutting and Horsley [P7] have indicated a simple procedure for determining the
range of pressure in which an azeot rope exists. A Cox diagram, having scales for log p
+
along the y axis and 1/(T 230) along the zaxis (Tin "C),is plotted with the vapour
pressure curves for the pure components and the azeotrope (Fig. 227). (Owing to the
linearity of vapour pressure curves in a Cox diagram, only two values are needed for
each line.) If the straight line for the azeotrope intersects the lines for the compo-
nents, the special point cannot esist beyond the points of intersection P and P'. i.e.

positive
P
-m

I C = binarv azeotroDe 1
1 I ( T'C +230) _+
Fig. 227
Determination of the azeotropic pressure range by the method
of Kutting and Horsley

when its temperature at a given pressure is between those of the components. If no

intersections occur, and the azeotropic line is above or helow those of both conipo-
nents, the azeotrope will remain present at all pressures. Joffe [48] has described a
method in which it is necessary to know only the azeotropic composition at one
temperature (pressure) for deriving the azeotropic compositions a t other tempera-
tures (pressures). Malesinski [49] has published formulae with which the boiling
points of ternary azeotropic mixtures may be calculated.
Should it prove impossible to effect a satisfactory separation by changing the
dietdation pressure, the next step will be to find a suitable additive with which one
of the constituents forms a.het.eroazeotrope, or a homoazeotrope that is easily split
rip [34]. An approximate method for separating heteroazeotropic mixtures based on
the mathematical model for the liquid-liquid-vapour equilibrium of two mdti-
component systems was elaborated by Bril et al. [49a].
To illustrate the procedure we can best take an example. The system water-pyri-
dine gives rise to a niinimuni boiling azeotrope, boiling at 92°C and containing 64Oj,
wt. of pyridine. The water is to be eliminated by the addition of an entrainer. Thc
requirements that the latter should satisfy are the following [34, 501:
1. It should form an azeotrope, boiling below 92"C, with water;
3. it must not yield an azeotrope with pyridine;
iranchembook.ir/edu

6.2 Azeotropic and extractive distillation 32 1

3 . no ternary aeeotrope should he produced;

1. the proportion of water in the azeotrope should be as large as possible:
3. the mutual solubility of the entrainer and water should be low, so that separation
takes place on cooling (i.e. they should form a heteroazeotrope).
Ivow minimum boiling azeotropes are formed when the forces of attraction he-
tween unlike molecides are smaller than those between identical ~nolecules: i n t h e
converse case maximum boiling azeotropes occnr. On the basis of data compilcd by
Leeat [20] and Horslej 1511. Ewcll, Harrison and Berg [521 have concluded that the
forces of interaction concerned are niainly those due to hydrogen bonds, coinpared to
which other forces such as dipolar forces and those due to induction are rrlntively

Tnhle 48
Cliissification of liquids according t o hydrogen-bonding tendency
it) The five classes according to Berg [34, 521

i'liisu Kind of molecule Examples

1. Liquids capable of furniirig threr- Water, glycol, timino alcohols, hydrosyla-

dimensional networks of strong hy- mine, hydroxy acids, polyphenols. aiiiides
drogen bonds et.c. Compounds such as nit.ronicthane and
acetonit,rile also form three-climensional
networks of hydrogen bonds, but t,he bonds
are much weaker than those inwlring
OH and N H groups. Therefore, t hcse types
of compounds a~replaced in class I1
11. Other liquids composed of molecules Alcohols, acids, phenols, primary and
containing both active hydrogen secondary amines, oximes, nitro com-
atoms and donor atoms (oxygen, pounds witah a-hydrogen at>oms,nitriles
nitrogen and fluorine) with a-hydrogen atoms, ammonia,
hydrmine, hydrogen fluoride, hydrogen
cyanide, etc.
111. Liquids composed of molecules con- Ethers, ketones, aldehydes. esters, tertiary
taining donor atom6 but no active amines (including pyridine type), nitro
hydrogen atoms compounds and nitriles without a-hydro-
gen atoms, etc.
1V. Liquids composed of molecules con- CHCi,, CH,CI,, CH,CHCI,, CH,CI -CH,CI,
taining active hydrogen atoms but CH~CI-CHCl-CH~C1, CH2CI -CHC12.
no donor atoms. These are mole- etc.
cules having two or three chlorine
atoms on the same carbon a s a
hydrogen atom, or one chlorine on
t h e same carbon atom and one or
more chlorine atoms on adjacent
carbon atoms
V. All other liquids - i.c. liquids having Hydrocarbons, carbon disulphide,
no hydrogen-bond-forming capabili- sulphides, mercaptans, halohydrocarbons
ties not in class TV, non-metallic elements
snch as iodine, phosphorus, sulphur, etc.

21 Krell, Handbook
iranchembook.ir/edu

322 6. Selective separating processes

Table 48 (Continued)
b) Summary of deviations from Raoult's law

Claaees Deviations Hydrogen bonding

Always +
deviations; I + V, H bonds broken only
1:;:: } frequently limited solubility
111 + IV Alnays - deviations H bonds formed only
Always i deviations; I + IV, H bonds both broken and formed, but
frequently limited solubility dissociation of class I or 11 liquid is
more important effect
Usually +
deviations, very H bonds both broken and formed
complicated groups, some
I + 111 - deviat.ions give some maxi-
I1 + II mum azeotropes
I1 - I11
I11 + 111 Quasi-ideal systems; always S o H bonds involved
JIL
IV +
V
JV
+
deviations or ideal;
LV -v azeotropes, if any, will be
minima
V L V

c) Solutions giving maximum boiling azeotropes

' Components Examples

IVater +-
strong acids Water + HCI, HBr, HI, HNO,
W\'ater -:- certain associated liquids Water +
formic acid, hydrazine, ethylene-
diamine
Donor liquids (class 111) + non-associated Acetone + chloroform, cyclohexanone
liquids containing active hydrogens A bromoform, butyl acetate
(class IV) + 1.2.2-trichloropropane
Organic acids - amines Acetic acid + triethylaminc, propionic acid
+ p-yridine
Phenols - an1inc.s Phenol + aniline, o-cresol + dimethyl-
aniline
Organic ucids
oxygen
- donor liquids containing Formic acid + diethylketone, butyric acid
+ cyclohexanone
Phenols - donor liquids containing oxygen Phenol f methylhexylketone, o-cresol
+ ethyl-oxalate
Phenols - <ilcohols Phenol L n-octanol, 0-cresol + glycol

uriiiuportant. On this basis liquids way be divided into five classes according to t h e
niimher and strength of the hydrogen bonds between the molecules. These five classes
have heen collected in Table 48, part (a). By looking up the hydrogen-bond classes to
which the two components belong it is possible, on referring to part (b) of the tahle.
to estimate the departure from ideal hehaviour of a mixture of other classes, and so to
iranchembook.ir/edu

6.2 Azeotropic and extractive distillation 325

select a suitable entrainer [52]. I n part (c) of the table the coniponents giving niaxi-
niuin boiling azeotropes have been listed (see p. 156 of [29]).
Following Ewell's classification of mixtures, Eduljee and Tiwari [52a] extended
the relations found by Yoqhimoto. They studied various properties of the interactions
of the components in the mixture and determined the temperature range within
which no azeotrope was formed. The mean standard deviations in predicting the
azeotropic composition and the boiling point (from the boiling points of the pure
components) were found to be 0.17 and 0.03, respectively, for 14 combinations.

Table 49
Suitable additives for removing water from pyridine by azeotropic distillation [50]

Class Additive Kp.760 Bzeotrope Parts of Solubility

with H,O additive reqd. in 100 parts
for removing of H,O
("C) ("C) 1 part of H,O

I11 Isobutyl formate 98.2 80.4 11.8 1.0

Ethyl propionate 99.0 81.2 9.0 2.4
Methyl butyrate 102.0 82.7 7.7 1.5
Ethyl isobutyrate 111.7 85.2 5.6 trace
Di-isobutylether 122.2 88.6 3.4 trace
9 m y l formate 132.0 91.6 2.5 trace
IV Ethylene chloride 83.0 72.0 11.0 0.9
1,2-dichloropropane 96.8 78.0 7.4 0.3
v Benzene 80.2 69.3 10.2 0.1
Toluene 110.7 84.1 6.4 trace

I n the system water-pyridine, the compounds mentioned in Table 49 can be

considered a4 possible entrainers. Water (class I) presents large departiires froiii
Raoult's law when mixed with liquids of classes 111, I V and V, whilst pyridine (class
111)gives sniall departures, if any.
It will he seen that the hoiling point of the azeotrope of water and the entrainer is
lower. the lower the boiling point of the entrainer. The ideal choice will be an entrainer
comhining a sufficiently low azeotropic boiling point, an econoniically low proport ion
in the azeotrope and a small solubility in water.
Berg and collaborators [34, 521 investigated a number of narrow-boiling hydro-
carbon mixtures. For azeotropic distillation they found the optiniuni additives listed
in Table 50. I t is obvious that the normal separation factor 3~ can be drastically
changed into xaz by optimum additives.
Whether or not a compound forms an azeotrope with a given component is also
important in extractive distillation, where the occurrence of azeotropes is generally
undesirable. Some further empirical methods, applicable both to azeotropic and
extractive distillation, will therefore he mentioned.
A sniiple graphical procedure for predicting azeotrope formation has been de-

21*
iranchembook.ir/edu

324 6. Selective separating processes

Table 50
Optimum additives for the azeotropic distillation of hydrocarbons

Mixture to be Beparated Xormal Optimum Szeotropic

separation additive separation
factor I factor a.,

neohexane-cyclopentane 1.006 n-propylamine 0.987

cyclohexene-2.4-dimet lylmethaiae 1.006 acetone 1.025
methylcyclohesane-iso-octane 1.046 ethanol or 1.4-diouane 1.084
3-methylpentane-herene-1 1.009 methylenechloride 1.159
3-methylpentane-
2-ethyl-I -but.enc 1.037 ethylformiate 1.156
2.2.4-trimethylpent,nnt'-
2.2.4-trimethylpeutene-1 1.040 isopropylacetate 1.129
n-heptane-2.2.4-trimethyl-
pentene- 1 1.045 isopropylaceta te 1.129
ethylbenzene-p-xylene 1.036 2-methylbutanol 1.079
p-xylene-m-xylene 1.020 2-methylbntanol 1.029
m-xylene-o-xylene 1.105 met hylisobutylcarbinol 1.150

scribed hy Mair, Glasgow and Rossini [ b l ] . Pig. 228 shows the boding points of
azeotropes formed between benzene and various hydrocarbons as a function of their
compositions [53]. If the boiling point of the pure hydrocarbon, plotted along the y
axis, is connected to the boiling point of the azeotrope on the curve, straight lines
with various slopes are obtained. From the diagram it may then be predicted that
hydrocarbons boiling below 68" and ahove 100°C will not give azeotropes wihh
benzene. Further, one can deteritline the boilingpoint and composition of the azeotrope
fornied by any hydrocarbon and benzene hy drawing a line p a d e l to R I I adjacent
ine already on the figure, starting from the boiling point of the pure suhstance on

100
"C

1.
:
P
c

580
+
c
.-
0
a
--
-3- 0
0
n
Fig. 228
Determination of azeotropic data (Mair,
60 Glasgow and R.ossini)
0 0.2 0.4 0.6 0.8 ' I
'Benzene
iranchembook.ir/edu

6.2 Azeotropic and extractgivedistillation 325

the y axis. (The gradients of these lines depend on the boiling pnints of the pure
hydrocarbons in question and their degree of branching). If, for instance, we draw
the line as described for the hydrocarbon 3-ethylpentane (B.P. 93.5 "C), we find on
the curve that the azeotrope with benzene boils a t about 80°C and contains 96 nlolo,
of benzene; these values agree closely with observed data. Methods reported by
Horsley [54]and Meissner and Greenfeld [55] yield similar diagrams.
X graphical method proposed by Novikova and Natradze [56] is based on a three-
tlicnensional co-ordinate system, in which the variation of the parameters is represen-
ted by a straight line. The abscissa corresponds to the azeotropic composition in
niol%, the ordinate to the reciprocal of the boiling point and the applicate (third
axis) to the logarithm of the pressure. If the azeotropic data are known at two
pressures it is possible to determine with accuracy for binary mixtiires :
1. the azeotropic data at another pressure,
2. the range of pressures over which an azeotrope exists, and
3 . the concentration range of azeotropic mixtures.
Litvinov [26] evolved a graphical method for determining the probability that,
azeotropj- occurs in ternary systems. He shows that ternary azeotropes may occur
even when the system in question has no binary azeotropes. The existence of a
quaternary azeotrope, however, is associated with the occurrence of binary and
ternary azeotropes. Marinichev and Suearev [56a] elaborated an extensive system of
eyriations in order to infer physical properties of quaternary azeotropes.
According to Skolnik [57] a purely arithmetical prediction is also possible. For the
nienibers of a homologous series (e.g., hydrocarbons) and a second component (cg.,
henzene), these linear relations hold:
log zaL= A(273.1 + f+c,z) + B, (193)
with xaz = concentration of the second component in the mixture i n itiolrJ0;
8,,= boiling point of the azeotrope, "C; A , B = constants; and
log [D- &az) = E - F(273.1 + a), (194)
where D,E and E are constants and 6 is the boiling point of the member of the horno-
logoixs series. The curves shown in Fig. 229 are derived from expressions (193) and
(194). Curves I and I1 correspond to (193) and (194), respectively. The azeotropic
data of hydrocarbons of the same homologous series can be read immediately from
the curves. Thus, e.g., the boiling point of 2.2-dintethyl pentane is at 79.2"C. Curve I1
gives an azeotrope composition of 50 mole/, and curve I, a boiling point of the
azeotrope with benzene at 75.8"C. Curve I1 also serves to determine the boiling range
over which an azeotrope with benzene can exist. This is the range from 65 to 98"C,
which 1s in good agreement with the values obtained by the graphical method of
Mair, Glasgow and Rossini. The boiling range of the paraffins which form honio-
geneous azeotropes with benzene is called the "azeotropic effect" of benzene [ 5 7 ] .,4
diagratii constructed by Nakanisi et al. [58] on the basis of infrared-spectroscopic
data predicts azeotrope formation for a niiinber of components in the temperature
range from 30 to 150°C in binary mixtures with methanol.
iranchembook.ir/edu

326 6. Selective separating processes

The use of limiting activity coefficients for the prediction of azeotropic compo-
sitions at constant temperatures and of the partial miscibilities in binary liquid
mixtures has been reported by Brandani [%a].
According to Martin [58b], hinarr systenis form azeotropes if these conditions
are satisfied:
1 . =In azeotrope with minimum boiling point is formed if the activity coefficient of
the high-boiling component is greater than the vapour pressure ratio of the pure
components.
3. A heteroazeotrope exists if the limiting activit?' Coefficient of the low-boiling
component is greater than 7.4.
The validity of these relations was tested on the system water-n-propanol.

g0.6
-8
0
L

0.4
0
.- 0.3
a,
0.2

ix Fig. 229
Determination of azeotropic

0.1 70 75.8 80 90
-
oc
t (boiling point of azeotropes)
100
data (Skolnik)

The calculation of the conditions required for separation by distillation, such as

the nuniher of stages and the reflux ratio, is performed as described in chapter 4. For
regular systems, having an azeotropic point but only a small heat of mixing, Kuhn
[59]gives the following formula as a means of calculating the number of theoretical
stages, corresponding to formula (1 14) of Fenske for batch distillat ion with w = 00 :

so = x - 1;
x,, = theconcentration a t the azeotropic point.
In this equation it is assumed that xE < xaz,that is to say that the desired en-
richment lies helow the azeotropic composition. When xE = xazthe number of stages
required, nmln= 00.
Szapiro [60] has derived forniulae for the separation of binary mixtures containing
low proportions of the heavy component and of mixtures approximating to the
iranchembook.ir/edu

6.2 Azeotropic and ext.ractive distillation 327

dzeotropic composition. Calculations for the mixtures water-ethyl acetate (1.750,,),

benzene-toluene (4%) and ethanol-water (11%)are given as examples. Hunsniann
[cjt a] demonstrates the procedure of calculating a n azeotropic column, taking as an
tsaiiiple the separation of 1.3-butadiene and 1-butene by azeotropic distillation
with methyl amine. First, binary and ternary equilibria are measured in the techni-
cally important pressure region. The results are expressed as a set of equations, which
<ireused to ralculate plate after plate, progressing from the bottoiii upward. The cal-
culations show the influence of temperature and an excess of amine. The method was
chrcked using a small colunin. Agreement was quite satisfactory.
The mode of operation and the dimensioning of a heteroazeotropic distillation as
cxseinplified by the separation of the system water-acetic acid has been described
by Wolf et al. [61b]. Morozova and Platonov [6l c1 analyzed the structure of phase
diagrams of multicoinponent mixtures using a digital computer. They studied thr.
rcquireinents for the separation of azeotropic mixtures. I n order to achieve optirniuii
coliunn combinations Ser;*finiovet al. [58c] studied the ternary mixture isopropanoli
benzene/water on the basis of a mathematical treatment of the separation of hetero-
azeotropic mixtures. In another paper [58d] a procedure was presented for t h c k
heparation into its components of the water-containing inixture with acetone,
ethanol, benzene and butyl acetate by means of the therniodynainic and topological
analysis of the phase diagram structure.
To conclude this section an example will be given of an azeotropic distillation in
which a ternary azeotrope is formed - the preparation of absolute alcohol by
distillation with benzene as auxiliary substance. If 95% wt. alcohol is dihtilled in
batch with the calculated quantity of benzene, there first passes over a t 649°C a
ternary azeotrope, separating a t 28°C into two layers; the conipositions of the azeo-
trope and these layers are as follows:
azeotxope upper layer lower layer
(at 28°C) (54.70;, (15.30/,)
benzene, yo wt. 74.1 85.6 8.1
ethanol, yo wt. 18.5 11.6 53.1
water, o/, wt. 7.4 2.8 40.6

When the water has been eliminated, there follows a t 683°C a binary azeotrope
containing 32.4% of ethanol and 67.6% of benzene, and finally at 78.i"C alcohol of
about 99.8% purity distils over. On an industrial scale the preparation takcs place
continuously, but in the laboratory it is often more convenient to operate in batch
and work up the two layers of the ternar- azeotrope and the binary azeotrope in
portions.

6.2.2 Extractive d isti I lation

Extractive distillation, like azeotropic distillation, is eniployed for concentrating
tnistures beyond an azeotropic point and for separating narrow-boiling or non-ideal
inistures having separation factors close to unity. The scope of extractive distillation,
however, is more extensive than that of azeotropic distillation, on account of the
iranchembook.ir/edu

328 6. Selective separating processes

almost unlimited number of additives that are available for use. The separating
efficiency is strongly dependent on the amount of solvent added and hence the latter
should he deteriiiined as exactly as possible.
The separation of n-heptane-methylcyclohexane, with aniline as auxiliary sol-
vent, has already been cited ~ b 8an example of the treatment of a narrow-boilingmix-
tvre. The system methylcyclohexane-toluene has an equilibrium curve which
approaches the diagonal ver3 closely, so that the relative volatility becomes very
nearly equal to unity at. high concentrations of methplcyclohcxane. An excessively
large number of stages would be needed to obtain niethylcgclohexane in a pure state,
but by the addition of 65 ~101%of the polar solvent aniline the relative volatility is
increased and the separation is consequently very greatly facilitated. Extract ire

Fig. 230
Isobaric equilibrium
HCI -HZS04 -HZ0 iLt
750 torr: x1 molar fraction
of H,SO, in liquid phaae
lines of constant
vnpour composition
- - _ isotherms

distillation is also useful for separating azedropic nr ixtnres. For exaniple, the azeotrope
of cyclohexane and benzene can be dealt, with, again, by the addition of aniline. to
yield pure cyclohexane as overhead product. Kortiim and Rittel[61] have reported t h r
separation of priniarp, secondary and tertiary aromatic atnines with glycerine. anti
j’araffin oil. It is notable that they used a largely automated apparatus for 110th
batch and continuous extractive processes. Extractive distillation has proved qiiite
useful even for the separation of multiconiponerit inixtures containing compoiivtit~
froiu various classes of substances. Thus, e.g.. aromatic hydrocarbons can be separa-
ted from tilisturea with non-aromatic hydrocarbons by extractive distillation osinp a
butane diol such as butane did-2.3 and adding octyl alcohol as solvent [62].
detailed report of the recovery of pure benzene by extractive distillation witli di-
iuethyl fornianiide from the point of view of process engineering is given by Lehmann
et al. [62a].
Another interesting exainple of extractive distillation is the separation of HCI
froin azeotropic hydrochloric acid with H,S04 as described by Grewer [63]. He nsed a
graphite colrinin. Fig. 230 shows the isobaric boiling equilibriiiru for his sgsteni.
iranchembook.ir/edu

6.2 Azeotropic and extractive distillation 329

Solids may also he used as additives. For the removal of small aniounts of methyleric
chloride from acetone, Ino et al. [94] foiind ZnBr, to be the most suitable salt, more
than 50 wt. O/, being soluble in acetone. The salt was removed from the reflux 111 H
stirring kettle. ,4 procedure for separating and purifying isoprene by extractive
distillation and subbequenf countercurrent distillation has been described in tictail
by Saraev et al. [94a].
The principle of extractive distillation consists in enhancing the individual effect
A x of the boiling equilibrium by introducing an additive. It is even conceivable to
change an azeotrope with minimum boiling point into one with maximuiii boiling
]mint, i . r . . t,o make a positive activity coefficient become negative [63]. The solvent
t o be added, which is norinally a high-boiling siibstance, has to he chosen such that

Fig. 231
Eqiiilibrium curves of the niixt iirp
acetone-chloroform t o which
various quantities of methyl
isobiityl ketone have been added

thr. relative volatility of the components is strongly increased. It is introduced con-

tiiiuoiisly near the top of the column so that a higher concentration in the liquid
phase is maintained in the whole column (see chap. 6.3).
Fig. 231 shows t,he equilibrium curves of the azeotropic mixture acetone-chloro-
foriii to which have been added various amounts of the extracting agent, inethvl-
isobntylketone. (The quantities are mole fractions.) The binary system give? an
azcotrope with niaxiiiiuni boilingpoint a t 34.5 mol%. This disappears when 30 nl(J1”
of the additive is present; further additions cause a still larger increaw in the relatirc.
volatility [SO].
Pavlov et al. [63a] studied extractive distillation on the basis of C, and C, hydrci-
carbon cuts, using 19 auxiliary liquids. They plotted the separation factors ohtaiiitd
in diagrams thns showing the influence of high-boiling hydrocarbons.
An auxiliary liquid for extractive distillation should comply with the following
rcqiiirenients :
1. the substance should not form a n azeotrope with the coniponents of the mixture
In question;
2 . the additive should be easily separated from the components;
iranchembook.ir/edu

350 0. Selective separating procebses

3. there should be as large a difference as possible between the degrees of departure

from Raoult’s law given by the additive with each of the two components SO that
the relative volatility of t,he two initialcomponents should be increased as much as
possible [MI.

The conditions for azeotrope formation and their avoidance have been described
above (see chap. 6.2.1). In addition, according to Scheibel [65], the separat,iou
coefficients of the mixt,ures to be separated are indicative in this respect,. Ease of
separation of the additive from the conryonent,s is generally ensured because the
additive is relatively high-boiling. According to Berg [34, 521, those liquids (of
Table 18)are suitable additives which belong to classes I and 11. These are chiefly
substances capable of st,rong hydrogen bonding and of acting as bot,h proton and
electron donors. They include phenols, aromatic amines (aniline and derivat,ives),
higher alcohols, glji:ol, etc.
Starting from the extractive distillation of ally1 chloride-propyl chloride and n-
but,ane-trans-butene-2 mixtures, Garber and Mironenko [65a] derived a universal
method for selecting suitable ausiliary liquids. For the charact.erization of addit,ives
in extractive distillation the gas-chromatographic vapour space met,hod was sug-
gested by Hachenbeg and Schmidt. Data are obtained in the form of peak area
ratios of components 1 and 2 in the vapour spaces over the non-boiling mixtures wit,h
and without additive.
The choice of an extracting agent having the required selective action will
involve a fair amount of preliminary experimental work. The most reliable method
is to determine the equilibrium curve of the mixture to be separated in the presence
of several quant.ities of the additive. Rose [36] has reported the results of such
measurements on the mixture n-hept,ane-methyl cyclohexane upon addit.ion of
different!liquids. He denoted t.he relative volat,ility with additive present by a,. For
the purpose of comparing the effect, of the various additives he set the relative
volatility of the components alone equal to unit,y a,nd then det.ermined the efficiency
or
2V = 2 . It is seen that, t.he addition of aniline (92%) gave the greatest effect on
a
the phme equilibrium. The effect of the additive on the relative volatility of the
c0mponent.s may also he estimabed from the integral heats of mixing [50].
Kogan [66] has shown that upon addit,ion of a t.hird component to a binary mix-
ture that component, increases in relat.ive volatility in which the t,hird component is
less soluble. Further principles applicable to t,he selection of additives have been
discussed by Kafarov and Gordijewski [67] and Kogan [68]. The connections be-
tween gas-liquid chromatography and extractive distillation have been elucidated by
Rijck [69] and Porter and Johnson [70].They have pointed out t,hat, the lat,t,er
inet,hodprovides a simple means of finding suitable additives for extractive distilla-
t ion.
A rapid estiinatte of the effect of additives can be obtained by determining
differences in boiling point. To the mixture of the component,s is added, as a first.
st.ep, a n equal volume of the auxiliary liquid in question and t.he boiling point. is
observed. The boiling point, of the niixtrire is then calcdat#ed,assuming additivity
iranchembook.ir/edu

6.2 Azeotropic and extractive distillation 331

on a inolar basis. If the difference between the calculated and the observed boiling
points is considerable, the additive will generally prove satisfactory [35].
Additives which are required to have a selective effect on the low-boiling coiii-
ponent may be chosen under the aspect of relative polarity. While liquids with the
same polarity give an ideal solution, inixtures of components with different polarity
deviate from ideal behaviour depending on the extent of this difference. Thus, e.q..
the highly polar mixture acetic acid-water can be influenced by the addition of a
non-polar high-boiling oil in such a way that the relative volatility of the water is
increased. I n the case of a non-polar mixture, such as cyclohexane-benzene, a polar
extracting agent, such as aniline, causes an increased volatility of the cyclohexane.
Berg et al. [52] and Hildebtand [71] have developed this method further.
Extractive distillation i q not limited to the separation of binary mixtures, but
1s also capable of removing particular classes of substances from multiconiponent
mixtures, as for instance benzene from mineral oil fractions. Mixtures of saturated
;ind unbaturated hydrocarbons having closely similar boiling points can be separated
hy extractive distillation with ketoesters [73]. Recently, the separation of lower
hydrocarbons C,. ..C, has been gaining ground [74]. Garner et al. [75] studied the
pfficiency of packed columns in the extractive distillation of the system nic.thj4
cyclohexane-toluene with the addition of furfiirol and derived equations for this
process.
After the optiiniirn additive has been chosen the liquid-vapour equili hria of t hc
ternary or iiiiilticoinponent systenis have to be determined as exactly as possihle 111
extractive as well as in azeotropic distillation. For this reason Null and Palnier [76]
looked for methods allowing equilibriuin values to be obtained from as few experi-
mental data as powible.
The number of stages needed for an extractive distillation can be calculated with
the Fenske formula (section 4.7.5.3) if the value of 01, in the presence of the additive
is taken. One can also applv the graphical construction of McCabe-Thiele, using thc
c>quilibriuincurve corresponding to ae, but eniploying the proportions in the solvent-
free mixture for the calculation (see chap. 4.7). Another simple way of determining
the number of stages necessary for an extractive distillation has been reported by
Gelbin [77] who nsed the pole height method after BosnjakoviC.
A modified McCabe-Thiele method employed in extractive distillation has betw
described by Nagel and Sinn “781. Kortiim and Faltusz [79] have dealt with a varlet\
of problems involved in selective separating processes ranging from the design of
a n automatic apparatus of special steel for continuous operation to the calculatioii of
the minimum reflax ratio and the required amount of additive.
The quantity of auxiliary liquid to be employed is dictated by the liquid reflux. If,
for instance, in our example of aniline-n-heptane we assume that the selected inolar
proportion of aniline to n-heptane is to he 83 niol%, we find from the molecular
weights and densities that the ratio by volunie is 3 : 1. Now if the calculation has shown
that for a load of 1100ml/h of n-heptane a rrflux ratio of 10 : 1is required, it follows that
+
3 x 1100 x lo/(10 1) = 3000 ml/h of aniline must be supplied at the colurnn head.
We then obtain 100inl/h of pure ri-heptane as distillate, while lOOOml/h of 11-
heptane, together with the 3000 ml/h of aniline, retnrn to the column as reflux. The
iranchembook.ir/edu

332 6. Selective separating procesrtes

extract may be allowed to accumulate in the still pot until the low-boiling component
ha29 been collected completely as distillate, after which the high-boiling component is
removed by batch distillation so as to leave the additive as residue. Alternatively,
the second coniponent can be distilled off continuously in another column as over-
head product, and the residue (the additive) drawn off at the base and returned to the
first column (Fig. 224).
To conclude the more theoretical considerations, some examples of extractive
distillation [64] are given in Table 51.

Table 51
Examples of the separation of ideal, non-ideal and trzeotropic mixtures by extractive distillation

Additive Yixt,ure to be separated ARp.,,, Remarks

(“C)

-4niline n-heptane-methylcyclohexane 2.7 Ideal mixture ( a = 1.07)

benzene-cyclohexane 0.7 Azeotrope
n-heptane-toluene
Non-ideal mixtures;
Pheuol n-heptane-to1uene asymptotic approach of
iso-octane-toluene 11.4 equilibrium curve to
met,hylcyclohexane-toluene 9.5 diagonal
Ethyleiieglgcol Methylethylketone-water 20.4 -4zeotrope
monohutylether
Diethylet her Ethanol-water 21.6 Azeotrope
Higher ketones and Acetone-methanol 8.5 Azeotrope
alcohols
Higher esters and Ethylacetate-ethanol 1.3 Azeotrope
alcoholn
Higher ketones and Acetone-chloroform .5.0 Xzeotrope
chloro compoundn

Nornial laboratory apparatus inay be used for azeotropic and extractive distilla-
t ion, whether performed batchwise or continuously. Packed columns and plate
coluiilns are equally suitable. Only rarely are additional components necessary (see
Figs. 223 and 224).
Let us consider azeotropic d~stillationfirst. If a homogeneous azeotrope is pro-
duced, no changes in the normal apparatus for countercurrent distillation are required.
When a heterogeneous azeotrope results the whole azeotrope is not returned to the
column as reflux, but only the phase rich in additive. For this purpose an “azeotropic
column head” is employed, as shown in Fig. 232, which allows either the heavy or
the light phase to be used aa reflux. The reflux is preferably returned to the column at
some distance below the head. In continuous operation, if a homogeneous azeotropr
is formed, the additive is mixed witb the feed; if the azeotrope is heterogeneous it ifi
iranchembook.ir/edu

6.2 L4zeotropicand extractive distillation :333

sufficient to supply the necessary quantity of auxiliary liquid through the “azeotropic
column head” (Figs. 223 and 232). More azeotropic colunin heads will he cleqcnhwl
in chap. 7 . 5 3 .
Extractive distillation needs a section of column to which the additive is conti-
nuously supplied. The additive should be admitted somewhat below the to11 of the.
colunin SO that a few plates are available to separate distillate and additive. I n
continuous operation the ft.ed is introduced a t about one third of the column height

Fig. 232
hzeotropic column head with magnetic reflux division and phase separation
a ) Take-off of the lighter phase
b) Take-off of the heavier phase

from its base (Fig. 224). As we have said previously, the procedure may- be made
fully continuous by drawing off the bottom product and passing it into a second
column; there it is separated into the high-boiling component and the auxiliary
liquid, the latter being recycled to the first column.
It should be noted that extracting agents may be employed which boil a t lower
temperatures than the initial mixture. I n such cases the auxiliary liquid is admitted
just above the still pot. This arrangement is necessary, for example, in the extractive
distillation of ethanol-water with diethyl ether.
iranchembook.ir/edu

334 6. Selective separating processes

6.3 Solution distillation and special methods

In the selective processes so far described an alteration in the phase equilibrium

is attained either by the use of a carrier vapour or by the addition of other liquids
(azeotropic and extractive distillation).
Another possibility is to dissolve salts in the mixture to he separated, so as to
bring about a change in the ratio of partial pressures of the components. This method
of solution dGtil2ation has been employed mainly for mixtures of which one compo-
nent is water. The mixture water-acetic acid, for example, at atmospheric pressure
has an equilibrium curve approaching the diagonal very closely above 96% wt. acid,
and it is consequently very difficult to eliminate the water entirely. From an industrial
point of view this distillation also possesses the economic drawback that water (the
distillate) has nearly five times the heat of evaporation of acetic acid; this becomes
important in the concentration of dilute acid. It would therefore be preferable if the
acetic acid could be obtained as overhead product and the water as residne. This can
1.0

0.8
1 0.6
Fig. 233
3'0.4 Equilibrium curve of the system ethanol-
water.
0.2
a = as such; b = with 10 g CaCl, in
101) ml of mixture
0 0.2 0.4 3.6 0.8 1.0
XB-

be realised by the addition of 80,b wt. or more of calcium chloride to the mixture [80].
With a 30°, wt. water-acetic acid mixture, for instance, the relative volatility OL
drops from a d i i c of 1.36 at 760 nuu to 0.525 in the presence of 20% wt. of CaC1,;
in other words the volatilities have been reversed and the acetic acid becomes the
iuore volatile component. It is essential, however, for the salt to be evenly distributed
over the whole length of the column. Daubach [81] has shown that the addition of
log of CaCl, to 100 lul of the azeotrope of ethanol and water caiises a disappearance
of the azeotropic point (Fig. 233); a distillate of high purity can thus be prepared by
introducing a solrrtion of ths salt at the top of a continuous column [82].Similar
experiments were carried out by TurBi and Thompson [83] with NaN03 and K,SO,.
Novella and Tarasso [a] used salts which are soluble in both components, such as
C'uC1, and KCI,H,O,. They found that they thus prevented azeotrope formation.
Schier [85] has discussed the influence of metal salts on the liquid-vapour equilibrium
of aqueous nitric acid. Using the apparatus of Othmer, Olefir et al. [85a]examined
the position of the azeotropic point in the system HN03/Hz0 on adding different
amounts of KN03. The azeotropic point was shifted from 68.2% HN03 (in the
system HN03/H20)to 81 HN03(in the system mo3/m(&/H@).
Similrtrly, the azeotropic mixture methanol-acetone can be enriched beyond
iranchembook.ir/edu

6.3 Solution distillation and mecia1 methods 33.5

the azeotropic point by the addition of 3.5 parts by volume of an aqueous calcium
chloride solution of density 1.2 (a 2.3-inolar solution at 20°C) [86]. Fig. 234 shows the
course of the equilibrium curve with and without the additive and also the flow
diagram of the operation. For the mixture water-phenol the addition of 17q, of
Sac1 results in a shift of the azeotropic point from 91 to 840&wt. of water, so that
the heterogeneous region of the system can be iitilized [87]. Ethanol-water inistrires
saturated with potassium nitrate in the concentration range from 15 to 70% yielded
higher enrichments than mixtures without additive [88]. Guyer et al. [89] investi-
gated the effect of sodium chloride on the vapoar-liquid equilibrium of the mixture
formic acid-water. In the presence of 35.5% of salt an azeotrope no longer occurs
$0 that a high concentration of formic acid can be obtained. Systematic investigations
of the salt effect in the system ethanol-water have also been conducted by Miro et
al. [go].
100
mol%
80
t LT o Fig. 234
Equilibrium curves of the systeiir
'I)
c 40 methanol-acetone
aJ
5 20 Q = in the presence of CaCI,

-
c
8 solution (2.3 molar)
Q
0 20 40 6CmOl% 100 b = without additive
Azetone in liquid b) Plow diagram of the operatioil
it ) b)

Furter [91] has analyzed the state of the art from the point of view of einployinp
the salt effect in industrial processes, especially in extractive distillation. I n addition.
he has made up a list of references covering the years 1966 to 1977 [91a]. Schuhert
et d.[92] investigated the effect of some metal chlorides and other salts on the iso-
thermal (8= 6O'C) phase equilibrium hehaviour of the systems n-propanol-water.
n-hutanol-water and methanol-water. Using CH30H/H,0/NaBr as an example, the
inethod of predicting salt effects for vapour-liquid equilibria as developed by Schu-
hcrth has been extended to unsaturated solutions [92a].
.Jaqiies and Furter [95] derived a n equation for the salt effect on the water-vapour
(quilibriuin in binary mixtures which correlates the teiiiperature and the liquid
concentration of the three coinponents ethanol, water and salt. The equation has
6 constants. The theory of the salt effect has been discussed by Furter and Meranda
[MI. On the basis of siinplifying assuinptions Sada et al. [97] have established n
relation for the calculation of vayour-liquid equilibria for non-aqueous binary
systems in which the salt is dissolved only in one coniponent (e.g., benzene-ethanol
with lithium or calcium chloride).
The above examples indicate that selective separating methods are of particular
importance for coinplicated separations. Systematic experimental work will tin-
doubtedly open up further applications. The coinhination of the distillation process
with adsorption effects led to the technique called adsorptive distillation. The in-
fluence of the nature of the jiacking inaterial on the efficiency of the separation of
the mixture water-acetic acid has been studied by Fuchs and Roth [93].
iranchembook.ir/edu

7. Constructional materials and apparatus

Various forms of apparatus have already been mentioned incidentally in chap-

ters 1 to ti. A auniniary of the characteristic properties of glass as a constructional
material together with important constants has been given by Pietzach and Rauten-
berg (see chap. 5, ref. [3Ob]). The present chapter will deal principally with recent
deve1opment)s in the components and equipment needed for different kinds of
distillation. A selection has purposely been made among the many types, often
identical in principle, that are available and this selection will also demonstrate that
flirt her standardimtion is possible. Standardization benefits both the user and the
prodiicer, but it must, of course, he applied in such a way that (.very process likely
to be employed can be perfornied in suitable conditions.

7.1 Constructional materials for distillation apparatus

The materials of constructiou for distillation equipment are nietals, porcelain,

silica and - most iruportant of all - glass. Metals are chosen for distillations at
high temperatures and pressures, where the resistance or strength of glass becomes
insufficient. For instance 18:8 stainless steel (187, chromium, 8% nickel) can be
used not only a t high temperatures and pressures but also with many corrosive
substances. Ceramic materials such as porcelain are not often used now that glassware
suitable for pilot-plant and industrial use is available. Silica finds its main application
in water stills and for operations at very high temperatures. Certain synthetic poly-
mers, such as Teflon, have been employed for the construction of various components
such as gaskets.
G b s is a material that was used 5000 years ago in ancient Egypt, not only for
household utensils, but also for alchemical equipment, and it, has retained its promi-
nent position in the laboratory to the present day. Its wide use follows from its very
favourable chemical and physical properties.
The transparency of glass gives it the undisputed advantage of allowing an
operation to be seen. I n processes that must proceed adiabatically an evacuated glass
jacket can be tisect up to 120°C or more for heat insulation. At higher temperatures
and in large apparatus a jacket of glass wool, provided with a slit for observation,
forms a satisfactory insulation (cf. section 7.7). Glass is also very resistant to corro-
+ion. There are, in fact, numerous separations and chemical reactions that can he
performed only in glass or ceramic apparatus. The hardness and smoothness of glass,
which reduce superficial deposits and allow the apparatus to be cleaned rapidly and
easily, together with its low coefficient of expansion, have favoured its use in the
iranchembook.ir/edu

7 . 2 Standard apparatus and unit parts 337

laboratory. In the preparation of pharmaceutical materials and of distilled water

glass has the valuable property of ensuring products free from odour, taste and
nietallic contamination.
The principal disadvantages of glass in comparison with metals are its fragility
and its lower resistance to changes in temperature. By special processes of annealing
it is possible to make glasses containing internal stresses of such magnitudes that they
counteract breakage. The present trend is to produce special glasses, certain proper-
ties of which have been improved for particular applications. I t will doubtless never
be possible to evolve an ideal glass offering the same range of application and ease in
machining that are presented by metal alloys, and it, will therefore remain necessary to
compromise by selecting among the many types of glass that are now available [ I , 51.
No useful purpose would be served by including in this book a detailed list of thr
chemical and physical properties of glasses. The testing conditions may vary appre-
ciably, which makes it difficult to arrive a t meaningful con~parisons.Besides, all the
specifications required are given in the catalogues of the glassware manufacturers [ 11.
Extensive tables of constants can be found in the publications of Monch [ 2 ] and
Eitel et al. [3]. A paper prepared a t the Central Laboratory of the \-T’B der Glas-
industrie (Association of the Glassware Manufacturers), Ilnienau, GDR [4], contains
the constants of chemical glasses as compared with those of porcelain, metals and
plastics. Also, applications of glass as a constiuctional material and the properties of
glass fibres and products iiiade of glass fibres are exhaustively dealt with.
For the uses of semi-technical glass apparatus in the laboratory and in pilot
plants the reader is referred to the papers of Rieger [ 5 ] . By using special types of
glass it is possible t o make apparatus capable of withstanding distillation pressures
u p to 4.3 kg/cm2 [GI. The use of porcelain in the laboratory has been discussed by
Kainrnel [7].

7.2 Standard apparatus and unit parts

Reference has already been made to standardized, conventional inet hods of

distillation in chap. 3.1 Fig. 235 shows the device used for Engler distillations which
has been improved by employing ground joints. Thus, the immersion depth of the
thermometer and the posit ion of the side tube for the distillate remain exactly fixed.
Automatic apparatus based on this device will be described in detail in chap. 8.1.
The use of such apparatus is to some extent justified in industrial analyses, where
rapid work is essential, but it is unsuitable in more exacting applications. Further-
more, it cannot be employed for vacuumdistillation. For simple unrectified distillation
at reduced pressures down to 10 mni Hg it is preferable to employ the Claisen flask
(Fig. 236). If the pressure is measured near the receiver, low values will often be found.
In accurate work the pressure and temperature should be measured close together.
If the pressure is below 10 m m it should be determined at the spot indicated by
dotted lines in Fig. 236. Another disadvantage of both the Engler and the Claisen
apparatus is that the bubble point cannot be defined exactly. According to the perti-
nent standard the temperature is measured in an Engler apparatus when the first
22 Krell. Handbook
iranchembook.ir/edu

338 7. Constructional materials and apparatus

drop of the distillate leaves the condenser. Here as in the Claisen apparatus the
distillate contains also higher-bow material which makes boiling appear to begin
at a higher temperature. Due to the small amounts (about 100 g) which are commonly
used in analytical work shifts of the bubble point affect the results strongly (cf.

Fig. 235
Engler distillation apparatus with standard ground joint

Fig. 135). Recently, test distillation apparatus (Fig. 237) for amounts of 200 to
500 ml has increasingly been used. It containsa short packed column and allows initial
operation at an infinite reflnx ratio in order to determine the bubble point exactly.
By incorporating 8 pressure regulator such apparatus gives quit,e reproducible
resiilts when used at reduced pressure (cf. section 8.3). If an auxiliary liquid (chaser),
having an initial boiling point about 5Odeg. C above the final boiling point of the
mixture, is added it is possible to collect the whole of the feed as distillate. Towards
t h r end of the distillation the auxiliary liquid then constitutes the hold-up of the
apparatus [8]. This technique is common in many forms of analytical distillation.
iranchembook.ir/edu

7.2 St,andard apparatus and unit parts 339

Simple unrectified distillation is used in the laboratory chiefly for the recovery
of solvents and other contaminated liquids, and for performing initial separations on
large amounts of mixtures. The only apparatus needed for this purpose is a flask, a
atill head with a ground-in therinonieter and components for condensation, vacuum
connection and collecting the distillate. It is frequently advisable to include a spray
trap consisting of a short section - say 5 to 10 cm - of column. Fig. 238 demon-
strates the wide range of possihilities offered by the use of readily available parts [9].

Fig. 237
Apparatus with column for analytical
work made from Destinorm components

The still head 1 is suitable both for low- and high-boiling substances, whilst the
“bridge” 13 is intended for high-boiling, solidifying materials. Air-cooling usually
suffices for condensing substances of high boiling point and is obtained with coin-
ponent 3 or the reflux condenser 16. If fractions solidify in an air condenser tlhe-may
easily be melted with the aid of a Bunsen burner or an electric heater. A safer course
in such a case is to use component 14 or 15 as condenser, since they have jackets that
may be kept warm and can be cleaned after use. I n this respect the screw-type con-
denser 14 is convenient, as in some versions the screw portion may be removed by
means of a ground joint for cleaning. When low-boiling substances are distilled,
cooling must be thorough and a normal Liebig condenser 5 may prove to be insuffi-

22*
iranchembook.ir/edu

340 7. Constructional materials and apparatus

?5

Fig. 238
Components for simple distillation
iranchembook.ir/edu

7.2 Standard apparatus and unit parts 341

cient; the spiral condenser 6 and the Diniroth condenser 7 provide more efficient
cooling and also allow a more compact, vertical arrangement. For vacuuni work the
Anschutz-Thiele receiver 8 is most usual, since it permits any number of fractions to
he taken without interruption of the distillation. It is also useful for operations at
atniospheric pressure in which gases are evolved. The distillate may be collected in
graduated cylindrical vessels 9 which allow the volunie t o be read off at any instant.

ti
Fig. 239 Fig. 240
Still heads for t h e distillation of *4rrangement of Friedrichs for the
foaming liquids distillation of inflammable substances
a ) Reitmeir b) Friedrichs

The adapters 10 and 17 can be einployed in cases where the distillate is not separated
into fractions. Sketch 12 shows a receiver with which three fractions may be collected,
while version 12 can take 7 fractions. If it is necessary to use a boiling capillary, the
Claisen still head (Fig. 236) may be chosen, or else a three-necked flask with a
capillary 18 inserted into one of the ground joints.
For foaming liquids special still heads are available, such as the I'ear-shaped type
for mild foaming (Fig. 239a), and the Friedrichs cyclone (Fig. 239b).
Very low-boiling or inflainniable substances can be more safely distilled with an
arrangenient due to Friedrichs (Fig. 240). The vent u is connected to a pipe leading
t o the open air.
Apparatus for countercurrent distillation can be assembled from the coniponents
to he described in the next paragraphs. Even complicated arrangements siniulatirig
industrial installations, having more than one col~ininand niiinerous auxiliaries, a$
shown in Figs. 134 and 162 present little difficulty [lo, 111. With the parts of the
Destinorm series designed by the author a large variety of apparatus, from the sinip-
lest equipment for straight distillation, vapour pressure measurenient and equilibrium
iranchembook.ir/edu

342 7. Constroctionel materials and e;pparatus

curve determination to exteneive apparatus for continuous distillation a t normal and

reducedpmsures (Fig.223 and 224), may be set up. A similar series (Labodest) w w
developed by Stage (manufactured by Fischer Labor- und Verfahrenetechnik,
Bonn-Bsd Godesberg), and the components me,nufaatured by the Normag, of
Hofheim, and by the Normschliff-Glaagerte GmbH, Wertheim, can be assembled
for me in the various separations.

$d
Fig. 241 Fig. 242
pi& components Chain of joints (Friedrichs)

Long lengths of piping may conveniently be built up of 1 - 11/, UI sections

(Fig. %l/r). 135" and 90"bends (Fig. 241/2 and 3), T- and Y-pieces (Fig.241/4 and
5 ) are available for constructing any aasembly of pipe, in which stopmcks (Fig.24116
and 7) may be included. To increase the flexibilit>yof lines, hall-joints may be used,
or a chain of joints (Friedrichs:Rg. 242), which gives flexibility in all directions, is a
possibility.
Connections between one Rim of ground joint and another can be made b r means
of adapters (Fig.M),which are obtainable commerciall~7in almost any practical
combination of sizes (cf. Table 52).
iranchembook.ir/edu

7.2 Standard apparatus and unit parts 343

Table 52
Commercially available adapters for various joint sizes

Socket NS 12.5 14.6 14.5 24 29 29 29 29 29 29 29 29 29

Cone NS 14.5 14.5 29 29 24 29 34.5 45 55 60 70 85 100

Joints may also be connect,ed with vacuum tubing by the use of parts such as
that illustrated in Fig. 24317.
-4s can be seen from Table 53 some of the components mentioned above have been
rtandardized.
Piping systems can be set up in a similar manner using plane and ball joint$.
Maximum lengths of the straight sections for semi-technical purposes are of the
order of 3 111. A great varietv of components similar to those shown in Fig. 331 are

I 2 3 4 5 6 7
Fig. 243
Adapters for various joint, sizes
1-4 = cone 2 socket
5 = cone < socket
6 = cone to cone
7 = NS 14.5 cone to rubber tubing

Table 53
Standardized components

Standard Component

TGL 13828 connecting pieces, straight

13 831 connecting pieces, bent (90")
13834 connecting pieces, S-shaped
13835 connecting piece with one-way tap
13836 connecting piece with steam pipe
13837 connecting piece with drop-coant,er
13838 connecting piece with tap funnel
9966 adapters with ground joints
9973 U-pieces with two joint sockets
9978 boiling capillary
All the above components are provided with
standard ground glass joints
DIN 12404/63 boiling capillary with standard ground joint
12565170 dropping funnel
iranchembook.ir/edu

344 7. Const.ructionalmaterials and apparatus

comuiercially available for semi-technical and pilot plants. For details the reader is
referred to Refs. [ I , 6, 71. By now, a number of standards have been issued
(TGL10192-10200, 11641-11644, 12023-12032, 20277 and drafts for DIN
28 800 -28 808).

7.2.1 Taps and valves

As can be seen froiii Table 54, a nuniber of glass taps and stopcocks have been
standardized.
Figs. 189 and 190 show taps for vaciiuiii work. Kinza has devised a three-way tap
[ 121 having the unusual feature that if the plug is always turned in the same direction
no mistakes in position can be made in racuuiii work (Fig. 244). Wolf [ 131 has pub-
lished an excellent review of standardized and special-purpose taps. The latter include
the “Vestale” rotary disk shown in Fig. 245. Straight-way taps, drain cocks, three-
way and four-way taps are supplied for semi-technical purposes.
While ordinary glass taps require grease, those having exchangeable Teflon plugs
can do without it. The Otto Fritz GmbH Normag, Hofheiui/Taunus, make various
tap8 of Duran glass with safety rings. The miniature taps of the French firni Gaflon
are entirely made of Teflon. To connect them with the glass tuhes Teflon gaskets
and screwings are used.
Even with ordinary taps the plug should be protected against coming off (Fig. 217).
As there is always a danger that the valve plug may be withdrawn accidentally when
it it? turned, it is advisable to provide the valve with a rrihber disc or ring, or better
still with a safety clip of steel wire (Fig. 246).
Glass taps with greased ground surfaces, in particular with dimensions > 40 cm,
are unsuitable for semi-technical or pilot plants. In these cases valves are used instead
of taps, especially since the comhination of glass and Teflon has made possible many

Table 64
Standardized glass taps for diBtilhtion apparatus

Standard Component
~

TGL 21 134 Sheet 1 taps, survey

Sheets 3-7 special t&ps
13843 Sheet 1 high-vacuum t a p
2 one-way t a p , dimensions
3 two-way taps, dimensions
DIN 12541/67 Sheet 1 preliminary standard one-way tap cocks, masses
/70 Sheet 1 preliminary standard one-way tap cocks, tests for
tightness
12551/67 Sheet 1 preliminary standard one-way tap cocks,
provided with non-interchangeable,
massive cocks
I2 553/61 two-way taps with non-interchangeable oocka
12 554167 three-way tap cocks with non-interchangeable cocks
12563167 three-way tap cocks with angular bore (120”),provided
with non-interchangeable, massive cocks
iranchembook.ir/edu

7.2 Standard apparatus and unit parts 345

Fig. 244
Kinza’s three-wily tap

a1 bl

Fig. 245
a ) The parts of the rotary disk “Vestale”
b) The rotary disk “Vestale” (DBP 1, 112, 846)

Fig. 246
Steel wire safety clip for tap
iranchembook.ir/edu

346 7. Conatruotional materiale and epparatus

new valve designs [l]. The tap plugs are replaced by Teflon spindles in the valves for
laboratory work rn shown schematically in Fig. 247. Messrs. Sovirel offer preoision
valvea of the “Torion” series for a range of diameters from 2.5 to 10 mm. The valves
are said to be vacuum-tight. down to 10-4 tom. In the valves of semi-technical plants

corner valve

Fig. 247
Precision valve of glass with PTF spindle for the control of liquid and gas flow
(face and tube diameter d = 4 and 7 mm)

Fig. 2411
Pneumatically controlled valve with nominal
widths 25, 32,40 and 50 mm
iranchembook.ir/edu

7.3 Columns 347

closure is achieved by Teflon bellows provided with a Teflon plate or cone (Fig. 248).
The valves may be operated manually by means of electromagnets, electroniotors or
pneumatic drive. This enables distillation plants to be controUed automatically to a
large extent. The glassware manufacturers provide a variety of valves with plane
or ball joints having nominal widths up to about 150 mm. These map be classified
(cf. chap. 5.1.3.2) as:
straight-way valves, drain valves, corner valves,
throttle valves, non-return valves with flaps or balls,
safety valves for overpressures up to 5 atm.

