Recent Advances in Gas Separation by Microporous Ceramic Membranes pdf

RECENT ADVANCES IN GAS SEPARATION BY
MICROPOROUS CERAMIC MEMBRANES
This Page Intentionally Left Blank
Membrane Science and Technology Series, 6
RECENT ADVANCES IN GAS
SEPARATION BY
MICROPOROUS CERAMIC
MEMBtlANES
Edited by
N.K. Kanellopoulos
NCSR "Demokritos", Membranes for Environmental Separations Laboratory,
15310 Aghia Paraskevi Attikis, Greece
2000
ELSEVIER
Amsterdam - Lausanne - New York - Oxford - Shannon - Singapore - Tokyo
ELSEVIER SCIENCE B.V.
Sara Burgerhartstraat 25
EO. Box 211, 1000 AE Amsterdam, The Netherlands
9 2000 Elsevier Science B.V. All rights reserved.
This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use:
Photocopying
Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher
and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or
promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to
make photocopies for non-profit educational classroom use.
Permissions may be sought directly from Elsevier Science Rights & Permissions Department, PO Box 800, Oxford OX5 1DX, UK;
phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: You may also contact Rights & Permissions
directly through Elsevier's home page (), selecting first 'Customer Support', then 'General Information', then
'Permissions Query Form'.
In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive,

Danvers, MA 01923, USA; phone: (978) 7508400, fax: (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid
Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 171 631 5555; fax: (+44) 171 631 5500.
Other countries may have a local reprographic rights agency for payments.
Derivative Works
Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for external resale or
distribution of such material.
Permission of the Publisher is required for all other derivative works, including compilations and translations.
Electronic Storage or Usage
Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part
of a chapter.
Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher.
Address permissions requests to: Elsevier Science Rights & Permissions Department, at the mail, fax and e-mail addresses noted above.
Notice
No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein.
Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.
First edition 2000
Library of Congress Cataloging-in-Publication Data
Recent advances in gas separation by microporous ceramic membranes / edited by N.K.
Kanellopoulos Ist ed.
p. em. (Membrane science and technology series ; 6)
Includes bibliographical references and index.
ISBN
0-444-50272-6 (alk. paper)
1. Gas separation membranes. 2. Ceramic materials. 3. Gases Separation. I.
KaneUopoulos, N. K. (Nick K.) II. Series.
TP159.M4 R43 2000
660'.2842 dc21
00-056192

ISBN: 0-444-50272-6
ISSN: 0927-5193
(~The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).
Printed in The Netherlands.
CONTENTS
List of Contributors
VII
Preface
Xl
1. Fundamentals and Sorption in Micropores
1.1
Membrane characterisation by combination of static and dynamic
techniques
Th. A. Steriotis, K. L. Stefanopoulos, A. Ch. Mitropoulos and
N. K. Kanellopoulos
1.2
In situ X-ray diffraction studies on micropore filling
T. liyama, T. Ohkubo and K. Kaneko
35
1.3 Neutron and ion beam scattering techniques
J. D. F. Ramsay
67
1.4 Application of pulsed field gradient NMR to characterize the transport
properties of microporous membranes
W. Heink, J. Karger and S. Vasenkov
97
1.5 Diffusion studies using quasi-elastic neutron scatttering
H. Jobic
109
1.6

Frequency Response methods for the characterisation of microporous
solids
L. V. C. Rees and L. Song
139
1.7 Measurement of diffusion in porous solids by Zero Length Column (ZLC)
methods
D. M. Ruth'ven and S. Brandani
187
1.8 Characterisation of microporous materials by adsorption microcalorimetry
P. Llewellyn
213
2. Modeling of Sorption and Diffusion in Microporous Membranes
2.1 Simulation of adsorption in micropores
D. Nicholson and T. Stubos
2.2
Molecular simulation of transport in a single micropore
D. Nicholson and K. Travis
2.3 Simulation of gas transport in a "network of micropores". The effect of
pore structure on transport properties
E. S. Kikkinides, M. E. Kainourgiakis and N. K. Kanellopoulos
231
257
297
3. Recent Advances in Microporous Membrane Preparation
3.1
Microporous carbon membranes
S. Morooka, K. Kusakabe, Y. Kusuki and N. Tanihara
3.2
Microporous silica membranes
N. Benes, A. Nijmeijer and H. Verweij

