Askari Mohammad-Thesis-23 July 2014

CO2/CH4 AND PROPYLENE/PROPANE SEPARATION
USING CROSS-LINKABLE POLYMERIC

MEMBRANES
MOHAMMAD ASKARI
B. Eng. (Shiraz University)
M.Sc. (Sharif University of Technology)
A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF

PHILOSOPHY
DEPARTMENT OF CHEMICAL AND BIOMOLECULAR

ENGINEERING
NATIONAL UNIVERSITY OF SINGAPORE

2013
DECLARATION
I hereby declare that the thesis is my original work and it has been written by
me in its entirety. I have duly acknowledged all the sources of information
which have been used in the thesis.
This thesis has also not been submitted for any degree in any university
previously.
Mohammad Askari
July 2014
Acknowledgment
This thesis would not have been possible without the support, patience and
guidance of several individuals who in one way or another contributed and
extended their valuable assistance in the preparation and completion of this
study.
I wish to first express my sincere appreciation to my academic advisor,

Professor Neal Chung Tai-Shung, for his teaching and guidance. His
unwavering support and relentless drive will remain a source of inspiration in
all of my future endeavors. Over the past four years, he has pushed me to
achieve beyond what I ever imagine and nurtured me as an independent
researcher.
The whole of the Prof. Chung group has been a pleasure to work with, and
friendships forged will remain with me for life. I would especially like to
express thanks to my mentors Dr. N. Peng, Dr. P. Li, and Dr. Y.C. Xiao and
also Ms. M.L. Chua for their assistance, conversation, and general
camaraderie.
I gratefully acknowledge the research scholarship by the National University

of Singapore. I would like to thank the Singapore National Research
Foundation (NRF) for the support on the Competitive Research Program for
the project Molecular Engineering of Membrane Materials: Research and
Technology for Energy Development of Hydrogen, Natural Gas and Syngas
(grant number: R-279-000-261-281).
Perhaps most influential in my graduate school success has been my best

friend, my most trusted confidant, and my wife, Zahra. I am incredibly
thankful for her endless patience and support through countless late-night lab
visits and disrupted weekends, all in the name of science. I cannot imagine my
time at NUS without her, nor can I thank her enough.
i
I owe my loving thanks to my mother, my father, my sisters and my sons,
Kian and Shayan. They have lost a lot due to my research abroad. Without
their encouragement and understanding it would have been impossible for me
to finish this work.
Lastly, I would like to thank God for always being there for me and giving me
strength and hope during difficult times.
ii
Table of Content
ACKNOWLEDGMENT.......................................................................................... I
TABLE OF CONTENT .........................................................................................III
SUMMARY ......................................................................................................... VII
LIST OF TABLES ................................................................................................ XI
LIST OF FIGURES ............................................................................................ XIII
NOMENCLATURE .......................................................................................... XVI
LIST OF ABBREVIATION ............................................................................ XVIII
CHAPTER 1: INTRODUCTION ....................................................................1
1.1 MEMBRANE TECHNOLOGY FOR GAS SEPARATION ..................................4
1.1.1 NATURAL GAS PURIFICATION .............................................................5
1.1.2 CARBONE DIOXIDE CAPTURE .............................................................5
1.1.3 OLEFIN/PARAFFIN SEPARATION ..........................................................6
1.2 MEMBRANE STRUCTURES AND MODULES ..............................................8
1.3 RESEARCH OBJECTIVE AND ORGANIZATION OF DISSERTATION ............10
1.4 REFERENCES .........................................................................................14
CHAPTER 2: THEORY AND BACKGROUND .........................................18
2.1 POLYMERIC MEMBRANES FOR GAS SEPARATION .................................18
2.1.1 GENERAL TRANSPORT THEORY ........................................................19
2.1.1.1 PERMEABILITY COEFFICIENT .....................................................20
2.1.1.2 DIFFUSIVITY COEFFICIENT .........................................................21
2.1.1.3 SOLUBILITY COEFFICIENT..........................................................22
2.1.1.4 PERMSELECTIVITY .....................................................................25
2.1.1.5 PLASTICIZATION BEHAVIOR ......................................................26
2.1.2 CROSS-LINKABLE POLYMERIC MEMBRANES FOR GAS
SEPARATION ...................................................................................................27
2.2 MIXED MATRIX MEMBRANES FOR GAS SEPARATION ...........................31
2.3 REFERENCES .........................................................................................34
CHAPTER 3: MATERIALS AND METHODOLOGY ..............................43
3.1 MATERIALS ..........................................................................................43
3.1.1 POLYMERS ........................................................................................43
3.1.1.1 6FDA-DURENE POLYIMIDE AND 6FDA-DURENE/DABA CO-
POLYIMIDE ..................................................................................................43
iii
3.1.1.2 6FDA-DURENE/DABA (9/1) CO-POLYIMIDE GRAFTED WITH
CYCLODEXTRIN ..........................................................................................44
3.1.2 ZEOLITIC IMIDAZOLATE FRAMEWORKS (ZIFS) NANOPARTICLE
SYNTHESIS ......................................................................................................46
3.1.3 GASES ...............................................................................................46
3.2 MEMBRANE FORMATION.......................................................................46
3.2.1 DENSE FLAT SHEET FILM FORMATION ...............................................46
3.2.1.1 THERMAL CROSS-LINKING OF DENSE FLAT SHEET
MEMBRANES ...............................................................................................47
3.2.2 FABRICATION OF DUAL LAYER HOLLOW FIBER MEMBRANES .............47
3.2.2.1 INNER LAYER AND OUTER LAYER DOPE PREPARATION ..............48
3.2.2.2 SPINNING CONDITIONS AND SOLVENT EXCHANGE ......................49
3.2.2.3 POST TREATMENT OF THE HOLLOW FIBER MEMBRANES .............50
3.2.3 PREPARATION OF DENSE FLAT SHEET MIXED MATRIX MEMBRANE.....51
3.2.3.1 THERMAL CROSS-LINKING OF MIXED MATRIX MEMBRANES .......51
3.3 PHYSICOCHEMICAL CHARACTERIZATION ..............................................52
3.3.1 FOURIER TRANSFORM INFRARED SPECTROMETER (FT-IR) ................52
3.3.2 THERMO GRAVIMETRIC ANALYSES (TGA) ........................................52
3.3.3 DIFFERENTIAL SCANNING CALORIMETRY (DSC)...............................52
3.3.4 DENSITY MEASUREMENT ..................................................................52
3.3.5 WIDE ANGLE X-RAY DIFFRACTION (WAXD) ....................................53
3.3.6 GEL PERMEATION CHROMATOGRAPHY (GPC) ..................................53
3.3.7 FIELD EMISSION SCANNING ELECTRON MICROSCOPY (FESEM) ........53
3.3.8 POSITRON ANNIHILATION LIFETIME SPECTROSCOPY (PALS) ............54
3.4 DETERMINATION OF GAS TRANSPORT PROPERTIES ................................55
3.4.1 PURE GAS SORPTION TEST .................................................................55
3.4.2 PURE GAS PERMEATION TEST ............................................................56
3.4.2.1 DENSE FLAT SHEET MEMBRANES ...............................................56
3.4.2.2 HOLLOW FIBER MEMBRANES .....................................................57
3.4.3 MIXED GAS PERMEATION TEST ..........................................................59
3.4.3.1 DENSE FLAT SHEET MEMBRANE .................................................59
3.4.3.2 HOLLOW FIBER MEMBRANES .....................................................60
3.5 REFERENCES .........................................................................................61
iv
CHAPTER 4: CROSS-LINKABLE 6FDA-DURENE/DABA CO-
POLYIMIDES GRAFTED WITH CYCLODEXTRINS .................................63
4.1 INTRODUCTION .....................................................................................63
4.2 RESULTS AND DISCUSSION ....................................................................65
4.2.1 CHARACTERIZATION .........................................................................65
4.2.2 PURE GAS PERMEATION EXPERIMENTS ..............................................72
4.2.3 MIXED GAS PERMEATION EXPERIMENTS ...........................................76
4.3 CONCLUSION ........................................................................................79
4.4 REFERENCES .........................................................................................81
CHAPTER 5: PERMEABILITY, SOLUBILITY, DIFFUSIVITY AND
PALS DATA OF CROSS-LINKABLE 6FDA-BASED CO-POLYIMIDES..85
5.1 INTRODUCTION .....................................................................................85
5.2 RESULTS AND DISCUSSION ....................................................................86
5.2.1 SOLUBILITY COEFFICIENT OF CO-POLYIMIDE MEMBRANES ................86
5.2.2 PERMEABILITY AND DIFFUSIVITY COEFFICIENT OF CO-POLYIMIDE
MEMBRANES ...................................................................................................94
5.2.3 PERMSELECTIVITY ..........................................................................100
5.2.4 COMPARISON OF DUAL SORPTION BEHAVIOR, SOLUBILITY, AND
DIFFUSIVITY COEFFICIENT OF CO2 IN POLYMERIC MEMBRANES ....................101
5.3 CONCLUSION ......................................................................................102

5.4 REFERENCES .......................................................................................104
CHAPTER 6: CROSS-LINKABLE DUAL-LAYER HOLLOW FIBER
MEMBRANES COMPRISING -CYCLODEXTRIN ..................................109
6.1 INTRODUCTION ...................................................................................109
6.2 RESULTS AND DISCUSSION ..................................................................111
6.2.1 MORPHOLOGY OF THE DUAL-LAYER HOLLOW FIBER MEMBRANES ..111
6.2.2 EFFECT OF TAKE-UP VELOCITY ON GAS SEPARATION
PERFORMANCE ..............................................................................................113
6.2.3 EFFECT OF OUTER-LAYER DOPE FLOW RATE ON GAS SEPARATION
PERFORMANCE ..............................................................................................117
6.2.4 EFFECT OF THERMAL TREATMENT AND SILICON RUBBER COATING

ON GAS SEPARATION PERFORMANCE .............................................................118
6.2.5 CO2 PLASTICIZATION AND MIXED GAS PERMEATION EXPERIMENTS 120
v
6.2.6 COMPARISON OF C3H6/C3H8 SEPARATION PERFORMANCE OF
HOLLOW FIBER MEMBRANES .........................................................................122
6.3 CONCLUSION ......................................................................................124

6.4 REFERENCES .......................................................................................126
CHAPTER 7: THERMAL CROSS-LINKABLE CO-POLYIMIDE /
ZIF-8 MIXED MATRIX MEMBRANES ........................................................129
7.1 INTRODUCTION ...................................................................................129
7.2 RESULTS AND DISCUSSION ..................................................................129
7.2.1 CHARACTERIZATIONS .....................................................................129
7.2.2 PURE GAS PERMEATION EXPERIMENTS ............................................136
7.2.2.1 EFFECT OF ZIF-8 LOADING ON GAS SEPARATION
PERFORMANCE ..........................................................................................136
7.2.2.2 EFFECT OF DURENE TO DABA RATIO ON GAS SEPARATION
PERFORMANCE ..........................................................................................138
7.2.2.3 EFFECT OF THERMAL TREATMENT TEMPERATURE ON GAS
SEPARATION PERFORMANCE ......................................................................139
7.2.3 PLASTICIZATION BEHAVIOR AND MIXED GAS EXPERIMENTS ............140

7.3 CONCLUSION ......................................................................................143
7.4 REFERENCES .......................................................................................146
CHAPTER 8: CONCLUSION AND RECOMMENDATION ..................149
8.1 CONCLUSION ......................................................................................149
8.1.1 CROSS-LINKABLE 6FDA-DURENE/DABA CO-POLYIMIDES
GRAFTED WITH CYCLODEXTRINS ..................................................................150
8.1.2 CROSS-LINKABLE DUAL-LAYER HOLLOW FIBER MEMBRANES
COMPRISING -CYCLODEXTRIN.....................................................................151
8.1.3 THERMAL CROSS-LINKABLE CO-POLYIMIDE/ZIF-8 MIXED MATRIX
MEMBRANES .................................................................................................152
8.2 RECOMMENDATION FOR FUTURE WORK ..............................................153
8.2.1 POTENTIAL FUTURE PROJECT DIRECTIONS .......................................153
8.2.2 HOLLOW FIBER SPINNING OF THE CROSS-LINKABLE MIXED MATRIX
MEMBRANES .................................................................................................154
vi
Summary
Membrane is an emerging technology that holds great promises and displays

attractive advantages over conventional methods. Polymeric membranes,
especially polyimide membranes, have been widely applied for gas separations
due to their attractive permeability, selectivity, and processing characteristics.
However, traditional membrane materials cannot always achieve high degrees
of separation performance and suffer from an upper-bound relationship for its
permeability and selectivity. Their use of natural gas and hydrocarbon
separations is also limited by plasticization-induced selectivity losses in feeds
with significant partial pressures of CO2 and C3+ hydrocarbons. This greatly
constrains the application of polymeric materials for industrial use. In this PhD
work, the main focus is to explicitly tailor the properties of cross-linkable
glassy polymeric membranes for gas separation application. Four aspects have
been thoroughly investigated.
Firstly, the new flexible and high performance gas separation membranes were
fabricated by grafting various sizes of cyclodextrin to the cross-linkable co-
polyimide (6FDA-Durene/DABA (9/1)) matrix and then decomposing them at
elevated temperatures. The gas permeability of thermally treated pristine
polyimide (referred as the original PI) and CD grafted co-polyimide (referred
as PI-g-CDs for 200 and 300 C and partially pyrolyzed membranes (PPM) -
CDs for 350, 400, and 425 C) has been determined. It was observed that
permeability of all tested gases increased with an increase in thermal treatment
temperature from 200 to 425 C. However, permeability increased more for
those grafted with bigger size CD. The permeability of the original PI
thermally treated at 425 C was about 4-6 times higher than that treated at 200
C. The permeability increase jumped to 8-10 times for PPM--CD and 15-17
times for PPM--CD due to CD decompose at high temperatures and bigger
CD creates bigger micro-pores. Interestingly, the permeability ratios of PPM-
-CD to PPM--CD and PPM--CD to PPM--CD at 400 and 425 C were
around 0.6 and 0.8, respectively. These numbers were almost the same as the
cavity diameter ratios of -CD to -CD and -CD to -CD. Permselectivity
decreased first with an increase in thermal treatment temperature up to 350 C
vii
and then increased. Permselectivity of thermally treated CD grafted co-
polyimide membranes were also slightly higher than that of the original PI due
to higher degrees of cross-linking in CD grafted co-polyimide membranes. In
addition, for co-polyimide membranes grafted by CDs, the higher thermal
treatment temperature resulted in membranes with the better plasticization
resistance to CO2 and the better separation performance in 50:50 CO2/CH4
mixed gases.
Secondly, with the purpose of better fundamental understanding of this class

of polymers and aid evaluation of their potential for use in industrial
application, the intrinsic gas transport properties of thermally treated cross-
linkable 6FDA-based co-polyimide membranes have been studied. Grafting
various sizes of Cyclodextrin (CD) to the co-polyimide matrix and then
thermally decomposing CD at elevated temperatures are an effective method
to micro-manipulate microvoids and free volume as well as gas sorption and
permeation. The pressure-dependent solubility and permeability coefficients
were found to follow the dual-mode sorption model and partial immobilization
model, respectively. Solubility and permeability coefficients of CH4, CO2,
C3H6 and C3H8 were conducted at 35 C for different upstream pressures. The
Langmuir saturation constant, C'H, increases with an increase in annealing
temperature. On the other hand, Henrys solubility coefficient kD and
Langmuir affinity constant b do not change noticeably. The CH4 permeability
decreases with pressure while some membranes exhibit serious plasticization
with an increase in CO2, C3H6 and C3H8 pressures. The diffusivity coefficient
of the Henry mode (DD) and Langmuir mode (DH) were calculated from the
permeability and solubility data, and their ratio, F, is higher for membranes
thermally treated at 425 C than those treated at 200 C. Data from positron
annihilation lifetime spectroscopy (PALS) confirm that free-volume and the
number of micro-pores increases while the radius of pore sizes decreases
during the high temperature annealing process. All CD grafted membranes
thermally treated at 425 C have almost equal or lower solubility selectivity
than the original membrane for CO2/CH4 and C3H6/C3H8 separations, but the
former has much higher diffusion selectivity than the latter. As a result,
diffusion selectivity plays a more important role than the solubility in
viii
determining the permselectivity of the CD grafted membranes thermally
treated at 425 C.
Thirdly, considering the importance of hollow fiber for industrial use,

thermally cross-linkable co-polyimide dual-layer hollow fiber membranes
grafted with -Cyclodextrin were fabricated. The fiber membranes were
thermally cross-linked at different temperatures and the performance of the
fibers before and after the silicon rubber coating was studied. It was observed
that the performances of all gases decreased with an increase in take-up
velocity and outer-layer dope flow rate. Selectivities of the membrane with
respect to the take-up velocity initially increased and after a take-up velocity
value of 7.4m/min started to decrease. This up and down trend was attributed
to the influence of the elongational draw ratio and change in surface porosity
of the membrane. Optimum take-up velocity and outer-layer dope flow rate for
as-spun fibers were 7.4 m/min and 0.5 ml/min, respectively. These conditions
resulted in CO2/CH4 selectivity of 6.22 and 14.3 before and after silicon
rubber coating, respectively. The results demonstrate that thermal treatment
improves membrane selectivities and decrease membrane permeances. The
enhancement of selectivities should be a result of cross-linking and reduction
in permeances due to densification of the hollow fiber membranes.
Selectivities of thermally treated fiber membranes at 350 C were slightly
higher than those of the precursor fibers, and this improvement was more
significant for membranes treated at 400 C. This enhancement demonstrates
that cross-linking is more severe at 400 C than 200 and 350 C. The best
separation performance of the annealed and silicone rubber coated hollow
fibers in this study had a CO2 permeance of around 82 GPU with a CO2/CH4
ideal selectivity of around 20 and a high C3H6 permeance of around 29 GPU
with a C3H6/C3H8 ideal selectivity of 15.3. It could also resist CO2 induced
plasticization until 25 atm. It is believed that, with these gas separation and
anti-plasticization properties, the newly developed membranes may have
highly prospective for natural gas purification and olefin/paraffin separation.
Fourthly, to continue from the previous works, the mixed matrix membrane
(MMM) were fabricated by using three 6FDA-based polyimides (6FDA-
ix
Durene, 6FDA-Durene/DABA (9/1), 6FDA-Durene/DABA (7/3)) and nano-
size zeolitic imidazolate framework-8 (ZIF-8) with uniform morphology
comprising ZIF-8 as high as 40 wt% loading. Permeability of all tested gases
increased rapidly with an increase in ZIF-8 loading. However, the addition of
ZIF-8 nano-particles into the polymer matrix increased the CO2/CH4
selectivity only for 6.87%, while the ideal C3H6/C3H8 selectivity improves
134% from 11.68 to 27.38 for the MMM made of 6FDA-Durene/DABA (9/1)
and 40 wt% ZIF-8. Experimental data demonstrated that the plasticization
resistance and gas pair selectivity of MMMs were strongly dependent on the
amount of cross-linkable moiety and annealing temperature. MMMs made of
6FDA-Durene did not show considerable improvements on resistance against
CO2-induced plasticization after annealing at 200-400 C, while MMMs
synthesized from cross-linkable co-polyimides (6FDA-Durene/DABA (9/1)
and 6FDA-Durene/DABA (7/3)) showed significant enhancements in
CO2/CH4 and C3H6/C3H8 selectivity as well as plasticization suppression
characteristics up to a CO2 pressure of 30 atm after annealing at 400 C due to
the cross-linking reaction of the carboxyl acid (COOH) in the DABA moiety.
The MMM made of 6FDA-Durene/DABA (9/1) and 40 wt% ZIF-8 possessed
a notable ideal C3H6/C3H8 selectivity of 27.38 and a remarkable C3H6
permeability of 47.3 Barrer. After thermal annealing at 400 C, the MMM
made of 6FDA-Durene/DABA (9/1) and 20 wt% ZIF-8 showed a CO2/CH4
selectivity of 19.61 and an impressive CO2 permeability 728 Barrer in mixed
gas tests. These newly developed MMMs may have good potential for
industrial nature gas purification and C3H6/C3H8 separation.
x
List of Tables
Table 1.1: Thermodynamic properties of propane and propylene .................... 6