7.3 Colu mns

It was pointed out in section 4.1 that the exchange of material and heat in R
column are processes taking place a t the boundary between the liquid and vapour
phases. A column should therefore present as large a surface as possible for this
exchange, but this surface should not be associated with a large hold-up (section
4.10.5). The best column is one having both a small HETP and a low hold-up.
I n an attempt to classify columns systematically, selection criteria were estab-
lished by Heckmann [la]. The number of column constructions that have been des-
cribed is large, so that only the characteristics of the fundamental types, listed below.
will be described.
Empty columns,
Packed columns,
Plate columns,
Columns with stationary elements,
Coluinnq with rotating elements.
Reichelt [14a] has reported a combination of packed and plate coluinn. The
countercurrent motion of the phases sets the packing (mostly light spheres) which it;
supported by grids (= plates) whirling. Thus, relatively high -throughputs are
obtained. Data on the lower and upper limits of operation of a packed colunin
apparatus with moving packing are given by Hand1 [14b].

7.3.1 Empty columns

Favourable features of empty columns are that they give a low hold-up and a
small pressure drop, so that their main applications are in the fields of micro-distilla-
tion (section 5.1. l ) , high-temperature distillation (5.3.2) and vacuum distillatio~~
(5.4.1). In such columns the exchange takes place between the vapour and the film of
liquid falling down the walls. The possibilities and problems involved in the use of
falling films have been studies in detail by Malewski [15].
Empty columns may be divided into the following types:
a) straight or spiral smooth tubes,
b) tubes with a n increased surface area,
c ) concent.ric-tube columns.
iranchembook.ir/edu

348 7. Constructional materials and apparatus

Owing to their low resistance to flow empty columns may be heavily loaded,
but as a result of the small surface for exchange the efficiency is as a rule low (with
the exception of group 3), particularly at high loads.
Smooth vertical tubes, with diameters of 20-50 nim, are now rarely used for
separations, except as spray traps in simple distilkions. Their HETP will not. be
lower than about IOcni unless the load is sniall, say less than 30ml/h. Attempts
have therefore been inade to improve their efficiency by inerewing their mvface area.
The Wurtz (ef. Fig. 15) and Young columns, consisting, respectively, of series of
simple and pear-shaped bulbs, produce a better separation, but have a larger hold-
up. Data for unfilled columns are given in Table 55.

Table 55
Data for empty columns

Diameter Length Load HETP Hold-up per Refer-

theoretical enms
stage Chap. 7
(mm) Win) W/h) (cm) (mU

Empty vertical tube 30 lo00 25 [I61

Tube: with sintered glass on walls 7 500 12.5 [I61
Vigreux 12 460 54 5.4 r.181
96 7.1 0.46
294 7.7 0.62
540 7.7 0.68
24 460 120 10.2 1.3
240 12.1
510 11.5 1.8
Vigreux, modification of Ray ti 1200 60 2.0 <0.02 [I71
Jantzen 4-6 50-200 4-6 0.2-0.6 [191

Other nirans for increasing the area are the application of an inner layer of
sintered glass [I61 and the provision of vertical and inclined indentations, as in the
well-known Vigreiix coluinn. Slanting indentations also have the advantage of
promoting the redistribution of liquid froni the walls to the centre (Fig. 249). A
modification of the Vigreux colunin by Ray [ 171 has the indentations a t right angles
to the walls. Its separating efficiency is better than that of an einpty tube (see
Table 55). The Vigreux coluim is useful for relatively simple separations, where it is
iniportaiit to keep the hold-up low, as in micro-distillations or high-temperature
distillations of tars and other high-boiling materials that have t o be evaporated to a
snitdl residue [ 181. Similar purposes are served by- the Jantzen spiral coluwm (Fig. 250),
which IS made commercially with inner tubes of 3 and 6 ni (straight) length and gives
a hold-rip of 0.2 to 0.8 1111per theoretical stage. In a test run with the niixtiire carbon
tetrachloride-benzene, perfomied with a Jantzen colunin of 6 m straight length,
iranchembook.ir/edu

7.3 Columns 349

6 n m tube diameter, 50 rrm spiral diameter and 12 rnm pitch, the following data
were determined:
Flooding load 900 ml/h ;
Hold-up per tlheoretical stage 0.4 ml a t 50 ml/h load;
HETP 4.3 cni (load 50 ml/h) :
5.4 cm (load 90 ml/h) :
5.6 em (load 190 ml/h).
For insulation, the Vigteux column is usually provided with a sheath, while the
Jantzen column has a vacuum jacket.

Fig. 249
Vigreux column wit.11air-insulating jacket.

In the use of all enipty columns i t is important for the wall temperature to he
lower than the teniperature of the reflux, since the niaintenance of a uniforni film IS
otherwise impossible.
All empty columns have a satisfactory efficiency only if the load is kept relatively
low. However, if a number of tubes are combined in parallel, the total load may be
raised to any required value. Such a column was patented in 1936 by Fenske [SO].
If a narrow, vertical tube, say, 0.6 cm in diameter is used at a load of 10 nd/h, an
HETP of 1.73 cm may be found [21]. Such a tube, having a length of 1 111, would
produce a pressure drop of 2.7 nim Hg and would have a hold-up of only 0.4 ml.
Starting from such data, Kuhn [22] developed the so-called h a i r p i n countercurrent
principk (Fig. 251). The basic idea is that a small separating effect, occurring under
stationary conditions in a direction a t right angles to the axis of the tube (vector a),
is rnultiplied by means of a flow along the length of the apparatus, changing its
direction at the ends (indicated by arrows u1 and u2).I n order that a high efficiency
may be attained, the vapour velocities must be kept as sniall as possible and the
reflux ratio must be very high. For a reasonable throughput it is then necessary, as
already stated, to combine a large number of tubes in parallel (Fig. 252). The reflux
is produced for all elements collectively by means of a thermostat c and the distillate
iranchembook.ir/edu

7. Constructional materials and apparatus

Fig. 250
Jantzen spiral-tube column with vacuum jacket
a) Straight length 3 m
b) Straight length 6 m
iranchembook.ir/edu

7.3 Columns 35 1

Table 56
Throughputs (grams per 24 h) of Kuhn's multiple-tube column, having 100 tubes, diam 0.2 em,
length 2 m

Initial to final concentrations, 40 + 60 10 + 90 10 -> 99 50 --f 90 1 -- 99

zB zE (mol%)

Difference in boiling 0.1 30 - - - -

point ("C) 0.31 300 5 2.9 22 0.2
1.0 3,000 50 29 220 2
3.1 30,000 500 290 2,200 20
10.0 300,000 8,000 3,900 22,000 200

is taken from the ends of the tubes through superheated capillary tubes d (cf. also
Fig. 157).
Table 56 shows the throughputs, with various initial and final concentration%
xiid various differences in boiling point, for a multiple-tube column containing
100 parallel tubes (diameter 0.4em, length 2 m). The values were calculated [22] for
a molecular weight of 100 and a boiling point of approximately 90°C.
Another ninltiple-tube column, containing 61 tubes of 0.2 cm diameter and 1.5 in
length, was eiiiployed for the separation of the isotopes 12C-13C, where the difference
in boiling point was only 0.03deg. C. The number of stages was determined Is>*
analysis of the products in the mass spectrometer. The hold-up of a tube 2 m long

Fig. 251 Fig. 252

Diagram of the multiplication of Multiple-tube column (Kuhn)
separating effects by the hairpin
countercurrent principle of Kuhn
iranchembook.ir/edu

352 7. Constructional materials and apparatus

and of 0.2 cm diameter amounts to about 0.5g [22]. The method has also been used
Ruccessfully for the separation of close-boiling isomers, such as the xylenes'(Fig. 263)
and the isomeric airipl alcohols. This type of column would appear to be promising for
many problems of separation requiring inore than 100 stages, since its HETP is
extremely small.

Fig. 253
Separation of the isomeric
xylenes with Kuhn's multiple-
tube column, length 0.8 m

take-off

4-

Fig. 254
Concentric-tube column (Kuhn)
B = still pot, H = electric heater, M = vacuum
jacket, S, and S, = concentric tubes, K = con-
denser. G = cold trap, F = vacuum connection

The calculating methods were discussed in chap. 5.1.4.2 for ideal mixtures.
Craig [23]has described a concentric tribe column with narrow gap which is suitable
for micro-scale work, its capacity being 0.25 ml. The annular space for distillation is
provided by a t.hh rod being placed in a capillary.
Kuhn's starting point for the calculations on the multiplication of the separating
effect [24,261 was a column having an annular cross section, formed by t,wo concentric
cylinders (Fig. 264). Such column8 are known as concentric-tube columns (compare
iranchembook.ir/edu

7.3 Columns 353

Fig. 257). They have also been studied extensively by Westhaver [26], Donne11 arid
Kennedy [37] and recently by Jantzen and Wieckhorst [28],who conipared them
with columns containing Stedman packing and wire spirals. As shown by Fig. 5 5 ,
the concentric-tube column is superior to the last-nanied types a t loads below
100 ml/h. The loss in pressure in a concentric-tube column can be calculated [21] hp

throughput --+

Fig. 255
Number of theoretical plates as a function of the throughput [28]
a = concentric-tube column, diam. 2.5 cm (gap, 0.4 cm), b = Stedman colurnn,
diarn. 2.5 cm. c = packed column with helices 4 x 4 mm, diam. 2.4 cm

"'I
10

-
Q 10l5
2 2
-0 10-z5
L
2 10-3;
$ J 2
QIO-4
5
10-5 2
2 4 6810 20 4060100 g/h 1 I00
throughput
Fig. 256
Pressure drop as a function of the throughput a t atmospheric pressure, referred
t o 50 theoretical plates [28]
a = concentric-tube column, b = Stedman column, c = packed column with
helices 4 Y 4 mm
iranchembook.ir/edu

364 7. &mtNctionai materials and apparatus

means of the formula :

~vJQ
1’1 - P2 = -
2 ba=
in which p , - p 2 = the pressure drop in dynes/cm2,
11, = the viscosity of the vapour in poises,
1 = the length in cm,
Q volume of vapour per unit time (cma/sec) at the average pres-
= the
+
sure (PI Pp), +
b = the average circumference of the gap in cm,
u = half width of gap in cm.

Fig. 267 a)
7. Section of a concentric-tube column (Fischer)
1 = gap for vapour flow, 2 = inner tube with profile
for spiralling reflux of liquid, 3 = outer tube with inverted
profile for reflux of liquid

Fig. 257 b)
Fully automatic concentric-tube column for semi-micro operation
Cotctrol eyatem: electronic control for oil bath according to temperature or
differential column pressure - electronic control or follow-up control for heating
jscket - reflux control with vapour division - vacuum control - fraction
collector control - recording of all d a t a - protection against superheating and
interruption of cooling water supply
2 = oil bath with variable height support, 2 = resistance thermometer oil
bath, 3 = still pot, 4 = thermometer residue, 5 = motor for stirrer, 6 = con-
centric-tube column, T = compensating heating jacket, 8 resistance thermo-
meter heating jacket, 9 = column head with condenser, 10 = resistance thermo-
meter for head, 11 = valve armature, I:! = magnet coil, 13 = vacuum receiver.
14 = distillate condenser, 25 = fraction collector, 16 = support for receivers,
27 = magnetic coupling, 18 = automatic drive, 19 = condenser for differential
p w u r o measurement, 20 = switch cupboard, 21 = 6-colour cornpensating
recorder, 22 = Minitron 5, control device for still pot, heating jacket a n d column
head, 23 = differential pressure controller and cooling water safety device,
24 = vacuum controller, 25 = fraction collector control, 26 = cupboard for
accessories
iranchembook.ir/edu

7.3 Columns 355

20

-2f

-22

-23

124

-25

- 26

23*
iranchembook.ir/edu

356 7. Constructional materials and apparatue

In comparison to the two other types of column mentioned, the concentric-tube

column, in the region examined, caused a preesure drop lower by a factor of about
10 (Fig.256). Koch and Feind [29] determined the pressure drops of concentric-tube
c~lumn ~ gap widths from 4.06to 19.68 mm, the diameter of the outer tube being
with
49.96 mm in all cases. The measurements were carried out under isothermal con-
ditions. Other studies concerned the heat transfer with the gap heated via the outer
tube.I h k [30]investigated the heat transfer and pressure drop in vertical concentric
and excentric gaps with fomed flow and free convection.
Donnell and Kennedy [27] had previously demonstrated that with concentric-
tube columns an HETP of 0.5 to 0.6 cm can be attained if certain requirements are
fulfilled :
1. a uniform gap must be present throughout the whole length of the column; this
may be attained by the me of precision bore tubes:
2. the mll~ must be uniformly wetted, for example by etching out a spiral;
3. adiabatic operation (obtainable by proper: insulation' or thermostatic jacket
heating) is essential ;
4. the column must be uniforndy loaded (e.g. by automatic control of the rate of
evaporation).
I n a number of papers Fischer [31] has reported his work on concentric-tube
columns and the experimental results obtained. He developed all-glass apparatus
which is particularly suited for a mild countercurrent distillation of high-boiling
substances, the pressure drop being of the order of 10-1 to 10-3 t<orrand the operat-
mg hold-up lying between 0.1 and 1.5 ml. The apparatus is highly automated and
can be used for pressures from atmospheric down to torr i o that still pot press-
ures below 0.1 torr are reached. Intense exchange of material is secured by an opti-
mum hydrodynamic design of the tubes as shown in Fig. 257a. The optimum vapour
velocity is given by

where D = diffusion coefficient of the vapour product under operating conditions

and a = gap width.
Fig.257b Rhows the schematic diagram of a fully automatic concentric tube
column for work on a semi-micro scale as developed by E'ischer. It ia manufactured in
7 versions with 10 to 80 theoretical stagea for still pot capacities from 0.6to 4,000 ml.
AcontroUed oil bath provides a constant temperature of about 0.1 "C. Further details
can be gathered from the legend (see also chap. 8.1.2).

7.3.2 Packed columns

This type of column consists of a tube filled with packing in a random manner
(cf. section 4.10.2). The packing is supported by some form of grid (Fig. 358) or by a
perforated funnel fused into the column; the latter arrangement is employed in the
colamns of Stage's Labodest series and the author's Dest inom series of components
iranchembook.ir/edu

7.3 Columns 357

(Pig. 289). The support should have a t least the same free cross-sectional area as the
packing resting on it. It is the usual practice t,o place a few larger-sized packing units
on the grid before introducing the actual packing in order to minimize flooding. As a
general rule the size of the packing should be about l/lOth of the column diameter.

a) b)
Fig. 258
Grids for supporting packing
a) spiral b) flat

Fig. 259 Fig. 260

Perforated funnel as Column with restrictions
packing support (Grosse-Oetringhaus)

In order to prevent the reflux from running down the walls of the column (sec-
tion 1.2) it is desirable to collect it and lead it back to the centre a t short intervals.
Grosse-Oetringhaus [32] attempted to do this by providing the column with restric-
tions, 2 inm deep, a t distances of 30 mni, as shown in Fig. 260. A better method is to
divide the coliimn into sections, the reflux from each section being collected and
redistributed (section 4.10.2).Alternatively, discs of wire gauze with conical depres-
sions mag be employed. They have the advantage of causing almost no dead volume
[33].
The diameter of the column should he chosen according to the required throughput
(section 1.11). Columns for normal laboratory use have diameters of 10-50 mm
( t - 2 inches); if the diameter is larger, the operation is generally classed as semi-
technical. In pilot plants, column diameters range from about 150 to 400 mm.
iranchembook.ir/edu

368 7. Constructional materials and apparatus

The effective lengths of these coluiiins were purposely restricted to 500 and
l000mm so as t o ensure that tests with various kinds of packing would be coni-
parable. If a larger number of separating stages are required several column sections
can he connected in series. Should this number of stages have to be fairly exact, as
for comparisons with an industrial installation, it may prove necessary to employ a
section of non-standard length : in analytical work, however, this will seldom occur.

7.3.3 Plate columns

In the laboratory and on a semi-technical scale, plate columns are used mainly for
special purposes, for instance in constantly-occurring separations where a distillate of
high purity has to be prepared. Distillations carried out for comparison with in-
dustrial plate columns may similarly require the use of laboratory plate columns. I n
analytical work at atmospheric pressure the sieve-plate column has proved reliable.
In industrial cont,inuous distillations plate columns are preferably used [34].
Recent developments of laboratory and denii-technical apparatus have been thorough-
ly discussed by Stage who has also dealt with the question of systematic tests for
efficiency [36]. The advantages of bubble-cap columns in continuous operation have
already been mentioned (cf. chap. 5.2.2.3 ; Fig. 167). Cyclic operation (Fig. 164)
yields efficiencies > 150%.
Frotn a constructional aspect plate colunin may be subdivided into IruMte-plate,
hlbble-uzp and trieve+.de columns. All of these are employed mostly a t atmospheric
pressure, as their relatively high resistance to vapour flow gives rise to an appreciable
pressure drop.
The action of a plate column is based on the fact that a certain auiount of liquid
IS present on each plate and that the ascending vapour is forced to pass through
thiR liquid. The reflux flows down from each plate to the next through tubes (down-
corners), which map be situated either inside or outside the column (cf. chap. 4.2
and 4.7).
It should be noted that valve-plate colunins with nominal widths of 60, 100, 200
and 300 mm have recently been employed for laboratory and pilot plant distillations
[35a]. They are made in Czechoslovakia, the constructional material being Sininx
glass. They can be used at reduced pressures for distillate boiling teniperatures up to
115 "C.
A bubble-plate column due to Keesoni [36] is shown in Fig. 261. 9 layer of liquid is
formed in trough a ; the reflux is conducted downwards by tube b. As in all bubble-
plate and bubble-cap columns the vaponr, in traversing the liquid, undergoes a
change in direction; it here passes under the rim of each section and bubbles up
through the liquid layer. Owing to its relatively low stage efficiency Keesoni's
column is now seldom used. Bubble-cap colunins, on the other hand, have been
increasingly used for various purpows, especially for the determination of data for
scahng-up. According to Stage [36], they are best suited for continuous operation on
the laboratory and semi-technical scales because they have a wide range of loading
and their efficiency is largely independent of the load and the liquid-to-vapour ratio
as well as of short-time fluctuations in the control. The Bruun bubbZe-eap colurrm
iranchembook.ir/edu

7.3 Columns 359

[R7] has found wide application in the laboratory (Fig. 262). It resembles industrial
bubble-cap columns in construction, but has the disadvantage that the downcomers
are outside the column. There is thus a tendency for the reflux to lose heat. I n order
to counteract this tendency to some extent the column is provided with an airor
vacuum insulating jacket (Fig. 263). The reflux tubes a and b are so constructed as to
maintain a liquid layer of about 10 mm on the plates. The vapour passes 1113 through

a
b

Fig. 261
Bubble-plate column (Keesoni)
a = weir for liquid level.
b = downcomer

a
b
4)

Fig. 262 Fig. 263

Bruun’s bubble-cap column Bruun’s bubble-cap columns
a = reflux tube (in), b = reflux a ) with 40 actual plates a n d
tube (out), c = vapour tube removable air jacket
(riser), d = serrated cap b) with 20 actual plates and
vacuum jacket
iranchembook.ir/edu

360 7. Constructional materiala and apparatus

risers c, which are covered by loose caps d ; the latter are serrated along their lower
rims, so as to break up the vapour into small bubbles.
A column accurately duplicating industrial columns is Krell’s bubble-cap column
(F‘ig. 139), which has internal downcorners. It is made in diameters of 50 to i0 iiim
and has proved particularly useful for carrying out investigationfidesigned to parallel

Fig. 264a) Fig. 264b)

Vspour path through collision plateF Vapour collision plate (Stage)

4
Fig.1265
Stage’s bubble-cap:column,
with shielded reflux t,iibes
iranchembook.ir/edu

7.3 Columns 361

large scale distillations. This column can also be equipped with sampling taps and
thermometer wells. Heated, hinged column jackets are available for insulation.
The multi-chamber column of Klein, Stage and Schultze [38] also has internal
downcomers. The vapour is distributed on each plate by a number of holes arranged
In a circle, the intention being to produce small bubbles giving a thorough mixing of
the liywd. Stage has also developed a bubble-cap column provided with so-called

200 3.0

80 2.5

1 60 LO 2
t
x
C
-ma
? I
u a
c a
al
a 40
*
1.5 g
-a
Q
E
In
I
?!
a
20 1.0

0 0.5
1 2 4 6 8
reflux Load for R = oc,
Fig.
u
266
Efficiency and pressure drop as functions of the reflux load (l/h) at 760 tori
for various bubble-cap columns with dimensions RS given in Table 57
I = Labodest bubble-cap column with vapour collision plates, 2 = Labodest
vaponr collision plate column, 3 = Labodest bubble-eap column with shielded
reflux tubes, 4 = bubble-cap column (Schmiickler-Fr‘itz), 5 = Brand buhhlc-
cap column, 6 = Normschliff bubble-cap column

“vaponr collision plates”, in diameters of 40 to 130 mni. The vapour enters a t the cir-
cumference of a rotating layer of liquid on the plate (Fig. 264a) and then impinges on
the opposite wall. The vortices produced in this way favour mass transfer, and thiq I >
further promoted hy the long path covered by the liquid from its entry to cwt
(Figs. 264b and 167). The plate efficiency varies little with the load and lies between
70 and 90°;,. Another type of bubble-cap column due to Stage (Fig. 265) has shieldtd
downcomers and permits high liquid loads t o be applied. Fig. 266 provides a coiii-
parison of pressure drops and plate efficiencies for the mentioned columns. Table 57
contains data for various bubble-cap columns and Fig. 267 displays their constrnc-
tiond differences.
iranchembook.ir/edu

362 7. Constructional materials and apparatus

Table 57
Cross-seotional area8 for vapour and liquid flows in 6 plate oolumns as compared by Stage
[35] (for the meaning of numbers 1-6 cf. Fig. 266)

No. Dimension Column No.

1 1 z 3 4 5 6
2 Outer diameter (mm) 55 55 55 55 62 58
:I Inner diameter (mm) 51.5 51.5 51.5 51.5 67.5 54
~ ~

4 b h m n cross-section (cm*) 21 21 21 21 26 23
.i Outer vapour riser cross-
section (em') 3.51 7.8 2.58 3.31 7.9 3.29
ci oo of 4 16.7 37 12.3 15.7 - 30.3 14.4
7 Inner vapour riser cross-
section (cm') 2.44 5.2 2.12 1.13 3.5 1.76
8 o/, of 4 11.6 24.7 10.1 5.4 13.7 7.7
9 .hof openings produced
by serration ( c d ) 5.58 3.1 3.1 2.25 1.82 2.1
10 O/, Of 4 26.65 14.7 14.7 10.7 7.0 9.2
11 Area covered by cap (mia) 5.58 1.83 1.83 0.85 4.33 1.0
12 ",of 4 26.65 8.7 8.7 4.06 16.6 4.36
13 Total cross-sectional area
for vaponr flow (em2) 11.16 4.93 4.93 3.10 6.15 3.10
14 oA of 4 53.3 23.4 23.4 14.8 23.6 13.6
I-i ('roes-sectional area for
vapour flow outside cap
(Cm2) 9.7 4.15 13.45 10.0 8.83 16.3
16 Oh of 4 46.3 20.0 64.0 47.5 33.9 72.5
17 Area for vapour flow within
the cap (em?) 9.74 7.17 6.38 6.6 7.22 6.4
id 04 of 4 46.2 34.1 30.4 31.5 27.8 28.0
19 Cross-sectionalarea. for
liquid flow (cm') 1.M 0.67 0.29 0.6 0.39 0.5
50 "A of 4 4.93 3.15 1.38 2.38 1.5 2.19

Sipre-plate columns have perforated plates on which IL small amount of liquid is

iriaintained by the pressure of the ascending vapour. They therefore require a uiini-
tniitn load, as the liquid otherwise falls through the perforations. This type of column
hit* 1trnvrd excellent for analytical distillations of low-boiling hydrocarbons. 111 thc
~ ~ t w ~ - p l roliimn
ate of Oldershaw, as modified by Groll [all, the holes (dim.0.75 to
I i i m ) are arranged in circles and pass vertically through the plates (Fig. 268).The
wflus flows down through a central tube to the next plate. (For the pressure losses in
these columns see Table 33, section 4.11.) The sieve-plate column of Sigwart, on the
(it her hand, has external downcoiners and cup-shaped plates with horizontal holes
(Fig. 569).
iranchembook.ir/edu

7.3 Columns 363

Fig. 267
Recent bubble-cap columns
a) column due to Schmiickler Nnd
Fritz [40]
b) Brand column [35]
c) Normschliff column [351

Fig. 268
Sieve-plate column (Oldershaw-Groll)
a) with vacuum jacket and expansion bellows
b) in operation
iranchembook.ir/edu

364 7. Constructional materids and apparatus

The plate efficiencies of sieve-plate columns have been determined by a number

of investigators under various conditions. Van Wijk and Thijssen [43] tested a
column of t>histype with an I.D. of 38 mm (1.5 in.), having 8 actual plates spaced a t
distances of 120 mm (5 in.). With the test mixture n-heptane-methylcyclohexane it
was found that the plate efficiency fell off markedly when the concentration of one of
the components was reduced to about 0.1 moloh. Depending on the vapour velocity
(10-27 cm/sec), the plate efficiency had a value between 86 and 75%. At a constant
t-apour velocity (17 cmlsec) and a concentration in the still pot of XB = 60 molyothe
plate efficiency varied with the reflux ratio (0.54 to 1.0) from 65 to 85%.

Fig. 269 Fig. 270

Sieve-platecolumn of Sigwart Perfo-dripsieve-plate column [46]

Uinho z and van Winkle [44]Carrie, out investigations on a sieve-plate column

of 25 nim (1 inch) diameter and 50 mm plate spacing. The free cross-section was
16.20,h. The hghest efficiency was obtained when the diameter of the holes in the
plates was 1.9 mm. Tests with an o-xylene-p-xylene mixture showed that the effi-
lciency dropped from 90 to 65% on reduction of the pressure from 7M) to 50 mm El&.
When the load waq increased the efficiency first rose, but from 200-250 g/cm2h on-
wards it remained nearly cqnstant,. In a later investigation [a51 the influence of the
plate thickness and of the properties of the mixture (surface tension, density of
rapour and liquid) were examined.
A recent form of perforated plate column, the “Perfo-drip” column [46], consists
of a tube of very uniform inside diameter into which are inserted a number of per-
forated plates of stainless steel, fitting onto a central rod so that the plate spacing can
he varied at will (Fig. 270). The insert can also be made in Teflon. The column is
stated to have a higher efficiency than the Oldershaw one and is easy to clean.
By making the vapour pulsate McGure and Maddox [47] could increase the
efficiency of an Oldershaw sieve-plate column of 32 mm diameter and 30 mm plate
4pcing (each plate having 82 holes of 0.85 mxn diameter). For a given system the
efficiency depended on the amplitude and frequency of pulsation.
iranchembook.ir/edu

7.3 Columns 365

A basic requirement for all plat,e columns is that the distance between the plates
should be sufficient to prevent mechanical entrainment of liquid. Liquid carried
upward by the vapour flow would markedly decrease the efficiency. Wagner et al.
[48]made experiments with radioactive tracers in a vacuum distillation apparatus
to determine non-volatile impurities in the distillate. I n connection with mechanical
entrainment Newitt et al. [49] made theoretical and experimental studies of the
mechanisni of droplet formation and of droplet size distribution.
The hydrodynamic conditions prevailing in sieve-plate and bubble-cap columns
have been discussed by Melikyan [50].
The properties of the plat>ecolumns dealt with above are summarized in Table 31
135,381. Since the plate efficiencies of the various columns have not been determined
under uniform conditions, the figures given are not directly comparable ; they serve
only as an indication of the magnitudes that can be attained.

7.3.4 Columns with stationary elements

In distinction to plate columns, in which the various parts are permanently built
in, and packed columns, which contain packing in a random arrangement, the columns
to be described now are provided with some loose, regularly-arranged form of solid
or perforated contact element.
On a technical scale those inserts are now called packings. On the paclung surfaces
a thin film is formed by the falling liquid which, in the case of perforated material,
gives high degrees of wetting (cf. chap. 4.2). These, in turn, result in the relatively
high efficiency of these "wetting columns".
They niay be divided into the following types:
a) Columns with continuous glass spirals ;
b) Coluriins containing wire-gauze spirals ;
c) Columns containing wire helices;
d) Colunins containing Stedman packing and inclined-film packing.
The gluss-spirtrl column of Widmer [51] (Fig. 271) was evolved from a column
having a syphon closure and concentric tubes described by Golodetz [52]. A basic
shortcoming is that it does not operate on the countercurrent principle, and it is not
to be recommended. A column designed by Dufton, containing a glass spiral with a
pitch diminishing towards its upper end (Fig. 272), has a small hold-up but a rela-
tively low efficiency. I t s HEW is generally more than 10 cm.
An appreciablj- higher efficiency is attained in a column due to Lecky and Ewell
[53], which is packed with a wire-gauze spiral formed by shaping a strip of gauze in
a screw form around a glass centre (Fig. 273). Its HETP is of the order of 1-5 cin,
and the hold-up approximately 0.5 nd per theoretical plate. The construction of the
spiral, which is not easy, has been described by StaUcup et al. [54]. An insert inore
easily made is that of Bower and Cooke [55], which is constructed to fit a diameter
of about 5 mm. A Monel gauze strip (approximately 120 mesh) is so bent that vertical
surfaces stand above each other a t a n angle of go", whilst the horizontal surfaces form
two 90" sectors (Fig. 274). A convenient way of inserting the spiral is to uioisten it
iranchembook.ir/edu

366 7. Constructional materials and apparatus

with oil, attach a copper wire to its end and pull it into the column, which should
preferably be made of precision bore tube. The oil is removed with a solvent and the
copper wire may be dissolved in concentrated nitric acid. Data for this column are
given iu Table 68.

Fig. 271 Fig. 272 Fig. 273

Widmer’s column with Dufton column with Wire-gauze spiral
concentric tubes and glass spiral column of Lecky and
glass spiral Ewell

The surprisingly small hold-up should be particularly noted. The pressure drop
is also relatively low, so that this column is especially valuable for vacuum distillation
a t pressures down to 1 mm with small charges (6-15 ml).
An insert that is more difficult to construct, but has a higher efficiency, is the
wire-wound type of Podbielniak [66], known aa “Hdi-grid” packing. It consists of a
helix of wire laid as a screw around a central core (Fig.275), here again it is essential
for the packing to make good contact with the wall, SO that the reflux does not flow
iranchembook.ir/edu

7.3 Columns 467

Table 58
Data for column with wire-gauze insert (Bower
and Cooke) (Fig. 274)

Load HEW Hold-qp per

theoretical plate
WIh) (em) (mu

38.5 1.88 0.045

83.0 2.28
84.0 2.54
110.0 2.79 0.066

down the outside tube. Heli-grid packing is made for colimin diameters of 8-30 inn1 .
the wire is 0.25 mm thick and is wound with a spacing of 0.26 mni between the turns.
Data for columns with this packing are collected in Table 59. Another forni of
packing due to Brezina consists of a central pin coated with a glass fabric around
which R tube of glass fabric is wound as a spiral. Thc HETP values lie between 1.60
and 2.37 cin. The high efficiency of Heli-grid packing is dne to a film of liquid retained
hg capillary action between the turns of wire [57].
On account of their extremely low HETP’s, even a t high loads, Heli-grid coliuiitih
itre of value for analytical distillations. Together with the concentric-tube and ~ n u l -
tiple-tube columns they constitute the most efficient group of coliitnns at prpscnt
known.
Another type of element, known as Stednian packing [58], consists of wire-gauze
pieces of double-conical shape (Fig. 276a). It also is characterized by a high effi-
ciency coupled with a low hold-up, and it can operate with large loads. The difficnlty,

Fig. 274 Fig. 275

Wire-gauze insert (Bower and Cooke) Podbielniak’s Heli-grid packing
iranchembook.ir/edu

388 7. Congtmctional materials and apparatus

Table 59
Data for columns with “heli-grid” paoking

Diameter Load HETP Hold-up per

theoretical plate
(mm) (ml/h) (4 (ml)

11.0 200 0.9 0.07

246 1.14
315 1.4
375 1.64 flooding point
25.0 500 1.0 0.33
lo00 1.26 0.53
1600 1.58 0.76
2000 1.90 1.07

again, is to pack it in such a way that the reflux cannot flow down the walls. The
use of precision-bore tubing is hence almost essential. The holes a, though which the
vapour y e s , are on opposite sides of the column. Koch and van Rmy [69] con-
struoted a modified form of this packing, conaieting of spherical segments and hence
possessing some ehticity (Q. 276 b). The rolled edge, on account of its springiness,
fits closely to the column wall. It is consequently possible to employ tubes varying as
much as 0.6 -1 mm in diameter. The original and modiiied f o m , nevertheless, yield-
ed about the same HETP when tested with n-heptane-methylcyclohexanemixture,
a8 will be seen from the figures lieted in Table 60.
According to Bragg [60] the theoretical plate number of columns containing Sten
man packing of 10-3300 mm diameter can be calculated by the followingformula:
6.5 0.5
118 = 2.8 + Lo.’; i L”.9 (197)

in which n, = specific number of theoretical plates per foot of effective column

length; L = amount of reflnx (gals/h) at the operating temperature.

a
d

Fig. 276
Stedman packing
a) Original form
b) as modified by Koch and van RaaF
- a)
iranchembook.ir/edu

7.3 Columns 369

Table 60
Data for a column with original Stedman packing and
Stedman packing as modified by Koch and van Raay
(Column diameter: 25 mm; column length: 1 m)

Packing Load Reflux HETP

(ml/cm2. h) ratio (cm)

Original Stedman 30 CQ 1.25

100 1.96
200 2.42
220 2.54
Koch and van Raay 25 40:l 1.43
50 100:l 1.59
120 100:1 2.35

A simple form of element is the inclined-filr packing [61]. Elliptical leaves of

stainless steel gauze are arranged as a eig-zag in the column. Here again the vapour
holes are diametrically opposite to each other (Fig. 277). The packing is made for
coliiizin diameters of 15,30and 50 mni and has the following characteristics (Table 61).

Fig. 277
Column with inclined-film
packing

24 Erell, Handbook
iranchembook.ir/edu

S O 7. Constructional materiele and apparatus

Table 61
Data for a column with inclined-film packing

Diameter Number of leaves Reflux HETP Hold-up per

theoretical plate
(ml/h) (cm) (mu

1.5 230 250 1.82 0.4

30 115 400 2.40 0.83

Packing5 are increasingly used in vacuum distillations down to 1 torr on a semi-

technical scale. Their chief advantage is high efficiency with low pressure drop. If the
packing can he shaped such that the aero- and hydrodynamic conditions remain con-
stant throughout the column operation (cf. chap. 4.2) the separating efficiency will
he indrpendent of column diameter over a wide range. Following the principle of the
Stcdnian packing inserts structured as fabrics began to be developed [62]. In a newly
developed distillation apparatus the Multifil paclung was tested by Juchheim [63].
This IS a special-steel wire net,ting with excellent capillary action for the liquid and a
free cross section of about 95% for the gae (Fig. 278). The packing is available in
standard lengths of 100 nun with diameters from 25 to 87 rum. The number of sections
used depends on the required separating length. With electrical heating of a 1 in
colnrnn 60 theoretical stages (BETP = 1.06 cm) were meamred using o-xxlene-m-
xylene as test mixture. For seniitechnical plants Hyperfil packings are supplied in
sections of 100 inni length and diameters from 100 t o 375 (as a maximuiu, 900) nim.
Their distributing power for the liquid phase is quite satisfactory up to :300 nun dia-
nwtrr. With greater diameters the distribution of the liquid should be enhanced hy
special arrangements (Fig. 279). For use on a technical scale Huber et al. [a] clevel-
oped the Sulzer packing which also consists of wire netting or glass-fibre fabric s w -
tions. It has regularly arranged flow channels (cf. chap. 4.2) which are at an angle to
t h r coluriin axis. Paokings consisting of thin perforated nietal or ceramic plates are

Fig. 278
“Multifil” wire-netting packing
iranchembook.ir/edu

7.3 Columns 37 1

based on the sanie principle [64a]. Sulzer packing made of plastic is available in the
form of cylindrical synthetic-fibre (polypropylene and polyacryl nitrile) fabric
sections of 200 nim length [64b].
A survey of the present state of the varions types of Sulzer packing given by
Meier [sac] includes a comparison of performances. Besides, the use of heat pumps
and the activities of a special laboratory for process engineering are reported. The
applications mentioned show the versatility of Sulzer packings, particularly in
vacuum distillation.

Fig. 279
“Hyperfil” packing

Systematic tests of wire netting inserts with dianieters of 28 and 150 nini were
made by Tiniofeev and Aerov [65] for vapour velocities from 0.07 to 0.47 m/s. They
varied, above all, the angle of inclination, the flow channel depth and the specific
surface.
Besides these wire netting elements which are mostly quite expensive a niiniber of
other packings have been developed all of which required that the liquid be evenly
distributed over the walls of the packing (Fig. 280). They may be made of expanded
metal, glass or some ceramic material. According to Stage (in a prospectus of the
ACV) packings with meshes of sizes 6 x 6 and 10 x 10 min having diameters from
35 to 150nini are available in short and standard sections of 200 and lOOOmm,
re~pectively.Pig. 381 shows an ACV packing of 200 mrn diameter. In Fig. 580
the scheniatic diagram of t h e distributing head of a laboratory column is represented.
24*
iranchembook.ir/edu

372 7. Constructional materials and apparatus

Fig. 280
Diagram of the distributing head of a laboratory
column for falling film operat,ion

Fig. 281
Wire-mesh packing8 with diameter 200 mm (ACV)
iranchembook.ir/edu

7.3 Columns 373

The packing developed by Kwasniak [65a] consists of zigzag strips of metal which
are oriented differently. EndEst et al. [65b] examined a packing made of expanded
metal with a mesh width of 10.5 nim. They used a glass column of 50 mm diameter
and 1.5 ni separating length which was operated a t various pressures and loads with
ch1orc)benzene-ethyl benzene as test mixture. At 50 torr they found HETP values
of 45 cin and ~ ~ r e s s udrops
re of 5 to 15 nini column of water per separating stage.
Packings consisting of glass-fibre fabric and metal foils with quartz sand displaying
R so-called capillaryv-hydrodS.namic effect were studied by Tiniofeev et al. [65c] for
liquid flow rates of 14 to 600 cin3/min.
Coliiiiins operating on the principle of simple distillation were developed by
Shavoronkov and Malyusov [66]. Exchange takes place in packings containing
- vertical channels of circular or rectangular cross section in which the liquid is
carried upward by the vapour;
- channels or other guiding structures in which turbulent flow and slmyinp result
in the two phases penetrating each other in their upward flow.
Experimental results and calculating methods are reported. Further it is pointed
out that it is particularly difficult to separate a sufficient amount of liquid after each
stage. Recent developments are directed a t providing uniforni channels h v introduc-
ing the packing into the columns as regularly ordered structures. Thus, iising 25 111111
Pall rings Billet et al. [66a] obtained considerably smaller pressure drops and twice as
high gas velocities at flooding onset as coinpared 1o irregular packing.

7.3.5 Columns with rotating elements

Colunirir provided with built-in, rotating elements were first proposed in 1925.
I n view of the r e d t s of Jost et al. [73] the qiiestion has arisen whether this t \ p e of
column really offers so great advantages over plate and packed columns as well as
over coluinns with qtationary elements that furtlrer development would seem proniis-
111g.
Their evolution was pronipted hy the need to improve the separating e€fect
without increasing the hold-up and the pressure drop. At that time an efficient
colrimn was lacking for certain classes of work, particularly for chargrs of 1-5 g.
The columns belonging to this group can he subdivided into the following t,ypes
(cf. chap. 5.1.1):
1 . Spinning hand columns, containing a flat, spiral or cross-shaped rotating n i c tal
strip (Fig. 282);
2. Rotating concentric-tube colunms, having a stationary and a rotating cylinder
separated by a gap of 1-2 rnm (Fig. 286);
3. Rotating condenser coliinins, of which the column wall is heated and in which a
cylindrical condenser revolves (Fig. 283):
4. Conical segment colunins, provided with rotor and stator elements (Fig. 281),
5. Columns containing rotating wire spirals (Fig. 285).
The first column with a rotating element, constructed by Myers and Jones, was
intended for laboratory use and contained conical srgments. Urey and Haffniann
iranchembook.ir/edu

374 7. Constructional materials and apparatus

built a column operating on the same principle for the enrichment of the oxygen iso-
tope 180; it had a diameter of 16 cm, a height of 10.7 metres and an efficiency corres-
ponding to Mx)tlieoretical stages. Lesesne and I m h t e introduced a narrow rotating
&rip into a tube and thus created the basic model of the spinning band column, which
wafi modified and improved by several const,mctors.

Fig. 282a) Fig. 282 b)

Principle of Abegg’s spinning bend Spinning band column (Abegg)
column

Fig. 282a gives a schematic diagram of the spinning hand colrrum of Abegg. The
liquid moves counter to the vapour in the forni of a fine spray. This yields HETY
values of 2.5 to Scm. The band rotates with between loo0 and 3600 r.p.m. The
total hold-up of a spinning band column of 375 mrn length and 6 mni inner diameter
is only 0.2 ml and the presslire drop for 1 ni of coliimn height. is of the order of 0.5 torr.
This make8 this type of column very useful for the countercurrent distillation of
extremely h i g h - b o w substances since under these conditions decoiiiposition is
avoided to a large extent. Spinning band column data are listed in Tables 37, 83,63
iranchembook.ir/edu

7.3 Columns 375

in I t out

condenser
driue motor
7 condenser water

out - 5
U

vacuum
t

column heating jacket

water- cooled
rotating condenser

Fig. 283
Botating condenser column (Byron, Bowman and Cooll)

Fig. 284
Conical segment column of Slyer8 and Jones
iranchembook.ir/edu

376 7. &nstructionel materide and appe~atus

Fig. 285
Wire-spiral centrifugal column of Podbielnink

Fig. 286
Rotating-oylinder column (Jost)
a = Flask, I = Rotor, c = Ball bearing, d = Drive shaft, e = Gap, f = Conden-
eer, g = Take-off, h = Flange, i = Thermowell, k = Differential thermocouple,
1 = Heating jacket, m = Insulation
iranchembook.ir/edu

7.3 Columns 37;

and 64. One disadvantage is that the efficiency decreases strongly with diniinishing
preasnre. Thus, from an operating pressure of 760 torr to 10 torr the drop aniountq
to 70°/,, [67].
3 largely automated apparatus after Abegg (Fig. 28213) with a column of 6.5 inin
diariiei er and 550 mni length has a spiral-shaped springy hand of stainless steel 01’
Teflon which rotates with 1400 r.p.ni. The reflux is regulated by nieans of a swing-
ing rod. The total hold-up is about 1.2 nil, the efficiency corresponds to 30 theoretical
stages for loads up to 70 mllh.
The automatic micro-vacuirm spinning band column of Messrs. Ernst Haage, of
Miilheini/Ruhr, is designed for capacities from 2 t o 100 i d . It is made of quartz a n d
has a length of 400 min (25 theoretical stages) or 1000 nini (50 theoretical stagey).
The “centrifugal superfractionator” of Podbielniak [68] contains a rotating spiral
and allows higher throughputs to be obtained than with the other coliimns of thi.
group (Fig. 285).
In the columns mentioned above the purpose of the rotating element io to fling
the liquid reflux against the column wall by centrifugal action, to stir the vapour 111 a
horizontal sense, and in some designs to scrape the liquid on the wall; all these factor\
increase the rate of transfer of matter and heat but they tend to produce a relativel?
high pressure drop. A different principle was ainitd at by Benner et al. [69]. The\
heated the outer coluinn and placed a cylindrical, rotating condenser inside it. The
resiilting separating effect is then produced by a continuous process of partial con-
densation and evaporation along the length of the coluinn. Shortly afterward6 .
Byron ct al. [70] described a similar column (Fig. 283) and developed the theory of
“thermal rectification” on which it is based (cf. the brush still, section 5.4.3).
Rotating concentric-tube colunins were evolved from stationary concentric-tuhc
apparatus (section 7.3.1). Like the latter they consist of two cylinders separated 1)) a
gap of 1-2 nnn. After Willinghzni et a1 [71] had published preliminary work, it v a r
principally Jost [72] who worked on the further development of these column^
(Fig. 286). HP formulated the hypothesis that the separating effect is not so much
deterniined by the turbulence of the vapour as by the regularity with which the film
clf liquid flows down. He therefore evolved rotors to which brushes were attached,
with the object of obtaining a better distribution of the liquid ([73], Fig. 201 j. Kuhn
[24] and Jost [73] showed that the highest separating effect is produced
1 . by homogeneous wetting and a uniform descending filni;
2 . by IL narrow gap, causing a short time of exchange with the vapour.
Further requirenients for the proper functioning of all columns with rotating
elements are a conipletely adiabatic operation and a constant speed of revolution of
the rotor. There are, however, wide differences in the values that have been rel~orted
for the optimum speed of rotation and in the dependence of the efficiency on thi3
speed. As will be seen from Table 62 the speeds of revolution employed vary between
250 and 6000 per minute. According to Jost [73] the number of theoretical stages is
dependent on the throughput, on the shape of the rotor and on its rate of revohtion.
I n Table 63 the optimum speeds of rotation at various throughputs have been taliu-
lated. Figs. 287 to 289 show the same data diagrammatically.
iranchembook.ir/edu

Table 62
Data of columns with rotatsingelements (rf. d m Table 37)

Ref. Type Column Column Theor. HETP Atu 11 r.p.m. Hold-up. N0bs
chap. 7 height diameter stages (om) load pressure
(cm) (mm) of drop

cylinder 43 1.2 (190 ml/h 2 tion 0.17 ml/theor. stage, gap width 1 inm
pressure drop: 0.5 torr
for 43 theor. RtageR
cylinder 12 1.7 gap width 1 -2nim
50 0.9 gap width 1 mm
cylinder 5n.4 74.1 62 0.935 17.5 ml, preswurc gnp width 1.09 mm
drop: 2.0 torr
rotating 80 in 4.45 pressure drop:
oondenwr 0.2 -0.5 torr
rotating 100 75 38.8 2.56 210ml/h 5 rotor length 1 m,
condeneer diameter 26 mm
heat ratio 20:l
conical 1070 .W) 2 2601 821 rotating
segment8 1500 elements
conical 1.28 80 ml/h 'MI 125 rotor elements
segments reflux 1700 124 sttitor elements
conical 0.93 280 ml/h
segment8 reflux
conical 4.51 200 to 320 175 pairs of conicnl
segments 240 segments
iranchembook.ir/edu

360
or
$50
d

.-240
+.’
g30
al
L
.r 20

10

a)
0

Fig. 287a)
I
1000 2000
I

r.p.m. -
3000 4000 51 D
b)
0

Fig. 287 b)
500
r.p.m. -
1000 1500 2000

Relation between the specific stage number n, and the speed of Relation between the number of stages and the speed of
rotation, with the load 8s parameter, for rotating cylinder revolution for columns with a star-shaped spinning
columns [71] band [75]
iranchembook.ir/edu

380 7. Constructional materials and apparatm

In the case of .Jost’s rotating-cylinder column the number of stages drops sharply
if the speed of rotation is increased beyond the optimum value (Figs. 288 and 289).
On the other hand Willingham et al. [71] found B continuous increase in the stage
nrrniher with the speed of revolution (Fig. 282a). In the USA a rotating cylinder

Table 63
Optimum speeds of revolution for columns wlth rotating elements

Column type Type of rotor Throughput Range Optimum References

examined speed
(ml/h) (r.p.m.) (r.p.m.) (chapter 7)

(‘ylinder with brushes 3 b 690 400-2600 2000

3b 1330 400-2600 2000
4r 6W 800-320 2600 [73]*
4r 950 400 -3 000 2 600
4c 1170 400-2400 ?
4c 1330 400- 1600 ?
Cyl Iode r smooth gap 1500 0-4OOO 2600 [71]**
1.09 mm 2000 0-400 2600
3000 0-4OOO 2600
4O W 0-4000 2600
Spinning band star-shaped 150 0-2000 1400 [7.5]***
* Efficiency drops rapidly above optimum speed of revolution.
** Efficiency remains constant from 1400 to 2000 r.p.m.
*** Above 2600 r.p.m. efficiency continues to increase, but the curve does not
rise as steeply as between 2200 aiid 2600 r.p.m.

0
kLLH rotor 4 c

800 1600
3200
r p . m -+
Fig. 288 Fig. 289
Dependence of the number of Dependence of the number of
stages on the load and the speed stages on the loadand the speed
of revolution for Joat’a rotating- of rotation for Jost’s rotating-
cylinder column with rotor 4 b cylinder column with rotor 4c
iranchembook.ir/edu

7.3 Columns 38 1

column having 500 theoretical stages a t 8000 r.p.m. and a throughput of 4 1/11 IS
said to have been constructed. The optimum speed of revolution is thus dependent
on the size and construction of the column and especially on that of the rotor
(Table 63). Gelperin and Khatsenko [74], for instance, found that in a column having
conical segments the stage nuniber remained unchanged above a speed of 320 r.p.ui.
This seems, indeed, to be about the optimum speed for many design$, since Huffniann
and Urey state the speed had little or no influence on the efficiency in the range of
250-1 500 r.p.ni.

Table 64
Pressure drop in columns with rotating elements
~~

Column Operating Pressure Pressure drop References

pressure drop per cm per t,heoretic. (chapter 7)
Type Length Diameter of column stage
length
(cm) (mm) (mmHg) (mmHg) (mmHg)

Spinning 545 6.7 1 1.3 x 10-3

band 82 20.0 1 2.0 x 10-3 34-710 x 10-3
145 i5.0 0.5 2.0 x 10-3 4.3 x 10-3 [75]
Rotating 11.5 x
cylinder
Rotating 80 0.005-1.5 4.4 x 10-3 20 x 10-3 [691
condenser

The optimum speeds reported for spinning band columns vary between 1000 and
3500 r , p . n ; the shape of the strip here seems to exert an appreciable influence a.:
well as the width. Murray [75], for example, found a n optimum speed of 1400 r.p.in.
with a star shaped band a t a load of 150 ml/h; the efficiency remained constant if the
speed was increased beyond this figure (Fig. 287b).
The time to establish equilibrium when starting up a spinning band column i.
approximately 'iZ-1 h ; for a rotating-cylinder column Jest, [73] also found aboiit
1 hour. Irlin and Bruns [76] go so far as to state that one of these columns, with an
HETP of 1 cni, requires an equilibration time of only 6 min. The pressure drops
caused by columns with rotating elements are listed in Table 64. For coniparison it
can be mentioned that a packed column will give a pressure drop in the range of
0.03-1.3 mm Hg per theoretical stage, or 0.005-0.2 mm per cm of column length,
depending on the size and shape of the packing, the load and the pressure.
This table shows that columns with rotating elements are clearly superior as
regards pressure drop, since the values for packed columns (and even more those for
plate columns) may lie several orders of magnitude higher.
The other characteristics of rotating colunins are comparable to those of con-
centric-tube and multiple-tube columns, and of columns with wire-gauze or wire helix
elements (sections 7.3.1 and 7.3.4).
iranchembook.ir/edu

382 7. Constructional materials and apparatus

From theae figures it. is seen that the rofating-cylinder colunin has the most
favourable properties for laboratory use, though the high speeds of rotation that are
necessary constitute a disadvantage (see Fig. 287a). In operations at reduced
pressures the additional difficulty of constructing a satisfactory stuffing-box for
the mpidly rotating journal also arises. Attempts have therefore been made to
employ an electromagnetic drive.
Hoher [77] investigated the dependence of the efficiency and the hold-up on the
speed and direction of rotation for a column with a rotating spiral.
The applicability of a rotating column of the Kirschbaum-Stora type with
600 mm diameter which has 8 interchangeable components to semi-technical work
was examined by Neumann [78].
Mamin et al. [78a] have reported work on a column for large-scale operation at
reduced pressures.

7.4 Condensers and dephlegmators

In laboratory and pilot plant distillation cooling is practised for the following
purposes :
I . the total condensation of vapours (condensers);
2. the partial condensation of vapours (dephlegmators);
3. the cooling of liquids (coolers).
I n these operations the cooling agent - usually water - takes up and removes
heat. The apparatus in all such cases can therefore be regarded as a heat exchanger,
qince it can obviously also be used to heat the liquid. In accordance with industrial
practice this method is sometimes adopted in continuous laboratory distillations for
heating the feedstock. If the freezing point of the distillate is higher than the tempe-
rature of the cooling water it is necessary to regulate the coolant temperature thermo-
statically so that no solidification occurs in the condenser. The term dephlegmator
is used when the apparatus, with a limited supply of water, condenses only a portion
of the vapour; this part is generally returned as reflux to the coltirun, whilst the
remaining vapour is hquefied in a subsequent condenser (cf. section 5.2.3).
The more important types of condensers have been fixed a8 to their designs and
dimensions by standardization, aa is shown in Table 65.
A dlstinction can be made between inclined and vertical Condensers (Figs. 238-5
and 238-7, respectively); in the latter the vapour can be made to enter either at the
upper or a t the lower end. If vapour enters at the upper end, cooling is very thorough,
since the substance passes over the whole surface, and in certain cases the intensity
of cooling may even be too great. This arrangement is, however, advantageous in the
separation of water from organic liquid$. The drops of water do not then remain in
the condenser, as often occura when the v a p u r enters at the lower end, but are washed
down 1))- the distillate. Vertical condensers with entry at the bottom have the ad-
vantage that suhcooling is prevented and that uncondensahlegases can escape with-
out dissolving in the distillate. The cooling water should be made to flow in counter-
currrwt to the vapour if the conptruction of the condenser allows it.
iranchembook.ir/edu

7.4 Condensers and dephlegmators 353

According to the form of the cooling surface we can distinguish between condcn-
sers with a smooth tube, augmented surface. or coils.
1. Condensers with a smooth tube
a) Air condensers (Fig. 338-16 and 3 ) for distillates having high boiling and/or
setting points.
h ) Condensers with a wide cooling jacket (Liebig; Pig. 238-5).
c ) Condensers with a narrow cooling jacket (FTest ; Fig. 291).

