Membrane Materials for Gas and Separation E-Book

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Sprache: Englisch

Seitenzahl: 782

Veröffentlichungsjahr: 2017

Beschreibung

Si containing polymers have been instrumental in the development of membrane gas separation practices since the early 1970s. Their function is to provide a selective barrier for different molecular species, where selection takes place either on the basis of size or on the basis of physical interactions or both.

• Combines membrane science, organosilicon chemistry, polymer science, materials science, and physical chemistry

• Only book to consider polymerization chemistry and synthesis of Si-containing polymers (both glassy and rubbery), and their role as membrane materials

• Membrane operations present environmental benefits such as reduced waste, and recovered/recycled valuable raw materials that are currently lost to fuel or to flares

Leseprobe

Membrane Materials for Gas and Vapor Separation

Synthesis and Application of Silicon‐Containing Polymers

Edited by

Yuri Yampolskii and Eugene Finkelshtein

A.V. Topchiev Institute of Petrochemical Synthesis, RAS, Moscow, Russia

This edition first published 2017© 2017 John Wiley & Sons Ltd

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Yuri Yampolskii and Eugene Finkelshtein to be identified as the author(s) of the editorial material in this work has been asserted in accordance with law.

Registered OfficesJohn Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USAJohn Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial OfficeThe Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty

In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication Data

Names: Yampolskii, Yuri, editor. | Finkelshtein, Eugene, editor.Title: Membrane materials for gas and vapor separation : synthesis and application of silicon-containing polymers / Yuri Yampolskii, Eugene Finkelshtein.Description: Chichester, West Sussex, United Kingdom : John Wiley & Sons, Inc., [2017] | Includes bibliographical references and index.Identifiers: LCCN 2016036752| ISBN 9781119112716 (cloth) | ISBN 9781119112730 (epub) | ISBN 9781119112723 (Adobe PDF)Subjects: LCSH: Gas separation membranes–Materials. | Silicon polymers.Classification: LCC TP248.25.M46 I35 2017 | DDC 660/.28424–dc23LC record available at https://lccn.loc.gov/2016036752

Cover image: ALFRED PASIEKA/SCIENCE PHOTO LIBRARY/Gettyimages

Contributors

Morteza AziziDepartment of Chemical EngineeringUniversity of LouisvilleKYUSA

Susanta BanerjeeMaterials Science CentreIndian Institute of TechnologyKharagpurIndia

Giuseppe BarbieriInstitute on Membrane Technology (ITM‐CNR)National Research Councilc/o The University of CalabriaCubo 17CVia Pietro BucciRende CSItaly

Debaditya BeraMaterials Science CentreIndian Institute of TechnologyKharagpurIndia

Maksim BermeshevA.V. Topchiev Institute of Petrochemical SynthesisRASMoscowRussia

Adele BrunettiInstitute on Membrane Technology (ITM‐CNR)National Research Councilc/o The University of CalabriaCubo 17CVia Pietro BucciRende CSItaly

Pavel ChapalaA.V. Topchiev Institute of Petrochemical SynthesisRASMoscowRussia

Enrico DrioliInstitute on Membrane Technology (ITM‐CNR)National Research Councilc/o The University of CalabriaCubo 17CVia Pietro BucciRende CS, Italy;Dipartimento di Ingegneria per l’Ambiente e il Territorio e Ingegneria ChimicaThe University of CalabriaCubo 44AVia Pietro BucciRende CS, Italy;Hanyang UniversityWCU Energy Engineering DepartmentSeongdong‐guSeoulSouth Korea

Timothy DubbsDepartment of Chemical EngineeringUniversity of LouisvilleKYUSA

Eugene FinkelshteinA.V. Topchiev Institute of Petrochemical SynthesisRASMoscowRussia

Joel R. FriedProfessor and Chair of Chemical EngineeringUniversity of LouisvilleKYUSA