Table 3.1: Properties of the three types of Cyclodextrin ................................. 46
Table 3.2: Spinning conditions for dual layer PI-g--CD/6FDA-Durene hollow
fiber membranes............................................................................................... 50
Table 4.1: Properties of the synthesized co-polyimides before and after
grafting different kinds of CD.......................................................................... 65
Table 4.2: Physical properties of co-polyimide membranes in different heat
treatment temperatures ..................................................................................... 72
Table 4.3: Pure gas permeability and selectivity of co-polyimide membranes
(10 atm and 35 C) ........................................................................................... 73
Table 4.4: Mixed gas permeability and selectivity of co-polyimide membranes
treated in 425 C .............................................................................................. 77
Table 5.1: Solubility, diffusivity, and dual mode sorption parameters of
different co-polyimide membranes for different gasses at 35 C and 1 atm ... 89
Table 5.2: Positron annihilation lifetime spectroscopy (PALS) data of different
co-polyimide membranes ................................................................................. 94
Table 5.3: Diffusivity parameters of different co-polyimide membranes for
different gasses at 35 C .................................................................................. 99
Table 5.4: Contributions of solubility selectivity and diffusion selectivity to
the permselectivity of each membrane for CO2/CH4 and C3H6/C3H8
separations at 35 C and 1 atm....................................................................... 100
Table 5.5: Comparison of dual sorption behavior, solubility, and diffusivity
coefficient of CO2 in polymeric membranes ................................................. 102
Table 6.1: Gas separation performance of dual-layer hollow fiber membranes
at different conditions before silicon rubber coating ..................................... 114
at different conditions after silicon rubber coating ........................................ 114
Table 6.3: Intrinsic gas transport properties of flat dense PI-g--CD co-
polyimide membranes heat treated at different temperatures ....................... 120
Table 6.4: Mixed gas permeance and selectivity of the dual-layer hollow fiber
membranes spun at condition B, thermally treated at 400 C after silicon
rubber coating ................................................................................................ 122
xi
Table 6.5: C3H6/C3H8 separation performance of polyimide and co-polyimide
hollow fiber membranes ................................................................................ 123
Table 7.1: Single gas separation performance of CPI (9/1) ZIF-8 MMMs
with different ZIF-8 loading (Annealed at 200 C during sample preparation)
........................................................................................................................ 131
Table 7.2: Comparison between experimental data and calculated data by
using the Maxwell equation ........................................................................... 137
Table 7.3: Single gas separation performance of pure polyimides and
polyimides with 20 wt% ZIF-8 at different annealing temperatures ............. 138
........................................................................................................................ 142
xii
List of Figures
Figure 1.1: Sources of carbon dioxide emissions globally ............................... 3

Figure 1.2: Global propane and propylene consumption .................................. 3
Figure 1.3: Simulated molecular dimension of propylene and propane ........... 7
Figure 3.1: The synthetic scheme and chemical structure of co-polyimide .... 45
Figure 3.2: Chemical structure of three kinds of Cyclodextrin ...................... 45
Figure 3.3: Chemical structure of a) 6FDA-Durene and b) 6FDA-Durene /
DABA (9/1) grafted with -Cyclodextrin ........................................................ 48
Figure 3.4: Shear viscosity vs. polymer concentration for inner and outer layer
dopes at 25 C .................................................................................................. 49
Figure 3.5: (a) Side view and (b) cross-sectional view of the dual-layer hollow
fiber spinneret ................................................................................................. 50
Figure 3.6: Schematic diagram of the bulk PALS system and the inset
indicates the sample packing structure. ........................................................... 54
Figure 3.7: Schematic diagram of the microbalance sorption cell ................... 55
Figure 3.8: Schematic diagram of the dense film gas permeation testing cell 57
Figure 3.9: Schematic diagram of pure gas permeance test apparatus for
hollow fibers ................................................................................................... 58
Figure 3.10: Schematic diagram of a mixed gas permeation cell for flat sheet
membranes ....................................................................................................... 59
Figure 4.1: A hypothesis of the cross-linked structure a) low temperature
(below 350C) b) high temperature (above 350C) ......................................... 65
Figure 4.2: FTIR-ATR spectra of co-polyimide membranes at different heat
treatment temperatures ..................................................................................... 67
Figure 4.3: Residual weight and decomposition rate vs. temperature for
original co-polyimide and PI-g-CDs ................................................................ 68
Figure 4.4: WAXD patterns of co-polyimide membranes at different treatment
temperatures ..................................................................................................... 69
Figure 4.5: Flexibility of the co-polyimide grafted with -CD membrane
thermally treated at 425 C .............................................................................. 71
Figure 4.6: a) Dimensions of , , and -CD [30] and b) Permeability ratios of
PPM--CD to PPM--CD and PPM--CD to PPM--CD at 400 and 425 C. 75
xiii
Figure 4.7: CO2 plasticization behavior of a) original PI, b) PI-g--CD, c) PI-
g--CD, and d) PI-g--CD at different heat treatment temperatures. .............. 76
Figure 4.8: Trade off lines for CO2/CH4 and C3H6/C3H8 separation (empty
symbols are pure gas results, solid symbols are mixed gas results) ................ 78
Figure 5.1: Sorption isotherms of CH4 and CO2 in different co-polyimide
membranes at 35 C ......................................................................................... 87
Figure 5.2: Sorption isotherms of C3H8 and C3H6 in different co-polyimide
membranes at 35 C ......................................................................................... 88
Figure 5.3: Solubility coefficient of CH4 and CO2 in different co-polyimide
membranes at 35 C ......................................................................................... 90
Figure 5.4: Solubility coefficient of C3H8 and C3H6 in different co-polyimide
membranes at 35 C ......................................................................................... 91
Figure 5.5: Relative solubility coefficient of different gasses for co-polyimide
membranes thermally treated at 425 C compared to Original PI-200 ............ 92
Figure 5.6: Relative kD, b, and CH value of different gasses for co-polyimide
membranes thermally treated at 425 C compared to Original PI-200 ............ 93
Figure 5.7: Permeabilities of CH4 and CO2 in different co-polyimide
membranes as a function of pressure at 35 C ................................................. 95
Figure 5.8: Permeabilities of C3H8 and C3H6 in different co-polyimide
membranes as a function of pressure at 35 C ................................................. 96
Figure 5.9: Plots of permeability coefficients of CH4 and CO2 against (1+bp)-1
for different co-polyimide membranes ............................................................ 97
Figure 5.10: Plots of permeability coefficients of C3H8 and C3H6 against
(1+bp)-1 for different co-polyimide membranes .............................................. 98
Figure 6.1: Cross-section morphologies of the dual-layer hollow fiber
membranes spun at different conditions ........................................................ 112
Figure 6.2: Bulk and surface morphologies of the dual-layer hollow fiber
membranes spun at condition B. .................................................................... 113
Figure 6.3: CO2/CH4 separation performance of dual-layer hollow fiber
membranes at different conditions. ................................................................ 115
Figure 6.4: C3H6/C3H8 separation performance of dual-layer hollow fiber
membranes at different conditions. ................................................................ 117
Figure 6.5: Comparison of cross-sectional morphologies of the fiber
membranes spun at condition B at different thermal treatment temperature . 119
xiv
Figure 6.6: Plasticization behavior of fiber membranes spun at condition B at
different thermal treatment temperature before and after silicon rubber coating
........................................................................................................................ 121
Figure 7.1: Thermo gravimetric analyses of pure CPI (9/1), ZIF-8 and CPI
(9/1) - ZIF-8 MMMs under air atmosphere ................................................... 130
Figure 7.2: ZIF-8 particle size distribution .................................................... 132
Figure 7.3: XRD pattern of pure CPI (9/1), ZIF-8, and CPI (9/1) - ZIF-8
MMMs with various ZIF-8 loading and CPI (9/1) 20 wt% ZIF-8 MMMs at
different annealing temperatures.................................................................... 133
Figure 7.4: FESEM morphologies of CPI (9/1) ZIF-8 MMMs with various
ZIF-8 loadings without annealing .................................................................. 134
Figure 7.5: SEM-EDX mapping for Zn from the cross-section of CPI (9/1)
ZIF-8 MMMs with various ZIF-8 loadings annealing @ 200 C .................. 134
Figure 7.6: FESEM morphologies of CPI (9/1) 20 wt% ZIF-8 MMMs at
different annealing temperatures.................................................................... 135
Figure 7.7: Gas permeability and ideal selectivity trend of CPI (9/1) - ZIF-8
MMMs with various ZIF-8 loadings for a) CO2/CH4 and b) C3H6/C3H8 ...... 136
Figure 7.8: Trade off lines of CO2/CH4 and C3H6/C3H8 separation .............. 139
Figure 7.9: CO2 plasticization behaviour of a) PI 20 wt% ZIF-8, b) CPI (9/1)
20 wt% ZIF-8, and c) CPI (7/3) 20 wt% ZIF-8 at different annealing
temperatures. .................................................................................................. 141
Figure 7.10: Mixed gas CO2 permeability as a function of feed pressure for
CPI (9/1) 20 wt% ZIF-8 membranes. a) thermally treated at 200 C, and b) at
400 C. The permeability curves obtained with CO2/CH4 (50/50) at 35 C. . 143
xv
Nomenclature
A Effective area of membrane (cm2)

b Langmuir affinity constant (atm -1)
bA Affinity constant of component A for the polymer (atm-1)
C Total sorption amount of polymer (cm3 gas (STP) /cm3 polymer)
CD Concentrations of the penetrant sorbed at the Henry site (cm3
gas (STP) /cm3 polymer)
CH Concentrations of the penetrant sorbed at the Langmuir site (cm3
gas (STP) /cm3 polymer)
C'H Langmuir saturation constant (cm3 gas (STP) /cm3 polymer)
C'HA Langmuir capacity coefficient for component A ((cm3 gas
(STP)) / (cm3 polymer))
d Dimension spacing ()

D Average diffusivity coefficient (cm2/s)
D(C) Concentration dependent diffusion coefficient (cm2/s)
DD Henrys law diffusion coefficient (cm2/s)
DDA Henrys law diffusion coefficient of the penetrant A (cm2/s)
DH Langmuir mode diffusion coefficient (cm2/s)
DHA Diffusion coefficient of the penetrant A in the microvoid
environment (cm2/s)
F Mobile fraction of the Langmuir mode species
FFV Fractional free volume
I3 Intensity of ortho-Positronium
J permeation flux (mole / cm2.s)
kD Henrys solubility coefficient ((cm3 gas (STP)) / (cm3 polymer -
atm))
kDA Henrys law constant for component A ((cm3 gas (STP)) /
(cm3 polymer -atm))
L Membrane thickness (cm)
Mn Number average molecular weight (g/mol)
n Number of fibers
Ni Flux of a penetrant i (kg/m2.s)
xvi
p Pressure of the feed (atm)
pA Partial pressure of gas A (atm)
pB Partial pressure of gas B (atm)
P Gas permeability (Barrer)
PA Permeability of gas A (Barrer)
PB Permeability of gas B (Barrer)
P Average permeability coefficient (Barrer)

P2 Upstream operating pressure (psia)
Q Gas rate (cm3/s)
R Mean free-volume radius R (nm)
S Solubility coefficient ((cm3 gas (STP)) / (cm3 polymer -atm))
T Temperature (K)
v Specific volume (cm3/g)
v0 Occupied volume (cm3/g)
vw Van der waals volume (cm3/g)
V Volume of downstream chamber (cm3)
Wp Mass with polymer sample (g)
WSP Mass without polymer sample (g)
x Distance across the membrane (cm)
X Mole fraction in feed stream
y Distance through the membrane (cm)
Y Mole fraction in permeate stream
A/B Ideal Selectivity
Diffraction Angle ()
X-ray Wavelength ()
3 Ortho-Positronium lifetime (ns)
xvii
List Of abbreviation
AFM Atomic Force Microscopy
ASRC After Silicon Rubber Coating
BSRC Before Silicon Rubber Coating
BTDA 3,3',4,4'-biphenyltetracarboxylic Dianhydride
CD Cyclodextrin
CH4 Methane
CMS Carbon Molecular Sieve
CO2 Carbon Dioxide
CPI Co-Polyimide
C3H6 Propylene
C3H8 Propane
DAB Diaminobutane
DABA 3,5-diaminobenzoic Acid
DAM 2,4,6-trimethyl-1,3-diaminobenzene
DDBT Dimethyl-3,7-diaminodiphenylthiophene -5,5-dioxide
DSC Differential Scanning Calorimetry
Durene 2,3,5,6-tetramethyl-p-phenylenediamine
EDX Energy Dispersive X-ray Spectroscopy
FESEM Field Emission Scanning Electron Microscopy
FT-IR Fourier Transform Infrared Spectrometer
FFV Fractional Free Volume
GC Gas Chromatography
GPC Gel Permeation Chromatography
H2S Hydrogen Solfide
ID Inner Diameter
MMM Mixed Matrix Membrane
MOF Metal Organic Framework
mPD 1,3-phenylenediamine
NMP N-methyl-2-pyrrolidinone
OD Outer Diameter
ODA 4,4'-oxydianiline
o-PS Ortho-Positronium
xviii
PALS Positron Annihilation Lifetime Spectroscopy
PAMAM polyamidoamine
PDI Polydispersity Index
PI Polyimide
PPM Partially Pyrolyzed Membrane
SSZ Surface Modified zeolite
Tg Glass Transition Temperatures
TGA Thermo Gravimetric Analyses
UV Ultraviolet
WAXD Wide Angle X-ray Diffraction
ZIF Zeolitic Imidazolate Frameworks
6FDA 4,4'-(hexafluoroisopropylidene) diphthalic anhydride
xix
Chapter 1: Introduction
Global warming, by means of an increase in the average temperature of earth

mainly attributed to the release of greenhouse gases (i.e., Water vapor, CO2,
methane and nitrous oxide) into the atmosphere, has become a great issue
since mid-20th century. According to studies from the Intergovernmental
Panel on Climate Change (IPCC), the immediate effect of global warming will
lead to the sea level rise and an increase in frequency of some extreme weather
events, e.g., heat waves, tropical cyclones, flood, etc [1]. Based on a recent
study by Yamasaki [2], among all greenhouse gases, the high proportion of
CO2 contributes about 80 percent of global warming effects. Thus, to preserve
our earth and let our descendants have a habitable world, the CO2 capture
instead of releasing them into the atmosphere has become an exigent task in
this century. Figure 1.1 shows the sources of carbon dioxide emission. As can
be seen, almost all CO2 emissions come from burning of fossil fuels. In order
to efficiently capture CO2 in petrochemical and refineries, one needs to
separate CO2 from other gas species. One of the main separation processes, in
prospect of CO2 captures, is natural gas purification (CO2/CH4 separation) [4].
Besides, natural gas plays an important role in todays energy production. It

has been widely used as the energy source for domestic appliances,
manufacturing of metals and chemicals, electricity generation as well as the
natural gas powered vehicles. Recent statistics showed that approximately
50% of electricity in the U.S. was generated by combustion of natural gas, and
moreover, it was expected to increase dramatically over the next 25 years
[5,6]. The demand for natural gas is continuously growing due to the fact that
not only natural gas is a clean and efficient fuel, but also a principal feedstock
for the manufacture of many essential chemicals. Removal of acid gases (CO2)
is one of the major steps in natural gas purification because it can (i) increase
the heating value of natural gas, (ii) decrease the volume of gas to be
transported in pipelines and cylinders, (iii) prevent corrosion of pipeline
during gas transport and distribution and (iv) reduce atmospheric pollution [7-
1
9]. Therefore, CO2/CH4 separation is a very necessary process for CO2 capture
and natural gas sweetening.
The other important separation process in industries is olefin/paraffin

separation. Propylene is the second highest petrochemical feedstock after
ethylene. Propylene is a raw material for a wide variety of products including
polypropylene, which is used in packaging and other essential applications
such as automotive components, textiles, and laboratory equipment [10-12].
Figure 1.2 shows the major uses of propylene and propane globally [13-14].
The production of polymers and other special chemicals from olefins such as
propylene requires extremely high purity olefins (> 99.9%). Since light olefins
are commonly produced together with corresponding paraffins, the
olefin/paraffin separation process in the petrochemical industries is crucial
[15].
Separation of gases is one of practical but very important unit operations in

chemical and petrochemical industries such as recovery of hydrogen from
product streams of ammonia plants, separation of methane from the other
components of biogas, enrichment of air by oxygen for medical or
metallurgical purposes, removal of hydrogen, water vapor, CO2, and H2S from
natural gas (natural gas purification) and olefin/paraffin separation [16].
Capital and operating cost of these separation processes can account for more
than 50% of the production cost in chemical and petroleum refining industries.
Currently, amine absorption and pressure swing adsorption are the major
methods for natural gas purification and the separation of olefin and paraffin
mixtures is typically performed using rectification, adsorption, and cryogenic
distillation [7-8,15-18]. However, these methods are expensive and energy
intensive. The increase in energy cost due to the resource depletion has raised
the demands for the development of low energy separation technologies.
Membrane separation, compared to amine absorption, pressure swing

adsorption, rectification and cryogenic distillation, has advantages, including
low energy consumption, easy operation and maintenance, environmental
benign and small footprint [19-21]. Many studies have been done on
2
membrane separation for the purpose of replacing traditional separation
technologies [8]. Membranes, especially polymeric membranes have been
explored in various kinds of gas separation applications such as natural gas
sweetening [7-9,20-21], and olefin/paraffin separation [22-29]. In order to
have a good gas separation performance, membranes must have high
permeability and permselectivity, excellent chemical resistance (for resistance
against corrosive materials such as H2S), great thermal stability (for high
temperature applications), good mechanical properties (for high pressure
applications), and superior plasticization resistance [20, 30-31].
Figure 1.1: Sources of carbon dioxide emissions globally [4]
Figure 1.2: Global propane and propylene consumption [13,14]
3
The following sections introduce the membrane technology in gas separation
applications, membrane structure and modules and research objective. A more
detailed discussion of previous and current research will be presented in
Chapter 2.
1.1 Membrane Technology for Gas Separation

Compared with conventional gas separation processes such as cryogenic
separation, physical adsorption and chemical absorption, membrane-based gas
separation has played a significant role in gas separation industries [8]. This is
fairly visualized with a less energy consumption for membrane gas separation
since it does not require phase displacements as does in cryogenic separation.
On the other hand, membrane gas separation is operated by continuous
separation and does not require intermittent cycles as does in physical
adsorption. Furthermore, it does not consume chemical that reveals the
characteristics of environmental friendliness, whereas, substantial amount of
toxic and corrosive chemicals are used in the chemical absorption process. The
lack of mechanical complexity with the absence of moving parts in membrane
systems is another advantage of membrane separation. Thus, membrane
separation allows a simpler system of operation and can be accomplished with
small footprints. This is particularly suited for use in remote applications such
as offshore gas-processing platforms [32].
Gas separation membranes currently comprise a market of over 150 million

U.S dollars per year. This number is expected to rise to around 750 million
dollars by 2020 [8]. Moreover, it is expected that the membrane gas separation
will play an increasingly important role in reducing the environmental impact
and costs of industrial processes, particularly with the concern of global
warming, a direct impact of fossil fuel usage with the increase of carbon
dioxide (i.e., CO2) concentration in the atmosphere. Among the different
options that can prevent carbon dioxide from build-up, such as processes with
enhanced energy efficiency, an increased use of renewable energy sources or
the development of non-CO2 emitting energy sources, carbon capture and
storage is considered a key issue. This, on the other hand, opens another
opportunity for membrane-based gas separation process and technology for
4
CO2 capture. Membrane technology has been used commercially for some gas
separation applications since 1980. The major applications that we focused in
this study are introduced below.
1.1.1 Natural Gas Purification

Natural gas is a gas mixture consisting primarily of methane with up to 20% of
other hydrocarbons as well as impurities in varying amounts such as carbon
dioxide. The composition of raw natural gas varies widely. Beside the main
component of methane, it contains significant amounts of low hydrocarbon
and a small amount of undesirable impurities such as carbon dioxide, nitrogen
and hydrogen sulfide. Consequently, treatment is required to meet the transfer
condition in the pipeline. Natural gas is usually produced at high pressure and
to conserve the energy membrane should be designed to remove impurities
into the permeate stream and to leave methane, ethane and other low
hydrocarbons in the high pressure feed side. This can eliminate the
recompression process which is highly energy intensive. The first membrane
systems for CO2 removal were installed in the early 1980s. Since then, the use
of membranes in natural gas processing has been dominated by CO2 removal.
Traditional membranes used for acid gas removal are constructed from glassy
polymers, such as cellulose acetate or polyimide. These polymer membranes
remove both CO2 and H2S from natural gas streams, but CO2 is the focus
molecule due to its small size and high contaminant concentration in many
streams [33].
1.1.2 Carbone Dioxide Capture

Carbon dioxide is the main greenhouse gas which causes the global warming.
With the high concern of global warming and severe climate changes, carbon
dioxide separation has paramount importance in the industry because the high
proportion of CO2 contributes about 80 percent of global warming effects [3].
Capturing carbon dioxide from both pre- and post-combustion then sequesters
it in a secured place is an important task to mitigate the global warming issue.
Amine absorption is the most widely used technology in removing carbon
dioxide from the mixture. The high capital, operational and maintenance costs
of this technology make people searching for an alternative technology.
5
Membrane technology shows great potential to remove carbon dioxide.
Unfortunately, it is only attractive for the small scale of operation (<5 million
SCFD), and costs are still too expensive to compete with amine absorption if
the system needs to handle more than 40 million SCFD [34]. In general,
despite the advantage of small footprint, low maintenance cost, membrane
technology is still not competitive with the current amine absorption
technology unless the performance of the membrane increases significantly. In
the application of carbon dioxide capture, the membrane is favored only in
offshore platforms where constrain of space is the main concern.
1.1.3 Olefin/Paraffin Separation