Table 65
St,andardized condensers for distillation apparatus

Shndard Component

TGL 40-344 Sheet 7 condensers, compound consensers

8339 Sheet 1 condensers, technological terms of delivery
Sheet 2 Liebig condenser
Sheet 3 spherical condenser
Sheet 4 Dimroth condenser
All the above condensers are available with or without
standard ground joints
DIN 12575/53 Liebig condenser, tube and cooling jacket ftised together
12576156 Liebig condenser with standard ground joint
12 580/53 spherical condenser
12581/56 spherical condenser with standard ground joint
12 585157 coiled condenser
12 586155 coiled condenser with standard ground joints
12 59 1/54 Dimroth condenser with standard ground joints

An efficient .air condenser suited to the distillation of substances boiling a t least

25 deg. C above room temperature has been described by Toeldte [79] (Fig. 290).
The vapour ascends in tube e and is cooled by the surfaces a and b. As the result of
being heated the air flows up the central tube b h; convection and constantlx
draws in cold air. The liquid seal a t the top is filled one-fifth full with a high-boiling
liquid; the prewure is equalized through the holes q. In all condensers having the
cooling surface and the jacket in the form of parallel tubeP the greatest rate of heat
transfer is obtained if the annular space between the two is kept as small as possible,
as is done in West’s condenser (Fig. 291), since the degree of turbulence is thengreatest
for a given flow of coolant. This type of condenser is also available with cooled socket
joint (Fig. 291).

2. Condensers with augmented pamllel cooling surfaces

In these condensers the cooling surface is enlarged hp the provision of indentn-
tions, undulations, bulbs or a screw structure.
Examples are :
a ) West’s condenser with indentations (Fig. 291) ;
iranchembook.ir/edu

384 7. Constrncional materials and apvarat,us

c=

Fig. 290 Fig. 291 Fig. 292

Air condenser (Toeldte) West’s condenser provided Condenser with undulai
with indentations, narrow cooling surface
jacket and cooled ground
joint (socket)

Fig. 293 Fig. 294

Staedeler’s condenser with Compound condenser:
veeeel for solid cooling Liebig and Dimroth
U agent condenser combined
iranchembook.ir/edu

7.4 Condensers and dephlegmators 385

b) Schott’s condenser with undulating cooling surface (Fig. 293) ;

c) Allihn’s condenser wit,h cooling surface formed by a series of bulbs (Fig. 238 -1.5):
d ) Friedrichs’ screw-type condenser (Fig. 238 -14).

3. Condenseis with coils

a) Condenser with coiled distillate tube (Fig. 238-6);
b) Dimroth condenser with coil for cooling medium (Fig. 238 -7) :
c) Staedelerk condenser, provided with a vessel for a cooling agent (Fig. 293) :
d) Compound condensers with niultiple cooling surfaces.
A coil-shaped condenser is particularly suitable for dealing with vapour at
atmospheric pressure. I n vacuum distillation this type causes an appreciable pressure
drop owing to its narrow cross-section; it is then better t o employ the Dimroth form,
which furthermore has a high coefficient of heat transfer. The Staedeler condenser,

Fig. 295 Fig. 296

Hen’kel condenser and still head Dephlegmator with variable
with condenser cooling surface [80]

which can be filled with ice or CO, snow, is used for condensing very low-boiling
substances. For the same purpose one niay employ the various forms of compound
condensers with niultiple cooling surfaces. These condensers consist of several cooling
elements placed inside one another, for instance straight tubes inside undulating
tubes (Fig. 292), or a coil inside a Liebig condenser (Fig. 294).
The condenser of Henkel [81] has a remarkably large cooling surface as related to
its height. It consists of a Liebig or West condenser wound into a coil (Fig. 295) and
is a typical conipound condenser with a design suitable from the point of view OI
fluid dynamics and for vacuum work. Having a height of 20 cin it provides 500 crn2
of cooling surface (cf. Table 66).
When a condenser with a cooled distillate tube is used as dephleginator the
problem is to exactly set the desiredreflux ratio. The cooling surface of the dephleg-
mator shown in Fig. 296 can be adjusted as desired by operating the knurled nut,
thus varying the position of the still pot [80].
With condensers made of glass one can count on a mininium heat transfer coeffi-
25 Krell, Handbook
iranchembook.ir/edu

386
~ ~~~~~~~~~~
7. Constructional materials and apparatus
~

Table 66
Characteristics of condensers (Eichhorn [86])

Type of Length Condens- Length External Internal Wall Heat transfer Rate of heat
condenser ing of jacket diameter diameter thickness coefficient k transfer E
area of jacket of tube of tube ( e e l . emax (cal.min-lx
(CIU) (cm2) (cm) (cm) (em) (mm) min-l.deg.C-') deg. C-I)

Liebig 20 80 19.5 2.5 1.3 0.9 1.08 86.4

40 159 39 2.5 1.3 0.9 0.99 167
TO 285 70 3.4 1.3 1.2 0.698 199
100 426 100 3.4 1.35 1.2 0.652 277
Gpherical 20 124.5 21 3.6 1.3 1.2 1.05 131
30 227 32 3.95 1.25 1.2 1.02 232
40 288 40 4.0 1.3 1.2 0.725 209
Coiled 20 120 20 3.9 0.75 0.9 1.24 149
30 200 30 3.9 0.75 0.9 0.87 174
Dimrorh coil 20 180 24 3.5 0.7 0.8 1.5 270
30 310 35 3.5 0.7 0.8 1.42 440
Stnrdrlrr 295 25 9 0.75 1.o 0.945 278
Compound 20 220 20 4.05 2.8/1.8 1.0 0.995 199
Soxhlet 200 12 12 7/33 1.0 0.87 174
spherical
Schott 1137 41 5.1 1.2 0.575 654
undulating
Fr?edrichA 20.5 17 4.5 1.25 256
6ClY.W

Bacbmann 435 25 4.5 0.35 0.7 0.533 232

double coil

cient k of 0.5cal/cni2 min. "C, and the required cooling surface area may be esti-
mated bv tlhe use of the formula

in which 0 = the heat transferred in unit time (cal/min);

F = the area of the cooling surface (cmZ);
A6, = the average difference in temperature between vapour and cooling
mediiini.
Strictly, J6, should be calculated by the following formula for the logarithmic
mean if the condensation is isothernial:

(199)

where 9 = the saturat,ion temperature of the vapour ("C);

81= the temperature of the cooling water on entering;
BZ = the final temperature of the cooling water.
iranchembook.ir/edu

7.5 Adapters; still and column heads 387

Extensive measurements on the efficiency of condensers have been perforiried by

Friedriche [83], Friedrichs and v. Kruska [84] and Mach and Herrrnann [85].Accord-
ing to recent investigations of Eichhorn [86] the heat transfer coefficients of glass
condensers with a length of 200-1000 mm are from 0.5 to 1.5 cal/cn? Illin. T. It is
obvious that the heat transfer coefficient decreases as the condenser length increases
so that it seems reasonable to have two small condensers instead of a large one. These
.diould he supplied with cooling water separately since the drop of the heat transfer
coefficient is due to the cooling water being more strongly heated with longer conden-
sers.
The Dirnroth condenser and the Friedrichs screw condenser, which give k values
of 1.5 and 1.25 respectively, can be considered to have a good efficiency. The effec-
tiveness of the Jena condenser with undulating tuhes is ti result of its very large cooling
area (1 186 cniz));its heat transfer coefficient, 0.575 cal/crnZ min. "C , is only moderate.

Table 67
Heat transfer coefficients for industrial glasses

Heat exchange k (kcill . m-2 . h-' . deg + C-1)

Condensation of vapours in condensers 250- 300

Cooling of liquids 150
Cooling of gases 50

Rrexina [82] dealt thoroughly with Liebig condenser designs and efficiencies. He
derived a new matheniatical relation for the examination of the efficiency of this
type of condenser and suggested a measuring device for condenser tests.
By using formula (198)and the values of k listed in Table 66 the reader can carry
out calculations for individual types of condenser; it is, however, always advisable to
allow a niargin of 50% above the area so found. Purchased condensers as a rule have
very liberal dimensions.
However, heat exchangers for semi-technical and pilot plants have t o be calcu-
lated exactly, the more so because the cooling surfaces are assembled from standardized
cotnponents. For this purpose, the manufacturers of glass cornponents give the heat,
transfer coefficients k as listed in Table 67.
For the assembly of larger cooling surfaces a variety of corr~poncnts,niostly in the
form of coils with nominal widths from 40 to 450 mm. are commwcially available.

7.5 Adapters; still and column heads

The vapours evolved in a boiler are led away, their ternyeratnrc is measured and
they are finally condensed. I n siniple distillation the auxiliary components used for
this purpose are adapters and still heads. I n countercurrent distillation it is also
necessary to return reflux to the column and to nieasure the reflux ratio. Special
iranchembook.ir/edu

a88 7. &netrnOtianal m8terbb and 8pparetnS

oolnnrn heads have been designed to perform these tasks. Of the numerous forms in
which thew componentsexist only a mtrictednumber can be mentioned, so chosen as
to illncctrate the main differences.

7.5.1 Adapters
Ahptera form the connection between the condenser and the receiver, and are
uBu&uy provided with a vacuum connection (Figs. 238-17 and 10). The vacuum
a d ~ p t e rof Anschiitz-Thiele (Fig.238-8) has given satisfaction and has been pro-
posed for stan&&tion in Germany. This adapter is also very useful for work at
atmospheric pressure, as it allows any number of fractions to be taken (cf. chap. 7.1).

7.5.2 Still heads

In simple distillation still heads form the connecting link between the still pot and
the oondeneer. As a rule they have provision for a thermometer. Fig. 238-1 and 13
ahow common types. Thoae having a ground joint for the thermometer are to be
preferred, as they ensure that the thermometer bulb will always be in the same
poaition in the vapour tube, a fact of importance in comparative distillations.
A number of stin heads have been standardized. Table08 lists the respective
specification sheets. Some examples are shown in Fig. 297.
The Claisen still head haa two tubes with B 14 sockets, of which t'hat on the right
is used for the thermometer, that on the left for a b o i l i i capillary, a stirrer or a
dropping funnel (Fig. 236). This head has proved satisfactory for distillations both
at normal and reduced pressures. The right-hand tube is sonietimes lengthened so as
to form a Vigreux column or a short column that may be packed with rings or spirals
(Fig.298). The oolumn functions principally as a spray trap, aa does the pear-shaped
head (Fig. 239a) and a bulb-shaped head filled with rings (Fig. 299). Special still
heade have been developed for dealing with foaming liquids (Fig. 239b). For steam
distillation Prahl's still head, which has a steam injection tube (Fig. 300), may be

Table 68
Standardized still heads

Standard Component

TGL 9974 still head with one lateral tube

9975 still head with one parallel tube
9977 bridge-type head with ground joint for thermometer
9 979 Claieen still head with lateral ground joint
9980 still head with fused-on Liebig condenser
9081 still head with fused-on Dimroth condenser
13840 still head after Prahl
40-363 still head for steam distillation
DIN 12694/64 still heeds with standard ground joints
iranchembook.ir/edu

7.5 Adapters; still and column heads 389

used. Some still heads having fused-on vertical or inclined condensers (Figs. 240 and
301, respectively) are still manufactured. Those having inclined condensers require a
good deal of lateral space, so that vertical condensers are frequently preferred. With
a bridge as still head (Sig. 238-13 and 297c) and a coil or screw-type condenser the
apparatus becomes appreciably shorter and more convenient.
The still head of Henliel (Fig. 295) takes up only little space. It is manufactured
in several versions [81]. Further possibilities have been described in detail by Gem-
meker and Stage [87].

sockets B sockets 8
after TGL 9967 after TGL9967

cone B cone B
3' after TGL 9Y67 after TGL 9967
b! U

x socket A N S l L . 5 / 2 3

TGL 9967

canes B

Fig. 297 Pig. 298

Still heads Claisen still head with short
Vigreux column
a ) with one lateral tube
b) with one parallel tube
c) bridge-type head with ground joint for
thermometer

Fig. 299 Fig. 300 Fig. 301

Bulb head filled with Still head for steam Still head with fused-on
Raschig rings distillation (Prahl) condenser, vacuum connec-
tion and joint for receiver
iranchembook.ir/edu

390 7. Conatrnctionai materials and apparatus

For semi-technical and pilot plants the still heads are usually made up of the same
componente as those required for the assembly of lengths of piping (cf. chap. 7.2).
Gases and liquids are introduced into the colunins by means of feed l i e s or tubes
provided with an annular rose (-8. 138 and 143).

7.5.3 Column heads

Column heads form the link between the column and the condenser. They are
used in countercurrent distillation and are therefore provided with devides for
regulating and measuring the reflux ratio.
According to the way in which the reflux is produced we can distinguish between
column heads for partial and total condensation [88].A column head for partial con-
demation has beendescribed in apreviouspart of the book (Fig. 171). In low-tempera-
ture distillation dephlegmators are commonly used since the product is often with-
drawn in the gaseous state. The column heads are designed to meet this requirement
(cf. chap. 5.3.1). AA was stated in section 5.2.3, total condensation is t8hesoundest
method and it is the one most used in laboratory disti{lation, especially if large reflux
ratios have to be maintained.
According to Thormann [191, Gemmeker and Stage [87]and Schneider and Schmidt
[89] the requirements that a column head should satisfy may be surnniarized as
follows:
1. The reflux ratio should he easily adjusted and measured.
2. The hold-up of the column head should he small.
3. The construction should he simple and sturdy; the device for distributing the
reflux should not be subject to leakage or obstruction.
4. It must be possible to measure the vapour temperature with accuracy.
5. The reflux should flow back into the column at (or only slightly below) its boiling
point,.
6. The column head should he suitable for use both at atmospheric and at a reduced
pressure.
It should be possible to measure the coluiiin load at any moment. The iiieasure-
mcnts should still be reliable at loads above 500 nil/li, when the liquid 9s a rule no
longer flows in drops but as a continuous streani. A particularly important point is
the accuracy of the temperature readings; toenslire this, the head should he so con-
stnicted that liquid from the condenser cannot reach the thermometer bulb and that
the pressure at the point of temperature measurenient is the same as that at which
the pressure is read. A flowmeter iuay be used to check the rate of the cooling water
(section 8.6), as excessive suh-cooling leads to a “false” reflux.
Fundanientd considerations concerning the distribution of miall amounts of
liquid in columns for the transfer of matter have been put forward by Kloss [89a].
The methodk adopted by designers of column heads in attempting to attain accu-
rate proportioning and uieasurement of the reflux are the following.
1. The production (and counting) of drops falling from points, from tubes having
having oblique ends or froin capillary tubes.
iranchembook.ir/edu

7.5 Adapters; still and column heads 39 1

2. The separate production of liquid reflux and distillate by condensation on two

condenser surfaces in parallel.
3. Regulation of the amounts of the reflux and the distillate in the required ratio by
causing these two portions to flow through capillary tubes of different lengths or
diameters.
4. Automatic reflux control by mechanical nieans or by electronic time switches.
The dropproduction method is that, which has been most used for coluinn heads.
Schneider and Schmid [8Y]have systematically investigated drop-counters of various
shapes (Fig. 302), and (using gasoline as liquid) have arrived at the results shown in
Fig. 305. The latter figure clearly indicates that some fornis of drop-counters give
a considerable variation of the drop size with the load.

symbol
diameter]
+-+I

-2 3.3 sharp-edged
I
-3 3.3 rounded
I
-4

x- 5i
-.-.95,
, 4 5 2
----u5, 7.5

a) b)
Fig. 302
a ) Shapes of drop counters examined by Srhneider and Schmid
h ) multiple tip for loads greater t h a n 500 ml/h

60

t
-E 40
50

ta
Lo

30
U

0 100 20 0 300 d/h 400

llquld load - +
Fig. 303
Relation between the number of drops and the liquid load for drop counters
of various shapes (liquid: gasoline)
iranchembook.ir/edu

392 7. Constructional materials and apparatus

In the author’s Destinorm column head (WP DDR Nr. 8234, DB-Patent-
Nr. 1011177) capillary tubes are used rn drop-counters, whilst the dietillate and
refluxcan also Be measured volumetrically. The use of a capillary tube that h m been
ground flat. and polished is in accordance with accepted stalagmoruetric practice and
the short length of the capillary eneures that the hydrostatic pressure remains s m d
(Figs. 128, 162).

n
U
Fig. 304 Fig. 305
Colnmn head for apparatus working Column head of Rehn and Theilig
at atmosphericand reduced pressures
I = t,aP, 2 = a~pcoc.

Since most organic liquids of technical iniportance have a surface tension in air of
20-40 dynes/cul at a temperatiire just below the boiling point, these capillary drop-
counters mill give drops of about 4-6 x ml, i.e. there will be 20 to 25 drop8 per
The higher the surface tension, the larger the drops formed.
1111.
Other examples of devices eniploying the drop-counting method are the Jeria
coluriin head and the Normag head 1911. In the redesigned Jena head (Fig. 304) the
vapour enters the condenser in the lower part, which may be coupled with additional
condenser units. The tap for distillate take-off has a PTFE plug which is prorided
with a conical slit. The plug is graduated and the rim of the tap bears a mark so that
a reproducible and precise setting of the reflux ratio is ensured. The Normag column
head due to Rehn-Theilig (Fig. (305) is based on a well-known form of construction -
, useful feature of this head is the provieion of two taps
the collecting collar or weir. 4
in the distillate tube. The one on the left serves to control the reflux,the one on the
right is for complete closiire. The regulating tap is usually of the fine-adjusting type,
iranchembook.ir/edu

7.6 Adapters; still and column heads 393

being provided for this purpose with conical slits or multiple channels [92]. For a
more accurate control, Scbneider and Schmid [89] in their column head (Fig. 306)
employ a ground-in glass needle valve. The column head shown in Fig. 305 ma1 also
be provided with needle valves the development of which has been described by
Kramer [go]. Further variants are Teflon spindles (Fig. 247) and bellow-type valves
(Fig. 248). The measurement of the reflux ratio is carried out with drop-counters
ground off to an angle of 60” (Fig. 302, 303).

Fig. 306
Column head of Schneider and Schuiid
with glass valve

-4 coluinn head due to Hiibner [93] incorporates an ingenious combination o,

condenser and distillate cooler. Sliding the condenser in the valve cone (Fig. 307)
which is cooled simultaneously, controls the amount drawn off. The slope of the con-
denser results in a sinall total height of the apparatus and a low dead volume. The
device used for changing the receiver also appears to be somewhat brtter than that of
Anschiitz-Thiele.
The principle of the “Corad” (constant ratio) head [04](Fig. 308) is as follows.
The reflux condenser has a t its lower end a number of drop-forniing tips a of various
widths and can be rotated as a whole in its ground joint, so as to bring any one of
these tips above the sniall collecting funnel b. I n this manner different amounts of
distillate, corresponding to certain reflux ratios, can be taken off, though accurate
control of reflux ratio depends on the absence of channelling on the condenser sur-
face. The removal of distillate can be interrupted by turning the condenser to a posi-
tion in which the liquid misses the funnel.
The automatic colunzn heads are based on a time cycle; bx electronic or mechanical
means (cf. section 8.4) the device is alternately switched for a definite period into each
of two positions ; the periods in question determine the reflux ratio. This requires
iranchembook.ir/edu

394 7. Comtructional materials and apparatus

that the evaporation rate ie a h kept constant by appropriate means (cf. chap. 8.4).
Two tspes of automatic c o l r m head can be diatinguished. In the first, the vapour is
chided into two parts, which are separately condensed as reflux and distillate. In the
second, the vapour is first completely condensed and the liquid is then divided in the
proper proportion.

s
Fie. 307
Hiibner’s column head with combined condenser, dist,illatecooler and valve cone

Pig. 308
‘Torad’’ head of Lloyd and Hornbacher with
separate cooling surfarea for reflor and distillnte
(I drop-forming tip, b = collecting funnel

U
iranchembook.ir/edu

7.5 Adapters; still and column heads 305

As opposed to all other column heads the automatic forms have the advantage
that they do not usually contain taps, so that there can be no contamination with
grease. Furthermore they allow the reflux ratio to be set rapidly and reliably at any
desired value. The hold-up is very small. Large reflux ratios, froin about 30: 1 up to
100: 1, may be adjusted accurately, whilst with valves (even of the fine control t,vpr)
the adjustment becomes extremely difficult above a ratio of 30: 1.

Fig. 309
Automatic vapour dividing head (Collins
and Lantz)
a = ground-in thermomet,er, b = reflux
condenser, c = solenoid, d = valve tube,
e = valve, f = connection to distillat,e
condenser

Fig. 309 (cf. also Fig. 257b) shows the vapour dividing head of Collins and Lantz
[95]. The ascending vapour flows past the thermometer a to the reflux condenwr b.
A solenoid c, actuated by a time relay, a t definite intervalspulls up the rodd connect-
ed to the small ball valve e , which is ground into seats a t the top and bottom, so that
a certain fraction of the vapour flows through tube f into the distillate condenser.
This method is useful for azeotropes demixing in the liquid state. By dividing the
homogeneous vapour phase the compositions of the azeotropic mixture remain un-
changed. The principle of vapour division is employed in the two Norniag azeotropic
column heads (Fig. 232a and b).
Eqeriments performed by Collins and Lantz have shown that vapour division
gives a smaller reflux ratio than liquid division when the time,ratios are the sanic.
The former method requires that, no “false” reflux shall be formed in the column head:
with this object it is carefully insulated by means of a vacuum jacket. On this account
it is desirable in very accurate work to perform preliniinary experiments to determine
the corrections that must be applied to the calculated reflux ratio.
In Stage’s [97] column head for vapour division which is provided with two
magnetic valves the reflux ratio corresponds exactly to the pre-set times the valves
are alternately open and closed (Fig.310).
The method of liquid division is most usually followed. Automatic equipment
utilizing this principle has been designed to operate with cone-and ball-valves, with
iranchembook.ir/edu

396 7. Constructional maCrbla and spperatua

swinging rods or with Swinging funnels, all actuated electromagneticdly. Fig.311

gives an example of a system employing valves [MI. The aacending vapour is lique-
fied in condenser a. When ball-valve b is closed the condensate returns to the column.
-4solenoid c raises the valve-rod d at intervals in accordance wit,h the reflux ratio and
the liquid then flows past valve b to the receiver. The two valves e serve instead of

U
Fig. 310 Fig. 311
Stage's column head [97] for vapour Column head of Kieeelbach for
division with two magnetic valves liquid division
and siphon a = condenser, b = ball-valve, c =
solenoid, d valve-rod, e = valve
--1

w c k s to aloee the receiver and are stated to function properly a t pressures down to
1 0 i n i n Hg. Stage [97] uses single and double valves as well aa sliding disks.
'rhe spinning hand coliirun of Abegg (Fig. 282) emplop the principle of the swing-
ing rod. The condensate produced on the surface of the vertical condenser flows down
from itR end along a short rod, which is pivoted from i t and contains a soft iron core.
Sornidly the liquid flows back into the column, but when attracted by a n electro-
inagnet the rod is deflectd end the condensate falls into a funnel leading t o the
receiver. -4 more satisfactory @?tern, capable of dealing with higher loads, is obtained
iranchembook.ir/edu

7.5 Adapters; still and column heads 397

I Jthe~ use of a swinging funnel; the author’s automatic column head (Fig. 312) is of
this type. It, was designed in such a way that the vapour flow by-passes the control
mechanism; the condenser is mounted a t a low level so that space CL can contain 110
vapour that might interfere with the mechanism. The swinging funnel c can be
detached from a short ground joint (not shown in the figure), whilst the connecting
rod and iron core can be removed through joint e. Any defects can thus be rectified
easily. The jacket g connected to funnel f allows low-boiling substances to he cooled
and distillates of high melting point to be warmed. The column head regulates the
reflux ratio with an accuracy of &2%.

Fig. 312
Krell’s automatic Destinorm
column head with swinging
funnel

Another column head makes use of mechanical regulation by means of a slowly

moving shaft. The pivoted rod which is moved to and fro by a periodically screened
jet of air delivers the distillate alternately to the column and the take-off [98].
Column heads of semi-technical plants are made up of separate condenser and
reflux units. The valves are operated mech~nicallyand, more commonly. electro-
magnetically. Their operation is similar t o that in laboratory column heads (Figs. 306,
247). The valves are controlled automatically by means of electromagnets, electro-
motors or pneumatic drives (Fig.248).For the automatic division of liquid, however, the
swingingfunnelis chiefly used (Fig. 142)which is actuated by an electromagnet placed
outside the columnwall. With the magnet switched on thedistillate is taken off laterally.
iranchembook.ir/edu

398 7. Constructional materials snd apparatue

7.6 Still pots, receivers and fraction collectors

The components to be described next are the first and last links in the chain of
distillation apparatus. The still pot contains bhe substance to be distllled and the
receiver and the fraction collector take up the purified and fractionated distillate,
respectively.

7.6.1 Still pots

For laboratory distillation ordinary standardized, round-bottomed flasks provided
with ground joints are used. They may have short or long necks and nominal capaci-
ties lip to 5001~1.The Engler (Fig. 235) and Saybolt flasks (TGL 0-12363)are
examples of special devices designed for standardized distillation techniques. I n
addition, a distilling flask with capacities from 25 t o 1OOO ml (TGL 0-12364) has been
standardized (Fig. 313).

Fig. 313 Fig. 314

Distilling flask, TGL 0-12364 Three-necked flask with thermo-
meter well and sampling device

Long-necked flasks are used mainly for simple distillation (Fig. 313);in the ab-
sence of a colunin the long neck acts as a spray trap. Flat bottomed flasks should not
he used for vacuum distillation because of the danger of collapse. For countercurrent
distillation short, round-bottomed flasks with three necks are the most suitable. The
two side-necks, provided with B 14 ground joints, are required for the thermometer,
for filling and emptying, for a boiling capillary, for a gas or v a p u r inlet, tube and for
the removal of samples. The axes of the side-necks meet the axis of the centre neck a t
the bottom of the flask, so that there is space for these components if the column is
surrounded by a n insulating jacket. The necks should preferably be provided with
hooks for holding the ends of retaining springs over the joints, to avoid loosening
owing to sudden pressure. If it is necessary to change the thermometer during a distil-
lation, or to avoid the use of therinonieters with ground joints, a thermometer well
may be employed (Fig. 314). A little oil should be put in the well to promote heat
transfer.
The charge should be at most two-thirds of the volume of the flapk in distillation at
atniwpheric pressure and not more than half in vacuuni distillation.
iranchembook.ir/edu

7.6 Still pots, receivers and fraction collectors 399

Pig. 315
Flat-bottomed flask with ball joint

Fig. 316
Jena. suspended vessels (16 to 150 1) as still pots
iranchembook.ir/edu

400 7. Constructional materials and apparatus

9 .w n
Fig. 317
Metal boiler with welded-on oil-heated jaoket (Stage [SS])
I = joining socket for oolumn, 2 = socket for pressure measurements, 3 , 4 =
sockets for temperature meeeuremente, 5 = outlet socket, 6 = socket for oil
expansion vessel, 7a, b = joining sockets for circulating pump, 8a, b = joining
sockets for cooling water, 9 = outlet socket for oil, I0 = socket for heater, 11 =
eockets for temperature measurements, 12 = inlet socket for oil

Fig. 318
Armoured glass flask with metal-
bath heating
iranchembook.ir/edu

7.6 Still pots, receivers and fraction collectors 40 1

For micro- and semi-micro-distillation smaller flasks are frequently necessary. As

a rule these are made with pointed bases, in order that distillation may be continued
down to a small residue (Fig. 127).
Flat-bottomed flasks have proved to be favourable (Fig. 315) since the evaporat-
ing surface remains almost constant down to the residue whereas in round-bottomed
flasks the liquid surface decreases more and more as the level sinks.
For continuous distillation a satisfactory arrangement consists of a flask provided
with an overflow tube, which maintains a constant liquid level and through which
the bottom product is continuously drawn off.
I n low-temperature distillation it is usual to employ cylindrical flasks which are
generally placed inside a Dewar vessel (Figs. 173, 175, 176,180). Other forms of flask
may sometimes be necessary, depending on the method of heating (section 7.7.2).
For semi-technical plants of glass, short-necked round-bottomed flasks of
capacities up to 4 1 and three-neck round-bottomed flasks of capacities up t o 10 1
(TGL 10 102) are available. The glassware manufacturers can now offer spherical
distilling flasks and cylindrical vessels with capacities up to 200 and 375 1, respec-
tlvely. Connection pieces may be placed a t the top, at the side or at the bottom.
Cylindrical vessels are usually provided with caps which hold the connection pieces.
Fig. 316 shows suspended vessels which are available in sizes of 16 to 150 1 [l,51.
Adapters for connecting the necks to ball and conical ground joints can be purchased
for bobh types. Flasks of this size are best heated by steam, by some kind of bath. or
by electric immersion heaters or mantles.
Bor combustible or explosive substances the use of a stainless steel boiler is ad-
visable. A metal vessel developed by Stage has a nominal capacity of 10 to 200 1
(Fig. 317) and is provided with a jacket for oil-bath heating. Here again a n adapter is
available for joining the ground flange t o a glass ball joint. If the use of glassware is
essential on account of corrosion, a n armoured glass flask heated in a metal bath gives
the greatest safeguard against explosions (Fig. 318). Such flasks are made in sizes
from 1 to 20 litres [99]. The flask is surrounded by a metal jacket and the inter-
vcning space is filled with bismuth or an alloy. The metal bath ensures an even
transfer of heat; two opposed sight-glasses allow the contents of the still to be scen.

7.6.2 Receivers and fraction collectors

Practically any form of vessel may be used to collect the distillate (Fig. 238). For
measurement of volume it is preferable to employ graduated cylindrical receivers
(Fig. 238-9) which can be provided with a jacket for dealing with high-melting or
low-boiling distillates; the jacket also encloses the drain cock (Pig. 319). I n accurate
work it is always advisable t o include a distillate cooler and to jacket the receiver, so
that the amounts of distillate can be measured a t the same temperature. I n distilla-
tion a t normal pressure the receivers should be in communication with the atmosphere,
at reduced pressures they are connected to the vacuuni line. A satisfactory arrange-
ment is that shown in Fig. 320 where the vacuum connection is also cooled [loo].
By means of Bredt's rotating receiver (Figs. 338/11 and 321) and that of Briihl
(Fig. 238/12; cf. also Pigs. 127, 130) it is possible to collect 4 and 7 fractions, respec-
26 B e l l , Handbook
iranchembook.ir/edu

402 7. Constructional materials and apparatus

3
Fig. 319 Fig. 320
Receiver with jacket Receiver with jacket and cooled
vacuum connection a

Fig. 321 Fig. 322

Rredt’s rotating receirer with Stage’s receiver for vnciium
graduated tubes distil la t ion
iranchembook.ir/edu

7.6 Still pots, receivers and fraction collectors 409

tively, 111 vacuum distillation without interrupting the operation. Any number can be
collected with the Anschiitz-Thiele adapter with interchangeable receivers (Fig. 238-8,
cf. section 7.5.1). Fig. 322 shows Stage’s receiver for vacuuni distillation and Figs. 137
and 112 illustrate the arrangement used for work on a larger scale; the latter again
obviates interruption by the provision of vacuuni and air connections. The receiver
of Rock and .Jantz is also of the “pig” type and has a magnetic switch-over device
[lo11 (Fig. 323). The magnet a attached to the ring c pulls the rod d, hanging from
hook b t o the appropriate receiver. This arrangement has the advantage that the
distillate does not beconie contaminated by tap grease. For taking off gaseous
distillates in low-temperature distillation special device8 are used (Figs. 183 to 185).
Of late clufoniutic fraction collectors, as employed in chrcmatographic ‘work, have
also become of iniportance in distillation a t atniospheric pressure. These devices
generally have a rotating table, on which test-tubes are arranged in a circle or spiral.
The principles according to which they can function are the following:
1. drop counting;
2. control of the volume by a syphon or photo-electric cell :
3 . control by time.
Modern fraction collectors can mostly be eniployed selectively according to these
three principles. Pneumatic devices are also available now for work in rooiiis where
the danger of explosions is great.
In the automatic collectors based on drop counting a falling drop brings about a n
electric contact or activates a photo-electric cell. The resulting pulses are passed
through a relay to a counter, which can be so set that after any desired nuinl~erof
drops a tiiechanism releasing thp table is set in motion, whereby the latter moves on
to the position of the next receiver. A t the same time the counter automatically re-
turns to zero.
I n analytical work it is common practice to collect fractions of a definite volume
and to read the distillation teniperature a t each cut. The volume of the fractions
del~endson t h e size of the charge and generally lies betneen 1 and 201111. Whilst
fractions of 1 to 5 nil can be collected satisfactorily by the drdp-counting method,
those above 5 ml are best obtained by syphon control (Fig. 324). To a ineasiiring
vessel 1 a syphon is connected, which draws off the liquid as soon as it has reached a
certain level. The stream of liquid flows past a photo-electric cell 2-3, that passe? on
its pulse to a relay. The latter, after a delay of a few secoiids, nioves the Iotating
table on. The collecting vessel 1 in the author’s apparatus is provided with a plunger
with which its capacity can be regulated to within 0.1 nil, so that it is unnecescary
to change the vessel for samples of different volunie.
The fully xutoniatic fraction collector of Grassrnann and Deffner [lo21 functions
hv photo-elec-tric volume rtieasurement. A pencil of light measures the volunie directly
in the test-tube (Fig. 325); this volume can be regulated between 0.5 and 8 nil (with
an accuracy of one drop, 0.03 nil) by raising or lowering the light source L. The
mechanism changing the receiver i s actuated by the photo-electric cell K when it is
struck by the light reflected down (via the walls of the tube) from the meniscus M as
this passes the light source, the round bottom of the tube acting as a collecting lens.
iranchembook.ir/edu

404 7. Constructional materiala and apparatus

'E1
Fig. 323 Fig. 324
Receiver with magnetic Syphon arrangement according to Krell
control (Rock and Jantz) with volume regulation by plunger
I = collecting vessel with plunger,
2 = photo-electric cell, 3 = light source,
4 = measuring pipette

Fig. 325
Photo-electrically operated fraction
collector of Gressmann-Deffner
dl = meniscus, L = light source,
K = lens and photo-electric cell
iranchembook.ir/edu

7.7 Insulation and heating devices 405

For the measurement of small volunies in the order of 1 ml the pulse required
for operating the table can also be obtained from the current passing hetw-een two
electrodes [lo?,].The current required is extremely weak (about 0.1 nlicro-anpke)
and the method can therefore he used for most liquids except the hydrocarbons. The
error is said to be not more than 0.1%.
I n the titne-reguluting method the table is merely made to rotate one step further
a t certain intervals of time. If fractions of the same volume are to be collected this
procedure obviously entails an accurate regulation of the speed of distillation. 1 t
has proved particularly satisfactory in micro-distjllations (Fig. 257b).

7.7 Insulation and heating devices

In order to start and maintain a distillation a continuous supply of heat is required.
This goes principally to preheat the feed, to form reflux and to evaporate the fractions.
A further part of the heat, is consunied in compensating for heat losses. When the
temperature of distillation lies below 100°C, these losses can as a rule be sufficiently
reduced by good insulation only, but a t higher temperatures some form of heating
inside the insulation beconies necessary.
Thus we distinguish these heating systems:
- the heating of still pots and flasks;
- the heating of feed-stock and bottom;
- insulation.
Stage and Gemineker [ 1041have dealt with all aspect? of evaporation and heating,
Including numerous constructional details.

7.7.1 The heating of still pots and flasks

St111 pots niay he heated either directly by gas or electricity or indirectly by a
heat carrier. Bunsen burners are now as a rule used only for heating small flasks, AS
in an Engler distillation (Fig. 235) or micro-distillation. When foaming liquids have
to be distilled a Bunsen burner is also useful, as it can be manipulated in such a way
that “puking” of the flask contents is counteracted. Large flasks are now rarely
heated by gas, firstly because of the danger of superheating, secondly because an
accurate temperature control is not easy to achieve with gas heating and thirdly
hesause of fire risk. These difficulties may to some extent be avoided by placing
asbestos wire-gauze below the flask or surrounding it by an air bath [loti]. Heat
transfer then takes place more uniformly by conduction from the hot gases of com-
bustion. By the use of a chimney (Fig.326) the heat of the flame is utilized econoniically.
Electrical heating, on the other hand, is easily regulated. It can he applied in
various forms :
1. hy open and encased heating elements;
2. hy shaped heaters;
3 . by immersion heaters ;
4. by infrared radiation heaters.
iranchembook.ir/edu

406 7. Constructional msteriala and apparatus

Hd@ates with hare heating elements have the advantage of providing sensitive
mntrol, as is necessary, for instance, when the heat input is controlled by the pressure
drop over the column (cf. section 8.4); a covered hot-plate in this case has too long a
tempemture lag, which may result in flooding. It is adviRable to leave an air space
hetween t,he hot-plate and the flask and to avoid heat losses by radiation as far as
poasible (Fig. 327).
Another form of heating are the so-called ‘‘niirror cookers” or electric Bunsens.
These Lrb~~mers” depend on the concentration of the radiation emanating from an

Fig. 326
Hiibner’s air bath

Fig. 327
Deatinorm flask-heating unit
iranchembook.ir/edu

7.7 Insulation and heating devices 407

rlectric heater to a point bjr reflection from a concave mirror [lOSl. The temperature
is itdjusted by means of a regulating shield (“electro-tap”). With Hoffniann’s [ 1071
birrner temperatures up to 800°C can be reached.
Heating elements with resistance wires conforniing to the shapes of the surfaces
to he heated are known as shaped heaters (Figs. 149, 184).If the contents of the flask
are t o be stirred electruniagnetically an apparatus such as the one shown in Fig. :328
IS convenient; it combines a hot plate and a magnetic stirrer.

,
Fig. 428 Fig. 329
Hot-plate magnetic stirrer Multi-heating jacket for
temperatures up to 460” C

When conibiistible or explosive sttbsbances have to be distilled, flexible heat.ing

niantles provide additional safety; they can be obtained in individual sizes or t,he
inillti-size design can be employed for various sizes of flask. The heating elenients are
secured in a, glass fabric; the maxiitiuni teniperatare is ahout 400 “C. In the multi-size
11iantlea‘ spst’eniof circular springs gives t)heflask a secure seating, and an opening in
the base prevents the formation of air cushions and allows the heater to be used for
flasks with a bottom drain and for funnels (Fig. 329) [lOSl. The proximity of the
heating jacket to the wall of t’he flask ensures a good transfer of heat and helps t,o
prevcnt humping, especially in the case of viscous liquids.
Efficient heat transfer is also a feature of immersion heaters (cf. chap. 7.6.1). This
tl-pe of heating requires tjhe presence of heating wells. Fig. 147 shows a water still
having a heating element made of quartz. In another construction Qheflask has a well
in its base cont’aining a vertical heater (Fig. 330). Junge’s “Intus” heater (Fig. 331)
is quipped with a n iinmersion heater passing through a central opening in its top
iranchembook.ir/edu

408 7. Conatructions1 meterisle and eppmratua

and projecting down into aa extrusion in its base, so as to allow distillation to be

continued to a small residue. The flask, however, is surrounded by a container filled
with insulating material and it, is consequently difficult to watch the contents. Good
mising of the liquid is ensured by the use of a flask with cimulation heating (Fig. 332).
The well for the immersion h a t e r has a layer of glass powder fused onto its surface to
promote regular ebullition. Immersion heaters such as the THQ 6 heater of VEB
Jenaer Glaawerk Schott 8c Gen., Jena, with a rating of 6 kVA (cf. chap. 5.1.3.3) may
also be used in semi-technical and pilot plants.

Fig. 330 Fig. 331 Fig. 332

Flask with immersion heater Junge’s flask with immer- Flask for circulation
in its base sion heater heating

At the present time the methods of heating based on conduction are being
replaced to some extent by the gentler and more uniform method of radiant heating
and especially by infrcrred lamp. Among the lat.ter we can distinguish between bright
radiators, which have the shape of a large electric light bulb and are most commonly
used in the 250 Watt rating, and dark radiators in the form of a metal tube, with
ratings up to loo0 Watt. By the use of several bright radiators temperatures up to
300°C can be established. Dark radiators are easier to adapt to the surfwe to be
heated and in particular have proved valuable for micro-distillation. A distinct
advantage of infrared heating lies in the extremely small differences in temperature
t>hatoccur between the outer and inner walls of the glass vessel heated. The methods
of infrared heating have been described by Klees [110].
Indirect heating of the flask by a heat carrier should always be considered if
direct heating is not advisable for reasons of safety, or if a very even transfer of
heat is essential for the prevention of local superheating. It should, for instance, be
employed for preference if a low-boiling light fraction is to be separated from a high-
boding residue, aa a drastic increase in temperature after removal of the “tops” can
otherwise scarcely be avoided. Indirect heating can be carried out by the w e of open
liquid baths (Fig.203), or by passing a heat carrier through a coil in the flask (Fig. 333)
or through a jacket surrounding it (Fig.334). If high-pressure steam is not available
for temperat,ures above 100°C, low-pressure steam may be superheated as described
in section 6. I . Liquid heat carriers, such as glycerin, lubricating oil or triglycol, are
iranchembook.ir/edu

7.7 Insulation and heating devices 409

heated by being passed through a directly heated coil (Fig. 317) or a coil placed in a
bahh. I n glass-made pilot and semi-technical plants steam and fuel oil are chiefly used
as heat carriers. Fig. 335 shows immersion heaters with and without stirrers. The
following substances are suitable for open baths.

Substance For temperatures

~~

Water lip to 80°C

Oil up to 330°C; cf. Table 46
Salt mixtures 150--500°C; cf. Table 46
Sand any temperature
Metal alloy above 70°C; cf. Fig. 318

Fig. 333 Pig. 334

Destinorm flask provided Destinorm flask with heating
with heating coil jacket

Fig. 335
Immersion heatera for pilot and semi-technical plants
a) without stirrer b) with stirrer
iranchembook.ir/edu

410 7. Constructional materials and apparatus

It should be noted that it is very difficult to control the temperature of a sand

h t h , and that when using salt or metal baths one should remove the flask before the
b a t h has solidified. Some suitable alloys for metal baths are:
Wood's metal (M.P. 71 "C): cadmium 1-2, tin 2, bismuth 7-8 parts;
Rose's metal (M.P. 95°C): bismuth 2, lead 1, tin 1 parts.
Mrreury and alloys containing large amounts of lead should not be employed, owing
to their poisonous nature.

7.7.2 The heating of feed-stock and bottom

In continuous distillation the inconung feed-stock has to be heated to a tenipera-
tiire corresponding to the point at which it, enters the colunin (Figs. 142, 167). Tf this
tenij~ratureis not higher than lOO"C, it is often possible to use an ordinarj- coil-
t 4 - p condenser (chap. 7.4)as heat exchanger, a liquid of constant temperature being
passed through the coil. Conversely, the liquid to be heated may be run through a coil
immersed in an oil bath controlled by R thermostat. Semi-technical and pilot plants
made of g b s usually employ coils which are protected against displacement. These
devices are operated with steam (3 atm. overpressure, 147°C) or fuel oil (maximum
tmiperature, 220°C) (Fig. 336). The upper connection serves as steani inlet, the lower
as condensate outlet. The left-hand upper connection is used for tenipemtnre mee-
surements.
Fig. 138 s h o w a pre-heater with external electric heating coil and Fig. 337, the
Labodest VD 2 pre-heater which is provided with a n electric inmersion heater.
These unitR map be operated at a base load, the regulating load being controlled via
contact thermometers and relays (cf. chap 8.2.2).

NW225

Fig. 336 Fig. 337

Heating coils with protection Labodest VD 2 pre-heat,er with
against displacement immersion heater flanged to it
iranchembook.ir/edu

7.7 Insulation and heating devices 41 1

I n continiioiis operation the bottom may be heated as shown in Figs. 328 to 335.
By employing circulation heating the amount of coluinn bottoin and particnlarly the
iiiean residence time can he decisively reduced [ 1041. Some possible designs rising w r -
tical or inclined iniinersion heaters are demonstrated in Fig. 338. The Labodest
horizontal circulation heater (Fig. 339) was designed for vaciiiini work with high-

a’
Pig. 33s
B b)

C‘ircwlatiori heaters with’verticsl and inclined immersion heat,ers

a ) “21Tormschliff” model h) ‘‘Noriniig” rnorlel r ) “Schmidt,” model

Fig. 339
L;xbodest horizontal circulation heater with separate pipes for vapour and reflax (Stage)
iranchembook.ir/edu

boiling liquids. I t has aeparate pipes for the vapour and reflux. This arrangement is
particularly suitable for large-scale plants. Circulation heating is appropriate for
pilot plants as well, as can be seen from Figs, 141 and 142.

7.7.3 Insulation
The vapour ascending from the boiler should in general reach the condenser with-
out the introduction or removal of heat, other than by exchange with the reflux.
When the usual insulating materials are used this requirement can be satisfied suffi-
ciently closely up to a temperature of about 100°C only; above thia temperature,
heat must be aupplied to make good what is lost by convection and radiation t o the
aurrounding air. The still pot and all parts of the apparatus up to bhe condenser

Fig. 340
Semi-circular glaes-wool eheathe for insulation

should therefore be insulated. Proper insulation not only improves the sharpness and
reproducibility of the separation in question, but also ensures that the heat is economi-
cally utilized (cf. section 4.12).
The following means of insulation for laboratory and semi-technical apparatus
are available.
a ) Coatings of a material having a low heat conductivity;
h) Vacuum jackets;
c) A jacket containing a heat carrier in circulation [ 1113;
d) Electrical heaters in an insulatiug jacket.
In simple forms of distillation, not requiring an accurate control of the reflux ratio,
asbestos cord is still largely employed aa insulation in the laboratory. As a rule it is
applied in too thin a layer; the thickness should amount to about 50-60mui
(2 incheu). More convenient in use are lengths of pipe insulation made of glass wool,
which are easily fitted to the column and can be cut to any desired length (Fig. 340).
hround these sheaths a layer of glass-fibre tape is wound. If a loose material such as
magnesia or mineral wool is to he used, a metal jacket is required as container. The
iranchembook.ir/edu

7.7 Insulation and heating devices 41 3

efficiency of insulation is improved if the column is first coated with a reflecting sheath
of aluminium foil. Insulating jackets of the above-mentioned type are, however, good
enough only for operating temperatures not exceeding 80°C; the same applies to air
jackets.
Vacuum jackets provide satisfactory insulation up to 150°C or more if they are
evacuated to 10-4-10-6 mm Hg [112]. They are usually silvered inside t o reduce
radiation. The quality of the silver layer has a great influence on its efficiency. A strip
of 5-10 mm width is often left clear in the silvering for purposes of observation, but
it should preferably be only 2 mm wide. A much better method is to place metal

%
'OOk

20 --
---- ---9--
7------
1
-- - ---
0
100 500 1000 2000 d/h 3000
Fig. 341
Relation between load and number of theoretical stages, n t h [la11
Test column: length 1200 mm, diameter 25 mm, with silvered vacuum jacket
I = wire-gauze packing with large mesh size, DNW 1, for small pressure drop,
2 = wire-gauze packing with small mesh size, DNE 2, 3 = packing as in 2 but
with electric compensation heating jacket

cylinders or curved aluminium foil in the jacket. Perforated foil is claimed to provide
the same insulation as a completely silvered surface and has the advantage of allow-
ing all parts of the column to be seen [113]. If it is essential for some reason to have an
unobstructed view of the column contents, no other course is open than to dispense
with the silvering or foil. Then it is advisable t o keep the jacket constantly evacuated
to the high vacuum mentioned above by connecting it to a diffusion pump. Commer-
cial vacuum jackets should be checked to ensure that they are sufficiently evacuated.
The strain produced by the difference in expansion of the outer and inner walls of
a vacuum jacket is reduced if the outer tube is provided with expansion bellows
(Pigs. 257b, 282), or the inner tube with an expansion coil. If the column has dia-
meter of less than 10 mm and a length of more than 500 mm, or if the temperature
exceeds 150°C it is advisable to wind an electric heating tape around the outer wall
of the vacuum jacket. A rough estimate of whether the insulation is sufficient can be
obtained by feeling i t ; the outer surface should not be appreciably above room
temperature.
iranchembook.ir/edu

414 7. Constructional materials and apparatus

Fischer and Weyand (chap. 8, Ref. [27]) could prove that providing 8 column
with an electric heat-cornpeneating jacket in addition to a vacuum jackeb may
iinprove the efficiency considerably (Fig. 341).
-4 warning must be sounded here against the use of jackets through which a
henting niediuni, such as steam or a thermostatically controlled liquid (cf. section

Fig. 342
Heating sections in glass-fibre cloth,
applied t o a molecular still

7 . i . l ) is passed. There is necessarily a temperature gradient in any column 11121;

an isothermal jacket therefore produces either a “false reflux” or a n additional
evaporation along the column wall. Kolling and Tramm [ 11.11, when using this method,
placed a layer of insulation between the column and the oil circr~lat~ing jacket and
also insiilated the latter externally.
.A satisfactory method for insulating laboratory and pilot plant colunms is to use
electric heat-co7npemating juckets which themselves should be effectively insulated.
This t-ype of jacket is particularly necessary in vacuum distdlation, where the heat
capacity of the vapour is small. Fig. 342 shows such insulation made in the form of
iranchembook.ir/edu

7.7 Insulation and heat,ing devices 115

Fections lagged with glass fibre; these are obtainable as components for insulating
flasks, colurms and molecular stills 11151. Von Weber [ 1121 demonstrated that the
flow of heat through the column wall is independent of the position of the heating
wire within the insulation. There is thus no drawback in winding it direct,ly around
the coluinn.
h I

0 20 40 60 80 cm 100

number of turns per 10cm

Fig. 343a
-
distonce from top of tube --L
1 3 5 6 7 810111321

Fig. 343 h
Temperature gradients a t various currents Diagram of a heating jacket rircuit with
in the heating system of Kortiim and Bittel thermocouple

The fact that a temperature gradient exists in the column must of course be taken
into account in designing an electric heat-compensating jacket. For this reason tlie
Destinorm jackets, mentioned above, are made in lengths of 50 cm, each of which can
he adjusted to the corresponding interior teniperatiire of the colunin. -4satisfactory
method of doing this is to have a thermometer indicating the inner temperature of
tlie column in each section and to adjust the jacket temperature to a value 1 or 2 dry.
C below this reading. It is incorrect to regulate the jacket temperature in accordance
with that in the column head, even if the temperature gradient in the column in
not more than 1 or 2 deg. C.
Kortuni and Bittel [ 1161 have evolved an arrangement of two heating elements.!~k
means of which an arbitrary temperature gradient can be maintained along the
column. Two heating circuits wound on an iron tube are employed, the first of which
is so spaced that the distance between the turns beconies smaller toward the lower
end of the column. The temperature gradient is ineasured a t distances of 10 ciii b j
therniocouples. I t will he seen that an almost linear temperature profile is obtained
at all current strengths (Fig. 343a). The second heating element, insulated by bead?.
iranchembook.ir/edu

416 7. Constructional materiala and apperatus

is wound with constant spacing over the first, the spacing being equal to that in the
centre of the “gradient” heater. The whole is insulated externally with thick asbestos
cord and by a jacket of magnesia or kieselguhr. Two contact thermometers are
fitted in such a way that their bulbs are in direct contact with the packed column a t
distances of 10 CIU from its upper and lower ends. The upper contact thermometer
controls the second element, the current in which can be set at any value; t,he lower
contact thermometer controls the “gradient” heater. Only a fraction of the total
current is controlled.
In Fig. 343b the schematic diagram of a heating jacket regulated by means of
thermocouples [117] is shown. A power of 86 W (220 V, 0.39 A) is sufficient for
temperatures up to 200°C whereas k h e r temperatures require jackets with 100 W
(220 V, 0.46 A).

Table 69
Calibration of a heat-oompeneatingjacket

Electrical input Watts 14 42 77 114

Temperature “C 75 157 219 297
Heat input per metre cal/m- h 9300 28700 51000 76000
Condensate correeponding -
mol/m h 1.27 3.2 5.0 6.3
to heat input

Jackets may be calibrated as follows. B column of the proper length is placed

inside the jacket; thermometers are so inserted into the upper and lower ends of the
column that their bulbs are at one third of the length from the top and the bottom.
The temperatures established a t various current inputs are then observed, the average
of the two thermometer readings being taken. As an illustration of such measure-
ments, the values found by von Weber [112] for a jacket are listed in Table 69.
The figures for the heat input correspond to the amount of heat that would be
lost at that temperature if the column were operated without heat compensation.
The heat input refers to one metre of column length and the amount of condensate
has been calculated for a typical liquid obeying Trouton’s rule and boiling at the
observed temperature.
Given such a calibration, a heat-compensating jacket can be regulated to the
inner temperatures occurring in any distillation with a variable resistance and an
ammeter or wattmeter. A more accurate, automatic heat control is obtained by the
use of thermocouples or gaa thermometers (cf. section 8.2.2).
It is also necessary to insulate the parts between the still pot and the column,
hetween the individual column sections and between the column and the condenser.
If these portions are not lagged they act as dephlegmators and produce additional
reflux. It ia usually difficult to get the vapour of high-boiling substances out of the
column without adequate insulation. To prevent condensation in the vapour tube it
may be surrounded by a vacuum jacket (Figs.309, 310), but this measure is effective
only up to about 150°C. For higher temperatures the vapour tube or the vacuum
iranchembook.ir/edu

7.7 Insulation and heating devices 417

jacket can be heated by a winding of resistance tape in glass-cloth insulation (Fig. 344).
Heating tapes of this type are made in lengths of 60-250 cin (2-8 ft.) and widths of
6-90 mm (1/2-31/2in. [118]).Their ratings range from 36 to 550 W a t t ; at anaverage
surface load of 0.4 Watt/cin2 they can produce temperatures u p to 450°C. For use
a t still higher temperatures, flexible heated tubes in a sleeve of heat-resisting stainless
steel plaiting can be obtained ; they are manufactured in various lengths, for maxi-
mnni temperatures of 450"-500°C. Their dianieter is 6 mni (1/4 in.) and the snrface
load ' 4.65 Watt /cm2.