3.3
Zeolite membranes
J. D. F. Ramsay and S. Kallus
3.4 Chemical vapor deposition membranes
M. Tsapatsis, G. R. Gavalas and G. Xomeritakis
3.5
Composite ceramic membranes from Langmuir-Blodgett and
Self-Assembly precursors
K. Beltsios, E. Soterakou and N. K. Kanellopoulos
3.6
Nanophase ceramic ion transport membranes for oxygen separation
and gas stream enrichment
C. G. Guizard and A. C. Julbe
323
335
373
397
417
435
4. Gas Separation Applications
4.1
4.2
Nanoporous carbon membranes for gas separation
S. Sircar and M. B. Rao
Microporous inorganic and polymeric membranes as catalytic reactors
and membrane contactors
E. Driofi and A. Criscuofi
473
497
vii

List of Contributors
K. Beltsios
MESL, Institute of Physical Chemistry, NCSR "Demokritos", 15310 Aghia
Paraskevi Attikis, Greece
N. Benes
Laboratory of Inorganic Materials Science, Department of Chemical
Technology, University of Twente, P.O. Box 217, 7500 AE Enschede, The
Netherlands
S. Brandani
Department of Chemical Engineering, University College London, Torrington
Place, London WC1E 7JE, UK
A. Criscuoli
Department of Chemical Engineering and Materials, University of Calabria, Via
Pietro Bucci, Cubo 17/C, Arcavacata di Rende (CS), 87030 Italy
E, Drioli
Research Institute on Membranes and Modelling of Chemical Reactors, and
Department of Chemical Engineering and Materials, University of Calabria, Via
Pietro Bucci, Cubo 17/C, Arcavacata di Rende (CS), 87030 Italy
G. R, Gavalas
Division of Chemistry and Chemical Engineering, 210-41, California Institute of
Technology, Pasadena, CA 91125, USA
C, G, Guizard
Laboratoire des Mat6riaux et Proc6d6s Membranaires, UMR CNRS 5635,
Ecole Nationale Sup6rieure de Chimie, 8, rue de I'Ecole Normale, 34296
Montpellier Cedex 5, France
W. Heink
Fakult~t f0r Physik und Geowissenschaften, Universit~t Leipzig, Linn6stral3e
5, D-04103 Leipzig, Germany
T. liyama
Physical Chemistry, Material Science, Graduate School of Natural Science

and Technology, Chiba University, Yayoi, Inage, Chiba, 263-8522 Japan
H. Jobic
Institut de Recherches sur la Catalyse, CNRS, 2 Avenue Albert Einstein,
69626 Villeurbanne, France
A. C. Julbe
Laboratoire des Mat6riaux et Proc~d6s Membranaires, UMR CNRS 5635,
Ecole Nationale Sup6rieure de Chimie, 8, rue de I'Ecole Normale, 34296
Montpellier Cedex 5, France
M. Kainourgiakis
MESL, Institute of Physical Chemistry, NCSR "Demokritos", 15310 Aghia
Paraskevi Attikis, Greece
S. Kallus
Laboratoire des Mat6riaux et des Proc6d~s Membranaires, UMR CNRS 5635,
Universit6 Montpellier II, 2 pl Eugene Bataillon, 34095 Montpellier, France
viii
K.
Kaneko
Physical Chemistry, Material Science, Graduate School of Natural Science
and Technology, Chiba University, Yayoi, Inage, Chiba, 263-8522 Japan
N. K. Kanellopoulos
MESL, Institute of Physical Chemistry, NCSR "Demokritos", 15310 Aghia
Paraskevi Attikis, Greece
J. K~irger
Fakult~t for Physik und Geowissenschaften, Universit~t Leipzig, Linn6stral~e
5, D-04103 Leipzig, Germany
E, S. Kikkinides
Chemical Process Engineering Research Institute, P.O. Box 361, Thermi-
Thessaloniki 57001, Greece
K, Kusakabe
Department of Materials Physics and Chemistry, Graduate School of