Maria GringoltsA.V. Topchiev Institute of Petrochemical SynthesisRASMoscowRussia

Stepan GuselnikovA.V. Topchiev Institute of Petrochemical SynthesisRASMoscowRussia

Yanming HuKey Laboratory of Synthetic RubberChangchun Institute of Applied ChemistryChinese Academy of SciencesChangchunChina

Victor KopylovMoscow Technological InstituteRussia

Alexander KovyazinPENTA‐91 LLC CoMoscowRussia

Toshio MasudaDepartment of Polymer MaterialsSchool of Materials Science and EngineeringShanghai UniversityNanchenShanghaiChina

Leonardo MeloneInstitute on Membrane Technology (ITM‐CNR)National Research Councilc/o The University of CalabriaCubo 17CVia Pietro BucciRende CSItaly

Igor RaygorodskyMoscow Technological InstituteRussia

Yulia RoganThomas Swan & Co. LtdRotary WayConsettUK

Toshikazu SakaguchiDepartment of Materials Science and EngineeringGraduate School of EngineeringUniversity of FukuiFukuiJapan

Tadashi UragamiFunctional Separation Membrane Research CenterOsakaJapan

Nikolay V. UshakovA.V. Topchiev Institute of Petrochemical Synthesis, RASMoscowRussia

Yuri YampolskiiA.V. Topchiev Institute of Petrochemical Synthesis, RASMoscowRussia

Preface

Organosilicon compounds possess a number of specific properties due to the presence of Si‐containing chemical bonds. In general, this makes organosilicon chemistry an effective tool for a planned macromolecular design. Thus, Si–Cl bonds are substantially more active in hydrolysis reactions and in interaction with Grignar reagents than their carbon analogues. Si–H bonds smoothly react with olefins, in contrast to rather chemically passive C–H bonds. The silicon atom has a very weak tendency to formation of multiple bonds under normal conditions. This prevents the possibility of numerous undesirable side chemical processes, such as dehydrochlorination of chlorosilanes, dehydrogenation of hydrosilanes, and some others. At the same time, Si–C and Si–O bonds are quite stable, chemically as well as thermally. These bonds are the main “building blocks” of polycarbosilanes and polysiloxanes. Therefore, carbosilanes and siloxanes form an attractive basis for development of various polymer materials.

Simplicity of incorporation of different organic substituents on the silicon atom, including polar and sterically hindered groups, allows fabrication of a series of desired structures unattainable for purely organic compounds. This is the case for low molecular weight compounds (monomers), as well as for high molecular weight polymers.

Organosilicon monomers allow carbochain glassy polymers possessing high glass transition temperature (Tg) to be obtained by means of polymerization on multiple bonds, according to addition and metathesis schemes. Some monomers can also be used for synthesis of elastomeric polymers with very low Tg by ring opening polymerization via breaking endocyclic Si–C or Si–O bonds. Numerous examples of organosilicon polymers are shown below.

Homochain polymers

Heterochain polymers

Therefore, special peculiarities of organosilicon chemistry, as noted above, allow incorporation of a great variety of substituents on the silicon atom. This makes molecular design of desired polymer materials as well as conscious adjustment of their physicochemical properties realistically feasible.

Among various actual directions of the use of Si‐containing polymer materials, materials for gas and vapor separation membranes form an important and prospective field. Thus, the key to successful development of separation membrane materials is in finding and elaborating convenient methods for synthesis of appropriate monomers and determination of their optimal polymerization conditions, resulting in polymers with good gas transport and film‐forming properties.

Study of gas permeation parameters (permeability, diffusivity, thermodynamic sorption parameters) and important related properties such as free volume is an independent and a wide field of research. Among other tasks, one is to make an appropriate selection of gas mixtures that can be separated by certain membranes. Membrane science and technology related to the problems of gas and vapor separation are in permanent evolution. In this regard, modification of existing polymer membrane materials, searching for optimal conditions of separation and development of original syntheses of novel polymers provide permanent challenges for researchers. Methodologies based on organosilicon chemistry may be quite useful for the modern membrane industry.