Fig. 344
Chemotherm heating tape

Heating tape manufactured by Electrothermal Engineering Ltd., London, is also

available in extended lengths on rolls. The manufacturer provides a nomogram for
d~terminingthe length needed for a given purpose, and the amount required is cut
off. Electric terminals are supplied and can be fixed without difficulty.
When heating elements are nsed on the column head care must, of course. he
taken that the vapour is not superheated, as this would vitiate the temperature
readings. A calibration as described for column jackets must therefore be carried out
first. If distillates having a high melting point are produced, they can easily be
handled by the use of heating tape on the connections to the receiver and on the
receiver itself.

27 Erell, Handbook
iranchembook.ir/edu

418 7. Constraotional materials and apparatus

Table 70
Test data for packed columns

Packing Column
Type of packing Material Diameter Height Wall Surface Column
thickness area of diameter
one litre
(mm) (mm) (-) (m') (mm)

Raschig rings Glass, smooth 3.0 3.0 0.7 1.320

Glass, smooth 4.0 4.0 0.6 1.109
Glass, smooth 4.5 4.5 0.5 1.382 24
GI-, smooth 6.0 5.0 35
G l m , smooth 6.0 6.0 20
Glass, smooth 6.5 6.5 0.5 1.109 24
Glass, smooth 10.0 10.0 60
Porcelain 5.5 5.5 1.1 24
Porcelain 8.0 8.0 1.1 24
Earthenware 10.0 10.0 1.7 24
Dixon rings Wire mesh 3.0 3.0 26
Wire mwh 5.0 5.0 20
Wire mesh 6.0 50
Prym rings Met a1 3.2 2.5
Wilson spirals Glass 3.0 1.5-3.0 4.0 24
Glass 5.0 5.0 0.65 1.66
Glass 6.3 25
FenRke helices Glass 2.1 10
Glass 3.1 13
Glass 3.1 20
Glass 6.3 25
Helices Stainless steel 1.6 1.6 0.2 40
Stainless steel 2.0 2.0 0.2 4.65 24
Stainlexs steel 4.0 4.0 0.4 1.49 24
Bead8 Glsm. smooth 3 24
Glass, smooth 4 0.9
Glass, ma tt 4 24
Glass, ma tt 7.5 24
Saddles Porcelnin 4 1.5 30
Porcelain 6 1.15 30
Porcelain 8 30
Porcelain 10 0.72 30

Sigwart: N o further data supplied.

Thormann: The lower HEWSare for smaller loads and conversely.
iranchembook.ir/edu

7.7 Insulation and heating devices 419

Height equivalent to a theoretical plate, cm

Cross- Packed Schultze Sigwart Thor- Myles

sectional height and Stage mann e t al. Other References
area Ch. 4. Ch. 4. Ch. 4. Ch. 4.
(cm2) (cm) [216] [217] {218] [192]

5-7
6-9
4.5 60 5.82-9.52
9.6 90 11.1 1-12
3.1 125 15.5 8-14
4.5 ti0 5.0- 14.3
16.9 95 18.2
4.5 60 3.69-7.5
4.5 60 7.07-11.3
4.5 60 7.23 - 12.8
4.9 100 1.4-2.8 chap. 7.8
3.1 60 4.35
19.6 100 1.4-3.5
3-5
4.5 60 8.0 -13.35
7-12
4.9 107 7-9.2
0.8 80 4.8
1.3 126 7.4 4-6
3.1 125 9.5
4.9 107 3.6-4.2
12.5 100 1.4-2.0 Author’s measurements
4.5 60 1.26-3.32
4.5 60 1.82- 5.0 2-5
4.5 60 6.00-6.06 5-6
6-7
4.5 60 5,82-7.22
4.5 60 3.31 -7.50
7.0 45 4.7-6.3 Author’s measurement,s
7.0 45 6.0-9.0 Author’s measurements
7.0 45 7.5- 11.4 Author’s measurements
7.0 45 9.0- 12.8 Author’s measurements

Myles et al. : Test mixtures are n-heptane -methylcyclohexane and n-dodecane- cyclopen-
tylcyclohexane; the HETP’s are for various pressures and a medium reflux flow.

27*
iranchembook.ir/edu

420 7. Constructional materials and apparatue

For glass piping steam-heated copper or lead tubes are used which are placed
along the glass pipes and isolated. The elastic Calorex profile tubes [119] of synthetic
rubber have a wide range of heating applications. Their base widths are 8, 15 and
30 mm. The heat transfer coefficient for Calorex on glass (wall thickness, 1.5 mm)
ranges from 166 to 260 kcal x h x ni2 x deg.

7.a Packings
Packings, p l d in the column in a random fashion (c/. section 4.10.2), are intend-
ed to provide the surface necessary to carry the thin film of liquid (section 4.2) with
which oountercurrentf exchange takes place. In a few cases a space-symmetric
arrangement may turn out to be advantageous [118a]. The effective surface of
packing is that t a b part in the exchange of material and heat. The smaller the
size of packing unit of a given shape, the larger is the area per unit of column volume.
There is, however, also a corresponding increase in the hold-up, which tends to
reduce the separating efficiency (section 4.105). As so often in distillation one must
therefore take several factors into account and make a compromise when deciding on
the shape, size and material of the packing to he chosen. Leva [120] has compiled
data and calculations concerning the choice of packing units, part8icularlyfor semi-
technical and technical columns.

7.8.1 Shape of packing units

Although bodies of various shapes have been proposed as packing, cylinders are
by far the most usual form. Of the non-cylindrical types, balls and saddles are those
most often encountered in the laboratory. Some of the other shapes that have been
Ruggested are: single-turn glass spirals, glass wool, chain links, tacks, nails, metal
bellows. Data for packings are listed in Table 70.
Kerdnyi [121] has made a systematic study of the methods of selecting highly
effective packings for particular distillations. The dependence of the separating
efficiency on shape and size of the packing units and on material and operating para-
meters is represented in a number of diagrams. Besides, a relation for characterizing
efficiency in terms of a quality factor G was derived:

where q,,= number of theoretical stages, H = liquid hold-up, A p = pressure drop,

t = throughput, ml x h-l, A and B = constahts.
A double logarithmio plot. of Q vs. load (ml x h-l) gives rectilinear dependences
according to which the most efficient packing units may be chosen. Fig. 345 exhibits
the data for new packing units developed by Ker6nyi (I to 4) as compared to those
for helices with close and wide spacing (5 and 6).
iranchembook.ir/edu

7.8 Packings 421

A study of ring-type packing units (Fig. 346a to i and m, spaced wire helices,
Perfo rings and expanded-metal rings) carried out by Reichelt [121a] has revealed
that while a great variety of ring-type packing units have been developed no satis-
factory comparative assessment of their efficiency in material exchange and of their
fluitl dynaniic behaviour can be made.
Fig. 346 shows various fornis of cylindrical packing units, all of which are based
011 tlw Kaschig ring, a. Some non-cylindrical types are illustrated in Fig. 347. Among
the latter we must include the wire “Heli-Pak” [ 1221 and wire-gauze “Octa-Pak”

6 6
m
5 5

4
t’
I
Q
3
L 3
I

2 2

1 1

throughput ---C

Fig. 345
Comparison of the separating efficiency of various packing units (Iieritnyi [121])
Column: 24 rnm diameter, effective separating length 50 cm
Pressure: atmospheric
Reflux ratio: cy)
Test mixture : n-heptane-methylcyclohelrane
iranchembook.ir/edu

432 7. Constractional msteriaki and apparatus

Fig. 346
Cylindrical packing units (based on the Raschig ring)
a ) Raschig. smooth and matt g) “Intos” ring
h) Idem, w i t h circular grooves h) Helix
c) Idem. with lengthwise grooves i) Wilson spiral
d) Idem, with internal structure k) Spool
e) Prym ring 1) Pulley
f ) Raschig ring of wire gauze (Dixon ring) m) Pall ring

Fig. 347
Son-cylindrical packing units (baaed on the Raschig ring)
11) Sphere, smooth and matt q) Berl saddles
0 ) Kirschbaurn’s packing r) Twin Bectors
p) Bead, smooth and matt s ) Cross packing

Fig. 348 Fig. 349

Podbielniak’s * ‘Heli-Pa,k” Podbielniak‘s “Octa-Pdi”
iranchembook.ir/edu

7.8 Packings 423

bodies of Podbielniak (Figs. 348 and 349)and the perforated metal semi-cylinders of
Cannon (Fig. 350),all of which are very efficient.
The cylindrical forms that are most commonly used in the laboratory and in
pilot plants are:
Fig. 346a, Raschig rings of glass (smooth and matt), of ceramic material or of
metal, 2 - 10 inm;
Fig. 346e, Prym rings of metal, 2 -10 mni ;
Fig. 346/, Raschig gauze rings, with or without partition, 3-10 nim:
Fig. 346h, Helices of glass or metal, 1.4-10 mm;
Pig. 346i, Wilson spirals of glass, 3-6 i n n
P r y n rings are as a rule made of metal and are used for comparisons with in-
dustrial columns. Data on their stage efficiency are scarce. According to Thormann
[lUl 2.5 x 2.2 inm Pryin rings give the same HETP as 4 x 4 nitn metal helices. The
metal Raschig ring is now available also with circular holes as Perfo ring [123].

Fig. 350
Cannon's perforated metal-sheet semi-cylinders

Dixon's Raschig rings of wire gauze (about 100-140niesh), with one or two
turns and with or without partitions, are iised to an increasing extent [124, 1263.
Brass-wire netting Dixon rings of sizes 4.3 x 3.3 inm and 5.1 x 5.1 inin were investi-
gated with chloroethane/benzene and H,O/D,O as test mixtures by Zelvenshi et al.
[125a]. The measurements were performed in a column of 43 mm diameter and
1000 nim packing height in the pressure range of 50 to 750 torr. Chemical pretrcat-
ment of the rings with 25 to 30% NaOH or KOH resulted in different permeabilities,
presslire drops and retaining powers as compared to unpretreated rings. The results
are represented graphically.
Data for wire-gauzepackings have beencollected by Ellis and Yarjavandi [ 1661.This
packing is not made in sizes smaller than 1.5 min (1/16 inch), but gives HETP values
of about 1.5-2.5 cni and better. A favourable feature of this type of rings is that the
efficiency drops hut little with increasing column diameter. Measurements carried
out hy Kirschbaum [127] showed that a column of 100 mm diameter, containing a
1-nictre layer of bronze wire-gauze rings (single turn without, partition, 3 x 3 mm),
gave an HETP of 1.67-2.8 cm a t vapour velocities of 0.02-0.27 iii/sec (Fig. 3 [51]).
The test was performed at infinite reflux ratio with an n-heptane-methylcyclo-
hexane mixture.
Helices of thin wire (Grosse-Oetringhaus) have an extremely high efficiencv. The
author's experience shows that they can also deal with high loads. If such hclices are
to be resistant t30corrosion they should he made of stainless steel ( p . g . 18: 8): fiirther-
iranchembook.ir/edu

424
~~
7. Constructional materials and apparatus

more, their manufacture to a constant length in sizes smaller than 2 mm is difficult.

Their price is consequently high. On the other hand the efficiency of the mall sizes
(1.3-2 mm)is exceptional; an HETP of 1.0-1.6 cm can be obtained with them. The
latter sizes are made of 0.15 -0.2 mm wire. For one litre of coluuin space the number
of rings required is approximately as follows:
2 iniii 115,OOO
4 niiii 12,600
so 111111 100
I t was an obvious step also to make such helices of glass (Fenske helices). The
writer carried out some tests on closely-wound glass spirals of 4 ~ 4 x 0 . 3and
7 x 7 x 0.3 nun at atmospheric pressure with benzene-CCl, mixture and obtained
the results shown in Table 71. They are available as both single and multi-turn
helices.
The glass paclung known as Wilson spirals diffcrs from metal and Fenske helices
(which have turns of close pitch) in being made with spaces of 0.6 mm between each
turn (Fig. 346i). Their efficiency is of the same order as that of equally large Raschig
rings, but they can handle large loads.

Fig. 351
Tests with bronze wire-gauze
rings, 3 x 3 mm, single turn,
no partit,ion,by Kirschbmm
Column diameter: 100 mm;
packed layer: 1000 mm; test
mixture: n-heptane-
methylcyclohcxane; = cy.
vapour velocity d

Table 71
Teats with 4 and 7 mm spirals

Load (ml/h) 100 200 400 800

HETP with 4 mm spirals (cm) 5.2 9.6 6.1

HE" with 7 mm spirals (cm) 5.4 6.4 8.0

Column height, 1 m; diameter, 30 mm; effective length, 920 mm; reflux ratio
1' = rr; charge: 40 mol%
iranchembook.ir/edu

7.8 Packings 485

Data for glass helices with one turn are given in Table 70. Single-turn wire 1ielic.t.h
have also been investigated; Thormann [19] reported the HETP’s listed in Table 7 2
for this packing. Single-turn wire helices hence have about the sanie efficiency as
glass Raschig rings or glass spheres of the same size.
Of the non-cylindrical forms of packing the following are employed in the labora-
tory and in semi-technical plants.

Table 72
Values of t,he HETP for single-turn
wire helices

1.6 2.5-3
2.5 4
4.0 4-8
5.0 12

Fig. 352
“Intdox” packing

Fig. 3 1 7 ) ~Spheres
, of glass, smooth and matt, 3-8 mnr;
Big. 347 q, Saddle-shaped packing of porcelain, special compositions or wire gawt’,
4-10 mm;
Fig. 348, “Heli-Pnk”; wire gauze, 1.8-4.5 nim;
Fig. 349, “Octa-Pak“; wire gauze, 5 and 7.5 mm.
Spheres, smooth or matt, are still used to some extent, particularl- in j m p r a t i v t ~
work. Their large weight, however, limits the length of the column.
Saddles (Fig. 347q) probably represent the best solution of the packing J U - O ~ I W I
froni the aspect of flow. Every surface of such a body lies at a slant, whatever its
position, so that there are favourable conditions for contact between vapour and
liquid. For the liquid flowing centrally the angle of distribution is 55-60” : the surface
area of a saddle exceeds that of a n equally large Raschig ring by 30 to 60°,, in the
size range of 4-10 mni. The resistance to vapour flow is very low; 8 x 8 inn1 Paddle.,
for instance, cause only half the pressure drop of Raschig rings of the sanie sizct. -1
iranchembook.ir/edu

426 7. Constructional materials and apparatus

requirement for high efficiency of these packings is, however, that they are properly
made and do not have a ridge. Saddle packing has also been manufactured of 100-
mesh braes wire gauze (MacMahon packing [ 1281. A development of the saddle is the
.;o-called “Intalox” packing [129] (Fig. 352); it has larger surface area and an in-
creased free space, so that. the pressure drop is still smaller than with saddles of the
original type. It is made in 6 and 12.6 m m (l/, and l / z inch) sizes for laboratory and
semi-technical use. Another variant is furnished by the Hydronil “Plastic Intalori”
[130] which has two holes (Fig. 353)and is equivalent in efficiency to ceramic Intalox

Fig. 353
“Hydronil Plastic Lntalor” parking, of various plastic materials
and sizes 1. 2 and 3 in.

Fig. 354 Fig. 356

‘3upersadd le” (Kitiirschba11 m ) Metal-made “Interpnck” unit
iranchembook.ir/edu

7.8 Packings 427

Table 73
HETP values for non-cylindrical packing bodies

Size HETP References

(mm) (om) chapter 5

Perforated sheet-metal semi-cylinders 4 1.3-2 [19]

(Cannon)
Hel-Pak 1.8 0.5

Type Size HETP (cm) at a load of

(mm)
200 400 600 800 ni1/11
28.3 56.6 84.9 113.2 ml/cmz. h

Porcelain saddles 4 4.73 5.29 6.00 6.43

6 6.00 6.92 8.18 9.00
8 7.50 8.18 9.00 11.24
10 9.00 10.00 11.24 12.85
Test mixture: CCl,-benzene, 40 mol%
Column diameter: 30 mm; effective length 460 mm; v = rn

packings. While the pressure drop is the same the throughput, is twice as large as
with Pall rings. A further evolution is found in Kirschbaum’s “Supersaddle” [131 J
which has partitions extruded from the saddle surface (Fig. 354).
Almost all metals are used in the manufacture of “Interpack” packing units
(Fig. 355) which are available in sizes 10/10/0.3; 20/25/0.4; and 30/35/0.8 mni [132].
They are characterized by a relatively smal pressure drop. Plastics are under con-
sideration as constructional material.
The HETP’s for non-cylindrical packing, which have not been included in Table 70,
are listed in Table 73. The values for procelain saddles were measured by the author..

7.8.2 Constructional material of packings

I n chap. 4.2 it was already pointed out in hour far the chemical nature of the
inaterial of which the column packing is made may influence the separation. The
materials in question are glass, porcelain, earthenware, various metals and their alloys
and recently also plastics. Glass or a ceramic material is in general preferred on
account of its resistance to corrosion and low price. Porcelain should be hard-baked
and free from iron, as catalytic effects may otherwise occur. For high efficiencies
packing made of 18:s stainless steel wire or wire gauze is unsurpassed.
The differences in the separating effect found with wire-gauze packing made of
various alloys are explained by Fuchs [133], on the strength of work done by For-
sythe et al. [134], as being due to differences in the wettahility of the metals and t,o
iranchembook.ir/edu

428 7. Constructional materials and apparatus

different adsorption effects that may arise as a result of their respective chemical
activities. Wolf and Gunther [135], however, later examined the siirface behaviour of
various metals with respect to test mixtures of increasing polarity and found that the
differences in activity were of the same order as the accuracy of the experiments,
t i % . 10 to 12%.
The author was able to demonstrate that the degree of wetting of packings - and
consequently also the separating efficiency - depend to a marked extent on the
average degree of roughness of the material in a given liquid-solid system [136]. It
was further found that polymers having swelling values of about loo/, show excellent
wetting properties.

Fig. 356
Teller’s polyethylene “Tellerette8”

The basic advantages of packings made of plastics as compared to other materials

are [137]:
- high accuracy to size;
- low weight;
- large relative free volume of packing (due to sinall wall thicknesses);
- high breakmg strength;
- high cleanability;
- relatively high resistance to corrosion.

The relatively high separating efficiency of packings made of plastic material

with a low wettability is interpreted by Teller [138] as resulting from the formation
of an interstitial hold-up due to the low energies of adhesion involved (cf. chap. 4.2).
This hold-up extends the residence time of the liquid phase in the column. Teller’s
polyethylene “Tellerettes” (Fig.356) yield a relative free voltme of 83%. They are
shaped such that the two phases encounter relatively small solid surfaces by which
they arc divided and diverted.
Ponter et al. [144] have suggested that packings with a low wettability inag be
advantageous for the separation of positive azeotropes. The admixture of 10 to 5006
Teflon r i n g s to glass Raschig rings of the size 0.6 x 0.6 x 0.4 cm yielded favourable
results. It was furkher shown [14!5] that with PTFE Reschig rings the pressure drop
IR twice aa large as with glass Raschig rings of the same size, using water 8 s liquid.
However, increasing the PTFE proportion in packing mixtures up to 80% leads to a
pressure drop increaae by only 20°/,. With Teflon rings, flooding starts a t smaller gas
loads than with g1-s rings.
Tests have shown that the polymers polytetrafluoroethylene and polypropylene
arc suitable packing materials for columns using concentrated sulphuric acid or
acid water as liquids [143].
iranchembook.ir/edu

7.8 Packings 429

Miniature rings are made of metals, porcelain or high polymers [139]. The shaping
of highly efficient packings made of thermoplastic material has been discussed by
Egberongbe [ 1401.
The use of polyethylene packings in separating columns of glass or steel involves
the risk of increased wall flow. The difference in adhesion between thc systems
liquid-column wall and liquid-packing unit gives rise to changes in the adhesion
process in favour of the column walls (cf. chap. 4.2). According to Stiirmann [141] wall
flow may be checked by the use of undulating column walls. Krell [137] has shown
that an opt,imum shaping of the inner column walls from the point of view of intcr-
face physics ensures favourable flow density spectra even when plastic packings a w
used. Polyethylene or Teflon sheet placed on the inner wall eliminates wall flow
[142].
iranchembook.ir/edu

8. Automatic devices, measuring and control equipment

Only by using accurate methods of measurement can high separating efficiency

and good reproducibility be obtained in laboratory and pilot plant distdlation. The
principal memurements to be performed in distillation are those of temperature,
pressure, reflux ratio and load. The determinationof data on the products of distilla-
tion will here be discussed only in so far as they are actually performed during the
course of distillation. In addition, reliable measurements are an essential requirement
for reliable control and recordmg of processes in distillation, and consequently for
automation. Control methods employed in industry are also applied to some extent
on a laboratory and pilot plant scale. The latest developments in the field of control
are regularly reported by a number of specialized journals.

8.1 Automatic equipment

In practically all modern apparatus for laboratory distillation certain functions
are now automatic. Heating is often controlled by contact thermometers, and the
rate of evaporation of the charge with differential manometers. Reduced pressures
are also frequently kept constant with automatic units. In fact, the present trend is
towards fully automatic control of the conditions of distillation and simultaneous
recording of the data measured, an object that has already been reached in many
industrial installations. The ideal procedure is for the mixture to be merely introduced
into the apparatus and the latter switched on ;it then automatically yields the required
products of distillation and at the same time prepares a diagram of the top and
bottom temperatures and other data that may be needed.
The extent to which measuring and control devices are employed in batch and
continuous distillation depends on the nature of the work and, to a considerable
degree, on financial considerations [12]. In practice a di8tinction can be made be-
tween senii- (or partly) automatic and fully automatic apparatus. In t>helatter all
operat ions except starting up are performed automatically; in partly automatic
apparatus only some of the functions are governed by regulating devices.
The following possibilities may be distinguished :
1. Partly automatic operation
a) Control
1. Stdl pot heating: by contact thermometers and contact manometers via electronic
proportional controllers (cf. sections 8.2.2 and 8.4.2) ;
2. Column heating jackets: by thermocouples, contact thermometers or sir thermo-
meters (sections 7.7, 8.2.1, 8.2.2);
3. Feed heating in continuous distillation : by contact thermometer (section 8.2.2);
iranchembook.ir/edu

8.1 Automatic equipment 431

4. Feed rate in continuous operation; by metering devices (section 8.6);

5. Boil-up rate (and thereby the load): by contact manometer (section 8 .4 2 ) ;
6. Reduced pressure: by pressure controller (section 8.3.) ;
7 . Amount of cooling water: by constant level device and flowmeter (section 8.G)
together with safety switch;
6 ) Recording (cf. section 8.2.1)
S s a rule for top temperature, feed teinpcrature, bottom temperature and wall
temperature in column heating jackets (section 7.7.3) by the use of therniocouplcs or
resistance thermometers instead of glass thermometers.
11. Fully automatic operation
a) Control
As for partly automatic operation, with the addition of:
5. Feed rate : by means of small metering pumps or automatic level control in a supply
vessel (section 8.6);
9. Reflux ratio: by an automatic colunin head, actuated by mechanical or electronic
means (sections 7.5. and 8.4.1);
10. Distillate take-off: by an automatic fraction collector (section 7.6);
b ) Recording (cf. section 8.2.1)
1. Top temperature as a function of the amount of distillate : by temperature recorder ;
2. Feed, bottom and heating jacket temperatures: by temperature recorder;
3. Reflux ratio: indirectly, by recording the difference between top temperature and
temperature in the middle of the column;
4. Pressure: by pressure recorder.

8.1 .I Fully automatic equipment for standardized boiling point analysis

A good example of a fully automatic apparatus is provided by that designed for
ASTM distillation method D 86 [l],a procedure which is employed to determine boiling
point curves of fuels etc. The operations that have to be carried out manually are
confined to the introductionof the charge into the flask, switching on the heating, the
reiiioval of the flask after the analysis and detaching the recorded diagram. Fig. 357
gives a block diagram of the apparatus. Besides the provisions already mentioned, the
apparatus includes the following devices.
1. A heating arrangement, responding to rapid changes in temperature and so set
that the first drop of distillate falls within the specified time of 5 .to 10 minutes
after the start.
2. -4device by which time markings are recorded a t every 2nd and every 10th
minute on the chart.
3. ,4n arrangement for the automatic determination of the 95 vol% distillate point
and the subsequent maintenance of a constant heat supply, as required by the
specification.
4. Automatic cooling of the receiver to 15 f 3°C.
5. A mechanism for setting the whole apparatus to its initial state after termination
of the distillation.
iranchembook.ir/edu

432 8. Automatic devices, measuring and control equipment

The accuracy of the temperatures recorded is 0.5 deg. C, that of the distillate vol-
ume# 0.1 ml. The apparatus maintains a speed of distillationadjustable at either4.501-
9 d / m i n and is designed to run 24 hours a day.
A similar apparatus, Asda Mark 3, of Messrs. Gallenkamp, London, is designed
to cover the following operations:
- ASTM D 86-JP 123, groups 1, 2 and 3, and D 1078, DIN 51 761 and 61 752;
- -STM D 86- 123, ~ O U 4;
P
- rapid distillation (no standard).

Thm it was possible to carry out 80,000 h o i h g point analyses within 8 months in
a laboratory equipped with 15 Asda apparatuses. This corresponds to 666 analyses
per month and apparatus.

U
Fig. 387
Block diagram of automatic apparatus for ASTM method D 86
2 = Thermocouple, 2 = Condenser ice bath, 3 = Heater, -i = Switch for 96y0
distillate point, which sets heater to fixed voltage, 5 = Timing motor for control
of distillation rate, 6 = Heater control, actuated b y 5, 7 and 8, 7 = Potentio-
meter for time measurement. 8 = Potentiometer for volume measurement,
9 = Contact for 950/, distillate point, 10 = Upper limit switch, t o stop volume
follow-up, I 2 = Lower limit switch, to stop return of volume follow-up, 12 =
Spring to return receiver to starting position, 23 & I4 = Light sources for photo-
electric cells, 15 = Motor for volume follow-up drive, 26 = Motor and switch for
“correction temperature”, 27 = Amplifier and switch for volume follow-up,
18 = Amplifier for first drop, 19 = Electromagnet for putting pen on chert and
moving receiver after first dropso that distillate rims down wall, 20 = Photoelec-
tric cell for recording firet drop, 21 = Photoelectric cell for volume measurement,
22 = Amplifier for thermocouple, 23 = Motor of recording drum, controlled by
temperature, 24 = Final boiling-point switch, 25 = Cut out, 26 = Potentiometer
slide-wire, 27 = Relay, 28 = Starting switch, 29 = Pen, 30 = Drum and chart
iranchembook.ir/edu

8.1 Aut,omatic equipment 433

It should be quite possible for a single analyst to attend to ten sets of automatic
apparatus. The main advantage, however, lies in the fact that subjective errors are
eliminated.
I n the Soviet. Union a piece of automatic equipment, due to Kanterinan et al. [3],
is employed for the distillation of petroleum products; it operates with a n accuracy
of 5 2 d e g . C. An automatic apparatus for the distillation of fuels and solvents
according to DIN 51751 developed by Salzer [4]is manufactured by AEG, of Berlin.
Fig. 358 shows another apparatus which has been described by Jirniann [5]. It
consists of the switch cupboard with compensation recorder for a resistance thermo-
meter and two distilling units for alternate operation. The apparatus covers the
temperature ranges +lo0 to +385"C and 0 to f285"C. The automatic Herzog appa-
ratus [6] also operates on the digital principle and prints the results iniinediately
on a strip of paper. For standardized boiling point analyses a t reduced pressure the
apparatus shown schematically in Pig. 359 [6] was developed by Zappe et a1. [ 7 ] .

8.1.2 Fully automatic equipment for fractionation

Basically, any apparatus for batch or continuous distillation may be autoiriated
to such an extent that it may be called fully automatic. Afully automatic apparatus
was developed by Ulusoy [S] especially for preparative work. In the section on low-
teniperature distillation, a fully- automatic apparatus due to Podbielniak was nien-
tioned (Fig. 178) the block diagram of which is given in Fig. 360. 9 number of

Fig. 358
Automatic apparatus for the distillation of fuels according to DIPU' 5 1751 on the basis of the
ASTM D 86 or the D 158, D 1078 and IS0 3405 apparatus with compensation recorder (record-
ing width, 210 mm)

28 Xrell, Handbook
iranchembook.ir/edu

434 8. Automatic devices, meamring and control equipment

I
Fig. 359
Apparatus for boiling point analpea at reduced preesum with automatic temperature control
(Zappe et id. [7])
1 - 3 = still pot heating and control, 4-9a = distillation apparatus with heating jacket, still
pot, Chisen etill head, adapter and receiver, 10,I1 = bubble counters with boiling oapillary,
12a, 18 = manometers with inolined tube, 17, I9 = photo-cells, 24-16 = recorder with ther-
mostat and amplifier, 20 = premure controller

automatic Labodeat plants were developed by Stage [9]. One of these is the con-
centric-tube column for work on a semi-micro scale (Fig. 267 b). Such designs offer the
advantage thmt the mlessuring and aontrol devices can be used for various columns
[lo], e.g., for the concentric tube column with charges of 5 to 100 ml, a t atmospheric
or reduced pressure down to 0.1 tom, with an efficiency of about 40theoretical stages;
or for the wire gauze column with charges of 100 to 1OOO ml, at atmoapheric or reduced
pressure down to 5 tom, with &11efficiency of &out 60 theoretical stages.
The front part of the apparatus (Fig.eslb) oarries the various units which a n
be pulled out by means of telescopic rails. For all functional g r o u p there are plug-in
printed circuit modules. The following circuits are distinguished:
- temperature control of the oil bath by electronic proportional controller: setting
the tempemture between 0 and 300°C or otherwise setting the load in drops per
minute;
- temperature control of the heat-compensating jacket : setting the teniperature
between 0 and 300°C or automatic follow-up control according t o the vaporir
tenipratiire in the column head;
- control of the reflux ratio with vapour division in steps of 0.5 to 200sec;
- collection of the fractions with change of receiver at preset intervals (in minutes);
iranchembook.ir/edu

8.1 Sutomatjc equipment 435

- vaciiuni control;
- protection against failures, such as interruption of cooling water supply, over-
temperature or overcharging of receivers;
- recording of the various temperatures and of change of fraction.

For packed columns, the load is relatively easy to control via the pressure drop
whereas concentric-tube columns with pressure drops of the order of low3torr re-
quired a novel control system. Fischer and Weyand [lo] solved the problem by
electronically measuring the reflux dropping rate. Depending on the desired load the
still pot heating is controlled linear-proportional to tjhe difference of actual and
preprogramined drop number. Thus it is possible to cover loads from 50 to 250 ml/h
exactly and absolutely reproducibly, the pressure in the still pot ranging from 760 to
torr.

I I

Fig. 360
Block diagram of the fully-automatic low-temperature apparatus “Thermocon” (Podbiel-
niak) (cf. also Fig. 178)
I = electronic control for distillate, 2 = distillate vapour, 3 = photo-electric measuring
cell, 4 = circuit for heat conductivity measurements, 5 = recording manometer. 6 = coil-
nection t o drive of recording drum, 7 = pressure and reflux condenser control, 8 = t8em-
perature, 9 = heat conductivity recording, 10 = motor, 11 = vapour outlet, 1,”= air, 13 =
connection to atmosphere, 14 = automatic introduction of gas samples, 15 = heated air bath

28”
iranchembook.ir/edu

436 8. Automatic devices, meaaurinp: a n d control eqmipment

The author’s automatic Destinorm apparatus “Automatic 62” (Q. 361)possesses

the following features [111 :
1. h k with electric Intus heater.
2. Means for sampling or measuring the bottom reflux just above the flask.
3. Automatic electronic control of the column heating jacket to the average inner
oolumn temperature (F‘ig. 368).
4. Automatic reflux control by a magnetically operated swinging funnel and timer
(Flg. 312).
6. An automatic fraction collector, containing 60 receivers, with syphon control
(Fig. 324).
6. An electronic thermostatted vacuuni controller, with a n accuracy of +0.1 nim Hg
(Flg. 377).
7. An automatic control of the boil-up rate (load) by contact manometer (Figs. 387,
388).
8. Temperature reading a t eye level by means of an optical system, or temperature
recording by an electronic recorder.
The distillate removed by the swinging funnel first passes through a heat ex-
changer (Fig. 361) where it is brought to the required temperature; low-boiling
fractions are cooled, solidifying fractions are heated thermostatically. The magnetic
valve 14 which follows measures off the fraction in the graduated tube. Next there
is an arrangement for taking single-drop samples 18 for analysis. At this point a
thermometer may also be introduced, or a funnel for calibrating the s-yphon system.
The latter is provided with a n adjustable plunger 20, allowing the fraction size to be
set with an accuracy of 50.1 ml. The change in pressure caused by the action of the
syphon acts on a manometer 22 containing a n organic contact liquid, and the resulting
pulse, after passing through an electronic relay, brings the mechanism of the fraction
collector into action and causes it to move up one step. The same pulse can be
used to actuate an auxiliary pen on the temperature recorder and so to record the
cut points between the fractions.

Fig. 361
Fully-automatic dktilletion apparatus “Automatic 62”
I = 1 still pot, 2 = 1 memuring device for bottoms reflux, 3 = 2 transition
pieces, 4 = 1 drop-forming tip, 5 = 1 column, 6 = 3 thermometers. 7 = 3 air
thermometers. 8 = column heating jacket, 9 = 3 differential pressure mano-
meters, 10 = 1 column head, 21 = 1 condenser, 12 = 1 vacuum connection,
13 = 1 cold trap, 24 = 2 magnet coils, 15 = 2 thermometers, 16 = 1 connecting
piece, 17 = 1 stopper, 28 = 1 sampling device, I 9 = 1 metering device, 20 =
1 sinking body, 21 = 1 vacuum receiver, 22 = 2 electrical contact manometers,
23 = 1 fraction collector, 24 = 60 test tubes, 25 = 1 vessel for fraction collector,
26 = 1 ground glass piece, 27 = 1 distillate distributor, 28 = 1 length of piping,
29 = 1 length of piping, 30 = 1 distillate tube, 31 = 1 distillate tube, 32 = 1
condenser for pressure drop measurementa, 33 = 1 heat exchanger, 3.1 = 1
vacuum receiver (BnschBtz-Thiele), 35 = 1 receiver pipette. 36 = 1 thermo-
meter, 37 = 1 funnel for feed, 38 = 1 stopper, 39 = 1 switchboard
iranchembook.ir/edu

8.1 Automatic equipment 437

9
’

5-

Fig. 361
iranchembook.ir/edu

@a 8. Automatic devices, measuring and coot.rol equipment

The fraction collector 23 normally contains sixty 20ml receiving tubes and is
built into a vacuum desiccator with a wire-gauze safety cover. It can be employed at
atmospheric pressure and at reduced pressures down to 1 rum. After 30 tubes have been
filled a signal sounds; alternatively, the apparatus may be niade to switch over
automatically to the next circle of tubes. When all the tubes are full there is anobher
sigml and the reflux controller is switched off so as t o stop the distillate take-off.
The part of the apparatus between the syphon and the fraction collector can
easily be taken apart and is so constructed that the distillate is not contaminated by
grease.
For preparative distdlationa correspondingly larger syphons and receiving tubes
are uaed; alternatively, receivers ma>*he arranged to ring a bell when filled to a cer-
tain level and at the same tinie switch over the column head to total reflux.
An automatic fractionating still designed hy Koukol et al. [13a] is eniployed in
the distillation after Micko for testing the quality of wines. The heating is controlled
ae a function of the aniount of distillate collected per unit tinie. The constant distilla-
tion rate enables a mixture to be separated into several fractions of equal voluniea in
a pre-set constant t h e of, say, 15 minutes. The volume is measured by means of a
thermistor filler which is adjustable as to the level and ensures a reproducibility of a
fraction volume of 25 rill to within fO.l ml.
Automatic apparatus for continuous distillation has been developed by Kolling
and Tramm (Fig. 166) and Stage (Fig. 167).
Rdck e t al. [13] have mentioned a niunber of very useful components for auto-
niating laborator1 columns, includmg protections against lack of cooling water and
against fire. Extensive surveys of the control of laboratory apparatus have heen
given by Abegg [14], Kossler and Vohdenal [16] and Kadlec [17].
Pilot plants for continuous operation whose flow chart is identical with that of the
subsequent large-scale plant should he largely automated in order that real c'oni-
parative values h a y he obtained. Fig. 362 shows a pilot plant for the simulation of
criide oil production conditions. The plant operates cont inuoualy and has R main
column and three additional strippers. It is used for the separation of niulticompo-
nent mixtures which are taken off in 4 distillate fractions. The bottom product is
pumped out of the main column. The colunins are designed as vapour-collision plate
columns. The results obtained with these coluinns are intended to be transferred
directly to a large-scale plant. The metering of the material flows is done by means
of pumps P (c/. chap. 8.6). The rates of evaporation, the heating jacket teniperatures
and the refliix ratios are controlled from the switch cupboard 1. The vacuum con-
troller 2 ensiires a constant reduced pressure and the safety relay 3 switches off the
plant as soon as the supplj- of cooling water is interrupted. Naturally, the more
important teniperatures are continuously recorded on electronic recorders [ 17 a]
Fischer Labor- rind Verfahrenstechnik have also developed automatic apparatus
for the determination of mineral oil coinpositions according to ASTM 2892 and
DIN 51567. The device is equipped with a falling-film column with wire-netting
insert. It permits charges from 500 to 4OOO rul and operates a t normal pressure or
vacuum to 2 torr and at temperatiires to 350°C. The separating power is reported to be
atmiit 14 to 17 theoretical plates.
iranchembook.ir/edu

8.1 Automatic equipment 439

1 111

.
I

Fig. 362
Continuously operating pilot. plant with vapot~r-collisionbubble-plate columns for t h e
Gmulation of crude oil production conditions
iranchembook.ir/edu

440 8. Automatic devices, measuring and control equipment

8.2 Temperature measurement and control

8.2.1 Temperature measurement
In distillation the most important temperatures to he measured are those of the
overhead vapour, close to the point of condensation, and of the liquid in the still pot
in the m e of readdy dewiuposable substances. In continuous operation it is necessary
to know the temperatureof the preheatedfeed at the inlet. Other temperatures often
required are those in heating jackets and heating media (liquid baths or vapour
jackets).

Table 7 4
Standardized glass thermometers for distillation

TGL 40-335 liquid thermometer, distillation thermometer

40-336 liquid thermometer, thermometer for testing fuel oils and fuels,
thermometer A and B for boiling course
10-339 laboratory thermometer with conical joint NS 14.5/23
DIS 12779/71 distillation thermometer
13784/55 Sheet 1 thermometer with standard ground joint, distillation thermo-
meter (to be calibrated)
12784163 Sheet 3 thermometer with standard ground joint, thermometer with
bulb (to be calibrated)
12785158 Sheet 1 thermometer for boiling course (to be calibrated)
12788/63 Sheet 1 thermometer well with standard ground joint 14.6123

In laboratory and pilot plant distillation glass thermometers are still largely
employed. If the temperatures lie within a restricted range it is, however, often inore
useful to carry out the measurement electrically, i.e. with thermocouples, thermistor
or resistance t.hermometers, since it is then possible t o record the temperature auto-
matically. The possibilities of temperature measurements in the laboratory have been
reviewed by Vanvor [lS]. The measurement, generation and control of temperatures
from high to low have been discussed by Laporte [15].
It is important to place the measuring point in the optimum position. I n the
c*olrinmor still head it should be about 10 nim below the junction of the vapour tube.
The temperature in the still pot should be measured at its lowest point, so that
superheating may immediately become apparent. I n flowing vapour or liquid the
measuring point Bhould be well insulated and should lie in the centre of the streani.
Standardized glass thermonieters for distilling purposes are listed in Table 74.
Besides those specified in the standards, there are numerous types made for
special purposes, such as angle t.hermometers (useful for measuring inner colirnin
t emnperatures), flanged thermometers and thermometers having capillaries backed
with another colour to facilitate reading [19].
iranchembook.ir/edu

8.2 Temperature measurement and control 441

The thermometers most usually employed in distillation are of the enclosed type,
in which the capillary and the scale are fused into a protective jacket (Fig. 363). For
the temperature range of -58" to +360°C, the normal thermometer liquid is iner-
cury or its alloys; for low-temperat,ure distillation (-58" to -200°C) it may be
pentane or a liquid with good wetting properties.

d b

Fig. 363
Distillation thermometer wit- standard ground joint 14.5
a = Richter's scale-mounting, b = Stift's scale-mounting,
c = safety bulb, 1, = stem, I , = immersion depth

It should he noted that all the thermometers listed above except the ground-joint
type (Fig. 363) are calibrated for total immersion; in other words, they read correctlp
only when the whole of the mercury thread is at the nieasured temperature. Short
thermometer wells can thus give rise to appreciable errors. According to Piatti and
Marti [20] the error is small only if the mercury bulb is just immersed in the oil in the
well. I n practice, total immersion thermometers should be mounted so as t o be coin-
pletely iirmiersed in the vapour or liquid, although the ease of reading is thereb?-
adversely affected. Otherwise a correction for emergent stein has to be applied (cf.
iranchembook.ir/edu

442 8. Automatic devices, measuring and control equipment

section 4.13). For this reason the use of ground joint thermometers is to be pre-
f e d . The data for such a thermometer are given in Table 75.
Ground joint thermometers are also comeroially available for ranges of 50°C
and a @uatrion t o 0.1"C and as angle thermometers. This type, officially gauged,-
yerves as standard for ground joint thermometers and for sharp distillations. The
gauging certificate specifies the constant depth of inirnersion and the mean t hread

Table 75
Data of the distillation t liernionieter with standard ground joint according to TGL 40-339,
DIK 12784/55

Tgpe Measuring Graduation Temperature Calibration

rmge ("C) (toT) range ("C) error ("C)

u to 111) 0.5 10.5

0 tn 2 2.50 1 O t o 50 +0.7
.Ior B 50 to 100 11.0
loo to 200 +1.5

11 to ,360 1
200 to 350
0 t o 50
*k0.7
_.

3.0

50 t o 100 9 1.0
100 to 200
200 to 300
300 to 360
*+2.0
1.5

*2.B
Stem 1, = 50 110 200 mm
Immersion depth lr = 32 92 182 mm

temperature during gauging. With a graduation to 0.1"C the reading error is of the
following orders:
0 to 50'C: T 0 . 0 5 T
50 to 100°c: ~ 0 . 0 5 " C
100 to 150°C: +0.2'C
160 to 200°C: *O.?Y'
200 to 250°C: A0.G c'
Since the errors of simple thermometers lie hetween 1 0 . 2 and &S deg. C and
$as8 iindergoes a process of aging, it is apparent that thermometers used for distilla-
tion need to be checked a t intervals. AS already stated, some thermometers are
provided with auxiliary markings with which corrections can be determined. Others
should be testrd by comparison with standard thermometers. (The apparatiis of
Junge and Riedel, designed for this purpose, functions in the same way as the Thielr
irielting point apparatus and is suitable for checking thermometers calibrated at total
itumersion up to 300°C.) For testing ground-joint thermometers the author employs
the equipment shown in Fig. 364; this can also be used for checking thermocouples.
Standard compounds in the vapour phase provide the temperatures required :
Table 76 lists a number of compounds that can he used in the temperature range of
+%lo to 210°C. For higher temperatures, glycerin or some other stable, high-boiling
iranchembook.ir/edu

8.2 Temperature measurement and control 4-13

Fig. 364
Krell's thermometer calibration apparatus for
ground-in thermometers and thermocouples

Ta.ble 76
Substances suitable for vapour baths for test.ing thermometers

Compound I.P.760
"C

Diethylether 34.6
Nethylenechloride 40.67
a4cetone 56.13
Methylacetate 56.96
Chloroform 61.0
Methanol 64.72
Carbontetrachloride 56.69
Ethylacetate 77.06
Cyclohexane 80.8
Water 100.00
Toluene 110.6
n-Butanol 117.75
n-Butylacetnte 126.2
Chlorobenxene 131.69
lsoamylaceta te 142.0
Bromobenzene 165.6
Decane 174.0
Aniline 184.4
Ethyleneglycol 197.4
Benzylalcohol 205.2
Nitro benzene 210.6
Naphthalene 217.96
Diphenyl 254.9
Benzophenone 305.9
iranchembook.ir/edu

444 8. Automatic devices, measuring and control equipment

substance is employed at a reduced pressure. Three thermometers can be tested a t

the -me time; the fourth is a standard one provided with an official calibration
certificate, stating its errors at various temperatures. Care should be taken that the
temperature of the exposed thread does not depart appreciably from that prevailing
during its calibration; otheryise a correction must be applied.
For electric temperature measurements t,hermocouples and resistance thermo-
meters are universally applicable. They are commercially available in a variety of
designs [ 181.
For automatic recording, thermocouples or resistance thermometers are used. An
advantage of thermocouples is that they can be used in small cavities and thin layers,
for instance in the film flowing down the wall of a heater in molecular distillation.
Thermocouples and small resistance thermometers give a rapid response and can be
used over a wide temperatlure range; for thermocouples this range is approximately
-200 to 1800°C.
For the cold junction zero thermostats operating on the J'eltier principle have
proved to be appropriate [Zl].
Resistance thermomieters usually contain a platinum resistance element' and can be
used from -220 to +750°C. They rely on the resistance of a platinum element, (a
thin layer, a cylinder, a straight or coiled wire) which changes by about 0.4OiO with
cvery degree centigrade. I t follows that if temperatures are to be determined to within
0.01 deg. C the resistance must be measured with a n accuracy of a few hundred-
thousandths of the total value. The methods used for this purpose depend on the
accuracy required. If a Wheatstone bridge iseniployed, the changes in resistance can be
used for recording and for control, for example to control the temperatures of column
jackets to follow column inner temperatures, and t o record the reflux ratio indirectly.
The evaporation of metal onto a surface forms a resistance element capable of
determining the temperature of the surface accurately [22]. Data for surface heaters
can be obtained in this way.
In pilot plants, resistance thermometers are commonly used to measure vapour
and liquid temperatures. They often make use of two coils, one for recording and the
other for control. The Quickfit all-glass version (Fig. 365) has joining sockets of
nominal widths 25 and 40 mm.
9emicondvctor.s (thermistors), in combination with a recorder/controller, allow
temperatures to be measured and controlled over a narrow range with a precision of
~O.OOl"C,and for this reason are particularly suitable for use in vapour pressure
measurements and the determination of phase equilibria. The "Thermophil" eletronic
therinometer appears to be useful for measuring temperatures on surfaces, as in thin-
film distillation and on outer column walls. The sensing point, encased in a glass or
metal coating, contains a germanium semi-conductor 0.25 mm in diameter, which
withdraws negligible amounts of heat from the place of measurement. The apparatus
is made for various temperature ranges between -50 and +450°C and is accurate to
1 -2 deg. C. The signal produced is amplified electronically and actuates a pointer on
a scale. Up to 10 points may be connected to the apparatus b y leads of any desired
lcmqth. The newly developed digital thermometers can be u ~ e dwith thermistor.;,
thermocouples and resistance wires for sensing.
iranchembook.ir/edu

~~ ~
8.2 Temperature measurement and control
~
445

Temperature recorders are generally designed to serve one, two, four, six or twelve
measuring points.
Since temperature ranges of 200 deg. C occur in laboratory distillation the recorder
should give a relative accuracy of 0.1% or better and an absolute accuracy of 0.25%
of the full scale, if the recording is to be as accurate as observation from glass

Fig. 365 Fig. 366

All-glass resistance thermometer Contact thermometer adjustable by
means of a rotating magnet
Measuring range: -200 t o +3OO0C

thermometers. These requirements are fulfilled by various types of electronic re-

corders which are connected directJlyto a platinum resistance thermometer (100 ohm
at 0 "C). I n continuous distillation the instrument can also be employed for recording
deviations from a specified temperature.

8.2.2 Temperature control

For controlling the temperature of baths or flowing liquids and of various portions
of an apparatus (such as column jackets, tube heaters, etc.) contact thermometers are
largely used; they are obtainable with standard ground joints (e.g., NS 14.5).Contact
thermometers can be obtained either with contacts set to definite temperatures or
with variable temperature. I n the variable type (Fig. 366) the upper thermometer
scale is used to set the required temperature, while the lower one carries the electrical
contact. A contact thermometer with a spiral bulb with a large surface area (Fig. 367a)
iranchembook.ir/edu

446 8. Automatic devices, m d g amd control equipment

gives a more rapid control. Fig. 367b shows the difference between the speed of
response of this type and the usual cylindrical bulb [23].
Contact thermometers of a maximum-minimum type are also manufactured;
these make contact both on exceeding an upper limit and on dropping below a lower
limit. By means of a fine adjusting screw the temperatures may be set with an
accuracy of 0.1 “C[231. Temperature control may be arranged to follow a variable
reference temperature hy moving the thermometer contact with R synchronous
mot or.

##

a)
L“

b)
1.5 2.5 5
time -
10 15 20 30s 40

Fig. 367
a) Contact thermometer with spiral bulb
b) speeds of response of contact thermometara with spiral bulb and with cylin-
drical bulb

Quite satisfactory experience has been gathered with contact thermometers

having a special three-pole plug and sliding ground joint,, the latter ensqing that
immersion can be continuously varied [a].
Between the contact thermometer 1261 and the controlled apparatus a relay is
placed, since a contact thermometer is not capable of carrying a normal heating
current. Relays with vertical mercury switches have proved the most, satisfactory in
laboratory use, as they contain no parts subject to wear, such as bearings, levers or
flexible leads [26].
Transistor relays may also be used in connection with contact manometers ( c f .
chap. 8.3.1 and 8.3.1.1). They are commercially available as a comb,ination of stand-
ing or supported device. They operate on the “normally open” and “normally
closed” principles where a circuit is switched on and off, respect,ively, on niaking a
contact. These relays me further employed for photoelectric circuits, e.g. in the
level control of liquids and in vacuum control (ci. chap. 8.6 and 8.3.1.1). As safety
relays they switch off the circuit if the contact thermometer fails.
iranchembook.ir/edu

8.2 Temperature ~neasurementand control 447

Universal electronic control devices have been discussed by Fischer [ 2 7 ] . He re-

ports the use of proportional controllers besides electronic relays for temperatures hr-
tween -200 and +8OO0C. I n this case the sensing is done by resistance thermometers
having ground joints or flanges. A special advantage is the low surface load since the
frdl electrical input is required only for starting up the heating.
A frequent requirement in laboratory distillation is the control of a heating jacket
temperature in accordance with that inside the column. Methods for the thermal in-
sulation of coluinns have been discussed in section 7.7.3.

E
!
I
!
!
!
h !

!
I/
I
i4 i!
Fig. 368
Aiytomatic control of t h e column jacket temperetiwe acrording t o the inner
column temperature
a = Column, b = Heating element, c = Air thermometer, d = Glass wool in-
sulation, e = Column jacket, f = Control manometer. q = Electronic relay.
h = XS 14.5, i = 3 8 29

I n batch distillation the column temperature increases during the course of the
operation, frequently by sudden increments; the adjustment of the jacket tempera-
tures by hand is consequently a matter of some difficulty. As the result of tempera-
ture lag, differences up to 30 deg. C between the jacket and column may occur, but
these differences may be reduced considerably by automatic control. For this purpose
air thermonieters, thermocouples or ,resistance t herniometers can be eniployed as
sensing devices. Thermocouples are placed a t the top and hottoin of the column :
these together act on,the coil of a galvanometer, the position of which is “sensed” at
short, intervals by an electric switching device. The latter opens or closes the circuit
iranchembook.ir/edu

448 8. Automatic devices, measuring and control equipment,

of the h a t i n g jacket [28]. A circuit for the control of the heating jacket based on the
internal column temperature has been described by Hutla [ZY].
The distillation controlling device Minitron V (Fig. 386) of Fischer [27]is equip-
ped with two electronic proportional controllers serving the still pot heater and the
heating jacket, respectively. Resistance thermometers of the type PT 100 do the
sensing in the temperature range of 0 to 350°C. The heating jacket may be controlled
hy adjusting the temperature on a helipot or by automatically following up the in-
ternal oolumn temperature. A resistance thermometer in the column head instead of
the helipot in the controlling device sets the desired value. The heating jacket
temperature ir then automatically regulated according to the vapour temperature
in the column.
With ajr tbrriwnteters, one limb of which is placed inside the column and the
other in the jacket, the author [ll] was able to control to within f0.5 deg. C. This
aocuracv could be achieved by the use of a contact manometer containing a conduct-
ing organic liquid, coupled by an electronic relay t o the jacket heating circuit
(Fig. 3%).Up to 400 switchings per hour could be obtained with this arrangement.
The temperatures in the jacket and the column remained completely parallel, even
during sudden increases in temperature. The advantage of the method is that the
jacket (which should preferably be subdivided into sections of 25-50 cni length) is
controlled by the inside temperature of the column section. If the uppcr jacket
section is controlled, in the more usual nianner, by the temperature in the still head,
during the distillation of a transition fraction the jacket will be cooler than the inside
of the column. The upper colunin section then acts as a dephlegniator and produces
additional reflux. The method of control just described also makes it possible to
maintain a fixed temperature difference between the jacket and the column, if
reqiiired.