Engineering, Kyushu University, Fukuoka 812-8581, Japan
Y. Kusuki
Polymer Laboratory, Corporate Research and Development, Ube Industries,
Ichihara 290-0045, Japan
P.
Llewellyn
Centre of Thermodynamics and Microcalorimetry - CNRS, 26 rue du 141 ~rne
RIA, 13331 Marseille cedex 3, France
A. Ch. Mitropoulos
Cavala's Institute of Technology, Department of Petroleum Technology, 65404
St. Lucas, Cavala, Greece
S, Morooka
Department of Materials Physics and Chemistry, Graduate School of
Engineering, Kyushu University, Fukuoka 812-8581, Japan
D. Nicholson
Department of Chemistry, Imperial College of Science, Technology and
Medicine, London SW7 2AY, UK
A, Nijmeijer
Laboratory of Inorganic Materials Science, Department of Chemical
Technology, University of Twente, P.O. Box 217, 7500 AE Enschede, The
Netherlands
T. Ohkubo
Physical Chemistry, Material Science, Graduate School of Natural Science
and Technology, Chiba University, Yayoi, Inage, Chiba, 263-8522 Japan
J, D. F. Ramsay
Laboratoire des Materiaux et des Proc~des Membranaires, UMR CNRS 5635,
Universite Montpellier II, 2 pl Eug6ne Bataillon, 34095 Montpellier, France
M. B. Rao
Air Products and Chemicals, Inc., 7201 Hamilton Boulevard, Allentown,
PA 18195-1501, USA

L. V. C. Rees
Department of Chemistry, The University of Edinburgh, West Mains Road,
Edinburgh EH9 3J J, UK
D. M. Ruthven
Department of Chemical Engineering, University of Maine, Jenness Hall,
Orono, ME 04469-5737, USA
S. Sircar
Air Products and Chemicals, Inc., 7201 Hamilton Boulevard, Allentown,
PA 18195-1501, USA
L. Song
Department of Chemistry, The University of Edinburgh, West Mains Road,
Edinburgh EH9 3J J, UK
E.
Soterakou
MESL, Institute of Physical Chemistry, NCSR "Demokritos", 15310 Aghia
Paraskevi Attikis, Greece
K, L, Stefanopoulos
MESL, Institute of Physical Chemistry, NCSR "Demokritos", 15310 Aghia
Paraskevi Attikis, Greece
Th. A. Steriotis
MESL, Institute of Physical Chemistry, NCSR "Demokritos", 15310 Aghia
Paraskevi Attikis, Greece
T. Stubos
MESL, Institute of Physical Chemistry, NCSR "Demokritos", 15310 Aghia
Paraskevi Attikis, Greece
N.
Tanihara
Polymer Laboratory, Corporate Research and Development, Ube Industries,
Ichihara 290-0045, Japan
K.

Travis
Department of Chemistry, Imperial College of Science, Technology and
Medicine, London SW7 2AY, UK
M. Tsapatsis
Department of Chemical Engineering, 159 Goessmann Laboratory, University
of Massachusetts, Amherst, MA 01003-3110, USA
S. Vasenkov
Fakult~t for Physik und Geowissenschaften, Universit~t Leipzig, Linn6stral~e
5, D-04103 Leipzig, Germany
H. Verweij
Laboratory of Inorganic Materials Science, Department of Chemical
Technology, University of Twente, P.O. Box 217, 7500 AE Enschede, The
Netherlands
G. Xomeritakis
Department of Chemical Engineering, 159 Goessmann Laboratory, University
of Massachusetts, Amherst, MA 01003-3110, USA
This Page Intentionally Left Blank
PREFACE
This book is dedicated to the rapidly grown field of microporous ceramic membranes
with separating layers of pore diameter less than 2 nm. In spite of the recent rapid growth
of the research effort directed towards the development of microporous ceramic
membranes the field is still considered to be at its infancy and exhibits a significant future
growth potential. The driving force for these efforts is of course the very promising
opportunities for cost-effective large scale gas separation applications, which can be
classified into two major categories: the "high temperature molecular sieving" and the
"low-temperature reverse molecular sieving" modes of separation techniques.
For high temperature applications, molecular sieving, based on the size exclusion of
large molecules, is very often identified as the major mode of separation by microporous
membranes. For low temperatures, the "reverse molecular sieving" is considered as a
very efficient mode of separation in which preferential sorption of heavy gas mixture