Two of the most important petrochemicals are the olefins ethylene and
propylene. In 2004, 146 and 82 billion pounds of ethylene and propylene,
respectively, were produced worldwide [35]. Both are feed stocks for many
other important chemical products; the most important being polyethylene and
polypropylene. Worldwide, 72 billion pounds of polyethylene and 42 billion
pounds of polypropylene were produced in 2004. Propylene and propane are
found in the low and middle fractions of the distillation process, albeit in small
amounts. Large amounts are formed when gasoline is made by cracking or
reforming. Other olefins and paraffins can be found throughout the refining
process. The separation of these two has been relatively more difficult than
that of other gas pairs, mainly ascribed to their close thermodynamic and
physical properties. The simulated molecular dimension of propylene and
propane is presented in Figure 1.3 while the thermodynamic properties of both
are tabulated in Table 1.1.
Table 1.1: Thermodynamic properties of propane and propylene [23]

Molecular Molecular Weight Boiling Point Critical Temp.
Formula (g/mol) (K) (K)
Propylene C3H6 42 226 365.2
Propane C3H8 44 230.4 369.9
6
Figure 1.3: Simulated molecular dimension of propylene and propane [23]
As can be seen, both gases have similar molecular dimensions, boiling points
and critical temperatures. The only difference is that propane is saturated
hydrocarbon and propylene is unsaturated hydrocarbon. Compounds of carbon
and hydrogen whose adjacent carbon atoms contain only one carbon-carbon
covalent bond are known as saturated hydrocarbons. They are called saturated
compounds because all the four bonds of carbon are fully utilized and no more
hydrogen or other atoms can attach to it. These saturated hydrocarbons are
called alkanes and the general formula for an alkane is CnH2n+2. Due to the
presence of all single covalent bonds, these compounds are less reactive and
the number of hydrogen atoms is more when compared to its corresponding
unsaturated hydrocarbon. Compounds contain single carbon-Compounds of
carbon and hydrogen that contain one double bond between carbon atoms
(carbon=carbon) or a triple bond between carbon atoms (carboncarbon) are
called unsaturated hydrocarbons. Unsaturated hydrocarbons can be divided
into alkenes and alkynes depending on the presence of double or triple
bonds respectively. The general formulae are CnH2n for alkenes and CnH2n-2
for alkynes. These compounds are more reactive and their high reactivity is
due to the presence of pi-bond in their structure. Then sorption and adsorption
of these unsaturated hydrocarbon (alkene) in polymer matrix or inorganic
particles could be higher than saturated hydrocaron (alkan). Although
membranes can lead to extensive energy and cost savings, many researchers
envision membrane separation units fitted in line with current separation
systems rather than completely replacing them.
7
1.2 Membrane Structures and Modules
There are many different ways to fabricate a membrane such as solution
casting, melt spinning, wet spinning, track etching and sol-gel process. The
way of membrane fabrication results in a different structure of the ultimate
membrane. Generally, membrane structure can be categorized into four
different types: (a) symmetric, (b) asymmetric, (c) asymmetric composite and
(d) micro-porous composite membranes [36].
The membrane has a symmetric structure always named dense membrane. It

has an identical structure over the entire cross section of the membrane and
this type of membrane, usually prepared by solution casting method with
controlling the evaporation rate of the solvent. Economically, symmetric
membrane is not commercially viable due to the thick, dense layer which
hinder the performance of the membrane. However, it can be used for
fundamental investigation on the intrinsic properties of the membrane
material. The valuable information obtained from the fundamental
investigation provides guidance in the subsequent fabrication of asymmetric
membranes.
The asymmetric membrane consists of a number of layers, each with different

structures and permeability. A typical asymmetric membrane has a dense
selective layer and a porous substrate. The thin dense selective layer separates
the gas molecules, while the porous substrate provides mechanical strength to
the membrane. The asymmetric structure can be obtained by either solution
casting or wet spinning technique. It has a graded pore structure and
frequently from the same material across its thickness. The porous structure in
the support minimizes the substructure resistance which in turn enhances the
gas flux of the membrane. Asymmetric membrane has very thin dense
selective layer. In other words, it has a much higher gas flux compared to the
membrane with symmetric structure. Consequently, the asymmetric membrane
is more prevalent in industrial applications. Under the circumstance that the
defect-free selective layer is not attainable upon optimizing the membrane
fabrication protocol, the membrane still can be repaired by sealing the minor
defects on the membrane surface using silicone rubber. This type of membrane
8
is categorized as asymmetric composite membrane. Asymmetric composite
membrane normally has two or more distinctively different layers made in
different steps. If the material of the dense selective layer is not compatible
with the substrate, the integrity between the two materials will affect the
mechanical stability of the membrane significantly. In this case, a gutter layer,
which is compatible with both materials, should be adopted to enhance the
adhesion of the two layers.
In order to handle a large amount of industrial gas feed and apply membranes
on a technical scale, large membrane areas are required. Thus, it is desired to
pack the membrane area into a small unit, called membrane module. The
design of the membrane module needs to be carefully chosen for each
separation process and operational conditions. Unlike the membranes in
water/wastewater treatment process, the cleaning is of less importance in gas
separation. The main interest of module design is a high packing density such
that a high ratio of membrane area to module volume is feasible to minimize
manufacturing costs. There are three major module types for gas separation
processes, plate-and-frame, spiral-wound, and hollow fiber [37].
Depending on the process applications, different types of membrane modules

can be applied with the consideration of cost, membrane fouling and
concentration polarization [38]. Hollow fiber modules by far, is the most
favorable design in the industry due to its high packing density up to 30000
m2/m3, and it has the lowest cost per unit membrane area [38]. The
shortcoming of this module design is the poor fouling resistance. However, the
gaseous feed streams can easily be filtered, and the fouling problem is not
applicable in gas separation applications. Although fouling is not a serious
issue in gas separation, concentration polarization does affect the separation
efficiency in the module. A cross-flow hollow fiber module is commonly used
to obtain better flow distribution and reduce concentration polarization. In a
typical hollow fiber module, there may be thousands of hollow fibers
assembled together. Each hollow fiber consists of a thin functional layer and a
porous, non-selective support layer. The free ends of the fibers are potted with
agents such as epoxy resins. Self-supporting is another big advantage in
9
hollow fiber membranes. However, the major drawback associated with
hollow fiber modules is the significant pressure drops at the permeate side,
which might require additional energy to compress the wanted permeate gas
for transportation.
1.3 Research Objective and Organization of Dissertation

In view of the above review, polymeric membranes should have high gas
permeability, permselectivity and good operational stability to compete with
conventional technologies in the industrial applications. Currently, only a
small share of the available polymer materials has been used to make at least
90% of the total installed gas separation membrane base [8] and most of them
are rigid glassy polymers. Several hundred new materials are not fit for
industrial application. To compete with well-established conventional
separation processes and extend their applications further, polymeric
membranes should also have good mechanical properties, thermal/chemical
resistance, superior plasticization resistance and physical aging. Most
importantly, polymeric membranes should have ultra-high permeability and
good selectivity in order to treat large volumes of industrial gases. However,
polymeric membranes still face two major challenges.
There is a limitation in achieving the desired performance of a high
permeability combined with a high permselectivity. There is a trade-off
relation between permeability and permselectivity for most gas pairs
[19].
The separation performance of polymeric membranes generally
degrades when separating highly soluble gases such as CO2 or
hydrocarbons such C3H6 at relatively high pressures. This phenomenon
is referred as plasticization [31,39,40]. In the occurrence of
plasticization, the gas pair selectivity is reduced and the ideal
selectivity measured by means of pure gas tests can no longer be used
to estimate the mixed gas membrane performance [41].
In an effort to achieve enhanced membrane gas separation performance, the
main objective of this research work was to tailor the membrane properties by
modifying glassy polymeric membranes for various gas separation
applications. The specific objectives of this research were to:
10
Synthesize co-polyimide grafted with thermal liable molecules for
CO2/CH4 and C3H6/C3H8 separation. Three Cyclodextrins (CDs) with
different sizes will be grafted on the 6FDA-based co-polyimide by
esterification [30]. Thermal treatment of the co-polyimide will be
conducted from 300 to 425 C to study the effects of different
annealing temperatures on the membrane separation performance.
After thermal treatment, the decomposed CD may leave different sizes
of microvoids in the membrane and effect on membrane separation
performance.
Fabricate of dual-layer hollow fiber membranes with this modified
glassy co-polyimide due to the great importance of hollow fiber for
industrial use.
Synergistically combine the strengths of ZIF-8 and cross-linkable
6FDA-based polyimide and molecularly design MMMs for natural gas
purification and olefin/paraffin separation. These cross-linkable 6FDA-
based polymers will be chosen because of their impressive
performance for CO2/CH4 and C3H6/C3H8 separation [30]. The effects
of (1) diamine ratio (i.e., The ratio of Durene to DABA monomers in
the co-polyimide structure), (2) annealing temperature (i.e., Different
degrees of cross-linking) and (3) plasticization phenomenon on
membrane separation performance will be systemically investigated.
The result of this present study may contribute to:
Exploring the science and engineering if one can incorporate different
sizes of CD and manipulate the free volume and its size with enhanced
permeability and selectivity of the resulting membrane via by thermal
degradation and partial cross-linking of the polymer matrix.
Investigating the effect of ZIF-8 nano-particle loading and thermal
treatment modification of gas separation performance of these 6FDA-
based co-polyimide membranes.
The focus of this study is on thermal cross-linking of this special 6FDA-based
co-polyimide, and the temperature of thermal treatment is not more than 425
C (Partially pyrolyzed). It is not the task of this study to investigate the high
temperature treatment that makes carbon molecular sieve membranes (CMS)
because carbon molecular sieve membranes are brittle and less flexible than
11
polymeric membranes and partially pyrolyzed membranes. Therefore, it is
difficult to use these kinds of membranes for industrial applications.
This dissertation is organized such that general information has been covered
before introducing the specific findings of this work.
Chapter 2 summarizes background information, a review of significant

literatures and the fundamental theory of the topics investigated in this work.
Chapter 3 documents the experimental approaches and methodologies along

with the materials involved in all areas. Additionally, a detailed description of
the membrane characterizations, including both physical properties of
membranes and gas transport properties is provided.
Chapter 4 reports the preparation of cross-linkable co-polyimide (6FDA-

Durene/DABA (9/1)) membrane grafted with various sizes of Cyclodextrin
(CD). The post treatment is carried out by thermal treatment at different
temperatures from 300 to 425 C. The changes in the physical properties of the
membranes by grafting various sizes of Cyclodextrin to the polyimide matrix
and then decomposing them at elevated temperatures are monitored by TGA,
FTIR, DSC and XRD analyses. Additionally, the gas transport properties were
discussed at different annealing temperature.
Chapter 5 investigates the intrinsic gas permeation properties of 6FDA-

Durene/DABA (9/1) co-polyimide grafted with different size of CDs over a
wide range of pressures and the effects of annealing temperatures on gas
permeability, diffusivity, and solubility coefficients.
Chapter 6 discusses the preparation of cross-linkable dual-layer hollow fiber

membranes with this modified glassy co-polyimide due to the great
importance of hollow fiber for industrial use. The structural properties of the
hollow fiber membranes at different annealing temperature were monitored by
FESEM analyses.
12
In Chapter 7, the effect of ZIF-8 nano particle on the gas separation
performance of cross-linkable 6FDA-based co-polyimide mixed matrix
membrane is investigated. The physical properties of the mixed matrix
membranes at different annealing temperature and different ZIF-8 loading
were monitored by TGA, XRD, FESEM, and EDX analyses.
The general conclusions drawn from this research study are summarized in
Chapter 8. Recommendations for future works are proposed to consider the
industrial applicability of the developed membranes.
13
1.4 References
[1] IPCC Fourth Assessment Report: Climate Change 2007 (AR4),
www.ipcc.ch/publications_and_data/publications_and_data_reports.ht
m.
[2] A. Yamasaki, An overview of CO2 mitigation options for global
warming - emphasizing CO2 sequestration options, J Chem. Eng. Jpn.
36 (2003) 361-375.
[3] IEA Energy Technology Essentials on CO2 capture and storage 2006,
http://www.iea.org/techno/essentials1.pdf
[4] Inventory of U.S. greenhouse gas emissions and sinks: 19902010,
EPA 430-R-12-001, (2012).
[5] Electric generation using natural gas, website for organization of
natural gas. http://www.naturalgas.org/overview/uses_eletrical.asp.
[6] International Energy Outlook 2013; DOE/EIA-0484 (2013), U.S.
Energy Information Administration.
[7] B.D. Bhide, A. Voskericyan, S.A. Stern, Hybrid processes for the
removal of acid gases from natural gas, J. Membr. Sci. 140 (1998) 27-
49.
[8] R.W. Baker, Reviews future directions of membrane gas separation
technology, Ind. Eng. Chem. Res. 41 (2002)1393-1411.
[9] Y.C. Xiao, B.T. Low, S.S. Hosseini, T.S. Chung, D.R. Paul, The
strategies of molecular architecture and modification of polyimide
based membranes for CO2 removal from natural gas-A review, Prog.
Polym. Sci. 34 (2009) 561-580.
[10] M. Das, W.J. Koros, Performance of 6FDA6FpDA polyimide for
propylene/propane separations, J. Membr. Sci. 365 (2010) 399-408.
[11] M.T. Castoldi, J.C. Pinto, P.A. Melo, Modeling of the Separation of
Propene/Propane Mixtures by Permeation through Membranes in a
Polymerization System, Ind. Eng. Chem. Res. 46 (2007) 1259-1269.
[12] R. Faiz, K. Li, Polymeric membranes for light olefin paraffin
separation, Desalination 287 (2012) 82-97.
[13] Consumes of propane,
http://www.conocophillips.com/EN/about/energy/energytypes/pages/pr
opane.aspx
14
[14] Dow chemical, product uses of propylene
http://www.dow.com/productsafety/finder/pro.htm
[15] R.B. Eldridge, Olefin/paraffin separation technology: A review, Ind.
Eng. Chem. Res. 32 (1993) 2208-2212.
[16] F.G. Kerry, Industrial gas handbook: Gas separation and purification,
Taylor and Franc. Group, New York, 2006.
[17] S.U. Rege, J. Padin, R.T. Yang, Olefin/paraffin separations by
adsorption: pi-complexation vs. kinetic separation, AIChE J. 44 (1998)
799-809.
[18] M. Azhin, T. Kaghazchi, M. Rahmani, A review on olefin/paraffin
separation using reversible chemical complexation technology, J. Ind.
Eng. Chem. 14 (2008) 622-638.
[19] L.M. Robson, Upper-bound revisited, J. Membr. Sci. 320 (2008) 390-
400.
[20] S.S. Hosseini, Y. Li, T.S. Chung, Y. Liu, Enhanced gas separation
performance of nanocomposite membranes using MgO nanoparticles,
J. Membr. Sci. 302 (2007) 207-217.
[21] P.M. Budd, N.B. Mckeown, Highly permeable polymers for gas
separation membranes, Polym. Chem. 1 (2010) 63-68.
[22] R.L. Burns, W.J. Koros, Defining the challenges for C3H6/C3H8
separation using polymeric membranes, J. Membr. Sci. 211 (2003)
299-309.
[23] M.L. Chng, Y.C. Xiao, T.S. Chung, M. Toriida, S. Tamai, Enhanced
propylene/propane separation by carbonaceous membrane derived
from poly (aryl ether ketone)/2,6-bis(4-azidobenzylidene)-4-methyl-
cyclohexanone interpenetrating network, Carbon 47 (2009) 1857-1866.
[24] K. Tanaka, A. Taguchi, J. Hao, H. Kita, K. Okamoto, Permeation and
separation properties of polyimide membranes to olefins and paraffins,
J. Membr. Sci. 121 (1996) 197-207.
[25] S. Sridhar, A.A. Khan, Simulation studies for the separation of
propylene and propane by ethylcellulose membrane, J. Membr. Sci.
159 (1999) 209-219.
[26] M. Yoshino, S. Nakamura, H. Kita, K. Okamoto, N. Tanihara, Y.
Kusuki, Olefin-paraffin separation performance of asymmetric hollow
15
fiber membrane of 6FDA,BPDADDBT co-polyimide, J. Membr. Sci.
212 (2003) 13-27.
[27] C.S. Bickel, W.J. Koros, Olefin-paraffin gas separations with 6FDA-
based polyimide membranes, J. Membr. Sci. 170 (2000) 205-214.
[28] S.S. Chan, R. Wang, T.S. Chung, Y. Liu, C2 and C3 hydrocarbon
separations in poly(1,5-naphthalene-2,2'-bis(3,4-phthalic)
hexafluoropropane) diimide (6FDA-1,5-NDA) dense membranes, J.
Membr. Sci. 210 (2002) 55-64.
[29] S. Basu, A.C. Odena, F.J. Vankelecom, Asymmetric Matrimid / [Cu3
(BTC) 2] mixed matrix membranes for gas separations, J. Membr. Sci.
362 (2010) 478-487.
[30] Y.C. Xiao, T.S. Chung, Grafting thermally labile molecules on cross-
linkable polyimide to design membrane materials for natural gas
purification and CO2 capture, Energy Environ. Sci. 4 (2011) 201-208.
[31] Y.C. Xiao, T.S. Chung, H.M. Guan, M.D. Guiver, Synthesis,
crosslinking and carbonization of co-polyimides containing internal
acetylene units for gas separation, J. Membr. Sci. 302 (2007) 254-264.
[32] P. Bernardo, E. Drioli, G. Golemme, Membrane gas separation: A
review/state of the art, Ind. Eng. Chem. Res., 48 (2009) 4638-4663.
[33] A. Tabe-Mohammadi, A review of the applications of membrane
separation technology in natural gas treatment. Sep. Sci. Technol. 34
(1999) 2095- 2111.
[34] D. Dortmundt, K. Doshi, CO2 removal membrane technology recent
developments, Chem. Eng. World 38 (2003) 55-66.
[35] Facts & Figures for the Chemical Industry. Chem. and Eng. News 83
(2005) 41-81.
[36] T.S. Chung, A review of micro-porous composite polymeric
membrane technology for air separation. Polym Polym Compos 4
(1996) 269-282.
[37] W.J. Koros, G.K. Fleming, Membrane-based gas separation, J. Membr.
Sci., 83 (1993) 1-80.
[38] S.A. Stern, T.F. Sinclaire, P.J. Gareis, N.P. Vahldieck, P.H. Mohr,
Recovery by permeation, Ind. Eng. Chem. 57 (1965) 49-58.
16
[39] S. Neyertz, D. Brown, S. Pandian, N.F.A. van der Vegt, Carbon
dioxide diffusion and plasticization in fluorinated polyimides,
Macromolecules 43 (2010) 7813-7827.
[40] T. Visser, M. Wessling, Auto and mutual plasticization in single and
mixed gas C3 transport through Matrimid-based hollow fiber
membranes, J. Membr. Sci. 312 (2008) 84-96.
[41] L.S. White, T.A. Blinka, H.A. Kloczewski, I. F. Wang, Properties of a
polyimide gas separation membrane in natural gas streams, J. Membr.
Sci. 103 (1995) 73-82.
17
Chapter 2: Theory and Background
2.1 Polymeric Membranes for Gas Separation

Polymeric membranes have been explored in various kinds of gas separation
applications such as natural gas sweetening [1-3], and olefin/paraffin
separation [4-10]. In order to have a good gas separation performance,
membranes must have high permeability and permselectivity, excellent
chemical resistance (for resistance against corrosive materials such as H2S),
great thermal stability (for high temperature applications), good mechanical
properties (for high pressure applications), and superior plasticization
resistance [11-13].
Depending on the glass transition temperature of the membrane material, the

polymers are generally classified as glassy polymers and rubbery polymers.
Generally, a polymer is recognized as rubbery polymer if its Tg is well below
the room temperature. Otherwise, the polymer is glassy in characteristics.
Glassy polymers are characterized by hard and brittle moiety with restricted
chain mobility due to their high glass transition temperatures. Unlike rubbery
polymers, glassy polymers are in a non-equilibrium state, which makes the
diffusion in glassy polymers more complex compared to that in rubbery
polymers. In glassy polymers, the mobility term is usually dominant, and gas
permeability increases with decreasing gas penetrant size. The gas separation
is mainly achieved by size discriminating of the gas penetrants, and the
diffusivity selectivity dominates the overall selectivity of the membrane.
Conversely, in rubbery polymers, the permeability increases with increasing
gas penetrant size and larger gas molecules permeate preferentially. Therefor
the solubility selectivity is dominant in rubbery polymeric membranes [14-16].
Among the polymeric membranes for gas separation application, polyimides

are one of the most attractive and favorable materials due to their excellent
properties such as high thermal stability, chemical resistance, mechanical
strength, and impressive performance for CO2/CH4 and olefin/paraffin
18
separation. Moreover, these polymers can be tailored to yield varying transport
properties for optimized separations [17-19].
2.1.1 General Transport Theory