8.3 Pressure measurement and control

The pressure ranges in which various forms of distillation are carried out can be
defined as follows:

Pressure range Type of distillation

Above 760 mni Hg Pressure distillation
Bet,ween 760 and 1 inin Simple and fractional distillation
‘‘Flash” distillation
Thin-film distillation
Between 1 and 10-6 mm Thin-film distillation
Molecular distillation

In these various ranges different methods for the measurement and control of
pressure are employed [2]; these will be discussed in the following sections [30]. A
good review of the methods of pressure measurement has been compiled by Leck [31].
iranchembook.ir/edu

8.3 Pressure measurement and control 449

8.3.1 Pressure measurement and control above 760 mm

and from 760 to 1 mm Hg
When distillation is carried out at pressures up to 2 atni absolut,e - cf. section
5.4.5 - mercury manometers of 1 m length are suitable for measuring the pressure
above atmospheric (Fig. 369).
The distillation prespure is then given by the relation

b .rip pi, (202)

where b = corrected barometer reading (torr), dp = pressure difference with
respect to atmospheric as indicated by the manometer, p1 = distillation pressure in
the apparatus.

b
App.*C
C

Fig. 369
Ahnometer fc determinin the differenee
between distillation pressure and atmospheric
pressure

At higher pressures, diaphragm or Bourdon gauges are used. 9method for con-
trolling pressures above atmospheric has already been described in section 5.4.5.
The aneroid absolute pressure controller of the FA 149 type (manufactured by
Wallace & Tiernau-Chlorator GmbH, Gunzburg) allows ahsohite pressures to be
controlled in the range of 20 to 1520 torr with an accuracy of 0.25 tom.
When distillation is performed a t atmospheric pressure it is necessary to read the
barometer, from time to time, so that boiling points may be corrected to 7601nrn
(see section 4.13). A reliable mercury barometer should be used [32]. The one after
Gay-Lmsac or that after Schrodt-Kiefer is often used in the laboratory. For an
accuracy of 30.1 mm the vernier should be employed in reading the mercury level.
The liquid pressure gauges developed by Nickel [32] make allowance for the surface
tension of the measuring liquid and enable corrections for zero deviations such as
those caused by temperature changes.

29 Erell, Handbook
iranchembook.ir/edu

450 8. Automatic devices, measuring and control equipment

It is advisable to have barometers checked occasionally against a standard in-

strtunent so that the corrections t o be applied a t various readings may be known.
Aneroid barographsshould be calibrated in the same way. These instruments - which
generally have an accuracy of not more than f0.5 mm - are provided with charts
running either for a day or a week. The charts should be filed, in order that the
atmospheric pressure a t any time can be subsequently found. Aneroid barographs are
&R a rule compensated for errors due to temperature in the range of -4OO" to +40"C
by binietallic strips.
The barometer reading is similarly required if reduced pressures are determined
by means of open manometers (Fig. 369). The mercury, after thorough cleaning, is
introduced, without air bubbles, through the Pafety vessel a, and the apparatus is
connected a t b. Vessel c is also a safety trap, functioning in the event of a sudden
decrease in pressure. By means of the adjustable scale d the difference with respect
to atmospheric pressure Jp is read off. The pressure p , is then found to be
pi = b - J P , (203)
with 11, = pressure in the apparatus (torr), b = corrected barometer reading (torr),
31) = pressure difference as indicated by the manometer (torr).
The use of open nianometers for pressure nieasurement has the advantage that
the wliole range from atmospheric pressure down to a few millinietres can be covcrcd.
For greater precision the mercury levels should be determined by cathetometer. An
arrangeiiient for reading and recording the mercury levels automatically has been
described hy Farquharson and Kermicle [33]. At a height of 80 cm the repeatability
of operation is better than f0.05mm and the absolute accuracy better than
-&O.ll mm.
Differential pressure determinations, as described above, are simplified con-
siderably if the pressure gauge and the barometer are combined in a single apparatus
(Fig. 370). Instruments of this kind are made for a barometer range of 600-820 mm
and a pressure range of 1-31Onlm EIg. By using tubing with an inside diameter
constant to hO.01 min and with automatic zero correction [32] high accuracy may
be obtained.
Short nrunottieters (so-called Bennert manometers) have a closed evacuated limb
(Fig. 371) and are usually designed for a pressure range of 0-180 rnm Hg. They can
be read to within 0.6 mrn. Frequently, however, the closed limb is not sufficiently
evacuated, and air bubbles tend to pass into this limb during use. I n practice there-
fore, they can give rise to appreciable errors, especially in the 0-10mm range.
Frequent calibration of these instruments against an open manometer is therefore
desirable. A manometer of this type, due to Strohlein, is provided with a mercury seal
at the end of the closed limb, whereby it is possible t o pump out any entrained air.
The major sources of error involved in work with liquid manometers have been
described by Dosch [34]. i f a precision scale and a vernier for the readings are
employed the pressure can be read off with a maximum deviation of 0.2 mm.
The vacuum ruanometer due to Holland-Nerten [35] give6 measurements with art
accuracy of 10.5 mni in the 200-20 nun range; by the provision of an inclined limh
its accuracy is j 0 . l tnm Hg between 20 and 0 inm (Fig. 372).
iranchembook.ir/edu

8.3 Pressure measurement and control 45 1

Fig. 370 Fig. 371

Bsro-vacuum met,er Short Bennert-type manometer,
range 0 to 180 torr

29*
iranchembook.ir/edu

452 8. Automatic devices, meaauring and oontrol equipment

If instead of mercury, a silicone oil is used as manometer liquid, the precision of

reading increases tenfold. Thus the Laporte manometer use8 such a filling for the
range of 0.2 to 20 tom. A vacuum gauge working on the displacement principle [35]
which serves for precision measurements between 0 and 10 tom exhibits an accuracy
of 0.01 torr. Vacuum pump oil, purified paraffin and Amoil S cannot, be recommended
as manometer liquids since they absorb water and thus yield appreciable errors [36].
Following a suggestion of Bschmann, in the case of water sonie additives should be
ueed which reduce the surface tension.

Fig. 372
Fine-readingvacuum menometer (Holland and Merten)

Compreswionmanm.etersoftheMcLeod type can also be employed in the 760- 1mm

range. Since, however, their main field of application lie8 in the range of 10-1to mm
pressure, they will be diecussed under the latter heading. For the range of 0 to 60 niin
water column electronic contact manometers are now-in iise [36a].
Diaphragm qaauges have the advantage of giving a reading independent of the
nature of the vapour or gas [37]. Their sensing component is a thin metallic mem-
brane. The diaphragm gauge VM-M of Heraeus (Fig. 373) indicates pressures between
50 and lo-' mm with the a m e accuracy as a mercury manometer. It is particularly
suitable for laboratory and pilot plant work. An essential feature of the device is
that its sensitivity is not changed by the entrainment, of air. A variety of precision
pressure gauges in the form of mercury and aneroid manometers, especially such as
used for calibrating purposes, are manufactured by Wallace & Tiernau-Chlorat'or
GtnbH, Waaserburg. For pilot plant distillations various models of spring niano-
meters are suitable [2].
Whatever kind of manometer is ueed a cold trap should be placed between it and
iranchembook.ir/edu

8.3 Pressure measurement and control 453

the apparatus to remove condensable vapours. Care must also be taken that there is
no pressure drop between these two points. This may be checked by carrying out
coniparative pressure measurements on the evacuated apparatus before use, the
manometer being connected first to the thermometer joint and snbaequently nftrr
the cold trap.
I n the control of reduced pressures in the 760-1 torr range mechanical devices
as well as those conibining the mechanical and electronic principles are in me [38].
Two methods can then be adopted to control the pressure: the method using an air
leak and that of controlled evacuation.
With the vacuum punip running continuously, enough air is admitted into the
buffer vessel through a tap for fine regulation or needle valve to maintain the re-
quired pressure. Requirements for this method are that the V ~ C I I I I I I I pump operates
very uniformly and that the apparatus is leak-free (Fig. 374a).

Fig. 373
Diaphragm manometer, type VJI-Jl

8.3.1.1 T h e m e t h o d of c o n t r o l l e d evacuation
In this method evacuation of the apparatus takes place only when the pressurc in
the buffer vessel begins to exceed the rrquired value. A pressure controller and relay
can either start the pump or open a valve. A schematic diagram is shown in Fig. 374h.
I n the 760-10nim pressure range a device containing a fritted glass disc and
inercurp as sealing liquid can be used for control [39] (Fig. 375). The mercury level
.;hould be about 3 111111 above the tipper level of the fritted disc. With all taps of the,
controller open, evacuation is allowed to proceed until the pressure in the apparat 11s
is 1-2 inin above the predetermined value. Tap G is now closed. The level of the
mercury surrounding the fritted disc then falls, owing to reduction of the pressure in
vessel B, until the disc is at the point of becoming uncovered. At this monient valve
D is closed. The pressure in €3 is now equal to that in the apparatus minus the difference
in level of the mercury in the two vessels. If the pressure in the apparatus then in-
creases, the level in A falls further and gas is withdrawn by the vacuum pump through
thefritteddiec untilitisagainsea1edoff.By this inethodavacuumconstant to & 0 . 5 m 1 ~
can be maintained, provided that the regulator is kept a t a uniform temperatiire.
iranchembook.ir/edu

464 8. Autometic devioee, memuring and control equipment

Gilmont's pressure regulator [a] operates on the principle of the Cartesian diver
and can be used for pressures down to 6 mm. fig. 376 shows this apparatus, mounted
on a console together with a Stock manometer and a Friedrichs-Antlingernon-return
valve [41]. Its accuracy is about, f0.5 torr.

a)

Fig. 374
Pressure control
a) air-leak method
b) method of controlled evacuation
1 = vacuum pump, 2 = buffer vessel, 3 = manometer,
4 = pressure controller, 5 = electronic relay

e C

Fig. 375
Pressure regulator with fritted glass disc
a = fritted glass disc, b = residual pressure space, c, d , e = taps
iranchembook.ir/edu

8.3 Pressure measurement, and control 455

A higher precision, viz. hO.1 mm, can be attained by electronic naetkods. The
mercury nianonieter is replaced by one containing a high-boiling organic liquid,
increasing its sensitivity by a factor of about ten. A pressure difference of I i i i u i Hg
then corresponds to a difference in level of 10-13 inni in the limbs. The lower limit
of the field of application is determined by the vapour pressure of the filling. A photo-
electric method may be used, in which a thin pencil of light is directed over the float’
of a Dubrovin gauge onto a photo-electric cell. An increase in the pressure in the

U
Fig. 376
Gilmont’s pressure regulator
I (principle of the Cartesian
diver)

apparatus results in an interruption of the light by the float ; the darkening of the cell
causes the opening of a relay, which in its turn switches on the puiiip [ G I .
Another electronic method is based on the use of a manometer liquid such as
hutyl glycol that has been made feebly conducting by thc addition of a sniall qiiantity
of a salt such as sodium nitrite. The current passing at 6-8 1- need be only about
9;with an electronic relay it can be made to switch a current of 10-15 -4at
220 V. The controller functions as follows. I n one limb of a U-shaped vessel, niaintain-
rrl at a constant temperature, the required control vacuum is estahllshed (Fig. 374h).
If the same pressure is present in the apparatus connected to the other limb, the
contact liquid is at the same level in the two limbs. An increase in pressure in the
apparatus causes the liquid in the limb joined to it to fall; contact is then made in the
other, and an electronic relay starts the pump, which continues to run until contact
iranchembook.ir/edu

466 8. Automatio devim, measuring and control equipment

Fig. 377
Erell’s automatic vacuum assembly

6
1 Pig. 378
Stage’s automatic presmre regulator
iranchembook.ir/edu

8.3 Pressure measurement and control 457

is again broken. In the author’s automatic vacuum assembly (Fig, 377) an accuracy
of +O. 1 nim Hg could be obtained in the 300- 1 mm range, shown by tests carried otlt
with pure substances [43]. With phenol, pressure deviations of &O.l mm correspond
to boiling point deviations of +O.l deg. C. I n the separation of the isomeric xyleiles
at 70 Rim pressure, where boiling points have to be determined to within 0.1 deg. C ,
the pressure should be controlled with an accuracy better than 50.15 i i i r n Hg [&I.
Stage’s pressure controller (Fig. 378) contains a float, which is so placed in the inner

apparatus

fill with Liquid until

about 2 m m below contact tip

Fig. 379
Vacuum controller, type VK 1

linih of an oil iiianoiiieter 2 that mercury switch 3 is just kept level. If the ~xessurei i i
the apparatus, connected at 4, rises, the oil level in 2, and with it the float, falls. B?
the action of a relay a magnetic valve 5 then opens, and the apparatus is connected
to the vacuum pump via 6. The vacuum controller due to Fischer [27] uses a differen-
tial nianoineter with inclined limb which is surrounded by a silvered racuuni jacket
(F1g. 379).

8.3.2 Pressure measurement and control from 1 to lo-* mm Hg

The nieasurenient of very low pressures is a matter of considerable difficulty :
since differential readings with respect to atmospheric pressure then become too
inaccurate, indirect methods must be adopted. Observations of glow discharge$
enable the pressure to he determined only approximately. The fluorescence vacuscope
due to Burger corers the measuring points and tom. Intermediate
values may he estimated. Reproducible measurements in thin-film and inolecular
diqtillations require methods which are summarized in Table 77.
Measuring techniques for ultra-high vacua of 2 x lO-’O to 10-ls torr have been
reported by Hoch [46]. Advantages and disadvantages of the gauges given in Table 77
h a w been discussed bv Steckelmacher [47].
iranchembook.ir/edu

458 8. Automatic devices, measuring and control equipment

The Alphatron is an ionization gauge in which ions are generated by the bom-
bardment of the residual gas with a-rays [GI. The mol-vacuum gauges (anabsolute-
manometer type due to Knudsen) make use of the radiometer effect. The design
developed by Gaede [48] and the quartz thread vacuum gauge after Langmuir [49]
are preferably employed. As is the case with the mol-vacuum gauge, the diaphragm
microgauge does not include the kind of measuring gas in the calibration parameters.
The latter device covers the range from to 1.6 x 10-l torr. The problems in-
volved in pressure measurements in vacuiini apparat 11s with mercury diffusion
pumps have been discussed by Miiller [50].

Table 77
Applicability and measuring range of various pressure gauges (after Monch [45])

Type of gauge Pressure measured Dependence on nature Range of

(total or pattial) of gas preaent applicability
(torr)

Diaphragm and Bourdon gauges tota. independent 760- lI)-2

Gas discharge manometer

(Philip vacunm gauge) total dependent 1 ~ lo-j
-
Damping and friction mano-
meters total dependent 10-1 - 10-4
Thermal conduction manometer
( Pirani gauge) total dependent 10-1 - 1 0 4
Compression manometer
(McLeod gauge) partial independent (only for 760- 10-l
(not vapoiir pressures) permanent gases) 10-l-
Ionization manometer tota.1 dependent 10-8 -10-8
Radiometer (mol-vacuum gauge) totel independent 10-1- 10-8

It would take us beyond the scope of this book to mention all the vacuum gauges
that are manufactured. They are fully described in the various monographs on
high-vacuum technique [49]. In the 10-1t,o range the Pirani gauge is much used
and in the to range the ionization manometer. Combinations of these two
instruments into a single unit are available.
Vacuum gauges functioning electrically have the advantage that the pressure ix
indicated by a point,er, and that recording and controlling the pressure are possible.
Thermal oonductivity and ionization gauges both yield the t d a l pressures of gases and
vapours. To determine the residual pressure of any uncondensable gas it is necessary
to place a cold trap before the gauge.
These instruments can be calibrated against a McLeod gauge; this is usually
carried out with dry air. Calibrating curves are supplied by the manufacturers. "he
accuracy of Pirani gauges amounts to about 5% in the 1-5 x 10-3 tom range, that
of ionivltion gauges is about 3% in the 10-3-10-6 torr range. The measuring tubes
are provided with either NS 14.5/23 ground joints or flanges.
iranchembook.ir/edu

8.3 Pressure measurement and control 459

8.3.2.1 The McLeod compression manometer

The principle of this manometer, which gives a reliable reading only for permanent
gases, is the compressionof a certain volume of the gas, measured a t the pressure to
he det,ermined, to a small volume in a capillary tube. The gas sample is in this way
brought to a higher pressure that can readily be measured. The pressure p of the
original gas can then be calculated with sufficient accuracy by means of the formula

in which p = the pressure of the original gas (torr),

Ah = the pressure a t which the compressed gas is finally measured (the difference in
mercury level in mm),
V , = the volume of sphere G and capillary tube K (measured from the junction at R ) ,
V , = the volume of the compressed gas in capillary tube K.
Fig. 380 illustrates one of the commercial models of the McLeod manometer, in
which the mercury level is not raised by means of a levelling bulb but by atmospheric
pressure. The storage vessel T' is partly filled with carefully purified mercury and the
apparatus in which the pressure is to be measured is connected a t A . P is connected
to a filter pump. Whilst the apparatus is being evacuated the space in V above the
mercury intist simultaneously he pumped out. This is done by opening valve H to
the filter pump to such an extent that there is not too large a difference in pressure

Fig. 380
McLeod gauge
iranchembook.ir/edu

460 8. Automatic devioea, measuring end control equipment

Table 78
Volume of sphere and required amount of mercury in
commercial McLeod gauges
~ _ _ _ _ _

Range Volume of sphere Amount of mercury

(torr) (mu (mu

2 - 10-5 300 450

2-10-4 100 250
15-1e3 16 50

bet ween the manonieter space and that in vessel V ; otherwise mercury would rise in
the manometer titbe M or air would be sucked from M into V . To determine the
pressure in the apparatus, valve H is turned SO aa to adnut air slowly through tube L
(the latter preferably being provided with an absorption tube to remove moisture and
dust); the mercury is then forced up gradually into tube M. H is closed when the
mercury has reached the inark C, level with the end of capillary tube R. The difference
in height h is then read and p is calculated by formula (204). The volume V , is marked
on sphere G and the Volume F, on the scale against capillary tube K . Most manometers
of ths type, however, are so calibrated that pressitrep may be read off directlyfroni the
scale. If the mercury is raised up to mark B the instrument provides measurements of
pressures down to lop2mm; these are then read off on a scale along t ahe D. Prior to
each subsequent pressure determination the mercury must again be brought back by
the filter pump to below the tube junction at I z . The volume of the sphere in commer-
cial itlode15 and the amount of mercury required are shown in Table 78.
The numerous modifications of the McLeod manometer are dealt with in Ebert’s
hook [51]. A pneumatically controlled model with automatic setting which is indepen-
dent of the atmospheric pressure is described in detail by Peche [62). A number of
recomniendat ions are made which are intended to ensure reproducible measurements.
A s an example, Fig. 381 shows the “Vacutwope’’, a combination of a closed-limb
IT-tubemanometer and a McLeod gauge. It covers the range from 80 to 1 mm as
U-tube manometer and the 4-10-2 range aa compression gauge.
.4 form of comprewion manometer that is v e q convenient for use in the distillation
laboratory is the rotatable manometer of Moser (Fig. 382).I t hafi three nieasuring

Fig. 381
Va cuscope ’

U-tube manometer: 80 to 1 torr

caonipression manometer: 4 to torr
iranchembook.ir/edu

8.3 Pressure measurement and control 46 1

ranges, together covering from 760 to mm, and contains only 6-7 nil of nier-
cury. By turning the manometer anticlockwise around a ground joint one can adjust
it for the ranges 760-1 inin, 1 mni and 10-2-10-4 mni. The volume of the
niercury must be so chosen that a t atmospheric pressure bulb G is half full in positioo
4. In each case the manometer is turned so far that the mercury meniscus is level with
the mark M . The pressure is then read off against tube n. As in the case of the full-

a) &-
--

Fig. 382
R,otatable McLeod gauge of Moser
a) Side vien d ) Position for range 1 -lo-: mm
b) Position a t start e) Position for range lO-*-lO-4 mi11
c) Position for range 700-1 mm

Fig. 383
Kammerer’s compression xnaiiometer
10 to tom: Hg-filling, 15 1111
1 to lo-* tom: Hg-filling,17 ml
iranchembook.ir/edu

462 8. Automatic devices, measuring and control equipment

sized McLeod gauge the device has to be reset to its orighal state before each measure-
ment. The Moser gauge can be checked and calibrated by means of the large McLeod
manometer.
The compression manomet,er due to Kamuerer [53]also uses only a small amount
of merciiry (Fig. 383). Another a d v a n t q e is that the filling is greatly facilitated by
a new type of pumping mechanism. The instrument is supplied for the ranges of
10 to 10-3 torr and 1 to to-* torr and has an additional U-tube manometer for
lwessnres up to 80 torr.
I t should be pointed out again that all cobpression manometers indicate only
the partial pressure of the gas component which is not condensable a t the existing
temperature. The presence of condensable substances is confirmed if different
values of the pressure are measured in overlapping parts of two ranges [51].

8.3.2.2 Vacuum control t o pressures of lo-@mm Hg

Information on vacuum controllers for the medium and high vacuum ranges is
scarce. Laporte [49] has described a n instrument working on the thermal conductivity
principle in which the Wheatstone bridge is connected t o a signalling device which
produces an acoustic signal when the pressure exceeds a given value.
Siebet [54] has described an arrangement for regulating the pressure to mni
in a vrssel containing air, and Melpolder [66] has published a method for controlling
the pressure in the mm range. Melpolder’s apparatus is shown diagram-
matically in Fig. 384. It contains a McLeod manometer having four fused-in contacts
A - D . By means of the device E the mercury in this manometer is forced up at
intervals of 1 niinute. Control takes place through the contacts A and B. When the
pres~iiredetermined by contact D has been reached, this contact - via relays R,

Fig. 384
hfdpolder’s apparatus for pressure control in the range to torr
iranchembook.ir/edu

8.4 Reflux and rate of evaporation 463

and R2 - closes the electromagnetic valve S1leading to the vacuuni punip F. The
apparatus to be evacuated is connected a t G ; flask H serves as buffer for smoothing
out the variations in pressure. Cold trap I,cooled in liquid nitrogen, retains any con-
densable vapours. As soon as the required pressure is reached switch Swz is put over
to the position for “autoniatic control”.
The electronic vacuum controller of Fischer [27] operates on the principle of
t h e r i d conductivity in the range of 100 to torr (Fig. 385). The eensor is inserted
in the vacuum line by means of a ground joint. The switching for the vacuum control
can be continuously varied t o cover any measuring range. The output of the potentio-
meter amplifier is connected to a directly controlled magnetic valve which is placed in
the connection between the pump and the apparatus.

valve

Q
0 test

mains

-
Fig. 385
Electronic vacuum controller, type VKH

8.4 Reflux and rate of evaporation

8.4.1 Time-operated devices for reflux control

As stated in section 7.5, automatic column heads are usually made to function bj-
some timing device. The reflux ratio is then determined by the ratio of the period5
during which the device is switched on and off. The timing apparatus used for thih
purpose should allow any required ratio to be established. At low ratios, say 1 : 1 to
5: 1, the time during which distillate is taken off should be adjusted, for instance, to
1 sec; the corresponding periods of reflux return are then 1 to 5 sec. At higher reflux
ratios, say 10: 1 to 50: 1, the period of distillate take-off should be lo* enough for the
amount of liquid removed not to be greater than the quantity of enriched fraction
present a t the top of the column. It is therefore desirable to reduce the take-off
periods progressively and to increase the periods of reflux return corresponding]\ ,
particularly during the distillation of transition fractions.
These requirements are fulfilled by the aut,hor’s mechanical timing device [ 111.
It has a contact disc driven by a synchronous inotor; the adjustment of the desired
iranchembook.ir/edu

464 8. Automatic devices, measuring and control equipment

reflux ratio is done by means of a slider. The terminals are intended for connection to
an elerrtronic relay, which causes the electromagnet at the column head to function
at B potential of 220 V. As the reflux ratio is increased, the period of distillate take-off
is progressively reduced :
at a reflux ratio of 5 : 1 it is 2 sec.
at a reflux ratio of 10:1 it is 1 sec.
at a reflus ratio of 20: 1 it is 0.5 sec.
at a reflux ratio of 50: 1 it is 0.2 sec.
If the electronic relay is equipped with a switch to convert it from ‘‘norinally closed”
t o “normally open”, the reciprocal reflux ratios (e. g. 1:2, 1:5 etc.) can also be
obtained for coarse separations.
The contact disc can be replaced by others yieldmg different periods. A more
accurate control may, for instance, be provided in the range froni 20: 1 to 100: 1 ; on
the other hand the apparatus may be adapted for take-off times up to 30 sees, so
that it can be used in pilot-plant and technical distillations.
The timers functioning ent irelj- on electranic principles allow take-off and reflux
periods of 0.1 sec to 20 niin. to be adjusted with a n accuracy of 194 [56]. For an
alteration of the reflux ratio they require a change to be made in both the take-off
and reflux times. A number of electronic devices are commercially available most of
which are provided with bubble-point control. The coluiiin head is set at infinite
reflux ratio as soon as the pre-set temperature of the contact thermometer placed in
the oolumn head is reached. When the temperature falls below this threshold the pre-
selected reflux ratio is antonlaticall?- switched on again. -4s a n example, Fig. 386

4 3 5 6 7 8

-9

PtlOO Ptioo
contact thermometer 5 .

Fig. 386
Distillation control device ,Minitron 5
I = mains key, 2 = switch, 3 = isotherm follow-up control for heating jacket,
4 = temperature control for still pot, 5 = temperature control for heating jaoket,
6 = time decade for reflux, 7 = time decade for take-off, 8 = ammeter, 9 =
change-over switch for maximum column temperature
iranchembook.ir/edu

8.4 Reflux and rate of evaporation 465

shows the distillation control device Minitron 5 which may simultaneously serve as
timer and for flask and heating jacket temperature control (cf. chap. 8.2.2). The time-
switching component has 11 steps for take-off periods from 0.5 to 10 sec and 11 steps
for reflux periods from 1 to 100 sec, with the possible addition of a fixed period of
100 sec t o the reflux time chosen. This yields reflux ratios between 1 : 1 and 400: 1.
Further, a transistor relay can actuate a limit contact which automaticallp interrupts
the take-off indicating this by an acoustic or optical signal or switches off the appara-
tus altogether.
The accuracy of reflux control by means of electronic timers has been thoroughly
studied by Rock et al. [13]. Qeinmecker and Stage [57] have found that constant,
reproducible and load-independent values for reflux can be obtained only with elec-
tromagnetically controlled components. Deviations may be due to the following
effects.
The reflux ratio is larger than the time ratio if
condensation occurs below the divider so that condensate thus produced is not
handled by the divider ;
the condensate contained in the condenser does not completely flow to the divider
so that again a partial stream is not handled by the divider;
with vapour division the reflux condenser is accessible also during distillate take-
off.
The reflux ratio is smaller than the time ratio if
the dead volume of a magnetic valve always contains a sniall residue of liquid ;
vapours condense in the distillate take-off pipe ;
with vapour division the flow of liquid is obstructed in the condenser because the
condensate cannot leave the reflux condenser during distdlate take-off.

8.4.2 , Control df boil-up rate

We have seen that the theoretical plate number is strongly dependent on load i n
some columns. I n addition, if a column is operated near to flooding, the control of
the boil-up rate should be close enough not to allow flooding to take place. Two
methods of control are available :
1. that of keeping the heat input constant, especially in simple distillation and
2. control of the flask heater by the pressure drop in the column in countercurrent
distillation.
The method of supplying the electric flask heater with a constant current can be
unreliable since it involves frequent checking of the load, especially if there are large
differences in the heats of evaporation of the components being distilled. As the
pressure drop in a column depends on the vapour velocity (section 4.11), it provides a
convenient means for controlling the heater. The method also has the advantage of not
being seriously affected by mains fluctuations or by variations in the gas pressure [13].
Pig. 387 shows the arrangement employed in the Destinorin series of apparatus
[11]. The bottom reflux measuring device, w , carries a side-tube connected to a

30 Erell, Handbook
iranchembook.ir/edu

466 8. -4utomatic devices, measuring and control equipment

guard condenser b. In order t o prevent condensate from remaining suspended in the

aide-tube or the condenser itself, a small current of nitrogen may be blown in through
a bubbler, or, more simply, the tube below the condenser may be heated by a resis-
tance winding to mch a degree that flooding cannot occur. The condenser is connected
to a manometer c, provided with a contact wire d that can be adjusted during opera-
tion. The second lunb e of this manometer is open if the distillation is a t atmospheric
pressure; for vacuum distillation it. is connected to the vacuum lead at the coluinn

Fig. 385 Lr;" 1 -

Ar&ngement for controlling the load by the pressure drop in a column

n = hottom reflux measuring device, b = condenser, c = contact manometer.
d = adjustable contact, e = to atmosphere or column head, f = relay

head. The contact manometer (Fig. 388) contains mercury if i t is used with a mechani-
cal relay (section 8.2.2), or an organic liquid when employed with an electronic relay.
An organic liquid gives about ten times the sensitivity obtainable with mercury. By
ineans of a resistance the flask heater is first adjusted to the desired load, producing
a certain pressure drop; the wire d is then brought into contact with the manometer
liquid and the heating current is increased slightly. The relay connected t,o the heater
switches it off when contact is made. It is advisable not to switch the total current,
h i i t to keep about two-thirds of the heating capacity in circuit and to control only
the remaining third. The use of a delayed relay can be recommended, since otherwise,
if humping occim, every blimp will produce a contact. Some fornis of contact, rnano-
inet'er are made with a hinged limb (Fig. 389).provided with a millimetre scale and
angular graduations.
iranchembook.ir/edu

8.5 Measurement of physical data during distillation 467

9much more elegant method has been described by Stage [58].Elect<roniccontrol

is performed following either the t,emperature or a pre-set reflux rate in d r o p or
nil/min. The reflux is optically scanned and controls the proportional and cont inuoria
heat.ing. The still pot temperature is set according to the boiling point. of the bot,torii
product in questmion.The column load remains constant even after change of the
fraction. An arrangement for automatic reflux rat.io control dependent. on t.he bottom
teinperature in batch distillat,ion has been described by Fieberg [58a].

Y
Fig. 388 Fig. 389
Contact manometer (Krell) C o n h c t manometer with hinged limb
a = filling mark
b = G 1 fritted filter

8.5 Measurement of physical data during distillation

A correct boiling point is often not sufficient evidence for the purity of a compo-
nent and i t is therefore desirable to determine other properties during the course of
A distillation. This should preferably be done without removing distillate, and the
deterniinations must then be performed in the stream of reflux at the column head.
Such measurements will allow the conditions of distillation to be altered if necessary,
a t the proper moment.
Continuous determinations of the following properties [59] are possible in la hora-
tory and pilot plant columns :
melting point,
refractive index,
dielectric constant.

30*
iranchembook.ir/edu

468 8. Automatic devires. meaarving ana control equipment

With the last two, however, difficulties are experienced in keeping the liquid at a
constant temperature, since the distillate temperature normally rise8 during distilla-
tion.

8.5.1 Melting point

In the separation of high-boiling fatty acids Jantzen and Tiedcke used a n in-
genious device for determining the melting point of the distillate (Fig. 390). At a ,
the distillat,e or reflux drops int.0 a capillary tuhe cooled by a thermosyphon loop, as
in Thiele's melting point apparatus, and solidifies. In the first place this gives the

Fig. 3!M
Apparatus of Jantzen and Tiedcke for the determi-
nation of the melting point,of high-melting fractions

freezing point, as indicated by thermometer b. The liquid in the mean t,irne flows
through c into receiver d. The temperature of the thermosyphon is t,hen raised and
the melting point is observed. An automatic apparatus on the same lines, yielding
values with an accuracy of f0.25 deg. C, has been developed by Miilter and Zen-
chelsky ")].

8.5.2 Refractive index

The refractive index is widely employed in conjunction with the boiling point
for oharacterising organic aubstances [59]. In the separation of close-hoiling sub-
stances a continuous determination of the refractive index allows the conditions of
distillation to he so chosen that the transition fractions remain as small as possible
and yields are consequently increased.
P h d m w r i c refraetometera indicate their reading on the scale of an electric measur-
ing instrument and hence perniit this reading t o be recorded.
iranchembook.ir/edu

8.5 Measurement of physical data during distillation 469

Automatic and continuously recording flow refractometers have been developed

by Thomas et al. [61] and by Latchuni [62].
The model Remat 10 of VEB Carl Zeiss Jena 1631 is a differential flow refracto-
meter. It compares the refractive index of the flowing saniple with that of the re-
ference liquid which can be arbitrarily chosen. Fig. 391 shows the front and back of

Fig. 391
Flow refractometer Remat 10
a) front part with 2 troughs K 51 and 2 linear scales
b) back with cables for power supply
iranchembook.ir/edu

470 8. Automatic devices, measuring and control equipment

the device. I t works at temperatures in the range of -10 to +120°C and has
4 troughs with different measuring ranges.
Stage’s arrangement (Fg. 392) designed for measurements of phase equilibria,
appears useful. Two commercial flow refractonletem are so arranged in the equilibriilm
apparatus that they indicate the refractive indice8 of the liquid and vapoiir phases
and thus enable their compositions to he determined.

Pig. 302
Stage’s apparatus for phase equilibrium measurements with flow refractometers

Interferometers can also be employed as automatic indicators, in this caw of

refractive index differences. Kegeles and Sober [a] used an apparatus containing a
%-nim cell; with the green mercury line as light uource this gave readings accurat,eto
3 iinits in the 6th decimal place.

8.5.3 Dielectric constant

The increasing use of physical data in laboratory work has also led to developments
in the technique of determining the dielectric constant. This constant is an especial-
ly useful quantity when the niixtrire contains water (diel. const. 80) or other compo-
nents having widely different values. Exaniplcs are the mixtures acetic acid (diel
const 6.13)-acetic anhydride (22.2) and met hanol-toluene. In the latter 8ysteni the
azeotrope ha@a dielectric constant of 26.8, whilst methanol and toluene have values
of 33.8 and 2.37, respectively [a]. The dielectric constant has also proved convenient
for determining toluene in benzene, in spite of the fact t.hat the difference in the figures
for t.hese two compcments is only 0.08 iinits.
iranchembook.ir/edu

8.5 Measurement of physical data during distillation 47 1

Fig. 393 shows a n apparatus for determining the dielectric constant during the
coiirse of a distillation, as used by Oehme [65]. The measuring cell may of course also be
placed in the reflux tube of a column head, where it will indicate the proper moment
for altering the conditions of distillation. A book by Oehme contains a complete
collection of values of the dielectric constant, together with the frequency and
temperature a t which they were determined and the temperature coefficient.
In Slevogt’s dielectrometer, the “Multidekameter”, the frequency is not restricted
to a single value, but can be varied between 100 kilocycles and 12 megacycles [66].
Grant [67] has described a recording dielectrometer that appears suitable for distilla-

inlet for

liquid

measuring c d l

Fig. 393
Arrangement for measuring the
dielectric constant during distilla tioii
(Oehme)

tion. The limitations of the dielectric constant technique in organic cheiiiistry have
been studied by Felloni and Tantillo [68]. The determination of the quasistatic
dielectric constant as a parameter for the identity and purity of liquid raw materials
is reported by Lertes and Kid [68a].

8.5.4 Various deter mi nations

Laboratory and pilot plant work allows the same methods to be employed for the
continuous determination of chemical and physical properties of the distillates and
bottonis as work on a technical scale [2]. I n many cases, however, the measuring
arrangements will have to be modified appropriately.
The density balance due to Clasen [69] should be suitable for analytical work. It
requires 8 ml of substance and has an accuracy of g ~iii-S ~ .n arrangement for
the continuous measurement of liquid density has been reported by Linford [TO].
I n large-scale work, particularly in the petroleum industry, considerable use is
made of continuous methods for viscosity determination. Such a method might con-
ceivably be modified for distillation on a laboratory scale. Fritzsche [71] has given a
snrvey of viscosimeters for continuous on-line operation in oil refineries.
iranchembook.ir/edu

472 8. .4utomatic devices, meseurhg end control equipment

Other forms of continuous nieasuring devices utilized in chemical industry, suoh

as the infrared and dtraviokt spedrqraphs and the m a s spectrometer, might also
find an application in small-scale distillation, particularly for analytical work [72].
A review of modern continuous analytical methods has been published by Siggia
[73]. Of course, automatic analyzers have also come into use [74].
I n chap. 4.2 it was pointed out how the surface tension of the liquid phase may
influence the separating process in a countercurrent column. The Layda tensiometer
(DIN 53914, manufacturers: MeSgeriitewerk Ihuda) allows liquid interface tensions
to be measured and recorded automatically.
At the present tinie, radionzetric nzethodv are beconling of increasing iniportance
for following the course of a distillation. In the separation of multiconiponent mix-
tures of compounds having similar physical properties these methods are both
convenient and time-saving. To the initial mixture a certain amount of a labelled
compound is added and its concentration in the distillate fractions is deterniintd.
Hughes and Malocoy employed the conipound WH30H t o find the relationship
between pressure and composition in the system methanol, ethanol and water and
to observe the progress of the distillat>ion.
Zelvenskij et al. [75a] labelled low-boiling components with the radionuclides
C-1-4 and Cl-36 in order to measure diffusion and material transfer rates in falling-
film distillations. It should be possible to test columns by similar methods [75].

8.6 Measurement and metering of gases and liquids

l h o r a t o r y work with gases has been extensively discussed in the books of Bern-
hauer “761, Wittenberger [77] and Mbller and Gnauck [78] so that no details need be
given here.
The measurement and Rupply of gases and vapours are required in low-tempera-
ture distillation (cf. section 5.3.1), in carrier vapour distillation (section 6.1) and in
adsorptive distillation (section 6.3). A neutral atmosphere (usually nitrogen) is often
necessary in the distillation of inflaiumable, oxidizable or hygroscopic material,
and here again gas volumes niay have to be measured.
Methods for the indirect iueasurement of the quantities of saturated or super-
heated water vapour have already been described in section 6.1. To determine the
flow rates of gases or liquids under pressure, rotary flow meters (“Rotanieter”, a
trade name), containing a float that is kept in rotation and suspension in a tapered
tube, are now widely employed. Whilst bubble counters and capillary flow meters
have to be calibrated against a wet gas meter, conimercial “Rotameters” are generally
suyplicd already calibrated for given gases or liquids. The forms of flow meters com-
nionly used for laboratory and semi-technical work are shown in Fig. 394. Especially
for pilot plant distillations, they niay also be designed for pneumatic or inductive
data tranainission. For checking the amounts of cooling water employed in distilla-
tion, devices of the type SW 16.1 (Fig. 395) have proved convenient. The tapered
glass tubes contain floats of a variety of materials, depending on the properties of the
iranchembook.ir/edu

8.6 Measurement and metering of gases and liquids 473

a) b) C)
-
Fin. 394
“ROta” flow meters
a ) wit,h olive for tubing
b ) with olive for tubing and with jacket
c) with standard ground joints
d) with armature

Fig. 395
Laboratory flow meter, type S W 16.1
Measuring range I: 0.5 to 3.0 I/min;
range 11: 3 . 0 to 12.0 l/min
iranchembook.ir/edu

474 8. Antomatic devices, measuring and control equipment

flnitls. An automatic cooling water control has been described b;* Hinzpeter [79].
There are also flow- nionitors for liquid or gaseous media comniercidy available
which switch off the power supply when the water supply is interniptd or is less than
the pre-set value and switch it on antomaticaUy when the water supply is restored
[801.

5-

Pig. 396 Fig. 397

Diagram of dropping funnel Stage’e automatic feed control
I = valve screw, 2 = opening,
3 = heating jacket, 4 = ball joint
valve, 5 = screw-cap, 6 = valve
spindle, 7 = drop-forming t.ip
881

I n continuous distillation the wual practice is to place a storage vessel for the
feedstock at such a height that the resulting hydrostatic pressure is greater than the
pressure in the column, and II steady supply is ensured. If an ordinary glam stopcock is
iised for control, it will have to be adjusted frequently, since the level in the storage
wsfirl falls continually and a corresponding reduction in the feed rate would other-
wise result. Although it, is possible to control the rate accurately by the use of R
needle valve and capillary drop counter. a constant-pressure (Mariotte) bottle
(Fig. 224) is preferable. When necessary this bottle can be refilled froni a large vessel
hp air pressure from a small hand blower.
The dropping funnel of Darnmer and Kramer [81] (Fig.396) employs ~b combina-
tion ( i f the constant-pressure principle and valve control. The modified Mariotte
iranchembook.ir/edu

8.6 Measurement and metering of gases and liquids 455

bottle due to Hanel [82] with 6 interchangeable capillaries allows low dropping rates
to be set quite exactly.
The various fornis of Lahodest automatic apparatus have been provided by
Stage with a feed arrangement containing a magnetic valve, controlled by means of
a timer (Fig. 597). The period during which the valve is open can be varied iwtwem
1 and 120scc and the period of closure between 1 and 2 sec. Whenever the valve

capacity (water) +

Fig. 398 Fig. 399

Jena circulation primp. type 100 Characteristics for the Jena circulation priinp 1 ~ 0

opens a defrnitcl qiiantity of liquid flows into the nieasuring device helow and drains
away during the period of closrire. The feed vessel and its iininersion tube function as
a constant-pressure bottle as long as the liqnid level is above the lower end of t h t b

iiumersion tube.
Pinkava and Wichterle [83] have described a pneumatically operated glass
valve with which, for instance, a water sitpply can be regulated in the 0.02-3500 nd/h
range with an accriracy of f1.5O/,. The book of Pinkava [84]deals thoroiighly with
metering probletns encountered in the laboratory.
An appliance that. finds inany nses in the laboratory is the Jena Circulation Pimp
100 (Fig. 398), a centrifugal pump of t,he simplest type constmcted entirely of glass.
Its lift and capacity as a function of the number of revohitions per minute of the
iranchembook.ir/edu

476 8. $utomatic devices, measuring and control equipment

motor are given in the diagram of Fig. 399. The electric spring bellows pump manu-
factured by Messrs. Haage serves for the continuously variable metering of gases and
liquids. Its lift may amount to 20 m water column. The standard models, which have
no stuffing boxes, are designed for capacities of 16 nil/h to 1,500 1/h and made of
stainless steel, tomhac-brass or plastic material (Fig. 400).
A special group is formed by peristaltic pumps with tubes made of various
materials. They are mainly used for metering low-boiling liquids, emulsions, sludges,

Fig. 4.00
Electric spring bellows metering pump for liquids and gases

Pig. 401
Peristaltic pump, type DP 2-2, with 3 different speeds
Capacity: 3 to 2400 ml/h
iranchembook.ir/edu

8.6 Measurement and metering of gases and liquids is7

gases and suspended solids [86]. Depending on the tube diameter (approximately
3 to 10 mm) and the revolving speed of the rollers amounts of a few ml/h to about
20 l/h may be handled in the temperature range of -90 to f150"C. As an exaniple
Fig. 401 shows the weakly pulsating peristaltic pump of the type DP 2-2 which may
be employed as pressure or suction pump and has 16 separate tubes. A contribution
to the quantitative description of the pulsation of peristaltic pumps has been made
by Engler and Kuhnlenz [86a].
Plunger pumps have proved satisfactory for larger plants with continuous
operation. Metering pump heads of glass with PTFE bellows were already mentioned
in chap. 5.1.3.1 (Fig. 143) [88]. Circulation pumps made of glass are put to a variety
of uses for a capacity range up to 20 m3/h and lifts up to 15 m (Fig. 402). Metering
pumps which may also be employed for the automatic mixing of various liquid
streanis (Fig. 403)in lahorntory and pilot plant distillation [85] are manufactured b!.

Fig. 402
Glass-made circulation pump with Teflon gasket

Fig. 403
Precision metering device
Normados P 31
iranchembook.ir/edu

478 8. Automatic devices, mdasuring and control equipment

I

Messrs. Bran & Liibbe,Hamburg. Their Normados K and J series meet the require-
ments of laboratory and semi-technical work, the capacities ranging from 0.8 to
1,121 l/h. The machines can be controlled manually, electrically or pneumatidy.
Ginm the module system is strictly adhered to the components may be combined in
many ways. The type Gnomodos is intended for small capacities (0.6 to 4.4 l/h) and
is provided with a knurled flap for the fine adjustment of the stroke.
The model GnomodofiS 02 has several metering pump heads either with plungers
or diaphragms. The drive is designed for strokes from 0 to 15 mm. Up to 5pump heads
can be operated simultaneonsly.
iranchembook.ir/edu

9. Arrangement of a distillation laboratory,

starting up distillation apparatus

9.1 Lay-out

Points to be considered in designing a new laboratory are dealt with by Behre [l],
Coleman [2] and Schramni [3], while the arrangement of a laboratory for analytical
distillation i 4 gone into by Kincannon and Baker [4].
9 distillation laboratory should have long, high wall surfaces for supporting the
frameworks for apparatus. Examples of satisfactory arrangements in various spaces
are given in Figs. 404 to 407. The location of pilot plants is illustrated in Figs. 140. 141,
146 and 160. There is a tendency for the manufacturers of distillation apparatus to run
test centres for work on problems of inaterial and heat transfer. The testing apparatus
is controlled froin a measuring room. An associated 18boratory is designed for coinple-
mentary distillation experiments and the necessary analytical work on samples [21].

corridor
Fig. 404
Arrangement of a distillation laboratory (long type) with a well-ventilated roo111
and a separate pump room
A = Laboratory bench, ;M = Framework, N A = Framework in fume c u p b o ~ r d .
E = Laboratory bench

Lay-out I , Fig. 404

For the reduction of pump noise in the actual laboratory the vacuum pullips are
placed in a separate pump room. The punips P are close to the wall between this I'OQI~I
and the laboratory, SO as to make the vacuum lines as short as possible. The other
side of this wall is provided with a framework M and is reserved for vacuum distilla-
tion. automatic pressure controllers and vacuum gauges are placed on or behind the
framework where they are accessible and can easily be watched. The pump room can
also serve as a small workshop.
The window side of the laboratory is occupied b;v benches A with water and
electricity supplies. This bench is intended for physical determinations, such as
iranchembook.ir/edu

480 9. Arrangement of a distillation laboratory. starting up distillation apparatus

refractive index, solidification point, melting point, dielectric constant and optical
rotation. The centralbench E is reserved for chemical work. The main wall adjoining
the passage is also provided with a framework. For the distillation of unpleasant or
poiaonous compounds such as ether, benzene, chlorinated hydrocarbons or organic
nitro-compounds, a separate space is provided. It may be regarded as a large futne
cupboard which can beentered; a 40-fold renewal of its air every hour ensures suffi-
cient ventilation. The screen separating it from the laboratory is of wire-reinforced
glaes, so that the apparatus can be supervised from the laboratory. If automatic
apparatus is used in this chamber the measuring and control equipment should be
mounted on the wall outside it.

windows

Fig. 406
lEkL
corridor
cupboaid
t
Arrangement of a distilletion laboratory (square type) with a
well-ventilated room and a separate pump room
(for legend, see Fig. 404)
I
I
1
I
I
iranchembook.ir/edu

9.1 La.y-out i s1
Lapout I I , Fig. 405
The principle of this arrangement is the same as that of Fig. 404. 4 s the longest
free surfacw are the dividing walls, the frameworks are attached to them. The
mclosed, ventilated rooin is again intended for the distillation of noxions substances.
'Chr wntral laboratory bench here starts from the wall adjoining the passage and can
b e fitted at its end with additional stands for small apparatus (Fig. 406).

windows

corridor
Fig. 407
Srrangement of a distillation laboratory (short type) with a separate
pump room a id high fume cupboards
(for legend, s' e Figs. 404 and 405)

1 Fin. 408
Fu&e cupboard containing framework
- for apparatus

31 Krell, Handbook
iranchembook.ir/edu

482 9. Arrangement of a distillation laboratory, starting up distillation apparatus

Lug-out III, F ig . 407

ThiR laboratory, which is intended for distillations involving more extensive
chemical work, is equipped with two large central benches. lnstead of the ventilated
room, it only has two fume cupboards MA containing frames. One of these cupboards
extends from the floor up to the ceiling (Fig. 408); the second starts from table level,
but also contains st,ands for small apparatus. The air speed a t the openings of the
cupboards is nmintained at 1-2 ftlsec.

9.2 Frameworks and services

Frameworks to carry distillation equipment can either be built with sets of parts
or he Immanently attached to the walls. A framework which may he used to build
any structure demanded by the apparatus, is shown in Fig. 409. Unit frames mounted
on castors (Fig. 410) are obtainable, and have the advantage that t h e apparatus may
be used anywhere in the laboratory. For extensive apparatus it is preferable to use
frameworks fixed to the walls. Jordal rails (Fig. 411 a) which are embedded in the walls
are designed to take up special clarnps (Fig.4 l l b ) at any desired height [ 5 ] . By the
use of cross-bars of various lengths the framework can be made to match the ap-
paratus (Fig. 408).
The arrangenient used by the author is illustrated in Fig. 412. The. framework
consists of ititerconnected unitsof 1 metre width, cemented into the wall; the wall is
tiled to a height of Bfeet. On the floor below this franiework there is a brickwork
trough, 2 ft. 6 inches in width, to catch any liquid that may he spilt by breakage. This
trough will also localize a fire.

Fig. 409 Fig. 410

Framew-ork for components Frame on castors
iranchembook.ir/edu

9.2 Frameworks and services 483

Fig. 41 1
Jordal rails
a) rail with holes for insertion of
special rlamps
b) special clamp

2 and 3
phase sockets

Fig. 412
Framework unit with
brickwork trough, piping
and electric panel

31*
iranchembook.ir/edu

484 9. Arrangement of a distillation laboratory, starting up distillation apparatus

The piping for water, gas, compressed air, vacuum and steam should be laid along
the walls, together with a drain. Bench instruments may be placed on shelves behind
the fmieworli. Electric power is taken froni panels; the most convenient position for
these is hetween each pair of frames (Fig. 412). The panels have 8 to 10 sockets, rach
with a switch and a neon laiiip, so that it can be seen at a glance which parts of the
apparatus are in operation.
-A main switch should be installed to cut off the whole electrical supply except
lighting 111 the event of danger.

Fig. 413
Slotted angle iron for the conet,ructioii
of frames

Narrow sliding ladders, which can be kept in a corner hehind the framework
when not in use, are employed for reaching all parts of the equipment.
The wntilation should be so designed that fresh air enters along the floor behind
the framework and the waste air leaves near the ceiling. Ten changes of air per hoiir is
generally considered sufficient. If an appreciable quantity of gas is produced in a
distillation it may be led awa? through the upper ventilation holes by tubes connect-
ed to condensers.
Semi-technical dktillrrtim~u p a r a t w should be provided with specially constructed
frames and platforms (Fig. 424), which may if desired be built up from unit parts.
Slotted angle iron (Fig. 413) greatly facilitates the construction.

9.3 Building up the apparatus

Before the equipment is assembled i t is advisable, especially if the apparatus is
complicated, to draw a sketch of it and, as a start, to arrange the components on
the floor according to this plan. Next, the corresponding ground joints are examined
iranchembook.ir/edu

9.3 Building up the apparatus 44.5

by putting them together without grease and testing them for fit. By moistening the
cone of the joint or drawing a circle of chalk on it, one can easily see whether irre-
gularities are present. Loosely-fitting joints should be replaced. In vacuuiii distillation
this procedure saves much time which would otherwise he spent in testing for lealib.
Assembling of the parts should start from the bottom. In the firqt place it shoiild
he possible to remove the still pot without dismantling other components. For thl.
purpose flask heaters with a vertical adjustment (Fig. 327) can be used. Plain retort
rings do not always give the flask a firin seating. In Tome cases the suspension of the
flask in a spring cradle (Fig. 414)offers a solution.
Thecolumn should bemounted exactly vertically by means of a plummet. The lateral
parts of the apparatus must not be connected to the main, vertical parts before these
have been tested for vacuum-tightness. Testing for leaks should, in fact, always he
performed step by step, as this method is by far the most rapid. Ground joints should

Fig. 414 Fig. 415 Fig. 416

Spring cradle for flask Ground joint with metal Clip for ball joint
clamps and springs

Fig. 417 Fig. 418

Clip for plane joints Clip for ground joints (Edward & Co. London)
iranchembook.ir/edu

486 9. Arrangement of a diatillation laboratory, starting up distillation apparatns

be secured by hooks and springs. Small joints usually have fused-on hooks; larger
joints may be frtstened with metal clamps (Fig. 416). A clip for securing ball join& is
shown in Fig. 416 and one for plane joints in Fig.417. The clip for ground joints
which is manufactured by- Messrs. W. Edwards & Co, London, (Fig. 418) is easily
detachable.