components results in the exclusion of the light components of the gas mixtures and the
permeation of the heavy components through the membrane. The idea of the reverse
sieving was initially introduced during the early sixties by professor R.M. Barrer and
coworkers and even though it has been demonstrated that the method can combine high
permeances with high selectivities, no major application has been developed as yet.
One of the reasons, of course, is that the production of large surface areas of
microporous separating layers with a minimum number of defects and a minimum
thickness is an extremely challenging task. Over the recent years, significant progress has
been made with respect to the development of novel microporous asymmetric
membranes, mainly involving modification by means of deposition of additional material
within the pores of the substrates. Most state-of-the-art technologies aiming in the
development of microporous ceramic membrane are presented in chapters 3.1, 3.2 and
3.3. These include several material deposition methods and techniques on macroporous
or mesoporous supports and substrates from the liquid or vapour phase, namely those
involving sol-gel, zeolite and chemical vapour deposition techniques. In addition to the
above-mentioned methods, the classical technique of carbonizing polymeric deposits
along with one of the novel techniques of plasma-treating, organically deposited
Langmuir-Blodgett films, are also presented. Finally, chapter 3.6 is dedicated to
nanophase mixed ionic-electron membranes for enhanced oxygen transport, which pose a
strong candidacy for a number of significant commercial applications.
Another significant factor that seriously hinders further development of the
microporous ceramic technology is the lack of comprehensive understanding of the
equilibrium and transport properties of molecules confined within nanopores. The
development of a satisfactory sorption and transport equation for the microporous
membrane performance requires the development of efficient characterization techniques
for the elucidation of the structural characteristics of the separating layer. Combining
xii
sorption techniques, scattering and differential permeability techniques, the characteri-
zation of the complex pore structure of the microporous layer, interpenetrated by a
network of larger pores can be obtained. These are analytically presented in chapter 1.1.

The study of the physical state of sorbed phase confined in micropores can be determined
by nondestructive scattering techniques. The recent advances of in situ X-Ray Diffraction
(XRD) are presented in chaptem 1.2, whereas the principles of Small Angle Scattering
techniques are outlined in chapter 1.3 along with recent developments employing the
contrast matching technique. Several chapters of the first section are dedicated to the
study of the diffusion processes in the micropores. "Microscopic" methods allow for the
direct determination of the self-diffusion coefficient under equilibrium conditions by
using two complementary methods, the pulse field gradient neutron magnetic resonance
(PFGNMR) and quasi-elastic neutron scattering (QENS) techniques. "Macroscopic" or
non-equilibrium methods, which allow the determination of transport diffusivities
measured under the application of a concentration gradient, are also explicitly presented.
The recently developed QENS technique allows the simultaneous determination of both
transport and self-diffusion coefficients. In combination with the methods of differential
permeability, frequency response and zero-length column chromatography, presented in
chapters 1.4, 1.5 and 1.6 respectively, understanding of the effect of micropore
confinement on the self- and transport diffusion coefficients may be obtained.
The second section is devoted to the modeling of the sorption and transport through
the complex porous structure of the microporous separating layer. Chapters 2.1 and 2.2
provide an overview of the recent advances in the simulation of sorption and transport
processes at the single pore level. In chapter 2.3 the theory of networks of single pores is
presented. Although the network theory is fully developed, insufficient description of the
transport process in a single micropore precludes explicit assessment of the effect of the
pore structm'al characteristics (pore size distribution, degree of connectivity etc.) to the
overall transport and selectivity performance of the membrane. Since a micropore
network model is currently under development, the analysis in chapter 2.3 is limited to
networks comprised of mesopores, which are necessary for the description of transport
through the larger pore network interpenetrating the network of micropores in the
separating layer.
In chapter 4.1 some of the most promising applications for the "low temperature
reverse molecular sieving" mode of separation are presented, namely the recovery of