With the purpose of understanding membrane mechanism and performance,
the structure of the membranes must be considered. Structural information on
the size and distribution of micro-voids or free volume in the glassy state has
been used to interpret the transport properties of polymeric membranes.
Depending on the membrane structures, the gas transport mechanism is
different from one to another. The separation of gases in membrane with
porous structure is based on molecular size through the small pores in the
membrane matrix. On the other hand, most of commercial polymeric
membrane is based on nonporous membranes, which follow the solution-
diffusion mechanism. It is theoretically assumed that penetrant molecule first
sorbs into a polymer before diffusing in the material then diffuse across it and
finally desorb into the downstream gas phase side [14]. The transport of gas
molecules through the nonporous polymer membrane occurs due to random
molecular motion of individual molecules. This process can be described in
terms of Ficks first law of diffusion.
Understanding the glassy state of amorphous polymers is very important in

understanding physical properties and applying these polymers into industrial
processes. Membranes are generally characterized in terms of their
productivity (Permeability coefficient) and efficiency (Permselectivity). The
gas permeability depends on two factors: one is a thermodynamic term,
solubility coefficient, which characterizing the number of gas molecules
sorbed into the polymer, and the other one is a kinetic term, diffusivity
coefficient, which characterizing the mobility of gas molecules as they diffuse
through the polymer[14].
Permeability, solubility, and diffusivity coefficient through the glassy

polymeric membrane could be considerably affected by different levels of
interaction between the polymer matrix and the penetrant species. Essentially,
the plasticization effect varies widely according to polymeric membrane and
19
penetrant molecule. The relatively strong interaction between penetrant
molecules and polymeric membranes show non-Fickian or abnormal diffusion
and also causes plasticization of the polymeric membrane. Glassy polymeric
membranes can convert from the glassy state to a rubbery state by
plasticization even at room temperature, indicative of glass transition pressure
at constant room temperature [20].
2.1.1.1 Permeability Coefficient

Fundamentally, in gas separation, permeability (Pi) is defined as flux of a
penetrant (Ni) through a membrane with a film thickness (l) and the pressure
drop across the membrane (pi). Permeabilities are commonly reported in
units of Barrer (1 Barrer = 10-10 cm3 (STP)-cm / cm2-s-cmHg), defined as
[14,21]:
Ni l N l
Pi = i (2-1)
p2 p1 p
At steady-state, we assume that the diffusion coefficient is independent of

penetrant concentration, and then from Ficks first law we find [22]:
Di dCi
Ni (2-2)
1 Mi dx
where Di is the local diffusion concentration of the gas in the membrane, Mi is

the mass fraction of gas in the membrane, Ci is the concentration gas A
dissolved in the membrane and x is the distance across the membrane. If we
then integrate over the membrane thickness and describe the penetrant
concentrations in the upstream and downstream of the membrane as x=0
(C=C2) and x=l (C=C1) respectively, we find:
C2 C1
Ni DAv. (2-3)
l
With this equation, the permeability of a gas can be expressed as:

C2 C1
Pi DAv. (2-4)
p2 p1
20
where DAv. is the average effective diffusion coefficient between C1 and C2.
When the upstream pressure and concentration (p2 and C2) are much greater
than their downstream analogs, Eq. (2-4) simplifies as follows:
C2
Pi DAv. S Av. DAv. (2-5)
p2
where the SAv. is the equilibrium solubility coefficient of a gas penetrant A in

the membrane, it is the ratio of the concentration of gas dissolved in the
membrane at equilibrium to the pressure of gas in the contiguous gas phase
[22].
2.1.1.2 Diffusivity coefficient

Diffusivity of gas molecules through the dense membranes is possibly
occurring through the transient gaps in the polymer chain matrix due to
thermal fluctuations. This diffusivity coefficient mainly depends on the size of
the penetrant, flexibility of the polymer chains and the free volume size and
distribution of the polymer [23]. The diffusivity of gases decreases as the
kinetic diameter of the gas molecule increases.
In rubbery polymer diffusivity coefficient involves the generation of a

sufficiently large gap in the polymer adjacent to the sorbed penetrant for the
penetrant to move into, with the subsequent collapse of the sorbed cage that
was previously occupied by the penetrant. However, diffusion in glassy
polymer is different from rubbery polymers because of the difference in the
characteristic scales of the micro-motions that are much less extensive than
rubbery polymers and are believed to be torsional oscillations. Some
differences arise because of the presence of trapped excess free volume in
glassy materials.
It has been noted that diffusion coefficients can actually be a strong function
of local penetrant concentration [24,25], so DAv. in Eqs. (2-3) (2-5) is
actually an average diffusion coefficient in the membrane and we can simply
define this average as:
21
C2
D dC
i i
DAv.
C1
(2-6)
C2 C1
2.1.1.3 Solubility Coefficient

Gas solubility in polymeric membranes is determined by condensability of the
gas molecule, the interaction between the gas molecule and the polymer
material and free volume of the membrane. The condensability of the gas
molecule is closely related to the critical temperature of the gas species. The
gas has a high critical temperature tends to be condensed, easier than the one
has a low critical temperature [26]. Sorption in rubbery polymers is similar to
the dissolution of gases in liquids. The gas sorption, C, can be described using
Henrys law as follows [27].
C kD p (2-7)
where kD is the Henrys law constant indicating gas dissolution into the
equilibrium-defined polymer matrix (cm3 (STP) / cm3-atm), p is the pressure
of the feed gas in contact with the polymer (atm). However, the solubility of
penetrant molecules in glassy polymeric membranes is denoted by different
sorption mechanisms into the polymer matrix and the micro - void region. The
main mechanism of sorption into the matrix component is Henry sorption
(CD), and Langmuir sorption (CH) governs in the micro-void region.
Therefore, the dual model theory of gas sorption is described by the sum of
these two contributions and total sorption (C) is therefore used to describe the
uptake of gases into both of these glassy polymer environments. In terms of
gas pressure, the model is written as:
CH bp
C CD C H k D p (2-8)
1 bp
where C is the concentration of the gas sorbed in the polymer with a unit of
gas volume (cm3) per polymer volume (cm3), CD and CH are the
concentrations of the gas penetrant sorbed at the Henry sites and Langmuir
sites, respectively, C'H is the Langmuir capacity of the glassy polymer that is
related to the un-relaxed volume, which is a measure of the departure from
22
equilibrium in the glassy state, and b is the Langmuir affinity parameter
describing the affinity of the gas to a Langmuir site. The Langmuir capacity
can be viewed as a measurement of the excess free volume of a polymer [28-
32].
The solubility coefficient (S) of the gas in a glassy polymeric membrane is

described as the ratio of equilibrium gas concentration of the applied gas
pressure, as shown in Eq. (2-9):
C C b
S kD H (2-9)
p 1 bp
Gas sorption may display an unusual isotherm if a gas is polar in nature and
highly condensable. For example, some glassy polymers can be plasticized to
the rubbery state by condensed CO2 gas, and this phenomenon depends on the
interaction between glassy polymers and CO2 gas [20]. Since one can conduct
the gas sorption isotherm experimentally, the three dual-mode sorption
parameters (kD, C'H, and b) can be obtained by curve fitting using a nonlinear
least squares method [32-36].
In the dual-mode sorption theory, the Langmuir mode species were originally
considered not to be mobile at all. In contrast, the partial immobilization
model is based on the independent dual diffusion of the Henry and Langmuir
modes, but assumes that species in the Langmuir mode can mobilize relatively
in glassy polymeric membranes. Since the concentration C of gas sorption is
assumed to obey the dual-mode sorption model as written by Eq. (2-8) and
permeation flux (permeation amount per unit time) J through the glassy
polymeric membrane is given by Eq. (2-10) and Eq. (2-12) the following
equations can be derived for the average D and the average P as follows
[20,28,29,33].
C C CH CH y CD
J D(C ) = DD D DH = DD 1
C y y
F
y y y D
FK CD
= DD 1 2
(2-10)
(1 bp) y
23
C CD CH CD bCH p CD K p
= = = kD
y y y y (1 bp) y
2
y (1 bp) 2
y
CD K CD K CD
= = 1+ 2
(2-11)
y (1 bp) y (1 bp) y
2
FK
1
C FK CD (1 bp) 2 C
J D(C ) = DD 1 2
= D (2-12)
y y
D
(1 bp ) y K
1 (1 bp) 2

FK
C 1 (1 bp) 2 K
C 0 DD K
k D 1
2
dp
0 D C dC
(1 bp )
1 (1 bp) 2
D

C C
K
0 dC 0 kD 1 (1 bp)2 dp
FK FK
DD k D 1 p 1
(1 bp) = D (1 bp) (2-13)
K
D
1 K
k D 1 p (1 bp)
(1 bp)
C K FK
P SD D k D 1 D k D DD 1
p 1 bp (1 bp)
CH bDH
=k D DD (2-14)
1 bp
where C is the concentration of the penetrant; y is the membrane thickness;

D(C) is the concentration dependent diffusion coefficient; DD is the diffusion
coefficient under the Henry mode; DH is the diffusion coefficient under the
Langmuir mode; F is the ratio of the two diffusion coefficients (DH/DD); K is

equal to CH b kD ; D is the average diffusion coefficient, and P is the average
permeability coefficient.
The Eq. (2-14) is the partial immobilization model [37-39]. This equation is
reduced to a fully Henry mode dominant process if F is equal to zero. In

addition, the permeability coefficient, P , decreases slowly with an increase in
system pressure, and then reaches to kD DD irrespective of the F value because
of the saturation of Langmuir sites. The increased permeability coefficient at
24
low pressure is due to a contribution of the diffusion attributed to the
Langmuir mode species. The permeability coefficient is improved, notably by
the increase in the value of F. In this model, the diffusivity coefficients of the
Henry mode (DD) and Langmuir mode (DH) can be readily calculated from the

slope and intercept of the experimental plot of permeability ( P ) versus
(1+bp)-1 when a straight line is obtained in the plot. As Langmuir sorption is
known to be chemical sorption, generally the Langmuir-sorbed molecule is
considered to be thoroughly settled into a sorption site. However, such an
immobile mechanism is not applied to glassy polymeric membranes, but rather
the concept that the Langmuir sorbed molecule is slightly mobile is accepted
for the gas permeation of such systems. Therefore, the permeation model of
glassy polymeric membranes is called the partial immobilization model [20].
2.1.1.4 Permselectivity
The permselectivity of a membrane, describes the efficiency of a membrane
material in separating one component over the other component in the gas
mixtures. For a binary gas mixture permeating across a membrane, the
permselectivity can be defined as the ratio of more permeate gas (i) to less
permeate gas (j)
Yi Y j
i / j = (2-15)
Xi X j
where Yi, Yj, and Xi, Xj refer to mole fractions of the components in the
permeate stream and feed stream, respectively. Under conditions of negligible
permeate pressure; the ideal separation factor i/j can be expressed as the ratio
of the permeability of the gas penetrants i and j as follows:
Pi Si Di
i/ j = = (2-16)
Pj S j D j
The ideal permselectivity can be expressed as the product of solubility

selectivity (Si/Sj) and diffusivity selectivity. Therefore, both solubility and
diffusivity selectivity should complement each other in order to achieve high
selectivity of the membrane. Generally, the diffusivity selectivity dominates
the overall gas separation in glassy polymeric membranes, where the gas
25
separation mainly depends on the size discriminating ability of the polymer
material [24]. However, the solubility selectivity dominates the performance
of the membrane if rubbery material is used.
2.1.1.5 Plasticization Behavior

Although the glassy materials exhibit attractive gas transport ability, the
separation performance of polyimides generally degrades when separating
highly soluble gases such as CO2 or hydrocarbons at relatively high pressures.
This phenomenon is referred as plasticization [10-12, 40-42]. In the
occurrence of plasticization, the penetrant species such as CO2 swell up the
polymer matrix, facilitate segmental motions, and accelerates diffusion
processes of all gas species [43-45]. This enhancement in chain mobility
results in the loss of size and shape discrimination, and leads to increased
diffusivity for all penetrants in the membrane. The slowest penetrant is
impacted the greatest, and thus, a drop in gas pair selectivity is observed
concurrent with an increase in permeability. Therefore, the ideal selectivity
measured by means of pure gas tests can no longer be used to estimate the
mixed gas membrane performance [45].
As previously discussed, permeability asymptotically decreases with

increasing pressure in glassy polymer membranes due to Langmuir saturation.
However, when a membrane plasticizes, a pronounced increase in
permeability can be observed. The pressure at which permeability begins to
increase is defined as the plasticization pressure. Hence, the structure of glassy
polymers needs to be further stabilized to improve the membrane stability.
Cross-linking has been recommended as a way to stabilize the membrane

against plasticization [12, 46]. The networked matrix in these materials,
exhibits greater stability at higher feed pressures of high-swelling penetrants.
This concept is of fundamental importance to the current work and will be
explored further in Section 2.1.2.
26
2.1.2 Cross-linkable Polymeric Membranes for Gas Separation
As briefly mentioned above, an effective method to reduce plasticization-
induced of membrane performance is to use a cross-linkable polymeric
membrane. CO2 and propylene are important gasses for industrial separations,
and given its tendency to swell polymers, researchers have been investigating
cross-linking solutions for several decades. The extent of cross-linking and
distribution of the resultant network can profoundly affect the physical
properties of polymeric membranes. Generally, it is believed that cross-linking
can be used to: 1) increase the membrane stability in the presence of
aggressive feed components; 2) suppress CO2-induced plasticization; and 3)
achieve better gas separation properties. Owing to the well-known tendency of

CO2, propane and propylene to plasticize membranes under industrially
relevant conditions, this technique has garnered much interest in the area of
natural gas purification and olefin/paraffin separation. Cross-linking
modifications of polyimides can be induced by several methods, such as
ultraviolet (UV) and ion beam irradiations, chemical modification and thermal
treatment.
Kita et al. [47] studied the effects of photo cross-linking on the permeability
and selectivity of BTDA-containing polyimides. They reported that
considerable improvement in gas pair selectivity was achieved by increasing
UV-irradiation time, but gas permeability was significantly reduced due to the
densification of membrane structure and reduction of polymer chain mobility.
Won et al. [48] and Hu et al. [49] investigated the effects of ion beam on
surface modification of polyimide membranes. They concluded that ion beam
irradiation could provide a useful mean for improving the selectivity for gas
separation membranes. Despite the potential use of ion beam irradiation in
enhancing gas separation pair permselectivity, the high cost and difficulty in
implementing uniform irradiation on large sheets of flat membranes and
hollow fibers limit the industrial application of this technique.
Koros and his co-workers synthesized polyimide membranes containing

carboxylic acid, which can be cross-linked by either a trans-esterification
27
reaction or reaction with a metal ion [50-53]. Bickel and Koros [51] studied
the chemical cross-linking between the free carboxylic acid group of 6FDA-
mPD/DABA and ethylene glycol. They found that the CO2/CH4 selectivity
increased with the increment in the degree of cross-linking. Interestingly, the
CO2 permeability did not significantly decrease because the ethylene glycol
group not only created more free volume, but also reacted with the carboxylic
acid group and inhibited the hydrogen bonding. Wind et al. [52] investigated a
series of cross-linkable polyimides and found an increase in permeability;
selectivity in several cases was enhanced. They also found that the amount of
permeability enhancement depended on the chemical structure of the diol
cross-linking agents and the specific annealing temperatures. It was shown
that unlike other cross-linking modifications which typically decreased the
permeability of the resultant polymeric membranes, cross-linking of
carboxylic moieties added free volume to the polymer, allowing for higher
CO2 permeabilities. It should be noted that trans-esterification occurs only at
o
elevated temperatures (above 200 C). The increase in the free volume may be
due to the coupling of thermal annealing and the removal of pendant diol
groups during the trans-esterification reaction.
The aging phenomenon of these cross-linked membranes (thermally treated at

o
140 C) has been investigated by Kim et al. [53]. It was found that the
polyimide membranes suffered more from physical aging after cross-linking
than before cross-linking. This problem can be rectified by performing the
cross-linking reaction at elevated temperatures. Nevertheless, the application
of high temperature treatments to asymmetric membranes alters the membrane
morphology and causes densification of the membrane structure. It should be
noted that cross-linking of the bulk polyimide matrix reduces the polymer
processability during membrane fabrication and increases the brittleness of the
resultant membranes. Therefore, the preferred chemical cross-linking method
should have the following characteristics: (1) it can be performed at room
temperature, and (2) it functionalizes only the thin selective layer of
asymmetric membrane which reduces substructure resistance and brittleness of
the support layer.
28
Another chemical modification method has been proposed to overcome the
difficulties of post-treatment processes. Chung and his co-workers contributed
a number of works related to the use of diamines and multi-amines to cross-
link polyimides [12, 41, 46, 54-56]. Liu et al. and Cao et al. used p-
xylenediamine as a cross-linking agent to modify different polyimide hollow
fiber membranes at ambient temperature [54, 55]. It is proven that the
chemically modified membrane does not plasticize easily since the cross-
linked structure prevents the material from swelling in the presence of
plasticizing agents. Generally, increasing the immersion time increases the
degree of cross-linking, which significantly reduces polymer swelling and
chain mobility. Hence, an enhancement in the CO2/CH4 permselectivity may
be achieved. Compared to the ethylene glycol approach, the diamine and

multi-amine approaches are more advantageous, especially for the
modification of hollow fibers because of easy operation and localizing
chemical modification at the outer selective skin without damaging the entire
membrane substructure.
Shao et al [46] employed different generations of diaminobutane (DAB)

dendrimers as cross-linking agent for 6FDA-Durene polyimide membranes to
induce chemical cross-linking reactions at room temperature. They reported
that membrane selectivity increased, but the permeability decreased. Xiao et
al. [41] modified 6FDA based polyimide by using different generation of
polyamidoamine (PAMAM) dendrimers. In their study, they concluded that
the modification time and the type of dendrimer are the determinant factors for
the ultimate performance of the membrane. The ideal selectivity of He/N2
increased to about 200% as compared to the original polyimide film and the
CO2/CH4 selectivity also predicted to surpass the Robson Upper Bound via
this modification. Compared with the simulated results, the morphology and
conformation of the grafted PAMAM dendrimers on the polymer surfaces are
disk-shaped molecular clusters which were further proven by AFM. In fact,
immersion time and different dendrimer generation are two important factors
in determining the characteristics of modified polyimide films. A longer
29
immersion time yields higher dendrimer loading and equilibrium may be
reached after modifying for at least 24 hours. Due to their big molecular size,
the cross-linking modification is believed to occur mainly on the membrane
surface. The modified polyimide films also exhibited better gas separation
performance, especially for CO2/CH4 because of the affinity between
PAMAM dendrimer and CO2. In another study by Xiao et al. [56] 6FDA-
Durene polyimide containing internal acetylene units was thermally cross-

linked. They reported that increasing amounts of cross-linking sites led to a
decrease in chain mobility and an increment in gas selectivity. However, gas
permeability decreased after thermal cross-linking and went lower with
increased molar ratio of cross-linking sites.
Although the chemical modification of polyimide membranes enhances

CO2/CH4 permselectivity, the long term operational stability of the membranes
remains unclear. One drawback of this approach is the reversible reaction

between the amino-compounds and the imide groups at high temperatures.
Nevertheless, the modified membranes can be employed for CO2/CH4
separation at relatively mild operating temperatures. In addition, since the

amidation reaction is carried out at room temperature, the reaction may be
incomplete, which result in time-dependent gas separation performance of the
resultant membranes.
Koros and his co-workers [57] examined thermal treatment of the 6FDA-
DAM: DABA (3:2) co-polyimide membranes. In this work, the thermal cross-
linking of 6FDA based co-polyimides at temperatures much below the glass
transition temperature (~387 C by DSC) was demonstrated. In their study,
they reported that annealing temperature and time influenced the degree of
cross-linking of the membranes, and resulted in a slight difference in
selectivity for membranes under various cross-linking conditions. They found
that sub-Tg thermal cross-linking of this 6FDA based co-polyimide membrane
could be carried out completely even at a temperature as low as 330 C.
Interestingly, the selectivity of the cross-linked membrane maintained even
30
under very aggressive CO2 operating conditions that are not possible without
cross-linking.
It is obvious from this summary that different cross-linking mechanisms

produce different effects on a given membrane system. Choosing the method
appropriate for ones research can be difficult, but a detailed review of the
literature, along with good knowledge of the application of the cross-linked
membrane, will narrow the cross-linking method and polymer selection
significantly. The first purpose of this study was to investigate the effect of
thermal cross-linking of modified 6FDA-based co-polyimide membranes on
CO2/CH4 and C3H6/C3H8 separation for both flat sheet and hollow fiber
membranes.
2.2 Mixed Matrix Membranes for Gas Separation