Fig. 419 Fig. 420

Boae. 14 mm Rubber joint collars

Fig. 421
“Specco” hook-clamp

Rubber collars can be used for fastening ground joints and are manufactured for
the sizes NS 12 to NS 29 (Fig. 420); they can be employed at temperatures up to
130°C [6]. Their main advantages are that the use of joint grease is unnecessary and
that qtrains are unifornily distributed around the circumference.
Bosses must be carefully machined so that, when the screws are tightened, the
object connected is not forced out of position. If this happens, strains will ensue
which may lead to the breakage of glass parts, usually after they have been h e a t d in
the course of distillation. A reliable form of boss is shown in Fig. 419. The “Specco”
hook clamp (Fig. 421) is intended specially for the construction of frameworks. The
double three-point bearing ensures a tight hold free from movement. The standardized
Gcru system consisting of a variety of components allows the constriiction of appa-
ratus r a q y g from simple devices to large pilot plantn [7].
For supporting components that have to be adjusted exactly to a certain level,
such au f h k heaters, levelling bottles etc., the variable height support due to Mann-
chen (Fig.422)is convenient, as it allows the position to be regulated to within a
ndlimrtre. Ring clamps (Q. 423) are better than the customary jaw clamps for
iranchembook.ir/edu

9.3 Building up the apparatus 48i

Fig. 422
Variable height support (Mannchen-Riedel)

F1 a

' Fig. 423 Fig. 424

Ring clamps Framework for a 200 1 distillation plant.
Galvanized tubes with galvanized
coupling sleeves
iranchembook.ir/edu

488 9. Arrangement of a distillation laboratory. starting up distillation apparatus

holding columns and other cylindrical parts. A systematic survey of supports for and
the suspension of, chemical apparatus has been given by Schwarz [ 5 ] .
Pilot plant and semi-technical distillation apparatus made of glass requires
special connections, clamps and frameworks. It may suffice here to refer the reader
to the catalogues of the glassware manufacturers. Some firms even send crews for
the aasemblv of apparatus. As a n example Fig. 424 shows a framework of galvanized
tubes joined by means of socket,s with Iinhus screws.

9.4 Sealing ground joints

In most cases the problem of ensuring that ground joints and taps are leak-
proof needs attent.ion. The substances to be distilled have widely different solvent
powers for joint greases, so that preliminary tests may be necessary to find the most
suitable lubricant. Some data on the subject are given in Table 79.
The experience obtained with silicone greases is not yet sufficient to allow a defi-
nite judgment. There is some danger that coniponents of the grease will enter the
column and affect the wetting of the packing, and it is difficult to remove the white,
powdery filler which remains when the silicone itself has been leached out.
,4s a general rule it can be said that taps and stop-cocks should be lubricated with
n softer grease than ground joints. Lithelen (Table 79) cont8ainslithium soaps and
undergoes little change in consistency with temperature. It, can therefore be used for
greasing taps at temperatures from below 0" up to 150°C. A review on the preparation
of special laboratory greases has been compiled by Brooke [8].The various lubricants

Table 79
Greases for ground j0int.s and taps

Type of grease Maximum Composition or manufacturer Application

operating
temp. ("C)

Ramsay greaae 25 6 pts. vaseline General work

1 pt. paraffin wax
6 pts. Para rubber
High vacuumgreaae P 25 E. Leybold's Xachf., Cologne, High-vacuum work
High vaouum grease R 30 Germany
Lit helm 150
kipsenberg grease 1 pt. soluble starch For powerful greave
1 pt. glycerin solvents. such as
diethylether or
BZ 6364; BZ 6563 E. Merck, Darmstadt, Germany chloroparaffins
Sirnpy phosphoric acid
Graphite As graphite powder or a mir- High-temperatnre
ture of graphite and paraffin distillation
wax, or as colloidal graphite
suspension in wat,er
iranchembook.ir/edu

9.4 Sealing ground joints 489

have been extensively discussed by Wagner [9]. Bernhauer [ 101, Wittenbergcr [ 111.
Wolf [12] and Friedrichs [13] have given detailed instructions for the handling of
ground glass joints and glass taps, with particular consideration of how ground
joints that have seized should be treated.
I n molecular distillation it is essential to use a thoroughly degassed lubricant
with a low vapour pressure, for instance the high-vacuum greases P and R. Ground
joints with a mercury seal (Fig. 425), though providing a n entirely tight seal, should
be employed only in exceptional cases, owing to the poisonous nature of mercury : the
cup is better filled with vacuum oil or grease.
An ungreased joint - even one having the finest ground surfaces - will never be
entirely vacuum-tight or, conversely, will not prevent the escape of some vapour.
Greasing is carried out by coating the upper half of the cone with a uniform layer of

n
u with mercury seal
Fig.425
Joint

the lubricant in question. This is best done with a smooth wooden spatula; greasing
with the finger gives a uniform coating, but has the disadvantage that dirt or moisture
may be introduced into the lubricant. Another method, particularly suitable for
Ramsay or Apiezon grease, is to warm the cone gently over a hot-plate (not an open
gas flame) until the lubricant flows, after which i t is distributed by rotating the cone.
It is then inserted into the socket and pressed in with a turning motion until the
joint appears uniform and transparent. If this does not happen at once the process
should be repeat<edafter rewarming the joint.
If a joint starts to leak during a distdlation owing to solution or decomposition
of the lubricant, it, may be possible to seal the leak by applying a suitable cement
around the edge. A grade of picein - melting point 80" or 105°C - is most used.
Sealing wax may he employed as a substitute; its disadvantage is that it tends to
become brittle. A cement that can be used up to 250°C is a mixture of glycerin and
litharge, which sets hard. Picein and sealing wax should not be melted in a gas flame;
the proper method is to heat the place'to be sealed with a Bunsen burner to about
lOO"C,and then to apply and spread out the cement by means of a spatula heated to
the same temperature. Finally a small flame is passed over the surface for a short
time so as to make the cement flow and penetrate into the crevices of the joint.
A different approach for ground joints is the use of Teflon sleeves. This gives the
joints a tight fit and does not allow them to seize. The Xrench firin Gbchot supply
32 Krell, Handbook
iranchembook.ir/edu

4!90 9. Arrangement of a distillation laboratory, aterting up distillation apparatus

those pieces, called “Rodets”, in sizes suited for the commonly used joints. In chap. 3.1
it wtm pointed out that precision ground joints may ale0 do without grease. In pilot
and semi-technical plants Teflon rings and sleeves are preferably used for sealing
ball and flat joints.

9.5 Starting up distillation apparatus

When the apparatus haa been assembled the chasge is put into the still pot. It
should be filled to at most two-thirds of ite nominatkvohune, or to one-half in vamum
distillation. Boiling chips are added to the contents. The volume of charge should
be measured at the same temperature 88 the distillate fractions to be collected later,
and this temperature may be relatively high if the fractions are solid at room tempera-
ture. Poisonous substances and those with a high vapour pressure are hest forced by
gw pressure from a bottle into the still pot, which should then immediately be
Htoppered. In order to emure good wetting of the packing it is customary to pour
the charge in through the condenser if possible. If it is essential for the apparatus
to be completely free from moisture it can be dried by blowing in hot air with the
jacket heater in operation.
Stress should again be laid on the importance of a thorough cleaning of the
packing material before it is put into the column. It is not sufficient to rinse it a few
times, say with carbon tetrachloride or trichloroethylene. The adhering layer can be
removed by a treatment with hot benzene (poisonous; use fume cupboard) followed
by a rinse with trichloroethylene [14].
For vacuum distillation, a preliniinary test for leakage is necessary. The apparatus
is evacuated; when the required pressure has been reached the vacuum pump is
fitopped and the mte of increaae of pressure is observed (section 5.4.1). A further test
for tightness is performed when the column is in operation and the thermal load has
been established. The still pot heater is switched on only when the desired pressure is
attained, and the column jacket heater as soon aa vapour evolution starts. The cool-
ing water should be turned on before the heaters are put into operation; during the
initial period its rate of flow should be checked at intervals since it usually tends to
vary somewhat at the start with the expansion of tap washers. When the vapour has
reached the colmm head the load is adjusted to the required value (section 4.10.7)
unless the column is to be flooded (section 4.10.8) to wet the packing. The apparatus
ifi then allowed to run for at least half an hour at total reflux, so as to allow the
initial h o w p i n t to be accurately determined, and only then is it adjusted to the
required reflux ratio (section 4.10.4).
As mentioned before, a few boiling c h i p are added to the still pot in order to
promote regular ebullition. Generally speaking the use of boiling capillaries cannot
he recommended, since their presence makes it difficult to determine the effective
distillation pressure and air sometimes gives rise to polymerization. The “boiling
chips” may be small pieces of earthenware (flower pots) or pumice, platinum tetra-
hedra and in some cmes even hits of wood. For alkaline materials, tin dust may be
employed. It should be noted that boiling stones lose their activity if a vacuum
iranchembook.ir/edu

9.6 Safety measures 491

distillation is interrupted and air is admitted. Their action is connected with the
small bubbles of air clinging to their surface; if the vacuum is interrupted it is there-
fore necessary to add fresh boiling chips [15].
If foaming occurs to a troublesome extent a t the beginning of distillation, a n
attempt may be made to reduce it by the addition of an anti-foaming agent [16].
Silicone oils have recently been used with success for this purpose. If the material to
he distilled is highly viscous it is advisable to keep the still pot contents in constant.
motion with a magnetic stirrer or the stirrer due to Hubner [17].

9.6 Safety measures

I n laboratory practice it is frequently necessary to distil combustible, poisonoils

or explosive substances; in vacuum distillation there is the danger that vessels may
collapse: I t is therefore essential to observe certain safety measures [ 181. Data on
combustible gases and vapours which are of importance from the point of view of
laboratory safety have been collected by Dittmar [19]. In every case it shonld be
decided whether or not an explosion-proof electric installation is required.
Very poisolzous niaterinls should always be handled and distilled in the well-
ventilated room or in a fume cupboard. Explosive substances such as ether should be
distilled only in a room especially designed for such operations unless work in the
open air is preferred. If the charge contains poisonous gases in solution, the conden-
ser should be connected to a tube leading to the atmosphere or to a fume cupboard.
When distilling substances which injure the skin the operator should wear rubber
gloves and goggles since it is difficult to avoid contact with traces of such materials;
it may also be advisable to wear protective clothing.
Vacuum distillations require special precautions. Safety spectacles or, rather, a
face shield should be worn for work at reduced pressures. Heaters give sonie protection
from flask and column. In addition, glass-wool sheaths rnay be applied to the other
parts for insulation and protection. If this is not possible for experimental reasons,
for instance because the separation process should be visible throughout its course
a wire gauze or plane glass screen must be placed in front of the apparatus.
Bnffer vessels can be fitted into wooden boxes lined with glass wool or completely
encased in a wire gauze cage. If the vacuum pumps have a belt transmission, the
latter miist he provided with an effective guard. The safety measures that are neses-
sary in a high-pressure laboratory and in work with toxic substances are considered
in detail by Craig and Dew 1201.
iranchembook.ir/edu

References

Chapter 1 - Introduction
[l] Rerhenberg, C. V., Einfache untl fraktionierw Destillation in Theorie und Praxis.
Schimmel u. CO. AG, Miltitz 1923
[Z] Young, S., Distillation Principles and Practice, MacMillan, London 1922
[3] Thormann, K., Destillieren und Rektifizieren, Verlag Spamer, Leipzig 1928
[4] Badger, W. L., and McCabe, W. L.. Elementa of Chemical Engineering, RlcGraw-Hill,
New York and London 1911
[ti] McCabe, W. L., and Thiele, E. W.,Ind. Eng.Chem. li (1925) 605
[6] Jantzen, E.. Dechema Monogrtiph, Vol. 6 (1932) No. 48; Angew. Chem. 80 (1923).
692
[7] KirRrhbaum, E., Distillation and Rectification, Chemical Publishing Co. Inc., New
l'ork, 1948
[8] Gyula, N.,L e p C l L (Distillat~ion),Muszaki Konyvkiad6, Budapest 1955
[9] Jacobs, J., Deatillier- und Rektifizieranlagen, R. Oldenbourg-Verlag, bliinchen 1950
[lo] Robinson, C. S., and Gilliland, E. R., Elements of Fract,ional Distillation, RIcGraw-
Hill. New York 1950
[lOa] Perry. J . H., Chemical Engineers Handbook, McGraw-Hill, New York, 3rd Ed. 1953
[ll] Ychultze, G. R., and Stage, H., Die Kolonnendestillation im Laboratoriuni. VDI-
Forschungeheft 424, VDI-Verlag, Berlin 1944
[123 Thormann, K., Dechema-Erfahrungsaustausch, Laboratoriumstechnik, Arbeits-
methoden rind &rate. Destillieren und Rektifizieren, Dechema, Frankfurt/Main.
September 1950
[13] Thormann, K., .Dechema-Erfahrungsaristnusch,Techn. Apparat'e und Anlagen, Trennen
fliissiger M i d u n g e n , Destillieren. Betriebstechnik, Dechema, Frankfurt/Main, August
1952
[14] Stage, H., and Schultze. G . R.., Die grundlegenden Arbeiten uber Theorie, Apparate
sowie Verfahren der Destillation und R,ektifikat.ion, VDI-Verlng, Berlin 1944
[I51 Rose, A. and E., Distillation Literature Index and Abstrarts 1941-1945, 1946-1952,
Applied Science Laboratories. Stil te College, Pennsylvania 1953
[16] ZIC Referstedienst Chemiache Verhhrenstechnik, ed. by Zentralstelle fur Information
der chemischen Induetrie, Berlin 1972ff.
Fortschritte der Verfahrensterhnik, ed. by Ingenieurwissenschsftliche Abteilung der
Farbenfsbriken Bayer BG, Leverkusen, Verltlg Chemie GmbH, Weinheim/Bergstr.
1952ff.; from Vol. 9 (1971ff.) ed. by Verfiihrenstechnische Gesellschaft im VDI, VDI-
Verlag, Diiseeldorf
Verfahrenstechnische Berichte (Vorankundigungs.blattm, Karteibliitter und halbjilhr-
liche Ssch-, Autoren- und Firmenregister), ed. by IngenieurwissenschaftlicheAbteilung
der Farbenfabriken Bayer AG, Leverkusen. Verlag Chemie GmbH, Weinheim/Bergst,r.
[I?] Walsh, T. 6..Ind. Eng. Chem. from 1956
[181 Orlicek, A. G., and Poll, H., Eilfaburh fur YineralCiltechniker, Vol. I: Die Eigenschsften
von Kohlenansserstoffen, Minernlolprod~ikteii und Hilfsstoffen, Springer-Vcrlag,
Vienna. 1951
ClH] Sigwart, K..Destillation und Rektifikntion, in: Ullmanns Encyklopadie der technischcn
Chemie, Vol. 1, Urban & Schwarzenherg, Miinchen/Berlin 1951, pp. 429-470
[20 I Kortum, G., and Buchholz-Meisenheimer, H.. Die Theorie der Destillation und Extrilk-
tion von Flussigkeiten, Springer-\'erlag. Berlin/Gottingen/Heidelberg1952
iranchembook.ir/edu

References 493

Bosnjakovi6, Fr., Technische Thermodynamik, Vol. 1 (6th Ed.) a,nd Vol. 2 (5t,h Ed.).
Theodor Steinkopff Verlag, Dresden 1972/1971
Gattermann, L., and Wicland, H., Die Praxis des organischen Cheniikers. 41st Ed..
Walter de Gruyter & Co., Berlin 1962
Wittenberger, W., Cheniische Laboratoriumstechnik, 6th Ed., Springer-Verlug, Vimna
New York 1963
Weygand, C., and Hilgetag, G., Organisch-chemische Experimentierkunst, 4th Ed..
.Johann Ambrosius Barth, Leipzig 1970
Bernhauer, K., Einfiihrung in die organisch-chemische Laboritt>oriumsterhnik,Springcr-
Verlag, Vienna 1947
Carney, Th. P., Laboratory Fract,ional Distillation, MacMillan, Xew Pork 1949
Rosengart., M. J., Die Technik der Destillation und Rektifikation irn Laborat,orium.
VEB Verlag Technik, Berlin 1954 (translated from Russian)
Rose, A. and E., e t al., Technique of Organic Chemistry, Vol. IV, Distillation, Inter-
science Publishers, New York 1951
Weber, U. yon, Gewinnung atherkcher Ole durch Destillation, in: Gildemeister, E.,
and Hoffmann, Fr., Die Ltherischen Ole (revised by W. Treibs), Bkadcmie-Verlag,
Berlin 1966
Gigwart,, K., Destillieren nnd Rektifizieren, in : Honben-Weyl, Methoden der orga-
nisohen Chemie (ed. by E. Muller), Vol. I / l , Georg Thienie Verlag, Stuttgart 19.58
Zuiderweg, F. J., Laboratory Manual of Batch Distillation, Interscience Publishers.
New York 1957
Coulson, E. A., and Herington, E. F. G., Laboratory Distilhtion Practice, Georg
Newnes, London 1958
Bukala, N., Majewski, J., and Rodzinski, W., Fraktionierte Destillation in] Laborato-
rium, Pailstwowe Wydawnictwo Naukowe, Warsaw 1958
Industrial and Engineering Chemistry, Chemical and Process Engineering, \*Dl-
Zeitschrift, Chemische Tecbnik
Hoppe, K., Kiinne, H., und Bendix, H., Fortschritte in der Destillation. Rektifikntion,
Sorption und Extraktion, VEB Deutscher Verlag fiir Grundstoffindustrie, Leipzig 1970
Treybal, E. R., Masstransfer Operations (International Student Edition). 2nd Ed.,
McGraw-Hill, New York 1968
GeiseIer, G., e t al., Ausgewahlte physikalische Met,hoden der organisclien Chemie.
Akademie-Verlag, Berlin 1963
Keil, B., et al., Laboratoriumstechnik der organischen Chemie (ed. by H. Viirst).
Skademie-Verlag, Berlin 1961
Wiberg, K. B., Laboratory Technique in Organic Chemistry, McGraw-Hill, New Yolk
1960
Organikum - Organisch-Chemisches Grundpraktikum, 15th Ed., VEB Deutscher
Verlag der Wissenschaften, Berlin 1977
Anorganiknm - Lehr- und Praktikumsbuch der anorgnnischen Chemie (ed. by L. Kol-
ditz), 8th Ed., VEB Deutscher Verlag der Wissenschaft,en, Berlin 1978
Rock, H., Destillation im Laboratorium - Extraktive und azeotrope Destillat ion.
Dr. Dietrich Steinkopff Verlag, Darmstadt 1960
Stage, H., and Bose, K., Die Belastungsverhiiltnisse in Fullkijrpersaulen unter Destilla-
tionsbedingnngen, Springer-Verlag, Berlin/Gottingen/Heidelberg 1962
Miiller, E., and Stage, H., Experimentelle Vermessung von Dampf-Fliissigkeits-Phasen-
gleichgewichten, Springer-Verlag, Berlin/Gottingen/Heidelberg 1961
HdIa, E., Pick, J., Fried, V., and Vilim, O., Vapour-Liquid Equilibrium, 2nd Ed.?
Pergamon Press Ltd., London 1968
HQla, E., Wichterle, I., Polsk, J., and Boublik, T., Vnpour-Liquid Equilibrium Data a t
Normal Pressures, Pergamon Press Ltd., London 1969
Wichterle, F., Linek, J., and HQla, E., Vapour-Liquid Data Bibliography. Elserier
Scientific Publishing Company, Amsterdam/Oxford/New York 1973, Supplement I.
1976; Supplement, 11, 1979

33 Krell, Handbook
iranchembook.ir/edu

494 References

[42] Cf. R.eferences Chap. 4 [lo51

Ogorodnikov. C. K., Lesteva, T. X.,and Kogan, V. B., Handbook of azeotropic mix-
t.ures, Khimiya, Leningrad 1971 (in Russian)
[43] Billet, R., Grundlagen der t.hermischen Fliiasigkeitazerlegung, Bibliogrtbphisches Insti-
tut, Mannheim 1961
Billet, R.. Verdampfer-Technik, Bibliographisches Institut, 3fannheim 1965
[43a] Billet, R., lndostrielle Destillation, Verlag Chemie GmbR, Weinheim/Bergstr. 1973
[44] Billet., R., Optimierung in der Rektifiziertechnik, unter besonderer Berucksichtigung
der Vakoumrektifikation, Bibliographisches Institut, Mannheim 1967
[45] Olirer. E. D., Diffusional Separation Processes: Theory, DeRign and Evaluation, John
Wiley and Sons. Sew York 1966
[46] Pratt, H. R. C., Countercurrent Separation Processes, Elsevier Publishing Company,
Amsterdam 1967
[47] Krell, E., Frey, H.-P.. Gawalek, G., and Werner, G., Einfuhrnng in die Trennverfahren,
VEB Deutscher Verlag fur Grundstoffindustrie, Leipzig 1975
[48] Hoppe, I<., and Mittelst.rass, hi., Grundlagen der Dimensionierung von Kolonnenboden,
Theodor Steinkopff Verlag, Dresden 1967
[49] Jungnickel, H., and Otto, E., Tiefe Temperaturen in der Verfahrenstechnik (Vol. 11 of
the series ..Beitrage zur Verfahrenstechnik"), VEB Deutscher Verlag fur Grundstoff-
industrie, Leipzig 1972
[.W]Frank, W..and Kutachc. D.. Die schonende Destillation, Otto Krausskopf-Verlag,
GmbH, Mainz 1969
[a1 I Olevskiy, V. M., and Rnchinskiy, V. R., Rektificntion of temperature-sensitive sub-
stances, Khirniya, Moscow 1972
[ 5 2 ] Ho116. J., Kururz. E., and Borbdi, 4 . , The Applications of Molecular Distillation, Akad6-
miai Kiad6, Budapest 1971
[53] Schubertli, H., Thermodynamische Grundlagen der Destillation und Extraktion, Vol. I,
VEB Deutscher Verlag der Wiesenschaften, Berlin 1972
[54] Brauer, H., and Mewes, D., Stoffaustauech, einschl. chemischer Reaktionen, Verlag
Sauerlander AG, Aarau and Frai&furt/Main 1971
Brauer, H.. Urundlagen der Einphasen- und Mehrphasenstromung, Verlag Sauerlander
AG. Awau and Frankfurt/Main 1971
[Xi] Holland, C. D., Multicomponent Distillation, Prentice-Hall International, Englewood
Cliffs 1963
[56] Chemie-1ng.-Tech. 42 (1970) 14, 946-957; Chem. Techn. 22 (1970), 117-120
[56a] Third International Symposium on Distillation 1979, I. Chem. Symposium Series 56.
1043 page8
[ 5 i ] Billet, R.. CZ-Chemie-Techn. 3 (1974) 10, 353-361
[58] Sharov. V. T., and Serafimov, L. A., Physico-chemical Fundamentals of Distilhtion
and Rectification, Khimiya, Leningrad 1975
[.%I Hirata, M., Ohe, S., and Naga.hama, K., Computer aided data book of vapor-liquid
equilibria, Elsevier Scientific Company, Tokyo/Amsterdam/Oxford/New York 1975 (cf.
aim Chem.-1ng.-Tech. 48 (1976) i , 634
[W] Fratzscher, \V.. Picht, H. P., and Bittrich, H. J., Chem. Techn. 29 (1977) 7, 361
to 367
[Sl] Stichlmair, J.. Grundlagen der Dimensionierung des Gas/Flussigkeit-Kontaktapparates
Bodenkolonne, Verlag Chemie, Weinheim 1979
[a21 Sattler, K.,Thermische Trennverfahren, Vogel-Verlag, Wiirzburg 1977
[62a] Kafaror. V. V.. Grundlagen der Stoffiibertragung, tra.nslated from Russian, revised and
edited by Hsrtmann, K. and R.eher, E. O., Akademie-Verlag, Berlin 1977
[ti31 Wiinsch. G., Haufe, S. and Kopel, R., Chem. Techn. 31 (1979) 6, 286-291
[ti41 Kaibel. G., Mayer, H.-H. and Seid, B., Chem.-1ng.-Techn. 60 (1978) 8, 586-592
[65J Winter/Merte/Herz, Recent Development in Boiling and Condensation, Verlag Chemie,
\Veinheim, New York, 1977
iranchembook.ir/edu

References 495

[66] Gmehling, J. and Onken, U., Vapour-Liquid-EquilibriumData Collection, in Chemistry

Data Series, DECHEMA, Vol. T
Part 1: Bqueous Organic Systems,
2a: Organic Hydroxy Compounds: Alcohols (1978)
2 b : Gmehling, J., Onken, U. and Arlt, W.
Organic Hydroxy Compounds: Alcohols and Phenols (1978)
3+4: Aldehydes and Ketones (1979)
Gmehling, J. and Onken, U., Chem.-1ng.-Techn. 61 (1979) 2, 127
[67] WeiB: S., Chem. Techn. 31 (1979) 9, 459-463

Chapter 2 - A Review of the History of Laboratory Distillation

Schelenz, H., Zur Geschichte der pharmazeutisch-chemischen Destilliergerat*e. Beilage
zum Bericht April 1911, Verlag Schimmel und Co. AG, Miltitz
Gildemeist,er, E., Die Ltherischen Ole, Verlag Schimmel und Co. AG, Niltitz 1928
Rechenberg, C. von, Einfache und fraktionierte Destillation in Theorie und Praxis,
Verlag Schimmel und Co. AG, Miltitz 1923
Underwood, A. J. V., Trans. Inst,n. chem. Engr. 13 (1935), 34
Forbes, R. J., Short history of the a r t of distillation, Verlag E. J. Bull, Leiden 1948
Forbes, R. J., and Dijksterhois, E. J., ,4 History of Science and Technology, Vol. I:
Ancient Times t o bhe Seventeenth Century; Vol. 11: The Eight,eenth and Nineteenth
Centuries. Pelican Books No. A 498 and A 499, Penguin Books Ltd., Harmondsworth
Middlesex 1963
Schneider, W., Lexikon alchemiatisch-pharmazeutischer Symbole, Verlag Chemie
GmbH, Weinheim/Bergstr. 1962
Lippmann, E. 0. von, Entstehung und Ausbreitung der Alchemie, Vols. 1 and 2. Verlag
Julius Springer, Berlin 1919, 1931; Vol. 3, Verlag Chemie GmbH, ~ ~ e i n h e i m / n c r , s t r .
1953
Lippmann, E. 0. von, Abhandlungen und Vortrage zur Geschichte der Naturwissen-
schaften 1 (1906), 2 (1913), Verlag Chemie GmbH, Weinheim/Bergst,r.
Lippmann, E. 0. von, Beitrage zur Geschichte der gatura.issenschafter1 nnd der Tech-
nik 1 (1923), 2 (1953), Verlag Chemie GmbH, Weinheim/Bergstr.
Stage, F., Angew. Chem., Ed. B 19 (1947), 181
Kreis, H., Liebigs Ann. Chem. %24 (1884), 259
Young, S., J. Chem. Soc. [London] 75 (1889), 679
Friedrichs, J., Z. angew. Chem. 32 (1919), 340
Glas-Instrumenten-Techn. 14 (1970), 196
Krell, E., Glas-Instrumenten-Techn. 11 (1967) 1-5, Beilage 1-20
Strube, W., Der historische Weg der Chemie, VEB Deutscher Verlag fiir Glrundstoff-
industrie, Leipzig 1976
Bittel. A, Chem.-1ng.-Techn. 31 (1959) 6, 365-378

Chapter 3 - Standardization and Data on Concentrations

[l] DDR-Standards und Fachbereichsstandards der DDR, Ed. : Amt fur Standardisieruug
der Deutschen Demokratischen Republik, Berlin 1968
[2] Verzeichnis der DIN-Normblatter, Stand vom 28. 2. 72, Beuth-Vertrieb GmbH, Berlin
(West)
[ 3 ] Telle, W., Chemische Laboratoriumsgerate, 2nd Ed., VEB Deutscher Verlag fur Grund-
stoffindustrie, Leipzig 1969
[4] Friedrichs, F., Die Glashutte 76 (1949), 3-6
[ 5 ] Fliedner, K., Glasapparatetechnik 5 (1958), 33-37; Suppl. of Chem. Techn. 10 (1968)
[6] Kuehnhanss, G., et al., Chem. Techn. I) (1957) 407-409
[7] TGL 10192-10200, 11641-11644, 12023-12032, 20277, DIN drafts 28800-228808

33"
iranchembook.ir/edu

496 References

[70] Hampel, B., Chemiker-Ztg. 96 (1971) 7, 316-325

[S] Krell, E., Chem. Techn. 8 (1954), 452-455
[9] Padelt, E., and Laporte, H., Einheiten und GroSenarten der Naturwissenschaften, 2nd
Ed., VEB Fachbuchverlag, Leipzig 1967
Fischer, R.. Padelt, E.,and Schindler, H., Physikalisch-technische Einheiten - richtig
angewendet, 7nth Ed., VEB Verhg Teohnik, Berlin 1979
[lo] Foratsr, H., Einheiten, GroOen, Gleichungen und ihre praktische Auwendung, VEB,
Fachbuchverlag, Leipzig 1968
[Ill Haeder, W.,and Gartner, E., Die gesetzlichen Einheiten in der Technik, Verlag Beuth-
Vertrieb, Berlin 1970
[Is] Krell, E., Chem. Techn. 7 (1965). 30-32
[13] Lessels, A. G., Chem. Engng. 88 (1956), 208-210
[14] Orlicek, A. F., Poll, H., and Walenda, H., Hilfsbuch fiir Mineraloltechniker, Vol. 1:
Die Eigenschaften von Kohlenwasserstoffen, Mineralolprodokten und Hilfsstoffen.
Springer-Verlag, Vienna 1951
[151 Baehr, H., Chemie-1ng.-Techn. 23 (1953), 676
[lS] Kramer. M. T.. Glas-Instrumenten-Techn. 16 (1971), 251 254; 17 (1973), 43-48,
~

137- 139
[17] Spiith, W.,Chem.-hg.-Techn. 50 (1978) 6, 446-448
Spath, W., Zahl-MaS-Bild, A. W. Gentner Verlag KG, St.ut.tgart 1977

Chapter 4 - Physical Fundamentals of the Separation Process

111 Sigwart. K., Deatillation und Rektifikntion, in: Ullmanns Encyklopadie der techni-
schen Chemie, Vol. 1, Urbnn A Schwarzenberg. Miinchen/Berlin 1951, p. 429/470
[2] TGL21120
[31 Kuhn, W., Kelv. chim. Bcta 25 (1942), 2152-295
Kuhn, W.. and Ryffel, K.,Helv. chim. -4cta 26 (1943), 1693-1721
[3a] Niselson, L. -4., Tretyakova, K. V.. and Yamko, M. A., Zh. prikl. chim. 16 (1973) 7,
1613- 1614
[3b] Graczyk, J., Nukleonika 1 (1973) 1, 25-33
[a] Mcier, W.,and Huber, M., Chemie-1ng.-Techn. 89 (1969), 797-800
[4a] Bemer, G. C . and Zuiderweg, F. J., Chem. Engng. Sci. 38 (1978) 12, 1637-11643
[4b]Farid, 51. M. and Gunn, D. J., Chem. Engng. Sci. 34 (1979) 4, 579-592
[5] Huber, M., and Hiltenbrunner, R., Chem. Engng. Sci. 21 (1966), 819-832
[6] Knfarov. V. V., Dorokhov, J. N..and Shestopsvlov, V. V., Hydrodynamic Parameters
of Packed Columns, Proceedings, Part 4, Distillation 1969, Brighton/Englmd, Ab-
stract: Chemis-1ng.-Techn. 43 (1970) 14, 953
[7] Jonas, Chr., Chern. Techn. 86 (1974) 12, 87-91
[HI Zuiderweg, F. J., and Harmens, A., Chem. Engng. Sci. 9 (1958) 2/3, 89-1103 ,
Medinn, A. G., McDermott, C. and Ashton. N.,Chem. Engng. Sci. 33 (1978) 11 1489 to
1493
[Sti] Oomez, R. A., and Strumillo, Cz., Iniynieria cheniiczna 2 (1975) 6, 267-281
[8b] Bauer. H., Chemie-1ng.-Techn. 48 (1976) 10, 891
[8c] Ponter, A. B., Trauffler, P.. and Vijayan, S., I & E C Process Des. Devel. 15 1976) 1,
196--199
[Sl Krell, E., Chem. Techn. 16 (1964) 10, 591-600
[gal Kolev, N., Chem.-1ng.-Techn. 48 (1976) 12, 1105-1112
[lo] Titov, A. A., and Zelvenekiy, Ya. D., khim, prom. (1974) 9, 690-693
1111 Krell, E., Plaste und Kautachuk 14 (1967) 4, 2151-254
[lla] Kwasniak, J., Verfahrenstechnik 8 (1974) 11, 315-322
[ll b] Yilmaz, T., and Brauer, H., Chemie-1ng.-Techn. 46 (1973) 14, 928-934
[ l l c ] Kiinne, El.-J.,Chem.,Techn. 36 (1973) 9, 533-538
[ I l d ] Dietz, H., Reinemann, G., and Comel, M., Chem. Techn. 25 (1973) 3, 143-148
[11 el Godau, H.-S.. Chem. Techn. 06 (1974) 9, 661-584
iranchembook.ir/edu

~
References 497

[ i l f ] Billet, R., CZ-Chem.-Techn. 3 (1974), 9-12

[ I l g l Adolphi, H.-V., Chem. Techn. 28 (1974) 10, 843-645
[11 h] Antonov, Y. E., Safin, R. Sh., and Reut, V. J., teoret. osnovy khim. tekhnol. 9 (1975)
1, 140-142
[ l l i ] Grabbert, G., Winsch, G., StraBberger, L., m d Meyer, B., Chem. Techn. 28 (1974) 8.
457 -458
[ i l k ] lMarschall, E., Chem.-1ng.-Techn. 47 (1975), 879
[ill] Wiinsch, G., Chem. Techn. 27 (1975) 12, 717-723
[ l l m ] Krell, E., and Heinrich, W.,Chem. Techn. 29 (1977) 11, 615-618
[ll n] Heinrich, W., Thesis TH ,,Carl Schorlemmer", Leuna-Merseburg 1975
[ l l o ] WeiW, S., and Srhmidt, E., Cheni. Techn. 26 (1974), 476
[11 p] Schmidt, E., Thesis TH ,,Carl Schorlemmer", Leuna-Merseburg 1975
[11 q] Vorontsov, E. G., 6. of Appl. Chem. USSR, N. Y.51 (1978) 4, Pt. 1, 754-759
[12) Staudinger, H., Anleitung zur organischen qualitativen Analyse, 6t'h Ed., Springer-
Verlag, Berlin/Gottingen/Heidelberg1955, pp. 12-13
[13] Stage, H.! Erdol und Kohle 3 (1950), 478-490
[141 Hildebrand, J. H., and Rotariu, S. J., Analytic. Chem. 24 (1952), 770-773
[1.5] Mauser, H., and Kortum, G., Z. Naturforsch. 10a/4 (1955), 317-322
[16] Mannchen, W., Einfuhrung in die Thermodynamik der Mischphasen, VEB Deutscher
Verlag fur Grundstoffindustrie, Leipzig 1965
1171 Schuberth, H., Thermodynamische Grundlagen der Destillation und Extra.ktion, Vol. 1,
VEB Deutscher Verlag der Wissenschaften, Berlin 1972
[17a] Bit,trich, H.-J., Gratzscher, W., Lempe, D. and Picht, H.-P., Chem. Techn. 31 (1979) 9.
488 -472
[ I 81 Dreisbach, R. R., Pressure-Volume-Temperature Rehtionship of Organic Compounds.
3 r d Ed., Handbook Publ., Inc., Sandursky, Ohio 1952
I191 Dreisbach, R . R., Physical Properties of Chemical Substances, The Dow Chemical
Company, Midland, Michigan 1951 - 1953
1201 Stull, D. R., Ind. Engng. Chem. 39 (1947), 517/550 and 1684
1211 Stage, H., Fette u. Seifen 53 (1951), 677-682
[ 2 2 ] Stage, H., Fette u. Seifen 55 (1953), 217-224
[23] Stage, H., Miiller, E., and Faldix, P., Erdol u. Kohle 6 (1953), 375-380
r24) Stage, H., Miiller, E., and Faldix, P., Chem. Techn. 6 (1954), 31/36
[25] Jordan, T. E., Vapor Pressure of Organic Compounds, 2nd Ed., Interscience Publishers
New York 1954
LP(i) Riddick, J. A., and Toops jr.. E. E., Organic Solvents, Physical Properties and Methods
of Purifickztion, Interscience Publishers, New York 1955
[27] Sage, B. H., and Lacey, W.X., Some Properties of the Lighter Hydrocarbons, Hydro-
gen Sulfide and Carbondioxide, American Petroleum Institute, New York 1955
1281 Ghser, F., and Ruland, H., Chemie-1ng.-Techn. 29 (1957), 772-775
1291 Jaeckel, R., Destillation und Sublimation in Fein- und Hochvakuum, in: Houben-Weyl,
Methoden der organischen Chemie (ed. by Eugen Miiller), Vol. I/1, Georg Thieme Verlag,
Stuttgart 1958
[30] Kienitz, H., Bestimmung des Dampfdruckes, in : Houben-Weyl, Met.hoden der organi-
schen Chemie (ed. by Eugen Miiller), Vol. IIIj1, Georg Thieme Verlag, Stuttgart 1958,
pp. 255-324
[31] Milazzo, G., Chemie-1ng.-Techn. 28 (1956), 646-654
[32] Swiqtoshwski, W., Ebulliometric Measurements, Reinhold, New York 1945
Swietoslawski, W., and Anderson, J. R., in: A. Weissberger: Physical Methods of
Organic Chemistry, Vol. I, Interscience Publishers, New York 1949, p. 107
r33] Weber, U. von, J. prakt. Chem. 1/4 (1955), 318-323
[34] Eblin, P., J. chem. Educst. 27 (1950), 67
[35] Herrington, E. F., and Martin, J. F., Trans. Faraday SOC.49/2 (1953), 154-162
[36] Natelson, S., and Zuckernianu, J. L., Ind. Engng. Chem., analyt. Edit. 1 7 (1945).
739 740
-
iranchembook.ir/edu

498 References
. ~~~~~

[37] Gould, C.. jr.. Holtmann, G., and Niemann, C., Analyt. Chem. 19 (1947), 204-206
1381 Wiberley, F. J., Siegfriedt, R. K., and Benedetti-Pichler. A. A., Mikrochem. verein.
Miomhim. Acta 118 (1951). 417-475
[39] Neameyanor. A. N., Atomic Energy (Russisn) 3 (1957), 1013-1025, Kernenergie 1
(1868), 579-587
[a]Skliarenko, S. I-.,J. phys. Chem. (Russian) 33 (1968), 1916- 1921
[all Henning, F.,and Stock, A., Z. Physik 6 (1921), 226
[a]Smith and Menzies, 2. physik. Chem. 76 (1911). 501
[43] Encken. A.. and Suhrmann, R., Phpikalisch-chemische Praktikumsaufgaben, 7th Ed.,
Akademische Verlagsgesellschaft Geest t Portig KG, Leipzig 1968. pp. 100-106
[U] Koftiim, C., Moegling, D.. Woerner, F., Chemie-1ng.-Techn. 82 (1950), 453-457
[45) Sohliinder, E. U.. Chemie-1ng.-Techn. 85 (1963) 7, 482-487
[46] Riick. H..Z. physik. Chem., N. F. 4 (1955), 242-245
[47] Pbter, hi., Hagpar KBmiai Folybirat 10 (1953). 295-300, Abstract in: Chem. Techn. 6
(1954), 307
[47a] Kolllr, 0.. and Nagy, E., Chem. Techn. 88 (1971) 1, 35-36
[47b] Smith. G.. &‘innick, J.. Abrams. D. S., and Prausnitz, J. M., Canad. J . chem. Engng. 54
(1976), 337-343
La81 Pearson, I). E.. J. ohem. Educat. SU (1951), 60-62
[49] English. W. D., b. Amer, chem. SOC. 74 (1952), 2927-2928
[50] Diihring. U.. Wiedem. Ann. 11 (1880), 163; 68 (1894), 556
[51] Rechenberg. G. von. Einfache und fraktionierte Destillation in Theorie und Praxis.
Verlag Schimmel Rr Co. AG, Mikitz 1923
[521 Jantzen. E., and Erdmann. R., Fet,te 11. Seifen 54 (1952), 197-201
1531 Khejev. V. A., J. appl. Chem. (Russian) i (1934)
1.541 -2ntaine. C., C. R. hebd. S b n c e s Aced. Sci. 107 (1888), 681, 883. 1143/45
[561 Thomson, G. W., Chem. Rev. 88 (1946), 1-39
1561 Kiedel, L.. Chemie-hg.-Techn. 86 (1954). 83-89
[K’] Reckhnrd, H., Erdol u. Kohle 116 (1958): 234-241; Chemiker-Ztg. S4 (1960), 33-36
[581 Wichterle. I., and Linek, 5.. Die Antoinschen Dampfdruck-Konsttlriten reiner Verbin-
dungen, Vcrlag Academia, Prague 1971 (in English)
[591 Berechnung thermodynemischer Stoffwerte von Gasen rind Fliissigkeiten, VEB
Deutscher Verlag fiir.Grundstoffindust,rie, Leipzig 1966, pp. 77-89
[MI Othmer, D. F.,and Yu. E.-S.. Ind. Engng. Chem. 60 (1968) 1, 22-36
[Ma] Zip. T., and Thodos, G., Canad. ,J. chem. Engng. 62 (1974) 5. 630-635
[Mb]Cmmez-Nieto, M. and Thodos. G., ATChEJ. 28 (1977) 6. 904-913
[ ~ O C Zia,
] T. and Thodos. G., Chem. Engng. J. 16 (1978), 41 -49
[mdl Gomez-Nieto. M. s n d Thodos,G., Ind. and Engng. Chom., Fundam. 17 (1978) 1,45-51
[6Ocl Weetmeier. S.and Picht. H.-P., Chem. Techn. 31 (1979) 2, 86--88
[6(lf] Waseda, J., Jacob, K. T.. Tsuchiyn, P. and Tarnaki, S.. Z. f. Xat. forsnh. TIA 33 (1978)
8,940-945
La11 Schwab, G. M., Z. f. angea. Chem. 4.3 (1930). 155-156
1621 Stange, R.. Chem. Techn. 5 (1953). 78-79,
[63] Othmer, D. F.,Maurer. P. W., Molinary, C. J., Kowalski, R. C., Ind. Engnp. Chem. 49
(1957). 125-137
[64] Wittenberger, W., Chemiker-Ztg. 76 (1952), 371 -373
[65] Begholm. A., Svensk kem. Tidskr. 60 (1948), 96
Hoffmann, W.,and Florin. F., Z. Ver. dtsch. Ing., Beih. Verftthrenstechn.. (1943).
47-51
[ti?] Cox, E. R., Ind. Engng. Chem. 15 (1932). 592-593
[68] Davis, D. S., Tnd. Engng. Chem. 17 (1925), 435
I691 Calingart, G., and Davis. D. S., Ind. Engng. Cheiii. 17 (1925), 1287
[TO] Fe~.ldix,P., and Stage, H., Chem. Techn. 6 (1953). 526-528, 718-719
[71] Stage, H.,end Jiiilfs, J., Chemiker-Ztg. 77 (1953), 511,538,575; 7.S (1954). 43. 112. 182.
217
iranchembook.ir/edu

References 499

[72] Thormann. K., Destillieren und Rektifizieren, Spamer-Verlag, Leipzig 1928

[7R] ?rIatz, W.,Die Therinodynamik des Warme- und Stoffaustausches in der Verfahrens-
technik, Vols. 1 and 2. Dr. Dietrich Steinkopff Verlag, Darmstadt 1949/1953
[74] Kortiim, Q., and Bnchholz-Meisenheimer, H., Die Theorie der Destillation und Ex-
traktion von Fliissigkeiten, Springer-Verlag, Berlin/Got’t,ingen/Heidelberg 195-3
[75] Bosnjakovii, Fr., Terhnische Thermodynamik, Part 11, 5t,h Ed., Verlag Tlieodor~Stein-
kopff, Dresden 1971
[76] Florinl F.. Brennxtoff-Chem. 33 (1952), 161 - 166
[77] YrCahe, W. L., m d Thiele, E. W.,Ind. Engng. Chem., ind. Ed. li (1925), 603
[7X] HQla, E.. Pick, J.. Fried, V., and Vilim, O., (2nd English Ed.:) Vaponr-Liquid K q u i l i -
brinm. Pergamon Press, London 1968
Wichterle, F., et al., Yapour-Liquid Data Bibliography, Elsevier Scientific Publishing
Companv, Amsterdam/Oxford/New York 1973, Snpplement I, 1976
[79] N. N., Chemie-1ng.-Techn. 11 (1969) 12, 735-736
[SO] Wilson, G. M., J. Amer, chem. SOC.86 (1964), 127
[Xi] Donth, E. J., and Weichmann, R., Chem. Techn. 10 (1968) 2, 93-95
[82] Krug, K., Haberland, D., and Bitt,rich, H.-J., Chem. Techn. 23 (1951) 7. 410-415
[S3] Renon. H., and Prausnitz, 5. M., American Instit,ute of Chemical Engineers Jouim+l 14
(1968), 135
[XSa] Onken, U., and Gmehling, J., Chemie-1ng.-Techn. 49 (1977) 5: 404-411
[83b] Wichterle, I., Fluid Phase Equilibria 1 (1977) 2, 161-172
[R3c] P r a u s n h , J. M., Ber. Bunsenges. phys. Chem. 81 (1977) 10, 900-908
Oellrich, L., Plocker, U.. Prausnitz, 5. M., and Knapp, H., Chem.-Iii~.-?’erhii.49
(1977) 12, 955-965
Plocker, U., Knapp, H., and Prausnitz, J. M., Ind. and Engng. Chem., Process l k s .
Dev. 17 (1978) 3, 324-332
Anderson, T. F., and Prausnitz, J. M., Ind. and Engng. Chem., PIocess Des. Dc,r. 1i
(1978) 4, 552-561
[ G d ] Drahos, J., Wichterle, I., and Hala, E., Fluid Phase Equilibria 1 (1977/78) 3. 1 7 3 - 1 8 4
Wichterle, I., Fluid Phase Equilibria. 3. (1977/78) 3, 225-245
[ U e ] Onken. IT.,Chem.-In&.-Techn. 50 (1978) 10, 760-766
[X3f] Bittrich, H.-J., Chem. Techn. 31 (1979) 2, 104-105
[S3g] Hauthal, H.. Chem. Techn. 31 (1979) I, 37-41
[n4] Rosengart. W.J., Die Technik der Destillation nnd R?ektifikation nn Laboratoiirim.
VEB Verlag Technik, Berlin 1954 (Transl. from Russian)
[SS] Orlicek, A. F., and Poll, H., Hilfsbiich fur Mineralbltechiiiker. 1701s. 1 and I t . Springer-
Verlag, Vienna 1951/1955
[SS] Voorwijli, J . F., Chem. Engng. 63 (1956), 201 -202
[X7] Hollo, J., and Lengyel, T., Ind. Engng. Chem. 51 (1959).957-960; Collect. Czeclioslor.
chem. Comm. 23 (1958). 1735-1743
Hollo, J., Periodica polytechn. [Budapest]; chem. Ingenieurwes. I(1958). 112 -- 1%:
Fette, Seifen, Anstrichmittel 62 (1960) 10, 913-918
[X8] Umstat>ter.H., and Flaschka, H., Erdol 11. Kohle 1 (1948), 161 - 157
[S9] Williams, R,. A.. and Henley, E. J., Chem. Engng. 5.1 (1970) 2. 145-151
[X9a] Hal& E., Chem. Techn. 26 (1974) 8, 492-493
[X9b] Lu, B. C.-J., and Polak. J., Advances Cryogenic Engng. 10 (1975) F-1. 2(t4-?1-7
[90] Melpolder, F. W., and Headington, C. E., Ind. Engng. Chem. 89 (1947), i K L 7 W
Rose, A , , and Biles, R..,
Chem. Engng. Progr. 61 (1955). 138-140
[Yl] Rose. E., Ind. Engng. Cheni. 53 (1941), 596
[Ye] Hawkins, J. E., and Brent, 5.A., Ind. Engng. Chem. 43 (1951): 2611-2621
[93] Zypkina, 0. Yn., J. appl. Chem. (Russian) 46 (1955), 185-192
[94] Kortiim. G.. Chemiker-Ztg. 74 (l950), 151-154
[95] Carlson, H. C., and Colburn, A . P., Ind. Engng. Chem. 31 (1942). 581
[961 Orlicek, A. F., Osterr. Chemiker-Ztg. 50 (1949), 86
iranchembook.ir/edu

500 ~
References

[97] Laar, J. 6. van, and Lorenz, R., Z. anorg. Chem. 146 (1925), 239; Z. phys. Chem. 72
(1910), 723; 83 (1913), 599
[SS] Margules, M.,Wiener Sitz. Ber. 104 (1895), 1243
[98a] Grewer, Th.. and Schmidt, A., Chemie-1ng.-Techn. 45 (1973) 17, 1063-1066
[99] Landolt, H., and Bdm~tein,R., Physikalisch-chemische Tabellen, 6th Ed., Springer-
Verlag, Berlin/~ott.ingen/Heidelberg 1950/55
[IOU] Perry, J. H., Chemical Engineers' Handbook, 4th Ed., McGraw-Hill Book Company,
Nea York 1963, pp. 1942ff.
(1011 YII Chin, Distillation Equilibrium Data, 2nd Ed., Reinhold, New York 1956
[102] Chu Wang, Levy and Paul, Vapour Liquid Equilibrium Data, Edwards, T. W., Ann
Arbor, Mich. 1956
[lo31 Kirschbaum, E., Deatillier- und Rektifiziert.echnik, 4t.h Ed., Springer-Verlag, Berlin/
Heidelberg/New York 1969
[103a] HBbner, W., and Schliinder, E. U., Chemie-Lng.-Techn. 45 (1973) 5, 247-253
[103b] Vogelpobl, A., and Ceretto, R., Chemie-1ng.-Techn. 44 (1972) 15, 936-938
[lo41 .Jacobs, J., Destillier-Rektifizier-Anlagen,R. Oldenbourg Verlag, Miinchen 1960
[lo61 Kogan, V. B., and Fridmann, V. F., Handbuch der Dampf-Fliissigkeih-Gleichgenichte,
VEB Deutacher Verlag der Wissenschaften, Berlin 1960 (Transl. from Russian)
[I061 Hal&, E., Wichterle, I., Polik. J., and Boublfk, T., Vapour-Liquid Equilibrium Data
a t Xormal Pressure, Pergamon Press, Loadon 1969
[106a] The lnstit,ution of Chemical Engineers, London, runs a data service, PPDS (Physical
Property Data Service), which covere also liquid-vapour equilibrium data
[lo71 Kortiim, G., and Freier, H.-J., Chemie-1ng.-Techn. 26 (1964), 670-675
[lo81 Weber. L!. von, Z. phys. Chem., N.F. 38 (1963) 3/4, 129-139
[lo91 Gelbin, D., Chem. Techn. 16 (1963) 1, 6--9
[110] Stage, H., Miiller, E., and Gemmeker, L., ('hemiker-Ztg./Chem. Apparatur 86 (196l),
11
[ l l l ] Othmer. D. F., Ind. Engng. C'hem.. ind. Edit. '10 (1928), 743
[112] Stage, H., and Fischer, W.G., Glas-Instrumenten-Techn. 12 (1968) 11, 1167- 1173
[113] Gillespie, D. C.. Ind. Engng. Chem., analyt. Edit. 18 (1946), 57.5
[114] Rock, H., and Sieg, L., Z. phyeik. Chem., N.F. 3 (1955), 355-364; Chemie-1ng.-
Techn. 2S (1956), 133
[115] Kortum, C . , and Biedersee, H. von, Chemie-1ng.-Techn. 42 (1970) 8, 552-560
[lie] Schmidt, R..,Werner, G., and Schuberth, H., Z. phys. Chem. 212 (1969), 381-390
[116a] Rafflenbeul, L., and Hartmann, H., Chem. Techn. 7 (1978) 4, 145- 148
[117] Wicht,erle, I., and Hal& E., Ind. Engng. Chem. Fundamentals 2 (1963) 2, 155-157
[118] Neumann, A., and Walch, W., Chemie-Ing.-Techn. 10 (1968) 5, 241-244
[118a] St.age, H., and Fischer, W. G., Verfahrenstechnik i (1973) 6, 1 - 3
[IlSb] Schmidt,. A. P., Chem.-Lng.-Techn. 50 (1978), 537-538
[119] Kumarkrishna Rao, V. N., Swami, D. R., and Sarnsingnrao, M., Amer. Ind. chcni.
Eng. 3 (1957), 191- 197
[I201 Jost, W.,Rock, H., Schrijder, W.,Sieg, W., and Wagner. H.G., Z. physik. Cheni.,
N.F. 1 0 (1957), 133-136;
Schroder, W., Chemie-hg.-Tecbn. 30 (1958), 523-525
[110a] Slocum, E. W., I L EC Fundam. 14 (1976), 126-128
[IZOb] Granso, L.. and Fredenslund, 9..Ber. Bunsenges. phjs. Chem. 81 (1977) 10, 1088 t o
1093
[120c Granse. L., F d e n s l u n d , A., and Mollerup, J . , Fluid Phase Equilibria 1(1977), 13-20
r 1211 Lydersen, A. L., and Hammer, E. A., Chem. Engng. Sci. 7 (1958), 241 -245
[122] Junghsns, W., and Weber, U.von, J. prakt. Chem. 4/2 (1955), 265-273
[ 1231 Mai, K. L., and Bebb, A. L., Ind. Engng. Chem. 47 (1955). 1749 -1757
[I241 Zelvenskiy, Y.D., and Shalygin, V. A., J . phys. Chem. (Russian) X X S I , 7 (1957).
1501 -1509
j114a Trybuta, St., and Bandrowski, J., Iniyneria chemiczna 6 (1975). 679-695
[ 1251 Redlich, O., and Kister, A. T., Ind. Engng. Chem. PO (1948), 341-348
iranchembook.ir/edu