paraffins and olefins from fluid catalytic cracking off gas along with the carbon dioxide
removal from natural gas. These are two major processes that merit further consideration
for full commercial exploitation. In addition to the above, some applications based on the
"high temperature molecular sieving" technique are presented in chapter 4.2.
It should be noted that the chapters of this book bring forward a wide range of issues,
namely fundamentals of complex sorption and transport processes in micropore
structures, highly innovative methods of preparation of microporous membranes and
examples of their possible commercial applications. It is hoped that the reader will find
useful and will take advantage of the insights presented by the distinguished
investigators, who have contributed significantly to the advance of research efforts in the
diverse topics presented in this book.
xiii
Acknowledgments
I would like to thank the staff of Elsevier Science and especially Drs. Huub Manten-
Werker and the members of our lab Drs. G. Papadopoulos, F. Katsaros, G. Romanos, V.
Kouvelos, G. Pilatos and N. Kakizis. The contribution of our secretary Ms. S. Botta is
gratefully acknowledged. Special thanks go to European Commission for the financial
assistance and to all our partners in several microporous membrane research projects
funded by the European Commission and especially to Steve Tennison from Mast
International Ltd.
Nick Kanellopoulos
This Page Intentionally Left Blank
Recent Advances in Gas Separation by Microporous Ceramic Membranes
N.K. Kanellopoulos (Editor)
2000 Elsevier Science B.V. All rights reserved.
MEMBRANE CHARACTERISATION BY COMBINATION OF STATIC AND
DYNAMIC TECHNIQUES
Th. A. Steriotis l, K. L. Stefanopoulos 1, A. Ch. Mitropoulos 2, and N. K. Kanellopoulos 1
1 Membranes for Environmental Separations Laboratory, Institute of Physical Chemistry,
NCSR "DEMOKRITOS", 15310 Aghia Paraskevi Attikis, Greece

2Cavala's Institute of Technology, Department of Petroleum Technology, 65404, St. Lucas,
Cavala, Greece *(current address)
The combination of static and dynamic techniques is a powerful tool which can provide
detailed information for the characterisation of membranes. We here present the type of
information obtainable by the adsorption, permeability, and small angle scattering (SAS)
techniques and important combinations of them. Additionally, examples of recent applications
of the methods and their selected combinations on a series of characteristic membrane
materials are presented. Important conclusions are drawn regarding the pore and internal
surface morphology of various types of membranes.
1. INTRODUCTION
Membrane technology is of great industrial interest, as in many cases it can replace
successfully traditional, pollution-prone and energy consuming separation processes. The
thorough assessment of structural features of the pore network and surface physicochemical
properties is probably the most important part of the characterisation of a membrane system
and a number of equilibrium and dynamic experimental techniques can be used for that
purpose. It is generally accepted that detailed description of a membrane system can only be
possible when different techniques are combined.
In the following, we briefly survey three independent methods of membrane
characterisation (adsorption, permeability, small-angle scattering (SAS) of x-rays and
neutrons) and two significant combinations of them, permeability in conjunction with
adsorption and adsorption in conjunction with SAS, with special emphasis on the type of
information obtainable in each case. Additionally, we demonstrate examples of the application
of the methods to the following materials:
i. y-Al203 mesoporous pellets, prepared by symmetrical compaction of Degussa aluminium
oxide of type-C (particle radius ~ 100 A) in eleven sections. During compaction, special care
was taken to avoid non-homogeneity effects by applying an appropriate level of compression
to each section.
ii. A Vycor 7930 type porous silica. According to the manufacturer (Coming) the sample has
a porosity of ~ 28% and pores of ~ 40 A diameter.
iii. Asymmetric gas-separating carbon membranes obtained through carbonisation of a

polymer resin precursor and subsequent activation (1). These membranes have a microporous
carbon skin on top of a macroporous carbon substrate. The pore diameter of the microporous
material appears to be in the 12-15 A range, while microporosity is about 38%.
iv. Silicalite-1 membranes prepared by in situ hydrothermal synthesis and crystallisation of
the zeolite inside the pores of ct-A1203 macroporous disk-shaped supports. Syntheses were
performed by dissolving pyrogenic silica (Aerosil 380, Degussa) in aqueous solutions of
tetrapropylammonium hydroxide (templating agent). After ageing the mixtures were heated
with the support in Teflon lined stainless steel autoclaves. Finally, the organic template was
removed by calcination (2).
2. DYNAMIC AND EQUILIBRIUM METHODS
2.1.
Adsorption
The adsorption isotherm, i.e. the quantity of gas (vapour) adsorbed on a solid at
different pressures, at constant temperature, is a function of the surface area and the pore
structure of the solid and thus can provide useful information about these two factors. To this
end adsorption isotherms (especially N 2 at 77 K) is a widely used technique for the
characterisation of porous materials.
The presence of certain types of pores (micropores, i.e. pore width, w, less than 20 A,
mesopores with 20 A < w < 500 i~ and macropores with w > 500 A) produce different shapes
of isotherms. The majority of these isotherms can be grouped into in five classes (six with the
stepped isotherm) after Brunauer, Deming, Deming and Teller (3) according to the pore size
of the solid and, the adsorbent-adsorbate interaction.
The analysis of adsorption data can in principle produce values for the surface area and
total (or micro) pore volume of the solid under investigation, by means of well-established
methods such as BET, Langrnuir or DR (4). The pore size distribution (psd) of mesoporous
solids can be derived through isotherm analysis methods, based on the Kelvin equation (5).
Clearly the above analysis can be only applied to pores accessible to the penetrating gas (open
pores), while inaccessible (closed) pores can only be detected with the aid of other methods
such as SAS. Furthermore, Kelvin equation is not applicable to micropores, due to their small
size (few molecular diameters) and the overlapping potential fields of neighbouring walls. In