Previous section has focused on investigating polymeric membranes for the
gas separation application. According to the Robeson tradeoff between the
permeability (productivity) and selectivity (efficiency) of typical polymeric
membranes [58], new types of materials appeared to be needed to maximize
both membrane permeability and selectivity.
To overcome material bottlenecks, mixed matrix membranes (MMMs)

consisting of easy processing polymers and non-polymeric materials were
proposed. Early works focused on the incorporation of traditional materials
such as zeolites [59-63], silica and its derivatives [64-68], carbon [69, 70] and
MgO [13, 71] while the recent works on the use of POSS [72], carbon aerogel
[73] and Zeolitic imidazolate frameworks [74-79] to surpass the
permeability/selectivity trade-off line. However, challenging issues have been
encountered [61, 78]. One of them is the poor compatibility between the rigid
particles and the matrix polymers. As a result, poor particle adhesion and
phase separation are critical issues which significantly limit MMMs
applications. To enhance MMMs compatibility and permeability, Zeolitic
imidazolate frameworks (ZIFs) have received global attention recently. ZIFs
are a sub-class of metal organic frameworks (MOFs) with superior thermal
and chemical stability. Comparing to other molecular sieves, the ZIF family is
31
inherently more compatible with glassy polymers, especially its nano-particle
size and organic components [79]. The structure of ZIFs comprises a transition
metal (e.g. Zn, Co) cations bridged by anionic imidazolate linkers. The pore
size of ZIFs can be varied by changing the anionic imidazolate linker [79-82].
Furthermore, the imidazolate linkers present in metal organic frameworks
make ZIFs more hydrophobic than aluminosilicate zeolite, which may imply
better interfacial interactions between the sieve and the polymer matrix. ZIF-8
(Zn (2-methylimidazole) 2) is one of the most extensively studied ZIFs [74, 77,
83-90]. It has a sodalite structure consisting of Zn (II) metal centers and 2-
methyl imidazole. Its six-ring cages possess a pore cavity of 11.6 and a
theoretical pore aperture of 3.4 . The small pore size of 3.4 offers
molecular sieving characteristics that would enhance the separation of small
molecules from larger molecules (e.g., CO2 from CH4) while the larger
internal cavities of 11.6 allow rapid diffusion, even in the presence of larger
molecules which might otherwise create blockages and inhibit permeation
[79,80]. The framework of ZIF-8 has been found flexible to enable the
adsorption of molecules with sizes larger than 3.4 A [88].
Ordoez et al. [75] studied the Matrimid/ZIF-8 MMMs for gas separation.
They found that the separation performance was significantly enhanced with
an increase in ZIF-8 loading up to 40 wt%. Song et al. [74] employed ZIF-
8/Matrimid MMMs for CO2/CH4 separation by incorporating as-synthesized
ZIF-8 (size ~60 nm and specific surface area ~13001600 m2 g-1) into
Matrimid via solution mixing. The permeability increased by three times with
a 30 wt% ZIF-8 loading while the selectivity showed negligible losses. Li et
al. [83], Lai and his coworkers [91, 92] reported that ZIF-8 has very high
selectivity for C2/C3 and propylene/propane separation; hence MMMs
comprising ZIF-8 may be promising for olefin and paraffin separation.
Recently, Zhang et al. [93] observed that the permeability and selectivity of
ZIF-8/6FDA-DAM MMMs improved 250% and 150%, respectively, for
propylene/propane separation with a 48 wt% ZIF-8 loading.
While the upper-bound curve has been surpassed using mixed matrix
membranes in some cases [54, 94, 95], there are still many aspects of this new
32
technology that are not well understood. Controlling matrix phase swelling is a
paramount requirement for heavily interacting feeds containing strongly
sorbing components, such as CO2. The cross-linkable mixed matrix membrane
concept represents a novel technology that promises both increased efficiency
and plasticization resistance for natural gas purification and olefin/paraffin
separation in comparison to traditional polymer membranes. Recently, Ward
and Koros [96] embedded a surface modified zeolite (SSZ-13) molecular sieve
into a cross-linkable polyimide for natural gas purification. The surface
modified SSZ-13 not only provided high selectivity and permeability but also
exhibited surface compatibility with the cross-linkable polyimide. As a
consequence, their thermally cross-linked dense-flat MMMs showed
enhancements in both permeability and selectivity. However, the size of their
particles was around 500 nm that was too big and may effect on the gas
separation performance of the membrane. The second purpose of this study
was to investigate the effect of ZIF-8 nano-particle (80 nm) loading in thermal
cross-linkable 6FDA-based co-polyimide mixed matrix membranes on
CO2/CH4 and C3H6/C3H8 separation.
33
2.3 References
removal of acid gases from natural gas, J. Membr. Sci. 140 (1998) 27-
49.
[3] P.M. Budd, N.B. Mckeown, Highly permeable polymers for gas
separation membranes, Polym. Chem. 1 (2010) 63-68.
separation using polymeric membranes, J. Membr. Sci. 211 (2003)
299-309.
propylene/propane separation by carbonaceous membrane derived
from poly (aryl ether ketone)/2,6-bis(4-azidobenzylidene)-4-methyl-
[6] K. Tanaka, A. Taguchi, J. Hao, H. Kita, K. Okamoto, Permeation and
separation properties of polyimide membranes to olefins and paraffins,
J. Membr. Sci. 121 (1996) 197-207.
[7] S. Sridhar, A.A. Khan, Simulation studies for the separation of
propylene and propane by ethylcellulose membrane, J. Membr. Sci.
159 (1999) 209-219.
212 (2003) 13-27.
[10] M. Das, W.J. Koros, Performance of 6FDA-6FpDA polyimide for
propylene/propane separation, J. Membr. Sci. 365 (2010) 399-408.
34
[13] S.S. Hosseini, Y. Li, T.S. Chung, Y. Liu, Enhanced gas separation
performance of nanocomposite membranes using MgO nanoparticles,
J. Membr. Sci. 302 (2007) 207-217.
[14] M. Mulder, Basic principles of membrane technology, Dordrecht:
Kluwer Academic Publishers, 1996.
[15] R.M. Barrer, J.A. Barrie, J. Slater, Sorption and diffusion in ethyl
cellulose. Part III. Comparison between ethyl cellulose and rubber, J.
Polym. Sci., 27 (1958) 177-186.
[16] D.R. Paul, W.J. Koros, Effect of partially immobilizing sorption on
permeability and the diffusion time lag, J. Poly. Sci. Poly. Phys. Ed. 14
(1976) 675-685.
[17] Y. Mi, S.A. Stern, S. Trohalaki, Dependence of the gas permeability of
some polyimide isomers on their intrasegmental mobility, J. Membr.
Sci. 77 (1993) 41-48.
[18] Y. Hirayama, T. Yoshinaga, Y. Kusuki, K. Ninomiya, T. Sakakibara,
T. Tamari, Relation of gas permeability with structure of aromatic
polyimides I, J. Membr. Sci. 111 (1996) 169-182.
[19] M.L. Chua, Y.C. Xiao, T.S. Chung, Effects of thermally labile
saccharide units on the gas separation performance of highly
permeable polyimide membranes, J. Membr. Sci. 415-416 (2012) 375-
382.
[20] Y. Tsujita, Gas sorption and permeation of glassy polymers with
microvoids, Prog. Polym. Sci. 28 (2003) 1377-1401.
[21] S.A. Stern, The barrer permeability unit, J. Poly. Sci. Poly. Phys. Ed.
6 (1968) 1933-1934.
[22] K. Ghosal, B.D. Freeman, Gas separation using polymer membranes:
An overview, Polym. Advan. Technol. 5 (1994) 673-697.
[23] A. Singh-Ghosal, W.J. Koros, Energetic and entropic contributions to
mobility selectivity in glassy polymers for gas separation membranes,
Ind. Eng. Chem. Res. 38 (1999) 3647-3654.
35
[24] W.J. Koros, G.K. Fleming, S.M. Jordan, T.H. Kim, H.H. Hoehn,
Polymeric membrane materials for solution-diffusion based permeation
separations, Prog. Polym. Sci. 13 (1988) 339-401.
[25] W.J. Koros, Model for sorption of mixed gases in glassy polymer, J.
Poly. Sci. Poly. Phys. Ed. 18 (1980) 981-992.
[26] H. Lin, B.D. Freeman, Material selection guidelines for membranes
that remove CO2 from gas mixtures. J. Mol. Struct. 739 (2005) 57-74.
[27] H. Czichos, T. Saito, L. Smith, Springer handbook of materials
measurement methods (2006) Vol 978.
[28] R. Wang, C. Cao, T.S. Chung, A critical review on diffusivity and the
characterization of diffusivity of 6FDA-6FpDA polyimide membranes
for gas separation, J. Membr. Sci. 198 (2002) 259-271.
[29] T.A. Barbari, W.J. Koros, D.R. Paul, Gas sorption in polymers based
on bisphenol-A, J. Polym. Sci., Polym. Phys. Ed. 26 (1988) 729-744.
[30] D.R. Paul, Gas sorption and transport in glassy polymers, Ber.
Bunsenges. Phys. Chem. 83 (1979) 294-302.
[31] W.J. Koros, D.R. Paul, Design considerations for measurement of gas
sorption in polymers by pressure decay, J. Polym. Sci. Polym. Phys.
Ed. 14 (1976) 1903-1907.
[32] V.I. Bondar, Y. Kamiya, Y.P. Yampolskii, On pressure dependence of
the parameters of the dual-mode sorption model, J. Polym. Sci., Polym.
Phys. Ed. 34 (1996) 369-378.
[33] W.J. Koros, Sorption and transport of gases in glassy polymers, Ph.D.
dissertation, University of Texas, Austin, TX, 1977.
[34] S. Kanehashi, K. Nagai, Analysis of dual-mode model parameters for
gas sorption in glassy polymers, J. Membr. Sci. 253 (2005) 117138.
[35] D.R. Paul, W.J. Koros, Effect of partially immobilizing sorption on
permeability and the diffusion time lag, J. Polym. Sci., Polym. Phys.
Ed. 14 (1976) 675-685.
[36] K. Toi, G. Morel, D.R. Paul, Gas sorption and transport in
poly(phenylene oxide) and comparisons with other glassy polymers, J.
Appl. Polym. Sci. 27 (1982) 2997-3005.
36
[37] G.R. Ranade, R. Chandler, C.A. Plank, W.L.S. Laukhuf, Evidence of
dual mobility in sulfur dioxidepolyarylate system: an integral sorption
analysis, Polym. Eng. Sci. 25 (1985) 164-169.
[38] V. Stannett, The transport of gases in synthetic polymeric membranes:
an historic perspective, J. Membr. Sci. 3(1978) 97-115.
[39] Y. Maeda, D.R. Paul. Selective gas transport in miscible PPOPS
blends. Polymer 26 (1985) 2055-2063.
[40] S. Neyertz, D. Brown, S. Pandian, N.F.A. van der Vegt, Carbon
dioxide diffusion and plasticization in fluorinated polyimides,
[41] Y.C. Xiao, T.S. Chung, M.L. Chng, Surface characterization,
modification chemistry and separation performance of polyimide and
PAMAM dendrimer composites, Langmuir 20 (2004) 8230-8238.
[43] W.J. Koros, M.W. Hellums, Gas separation membrane material
selection criteria: differences for weakly and strongly interacting feed
components, Fluid Phase Equilib., 53 (1989) 339-354.
[44] A. Bos, I.G.M. Punt, M. Wessling, H. Strathmann, Plasticization-
resistant glassy polyimide membranes for CO2/CO4 separations,
Separation and Purification Tech. 14 (1998) 2739
Sci. 103 (1995) 73-82.
[46] L. Shao, T.S. Chung, S.H. Goh, K.P. Pramoda, Transport properties of
cross-linked polyimide membranes induced by different generations of
diaminobutane (DAB) dendrimers, J. Membr. Sci. 238 (2004) 153
163.
[47] H. Kita, T. Inada, K. Tanaka, K. Okamoto, Effect of photocrosslinking
on permeability and permoselectivity of gases through benzophenone-
containing polyimide, J. Membr. Sci. 87 (1994) 139-147.
37
[48] J. Won, M.H. Kim, Y.S. Kang, H.C. Park, U.Y. Kim, S.C. Choi,
Surface modification of polyimide and polysulfone membranes by ion
beam for gas separation, J. Appl. Polym. Sci. 75 (2000) 15541560.
[49] L. Hu, X. Xu, M.R. Coleman, Impact of H+ ion beam irradiation on
Matrimid. II. Evolution in gas transport properties, J. Appl. Polym. Sci.
103 (2007) 16701680.
[50] A.M.W. Hillock, W.J. Koros, Cross-linkable polyimide membrane for
natural gas purification and carbon dioxide plasticization reduction,
[51] C.S. Bickel, W.J. Koros, Improvement of CO2/CH4 separation
characteristics of polyimides by chemical crosslinking, J. Membr. Sci.
155, (1999) 145-154.
[52] J.D. Wind, C. Staudt-Bickel, D.R. Paul, W.J. Koros, Solid-state
covalent cross-linking of polyimide membranes for carbon dioxide
plasticization reduction, Macromolecules 36 (2003) 1882-1888.
[53] J.H. Kim, W.J. Koros, D.R. Paul, Effects of CO2 exposure and physical
aging on the gas permeability of thin 6FDA-based polyimide
membranes Part 2. With crosslinking. J. Membr. Sci. 282 (2006) 32
43.
[54] Y. Liu, R. Wang, T.S. Chung, Chemical cross-linking modification of
polyimide membranes for gas separation. J. Membr. Sci. 189 (2001)
231239.
[55] C. Cao, T.S. Chung, Y. Liu, R. Wang, K.R. Pramoda, Chemical cross-
linking modification of 6FDA-2, 6-DAT hollow fibermembranes for
natural gas separation, J. Membr. Sci. 216 (2003) 257268.
[56] Y.C. Xiao, L. Shao, T.S. Chung, D.A. Schiraldi, The effects of thermal
treatments and dendrimer chemical structures on the properties of
highly surface cross-linking polyimide films, Ind. Eng. Chem. Res. 44
(2005) 305967.
[57] W. Qiu, C.C. Chen, L. Xu, L. Cui, D.R. Paul, W.J. Koros, Sub-Tg
cross-linking of a polyimide membrane for enhanced CO2
plasticization resistance for natural gas separation, Macromolecules 44
(2011) 6046-6056.
38
400.
[59] S. Kulprathipanja, R.W. Neuzil, N.N. Li, Gas separation by means of
mixed matrix membranes, US Patent 4,740,219 (1988).
[60] M.D. Jia, K.V. Peinemann, R.D. Behling, Preparation and
characterization of thin-film zeolite-PDMS composite membranes, J.
Membr. Sci. 73 (1992) 119-128.
[61] R. Mahajan, R. Burns, M. Schaeffer, W.J. Koros, Challenges in
forming successful mixed matrix membranes with rigid polymeric
materials, J. Appl. Polym. Sci. 86 (2002) 881-890.
[62] L.Y. Jiang, T.S. Chung, S. Kulprathipanja, Fabrication of mixed matrix
hollow fibers with intimate polymer-zeolite interface for gas
separation, AIChE J. 52 (2006) 2898 - 2908.
[63] Y. Li, T. S. Chung, Z. Huang, S. Kulprathipanja, Dual-layer
polyethersulfone (PES)/BTDA-TDI/MDI co-polyimide (P84) hollow
fiber membranes with a submicron PESzeolite beta mixed matrix
dense-selective layer for gas separation, J. Membr. Sci. 277 (2006) 28
37.
[64] H. Lin, E.V. Wagner, B.D. Freeman, L.G. Toy, R.P. Gupta,
Plasticization-enhanced hydrogen purification using polymeric
membranes, Science. 311 (2006) 639-642.
[65] H.B. Park, J.K. Kim, S.Y. Nam, Y.M. Lee, Imide-siloxane block
copolymer/silica hybrid membranes: preparation, characterization and
gas separation properties, J. Membr. Sci. 220 (2003) 5973.
[66] D. Gomes, S.P. Nunes, K.-V. Peinemann, Membranes for gas
separation based on poly (1-trimethylsilyl-1-propyne)-silica
nanocomposites, J. Membr. Sci. 246 (2005) 1325.
[67] L. Shao, T.S. Chung, In-situ fabrication of cross-linked PEO/silica
reverse-selective membranes for hydrogen purification, Int. J.
Hydrogen Energy 34 (2009) 6492-6504.
[68] C.H. Lau, S.L. Liu, D.R. Paul, J.Z. Xia, Y.C. Jean , H.M. Chen, L.
Shao, T.S. Chung, Silica nanohybrid membranes with high CO2
affinity for green hydrogen purification, Adv. Energy Mater. 1 (2011)
634-642.
39
[69] D.Q. Vu, W.J. Koros, S.J. Miller, Mixed matrix membranes using
carbon molecular sieves I. Preparation and experimental results, J.
Membr. Sci. 211 (2003) 311334.
[70] M.G. Garcia, J. Marchese, N.A. Ochoa, High activated carbon loading
mixed matrix membranes for gas separations, J. Mater. Sci. 47 (2012)
3064-3075.
[71] S. Matteucci, V.A. Kusuma, S.D. Kelman, B.D. Freeman, Gas
transport properties of MgO filled poly (1-trimethylsilyl-1-propyne)
nanocomposites, Polymer 49 (2008) 1659-1675.
[72] Y. Li, T.S. Chung, Molecular-level mixed matrix membranes
comprising Pebax and POSS for hydrogen purification via
preferential CO2 removal, Int. J. Hydrogen Energy 35 (2010) 10560
10568.
[73] Y. Zhang, I.H. Musselman, J.P. Ferraris, K.J. Balkus Jr., Gas
permeability properties of mixed-matrix Matrimid membranes
containing carbon aerogel: a material with both micropores and
mesopores, Ind. Eng. Chem. Res. 47 (2008) 2794-2802.
[74] Q. Song, S.K. Nataraj, M.V. Roussenova, J.C. Tan, D.J. Hughes, W. Li,
P. Bourgoin, M.A. Alam, A.K. Cheetham, S.A. Al-Muhtaseb, E.
Sivaniah, Zeolitic imidazolate framework (ZIF-8) based polymer
nanocomposite membranes for gas separation, Energy Environ. Sci. 5
(2012) 8359-8369.
[75] M.J.C. Ordoez, K.J. Balkus Jr., J.P. Ferraris, I.H. Musselman,
Molecular sieving realized with ZIF-8/Matrimid mixed-matrix
membranes, J. Membr. Sci. 361 (2010) 2837.
[76] T.X. Yang, Y.C. Xiao, T.S. Chung, Poly-/metal-benzimidazole nano-
composite membranes for hydrogen purification, Energy Environ. Sci.
4 (2011) 4171-4180.
[77] G. M. Shi, T. X. Yang, T.S. Chung, Sorption, swelling, and free
volume of polybenzimidazole (PBI) and PBI/zeolitic imidazolate
framework (ZIF-8) nano-composite membranes for pervaporation,
Polymer 54 (2013) 774-783.
[78] T.S. Chung, L.Y. Jiang, Y. Li, S. Kulprathipanja, Mixed matrix
membranes (MMMs) comprising organic polymers with dispersed
40
inorganic fillers for gas separation, Prog. Polym. Sci. 32 (2007) 483-
507.
[79] J. Caro, Are MOF membranes better in gas separation than those made
of zeolites, Curr. Opin. Chem. Eng. 1 (2011) 77-83.
[80] K.S. Park, Z. Ni, A.P. Cote, J.Y. Choi, R. Huang, F.J. Uribe-Romo,
H.K. Chae, M. O Keeffe, O.M. Yaghi, Exceptional chemical and
thermal stability of zeolitic imidazolate frameworks, PNAS 103 (2006)
1018610191.
[81] A. Phan, C.J. Doonan, F.J. Uribe-Romo, C.B. Knobler, M. OKeeffe,
O.M. Yaghi, Synthesis, structure, and carbon dioxide capture
properties of zeolitic imidazolate frameworks, Acc. Chem. Res. 43
(2010) 5867.
[82] W. Morris, C.J. Doonan, H. Furukawa, R. Banerjee, O.M. Yaghi,
Crystals as molecules: post synthesis covalent functionalization of
zeolitic imidazolate frameworks, J. Am. Chem. Soc. 130 (2008)
1262612627.
[83] K. Li, D.H. Olson, J. Seidel, T.J. Emge, H. Gong, H. Zeng, J. Li,
Zeolitic imidazolate frameworks for kinetic separation of propane and
propene, J. Am. Chem. Soc. 131 (2009) 1036810369.
[84] H. Bux, C. Chmelik, R. Krishna, J. Caro, Ethene/ethane separation by
the MOF membrane ZIF-8: molecular correlation of permeation,
adsorption, diffusion, J. Membr. Sci. 369 (2011) 284289.
[85] S.R. Venna, M.A. Carreon, Highly permeable zeolite imidazolate
framework-8 membranes for CO2/CH4 separation, J. Am. Chem. Soc.
132 (2010) 7678.
[86] H. Bux, F. Liang, Y. Li, J. Cravillon, M. Wiebcke, J. Caro, Zeolitic
imidazolate framework membrane with molecular sieving properties
by microwave-assisted solvothermal synthesis, J. Am. Chem. Soc. 131
(2009) 1600016001.
[87] D. Fairen-Jimenez, S.A. Moggach, M.T. Wharmby, P.A. Wright, S.
Parsons, T. Duren, Opening the gate: framework flexibility in ZIF-8
explored by experiments and simulations, J. Am. Chem. Soc. 133
(2011) 89008902.
41
[88] S.R. Venna, J.B. Jasinski, M.A. Carreon, Structural evolution of
zeolitic imidazolate framework-8, J. Am. Chem. Soc. 132 (2010)
1803018033.
[89] T.X. Yang, T.S. Chung, High performance ZIF-8/PBI nano-composite
membranes for high temperature hydrogen separation consisting of
carbon monoxide and water vapor, Int. J. Hydrogen Energy 38 (2013)
229-239.
[90] G.M. Shi, T.X. Yang, T.S. Chung, Polybenzimidazole (PBI)/zeolitic
imidazolate frameworks (ZIF-8) mixed matrix membranes for
pervaporation dehydration of alcohols, J. Membr. Sci. 415416 (2012)
577586.
[91] Y.C. Pan, Z.P. Lai, Sharp separation of C2/C3 hydrocarbon mixtures by
zeolitic imidazolate framework-8 (ZIF-8) membranes synthesized in
aqueous solutions Chem. Commun. 47 (2011) 1027510277.
[92] Y.C. Pan, T. Li, G. Lestari, Z.P. Lai, Effective separation of
propylene/propane binary mixtures by ZIF-8 membranes, J. Membr.
Sci. 390391 (2012) 9398.
[93] C. Zhang, Y. Dai, J.R. Johnson, O. Karvan, W.J. Koros, High
performance ZIF-8/6FDA-DAM mixed matrix membrane for
[94] P.S. Tin, T.S. Chung, Y. Liu, R. Wang, S.L. Liu, K.P. Pramoda,
Effects of cross-linking modification on gas separation permformance
of Matrimid membranes, J. Membr. Sci. 225 (2003) 77-90.
[95] J. Ren, R. Wang, T.S. Chung, D.F. Li, Y. Liu, The effects of chemical
modifications on morphology and performance of 6FDA-ODA/NDA
hollow fiber membranes for CO2/CH4 separation, J. Membr. Sci. 222
(2003) 133-147.
[96] J.K. Ward, W.J. Koros, Crosslinkable mixed matrix membranes with
surface modified molecular sieves for natural gas purification: I.
Preparation and experimental results, J. Membr. Sci. 377 (2011) 75-81.
42
Chapter 3: Materials and Methodology
In this chapter, the materials that were studied, the membrane fabrication
techniques, characterization techniques, and the equipment used to
characterize these materials will be detailed. The specific materials selected
for this work are described in Section 3.1. Membrane formation techniques are
described in section 3.2 and material characterization techniques and
equipment are outlined in Section 3.3.
3.1 Materials
3.1.1 Polymers
3.1.1.1 6FDA-Durene polyimide and 6FDA-Durene/DABA co-polyimide
The polymers selected for this work was a 6FDA-Durene polyimide and
6FDA-Durene/DABA cross-linkable co-polyimide. Figure 3.1 shows the
chemical structure of the monomers that was used in this study. 4,4'-
(hexafluoroisopropylidene) diphthalic anhydride (6FDA), supplied by
Clariant, and 3,5-diaminobenzoic acid (DABA) supplied by Aldrich
(Singapore) were purified by vacuum sublimation before usage. 2,3,5,6-
Tetramethyl-1,4-phenylenediamine (Durene diamine) supplied by Aldrich
(Singapore), was recrystallized two times from methanol. 1-methyl-2-
pyrrolidone (NMP) purchased from Merck Chemicals (Germany), was
distilled at 42 C under 1 mbar and was dried with molecular sieve before use.
Acetic anhydride was received from Aldrich (Singapore) and dried with
molecular sieve before usage. Other chemicals and solvents including
triethylamine, methanol, and p-toluenesulfonic acid were all reagent grades or
better from Aldrich (Singapore) and were used without further purification.
The 6FDA-Durene polyimide, 6FDA-Durene/DABA (9/1) co-polyimide, and