References 501

[lee] Herington, E. F. G., Nature pondon] 161 (1947),610-611

Hamer, D. F. O., e t al., Ind. Engng. Chem. 42 (1950),120
Haase, R.,and Jost, W., Z. phys. Chem., N.F. 9 (1956),300-301
Tierney, J. W., Ind. Engng. Chem. 50 (1958), 707-710
[127] Herington, E . F. G., 5. appl. Chem. 2 (1952),11-23
[128] Bittrich, H.-J., Chem. Techn. 14 (1962)9, 527-533
[l291 Tao, L. C., Ind. Engng. Chem., Fundamentals 1 (1962)2, 119-123
[130] Kolling, H., Chemie-1ng.-Techn. 24 (1952),405-411
[131] Bowman, J. R., and Cichelli, M. T., Ind. Engng. Chem. 41 (1949),1985
[132] Junge, C., Chem. Techn. 8 (1956),579-588
[133] Anschiitz, R., Chem. Techn. 9 (1957),516-519
[134] Fisher, G. T., Ind. Engng. Chem., Process Des. and Devel. 2 (1963),284-288
1 1357 Billet,, R., Chemie-1ng.-Techn. 30 (1958),513-518
[136] Thuni, O . , Chemie-1ng.-Techn. 29 (1957),675-678
[137] Neretnicks, I., Ericson, I., and Eriksson, S., Brit. chem. Engng. 14 (1969) 12, 1711 - 1712
[138] Weber, U. von, Chem. Techn. 2 (1950),241-246
[1391 Vogelpohl, A., Chemie-1ng.-Techn. 43 (1971)20,1116-1121
[139a] Domenech, S., Guiglion, C., and Enjalbert, M., Chem. Engng. Sci. 29 (1974)7, 1519 t o
1528,1529-1535
[140] Billet, R.,Chemie-1ng.-Techn. 30 (1958),407-416
[140a]. Wilcox, W.R., Ind. Engng. Chem., Fundamentals 3 (1964)1,81-83
[141] Thormann, K.,Chem. Techn. 2 (1950), 255-256
11421 Cavers, S.D., Ind. Engng. Chem., Fundamentals 4 (1965) 2, 229-230
[t43] Pohl, K.,Chemie-1ng.-Techn. 28 (1956),562-564
[144] Horvath, P. J., and Schubert, R. F., Chem. Engng. 10 (1958)
[145] Flatt, R., Chimia [Zurich] 9 (1955),232-237
[146] Stage, H., and Schultze, Gg. R., Oel u. Kohle 6/6 (1944),90-95
[147] Zuiderweg, F. J., Laboratory of Batch Distillation, Interscience Publishers. Xea- \-ark
1957
[148] Schlfer, W., Angew. Chem., Part B 19 (1947), 251-253
[149] Fischer, W., Chemie-1ng.-Techn. 23 (1951),116
[150] Xatz, W., Angew. Chem., P a r t B 19 (1947), 131-134
[l5l] Hilberath, F., Oel u. Kohle 39 (1943),875-886
[id21 Bragg, L.B., and Lewis, J. W., Wld. Petroleum Rep. 14 (1943),61
'

[152a] Strangio, V. A, and Treybal, R,. E., Ind. Engng. Chem., Process Des. and Devel. 1 3
(1974)3, 279-285
[ l 3 ] Rose, A. and E., e t al., Distillation, Interscience Publishers, New York 1951
[I541 Rose, A., Ind. Engng. Chem. 33 (1941),594
[I551 Fenske, M.R., Tongberg, S. O., and Quiggle, D., Ind. Engng. Chem. 26 (1934),IltiO
[l56] Richter, H., Oel u. Kohle 40 (1944),282-288
l157] Pohl, H., Erdol 11. Kohle 6 (1952),291-294
[l58] Zuiderweg, P. I., Chemie-1ng.-Techn. 26 (1953),297-308
[159] Sizmann, R., Chemie-1ng.-Techn. 33 (1961),659-668
[160] Manning, R.E., and Cannon, M. R., Ind. Engng. Chem. 49 (1957),346-349
[161] Stuke, B., Chemie-1ng.-Techn. 25 (1953),677-682
[162] Heise, F., Hiller, G., and Wagner, H. Gg., Chemie-1ng.-Techn. 41 (1969)20, 1100
[162a]KLutar, A., and Wagner, H. Gg., Chemie-1ng.-Techn. 42 (1970),1127;46 (1974).997
[162b] Borchardt, E., and Wagner, H. Gg., Chemie-1ng.-Techn. 43 (1971),956-962: 4s
(1976),725
Chilton, T.H., and Colburn, A. P., Ind. Engng. Chem., ind. Edit. 35 (1935),255-2W).
904
Carney, Th. P., Laboratory Fractional Distillation, MacMillan, New Pork 1949
Danatos, S., and Osburn, J. O., Chem. Engng. 6.5 (1958),147-150
drkenbout, G. J., and Smit, W. M., Separation Sci. 2 ( 5 ) (1967),575-596
Merkel, F., Arch. Wgrmewirtsch. Dampfkesselwes. 10 (1929).1% 17
iranchembook.ir/edu

502 References

[l68] Weber, U. von, in: Gildemeieter, E., and Hofmann, Fr., Die iitherischen Ole (revised
b y W. Treibs). Vol. I, Akademie-Verlag, Berlin 1966
[it391 Robinson, C.S.. and Gilliland, E. R., Elements of Fractional Distillation, 6th Ed.
JlcGraw-Hill Book Company. New York 1950, pp. 476-478
[17U] Weber, U. von, Chem. Techn. 2 (1950),241-246
[171] Kolling, H..Chemie-1ng.-Techn. 21 (1952),405-411
[172] Kohrt, H.U.,-4ngew. Cheru., Part B 130 (1948),117-124
[1731 Hausen, H.,Z.anpew. PhyRik 4 (1962),41 -51
[174] Wijk, W.R. van, and Thijssen, H.A. C.. Chem. Engng. Sci. 1 (1962),121-123: 3
(19M),145-152
[I731 Lwow, S. W.,Ber. Aknd. \Vies. UdSSR 67 (1947),375-378
[175a]Jlmtafa, H.A., Chemie-1ng.-Techn. 47 (1975)2. 63
[175blChien, H.H.J . . C'hem. Engng. Sci. 8 (1973)11, 1967-1974
[17.?c] Serov, V. V., Abrsmenko. . '1 P.. and Zykov, D. D., Int. chem. Engng. 13 (1973)3,
514 -6 tti
[175d] Wagner, H..and BlaS, E., Chemie-1ng.-Techn. 4s (1967)3, 220-227
[176] €bell, E.. Chem. Techn. 4 (1952),200-205
[176a]Faldix, P.. and Stage, H.. Chemie-Ing.-Techn. 41 (1969)23, 1265 -1269

t 176bl Yiiller, if'.. Chem. Techu. 26 (1974)1, 16-20

176~1Billet. R..verfahrensterhnik 8 (1974)3, 65-72
[17Bd]Jobst, W.,K6llner. H., and Balke, M., Chem. Techn. 81 (1979)1, 19-22
[1771 Thormann, K., Dechema-Erfahrungsaustauech, Techn. Apparnte und Anlagen, Trennen
flussiger Mischunpen. Destillieren, Betriebstec4inik. Dechema, Frankfurt/Main, August.
1932
~ l i k ] Ellis. S. R. %I., and Freshwater, D. C., Perfurn. esuent. Oil Rec. 16 (1954).3SO--AR(i
[I791 Colburn, .4.P., Trans. Amer. Inst. chem. Engr. 37 (1941).805
[l80] Undcr\vood, A. .J. V.. Cheni. Engng. Progr. 44 [1948),60.1
LlKll Fenskc. M. R., Lnd. Engng. Chern. 24 (193'7). 482
[182) Harbert, W. D.,Ind. Engng. Chem. 37 (1945),1162
[la31 Rruijn, P. .J.. hledcd. Landhouwhogeschool. Wageningen 61 (9)(1961).1 -94
[1R3n I \Vagner, H., Chemie-1ng.-Techn. 4% (1976)8, 705-708; 9, 790-792; 10, 875--876
11, 10.59-1062; 12, 119?--1198; 49 (1977)1. 45-48
[lS4l Underwood, A. J. V., Trans. Instn. chem. Engr. [London] 10 (1932),112-192
[it351 Miireh, D. P.. Ind. Engng. Chem. 4-5(1953), 2616-2621
[186] Mullin, J. W.,Ind. Chemist. 38 (1957)390,408-417
[186al Groenhof, H.C.,Chern. Engng. J. 14 (1977)3, 181-191 ; 193-201
Grwnhof. H. C., nnd Stemerding, St.. Chem.-1ng.-Techn. 49 (1977)10, 835
Kirschbaum, E., hngew. Chem.. Part B 19 (1947),13-14; Chemie-1ng.-Techn. 2%
(1958).(i39-644
Kssaneki, B.A. .J. org. chem. (Russian) 12 (1942).112
Buehmakin. T.S., and Lplova. R. V., J. appl. Chem. (Russian) 26 (1952),303-312
Perktold. F., -4ngew. Chem., Part R 19 (1947),184-185
David. .\., Dechema-Monogr. 0 (1954).126-175
Myles, M.,et (11.. Tnd. Engng. Chem. 43 (1951), 1452-1456
Allenby, 0.C. W., and L'Heureux, C.. Analytic. Chem. 29 (1950).1340
Haldenwanger, K.,Chemie-1ng.-Techn. 23 (1951),437-440
Zuiderweg, F.J. (edit.) Recommended TeRt Mixtures for Distillation Columns. The
Institution of Chemical Engineers. London 1969
Brandt, H.. and Rack. H., Chemie-1ng.-Techn. 09 (1957),397-402
Feldmann, J..et al.. Ind. Engng. Chem. 46 (1953),214
Zelvenskiy. Y.D.. and Shnlgin. B. -1..Chem. Ind. (Russian) 6 (1962),38 (424)-41
(427)
Schultze. Gf. R., and Stage. H.. Oel u. Kohle 10 (1944), 68
Collins, F.C., and Lantz, I-., Ind. Engng. Chem.. analyt. Edit. 1s (1946),673-677
Saumann, K.,and Leibnitz. E., Chem. Techn. 8 (less),458-471
iranchembook.ir/edu

References ,503
~

[202] Obolonzev, R. D., and Frost, A. V.. Petroleum Econ. (R.ussian) 25 (1947). :33-13
"7031 Struck, R . T., and Kinnep, C. R., Ind. Engng. Chem. 41 (1950), 77-82
[204] Weber, U. von, Personal communication.
[205] Zuiderweg, F. H., Chem. Engng. Sci. l ( 1 9 5 2 ) . 174-193
12061 Kirschbauni. E., Bnsch, W..and Billet, B., Chemie-1ng.-Techn. 26 (1956), -173 -480
[207] Szapiro, S., Zeszytp naukowe Polit~echn.Lodzkiej, Chem. 2 (1955), 33-37
[208] Brauer. H., Chemie-1ng.-Techn. 29 (1957), 520-530, 785-790
[209] Zelvenskiy, Y. D.. Tit'ov, A. A., and Shalgin, B. A.. Cheni. Ind. (Russian) 2 (1963). 3;
(116) -42 (122)
[210] Wepgzznd, C., and Hilgetag, G., Organisch-chemische Experimentierknnst. .it,h Ed..
Johann Ambrosius Barth, Leipzig 1970
[211] Bernhauer, K.,Einfiihrung in die organisch-chemische Laborat~oriiimstechnik,Springer-
Verlag, Vienna 1947
[2121 Pestemer, M., Angew. Chem. Wa (1961), 118-122
[213] Brandt, P. L., Perkins, R. B., and Halverson, L. I<., Oil Gas J. 7 (1946)
[214] Lentz, H., Chemie-1ng.-Techn. 41 (1969) 15, 865-861
[214a] Leung, L. S., Hutton. B. E. T.. and Nicklin, 11. J..Ind. Engng. Chem., b'izndamentals
14 (1975) 1, 63-67
r214b] Kafarov, V. i'., Blyakhmnn, L. F., and Planovsky, A. N., t.eoret. osriovy khim. telthool.
10 (1976) 3, 331-339
[215] Stage, F.. dngew. Chem., Part B 19 (1947), 182
12161 Ychultze, Gg. R.. and Stage, H., Uber die Wirksamkeit, der nciien Xormalkolonne
des ,,Ausschusses fur I,aboratoriumsdest,illation", Fiat-Report No. 1133 of 23 Dw. 1947
[217] Sigwart, K., Destillation und Rekt,ifikat,ion, in: Ullmanns Encyklopadie der technischeii
Chemie, Vol. 1, Urban R: Schwnrzenberg, Miinchen/Berlin 1951, pp. 429-470
12181 Thormann, K., Dechema-Erfahrungsanstausch, Laboratoriomstechiiili, Arbeits-
methoden und Gerate, Destillieren iind Rektifizieren, Dechema. Fraiikfiirt,/~Iiliii.
September 1950
[219] Mach, E., VDI-Forschnngsheft 375 Nov./Dec. 1935
[220] Barth, W., Chemie-1ng.-Techn. 23 (1951), 289-293; 26 (1954), 29-34
[2211 Leva, M., Flow through Packings and Beds, Part I, New Pork 1957
[222] Kast, W.,Chemie-Ing. Techn. 36 (1964) 5, 464-468
[223] Teiitsch, T., Cheniie-1ng.-Techn. 86 (1964) 5, 496-523
12241 Gelbe, H., (.'hemie-1ng.-Terhn. 40 (1968). 528-530
[225] Glaser, H., Cheniie-1ng.-Techn. 41 (1969) 7, 444-450
[226] Heckmann, C. J., and Krell, E., Chem. Techn. 5 (1953), 609-704
[227] Neumann. F., Chemie-1ng.-Techn. 33 (1961), 439-442
[228] Reed, T. M., and Fenske, M. R . , I d . Engng. Chem. P2 (l950), 654-660
[229] Stage, H., From a prospectus on Labodest distillitt,ion apparatus from standard coinp-
nents, drawing No. 741, 98, 156
Stage, H., and Bose, K., Chemie-Ing.-Techn. 17 (1955), 439
[230] Umholtz. C. L., and Winkle, YI. van, Petroleiini Refiner 34 (195.5), 114-117
[231] Schumarher. R., Erdol u. Kohle 2 (1949), 501-508
[231a] Schumacher, R., Verfahrensbechnik 10 (1976) 11. 727-732
[232] Borchert, R., and Jubitz, W., Infrarotstrahler zur Erwiirmnng, Trockniuig t i n t 1
Aushartung, SVT Vol. 5, VEB Verlag Technik, Berlin 1951
[233] Junge, C., Chem. Techn. 6 (1954), 37-38
[234] Grosse-Oetringhaus. H.. Erdol u. Kohle 1 (1948), 273-280
[235] Reid, It. C., and Sherwood, T. K., The Properties of Gases and Liquids. 2nd Ed..
McGraw-Hill Book Company, New York 1966
[236] Hecht, G., et al.. Berechnung thermodynnmischer Stoffwerte von Gasen und Fliissig-
keiten, VEB Dcutscher Verlag fur GrundsDoffindust.rie, Leipzig 1966
[237] Stage, H., Fette 11. Seifen 56 (1953), 284-290
[238] Kharbanda, 0. &I.P., Nomograms for Chemical Engineers, Hepwood & Company.
London 1958
iranchembook.ir/edu

504 References

[239] Orlicek, A. F., and Poll, H., Hilfsbuch fur Mineraloltechniker. Vols. I and IT, Springer-
Verlag, Vienna 1951/55
[$Mu)]Haussler. W.,Technik 9 (1954), 51 -52
[241] Cole. P. .J., Analyt. chem. 52 (1950). 946-947
12421 Krell, E., Chcm. Techn. 4 (1952), 200-207
[243] Meier-Grolman, F. W., and Wesolofsky, F., Oel u. Kohle 39 (1941), 297-3300
p2441 Beck. R., Chcmie-1ng.-Techn. 44 (1972) 8, 596-598
[244n] SarkoviE-Stefanovir, J., SimonoviE, D., and PopoviE, G., Chem.-1ng.-Techn. 49
(1977), 825-826
[245] Gemmeker, L., and Stage, H., Olas-Instrumenten-Techn.7 (1963) 1, 18-23
Kaplick, K., and Schubert, M.,Datenverarbeitung in der Verfahrenstechnik - Digitalc
Rechenaotomaten, Vol. 6 of the aeries Beitriige zur Verfahrenstechnik, VEB Deutscher
Verlag fur Grundstoffindustrie, Leipzig 1969
[!?471 Wagner. M., -4nalogrechner in der Verfahrenstechnik, Vol. 10 of the series Beitrage'
zur Verfehrenstechnik, VEB Deiitecher Verlag fur Grundstoffindustrie, Leipzig 1972
[ S l S l Adler, H., Elektronische Analogrechner. 3rd Ed., VEB Deutscher Verlag der Wissen-
schaften, Berlin 1970
[249] Savas, E. S., Computer Control of Industrial Processes, McGnrw-Hill Book Company,
Ziea York/Toronto/London 1965
[%I q Roth, H.,Chem. Techn. 20 (1968) 2, 77-82
r2511 Schubring, A., Chcmie-1ng.-Techn. 32 (1960) 3, 223-227
[26113 Chien. H. H.J.. -4IChEJ 84 (1978) 4, 606--613
(251 b] Rose, L. M.,Chimia. Aarau 33 (1979) 2, 59-G5
[252] Sealey, C. J., Chem. Engng. Sci. 56 (1970), 561-568
[253] Leibnitz. E..and Schuhler, C., Z. physik. Chem. 217 (1961), 217 - 229
Schuhler, C.. Z. physik. Chem. 217 (1961), 263-270
[251] Nagel, O., Sniita. H., Wolf, D., and Gunzler, H., Chemie-1ng.-Techn. 41 (1969) 17,
983-987
[255] Pilhofer, Th., Chemie-1ng.-Tech. 42 (1970) 1, 509-515
.Johnson, hl. L., and Lupfer, D. E., Chem. Engng. Progr. 63 (1966) 6. 75-79
[2561 Wang, I. C., and Henke, G. E., Hydrocarb. Process. 46 (1966) 8, 155
[257] Sargent,. R. W.H., and Murtagh, R. A.. Trans. Instn. chem. Engr. 1 7 (lY69), 85
[238] Stainthorp, F. P., e t sl.,in: Piric, J. M.: Efficient Computer Methods for the Practising
Chemical Engineer. London 1967, pp. 181-189
[%9] Blatt. E., and Fleiesner, H., Glas-Instrumenten-Techn. 15 (1971), Sondcrhcft, 1333 to
1337
[260] Brauer. H., and Mewes, D., Chemie-1ng.-Techn. 44 (1972) 1/2, 93-96
[26l] Bitter, R. C.. Chemie-1ng.-Techn. 40 (1968) 13, 651-656
[262] Wagner, H.. and Blass, E., Chemie-1ng.-Techn. 46 (1973), 63-67
[263] Reichelt.. W.,and Rlass, E., Cbemie-1ng.-Techn. 48 (1971) 17, 949-956
h i c h e l t , \Y.. Chemie-1ng.-Techn. 48 (1971) 22, 1196-1203; 44 (1972) 18, 1068-1071
[264] Kohler, R.., nnd Schober, G., Chem. Techn. 26 (1973) 12, 718--721; 56 (1974) 8, 482
to 483, and 12, 747-750
[P65] Kirbitch, V.. Frenzel, L., and Strankmuller, El., Chem. Techn. 26 (1973) 12, 721-722
[2(lfi] Xorozova, L. V., and Platonov, V. M.,Th'eor. Found. chem. Engng. 8 (1974). 453-458
[267] Grosshenuig, P., Thesia Tu M h c h e n 1976
[268] BlaB, E., and Kurtz, R., Verfahrenstechnik 10 (1976) 11, 721 -724
[269] Achwal, S. B., and Stepanek, J. B., chem. Engng. J. 12 (1976) 1, 69-76
[270] Zech, J. B., and bfersmann, A. B., Chem.-hg.-Techn. 60 (1978) 7, 549
12711 Pouter, 3. B., Peier, W.,and Fabre, S., Chem.-Ing.-Techn. 60 (1978) 6, 444-446
[272] Fabre, S., Ponter. A. B., and Huber, M.,Ind. and Engng. Chem., Process Des. Dev. 1s
(1979) 2, 266-269
[273] Gohrke, E., and Kleemann, G., Chem. Techn. 28 (1977) 11, 604--607
1.2741 \VeiO, S.. and Kirbach, V.. Chem. Techn. 80 (1978) 11, 562-565
iranchembook.ir/edu

References 50.5

Chapter 5 - Separating Process

[l] Stage, H., and Gemmeker, L., Chem. Labor Betrieb 16 (1964) 5,177-188; 6,233-242;
7, 290-298; 8, 325-338
[ l a ] Stage, H., CZ-Chemie-Technik 1 (1972), 263-272
[2] Bodenlieimer, W., Nature [London] 173 (1954), 124
[3] Jantzen, E., and Witgert, H., Fette u. Seifen 46 (1939), 563
13a] Fischer, W. G., GIT Fachz. Lab. 18 (1974) 7, 689--691
[3b] Bertcle, E., and Humbel, F. A., GIT Fachz. Lab. 1 7 (1973) 6, 667--674
[4] Bitbcock, M. J., Analytic. Chern. 21 (1949), 632-633
151 Shrader, S. A., and Ritzer, J. E., Chem. Fabrik 12 (1935), 173
[6] Klenk, E., Hoppe-Seyler’s Z. physiol. Chem. 242 (1936), 250
[7] Fischer, W. S., Chemikcr-Ztg. 94 (1970) 5, 157-161; Glas-Instrumenten-Tecliii. 12
(1969), 535-540
[8] Winters, J. C., and Dinerstein, R. A., Analytic. Chem. 27 (1953), 546-550
Nerheim, A. G., arid Dinerstein, R. A., Analytic. Chem. 28 (1956), 1029-1033
[9] Spiegelberg, C., Thesis P.U. Berlin 1956
[lo] Gould jr., C. W., Holzmann, G., and Niemann, C., Analytic. Chem. 20 (1948). 3 6 1
[ I l l Breger, 5. A., Analytic. Chem. 20 (1948), 980
[I21 Paschke, R. F., et al., J. Amer. Oil Chemists’ SOC.31 (1954), 5-7
[13] Pfeil, E., Angew. Chem. 54 (1941), 161-184
1141 Eigenberger, E., Xikrochernie 26 (1939), 204
[l5] Clasen, H., Chemie-1ng.-Techn. 28 (1956), 45-48
[15a] Bock, R., Methoden der Analytischen Chernic, eine Einfiihrung, Band 1 : Trennungs-
methoden, Verlag Chemie, Weinheim/Bergstr. 1974
Stage, H., Fette u. Seifen 56 (1953), 217-224
Giinther, G., Kiihnhanns, K., and Hiittig, E., Chem. Techn. 7 (1955), 656-662
Profft, E., and Buchmann, G., Chem. Techn. 8 (1956), 417-418
Buchmann, G., Chem. Techn. 14 (1962) 10, -617-620; 12, 741-744
Fiirst, H., and Schwetlick, K., Chem. Techn. 8 (1956), 458-471
Olechnowitz, A. F., Thesis Halle-Wittenberg 1956
Fiebach, K., Thesis Leipzig 1957
Gutwasser, H., and Miiller, H.,Chem. Techn. 21 (1969) 11, 756-759
Gutwasser, H., Chem. Techn. 22 (1970) 8, 467-471
Mair, B. J., e t al., Analytic. Chem. 30 (1958), l(314-1817
Naumann, K., and Leibnitz, E., Chem. Techn. 8 (1956), 458-471
Oldroyd, D. M., and Goldblatt, L. A., Ind. Engng. Chem., analytic. Edit. l h (1!+4-6).
761 -763
Petrocelli, J. A., Puznirtk, T. J., and Clark, R. O., Analytic. Chem. 36 (1964), 1008 1011
~

Prospectus “Aerograph Research Notes 1965”, Wilkens Instrument and Research .A(;,
Basle/Switzerland
Stage, H., CZ-Chemie-Technik 2 (1973) 1, 35-40
[29j Hampel, B., Chemiker-Ztg. 95 (1971) 7, 316-325
“1 Rieger, H., Chcm. Techn. 22 (1970) 1, 36-40
[30a] Wiikelsesser, G., CAV (1976)~ 6 122-130 ,
[30b] Pietzsch, W., and Rautenberg, K.-H., Chem. Techn. 31 (1979) 3, 135-137
[31] Carpenter, J. K., and Helwig, R. W., Ind. Engng. Chem. 42 (1950), 571-578
[32] Jordan, D. G., Chemical Pilot Plant Practice, Interscience Publishers, New Yolk 1933
[33] Dietz, H., and Kardos, J., Chem. Techn. 19 (1967) 12, 713-719
r.141 Billet, R., Chemie-1ng.-Techn. 42 (1970) 14, 910-917
[ 3 5 ] Kullrnann, W. W., Pharmaz. Ind. 15 (1953), 159-162
[36] Job, C., and Zellner, H., Pharmaz. Ind. 17 (1955), 161-163. Manufacturer: Srhott 11.
Gen., Maim
[%a] Prospectus “Quickfit, Single-stage water distillation apparatus”. Throughput 3.5 to
4 l/h of distilled water
iranchembook.ir/edu

506 References

[36b] Schulze, F.-W., Petrick, H.-J., Cammenga, H. K., and Baer, K., GIT Fachz. Lab. 2%
(1978) 6,475-483
[37] Schpolski, E. W., Atomphyaik, Part 1, 12th Ed., Part 2, 9th Ed., VEB Dentsoher
Verlag der Wissenschaften, Berlin 1974
[38] Brodsky, A. E., Isotopenchemie, ed. by J. Miihlenpfordt, Akademie-Verlag, Berlin
1959 (Transl. from Russian)
[39] Brun.de, K., and Voshage, H., Massenspektrometrie - Physikalische und apparative
Grundlagen sowie Anwendungen, Verlag Karl Thiemig, Xiinchen 1964
Birkenfeld, H., Haase, C., and Zahn, H., Massenspektrometriche Isotopenanalpe,
2nd Ed., VEB Deutilcher Verlag der Wissenschaften, Berlin 1969
Wet@, K., Hiibner, H., Hubner, G., and Maass, I., Anwendung stabiler laotope in
Forschung nnd Technik, VEB Deutscher Verlag fiir Grundstofhduatrie, Leipzig 1983
[40] Conn, A. L., and Wolf, J. E., Ind. Engng. Chem. 50 (1958), 1231-1234, DBP No.
944848, granted 28 June, 1958
I413 Xercea, V., Isotopenpraxis 2 (1966) 8, 313-322
[a] 3UiIler, G., Mauersberger, K., and Sprinz, H., Analyse stabiler Isotope durch spezielle
Xethoden, Akademie-Verlag, Berlin 1989
1431 Baertschi, P., Kuhn, W., and Kuhn, H., Nature [London] 171 (1953), 1018
[44] Sellers, E. S., and Augood, D. R., Trans. Instn. chem. Engr. 34 (1956), 53-78
[45] Timmerhaus, K. D.. Weitzel, D. H., and Flynn, T. M., Chem. Engng. B o p . 64 (1958),
35-46
[46] Weiaser, R. H., Techn. Rundschau Sulzer 40 (1958), 71-73
[47] Shavoronkov, 3.31.. and Sakodynskiy, K. I., Kernenergie 4 (1961) 12, 893-904
[48] Clusius, K., and Meyer, H., Helv. chim. Acta 36/7 (1953), 2045-2055
[49] Groth, W.,Ihle, H., and Murrenhoff, A., Angew. Chem. 66 (1956), 644-648
[SO] Cluuius, K.,and Schleich, K., Helv. chim. Acta 42 (1959), 232-239
SlcInteer, B. B., and Potter, R. M., Ind. Engng. Chem. Proc. 4 (1965) 1, 35-42
[51] Devyatikh, G. G., Sorin, A. D., and Nikolaev, S . Y., J. appl. Chem. (Russian) 31
(195$), 368-375
[32] Sevrugova, N. N., Uvarov, 0. V., and Shavoronkov, N. M., Atomic Energy (Russian) 4
(1956), 113-117
[53] Kistemaker, J., Bigeleisen, J. B., and Sier, A. O., Proceedings of the International
Symposium on Isotope Separation (Amsterdam, April 23-27, 1957), North-Holland
Publishing Company, Amsterdam 1958
[34] Yagodin, G. A., Uvarov, 0. V., and Shavoronkov, N. M., Ber. Akad. Wiss. UdSSR
111 (1956), 384-387
[.%I Clusius, K., and Schleich, K., Helv. chim. Acta 41 (1958), 1342-1358
[56] Kirshenbaum. J., Physical Properties and Analysis of Heavy Water, McGraw-Hill
Book Company, New York 1951
[a71 Urey, H. C., J. chem. SOC. [London] 1347, 562
[58] Benedict, M., Survey of Heavy Water Production Processes, Conference, Switzerland,
30 June, 1955
[59] Kuhn, W., Chemie-1ng.-Techn. 25 (1953), 12-18; 29 (1957), 6-16
Kuhn, W., Buertschi, P., and Thurkauf, M., Chimia [Zurich] 8 (1954), 109-122,
145-156
Kuhn, W., Helv. chim. Acta 36 (1952), 1684-1736
Kuhn, W., Helv. chim. Acta 87 (19M), 1407-1422
[59a] Lumb, P. B., Nuclear Engng. 21 (1976) 244, M-69
[60] Baertachi, P., and Kuhn, W., Final Concentration of Heavy Water by Rectification,
Conference, Switzerland, 8 July, 1955
1611 Becker, E. W., h g e w . Chem. 68 (1956), 6-13
Huber, M., Die Rektifikation als Spezialfsll der Theorie der Isotopentrennung, Sulzer-
Verfahrenstechnik, Achema 1961
[ti21 Hiinny, J., Hartmann, F., and Hiltbmnner, R., Techn. Rundschau Sulzer 4 (1961),
21-26
iranchembook.ir/edu

References 60 T

Dixon, 0. G., J. SOC.chem. Ind. 68 (1949), 88-91

Krell, E., Chemie-1ng.-Techn. 32 (1960). 233-240
Uvarov, 0. V., Sokolski, V. A., and Shavoronkov, S. 31., Chem. Ind. (Russian) 7
(19563, 20-21
Dostrovsky, J., Llewellyn, D. B., and Vromeu, B. H., J. chem. Soc. 4 (1952), 3509-3521
Jackson, R. F., and Pigford, R. L., Ind. Engng. Chem. 4s (1956), 1020-1026
Rosen, A. M., Ber. Akad. Wiss. UdSSR 107/2 (1956), 295
London, H., Separation of Isotopes, George Sewnes Ltd., London 1961
Cohen, K., Theory of Isotopic Separations, McGraw-Hill Book Company, New York 1951
Staschewski, D., Chemie-1ng.-Techn. 40 (1968), 748; 43 (1971), 14, 810-816
Chem. Techn. 4 (1975), 269-280
Production of Oxygen-18 and Oxygen-17 by Countercurrent Distillation of Reactor
Grade Heavy Water, in Panel Proceedings Series: "Stable Isotopes in the Life Sciences''.
IAEA Vienna 1977 43 (1971), 14, 810-816
Suroniec, A. J., Ind. Engng. Chem. 62 (1960) 9, 798
Gupta, D., and Ray, S. N., Ind. Engng. Chem. Proc. 1 (1962), 255-262
Petgluk, F. B., Platanov, V. M., and Girsanov, I . V., Ind. Engng. Chem. 6 (1965) 7 .
309 - 317
Baertschi. P.. and Kuhn, W., Helv. chim. Acta 40 (1957).
, I .
1084-1103
Stage, H.; Chemische Arbeitsgemeinschaft Verfahrenstechnik, Jan./Feb. 1970, 44-51.
p 3 a ] Schwab, H., Chemie-1ng.-Techn. 44 (1972) 11, 704-iO7
[74] Gelbin, D., Chem. Techn. 21 (1969) 6, 329-334
[75] Cannon, N.R., Ind. Engng. Chem. 53 (1961), 629
1761 McWhirter, I. R., and Lloyd, W., Chem. Engng. Progr. .59 (1963) 6, 58
[77] Burstin, H., and Winkler, J., Chemiker-Ztg. 22 (1931), 211
[is] Kolling, H., and Tramm, H., Chemie-1ng.-Terhn. 21 (1949), 9
[79] Capitani, C., and Mioani, E., Ind. chimica 36 (1954), 671-684
[SO] Stage, H., Dechema-Monogr. l b (1950), 156- 170
[81] Sigaart, K., Destillieren und Rcktifizieren, in : Houben-Weyl, Nethoden der organi-
schen Chemie, ed. by E. Muller, Vol. I/l, Georg Thieme Verlag, Stuttgart 1955
[Y2] Harrsen, H., and Schlabterer, R., Cliemie-lng.-Techn. 91 (l949), 463-460
Kirschbaum, E., and Lipphardt, G., Cheniie-1ng.-Techn. 09 (1957), 393- 397
[83] Otto, W., Chem. Techn. 14 (1962) 4, 228-231
[84] Kirschbaum, E., Destillier- und Rektifiziert>echnik,4th Ed., Springer-Verlag, Berlin,'
Heidelberg/ISem Tork 1969
[853 Fahuss, B.. Z. Ver. ung. Chemiker 8 (1953), 323-324
[86] Fischer, V., Ann. Physik 6 (1947), 139-152
[87] Herrmann, E., Chemiker-Ztg. 80 (1956), 102-104, 131-134, 169-173
[SS] Trost,er, E., Chemie-1ng.-Techn. 32 (1960) 8. 525-531
[?&a] Wondrak, G. A. M., Ber. Bunsenges. 81 (1977) 10,1033-1036
Chem.-1ng.-Techn. 46 (1974) 65 (Synopsis)
[89] Junge, C., Chem. Techn. 6 (1954). 37-38
[90] Trenne, H.. Erdol u. Kohle 2 (1951), 63-66
[Yl] Kuhn, W., Helv. chim. Acta 05 (1952), 252-295
[92] Weber, U. von, Chem. Techn. 2 (1950), 241-246
[93] Voigt, G., Kernenergie 2 (1959), 14-39
[93a] Blass, E., Chemie-1ng.-Techn. 45 (1973) 13, 865-872
[93b] Wondrak, A. M., and Rohler, F., Chemie-1ng.-Techn. 46 (1974) 1, 37
[93c] Anderle, F., Frey, H., and Lerch, G., Kerntechn. 19 (1977) 11, 483-487
[93d] BlaB, E., and Sauer, H., Chem.-1ng.-Techn. 60 (1978), 525-532
[94] Podbielniak, W.J., Ind. Engng. Chem., analytic. Edit. 3 (1931), 177; 5 (1933), 119
[95] Grosse-Oetringhaus, H., Erdol u. Kohle 2 (1949), 286-299
[96] Wustrow, W.,Z. analyt. Chem. 108 (1937), 305
[97] St'okes, Ch. S.. and Hauptschein, M.,Analytic. Chem. 24 (1952), 1526
Drawin, H. IF'., Vakuum-Techn. (1961), 2
iranchembook.ir/edu

508 References

[98] Lerer, E., and Kuss, E., Z. physik. Chem., Sec. A, Vol. 163 (1933),73
[!%I Miller, G. H., Operations with Gases in Technique of Organic Chemistry, Vol. 111,
Part U, Chap. V, Interscience Publishers, New York 1957
[loo] Maler, G., and Gnauck, G., Reinste Gase, VEB Deutscher Verlag der Wissenschaften,
Berlin 1965
[101] Koch, H.,and Hilberath, F., Brennstoff-Chem. 21 (lW) ,
197-203
Hammerich, Th., Erdol u. Kohle 5 (1952),361-366
[1(,2] (?remmeker,W.,and Stage, H., Glas-Inatrumen~n-Techn.6 (1962)10, 372-376
[lo31 mein, E., and Walz, H., Gfas-Inthrumenten-Techn.11 (1967)5, 405-411
[lo41 Podbielniak, W.J., and Preston,S. T., Petroleum Engr. 6 (1955),205-211
rlC6] Manufacturer: Ernst Haage, Miilheim/R+r
[I061 Simone, G., Ind. Engng. Chem., analytic. Edit. 10 (1938),648
[lo71 Podbielniak, W. J., Apparatus and Methods for Precise Fractional Distillation Ana-
lyses, Smerican Chemical Society, Atlantic City Meeting 1941
[lo81 Rose, A. and E., et al., Distillation, Interscience Publishers, New York 1951
[log] Preston, W.,e t al., Freibetger Forschungshefte 5, Akademie-Verlag, Berlin 1953,
pp. 231-235
/
[I101 Stage, H.,Destillieren und Rektifizieren, Sec. 19, in: Fortschritte der Verfahrens-
technik 1952/53,Verlag Chemie GmbH, Weinheim/Bergstr. 1954
El11 1 Bprnhauer, K., Einfiihrung in die organisch-chemische Laboratoriumstechnik, Springer-
Verlag, Vienna 1947
[112] Gemmeker, L.,and Stage, H., Glas-Instmmenten-Techn. 6 (1962)11, 455-458
I1131 Holland-Merten, E. L.. Tabellenbuch der Vakuumverfahrenstechnik in der Grund-
stoffindustrie, VEB Deutscher Verlag fur Grundstoffindustrie, Leipzig 1968
[I141 ,Jaeckel, R., Kleinste Driicke, ihre Messung und Erzeugung, in: Technische Physik
in Einzeldaretellungen, Springer-Verlltg, Berlin/Gottingen/Heidelberg1950
Leybold, Vekuum-Taschenbuch, ed. by K. Die18 and R. Jaeckel, Springer-Vcrlag,
Berlin/Gottingen/Heidelberg1962
Keunecke, E.,Chemiker-Ztg. 81 (1957),239-244
Ymwood, J., Hochvakuumtechnik, 3rd Ed., Rudolf A. Lang Verlag, Berlin 1955
(Trans]. from English)
Pupp, W., Vakuumtechnik, Grundlagen und Anwendungen, Thiemig-Taschenbuch,
Vol. 43, Verlag Karl Thiemig, Miinchen 1973
[I 151 Miiller, L., Freiberger Forachungshefte, Edit. B 6, Akademie-Verlag, Berlin 1964,
pp. 81- 115; Freiberger Forschungshefte 11, Akademie-Verlag, Berlin 1957, p. 575
11161 Spendlove, M. J.. Vakuum-Techn. 8 (1957),15-18 and 36-44
[I171 Eorsley, W.,J. appl. Chem. 8 (1968),13;Chemie-Ing.-Teehn. 30 (1958), 736
[f 181 Parker, S. G.,and Wilson, 0. W., Ind. Engng. Chem. Roc. 4 (1965)4, 365-368
[118a] Bezobrazov, E. G., et al., khim. prom. (1974)4,301-302
[ll8b] Schmiicker, H., and Grigull, U., Heat Transfer in Boiling, Academic Press, 1977,
pp. 361-387
[I191 Laporte, H.,Hochvakuum, VEB Wilhelm Knapp Verlag, Halle 1951
[l20] EchoIs jr., L. S., and Gelus, E., Analytic. Chem. 19 (1947),668
[I211 U b o y , E.. and Sakaloz, T., Theoretische Grundlagen der Isotherm-Destillation
(Mitt. I-III), Revue de la facult& des sciences de I’universitB’d’Istanbul, Sbie C,
Tome XXITI/XXIV, Fasc. 3-4 (1958/59)
El221 Frank, W..and Kutsche, D., Die schonende Destillation, Otto Krausskopf-Verlag
GmbH, Mainz 1969
[I231 Billet, 8..and Raichle, L., Ghemie-1ng.-Tech. 89 (1967)3, 133-137
[124] Gelbe, H.,Chemie-Ing.-Techn. 40 (1968)11, 528-530
11251 Harries, W., Chemie-hg.-Techn. el (1949),139-142
11261 Monch, G. C., Hochvakuumtechnik, Rudolf A. Lang Verlag, PoDneck 1960; Keues
und Bewkhrim aus der Hochvakuumtechnik, 2nd corr. Reprint, VEB Verlag Technik,
Berlin 1961
iranchembook.ir/edu

References 509

[1271 Holland-Merten, E. L., Die Vakuum-Purnpen in der Verfahrenstechnik, VEB \l~illlelm

Knapp Verlag, Halle 1950
[128] Brdenne, M. von, Technik 13 (1958), 275-285
[128 a] Prospectus “Quickfit-Wiegand Glas-Dampfstrahlpumpen”, QVF Glastechllik GmbH,
Wiesbaden
[I291 Kienel, G., Glas-Instrumenten-Techn. 10 (1966) 9, 761-770
[I301 Krell, E., Chem. Techn. 2 (1950), 262-255; 4 (1952), 443-450
[131] Wittka, F., Angew. Chem. 53 (1940), 557
[132] Gemmeker, L.. and Stage, H., Glas-Instrumenten-Techn. S (1964) 6, 413-@28: 7,
503-508
Stage, H., Chemieanlagen $- Verfahren (1972), No. 2
[132a] W h e l m , G., and MiB, H., Chem. Techn. 8 (1979) 8,377-383
[133] Utzinger, G. E., Chem. Techn. 7/8 (1943), 61
El341 Sanosfia, P., Dechema-Monogr. 55 (1965)
[135] Fiirst, H., and Schimke, D., Chem. Techn. 22 (1970) 9,553-555
[136] Dieter, K., Chemiker-Ztg. 94 (1970) 12, 445-451
[137] Billet, R., Chemie-Ing.-Techn. 29 (1957), 733-742
[138] Egli, R., Glas-Instrumenten-Techn. 11 (1967) 6, 509-516
[1:HaJ Krainer, M. T., GIT Fachz. Lab. 1!3 (1974) 7, m2-868
L138bj Sorbe, G., GIT Fachz. Lab. 21 (1977) 14, 1315
[139] Burrows, G., Molecular Distillation, Oxford-Clarendon Press, Oxford 1960
Ho16, J., e t al., The Spplications of Molecular Distillation, Akademiai Iiiadb, Budapest.
1971
[I401 Ridgway-Watt, P., Chem. and Tnd., 27 May, 1961, 680-692
[140a] Habendorff, R., CZ-Chem.-Techn. 2 (1973), 21-24
Hickmann, K. C. D., and Trevoy, T. J., Chemie-1ng.-Techn. 25 (1953). 672-676
Langmnir, S., Phys. Rev. 8 (1916), 176
Burrows, G., J. appl. Cheni. 7 (1957), 375-384
Jaeckel, R., Erdol u. Kohle 4 (1951), 1761
Perry, E. S., Methods in Enzymol. 3 (1957), 383-391
Jaeckel, R., and Oetjen, G.-W., Chemie-Tng.-Tech. 21 (1949), 169-176
Jaeckel, R., e t nl., Vakoumtechnik und Molekulardestillation, in: Fortschrit,tc tler
Verfahrenstechnik, Vol. 2, 1954/55, pp. 29-43; Vol. 3, 1956/57, pp. 62-73. Vcrlag
Chemie GmbH, Wcinheim/Bergstr. 195-1/57
Jaeckel, R., Destillation und Sublimation im Fein- und H o c h v a h u m , in: Houlsen-
Weyl, Methoden der organischen Cheinie, ed. by E. Miiller. Vol. 1/19 Geog Thiemc
Verlag, Stattgart 1958
Frank, W., Glas-Instrumenten-Techn. 9 (1956) 11, 777-782
Utzinger, G. E., Chemie-hg.-Techn. 26 (1954), 129-131
Melpolder, F. W.,e t al., Analytic. Chem. 27 (1955), 974-977
Malyusov, V. A4., et al., Chem. Ind. (Russian) .5 (1958), 36-42
Malyusov, V. 9., Umnik, N. N., and Shavoronkov, N. M., Chem. lnd. (Russian)
July/Aug. (1958), 296-302; Collect. Czechoslov. chem. Commun. 23 (1958), 1720- 1726
Hickman, K . C. D., Ind. Engng. Chem. 39 (1947), 686-694
Hickman, K. C. D., Ind. Engng. Chem., malyt. Edit. 29 (1937). 968
Hickman, K. C. D., U.8. Patent No. 2117802 of 17 May. 1938, and S o . 2210928 of
13 Aug., 1940
Embree, K.D.. Ind. Engng. Cheni. 29 (1937), 975
Gorriz, 4 . M.,e t al., An. Real. Soc. espan. Fisica Quim., Ser. B 48 (1952): 825-850:
Ser. B 49 (1953), 19-22, 107-114, 579-586
Fawcett, E. W., Kolloid-Z. 86 (1939), 34-91
Masch, L.-W., Chemie-1ng.-Techn. I 2 (1950), 141- 146
Hickman, K. C. D., Ind. Engng. Chem. 48 (1956), 496-499
Carney, T. P., Laboratory Fractional Distillation, MacMillan. New York 1949

34 Krell, Handbook
iranchembook.ir/edu

610 References

%lyusov, V. A., Malafoev, X. A., and Shavoronkov, N. X., J. phys. Chem. (Russian)
3% (1958) 10, 2403-2409
Schneider, K. W., Erdol u. Kohle 4 (1951), 4-9
Simons, J. H., Ind. Engng. Chem., analyt. Edit. 10 (1938), 30
Uyeha, H., and Hagihara, Y., J. chem. Engng. Japan 3 (1970) 1, 13-17
Bewilogua, L., Verges, P., and Vinzelberg, H., Isotopenpraxis 9 (1973) 3, 97-100
Liedel. K. H., Glas in der Chemie - Laborgeriite, GroSapparate, Rohrleitungen,
0th Krausskopf-Verlag GmbH, Maim 1973
Stage, H., and Fischer, W. G., GIT Fachz. Lab. 23 (1979) 8, 724-731
Matwiyoff, N. A., et al., The Stable Isotopes (ICONS) Research and Development
Program a t the Los Alamoe Scientific Laboratory, in Proceedings Series “Isotope and
Radiation Techniques in Soil Physics and Irrigation Studies”, IAEA, Vienna 1977
Asatiani, P. Ya., et al., Industrial Pilot Plant for the Production of Oxygen-18, in
Panel Proceedings Series: “Stable Isotopes in the Life Sciences”, IAEA, Vienna
1977
Block, U., Schoenemakers, H., and Wolf, D., Verf. tech. 12 (1978) 10, 678-681

Chapter 6 - Selective Separating Processes

Findlay, A., Die Phasenregel und ibre Anwendungen, Verlag Chemie GmbH, Wein-
heim/Bergstr. 1958 (transl. from English) I

Badger, W. L., and McCabe, W. L., Elemente der Chemie-Ingenieur-Technik, Springer-

Verlag, Berlin 1932 (transl. from English)
Stage, H., Fette u. Seifen 66 (1963). 613-516
Stage, H., Erdol u. Kohle 3 (1950), 478-490
Hoffmann, E. J., Bzeotropic and Extractive Distillation, John Wiley & Sons, London
1965
DRP No. 767 162 KI. 12r, of 27 Nov., 1936, issued 3 Dec., 1951
Sizmann, R., and Stuke, B., Chemie-Ing.-Tech. 27 (1955), 669-675
Algermissen, J., Chemie-Ing.-Tech. 30 (1968),502-510
Merkel, H., Chemie-1ng.-Techn. ”A (1962), 511-512
Stage, H., Biinger, R., and Jonas, A., Fette u. Seifen 66 (1963), 580-583
Pozzi and Escot, Bull. SOC. chim. France 31 (1904), 932
Pfeil, E., Angew. Chem.64 (1941), 161-184
Bernhauer, K., Einfiihrung in die organisch-chemischc Laboratoriumstechnik, Springer,
Verlag, Vienna 1947
Thormann, K., Chem. Fabrik 13 (lsaO), 3-9
Weber, U. von, Gewinnung iitherischer ole durch Destillation, in: Gildemeiater, E.,
and Hoffmann, Fr., Die iitheriechen ole (revised by W. Treibs), Akademie-Verlag,
Berlin 1966
Rigamonti, R., Chemiker-Ztg. 80 (1966), 628-631
[16a] Prenosil. J. E., Chem. Engng. J. 12 (1976) 1, 69-68
[leb] Berkes, R., Blickle, T., and Ujhidy, A., Chem. Techn. 31 (1979) 2, 69-72
[17] Rock, H., Destillation im hboratorium, Extraktive und azeotrope Destillation,
Dr. Dietrich Steinkopff Verlag, Darmstadt 1960
[lS] Kiimmerle, K., Chemie-hg.-Techn. 32 (19eo), 513-617
[19] Gerster, J. A., Chem. Engng. Progr. 66 (1969) 9, 43
[19ct] Miiller, E., Verfahranstechnik 8 (1974) 3, 88-93
[19b] Helms, G., and John, P., inform. chimie (1976) 161, 105-109
[19c] Schober, G., Orschel, K.-D., Sichting, H., Sasse, H., Kohler, R., and Kleinschmidt, G.,
Chem. Techn. 31 (1979), 73-75
[20] Lecat, M., Tables azhtropiques, Azhtropes binaires orthobares, 2nd Ed., Brussels
1949 (published by the author)
[21] Ogorodnikov, C. K., Lesteva, T. M., and Kogan, V. B., Azeotropic Mixtures (Russian),
Khimiya, Leningrad 1971
iranchembook.ir/edu

References
~ ~~
51 1

[22] Stuke, B., Chemie-Ing.-Techn. 26 (1953). 133- 136

[23] Coulson, E. A., and Herrington, E. F., J. Chem. SOC.[London] 1947, 597-607
[24] Kuhn, W., and Massini, P., Helv. chim. Acta 33 (1950), 737
Kuhn, W., and Kuhn, H. J., Helv. chim. Acta 40 (1957), 2433-2454
[25] Eniistun, B. V., Commun. Fac. Sci. Univ. Ankara, Ser. B 6 (1951), 1-21
[26] Litvinov, N. D., 5. phys. Chem. (Russian) 27 (1953), 476-490
[27] Swigtoslawski, W., Przemysl. chem. 7 (1951), 363-370; Bull. SOC. chim. belges fi2
(1953), 10-20; Bull. Acad. polon. Sci., CI. 111, Vol. I , 1-2, 1953
Swigtoslawski, W., Physikalische Chemie des Steinkohlenteers, N. J. Hoffmann Verlag,
Koln 1954
Swigtoslawski, W., Azeotropie und Polyazeotropie, Vol. 1, Panstwowe Wydawnictwo
Naukowe, Warsaw 1957
[ZS] Lecat, M., Azhotropisme e t Distillation, in: Grignard, Trait6 de Chimie Organique 1;
Paris 1935
[28a] Othnier, D. F., Verfahrenstechnik 8 (1974) 3, 94-99; 4,118-123
[29] Schuberth, H., Thermodynamische Grundlagen der Destillat,ion und Extraktion, Vol. J.
VEB Deutscher Verlag der Wissenschaften, Berlin 1972
[29a] Avetym, V. S., Mashkona, V. F.,Petlyuk, F. B., and Serafimov, L. -4.,Theor. Found.
chem. Engng. 8 (1974), 253-258
[29b] Svoboda, V., VonkovB, J., Holub, R. and Pick, S.,Coll. Czech. chem. Com. 42 (1977) 9,
2607 -2614
[30] Zieborak, I<., Prace Glbwnego Inst. Chem. przemyslowej, (1961), 1-44
[31] Orsza.gh, A., Przemysl. chem. 9 (32) (1953), 333-336
[31a] Avetyan, V. S., Karp, L. E., Petlyuk, F. B. and Sernfimov, L. A., Russ. J. of phys.
Chem., London 68 (1978) 10, 1425-1427
[31b] Tamir, A. and Wisnink, J., Chem. Engng. Sci. 33 (1978) 6, 657-672
[32j Malesinski, I. W., Bull. Acad. polon. Sci., S6r. Sci. chim, geol., geogr. 6 (1958), 105- 109
[33] Stecki, J., Bull. Scad. polon. Sci., S6r. Sci. chim., geol., geogr. 6 (1958), 47-53,
[34] Berg, L., e t al., Hydrocarbon Processing 26 (1966), 103; Ind. Engng. Chem. 37 (1945)
585; 38 (1946), 1249
Berg, L., Chem. Engng. Progr. 66 (1969) 9, 52-57
[35] Rose, A. and E., e t al., Distillation, Interscience Publishers, New York 1951
[%a] Sharov, V. T., and Serafimov, L. A., Physico-chemical Fundamentals of Distillation and
Rectification (R,ussian),Hhimiya, Leningrad 1975
[35b] Schliinder, E. U., Chem.-Ing.-Techn. 50 (1978) 10, 749-7.54
[46] Franck, H. G., Angew. Chem. 63 (1951), 260-263
j37] Coulson, E. A., and Jones, J. J., J. SOC.chem. Ind. 65 (1946)- 163-175
[38] Dummett, G. A., Petroleum [London] 18 (1955), 41-45
[$9] Hunsmann, W., and Simmrock, K. H., Chemie-1ng.-Techn. 38 (1966) 10, 1053-11059
139al Petlyuk, F. B., and Avetyan, V. S., teoret. osnovy khim. tekhnol. 7 (1973) 3, 307-312
[39b] Naka, T., Komatsu, T., Hashimoto, I.,and Takamatsu, E.. Int,. chem. Engng. 16 (1976)
2. 272-270
LjYc] Susarev, M. P., and Toikka, A. M., J. phys. Chem. (Russian) 60 (1976) 3, 458-460
[39d] Kudryavtseva, L., Toome, M., and Eisen, O., Chem. Techn. 89 (1977) 11, 622-624
1401 Lino, M., e t al., J. chem. Engng. Japan 4 (1971) 1, 33-36
[41] Mair, B. J., Glasgow, A. R., and Rossini, F. D., J. Res. Nat. Bur. Standards 87 (1941).
39
[42] Mair, B. J., et al., Analyt. Chem. 30 (1958), 393-400
[43] Feldman, J., and Orchin, M., Amer. Pat. No. 2583554, of 9 Feb., 1950, issued 29 Jan.,
1952
[44] Schubert, H., Chem. Techn. 21 (1969) 1, 28-34, 51
(451 Sheinker, Y. N., and Peresleni, Y. M., J. phys. Chem. (Russian) 26 (1952), 1103-1109
L46] Ruether, J. A., and Lu, B. C. Y., Ca,nad. J. chem. Engng. 50 (1972) 2, 266-270
1471 Nutting, H. S., a n d Horsley, L. H., Ind. Engng. Chem., analyt. Edit. 19 (1947),602-603
[48] Joffe, J., Ind. Engng. Chem. 47 (1955), 2533-2536
iranchembook.ir/edu