All these issues form the subject of this monograph. In it, for the first time, we tried to consider jointly the questions of organosilicon chemistry and membrane science, giving historical backgrounds, outlining the trends of development and providing the contemporary state of the art of both fields.

In Chapter 1 the main parameters of membrane gas separation are defined and explained. Since gas permeation in non‐porous polymer membranes proceeds according to the solution–diffusion mechanism, the role of kinetic and thermodynamic factors in mass transfer through membranes is outlined. The role of the combination of high permeability and selectivity is stressed as a prerequisite of highly efficient membrane materials. Special attention is devoted to the effects of the nature and properties of gas and polymers on the observed gas permeation parameters.

From Chapter 2, consideration of the synthesis and properties of Si‐containing polymers is started. The subject of this chapter is rubbery polymers with flexible Si–O–Si bonds: organosiloxanes and block copolymers containing flexible siloxane blocks. The main feature of siloxanes is their extremely low Tg and, consequently, the very high mobility of their main chains. The chemistry and applications of polyorganosiloxanes and their copolymers have been intensively studied since the 1940s. They have found numerous applications, and one of them is their use as membrane materials. For a long period polydimethylsiloxane (PDMS) was considered as the most permeable polymer among all those known. A great impact on applications of siloxane‐containing polymers started 20–30 years later when block copolymers with rigid and flexible blocks were created and studied. The chapter gives a detailed description of the developed methods of synthesis of the polymers of this class, and numerous results of the studies of their membrane properties.

Interesting analogs of polyorganosiloxane are known; these are polymers where the flexible Si–O bond is replaced by the structurally similar Si–C bond: polysilmethylenes, which are the subject of Chapter 3. A comparison of these two types of polymer permits further elucidation of the role of the flexibility of the main chains of Si‐containing polymers and its effects on permeability and diffusivity. Approaches to the synthesis of polysiloxanes and polysilmethylenes have common features: in both cases it is a scission of strained cycles. However, there are differences between the polymers obtained: the latter have less flexible chains and, hence, their permeability is not that high. The polymers of both classes are rubbers, so the problems that can be solved using the membranes based on them are similar. This is mainly the separation of gaseous hydrocarbons; however, in many cases their relatively high gas permeability justifies consideration of the separation of light gases such as O2/N2 or CO2/CH4.

Since the 1960s a new era has started in the chemistry and physical chemistry of Si‐containing polymers as membrane materials. A big stride was made by creating poly(vinyltrimethyl silane) (PVTMS) and its structural analogs. A general feature of these vinylic polymers, described in Chapter 4, is that they contain Si in side groups and are glassy materials. On the basis of PVTMS the first industrially produced gas separation membrane was fabricated and produced from the end of the 1970s in the Soviet Union. The properties of this polymer, which seemed rather unusual when it was prepared and studied, undoubtedly influenced further activity in the field of Si‐containing membrane materials. The chapter gives a brief review of polymerization chemistry of vinylorganosilanes and emphasizes the role of anionic polymerization. Other vinylic polymers, e.g. Si‐substituted polystyrenes, are also briefly considered.

The theme of glassy Si‐containing polymers obtained an exceelent development in studies of disubstituted Si‐containing polyacetylenes, the subject of Chapter 5. These materials show a wide range of permeability and have demonstrated diverse manifestations of structure–permeability effects. As often occurs, even the first prepared polymer of this class, poly(trimethylsilyl propyne) (PTMSP), revealed record‐breaking permeability. It was with PTMSP that the phenomena of solubility controlled permeation were observed for the first time using glassy membranes. Another interesting reaction was discovered with polyacetylenes – desilylation. It resulted in formation of highly permeable materials that do not contain silicon (solid state elimination of Si‐containing groups with formation of additional free volume elements within the membrane). It is likely that the same concept can be applicable to other classes of glassy polymers that contain C(arom)–Si bonds in side groups.