6FDA-Durene/DABA (7/3) co-polyimide were synthesized via a two-step
chemical imidization in an NMP solution. For the preparation of polyamic
acid, a designated molar ratio of Durene and DABA diamine was dissolved in
NMP under a nitrogen atmosphere and then an equimolar of solid 6FDA was
gradually added to the solution in a moisture free flask with a magnetic
stirring under a nitrogen atmosphere at room temperature. The reason of
43
adding solid 6FDA gradually is to suppress the reaction between water and
6FDA. Aromatic dianhydrides can react with water and other impurities in the
amide solvents, but have a slower reaction rate relative to their reacting with
diamines. Therefore, gradually adding the solid 6FDA reduces its availability
for competing reactions with water or other impurities, thus leads to a higher
conversion and molecular weight [1]. After reacted for 24 h, a high molecular
weight polyamic acid was formed. Then for the chemical imidization step, a
mixture of acetic anhydride and triethylamine at 4:1 molar ratio was slowly
added to the polyamic acid solution to perform imidization for 24 h under
nitrogen atmosphere. Finally, the co-polyimide was precipitated in methanol,
filtered, washed and dried at 120 C in vacuum for 24 h. The schemes of co-
polyimide synthesis and chemical structure of this co-polyimide are shown in
Figure 3.1.
3.1.1.2 6FDA-Durene/DABA (9/1) co-polyimide grafted with Cyclodextrin

In this work, three Cyclodextrins (CDs) were grafted on the 6FDA-
Durene/DABA (9/1) co-polyimide by esterification [2]. The three CDs,
namely, -, -, and -CD, formed by six to eight glucopyranoside units, were
all reagent grade or better from Aldrich (Singapore) and were used without
further purification. Their chemical structure and basic properties have been
presented in Figure 3.2 and Table 3.1 [3]. For this purpose, co-polyimide and
, , and -CD were first dried at 120 C in vacuum for 12 h. After that co-
polyimide was dissolved in NMP and then a large excess amount of CD (3-5
times more than stoichiometric balance to the carboxylic acid group) was
added to the NMP solution. The esterification was carried out by adding a
catalytic amount (5 mg per gram of polymer) of p-toluenesulfonic acid and
heating to 120 C under nitrogen atmosphere for 18 h. Finally, the resultant
co-polyimide grafted with , , or -CD (referred as PI-g--CD, PI-g--CD, or
PI-g--CD) was precipitated in methanol, washed to remove unreacted CD and
dried under 120 C in vacuum for 24 h. The chemical structure of co-
polyimide after grafting CD is shown in Figure 3.1.
44
Figure 3.1: The synthetic scheme and chemical structure of co-polyimide
Figure 3.2: Chemical structure of three kinds of Cyclodextrin [3]
45
Table 3.1: Properties of the three types of Cyclodextrin
CD Type No. of Mw Cavity Cavity Decomposition
Glucose (g/mol) [3] Diameter Height Temp. (C) a
unit () [3] () [3]
-CD 6 972 4.7-5.7 7.8 3211
-CD 7 1135 6.0-7.8 7.8 3281
-CD 8 1297 7.5-9.5 7.8 3221
a
Td (5%), at which the mass of the sample is 5.0% less than its mass measured at 50 C.
3.1.2 Zeolitic imidazolate frameworks (ZIFs) nanoparticle synthesis

The nanoparticle used in this work to make mixed matrix membranes was
zeolitic imidazolate framework-8 (ZIF-8). The dispersed ZIF-8 nanoparticles
were synthesized at room temperature. In summary, a solution of
Zn(NO3)26H2O (7.332 g, 24.67 mmol) in 500 ml methanol was rapidly
poured into a solution of Hmim (16.222 g, 197.6 mmol) in 500 ml methanol
under stirring with a magnetic bar. The resultant solution slowly turned turbid
and after 1 h the nano-crystals were separated from the milky dispersion by
centrifugation and washing with fresh methanol. After washing and the 2nd
centrifugation, the particles were re-dispersed in fresh NMP before use. The
yield of ZIF-8 was ~ 50% based on Zinc [4].
3.1.3 Gases
All pure gases and gas mixtures used throughout this research (i.e., CO2, CH4,
C3H6, and C3H8) were research grade (99.999%) and supplied by SOXAL
(Singapore).
3.2 Membrane formation

3.2.1 Dense flat sheet film formation
While asymmetric hollow fiber membranes are the industrially preferred
membrane geometry, dense film membranes are much simpler to prepare and
characterize when developing new technology. As such, dense film polymer
membranes were exclusively investigated over the course of this work. Dense
films were prepared using the solution casting method. Prior to use, the
original co-polyimide (PI) and PI-g-CDs were dried overnight at 120 C under
vacuum. A 2 % (w/w) polymer solution was prepared by dissolving the
original PI and PI-g-CDs together in NMP with a weight ratio of 1:1. The
46
polymer solution was magnetically stirred overnight and filtered using a
Whatmans filter (1 m) to remove dust and insoluble particles before it was
cast onto a silicon wafer covered with a steel ring. The polymer casting
solution was heated at 40 C for 12 h and vacuum heated at 70 C for 24 h,
respectively. Then, the dense films were formed and pealed off from the
silicon wafer and placed in a vacuum oven to be further vacuum heated at 200
C for 24 h in purpose of removing a trace amount of residual solvent. The
absence of residual solvent was confirmed by the TGA. Finally, films with a
thickness of 50 10 m were prepared for testing and characterization in the
following studies.
3.2.1.1 Thermal cross-linking of dense flat sheet membranes

The thermal treatment was performed using a CenturionTM Neytech Qex
vacuum furnace (Yucaipa, CA, USA). The films were heated under vacuum
from room temperature to specific temperatures such as 300, 350, 400, and
425 C. The heating rate was 10 C/min and the films were held for 120 min at
the final temperatures. After completing the thermal treatment process, the
membranes were cooled naturally in the vacuum furnace to room temperature
and used for permeation tests. All thermal treated membranes were flexible
and mechanically strong enough for gas permeation tests. For sample
classification, the treated samples at 200 and 300 C were named as
polyimide-grafted-, , or -CD-final treated temperature, for example PI-
g--CD-300 and from 350 to 425 C were named as partially pyrolyzed
membrane-, , or -CD-final treated temperature, for examplePPM--CD-
400.
3.2.2 Fabrication of dual layer hollow fiber membranes

In this study, dual-layer hollow fiber membranes consisting of 6FDA-Durene
as an inner layer and 6FDA-Durene/DABA (9/1) grafted with -Cyclodextrin
(PI-g--CD) as an outer layer were prepared. The chemical structures of these
polymers are shown in Figure 3.3.
47
a)
b)
Figure 3.3: Chemical structure of a) 6FDA-Durene and b) 6FDA-Durene /

DABA (9/1) grafted with -Cyclodextrin
3.2.2.1 Inner layer and outer layer dope Preparation

A binary PI-g--CD/NMP system and a 6FDA-Durene/NMP system were
selected for the outer layer and inner layer dopes, respectively. To determine
the critical concentrations of both dopes, solutions with different polymer
concentrations were prepared and their shear viscosity was obtained from a
cone and plate rheometer (Rheometric Scientific ARES) using a shear rate of
10s-1 at 25 C. Figure 3.4 shows the relationship between the shear viscosity
and polymer concentration for both dopes. Based on the critical
concentrations, the polymer compositions in the outer layer and inner layer
dopes were chosen as 24 wt% PI-g--CD and 21 wt% 6FDA-Durene in NMP,
respectively. The outer layer dope has a polymer concentration higher than the
critical concentration of about 22.5 wt% to ensure the minimization of defects,
while the inner layer has a polymer concentration slightly lower than the
critical concentration of about 21.5wt% to ensure a low substructure
resistance. For preparation of dope solutions, the respective polymer was
added gradually to the solvent under continuous agitation. The solutions were
stirred at 25 C for 24 h to ensure complete dissolution and left standing for
another 24 h for degassing before pouring into the respective ISCO pumps.
Because of the high dope viscosity, spinning was performed on the following
day in order for degassing to take place in the ISCO pumps.
48
Figure 3.4: Shear viscosity vs. polymer concentration for inner and outer
layer dopes at 25 C
3.2.2.2 Spinning conditions and solvent exchange

The PI-g--CD/6FDA-Durene dual-layer hollow fiber membranes were
fabricated by a dry-jet/wet spinning process via co-extrusion of polymeric
dopes by a triple-orifice spinneret. The experimental set-up and procedure
have been described in our previous reports [5-7]. All the spinning conditions
except for the take-up velocity and outer-layer dope flow rate were kept
constant. Both the outer-layer dope flow rate and take-up velocity were varied
to obtain optimal spinning conditions for the fabrication of dual-layer hollow
fiber membranes, while the bore fluid composition was NMP/water (i.e., 95/5
wt %) to ensure a fully porous inner layer. The maximum take-up velocity was
11.8 m/min because the spinning process became unstable at a higher take-up
velocity and filament broke up. During the spinning, the inner layer and bore
fluid flow rate were kept unchanged at 1.5 and 0.8 ml/min, respectively. The
spinning experiments were carried out at ambient temperature. Table 3.2
summarizes the spinning and conditions Figure 3.5 shows the schematics of
the dual-layer spinneret used in this study.
The as-spun fibers were immersed in tap water for three days to remove most
of the solvent from membranes. The fibers were then solvent exchanged with
the following procedure. (1) Immersed in fresh methanol three times, 30 min
49
each time, (2) washed in n-hexane three times, 30 min each time, and (3) air
dry at ambient temperature for 24 h [8].
Table 3.2: Spinning conditions for dual layer PI-g--CD/6FDA-Durene

hollow fiber membranes
Condition
Parameters
A B C D E
Take up velocity (m/min) 4.0 (Free fall) 7.4 11.8 7.4 7.4
Outer Layer Dope extrusion rate (ml/min) 0.5 0.5 0.5 0.3 0.2
Inner Layer Dope extrusion rate (ml/min) 1.5
Inner Layer Dope Composition 6FDA-Durene/NMP (24/76 wt %)
Outer Layer Dope Composition PI-g--CD/NMP (21/79 wt %)
Bore fluid NMP/DI water (95/5 wt %)
External Coagulant Tap water
External Coagulant Temperature (C) 25
Air Gap (cm) 2.5
Bore fluid flow rate (ml/min) 0.8
Figure 3.5: (a) Side view and (b) cross-sectional view of the dual-layer
hollow fiber spinneret [8]
3.2.2.3 Post treatment of the hollow fiber membranes

The precursor hollow fibers were subjected to thermal treatment and silicon
rubber coating modifications. Thermal treatment was carried out with a
controlled thermal treatment protocol as follows: (1) heating at 10 C/min to
the final temperature, (2) isothermal soaking at the final temperature for a
given period of time (1 h), (3) cooled naturally to room temperature. For the
coating of the hollow fibers, a 3 % (w/w) of silicone rubber in a n-hexane
solution was used for a duration of 3 min.
50
3.2.3 Preparation of dense flat sheet mixed matrix membrane
Dense mixed matrix membranes were prepared using the solution casting
method. Prior to use, all the polymers were dried overnight at 120 C under
vacuum and dissolved in NMP by stirring it for 24 h at room temperature. A 2
% (w/w) polymer solution was filtered using a Whatmans filter (1 m) to
remove dust and insoluble particles. The ZIF-8/NMP suspension was stirred
and sonicated alternatively for 30 min for three times each, and then added
into the polymer/NMP solution followed by thoroughly stirring. The solution
was cast onto a silicon wafer covered with a steel ring. The polymer casting
solution was heated at 40 C in a vacuum oven for 12 h and the temperature
was further raised to 70 C for 24 h. Then, the dense films were formed and
pealed off from the silicon wafer and placed into the vacuum oven to be
further vacuum heated at 200 C for 24 h to remove any residual solvents. The
resultant membranes have a thickness of 5010 m. The ZIF-8 loading in the
membrane was calculated based on Equation 1. The actual particle weight in
the membrane was calculated from the residue weight of zinc oxide according
to the stoichiometric relationship. The weight of zinc oxide was obtained by
combusting the membrane in a thermo gravimetric analysis (TGA) instrument
under air atmosphere.
Particle weight
Particle loading wt% = 100 (3-1)
Particle weight + Polymer weight
3.2.3.1 Thermal cross-linking of mixed matrix membranes

The thermal treatment was performed using a CenturionTM Neytech Qex
vacuum furnace (Yucaipa, CA, USA). The films were heated under vacuum
from room temperature to a specific temperature, such as 350 C or 400 C.
The heating rate was 10 C/min and the films were held for 120 min at the
final temperature. After completing the thermal treatment process, the
membranes were cooled naturally in the vacuum furnace to room temperature
and used for permeation tests. All thermal treated membranes were flexible
and mechanically strong enough for gas permeation tests.
51
3.3 Physicochemical characterization
3.3.1 Fourier transform infrared spectrometer (FT-IR)
Fourier Transform Infrared Spectroscopy (FT-IR) was applied to investigate
the evolution of chemical structure changes. The FTIR measurements were
performed using an attenuated total reflection mode (FTIR-ATR) with a
Perkin-Elmer Spectrum 2000 FTIR spectrometer (Cambridge, MA, USA).
3.3.2 Thermo gravimetric analyses (TGA)

A TGA 2050 Thermo gravimetric Analyzer (TA Instruments, New Castle,
Delaware, USA) was employed to characterize the weight loss of the
polyimide. The analysis was carried out with a heating ramp of 10 C/min
from room temperature to 800 C in a nitrogen atmosphere, using ~10 mg
sample.
3.3.3 Differential scanning calorimetry (DSC)

A differential scanning calorimeter DSC 822e (Mettler Toledo, Singapore) was
used to monitor the thermal properties of polymers and measure their glass
transition temperatures (Tg). For this purpose, testing samples were heated to
450 C at a heating rate of 30 C/min in nitrogen environments during the first
heating, quenched at a rate of 30 C/min, and reheated at 30 C/min. All Tg
were measured at the second heating cycle.
3.3.4 Density measurement

Density was measured according to the Archimedean principle using a Mettler
Toledo balance (Singapore) with a density kit. At least three density measures
were recorded for each membrane film. Then, the fractional free volume
(FFV) is calculated by the following equation [9]: FFV= (V-V0)/V, where V is
the observed specific volume calculated from the measured density and V0 is
the occupied volume calculated from the correlation, V0 = 1.3Vw where Vw is
the Van der Waals volume. Vw is calculated from the group contribution
method of Bondi [10].
52
3.3.5 Wide angle x-ray diffraction (WAXD)
A Brukel wide-angle X-ray diffractometer (Bruker D8 advanced
diffractometer) was employed to determine the d-space value that indicates the
average inter-segmental distance of polymer chains. The measurement was
completed with a step increment of 0.02 and Ni-filtered Cu-K radiation at a
wavelength = 1.54 was used. The d-space was determined based on the
Braggs law [9] as follows: n=2d sin, where d, , and are the dimension
spacing, diffraction angle, and X-ray wavelength, respectively, and n is an
integral number (1, 2, 3,. . .).
3.3.6 Gel permeation chromatography (GPC)

Gel permeation chromatography (GPC) was used to measure the molecular
weights of the co-polyimides. GPC measurements were carried out on an HP
1100 HPLC system equipped with the HP 1047A RI detector and the Agilent
79911GP-MXC columns. Tetrahydrofuran (THF) was used as the solvent,
with a controlled flow rate of 1.0 ml/min. The column was calibrated using a
standard polystyrene sample and testing sample was 0.005 wt%. The
molecular weights were estimated by comparing the retention times in the
column to those of standard polystyrene.
3.3.7 Field emission scanning electron microscopy (FESEM)