this case the processing of adsorption data using the potential theory (6) is possible and can
lead to psd estimates of semi-quantitative validity.
Typical examples of N2 adsorption isotherms (77 K) for mesoporous alumina and
Vycor and microporous carbon membranes are shown in Figure 1, while the corresponding
pore size distributions are presented in Figure 2.
While a number of more or less established characterization methods exist for
mesopores and macropores, the assessment of microporosity is much less advanced, due to
experimental difficulties and the lack of an appropriate model for the interpretation of
adsorption data. N2 adsorption at 77 K is probably the most studied technique, however
obtaining accurate experimental isotherms is hampered by the long equilibration times
required at the low liquid nitrogen temperature. In order to overcome this limitation the
micropore structure evaluation can be based on isotherms of carbon dioxide or other vapours
obtained at higher temperatures, provided that suitable equilibrium models for the sorption of
non-spherical molecules are available.
200
180
160
140
n 120
I-
or)
100
E
0
v 80
60
40
20
J
._._ Ca r~n _,,iA~

_._v, cor
-o- Alumina :" / j/
i r i ,
0 0.2 0.4 0.6 0.8 1
piP0
Figure 1. N2 adsorption isotherms of Vycor, AI203 and microporous carbon membranes, at
77K.
70
60
"~ 50
or)
%
~ 30
>*
-~ 20
10
4- Carbon
Vycor
II
-o- Alumin
10 100
r (A)
Figure 2. Pore size distributions based on the Kelvin equation (Vycor, alumina) and Dubinin-
Astakov method (carbon).
The Grand Canonical Monte Carlo (GCMC) method is ideally suited to adsorption
problems because the chemical potential of each adsorbed species is specified in advance
(7,8). At equilibrium, this chemical potential can be related to the external pressure by making
use of an equation of state. Consequently, the independent variables in the GCMC simulations
are the temperature, the pressure and the micropore volume, i.e. a convenient set, since
temperature and pressure are the adsorption isotherm independent variables. Therefore, the

adsorption isotherm for a given pore can be obtained directly from the simulation by
evaluating the ensemble average of the number of adsorbate molecules whose chemical
potential equals that of the bulk gas at a given temperature and pressure. To this end, a method
for the determination of the micropore size distribution based on Monte Carlo simulation has
been developed (9). In this work the mean CO2 density inside a single slit shaped graphitic
pore of given width, is found on the basis of G.C.M.C. simulations for a pre-defined
temperature and different relative pressures. Starting from an initial PSD guess, it is then
possible to produce a computed CO 2 sorption isotherm and compare it to the measured one.
After a few iterations, the procedure results in a PSD which, if desired, can be further refined
at the cost of additional computational effort.
2.2 Permeability
The fluid flow properties of porous media, are extremely sensitive functions of the
pore size distribution (psd) and additional pore structural characteristics (shape, connectivity).
To this end permeability is a dynamic technique, which can provide useful data concerning
the structure of membranes and evaluate their overall quality simultaneously. We may note
that open pores can be either conductive or blind (dead-end). Both open pore types contribute
to adsorption, while permeation occurs through conducting pores only.
The measurement of the permeability, P, of a weakly adsorbed gas (e.g. helium)
through a membrane can be used for the calculation of gas diffusion coefficient, Dg (in this
case P=Dg.e where e is the porosity). As the pressure gradient across the membrane increases,
the flow regime changes from Knudsen to viscous (Poisseuille). Additional surface or
condensate flow occurs when the gas (vapour) is adsorbed or condensed in the pores of the
membrane. On Knudsen flow, which occurs when the pore radius, r, is sufficiently smaller
than the mean free path, ~., of the flowing molecules (r<0.05)~), Dg (and P) is independent of
the pressure gradient. Thus, the existence of membrane defects can be detected by examining
whether
Dg
is constant over a range of pressure gradients or not. A typical example is
presented by Romanos et al. (2) for two zeolite membranes on supports of different porous
size (0.08 ~m and 0.15ktm) synthesised at the same hydrothemal conditions. The membranes