A wealth of information is reported in Chapter 6. There, the authors deal with numerous Si‐containing glassy polymers of norbornene and polytricyclononenes. An unusual peculiarity of these polymers is that the same monomers can produce materials with entirely different structures, chain rigidities and other properties depending on the selection of the polymerization catalyst. Metathesis polynorbornenes have relatively flexible chains and rather modest gas permeability. Nonetheless, after preparation and investigation of a large number of polynorbornenes with different structures many important observations were made regarding the structure–permeability relationship. Addition Si‐containing polynorbornenes have very rigid main chains (high Tg) and demonstrated high gas permeability, similar to that of polyacetylenes. For this class of polymers solubility‐controlled permeation was also observed.

The subject of Chapter 7 is the description of synthesis and investigation of polyimides and polyamides with bulky side groups (e.g. tert‐butyl or adamantyl). The idea of this chapter is a demonstration that not only Si‐containing side groups but also other bulky substituents can result in significant increases in permeability, often not at the expense of permselectivity. The chapter contains much information on the details of synthetic chemistry of these polymers and the data on their gas permeation properties, using the O2/N2 pair as an example.

General questions of membrane science are considered in Chapter 8. In it gas permeability and diffusivity of diverse Si‐containing classes of membrane materials are discussed. The chapter starts with consideration of rubbery polymers (polysiloxanes and polysilmethylenes) and then proceeds to discussion of properties of glassy Si‐containing polymers that have played such an important role in the development of membrane gas separation. Structure–property relations are again at the focus of deliberation. The role of free volume in membrane materials is also outlined.

There are many examples where the structure and properties of Si‐containing polymers have been the subject of theoretical works and computer simulations. These questions are considered in Chapter 9. The most extensive work has been performed for PDMS among rubbery polymers and for PTMSP among polymer glasses. The authors of this chapter focus on the role of main chain stiffness, mobility of side groups and the effects of these properties on the diffusivity, solubility and permeability coefficients of various gases. A large appendix is included in this chapter; it contains numerous technical details used in these simulations and hence may be useful for future researchers.

It is known that Si‐containing polymers have proved their efficiency not only in gas separation but also in separation of liquids – pervaporation (PV). This is the subject of Chapter 10. It demonstrates the usefulness of siloxane polymers and PTMSP in various PV processes. Another method has also been developed – evapomeation, where the liquid mixture to be separated does not contact the membranes directly. Instead, vapor phases formed by evaporated components of the liquids are separated. Numerous examples of different separation processes are given.

This book would not be complete had it not included a chapter on practical implementation of membranes based on Si‐containing polymers. This task is accomplished in Chapter 11. It can be considered as a brief introduction to membrane technology. Different types of membrane (flat sheet and hollow fiber) are described, as well as different designs of membrane modules. Special attention is devoted to general advantages of membrane technology in comparison with other, more traditional methods of separation. Actual examples are given on separation of particular gas mixtures.

The editors would like to extend their sincere gratitude to all contributors and the reviewers of this book. We were sure, and the work on this monograph confirmed, that the contributors of this book are world‐class experts in their specific fields. Furthermore, we would like to express our thanks to the publishers of this book, John Wiley & Sons, Ltd, Chichester, UK, for their support and guidance. The editors of this book want to express their gratitude to Russian Science Foundation for support publishing of this volume and in particular Chapters 1, 3, 4, 6, 8.