To observe the cross-sectional morphology of the membranes, the samples
were immersed and fractured in liquid nitrogen. The fractured samples were
coated using a JEOL JFC-1300 Auto Fine Coater fitted with Pt target (i.e., 20
mA and 40 s). The morphology of the membrane was observed using field
emission scanning electron microscopy (FESEM JEOL JSM-6700LV). The
detailed sample preparation procedures can be found elsewhere [5]. The
FESEM samples were also used to measure their surface or cross-section
elemental content by Oxford Inca energy dispersion of X-ray (EDX, Oxford
Instruments Inca, Edinburg, UK). The mapping and line-scan modes were
selected in the analysis.
53
3.3.8 Positron annihilation lifetime spectroscopy (PALS)
The assembly for bulk PALS tests is illustrated in the bottom right of Figure
22
3.6 where Na isotope sealed in Kapton films was used as the source of
positrons and sandwiched by membrane samples. The flat-sheet membrane
samples were cut into circle pieces with the dimension of 1 cm. The thickness
of the membrane on each side of the source is about 0.5 mm. Positron lifetime
was measured as the time difference between the 1.27 MeV -quanta (start
signal) and the 511 keV -quanta (stop signal) generated by the annihilation of
positronium.
Figure 3.6: Schematic diagram of the bulk PALS system and the inset
indicates the sample packing structure.
In this experiment all PALS spectra were recorded with a counting rate of
around 200 counts per second, and a total count of 106 was fixed to collect
each spectrum. Positron lifetimes and their intensities were analyzed by a
computer program PATFIT from PALS spectra. The o-Ps lifetime 3 (ns) was
used to calculate the mean free-volume radius R ( to nm) based on an
established semi-empirical correlation equation from a spherical-cavity model
[11]:
54
1
1 R 1 2 R
3 = 1 sin (3-2)
2 R R 2 R R
where 3 and R are expressed in the units of ns and , respectively, and R

was determined to be 1.656 . The intensity I3 is directly related to the amount
of free volume in the membrane since the quantity of o-Ps annihilation with
electrons in micro-cavities may correlate to the free volume.
3.4 Determination of gas transport properties

3.4.1 Pure gas sorption test
Gas sorption tests were conducted using a Cahn D200 microbalance sorption
cell at 35 C over a pressure range of 0-30 atm for CO2 and CH4 and 0-7 atm
for C3H6 and C3H8. Figure 3.7 depicts the schematic diagram of the pure gas
sorption testing apparatus for the membranes [12].
Sample Tube Reference Tube
.
Figure 3.7: Schematic diagram of the microbalance sorption cell
For each run, films with a thickness of around 50 10 m, and total mass of
approximately 70-80 mg were cut into small pieces and placed on the sample
pan. The system was evacuated for 24 h prior to testing. The gas at a specific
pressure was fed into the system. The mass of gas sorbed by the membranes
was recorded at equilibrium state. Subsequent sorption experiments were
employed by further increasing the gas pressure. The equilibrium sorption
55
value obtained was corrected by the buoyancy force as shown in Eq. (3-3) and
(3-4), where WP and WSP are the balance readings with and without polymer
sample, respectively.
Gas Dissolved= WP - WSP+ Buoyancy (3-3)
Buoyancy= Volume of Polymer (cm3) Gas Density (g/cm3) g (3-4)
3.4.2 Pure gas permeation test

3.4.2.1 Dense flat sheet membranes
The pure gas permeation properties of the membranes were evaluated by a
variable-pressure constant-volume method at 35 C and 10 atm for O2, N2,
CH4, and CO2 and 3.5 atm for C3H8, and C3H6, by sequence. Figure 3.8
reveals the general setup of the permeation cell. The feed side of the cell
comprised a controlling valve, a pressure gauge and a gas reservoir with a
volume of 1000 cm3 to prevent or minimize any large fluctuations of gas
pressure during the experiments. The permeation cell was calibrated and the
volumes of the downstream compartments were calculated using standard
polycarbonate membrane and its published permeability data. Calibration was
also performed from time to time to ensure the accuracy of the data obtained.
The experimental temperature is controlled with a built-in temperature
controller system. The experimental temperature could be controlled since the
permeation cell was thermostatic. The film of dense membrane was mounted
onto the permeation cell and vacuumed at 35 C for more than 12 h before the
gas permeation test was carried out. The thickness of the membrane was
measured using a digital film measurement tool and the error is 1
micrometer. The slope of downstream pressure vs. time (dp/dt) at the steady
state condition was used for the calculation of gas permeability with
accordance to the below equation:
273.15 1010 V L dp / dt
P (3-5)
76
760 A T p
2 14.7
where P refers to the gas permeability of a membrane in Barrer (1 Barrer =

11010 cm3 (STP) cm/cm2 s cmHg), V is the volume of downstream chamber
(cm3), L is the membrane thickness (cm), A refers to the effective area of
56
membrane (cm2), T is the operating temperature (K) and p2 is the upstream
operating pressure (psia).
The permeability of each gas was obtained from the average value of at least 3
tests with a deviation of smaller than 1%. The ideal selectivity is defined as
follows: A/B=PA/PB, where PA and PB are the permeability of gases A and B,
respectively.
Figure 3.8: Schematic diagram of the dense film gas permeation testing cell
3.4.2.2 Hollow fiber membranes

The gas separation performance of the hollow fiber membranes is determined
using the tube permeation setup as depicted in Figure 3.9 [13]. Each
membrane module was composed of 10 fibers with an effective length of
about 15 cm per fiber. In each spinning condition, three modules were tested
at room temperature. The gas was tested in the order of CH4 (200 psi), CO2
(200 psi), C3H8 (50 psi), and C3H6 (50 psi) at ambient temperature. The
detailed module preparation and pure gas permeation test procedures have
been described elsewhere [14]. The gas permeance and ideal gas pair
selectivity of dual-layer hollow fiber membrane were calculated with
accordance to the following equations:
P 273.15 106 Q 273.15 106 Q

(3-6)
L T P A T P n Dl
57
A =
P
L A
B P L
(3-7)
B
where P/L is the gas permeance of dual-layer hollow fiber membrane in GPU
(10-6 cm3 (STP)/cm2.s.cmHg), T is the operating temperature (K), Q is the pure
gas rate (cm3/s), P is the gas pressure difference across the dual-layer hollow
fiber membrane (cm Hg), n is the total number of fibers in one testing module,
D is the outer diameter of the testing fiber (cm), and l refer to the effective
length of the testing fiber (cm). (P/L)A and (P/L)B are the pure gas permeance
of gasses A and B, respectively.
Figure 3.9: Schematic diagram of pure gas permeance test apparatus for
hollow fibers [13]
The apparent dense-selective layer thickness (apparent skin thickness) was

calculated according to the following equation:
P 103
L

(3-8)
P
L
where L is the apparent dense-selective layer thickness (nm), and P is the gas
permeability of the dense film made from the outer-layer material in Barrer (1
Barrer =1*10-10 cm3 (STP) cm/cm2.s.cmHg)
58
3.4.3 Mixed gas permeation test
3.4.3.1 Dense flat sheet membrane
Binary CO2/CH4 (1:1) gas permeation tests were obtained by a modified
constant volume permeation system in which an addition valve at the upstream
segment (C5) is included to adjust the stage cut (5%) and another valve in the
downstream port (C6) is installed to introduce the accumulated permeate gas
to an Agilent 7890 gas chromatography (GC) for the analysis of gas
composition. The mixed gas permeation cell is shown in Figure. 3.10. For easy
comparison with pure gas tests, the testing pressure of mixed gas CO2/CH4
permeation tests is 20 atm.
Figure 3.10: Schematic diagram of a mixed gas permeation cell for flat sheet
membranes
The permeability (P) of CO2 and CH4 for each membrane was determined
using the following equations:
273.15 1010 V l Y CO dp
2
PCO (3-9)
2 760 AT po 76 14.7 X CO dt
2
273.15 1010 V l Y CH dp
4
PCH (3-10)
4 760 AT po 76 14.7 X CH dt
4
59
where PCO2 and PCH4 are the CO2 and CH4 permeability, respectively, p0 is the
upstream feed gas pressure (psia), p is the downstream permeate gas pressure
(psia), V is the volume of the downstream chamber (cm3), l is the film
thickness (cm) and X and Y denote the gas molar fraction (%) in the feed and
the permeate stream, respectively. The separation factor for mixed gas tests is
determined from the Eq. (3-11), which represents the ratio of mole fraction
ratios of CO2 and CH4 in the downstream to the upstream.
PCO Y Y
CO = 2 = CO 2 CH 4 (3-11)
2 P X CO X CH
CH 4 CH 4 2 4
3.4.3.2 Hollow fiber membranes

In the CO2/CH4 mixed gas tests, the gas permeances of the hollow fibers were
measured using a modified variable-pressure constant volume gas permeation
cell. The compositions of the permeate and retentate were analyzed with a gas
chromatograph (Hewlett Packard (HP) 6890 Series GC) using a Carboxen
1010 column and a thermal conductivity detector (TCD), while their flow rates
were measured by a digital bubble flow meter. The testing procedure is a
standard one in our lab. First, the feed pressure was set at 50 psi, and the fibers
have been conditioned for 1.5 hours; at the desired pressure, the conditioning
time is at least 30 minutes. In each pressure, the flow rate of both permeate
and retentate were first determined and adjusted by the bubble flow meter,
thus achieving a stage cut around 5%. Thereafter, the permeate or the retentate
will be connected to the GC for sampling of 15 minutes with the gas mixture
being carried to the GC system by the inert gas Helium. Subsequently, the
composition of the mixture was determined by calculating the peak area for
each gas that was eluted out from the column at different retention time.
Finally, the composition of gas from the cylinder will also be tested by GC.
[8]. The hollow fibers were made into modules by using 1/4 in. standard steel
union tees supplied by Swagelok. Each module consisted of one fiber with an
effective length of approximately 10 cm. Two modules were tested under the
same condition and the averages were taken for calculation. A binary
CO2/CH4 (1:1) mixture with a pressure of 20 atm was employed as the
upstream feed gas and the tests were conducted at room temperature.
60
3.5 References
[1] C.E. Sroog, Polyimides, Prog. Polym. Sci. 16 (1991) 561-694.
[3] Y. Wang, T.S. Chung, H. Wang, S.H. Goh, Butanol isomer separation
using polyamideimide/CD mixed matrix membranes via pervaporation,
Chem. Eng. Sci. 64 (2009) 5198-5209.
229-239.
[5] D.F. Li, T.S. Chung, R. Wang, Morphological aspects and structure
control of dual-layer asymmetric hollow fiber membranes formed by a
simultaneous co-extrusion approach, J. Membr. Sci. 243 (2004) 155-175.
[6] L.Y. Jiang, T.S. Chung, D.F. Li, C. Cao, S. Kulprathipanja, Fabrication
of Matrimid/polyethersulfone dual-layer hollow fiber membranes for gas
separation, J. Membr. Sci. 240 (2004) 91-103.
[7] L.Y. Jiang, T.S. Chung, S. Kulprathipanja, An investigation to revitalize
the separation performance of hollow fibers with a thin mixed matrix
composite skin for gas separation, J. Membr. Sci. 276 (2006) 113-125.
[8] B.T. Low, N. Widjojo, T.S. Chung, Polyimide/polyethersulfone dual
layer hollow fiber membranes for hydrogen enrichment, Ind. Eng. Chem.
Res. 49 (2010) 8778-8786.
[9] Y.C. Xiao, T.S. Chung, M.L. Chng, S. Tamai, A. Yamaguchi, Structure
and properties relationships for aromatic polyimides and their derived
carbon membranes: experimental and simulation approaches, J. Phys.
Chem. B 109 (2005) 18741.
[10] A. Bondi, Van der waals volumes and radii, J. Physical Chemistry 68
(1964) 441-451.
composite membranes for hydrogen purification, Energy Environ. Sci. 4
(2011) 4171-4180.
61
[13] H.Z. Chen, Polymeric membranes based on CO2-philic materials for
hydrogen purification and flue gas treatment, Ph.D. thesis, 2012.
[14] T.S. Chung, S.K. Teoh, X.D. Hu, Formation of ultrathin high-
performance polyethersulfone hollow-fiber membranes, J. Membr. Sci.
133 (1997) 161-175.
62
Chapter 4: Cross-linkable 6FDA-Durene/DABA co-polyimides
grafted with Cyclodextrins
4.1 Introduction
Aromatic polyimides, especially the ones containing 4,4'-
(hexafluoroisopropylidene) diphthalic anhydride (6FDA) moiety, have shown
good mechanical and thermal properties, high gas permeability as well as good
permselectivity. It has been found that the - (CF3) - group restricts the
torsional motion of phenyl rings and chain packing in 6FDA dianhydrides so
that both permeability and permselectivity may be improved simultaneously
[1-3]. Many 6FDA-based polyimides exhibit good separation performance for
CO2/CH4 and C3H6/C3H8 gas pairs [1-8]. However, similar to other polymers,
6FDA polyimides still face two major challenges. First, there is a limitation in
achieving the desired performance of a high permeability combined with a
high permselectivity. There is a trade-off relation between permeability and
permselectivity for most gas pairs [9]. Secondly, the separation performance
of 6FDA polyimides generally degrades when separating highly soluble gases
such as CO2 or hydrocarbons at relatively high pressures. This phenomenon is
referred as plasticization [7-12]. In the occurrence of plasticization, the
penetrant species such as CO2 swell up the polymer matrix, facilitate
segmental motions, and accelerates diffusion processes of all gas species [12-
14]. The gas pair selectivity is therefore reduced and the ideal selectivity
measured by means of pure gas tests can no longer be used to estimate the
mixed gas membrane performance [15].
Cross-linking has been advised as a tactic to improve the membrane

performance by decreasing plasticization effects, improving the
permselectivity, as well as improving the thermal and chemical stability of the
membranes [4, 8, 10, 16-21]. Since cross-linking normally causes a severe
decrease in polymer permeability by reducing FFV and chain flexibility,
investigations have been carried out to design high free volume polyimides to
alleviate the decrease in permeability. One of the methods to prepare high free
volume membrane is the decomposition of labile components in block or graft
63
co-polymers as low k materials used in semi-conductor industries [21, 22].
Recently in our group [8], -Cyclodextrin (-CD), a thermal labile molecule
with a large molecular size [23], has been grafted on a polyimide. After
thermal treatment, the thermal labile molecules decomposed and formed free
volumes in the polymer matrix and the resultant membrane showed extremely
enhanced permeability. In addition, carboxylic groups of the polyimide
backbone underwent cross-linking reactions at high temperatures, tightened d-
space, and increased the gas selectivity and plasticization resistance. In
addition, the free volume radius (mean cavity radius) of -CD grafted co-
polyimides after heat treatment at 425 C was around 3.8 (cavity diameter is
around 7.6 ) which is higher than CO2, CH4, C3H6, and C3H8 kinetic
diameters. Since the kinetic diameters of CO2, CH4, C3H6, and C3H8 are 3.3 ,
3.8 , 4.5 , and 4.3 respectively, better selectivity for CO2/CH4 and
C3H6/C3H8 separation may be achievable if one can manipulate the free
volume radius and increase the total free volume by changing the Cyclodextrin
size and optimizing the heat-treatment conditions.
In this chapter, three Cyclodextrins (CDs) were grafted on the 6FDA-

Durene/DABA (9/1) co-polyimide by esterification [8]. The three CDs,
namely, -, -, and -CD, are formed by six to eight glucopyranoside units.
Thermal treatment of the co-polyimide was conducted from 300 to 425 C to
study the effects of different annealing temperatures on the membrane
separation performance. The functions of thermal treatment are as follows (1)
decomposing the structure of CD to create more pore volumes in the
membrane matrix, and (2) inducing thermally cross-linked rigid-chain
structure that avoids pore collapsing [8, 24]. The desired cross-linked structure
has been shown in Figure 4.1. As can be seen in Figure 3.2, the sizes of these
Cyclodextrins are different. After thermal treatment, the decomposed CD
would leave different sizes of microvoids in the membrane. Therefore, the
purpose of this work is to explore the science and engineering if one can
incorporate different sizes of CD and manipulate the free volume and its size
with enhanced permeability and selectivity of the resultant membrane via by
thermal degradation and partial cross-linking of the polymer matrix.
64
Figure 4.1: A hypothesis of the cross-linked structure a) low temperature
(below 350C) b) high temperature (above 350C)
4.2 Results and discussion

4.2.1 Characterization
The inherent viscosities of original PI (6FDA-Durene/DABA (9:1)), PI-g--
CD, PI-g--CD, and PI-g--CD is given in Table 4.1. The inherent viscosities
decrease after CD grafting in all co-polyimides. The number average
molecular weight (Mn) and polydispersity (PDI) of original PI, PI-g--CD, PI-
g--CD, and PI-g--CD co-polyimides are given in Table 4.1. Both Mn and
PDI of co-polyimides after grafting CDs decrease, possibly due to chain
scission during esterification because of high reaction temperature and acid
catalyst.
Table 4.1: Properties of the synthesized co-polyimides before and after

grafting different kinds of CD
Mn PDI Inherent Viscosity (dL/g)
Polymer
(g/mol) (Mw/Mn) (0.5 % in NMP 25C)
Original PI 82345 3.25 1.01
PI-g--CD 20359 2.96 0.54
PI-g--CD 17149 2.92 0.55
PI-g--CD 19299 2.95 0.47
65
In this study it was found that the heat treatment changes membrane color,
flexibility, density and permeation characteristics. The FTIR-ATR spectrum of
6FDA-Durene-DABA (imide group) can be characterized by bands around
1351 cm-1 (CN stretch), 1716 cm-1 (asymmetric C=O stretch), 1786 cm-1
(symmetric C=O stretch) and 741 cm-1 (bending of C=O). As shown in Figure
4.2, in general the imide groups of these co-polyimides are thermally stable
until 400 C and the intensities of these peaks below 400 C are constant. This
implies the decomposition of CD structure may not have a significant effect on
the imide group until 400 C. These results are similar to the results obtained
by Wieneke and Bickel [25], Shao et al [26], and Xiao et al [8]. Wieneke and
Bickel [25] showed that there was a strong relationship between the intensity
of peaks in the region of 3200-3400 cm-1 which represented the OH vibration
of the acid group and the amount of DABA in the 6FDA-copolyimide and
these peaks decreased after treatment at 400 C.
The intensities of the peaks at 15801600 cm-1, representing CC stretches

vibrations in the aromatic ring, increase when the membranes are treated at
425 C. This phenomenon may arise from the formation of new covalent
bonds between benzene rings [8] because radical DABA groups are formed
due to decarboxylation over 400 C [24]. In other words, two radical DABA
groups may form linkages as biphenyl [27]. For the membranes treated at 425
C, the intensities of imide groups are reduced visibly because the imide ring
starts to decompose at this temperature. On the other hand, the intensities of
peaks at 1580-1600 cm-1, representing CC stretching vibrations in the
aromatic ring increase, indicating the stronger cross-linking reaction between
aromatic rings. Compared with the original co-polyimide, the co-polyimide
grafted by CDs shows a stronger intensity of the characteristic peaks of OH
groups at 1620 and 3650 cm-1, indicating the success of esterification between
the co-polyimide and CD. Heat treatment at 200 C does not decompose the
grafted CD because the characteristic peaks of OH groups at 1620 and 3650
cm-1 are still clear. By increasing temperature, CD decomposes and the
intensities of these peaks decreases.
66
Figure 4.2: FTIR-ATR spectra of co-polyimide membranes at different heat
treatment temperatures
The intensities of the characteristic peaks of OH groups at 1620 cm-1, decrease

by decreasing the number of OH groups of CD. As a result, PI-g--CD has a
67
stronger intensity than that of the PI-g--CD and PI-g--CD has a stronger
intensity than that of the PI-g--CD. Compared to the original PI, the co-
polyimides grafted by CDs at 200 C shows a weaker intensity of the
characteristic peaks of COOH groups at 3200-3400 cm-1 because some COOH
groups have reacted with the CD during the esterification reaction.
The weight losses and derivatives of weight losses (decomposition rates) of

the original co-polyimide and PI-g-CDs as a function of temperature in
nitrogen atmosphere are shown in Figure 4.3. The original co-polyimide
shows two stages of decomposition. The first stage of weight loss of around 6
% is detected from 400-500 C, which comes mainly from the decarboxylation
of a DABA moiety of the co-polyimide [23]. In the second stage, additional 40
% weight loss occurs from 500 to 700 C mainly because of the
decomposition of imide and 6FDA groups in the polymer backbone.
Figure 4.3: Residual weight and decomposition rate vs. temperature for
original co-polyimide and PI-g-CDs
Co-polyimides grafted CDs have one more stage of decomposition. In this

stage, around 10 % weight loss takes place in the range of 200 to 400 C
mainly due to the decomposition of CDs structure [8, 23]. The weight losses
for the second and third stages are around 4 % and 35 %, respectively. The
68
weight losses of PI-g-CDs polymers due to the decarboxylation of DABA at
400-500 C are smaller than that of the original PI. This is because the former
has less COOH groups than the latter. Theoretically, if all COOH groups react
with CD, the weight losses of PI-g--CD, PI-g--CD, and PI-g--CD, owing to
CD decomposition at 200-400 C should be around 14.5, 16.5, and 18.5 %,
respectively. However, the real weigh losses are around 9, 10, and 11.5 % for
PI-g--CD, PI-g--CD, and PI-g--CD, respectively because only around 60
% of COOH groups have reacted with CDs.
The XRD patterns of original co-polyimide and co-polyimide membranes

grafted by CDs in different treatment temperatures are shown in Figure 4.4.
All of the wide angle X-ray diffraction (WAXD) patterns of these co-
polyimides are broad indicating that they have an amorphous structure.
Figure 4.4: WAXD patterns of co-polyimide membranes at different