were examined by helium differential permeability experiments at 308 K and exhibited a
typical crack (macropore) free behaviour. From Figure 3a it is clear that the change from
molecular to viscous flow, for both membranes, occurs at approximatelly 12 bar, where the
mean free path of helium for the aforementioned temperature is 168 ,~. By making the usual
assumption that Knudsen regime changes to viscous when 0.01<r/)~<0.05 (10), one can
calculate that the pore diameters range between 3.4 and 17/~, excluding thus the existence of
macro and mesopore intercrystalline voids. Furthermore, when macropores (cracks) are
present, the flow is expected to be Poisseuille even at very low pressures, as presented in
Figure 3b, which contains the differential permeance results for a membrane developed on a
support of 9~tm porous size, and under different hydrothermal conditions.
7
oO
~
Wb
i ~ . ~ O ~
o
o
~" 6.~:IDO- -O O O ~-O ~" " " "
o support 0.15~
x
m support 0.08+
I
3 ~ ~ , ,
0 5 10 15 20 25 30
Pressure (bar)
7 7
I
o
o
'~

5
x
8
support 9p
az
0 0
support O.081a
35 0 0.5 1 1.5
Pressure (bar)
(a) (b)
Figure 3. Helium differential permeances at 308K of the zeolite membranes on ct-A1203
substrates with mean pore diameters (a) 0.08 ~tm and 0.15 ~tm, (b) 0.08 ~tm and 9 ~tm.
Additionally, the structure factor kg (the ratio of the experimentally measured
Dg over
the theoretical one for cylindrical pores), introduced by Barrer and co-workers (11,12), can
easily be calculated from experimental helium permeability values. The magnitude of this
factor, in combination with other available structural information (porosity, inhomogeneity)
leads to some general conclusions about the pore network geometry (12,13), as in the case of
mesoporous alumina and microporous carbon membranes. The measured helium permeability
values were constant for both membranes over a wide range of pressure heads (10-900 mbar)
ensuring that the membranes were crack-flee and that mass transport occurred in the Knudsen
regime (14). The value of kg (0.51), calculated for the alumina membrane, was higher than the
expected ideal value of 1/3 for an isotropic, homoporous medium. This can be attributed to
deviations from the cylindrical geometry (flattened pores) and/or the assumption of random
reflection of molecules (15). On the other hand the structural factor kg for the asymmetric
carbon membrane was estimated (1) to be in the range of 8xl 0 5. Such an extremely low value
for a carbon membrane having a skin with a microporosity of approximately 0.38, may signify
an unusually long (tortuous) path necessary for permeability and therefore, the porous
structure should be in the vicinity of its percolation threshold. As the percolation threshold
should be near 0.15 for a three dimensional pore arrangement and near 0.50 for a two

dimensional arrangement, a percolation threshold in the vicinity of 0.38 indicates that the pore
arrangement has a partial two dimensional character. Nevertheless, an alternative explanation
for the observed very low structure factor is the presence of constrictions. To this end, with
the aid of information obtained from SAS and N2 adsorption, structural models for the carbon
membrane have been developed (16).
Furthermore, the dependence of Dg on temperature is of major importance mainly in
microporous systems, revealing whether diffusion is activated or not. Activated diffusion
takes place when the molecules must surmount an energy barrier, usually produced by
constrictions at the pore mouths. For alumina and Vycor the Knudsen mechanism was
supported by measurements of helium permeability at different temperatures (Dg oc T~), while
for all the microporous membranes the temperature dependence of the integral permeance of
helium was not Knudsen. The process was found to be activated and activation energies were
derived assuming an Arrhenius type (Poce Ea/Rx) of behaviour. The activation energies, E~, can
-16.5
-16.7
[ -16.9
r~
"~ -17.1
o
~ -17.3
x
~" -17.5
-177
2.4
o~
support 0.15Ix
support
0.08Ix
2.6 2.8 3 3.2 3.4
1000/T