Eugene FinkelshteinYuri Yampolskii

1Permeability of Polymers

Yuri Yampolskii

A.V. Topchiev Institute of Petrochemical Synthesis, RAS, Moscow, Russia

1.1 Introduction

Virtually all the membrane processes realized for separation of gases and vapors employ non‐porous polymeric membranes. The phenomena of permeation of gases and vapors through plastic films were known even in the eighteenth century [1]. However, the mechanism of these phenomena became clear only in the middle of the nineteenth century due to the works by K. Mitchell and T. Graham [1, 2]1, who advanced the so‐called solution–diffusion model. According to them the presence of microscopic pores within the films is not a prerequisite for realization of mass transfer. Instead, the dissolution of gaseous molecules in the film and their diffusion through it can be a basis for gas transport in membranes. The sorbed gas, as T. Graham wrote, ‘comes to evaporate… and reappears as gas on the other side of the membrane. Such evaporation is the same into vacuum and into another gas, being equally gas‐diffusion in both circumstances’ [2].

An empirical observation made approximately during the same period was that the flux of gas J through a film (mol/m2 s) is directly proportional to the gas pressure drop across this film Δp and inversely proportional to the thickness of the film l, i.e.

(1.1)

The proportionality coefficient in this equation, P, was defined as the permeability or permeability coefficient. However, this empirical equation does not reveal the molecular basis of permeation and the complicated nature of this quantity.

In order to understand the ‘solution–diffusion model’ let us consider a steady‐state isothermal flux through a homogeneous polymer film with thickness l that separates two gas phases containing a single gas with pressure .2 The transport within the film can be described by Fick’s first law:

(1.2)

where C is concentration, x is the coordinate across the film, and the diffusion coefficient D in the first approximation does not depend on C or x. It can easily be integrated, but the boundary conditions are usually unknown. On the other hand, pressure p1 and p2 can easily be measured and established in the experiment. Therefore, one must consider the relationship between C and the pressure or sorption isotherm. As will be discussed later on in this chapter, the form of sorption isotherms in polymers can be complicated, but now it is sufficient to consider the simplest case, that is, the Henry’s law isotherm:

(1.3)

where S is the solubility coefficient. It is commonly expressed as cm3(STP) cm−3 atm−1. Here the term cm3(STP) characterizes the volume of gas in the standard conditions (273.15 K and 1 atm or 101.3 kPa) and the term cm−3 characterizes the unit volume within the film. By replacing C by p in Eq. (1.2) one obtains

(1.4)

so it is obvious that

(1.5)

This is a key equation in membrane science. It indicates that the empirical parameter P includes two components: the diffusion coefficient D, which characterizes the mobility of dissolved gas molecules, and the solubility coefficient S, which characterizes the affinity between the polymer material and the diffusing gas. It is evident that S is a thermodynamic property of the gas–polymer system.

In the SI system, permeability coefficients are expressed in the following units:

However, a more widely used and accepted unit for P is the Barrer:

All the gas–polymer systems are characterized by permeability or permeability coefficients in the range 10−4–104 Barrer.

Equations (1.1) and (1.4) include the thickness of a polymer film l. In membranes this parameter is unknown, or different parts of the membrane have different thicknesses. Therefore, in the important case of membranes, pressure normalized steady‐state flux or permeance (Q or P/l) is used to characterize the gas transport rate. The accepted units for P/l are mol m−2 s−1 Pa−1 or m3(STP) m−2 h−1 atm−1. Permeance is often expressed using the gas permeation unit (GPU), where .

Since membranes are used for separation, another key characteristic of gas separation membranes is their selectivity. The ideal selectivity is defined as follows:

(1.6)

where PA and PB are the permeability coefficients of gases A and B, respectively, measured in runs with permeation of individual gases. Commonly, the more permeable gas is taken as A, so that . In the literature different terms (synonyms) are used for αAB: selectivity, permselectivity, ideal separation factor. They will be used throughout the chapters of this book. Ideal separation factors vary in much narrower ranges than permeability coefficients and in strong dependence on the gas pair: thus α(O2/N2) changes in the range from 2 to very seldom 20, while α(He/N2) in the much wider range of 2–104 [5]. Bearing in mind Eq. (1.5), the ideal selectivity can be partitioned into diffusivity and solubility selectivities as follows:

(1.7)

so it is possible to speak of the selectivity of diffusion and sorption. The analysis of and is very helpful in understanding the mechanism of gas permeation in polymers. Ideal separation factors can also be considered for membranes as the ratios .