512
--
References
~~ ~

[49] Malesinski, W., Bull. Acad. polon. Soi., C1. I11 4 (l956),371-379
[49a]Bril, Zh. A., Moszhukin, A. S., Petlyuk, F. B., and Serafimov, L. -4., teoret. osnovy
khim. tekhnol. 9 (1975)6, 811-821
[W] Kortiim, G., and Buchholz-Meisenheimer, H., Die Theorie der Destillation und Ex-
traktion von Fliissigkeiten, Springer-Verlag, Berlin/G6ttingen/Heidelberg 1952
[51] Horsley, L. H.,-tropic Data, Vol. I and 11, American Chemical Society, Washing-
ton 1952/1962;Advances in Chem., Series No. 6 and 36
(521 Ewell, R.H.,Harrison, J. M., and Berg, L., Ind. Engng. Chem. 36 (1944),871
[ B a ] Eduljee, G. H., and Tiwari, I(.K., Chem. Engng. Sci. 31 (1976)7,535-540
1531 Marachner, R. F., and Cropper, W. P., Ind. Engng. Chem. $8 (1946),262-268
[MI Hordey, L. H., Analytio. Chem. 19 (1947),603
[56] Meissner, H. P., and Greenfeld, S. H., Ind. Chemist 50 (1948),438-442
[SS] Novikova, K.Y., and Netradse, A. G., Chem. Ind. (Russian) 2 (1958),102-105
[56a]Marinichev, A. N. and Susarev, M. P., Zh. prikl. khim. 61 (1978) 10, 2221-2225
[57J Skolnik, H.,Ind. Engng. Chem. 40 (1948),442
[58] Nakanisi, K..Ichinose, S., and Skirai, H., Ind. Engng. Chem., Fundamentals 7 (1968)3,
381 -387
[58a]Brandani, V.. Lnd. Engng. Chem., Fundamentals 13 (1974)2, 154-156
[58b] Martin, G. Q.,Hydrocarbon Process. 54 (1975)11, 241-246
[580]Bril, Zh. A., Mozzhukhin, 8. S., Petlyuk, F. B. and Serafimov, L. A, teoret. osnovy
khim. tekhnol. 11 (1977)6, 811-817
[58d] Timofeev, V. S., Babich, S. V., Tikhonova, N. K., Anton, N. D. and Serafimov, L. A.,
khim. prom. (1977)9, 665-669
[59] Kuhn, W.,Helv. chim. Scta 87 (1954),1586-1605
[60] Szapiro, S., Przemysl. chem. 12 (36)(lQ56),106-110
LSlJ Kortiim, G., and Bittel, A.. Chemie-1ng.-Techn. 28 (1956),40-44, 282-284
[61a] Hunsnann, W.,Chemie-lug.-Tech. tl8 (1961)8, 537-540
[el b] Hegencr, B., Hesse, D., and Wolf, D., Chemie-1ng.-Tech. 45 (1973)14,942-945
[61c] Marozovr, L. V.. and Platonov, V. M., Theor. Found. chem. Engng. 8 (1974),453-458
[62] Houdrg Process Corp., Amer. Pat. No. 2776936 of 25 June, 1953,isaued 8 Jan., 1957
16281 Lehmann, P.. Oehler, R., Hennig, R. and Ulbrecht, H., Chem. Techn. 30 (1978)8,
403-406
Hetmig, R.. Ulbrecht, H. and Lehmann, P., Wiss. Z. d. TH Leuna-Merseburg 21 (1979)
2,185-195
[Sd] Grewer, Th.,Chemie-Lug.-Tech. 4.3 (1971),655-658
[63a]Pavlov, S. Yu.,Skorikova, V. V., and Borkova, S. B., Theor. Found. chem. Engng. ti
(1972)3,425-428
[64] Stage, H.,Erdol u. Kohle 3 (1950),478-490
[65] ScBeibel, E.G., Chem. Engng. Progr. 44 (1948),927-931
[ G a l Garber, Yu.N., undMironenko, V. F., teoret. osnovy khim. tekhnol. 8 (1974)6,846-852
[Mi] Kogan, V. B.,J. phys. Chem. (Russian) 29 (1953),1470-1476
[67] M a r o v , V. V., and Gordiyevski, L.A., J. appl. Chem. (Russian) 29 (1956),176-183
[68] K w n , V. B., Chem. Ind. (Russian) 6 (1957),356-361
[69] Wck, H..Chemie-1ng.-Tech. 28. (1956),489-495
Horsley, L. H., Analytic. Chem. 19 (1947),508-600; 21 (1949),831
[7U] Pork, R.S., and Johnson, J. F., ing. Engng. Chem. 62 (1960)8, 691-694
1711 Hildebrand, Solubilities, 2nd Ed., New York 1936
c72] Brit. Pat. No. 613867,issued 2 Dec., 1948
[73] h e r . Pat. No. 2549586 of 7 Nov., 1945, issued 17 April, 1961
Chiswold, J., et al., Ind. Engng. Chem. 38 (1946),65-70
[74] , h e r . Pat. No. 2610704 of 3 July, 1947,issued 16 Sep., 1952,Amer. Pat. No. 2588063
of 30 Sep., 1947, issued ,4Maroh, 1.952,Amer. Pat. No.2371 342 of 13 March, 1945,
issued 13 Maroh, l a 5 .
[751 Garner, F. H., Ellis, S. R. M., and Granville, W.H., J. Inst. Petroleum 43 (1956),
148-154
iranchembook.ir/edu

References 513

[76] Null, H. W., and Palmer, D. A., Chem. Engng. Progr. 66 (1969) 9, 47
[77] Gelbin, D., Chem. Techn. 11 (1959), 70-73
[78] Nagel, 0. and Sinn, R., Chemie-Ing.-Techn.49 (1967) 5/6, 275-282
1791 Kortum, G., and Faltusz, E., Chemie-1ng.-Techn. 33 (1961), 599-6606
[801 Garwin, L., and Hutchinson, K. E., Ind. Engng. Chem. 43 (1950), 727-730
[81] Daubach, E., Final Examination Paper, Marhurg 1948
[82] Jost, W., Chemie-1ng.-Techn. 23 (1951), 64
[83] Tursi, R. R., and Thompson, R., Chem. Engng. Progr. 47 (1951), 304-308
[84] Novella, E. C., and Tarasso, J. M., An. Real. SOC.espan. Fisica Quim., Ser. B if4 (1952).
441-448
[85] Schier, K. H., Thesis, Halle- Wittenberg 1959
[85a] Olefir, L. I., Knyshenko, M. I. and Priima, L. K., khim. prom. (1978) 2, 123-114
[86] Belck, L., Chemie-1ng.-Techn. 23 (1951), 90-91
[87] Bogart, M. J. P., and Brunjes, A. S., Chem. Engng. Progr. 41 (1948), 95
[88] Rieder, R. M., and Thompson, A. R., Ind. Engng. Chem. 42 (1950), 379
[89] Guyer, A., Guyer jr., A., and Karth Johnson, B., Helv. chim. Acta 38 (1955), 946 -954
[90] Mirb, A. R., de la G h d a r a , J. L. O., and GonzBles, J. R. A., -4n. Real. Soc. espan.
Fisica Quim., Ser. B 6 3 (1957), 171-184, 185-194
[91] Flirter, W. F., Chem. Engng. (1968), June, CE 173-Ce 177
[91a] Furter, W. F., Canad. J. chem. Engng. 66 (1977) 3, 229-239
[92] Schreiher, E., Schuttau, E., Rant, D., and Schuberth, H., Z . physik. Cheui. [Leipzig]
247 (1971) 1/2, 23-40
[92e] Schubert, H. and van Nhu, N., Z. phys. Chem., Leipzig, 258 (1977) 1, 106-112
[93] Fuchs, O., and Roth, F. H., Chem. Fabrik 11 (19:38), 401-406
[94] Ino, M.,et al., 5. chem. Engng. Japan 4 (1971) 1. 22-26
[94a] Sariiev, B. A., Pavlov, A. Yu., Bolkhova, L. A. ct al., Khim. prom. (1975) 11. 815-819
[95] Jaques, D., and Furter, W.F., Canad. J. chem. Engng. 60 (1972) 4, 502-503
[96] Furter, W.,.'!I and Meranda, D., GVC/,41 Che-J.. Joint Meet. 2 (1974) D 1-3. 1- 10
[97] Sada, E., Kito, Sh., Yamaji, H., and Kimura, hl.. J . appl. Chem. Biotechnol. 24 (1974)
4/5, 229-238

Chapter 7 - Constructional Materials and Apparatus

[l] Liedel, K. H., Glas in der Chemie - Laborgerate, CrroBapparate, Rohrleitungen. Otto
Krausskopf-Verlag GmbH, Mainz 1973
[2] Monch, G. Ch., Hochvakuumtechnik, Rudolf A. Lang Verlag, PoBneck 1950
[3] Eitel, W., e t al., Glastechnische Tahellen, Springer-Verlag, Berlin 1932
[a] VVB Westglas, Ilmenau, Technik 8 (1953), 675-680
[5] Rieger, H., Silikattechnik 4 (1953), 339-342; Chem. Techn. 22 (1970) 1, 36-40
[6] Pomeroy, H. H., Analytic. Chem. 24 (1952), 1076
[7] Kammel, H., Silikattechnik 2 (1951), 227-231
IS] Krell, E., Chem. Techn. 4 (1952), 200-207
[91 Catalogues of Glassware Manufacturers
Priedrichs, F., Das Glas im chemischen Laboratorium, 3rd Ed., Springer-Verliig:,
Berlin/Gottingen/Heidelberg 1960
Telle, W., ABC der Laboratoriumsgerate, VEB Fachbuchverlag Leipzig 1953
Telle, W.,Chemische Laboretoriumsgerate, 2nd Ed., VEB Deiitscher Verlag fur Orund.
stoffindustrie, Leipzig 1969
[lo] Stage, H., Erdol u. Kohle 3 (1950), 377-383
[ l l ] Krell, E., Chem. Techn. 4 (1952), 443-450
[12] Kinza, V., Registered design (GDR), S o . 10175
[13] Wolf, F. W. D., Glas-Instrumenteii-Techn. 4 (1960). 43-52, 126-131, 343-349
[la] Reckmann, C. J., Chem. Techn. 24 (1972) 9, 533-536; 25 (1973) 1, 20-23
[14a] Reichelt, W..Chemie-1ng.-Techn. 26 (1973) 1, 25-99
[lPb] Handl, R., Chem.-1ng.-Techn. 60 (1978) 1, 45
iranchembook.ir/edu

514 References

[15] Malewski, W., Chemie-Ing.-Tech. 42 (1970) 6, 385-390

[l6] Sigwart, K., Deatillation und Rektifiiation, in: Ullmanas Encyklopiidie der technischen
Chemie, Vol. 1. Urban & Schwarzenberg, Munchen/Berlin 1951
[17] Ray, J. D., Rev. sci. 28 (1957), 200-201
1181 Ckwney, Th. P.. Laboratory Fractional Distillation, MacMillan, New York 1949
[19] Thormann, K., Dechema-Erfehrungseustausch, Arbeitsmethoden und GerBte, Destillie-
ren nnd Rektifizieren, Dechema, Frankfurt/Main, Sep. 1950
[20] Fenske, M. R., h e r . Pat. No. 2037316 of 4 April, 1936
[21] Rose, ,4. and E., et al., Technique of Organic Chemistry, Vol. I\’, Interscience P u h -
lishers, New York 1951
[22] Kuhn, W., Chemie-1ng.-Techn.26 (1953), 12-18
Kuhn, W., et al., Chimis [Zurich] W (1954), 109-122
1231 Craig, L. C., Ind. Engng. Chem., analytic. Edit. Y (1936), 219; 9 (1937). 441
[24] Kuhn. W.,Helv. chim. Acta 05 (1942). 252-295
[25] Kuhn, W., and Ryffel, K., Helv. chim. Acta 26 (1943). 1693 to 1731
Jost, W., Angew. Chem., Part B 90 (1948), 231
Coenen. A., Thesis, T. H. Aachen 1941
Mair. B. J., and Willingham, Ch. B., J. Res. Nat. Bur. Standards 22 (1937). 5t9-527
[26] Westhaver. C. J.,Ind. Engng. Chem. 34 (1942), 126
[271 Donnell, C. K., and Kennedy, R. M., Ind. Engng. Chem. 42 (1050), 2327-2330
[28] Jantzen, E., and Wieckhorst, O., Chemie-1ng.-Terhn. 16 (1954). 302-396
[%I Koch, R., and Feind, K., Chemie-Ing.-Techn. 30 (1958), 577-584
[30] Beck, F., Chemie-1ng.-Techn. 36 (1963),837 4 4 4
[31] Fischer, W. G., Glas-Instrumenten-Techn. 13 (1969) 5,535-5540; 14 (t970) 8 , 933-936;
16 (1971) 7,753-759; 16 (1971) 8,832-838
Chemiker-Ztg. 94 (1970) 5, 157-161
Fette, Seifen. Anstrichmittel 70 (1970), 444-454
Fette, Seifen, Anstrichmittel 76 (1973), 131- 133
Chemie-Labor Betrieb 27 (1976) 2,42-46
[321 Nacura, H., and Grosse-Oetringhaus, H., Brennstoff-Chem. 19 (1938). 437-4439
[333 Manning. R. E., and Cannon, M. R., Ind. Engng. Chem. 49 (1957), 346-349
[341 Hoppe, K., and Mittehtrasa, M., Grundlagen der Dimensioniening von Kolonnenbdden
(Technieche Fortschrittsberichte, Vol. e l ) , Verlag Theodor Steinkopff, Dresden 1967
[35] Stage, H., Glw-Instrumenten-Techn. 13 (1969) 12, 128ci--1299
Stage, H., Chemieanlagen und -7erfahren (1970), No. 1/2
[35a] Hutla, V., and Strnal, M., Chem. Techn. 28 (1976). 546-547
[36] Beesom, W. H., e t al., Physics 1 (1934). 1109-1114
(371 Bruun, J. G., and Sehicktanz, S. T., J: Res. Nat. Bur. Standards 7 (1931), 851-882
C381 Stage, F., Angew. Chem., Part B 19 (1947), 182
[39] Stage, H., DBP No. 862598 of 4 July, 1950, granted 20 Nov., 1952
Kirschbaum, E., Chernie-1ng.-Tech. 26 (1953), 73-80
1401 Fritz, P. W., Glas-Instrumenten-Tech. 11 (1967) 3, 161- 162; Registered design (FRG)
No. 1957085
[41] Oldershaw, C. F., Ind. Engng. Chem., analytic. Edit. 18 (1941), 265-268
[42J Hilberdth, F., Oel u. Kohle 39/40 (1943), 875-880
[gal Wijk, W.R. van. and Thissen, H. A. C., Chem. Engng. Sci. 3 (1954), 153-160
[443 Umholtz, C. L.. and Winkle. M. vsn, Petroleum Refiner 34 (1955), 114-117
[45] Jones, P. D., and Winkle, M. van. Ind. Engng. Chem. 49 (19571, 232-238
[46] Developed in the Research Laboratory of Messrs. C. F. Bram and Co.; Manufacturer:
Messrs. Emil Greiner and Co.
[47] McGure, G. V., and Maddox, R. N., Ind. Engng. Chern. Proc. 6 (1971) 1, 9
[48] Wagner, C. D., Lukens jr., H. R., and Hughes, R. R.. Nucleonics 14 (1956). 78-83
[49] Newitt, D. M., Dombrowski, N., and Knelman, F. H., Trans Instn. Chem. Engr. 32
(1954), 244-261
[SO] Melikyan, R. A, J. appl. Chem. (Russim) 31 (1958),550-558
iranchembook.ir/edu

References 515

[51] Widmer, G., Helv. chim. Acta 5 (1924), 194-196

[52] Golodets, A., Chem. Ind. (Russian) 35 (1912), 102-108, 141-145
[53] Leeky, H. S., and Ewel!, R. H., Ind. Engng. Chem., analyt. Edit. 14 (1942), 503 - 547
[54] Stallcup, W. D., e t al., Ind. Engng. Chem., analyt. Edit. 14 (1942), 503-547
[ 5 5 ] Bower jr., J. R., and Cooke, L. M., Ind. Engng. Chem. 15 (1943), 290-293
[56] Podbiclniak, W. J., Ind. Engng. Chem. 13 (1941), 639-645
[57] Brezina, V., Glas- Instrumenten-Techn. 14 (1970) 5, 493-494
[5X] Stedman, D. F., Trans. Amer. Inst. chem. Engr. 33 (1937), 153-161
[59] Koch, H., a n d Raay, H. van, Chemie-1ng.-Techn. 8 (1950), 172
[GO] Bragg, L. B., Ind. Engng. Chem. 46 (1953), 1676
[el] DR,P, according t o prospectus from Messrs. Ernst Haage, Miilheim/Ruhr
[621 Ellis, S. R. M., a.nd Varjavandi, J., Chem. Process Engng. 39 (1958). 239
[63] ,Juchheim, G., Glas-Instrumenten-Techn. 10 (1966) 10, 914-917
[64] Sperandio, A., Richard, M., and Huber, M., Chemie-1ng.-Techn. 97 (196.5) 3, 322--328
Zogg, M., Chemie-Ing.-Techn. 46 (1973) 2, 67
[ ( i l a ] Sulzer techn. Rdsch. 59 (1977) 3, 138
I M b ] Bonio, P., Chem.-1ng.-Techn. 49 (1977) 11. 895-897
Xeier, W., Hunkeler, R. and Stticker, W.-D., Chem.-1ng.-Terhn. 51 (1979) 2, 119-122
[64c] Meier, W., Sulxer techn. Rdsch. 61 (1979) 2, 49-61
[63] Timofeev, A. V., and $erov, M. E., Chem. Ind. (Russian) 5 (1965), 381-383
[ M a ] Kwasniak, J., Verfahrenstechnik H (1974) 11, 315-318
[CiSb] Endfst, M., Cervenka, J., and Kblar, V., Coll. Czech. chem. Com. 41 (1976) 9,2676-2681
[ti5c] Timofeev, A. V., Aerov, M. E., and Gursky, M. M., khim. prom. (1977) 2. 140-142
[661 Shavoronkov, N. M., and Malyusov, V. A., Chem. Techn. 24 (1972) 7, 397 -401
[GGa] Billet, R.., Maekowiak, J. and Suder, St., Chem.-1ng.-Techn. 60 (1978) 7, 850-551
[A71 Zuiderweg, F. I., Chem. Engng. Sci. 1 (1952), 164-174, 184-194
[HHI Podbielniak, W.J., D R P No. 595250 K1. 12a of 18 Bug., 1931, granted G April. 1934.
[G9] Benner, F. C., Dinardo, A., and Tobin, D. J., Ind. Engng. Chem. 43 (1951), 722-727
[70] Byron, E. S., Bowman, J. R., and CouI1, J., Ind. Engng. Chem. 43 (1951), 1002-1010
[711 Willingham, C. B., Sedlak, V. A., Rossini, F. D., and Esthaver, J. W., Ind. Engng.
Chem., ind. Edit. 39 (1947), 706-712
[??I Jost, W.,Chemie-Ing.-Techn. 26 (1953), 356
Jost, W., Rektifika.tion unter extremen Bedingungen, in: Dechema-Monogr. 2‘1 (1964),
30 - 72
[73] Jost, LT., Sieg, L., and Brandt,, H., Chemie-1ng.-Techn. 25 (1953), 291-292
[74] Gelperin, N. W.,and Khatsenko, M. S., J. appl. Chem. (Russian) 25 (1952). 610--615
[75] Murray, K. E., J. Amer. chem. Soc. 28 (1951), 235
\:76] Irlin, A4.L., and Bruns, B. P., J. analyt. Chem. (Russian) 5 (1950), 44-47
[77] Hoher, G . , Untersuchungen iiber die Trennfahigkeit einer Labor-Rektifikationskolonne
mit rot>ierenderWendel als Austausrhflache, Examination paper, Ingenieurschule fiir
Chemie, IApzig 1958
[is] Neumann, F.. Chemie-1ng.-Techn. 33 (1961) 7, 485-491
[78a] Mamin, V. S . ,e t al., khim.-farm. Zh. 8 (1974) 3, 51-55
[79] Toeldt,e, W.,Seifen-Ole-Fette-Wachse74 (1948), 126
[80] N. N., Glas-Instrumenten-Techn. 10 (1966) 4, 259-260
[81] Kramer, M. T.. Glas-Instrumenten-Techn. 15 (1971) 8, 856-860
[82] Brezina, V., Glas-Instrumenten-Techn. 10 (1966) 3, 162- 172
[831 Friedrichs, F., Z. angew. Chem. 33 (1920), 29-32
[84] Friedrichs, J., and v. Kruska, H., Chem. Fabrik 7 (1934), 284-287
[85] Mach, F., and Herrmann, R., Chem. Fnbrik 6 (1933), 211-213
[86] Eichhorn, E., Untersuchiing der Wirkung von Laboratoriumsk~lilern,Esaminat,ion
paper, Ilmenau 1955
[8T] Gemmeker, L., and Stage, H., Gla,s-Instrument,en-Techn. 6 (1962) 3,69-78; 1.114- 119;
5,169-174; 6,206-208; 9,327-32s; 10,372-376; 12,494-497
[ 8 8 ] Krell. E.. Chem. Techn. 5 (1953), 581--587
iranchembook.ir/edu

518 References

[89] Schneider. K. W.,and Schmid, Ch., Erdol u. Kohle 2 (1949), 508-510

[89s] Kloss, H.-J., CZ-Chem.-Techn. 3 (19741, 89-93
[W] Kramer, 31. T., Glas-Instrumenten-Techn. 16 (1972) 4, 445-448; 5, 571-574; 8,
943-946
[91] Prospectus from Xormug, Hofheim/Ts., List No. 10; Chemie-1ng.-Techn. 26 (1954)
231
[92] Srhulz, E. O., Glasmstrumentenkunde, Wagner u. Sohn, Weimar 1937
[93] Hdbner, G . . Chemiker-Ztg. 88 (1959). 581-583
[94] LI3y-d. L. E., and Hornbacher, H. G., Analytic. Chem. 19 (1947), 120-122
[95] Collins, F. C.. and Lantz, V., Ind. Engng. Chem., analyt. Edit. 18 (1946), 673-677
[96] Kieselbach, K.. -4nalytic. Chem. 19 (1947), 815
[97] Stage, H.. itild Gemmeker, L., Crlaa-Instrumenten-Techn.6 (1962) 12, 494-497
[$8] N. N.,Glas-In&rumenten-Techn. 10 (1966) 3, 178-179
[W]Stage, H., and Cemmeker, L., Glas-Instrumenten-Techn.7 (1963) 11, 608-614; 12.
657-7704: S (1964) 6, 413-428; 7, 503-508
[1(W1 Munnfacturer: Glaewerk Schott und Gen., Mainz
[loll Rock, E. J.. and Jantz, G. F.. Analytic. Uhem. t22 (1950). 626
[lo21 Grsssmann, W.,and Deffner, C., Chemiker-Ztg. 76 (1952), 623
[lo31 Sclirarn, E., ilnalytic. Chem. 26 (1963), 1424
11041 Stage. H., and Gemmeker, L., Glas-Lnstrumenten-Tech. i (1963) 11, 608-614; 12,
687-704; 8 (1964) 6, 413-428; 7, 503-508
Stage, H.. Glas-Icstrumenten-Techn. 14 (1970) 3, 213-220; 4, 337-349
[lo53 Hubner, (;., Chemiker-Ztg. 76 (1952), 67
[ 11IS] ZIIannfwturer : Edniund Biihler, Tiibingen
[lIU] Manufacturer: Dr. Hofmann and Roth. Stuttgart-Hedelfingen; cf. also Krell, E., Chem.
Tt%rhn.7 (1965), 599-608
[lOS] h n u f a c t u r e r : U’.C. Heraeus GmbH, Hanau
[log] Rorchert. It., and dubitx, W., Irifrilrotatrahler z i u Erwarmung, Trocknung und Aux-
hartung. \-EB Verlag Technik. Berlin 1951
[110] Klees, K.F., Chem. Rdsch. [Solothurn] 7 (1954), 358-359
[l I t ] Schulze. H., Chemie-1ng.-Techn. 23 (1951), 565-569
[112] M’eber, I-. von, Chem. Techn. 1 (1949), 159-160
[113] Zampachovit. L., Chem. pdimysl. (Czech) 7/32 (1957), 408-413
[114] Kolling, H., and Tramm, H., Chemie-Ing.-Techn. 81 (1949), 9-14
[116] Manufacturer: Bronwill Scientific Ind., New York
[116] Kortum, (4.. and Bittel, A, Chemie-Ing.-Techn. $8 (1952), 40-44
11171 Rose, A. and E.. et al., Technique of Organic Chemistry, Vol. IV, Interscience Publi-
1 shers, Sew Tork 1951
El181 Mtinufactnrer: W.C. Heraeus GmbH, Hanau, Electrothermal Engineering Ltd.,
London, aiid VEB Xsbestdraht, Berlin
[118a] Sellin, J..chemir-anlagen + verfahren 4 (1976), 87-91; see also [66a]
[rig] Exclusi\ e right of sale for “Calorex”: Heraeus-Wittmann GmbH, Reidelberg
[I201 Leva, >I., Tower Packings and Packed Tower Design, 2nd Ed., The United States
Stoneware Company, Akron/Ohio 1953
11311 B e r h p i , E.. and SchultheiB, I., Proc. and Conf. on Appl. Phys. Chem., Veszprhi,
Aug. 2-5, 1971, Akademiai Kiado, Budapest 1971, Vol. 2, pp. 67-74
[121n] Reichelt. M’., Chemie-1ng.-Techn. 46 (1973), 2, 54-57
[ 1221 Manufacturer: Podbielniak Inc., Chicago 11
[I231 Bmun, R.. Chemie-1ng.-Tech. .33 (1961), 350-3352
[124j Dixon. 0. G., J. SOC. chem. Ind. 68 (1949), 88, 119, 299
[1%] Huber, BI., and Sperandio, A.. Chemie-Ing.-Techn. 36 (1964) 3, 221-227
[ i l i j t t ] Zelrenskii, Yn. D., Kovalenko, A. E., Mukhina, T. V., and Panfilov, V. I., khim.
prom. (1977) 10. 783 - 786
[Inti] Ellis, 8. R. M., and VarJavnndi. J., Chem. Process Engng. 7 (1958), 239-243
iranchembook.ir/edu

References 51 7

Unpublished measuring results of the Institut fur Apparatebaa und Verfnhrens-

technik, TH Karlsruhe (personal communication)
MncNahon, H. O., Ind. Engng. Chem., ind. Edit. 39 (1947), 712-718
Manufacturer: The United States Stoneware Company, BkronlOhio
Manufacturer: Hydronil Ltd., Chemical Plant Division, Fenton/England
Kirschbanm, E., Z. Ver. Dtsch. Ing. 98 (1956), 1801-1802
Alt, C., Meier, A., and Richter, G., Chemie-1ng.-Techn. 27 (1955), 502-506
Manufacturer: Vereinigte Fiillkorperfabrik GmbH 11. Co., Banmbach/Westerwald
Fuchs, O., Chemie-1ng.-Techn. 23 (1951), 537-540
Forsythe, W. L., e t al., Ind. Engng. Chem. 39 (1947), 714
Wolf, F., and Gunther, W., Dechema-Monogr. 29 (1957), 364-377
Krell, E., Chem. T e c h . 16 (1964) 10, 591-600
Krell. E., Plaste und Kautschuk 14 (1976) 4, 251-254
Teller, A. J., Chem. Engng. Progr. 60 (1954) 2, 65-71; Ind. Engng. Chem. 50 (1958)
8, 1201-1206
N. N., Cheni. Engng. 80 (1973) 19, 88
Egberongbe, S. A., Process Engng. (1973) 2, 82-83, 85
Stiirmann, H., Leistungssteigerung von Fullkorperkolonnen durch nene Wandforinen.
Thesis, Hamburg 1937
Krell, E.,Patent (GDR) No. 63761. Kl. 12a8. granted 20 Sep., 1968, “Column wit11
reduced wall flow-”
Laukner, E.,Wichmann, H., and Stabler, U., Cheni. T e c h . 28 (1976) 12, 718-721
Ponter. A. B., Shaver, C., and Trauffler, P., Chemie-1ng.-Techn. 48 (1976) 7, 635--63(;
Ponter, A. B., Taymour, X., and Dankyi, S. O., Chemie-1ng.-Techn. 48 (1976) 7 .
635 638
-

Chapter 8 - Automatic Devices; Measuring and Control Equipment

[l] Rolfson, F. B.,e t al., Analytic. Chem. 20 (1948), 1014-1019
[2] Baperl. V.. and Quarg, M., Taschenbuch des Chemietechnologen, 3rd Ed., VEB Deut-
scher Verlag fur Grundstoffindustrie, Leipzig 1968
[3] Kanterman, 1,. B., et. al., Chem. Technol. Brennstoffe Ole (Russian) 8 (1958), 57.- t i 1 1
[4] Salzer, F., Chemie-1ng.-Techn. 31 (1989). 405-409
[.5] Jirmann, R.. Glas-Instrumenten-Techn. 1 4 (1970) 1, 11-12
[6] Stage, H., CZ-Chemie-Techn. 1 (1972), 263-272
[i] Za,ppe, D., Heckmann, M., Gutwasser. H., and Dornbusch, E., Heckrnann-Mitt .
No. 5 (1968)
[ X I U l ~ s o y ,E., Chemiker-Ztg. 59 (l955), 46-48
191 Stage, H., Gemmeker, L., Fischer, G., and Richter. R., Glas-Instrumenten-Techn. 5
(1961), 138-145 and 204-211
[lo] Fischer, W. G., and Wepand, H., Glas-Instrumenten-Tech. 14 (1970) 1, 23-26
1111 Krell, E., Chem. Terhn. 9 (1957), 266-274
[12] Kortiim, Q., and Bittel, A., Chemie-1ng.-Techn. 18 (195G),40-44
[12a] Iioukol, H., Kupka, H. J., and Henning, G . , GIT Fachz. Lab. 19 (1975), Special
Sumber Labormedizin, 493-497
[13] Brandt, H., a n d Rock, H., Chemie-1ng.-Techn. 25 (19531, 511-514
Urandt, H., Rock, H., and Langers, F., Chemie-1ng.-Techn. 29 (1957), 86-91
[14] Abegg, H., Chimia [Zurich] 6 (1952), 258-267
[1.51 Laporte, H., Messung, Erzeugung und Konstant.haltung hoher bis tiefer Temperaturru.
VEB Fachbuchverlag Leipzig 1961
[l6] Kossler, I., and Vodehnal, J.. Chem. Prbmysl. (Czech) 7 (32) (1957), 292-297
[17) Iiadlec, &I..Chem. Listy (Czech) 61 (1957), 987-990
[17 a] Stage, H., CZ-Chemie-Technik 5 (1973), 203-208
[17 b] Nitteilung der Fischer Labortechnik in GIT Fachz. Lab. 21 (1977) 4, 392
[IS] Vanvor. H., (;las-Instr~unenten-Techn.8 (1964), 415
iranchembook.ir/edu

518 References

Ll9] Telle, W., Chemische Laboratoriumsgerate, 2nd Ed., VEB Fachbuchverhg Leipzig
1969
[20] Piatti, L., and Marti, W., Chemie-Ing.-Techn. 28 (l956), 473-475
[2l] Manufacturer: Siemens AG, Erlangen - Dr. Neumann GmbH, Miinchen 80
[22] Lijffler, H. J., and Henrici, H., Chemie-Ing.-Tech. SO (1958), 718
Soble, A. B., Chem. Engng. Progr. 56 (1967), 444-451
[23] Xanufacturer : VEB Thermometerwerk, Geraberg
[24] Manufacturer: Juchheim KG, Solingen
[%I Kleinteich, R., Glas-Instrumenten-Techn. 3 (1959), 89-93
[26] Walter, M., Relaisbuch, Franck’sche Verlagshandlung, Stuttgart 1951
[27] Fischer, W., Glas-Instrumenten-Techn. 14 (1970) 5, 513-516
[%I Thormann, K., Dechema-Erfahrungsaustmmch,ilrbeitsmethoden und Gerlte, Destillie-
ren und Rektifizieren. Dechema, Frankfurt/Main, Sep. 1950
[as] Hutla, V.. Chem. Listy (Czech) 61 (1957), 1964-1965
1:30:) Jaeckel, R.. Dechema-Erftbhrungsa.ustausch,Vakuumtechnik, Dechema, Frankfurt/
Yain 195ti
[31] Leck, J. H., Presdure Measurement. in Vacuum Systems, The Institute of Physics,
London1957
[32] Nickel, E., Gas- und Wilsserfach, Beilage Bau und Betrieb 6 (1954), No. 5 (Edit. Gas)
[34] Farqtiarson, J.. and Kermirle, H. A., Re)-. sci. Instr. 88 (1957), 324-325
[34] Dosch, R., Glas-Instrumenten-Techn. 4 (1960), 9- 17
r3.51 Holland-Merten, E. L.. Ta.bellenbuch der 17akuumverfahrenstechnik in der Grund-
st,offindustrie, VEB Dsutscher Verlag fiir Grundstoffindustrie, Leipzig 1964
[:%I Hohlschiitter, H. W., and Kampf, G., Angea. Chem. 66 (1956). 355
[3Ga] Xmufactiirer : Airflor-Lnftterhnik GmbH. Rheinba.ch
[371 Crompt,on, R’. W., and Elford, M. T., J. sci. Instr. 34 (1957), 405-407
[:Hl Stiirzba.rh, W., Glas-Instromenten-Techn. 14 (1970) 1, 36-37
[39] Eirell, E.. Chem. Techn. 3 (1951), 333-336
[40] Gilmont, R., Ind. Engng. Chem., analyt. Edit. 18 (1946), 633-636
[-Ill Manufacturer: Dr. Friedrichs- Dr. Matachke KG Normschliff-Glasgerite, Wert,heini/
Main
[42] Young, C. G.. et al., Canad. J. Techno]. 29 (1951). 447-450
1431 Krell, E., Chem. Techn. 10 (1958), 327
[Wl Coulson. E. A., and Warne, R. J., J. sci. Instr. 21 (1944), 122.-123
[W] Monch, G . C., Hochvakuumtechnik, Rudolf A. Lang Verlag, PoBneck 1950; Neues
und Reaiihrtes aus der Hochvakuumtechnik, VEB Wilhelm Knapp Verlag, Halle 1959
[41iI Hoch. H.. Vakuum-Techn. 16 (1966) 1/2, 8-13
1471 Steckelmacher, W., J. sci. Instr. 1 (1951), 10-19
1481 Gaede, W., Z. techn. Physik 16 (1934), 664
1491 Jaeckel. R., Kleinste Driicke, ihre Messung und Erzeugung, in: Technische Physik in
Einzeldarstellungen, Springer-Verlag, Berlin/GWingen/Heidelberg 1950
Laporte, H., Hochvakuum, seine Erzeugung, Messung und Anwendung im Laborato-
rium, VEB Wilhelm Knapp Verlag, Halle 1951 ; Vakuummessungen, VEB Verlap
Technik. Berlin 1955
Yarwood, J., Hochvakuurntechnik, Rudolf A. Lang Verlag, Berlin 1955
Leybold, Vakuum-Taschenbiich. ed. by K. Diels and R. Jaeckel, 2nd Ed., Springer-
Verlag, Berlin/Gottingen/Heideiberg1962
Frank, W.. and Kutsche, D., Die schonende Destillation, Otto Krausskopf-Verlag
GmbH, Mainz 1969
[50] Niiller, B., Glas-Instromenten-Techn. 14 (1970) 7, 798- 807
[Sl] Ebert. H., Kompressionsvakuummeter, Friedr. Vieweg + Sohn GmbH, Braun-
schweig 1951
[.XI Peche, G., Glas-Instrumenten-Techn.7 (1963) 1, 4,7, 48-49
[533 Manufacturer: Rudolf Brand, Wertheim/M., Glashiitte
[54] Xishet, J. S., J. sci. Instr. 86 (1949), 271-273
iranchembook.ir/edu

References 519

[%I Nelpolder, F. W., Ind. Engng. Chern. 39 (1957), 617

[56] Fischer, H. E., Analytic. Chem. 20 (1948), 982
[57] Gemmeker, I,., and Stage, H., Glas-Instrumenten-Techn. ti (1962) 12, 494 -497; i
(1963) 1, 18-23
[58] Stage, H., Chemie-1ng.-Techn. 22 (1950),374-375
Stage, H.. GIT Fachz. Lab. 19 (1975) 3, 182
[ S a ] Fieberg, D.. CAV (1977) 3, 64-65
[691 Houben-Weyl, Methoden der organischen Chemie. ed. by F,. Niller, Physikn lisclie
Methoden, Part, 1 and 2, Oeorg Thieme Verlag, Stuttgart 1955
[GO] 3~uIlsr.R. H., and Zenchelsky, S. T., Analytic. Chem. 24 (1952), 844-847
[6l] Thomas, G. R., e t nl., Analytic. Chem. 22 (1950), 1221-1223
[62] Latchnm, J. W.. Amer. Pat. No. 2529030 of 7 Nov., 1950
[63] Nehe, W., Riegler, H., and Menzel, W.,Jenaer Rundschau 18 (1973). Special Lssur,
87 - 89
[64] Kegeles, G., and Sober, H. A., Analytic. Chem. 24 (1952), 654-660
1651 Oehme, F., Chemische Analyse durch Messung von Dielektrizitats-Konstirntcn. Veda;
VEB Laborchemie. Apolda 1953
Oshme. F., Dielektrische MeBmethoden, 2nd Ed., Verlag Chemie GmbH, \Veinlirimi
Bergstr. 1962
[66] Slevogt', K. E., Dechemn-Monogr. 14 (1950), 97-120; 17 (1951). 95-107
[G7] Grant. R. A., J. appl. Chem. 8 (1958), 136; Chemie-1ng.-Techn. 30 (1958). 738
(681 Felloni, L., and Tantillo, E., Ann. Chimica 47 (1957), 751-758
[CiStt] Lertes, E., and Niel, F., GIT Fachz. Lab. 93 (1979) 5, 376
[GY] Clasen, H., Chemie-1ng.-Techn. 28 (1956), 45-48; 29 (1957), 110-111
[70] Linford, A., Ind. Chemist. 34 (1958) 403, 481-488
[71] Fritzsche, R. W., Petroleum Processing 7 (1952), 1138-1143
[72] Koch, O., Chemie-1ng.-Techn. 30 (1958), 181-187
Kienitz, H., Chemie-1ng.-Techn. 35 (1960) 10, 641 -650
[73] Siggia, S., Continuous Analysis of Chemical Process Systems. J o h n 1Vile.i ff Sons.
London1959
[74] Chemie-1ng.-Techn. 36 (1964) 9 8 , 1455- 1457
[75] Herforth, L., and Koch, H., Praktikum der angewandten Radioaktivitst. 2nd Ed..
VEB Deutscher Verlag der Wissenschaften, Berlin 1972
[75al Zelvenskij, Ya. D., Malinov, 8. A., and Shalyzin, V. A.. teoret. osnovy khim. tekhnol.
10 (1976) 2, 184-189
[76] Bernhauer, K., Einfuhrung in die organisch-chemische Laboratoriumsteclinik. Springci,-
Verlag, Vienna 1947
[77] Wittenberger, W., Chemische Laboratoriumstechnik, 6th Ed., Springer-Vcrlap.
Berlin/Gottingen/Heidelberg1963
[7S] Muller, G., and Gnauck, G., Reinste Gase, VEB Deutmher Verlag der Wisscnschiiften.
Berlin 1965
[79] Hinzpeter, G., Valinum-Techn. 11 (1962), 2
[8tj] Xanufacturer: Fischer, Labor- 11. Verfahrenstechnik, Bonn -Bad Godesbrrg
Schulze, F.-W., Hammerschmidt, H., and Camnienga, H.-K., GIT Fachz. Lab. I I
(1977) 12, 1033
[81] Dammer, H. J., Ghs-Instrumenten-Techn. 15 (1971) 2, 115-116
[82] Hanel. R., Chem. Techn. 15 (1963) 3, 176-177
[83j Pinkava, J., and Wiehterle, O., Collect. Czechoslov. chem. Comrnun. 90 (1955).B97-iU7
Pinkava, J., and Kaltib. V., Collect, Czechoslov. chem. Commun. 20 (1955). 707-712
[84] Pinkava, J., Laboratory technique of continuous chemical processes (Czech). Kakla-
datelstvi technicki: Literatury (SNTL), Prague 1958
[85] Schmidt, W., Chem. Techn. 10 (1958), 131-134
[86] Heusch, R., Pette, Seifen, Anstrichmittel 72 (1970) 11, 969-977
[8Ba] Engler, J., and Kuhnlenz, H., Chem. Techn. 31 (1979) 5 , 251-254
iranchembook.ir/edu

320 References

Pumpen fiir Fliiesigkeiten, VEB Verlag Technik, Berlin 1970

[87] Pohlenz, ITT.,
($81 Liedel, K. H.; a l a s in der Chemie - Laborgeriite, GroSapparate, Rohrleitungen,
Otto Krausskopf-Verlag GmbH, Mainz 1973

Chapter 9 - Arrangement of a Distillation Laboratory;

Starting U p Distillation Apparatus
h h r e , A., Chemisch-physikalische Laboratorien und ihre neuzeitlichen Einrichtungen,
-kademische Verlagsgesellschaft Geest & Portig, Leipzig 1950
Colemann, H. S., et a]., Laboratory Design, Reinhold, New York 1952
Schramm, W., Chemische und biologische Laboratorien, 3rd Ed. Verlag Chemie GmbH,
Weinheim/Rergstr. 1969
Kincannon, C. B., and Baker, hf. 0.. Analytic. Chem. 29 (1957), 1189-1193
Schwarz. H., Stative, Aufhiingung von Apparaturen, in: Houben-Weyl, Methoden
der organischen Chemie (ed. by E. Miiller), Vol. 1/1, Georg Thieme Verlag, Stuttgart
1958
Manufacturer: Ernst Hctage, Miilheirn/Ruhr
Xanufacturer : Geru, Geroldwill, Switzerland
Brooke, M., J. ehem. Educat. 28 (1951), 602-603
Wagner, G., osterr. Chemiker-Ztg. 43 (1940), 229
Bernhauer, K., Einfiihrung in die organisch-chemische Laboratoriumstechnik, Springer-
Verlag, Vienna 1947
Wittenberger, W., Chemische Laborcttoriumstechnik, 7th Ed., Springer-Verlag,
Berlin/Heidelberg/Sew York 1973
Wolf, F. W. D., Glas-Instrumenten-Techn. 4 (1960), 34-52, 126-131
Friedrichs, F., Das Glas im chemischen Lsboratorium, 3rd Ed., Springer-Verlag,
Berlin/Gijttingen/Heidelberg1960
Brauer, H., Chemie-1ng.-Techn. 28 (1957), 520-530
Greinacher, H., Ausgewahlte Aufgaben und Probleme aus der Experimentalphysik
(I 28: Der Siedeverzug), Springer-Vrrlag, Vienna 1953
dmer. Pet. No. 2680880, issued 1 Jan., 1952; No. 2580923, issued 1 Jan., 1962
Hubner, G., Chemiker-Ztg. 76 (1951), 229-230
The General Safety Committee of the Manufacturing Chemists’ Association, Guide
for Safety in the Chemical Laboratory, Van Pu’ostrand. S e w York 1954
Ditt,mnr, P., Erdol u. Kohle 11 (1958), 25-30
Craig, 1,. E., and Dew, J. E., Ind. Engng. Chem. 51 (1959), 1249-1252
Kulbe, B., Hoppe, K., Kunze, R., and Noack, J., Chem. Techn. 23 (1977) 11,600-603
iranchembook.ir/edu

Subject index

activity coefficient 90, 317. 329 columns

adapter 343 bubble-cap 29, 2.53, 358
additive 323, 329, 334 bubble-plate 358
adiabatic operation 163 concentric-tube 352
air condenser 383 for falling film distillation 46
air thermometer 448 gauze-disc 137
amount of reflux 156 Jantzen 25, 350
analog computer 201 low-temperature 232, 263, 435
aniilytical distillation 208, 210 multi-tube 239, 351
Antoine equation 68 packed 46, 128, 356
ASTM method 431 plate 29, 44, 104, 128, 217, 358
automatic column head 393 rotating condenser 275
automatic distillation apparatus 432 sieve-plate 177, 358. 362
azeotrope formation 313, 323 spinning band 374
azeotrope with maximum boiling point 79, Vigrenx 206, 348, 389
313 wetting 46, 365
azeotrope with minimum boiling point 79, with stAtionary or rotating elements 46
313 with wire-gauze inserts 365, 367, 371
azeotropic data 313 column comesion 143, 245
azeotropic distillation 55, 315 column diameter 48, 145, 177
azeotropic point 319 column dynamics 55
column efficiency 105, 129. 151
ball joints 35 column head 258, 333, 390
barometer reading 189, 449 column testing 144, 162, 168
batch distillation 45, 108, 193 components 340
Bennet manometer 450 compression manometer 459
binary mixture 79 concentration 39, 93
boiling capillary 341, 343, 490 concentric-tube column 352
boiling point analysis 213, 433 condensation 25, 215, 256. 385
boiling point correction 189 condenser 25. 44, 382
boiling point diagram 80, 254 contact manometer 466
boiling point difference -45, 120, 330 contact thermometer 445
bottom heating 411 continuous distillation' 29, 45. 183. 197, 246,
bubble-cap column 29, 253. 358 251
bubble-plate column 358 correction 188
Cottrell pump 94
countercurrent distillation 37, 44, 131, 183
calibration 163, 189, 441, 450 Cox diagram 70, 74, 320
carrier vapour distillation 55, 307 cylindrical packing units 422
channelling 48, 129
rirculating-type apparatus 220
circulation pump 475 dead space liquid volume 50. 21.2
classification of mixtures 321 degasser 298
clips 485 degree of mixing 49
coil-shaped condenser 385 degree of purity 139
iranchembook.ir/edu

*22 Subject index

degree of wetting 51 equilibrium thermodynamic consistency 91,

dehydration 282, 323 101
dephlegmation 218. 256 extractive distillation 316
dephlegmator 256, 385 evaporator 284
diagram 187, 191, 192
djaphragm manometer 452 feed-temperature 114
dielectric constant 470 Fenske equation 125, 235
differential curve 118 . filling of column 162
distillate withdrawal 213 film formation 54, 285, 289
distillation flash distillation 281
analytical 208, 210 flwk 337, 398
azeotropic 55, 315 flask heating 405, 465
batch 45, 108, 193 flooding 165, 275
carrier vapour 55, 307 flow factor 181
continuous 29,45,183, 197, 246, 251 flow meter 472
countercurrent 37, 44, 131, 183 fraction collector 403
Engler 32, 45, 337 fractional distillation 245
extractive 316 fractionating factor 121
flash 281 frameworks 482
fractional 246 free volume 175
high-temperature 269 f ully-automatic apparatus 435
isothermal 272
low-temperature 231, 261 gauze-disc column 137
molecular 207, 292 glass tap 278, 307, 344, 488
pressure 305 glass thermometer 189, 440
short-path 295
simple 37, 44, 183, 339 hairpin countercurrent principle 349
solution 344 heat calculations 182
stmdard 122 heat requirements 183
ateam 55, 307 heat transfer coefficient 185, 387
thin-film 283 heat of vaporization 106
vacuum 198, 205 heating jacket 415, 448
distiilcttion areas 318 heat.ing system 23, 405
distillation control device 464 henting tape 417
distillation curve 80, 254 heavywater 232
distillation laboratory 27, 32, 479 HETP-value 105, 134, 151
distillation test 248 high-temperature distillation 269
distillation thermometer 441 high-vacuum grease 488
dropping funnel 474 HTU-value 134
Duhring’s rule 67 hydrogen bond 314, 321

e I)ullionieter 60 immersion heater 409

vliniiii&ion curve 304 inflection point 115
Enplerdistillation 32, 45, 337 insulation 25, 185, 412
riitlialpy-concentrationdiagram 136 intensity 136
introduction of packing 147
equation of state 76
isoteniscope 64
equilibrium apparatus 93, 97 isothermal distillation 274
cquihbrium curve isotopes 228, 242
additives 323, 330, 334
calculation 82 Jantzen column 25, 350
cotme 57, 79, 92 Jordal rail 482
experimental determination 91
gr,iphical determination 84 Kapsenberg grease 488
prcwuw dependence 319 key component 141
iranchembook.ir/edu

Subject index 523

Liebig condenser 383 piping 276, 342

limiting velocity 38, 179 plate column 29, 44, 104, 128, 217, 358
liquid distribution 130, 145 plate efficiency 250
liquid flow density 49 plate equivalent 154, 166
load 130, 160, 466 pole height 102
low-temperature column 232, 263, 435 preheater 215, 410
low-temperature distillation 231, 261 pressure 38
low-temperature micro-column 265 pressure differential over column 167
lubricant 488 pressure distillation 305
pressure drop 50, 167, 177, 276, 378, 381
Iualdistribution 48, 130 pressure regulator 454
manometer 450, 458 product.ion of reduced pressure 279
AMcCabe-Thielediagram 107, 108, 111, 132
measurement of quantities 472 radiometric methods 472
melting point 468 Raschig ring 422
mercury 272 rate of evaporation 292, 465
metal dist,iIlation 272 reboiler charge 255
metering 219, 222, 472 receiver 401
inet>hodof increments 66 reflux ratio 105, 110, 117, 124, 152, 463
micro-technique 203 refractive index 93, 468
milligram quantities 203 refractonieter 469
minimum plate number 121, 125, 144, 267 resistance thermometer 444
minimum reflux ratio 108, 116, 127, 267 resistance t'o transfer 129 .
miscibility 55 ring-type packing units 421
mixed-film drum 288 Rose formula 123
molecular distillation 207, 292 rotating condenser columii 275
molecular distillation apparatus 209, 291 rotating cylinder 377
multicomponent mixture 87, 140, 143, 201, rotating evaporator 291
210, 249, 331 running-in time 136
multi-tube column 239, 351
saddle packing 173, 425
narrow-boiling mixture 314, 317 salt effect 335
nomograms sampling 93, 163, 261
barometer correction 189 separating efficiency 102, 298
concentration calcuIat,ions 41 separating length 48, 146
Cox diagram 73 separation factor 87, 305, 314, 324
equilibrium curve 86 separation parameter 236
limiting velocity 179 short-path distillation 298
minimum plate number 121, 268 sieve-plate column 177, 358, 362
separation factor 89 simple distillation 37, 44, 183, 339
t,emperature correction 190 SI units 37
vacuum lines 277 slotted angle iron 484
non-cylindrical packing units 425 solubility 55, 78
solut,ion distillation 334
operating hold-up 136, 157 specific stage number 159, 161, 173
operating line 106, 108, 109 spinning band column 374
operating pressure 159 spiral 172, 422
stage efficiency 145
packed column 46, 128, 356 stand for apparatus 480
packing inserts 366 standard distillation 122
packing units 51, 168, 173, 420 standard ground-glass joints 32, 278. :$a::.
Pall ring 422 489
partial condensnt>ion 259 standardized apparatus 32, 222, 483. :Iss.
partial pressure 78, 308 432
pilot plant 219 static hold-tip 50, 157
iranchembook.ir/edu

524 Subiect index

steam dietillation 55, 307 transfer unit 108, 132

steam production 310 t,ransition fraction 102
Stedrmtn packing 367 tubing 276, 342
still pot 398 tubular heater 251
etill pot contents 109, 138, 153
still pot heating 405, 465 units 37
substance properties 176, 187. 467, 471
Huspended vessel 309 vacuum assembly 457
systems of joints 35 vacuum controller 437, 463
vacuum distillation 198. 205
bake-off ratio 238 vacuum jacket 413
temperature control 445 vacuum line 277
temperature measurement 440 vacuum measurement 457
test distiuation apparatus 213, 338 valves 346
test mixture 137, 150, 162 vrtpour density 77
theoretical plate 102, 145, 199 v a p u r pressure 58
thermal rectification 377 Fapour pressure calculation 66
thermocouple 416 vapour pressure measurement 59
thermodynamic consistency 91, 101 vapour velocity 38, 140, 160
thermometer calibration eppnratus M 3 vapour volume 38
thermometer correction 189 velocity constant 177
thin-film distillation 283 vigreux column 206, 348, 389
thin-film still 211, 284 volatility 87
tightness 280, 489
tightness test 279 wall flow 48, 146, 357
time requirements 136 water still 225
timer 464 wetting column' 46, 365
total error 201 Wilson spiral 174, 422
t ) t d hold-up 15i, 210, 348, 367, 370. 378 wire-gauze packings 175, 423