(K l)
Figure 4. Calculation of Ea for helium on zeolite composite membranes.
3.6
be easily calculated from the slope of the In Pe vs. 1/T curves (Figure 4). The activation
energy for the carbon membrane was 4 kJ/mol, while in the case of zeolite membranes,
although the permeability values were different, Ea was 6 kJ/mol for both of them (Figure 4).
These values imply that the pore network of the carbon membrane is constricted while the
zeolite porous system, developed during synthesis, is the same in the two substrates (in terms
of pore size, connectivity, constrictions etc.). The latter can only happen if the membranes are
defect free and transport occurs only through silicalite-1 channels (2).
On the other hand, differential permeability measurements of adsorbable gases or
vapours through mesoporous media, performed at relative pressures (P/P0) ranging from 0 to
1, exhibit a maximum at a certain P/Po. This fact is attributed to the occurrence of capillary
condensation in the mesopores and by using appropriate models, structural characteristics of
the porous media can be deduced (17). Differential permeability experiments of CO2 on
microporous carbon membranes (1) at slightly supercritical conditions (T=35 ~ have shown
that at a pressure of about 35 bar a peak analogous to the mesopore case appears (Figure 5).
Further experimentation with nanoporous Vycor membranes having their surface modified by
various types of continuous or near continuous microporous layers, show that for materials
with similar physical chemical characteristics the maximum appears to shift to smaller values
with smaller pore sizes (18). Thus, the technique may eventually lead to a method appropriate
for the assessment of the pore structure of microporous membranes. However, the correlation
of such peaks with the pore structure is, at present, rather qualitative, due to the limited
knowledge regarding the behaviour of the fluid in micropores. The sharp maxima which are
observed can be attributed to the special arrangement of the CO2 molecules inside the
micropores. Simulation studies have predicted an orientational ordering transition of the
molecular axis of supercritical CO2, inside individual pore models, with size of the order of
few molecular diameters (19). To this end recently (20), neutron diffraction experiments in
conjunction with in situ adsorption are used to monitor changes in the nature of the adsorbed
gas.

350
300
25O
200
~g
2:~ 150
100
0 10 20 30 40 50 60
Pm (bar)
Figure 5. Differential CO2 permeability curve at 308 K for the composite carbon membrane.
A qualitative explanation for the observed maxima can also be given following
Nicholson et al. (21). The flux in the x-direction is given by the expression:
J=Jo +Jv = ~ 1] ~,Ox)
(1)
10
with Jo : diffusive flux, driven by the chemical potential, tg~t/c~, Jv : viscous flux, driven by
the pressure gradient, c~/o~x, c : molecular concentration inside the pore space, Do : diffusion
coefficient, k : the Boltzmann constant, T : absolute temperature, B0 : geometrical term and 1"1
:the viscosity coefficient. Non Equilibrium Molecular Dynamics (NEMD) studies revealed
that for certain micropores viscous contribution is activated above a critical density (and
therefore pressure) of the fluid inside the pores. At lower pressures the diffusive flux is
predominant, while in higher pressures the viscosity coefficient increases rapidly. This
situation will eventually lead to a maximum in the permeability, as predicted for the
permeability of supercritical CH4 through a 0.95 nm width slit-shaped carbon micropore (21).
On the other hand a Monte Carlo (MC) simulation (9) for carbon dioxide in supercritical
conditions is currently used to estimate the density profile (and, as a result, the mean density)
of the fluid inside carbon slits at various pressures (Figure 6). We expect that ultimately a
combination of such NEMD and MC simulations will permit the determination of pore width
from the high-pressure differential permeability data.
c i

A
6
o
O ; i
E
v
2
0 t ~ ,.+ t t t I
0 10 20 30 40 50 60 70
p (bar)
Figure 6. Mean density of CO2 at 308 K inside micropores vs. pressure (Monte Carlo
simulation) A. pore width = 7.5 A, B. pore width = 10.5 A, C. pore width = 15.5 A