Equations (1.6) and (1.7) hold when interactions between diffusing molecules in mixed gas permeation can be neglected, and also when they do not noticeably affect the properties of the polymeric matrix. In such situations the ideal selectivity measured in experiments with pure gases only approximately characterizes the actual selectivity of a membrane. The separation factor determined from the ability of a membrane to separate a binary feed gas mixture is defined as follows [6]:

(1.8)

where yA and yB are the mole fractions of the components produced in the permeate, and xA and xB are their corresponding mole fractions in the feed. In mixture separations with large stage‐cuts (or when the fraction of the permeate stream is comparable with that of the feed stream) sometimes xA and xB are taken as the mole fractions of these components in the retentate.

1.2 Detailed mechanism of sorption and transport

The solution–diffusion mechanism provides an overall principle of the mass transfer through non‐porous polymer membranes. However, in depth understanding of the mechanism of gas transport is impossible without more detailed, desirably atomistic models of what occurs when gas molecules are dissolved in polymers and diffuse through polymer films or membranes. Two approaches can serve for this aim.

1.2.1 Transition‐state model

It is well known that diffusion in condensed media is an activated process. When a molecule of dissolved gas (penetrant) permeates through the membrane it performs numerous elementary acts: ‘jumps’ from one equilibrium position in the polymer matrix into another (neighboring) one. The passage between these ‘microcavities’ or ‘cells’ implies overcoming forces of attraction between more or less aligned chains of macromolecules. This means that the diffusing dissolved gas molecule must overcome an energy barrier [7]. This barrier can be considered as an activation energy of diffusion ED, so for the diffusion coefficient the following Arrhenius equation holds:

(1.9)

By combining this equation with Van’t Hoff’s formula for the solubility coefficients

(1.10)

one obtains the Arrhenius equation for permeability coefficients:

(1.11)

The parameters of these equations are related by a simple formula: .

The most important application of this approach is that it logically explains the temperature dependence of the diffusion and permeability coefficients of gases in polymers. The values of ED are always positive, so the diffusion coefficients always increase when temperature increases. However, the sign of EP depends on the relative magnitudes of EP and ΔHs. For light gases typically , so the resulting values and permeability increases when temperature increases. However, there are rather frequent situations when and enhanced temperature causes decreases in the permeability coefficients. This occurs in vapor permeation, when ΔHs have large absolute values, or in polymers with rigid main chains and unusually low energy barriers for diffusion (small ED values) such as Si‐containing disubstituted polyacetylenes [8] or addition polynorbornenes [9].

The above interpretation of the mechanism of diffusion of small molecules in amorphous polymers suggests a description of diffusion as a sequence of successive, infrequent jump events, with the rate constant for each jump being estimable from the transition‐state theory. The theory well accepted for description of elementary chemical reactions [10] is also applicable for dissolved gas transport in membranes [7, 11].

The parameters ED, EP, and ΔHs are strong functions of molecular size of penetrants. A simple interpretation of this phenomenon was given by Meares’ equation [12]:

(1.12)

where N0 is the Avogadro number, d is the kinetic cross‐section of a diffusant, CED is the cohesion energy density in a polymer, and λ (adjustable parameter) is a diffusion jump length. This equation explains not only the dependence of ED on Vc or d2 but also decreases of diffusivity for gases with larger sizes d or critical volume. Relatively recently it was shown [13] that analysis of Meares’ equation in conjunction with the data of positron annihilation on the size and concentration of free volume elements in polymers can lead to a conclusion that the diffusion jump length λ is close to the average distance between adjacent free volume elements in glassy polymers: that is, this quantity acquired specific physical meaning.

1.2.2 Free volume model

Another alternative model for description of gas transport in polymeric membrane is the free volume model. The notion of free volume is of paramount importance for