The WAXD peak in the amorphous glassy polymer is often used to estimate
the average inter-chain spacing distance (d-spacing). As shown in this figure,
the d-spacing value increases with increasing post-treatment temperature up to
69
350 C and a further heat treatment at 425 C reduces the d-space. For the
original co-polyimide, the initial increment in d-spacing is possibly due to an
increase in chain mobility by increasing temperature [26]. However, the
decomposition of Cyclodextrin structure, which makes a lot of microvoids and
channels between polyimide chains, is the drive mechanism for the initial
increment in d-spacing in co-polyimides grafted by CDs [8]. The possible
cause of decreasing d-spacing above 350 C arises from the cross-linking
reaction at this high temperature that leads to higher chain packing; hence, the
distance between chains becomes smaller. Interestingly, the decreasing trend
in d-spacing is severer in co-polyimides grafted by CDs (decreasing from 6.40
to 5.84) than in the original PI (decreasing from 6.30 to 6.10). As illustrated in
Figure 4.4, the d-spacing values of PPM--CD-425, PPM--CD-425 and
PPM--CD-425 are lower than that of the original PI. This phenomenon may
be attributed to the factor that the decomposition of the CD may promote
cross-linking reactions due to the presence of more free radicals.
Figure 4.5 shows the evolution of color change of co-polyimide membrane

grafted -CD from colorless to black due to heat treatment. However, the
membrane thermally treated 425 C still has very high flexibility and
mechanical strength that are much better than conventional carbon molecular
sieve membranes. Additional physical properties of the original PI and co-
polyimide membranes grafted by CDs after heat treatment at different
temperatures are provided in Table 4.3. The values for the glass transition
temperature (Tg), density, and fractional free volume (FFV), which refers to
the ratio of the so-called "expansion volume" to the observed volume are
shown. In all cases, Tg increases with increasing heat treatment temperature
especially at 400 C. Clearly, an increase in Tg means an increase in polymer
chain rigidity due to occurrence of the cross-linking reaction at higher
temperatures. All co-polyimides at 425 C do not show a Tg by means of
running DSC up to 450 C, though there are some thermal events. Moreover,
compared to -CD and -CD, density of co-polyimide grafted -CD decreases
more. Also, the order FFV of these co-polyimides follows PI-g--CD > PI-g-
-CD > PI-g--CD due to the size effect of various CD. Since -CD is bigger
70
than -CD and -CD is bigger than -CD, the micro-pores created by -CD
after decomposition is bigger than those created by -CD and -CD.
Figure 4.5: Flexibility of the co-polyimide grafted with -CD membrane

thermally treated at 425 C
In case of density, it can be seen that it decreases with increasing heat

treatment temperature and as a result, the FFV value increases. This decrease
in density and increase in FFV, with the aid of heat treatment, is mainly
attributed to the decomposition of CD structure. According to TGA, these
polymers lose CD in the temperature range from 300 to 400C and if polymer
chains are adequately rigid, the spaces occupied by CD might be kept as
micro-pores after heat treatment [20]. Since it is still unclear that how much
CD has decomposed and been lost between 300 to 425 C, calculating FFV for
the co-polyimide grafted CDs in this temperature range is difficult.
71
Table 4.2: Physical properties of co-polyimide membranes in different heat
Tg (C) Density (g/cm3) V (cm3/g) FFV
Original PI-200 399 1.332 0.751 0.184
Original PI-300 400 1.327 0.754 0.187
Original PI-350 402 1.329 0.753 0.187
Original PI-400 408 1.325 0.755 0.189
Original PI-425 --- 1.304 0.767 0.201
PI-g--CD-200 400 1.325 0.755 0.203
PI-g--CD-300 402 1.315 0.760 ---
PPM--CD-350 404 1.304 0.767 ---
PPM--CD-400 410 1.291 0.774 ---
PPM--CD-425 --- 1.284 0.779 ---
PI-g--CD-200 401 1.327 0.754 0.205
PI-g--CD-300 402 1.316 0.760 ---
PPM--CD-350 405 1.305 0.767 ---
PPM--CD-400 410 1.287 0.777 ---
PPM--CD-425 --- 1.278 0.783 ---
PI-g--CD-200 401 1.324 0.755 0.208
PI-g--CD-300 403 1.315 0.761 ---
PPM--CD-350 405 1.298 0.770 ---
PPM--CD-400 412 1.281 0.780 ---
PPM--CD-425 --- 1.268 0.788 ---
4.2.2 Pure gas permeation experiments

The pure gas permeability and ideal selectivity of co-polyimides before and
after CD graft and then heat treated at different temperatures are shown in
Table 4.3. As can be seen, the permeability of all gases increases with heat
treatment temperature. At low temperatures, hydrogen bonds among COOH
groups of DABA tighten the polymer segmental packing and reduce the free
volume within the polymer matrix [28]. Hydrogen bonding is broken and
chain packing becomes a loser with heat treatment temperature and this can be
a reason that the permeability of original PI increases with heat treatment
temperature. Another reason may be due to the thermal expansion of the
polymer matrix. The chain mobility increases with heat treatment temperature
due to the thermal expansion of the polymer matrix and permeability increases
[26]. As mentioned before, after heat treatment, the spaces occupied by CD
can be kept as micro-pores that increase fractional free volume and gas
permeability. The degree of increase in permeability generally follows the
72
order of -CD > -CD > -CD due to different CD structures and thermal
decomposition behaviors.
Table 4.3: Pure gas permeability and selectivity of co-polyimide membranes

(10 atm and 35 C)
Permeability (Barrer) Selectivity
Membranes CO2 CH4 C3H6 C3H8 CO2/CH4 C3H6/C3H8
Original PI-200 235 10 --- --- 22.58 ---
Original PI-300 326 17 --- --- 19.18 ---
Original PI-350 392 23 11 0.81 17.23 13.65
Original PI-400 519 29 30 2.1 18.01 14.42
Original PI-425 1302 63 73 4.1 20.83 17.97
PI-g--CD-200 241 12 --- --- 20.14 ---
PI-g--CD-300 373 23 --- --- 16.79 ---
PPM--CD-350 416 26 17 1.2 16.06 13.32
PPM--CD-400 568 31 61 4.6 18.51 13.24
PPM--CD-425 2423 112 295 17 21.70 17.77
PI-g--CD-200 238 12 --- --- 20.67 ---
PI-g--CD-300 403 25 --- --- 16.05 ---
PPM--CD-350 593 37 18 1.2 15.91 14.09
PPM--CD-400 772 41 81 5.9 18.65 13.88
PPM--CD-425 3112 140 396 22 22.19 17.86
PI-g--CD-200 251 12 --- --- 21.21 ---
PI-g--CD-300 416 24 --- --- 17.46 ---
PPM--CD-350 663 41 22 1.5 16.34 14.22
PPM--CD-400 929 48 102 7.3 19.28 13.98
PPM--CD-425 4211 187 521 28 22.44 18.09
The selectivity of these co-polyimides decreases and then increase and, this
trend is consistent with the evolution of d-spacing of the co-polyimides. As
mentioned before, cross-linking reaction takes place at high temperatures and
makes chain packing denser. According to Steel and Koros [29] and Xiao and
Chung [8], ultramicropores and micro-pores are possibly formed in the
polymeric membrane treated at high temperatures. The ultramicropores have
dimensions equivalent or close to the d-spacing of the co-polyimide and
determine gas selectivity, whereas the micro-pores have bigger sizes created
by the CD decomposition and enhance gas permeability. Since the dimension
of ultramicropores decreases at high temperatures because of cross-linking
reaction, the gas pair selectivity is increased. According to WAXD results, the
73
degrees of cross-linking reaction are higher for co-polyimide membranes
grafted by CDs than that for the original PI, thus the former has a higher
selectivity than the later after thermal treatment.
One observation from Table 4.3 is that the permeability of the membranes heat
treated at 400 and 425 C are related to the cavity diameter of the CDs. Figure
4.6 shows the dimensions of CDs [30] and permeability ratios of PPM--CD
to PPM--CD and PPM--CD to PPM--CD at 400 and 425 C. As can be
seen in this figure, the cavity diameters of -CD, -CD, and -CD are around
5.7, 7.8 and 9.5 , respectively, and the cavity diameter ratios of -CD to -
CD and -CD to -CD are around 0.6 and 0.8, respectively. These ratios are
almost the same as the permeability ratios of PPM--CD to PPM--CD and
PPM--CD to PPM--CD at 400 and 425 C. Thus, the micro-pores size is
strongly related to the CD cavity size, while the permeability is strongly
associated with the micro-pores size. The permeability of the original PI
thermally treated at 425 C is 4 to 6 times higher than that treated at 200 C.
This increase is 8 to 10 times for PPM--CD-425 and 15 to 17 times for PPM-
-CD-425. These rapid permeability jumps have resulted from the fact that CD
decomposes at high temperatures and also bigger CD creates bigger micro-
pores. Another observation is that all membranes heat treated at 200 C have
almost the same permeability. This trend is due to the fact that the CD does
not decompose at this low temperature and all membranes may have almost
the same size of micro-pores. The order of gas permeability for all membranes
follows this sequence: C3H8 (kinetic diameter: 4.30 ) <CH4 (3.80 ) <C3H6
(4.5 ) < N2 (3.64 ) < O2 (3.46 ) < CO2 (3.30 ). Except for C3H6 gas, the
increasing order is consistent with decreasing kinetic diameter of gas
molecules that shows the separation mechanism may follow the solution
diffusion mechanism. The high permeability of C3H6 over CH4 and C3H8 may
be due to the effect of solubility. In other words, C3H6 has a higher solubility
than C3H8 and CH4. In addition, CO2 has the highest permeability because of
its high condensability, sorption capability and unique molecular linearity [26].
74
Figure 4.6: a) Dimensions of , , and -CD [30] and b) Permeability ratios of
PPM--CD to PPM--CD and PPM--CD to PPM--CD at 400 and 425 C.
As aforementioned, plasticization is not desirable in gas separation processes.

Plasticization is a pressure dependent phenomenon caused by the dissolution
of certain components within the polymer matrix which disrupts the chain
packing and enhances inter-segmental mobility [4, 5, 31-33]. The pressure at
which the corresponding gas permeability exhibits a minimum is known as the
plasticization pressure. Figure 4.7 illustrates the CO2 plasticization behavior of
the newly developed co-polyimide membranes thermally treated at different
temperatures. As shown in this figure, all membranes before heat treatment
appear to undergo plasticization at around 5 atm CO2 feed pressure and by
increasing the thermal treatment temperature, plasticization pressure increases.
Its clear that cross-linking reaction at high-temperatures confine the
movement of polymer chains and enhances anti-plasticization performance.
Another reason of this phenomenon may be due to higher Langmuir sorbed
molecules in contrast to Henrys law contribution at high-temperatures. PPM-
-CD-425 and PPM--CD-425 show superior anti-plasticization characteristics
up 30 atm possibly because higher degrees of cross-link reaction.
75
Figure 4.7: CO2 plasticization behavior of a) original PI, b) PI-g--CD, c) PI-
g--CD, and d) PI-g--CD at different heat treatment temperatures.
4.2.3 Mixed gas permeation experiments

Since the crucial applications of membranes are in the presence of mixed gas
streams, it is important to achieve a high selectivity for mixed gas experiments
to show commercial feasibility. In this study, mixed gas experiments were
conducted using a 50:50 feed mixture of CO2/CH4 with a total feed pressure of
20 atm at 35 C. Table 4.4 summarizes the gas permeability and selectivity of
co-polyimides thermally treated at 425 C under pure and mixed gas tests. For
the original PI-425, the mixed gas selectivity decreases tremendously when
comparing with pure gas results due to the effect of plasticization. Clearly, the
thermal treatment at 425 C does not enhance its anti-plasticization
characteristics. Consistent with Figure 4.7, the permeability of PPM--CD-425
and PPM--CD-425 in mixed gas tests shows slightly decreases, while their
selectivity is almost the same as those in pure gas tests.
The differences in pure gas and mixed gas systems have been studied by
Chern et al [34]. Assuming the permeability of a pure gas in glassy polymers
can be described by the dual mode or partial immobilization model, the
76
permeability of gas A may be expressed by Eq. (4-1) when the downstream
pressure is zero.
FA K
P kDA DDA 1 (4-1)
(1 bA pA )
where KDA, DDA, bA and pA refer to the Henrys law constant, Henrys law
diffusion coefficient of the penetrant, Langmuir affinity constant, and
upstream pressure of pure gas, respectively. FA is the ratio of the diffusion
coefficient of the penetrant in the Langmuir sites (i.e., microvoids) to that in
the Henry sites for gas A (FA = DHA/DDA), and K = CHA bA/k, is a useful
collection of the sorption parameters introduced above when CHA refers to the
Langmuir capacity constant for pure gas A.

treated in 425 C
Pure Gas a Mixed Gas b
Permeability Permeability
Membranes Selectivity Selectivity
(Barrer) (Barrer)
CO2 CH4 CO2/CH4 CO2 CH4 CO2/CH4
Original PI-425 1302 62 20.83 1221 91 13.41
PPM--CD-425 2423 112 21.70 2392 117 20.44
PPM--CD-425 3112 140 22.19 2911 133 21.85
PPM--CD-425 4211 187 22.44 3976 174 22.85
a
measured at 10 atm and 35 C
b
measured at 20 atm and 35 C
The permeability of gas A in mixed gas tests are given by Eq. (4-2). For
simplicity, gas A is to the component of primary interest, while gas B is the
second competing component. PA and PB refer to the partial pressures of
components A and B, respectively.
FA K
P kDA DDA 1 (4-2)
(1 bA pA bB pB )
As can be seen, Eq. (4-2) has one extra term in the denominator compared to
Eq. (4-1), which decreases the permeability amount for gas A. Therefore, gas
permeability obtained under mixed gas test is usually lower than that under
77
pure gas test. However, selectivity may stay almost constant, but still
somewhat depends on many other factors.
Figure 4.8: Trade off lines for CO2/CH4 and C3H6/C3H8 separation (empty
symbols are pure gas results, solid symbols are mixed gas results)
78
A comparison of gas separation performance of the original PI and co-
polyimide membranes grafted by CDs thermally treated at 425 C is shown in
Figure 4.8. The separation performance of the thermally treated original PI is
much inferior to those CD grafted PI. For CO2/CH4 separation, the entire PI
grafted CDs membranes surpass the trade-off line, while the thermally treated
original PI is still within the trade-off line. Clearly, the CD decomposition
facilitates micro-pore formation and enhances the permeability. For
C3H6/C3H8 separation, all membranes thermally treated at 425 C can surpass
the trade-off line. They have almost the same permselectivity, but permeability
follows the order of PPM--CD-425 > PPM--CD-425 > PPM--CD-425 >
Original PI-425.
4.3 Conclusion
Using a cross-linkable co-polyimide (6FDA-Durene/DABA (9/1)), we have
developed new high-performance gas separation membranes that can enhance
both membrane permeability and resistance to plasticization simultaneously by
grafting various sizes of Cyclodextrin to the polyimide matrix and then
decomposing them at elevated temperatures. The following conclusions can be
made from this work:
The decomposition of the CD may convert the spaces originally
occupied by CD to micro-pores and thus increase fractional free
volume and gas permeability.
Permeability increase is strongly related to the cavity size of a CD; the
bigger CD size, the higher permeability jumps after thermal treatments
at elevated temperatures.
Permselectivity increases if the thermal treatment is conducted over
400 C due to the occurrence of cross-linking reaction in the polyimide
matrix that tightens the d-space.
The separation performance of thermally treated polyimides that were
originally grafted with the CD is much better than that without the CD.
For CO2/CH4 separation, all the co-polyimide membranes grafted by
CDs surpass the trade-off line, while the thermally treated original
polyimide cannot.
79
For C3H6/C3H8 separation, all membranes thermally treated at 425 C
can surpass the trade-off line with almost the same permselectivity.
However, their permeability follows the order of PPM--CD-425 >
PPM--CD-425 > PPM--CD-425 > Original PI-425.
The thermally treated polyimides that were originally grafted with the
CD show much better anti-plasticization characteristics than that
without the CD. The plasticization pressure of the PPM--CD-425
membrane is over 30 atm, while the original polyimide is less than 15
atm.
The best result was achieved from the PPM--CD-425 membrane. It
has a CO2 permeability of 4211 Barrers with a CO2/CH4 ideal
selectivity of 22.44 and a C3H6 permeability of 521 Barrers with a
C3H6/C3H8 ideal selectivity of 18.09. The CO2 permeability drops
slightly to 3976 Barrers with almost the same CO2/CH4 selectivity of
22.84 in mixed gas tests.
The newly developed PPM--CD-425, PPM--CD-425, and PPM--
CD-425 membranes are highly flexible with good mechanical strength.
80
4.4 References
[2] S.S. Chan, R. Wang, T.S. Chung, Y. Liu, C 2 and C3 hydrocarbon
separations in poly (1,5-naphthalene-2,2'-bis(3,4-phthalic)
Membr. Sci. 210 (2002) 55-64.
[3] S. Neyertz, D. Brown, S. Pandian, N.F.A. van der Vegt, Carbon dioxide
diffusion and plasticization in fluorinated polyimides, Macromolecules
43 (2010) 7813-7827.
strategies of molecular architecture and modification of polyimide based
membranes for CO2 removal from natural gas-A review, Prog. Polym.
Sci. 34 (2009) 561-580.
separation using polymeric membranes, J. Membr. Sci. 211 (2003) 299-
309.
212 (2003) 13-27.
[7] M. Das, W.J. Koros, Performance of 6FDA-6FpDA polyimide for
400.
[11] Y.C. Xiao, T.S. Chung, M.L. Chng, Surface characterization,
modification chemistry and separation performance of polyimide and
PAMAM dendrimer composites, Langmuir 20 (2004) 8230-8238.
81
[13] W.J. Koros, M.W. Hellums, Gas separation membrane material selection
criteria: differences for weakly and strongly interacting feed
components, Fluid Phase Equilib. 53 (1989) 339-354.
[14] A. Bos, I.G.M. Punt, M. Wessling, H. Strathmann, Plasticization-
Sci. 103 (1995) 73-82.
propylene/propane separation by carbonaceous membrane derived from
poly (aryl ether ketone)/2,6-bis(4-azidobenzylidene)-4-methyl-
[17] H. Kita, T. Inada, K. Tanaka, K. Okamoto, Effect of photocrosslinking
on permeability and permoselectivity of gases through benzophenone-
containing polyimide, J. Membr. Sci. 87 (1994) 139-147.
[18] C.S. Bickel, W.J. Koros, Improvement of CO2/CH4 separation
characteristics of polyimides by chemical crosslinking, J. Membr. Sci.
155 (1999) 145-154.
diaminobutane (DAB) dendrimers, J. Membr. Sci. 238 (2004) 153163.
[20] C.E. Powell, X.J. Duthie, S.E. Kentish, G.G. Qiao, G.W. Stevens,
Reversible diamine crosslinking of polyimide membranes, J. Membr.
Sci. 291 (2007) 199-209.
[21] B. Cruden, K. Chu, K. Gleason, H. Sawin, Thermal decomposition of
low dielectric constant pulsed plasma fluorocarbon films: ii. Effect of
postdeposition annealing and ambients, J. Electrochem. Soc. 146 (1999)
4597-4604.
82
[22] B. Lee, Y. H. Park, Y. T. Hwang, W. Oh, J. Yoon, M. Ree, Ultralow-k
nanoporous organosilicate dielectric films imprinted with dendritic
spheres, Nature Materials 4 (2005) 147-151.
[23] L.X. Song, P. Xu, A comparative study on the thermal decomposition
behaviors between -cyclodextrin and its inclusion complexes of organic
amines, J. Phys. Chem. A 112 (2008) 11341-11348.
[24] A.M. Kratochvil, W.J. Koros, Decarboxylation-induced crosslinking of a
polyimide for enhanced CO2 plasticization resistance, Macromolecules
41 (2008) 7920-7927.
[25] J.U. Wieneke, C.S. Bickel, Thermal stability of 6FDA-(co-)polyimides
containing carboxylic acid groups, Polymer Degradation and Stability 95
(2010) 684-693.
[26] L. Shao, T.S. Chung, K.P. Pramoda, The evolution of physicochemical
and transport properties of 6FDA-durene toward carbon membranes;
from polymer, intermediate to carbon, Microporous Mesoporous Mater.
84 (2005) 59-68.
[27] T.P. Eskay, P.F. Britt, A.C. Buchanan, Does decarboxylation lead to
crosslinking in low-rank coals?, Energy Fuels 10 (1996) 1257-1261.
[28] W. Qiu, C.C. Chen, L. Xu, L. Cui, D.R. Paul, W.J. Koros, Sub-Tg cross-
linking of a polyimide membrane for enhanced CO2 plasticization
resistance for natural gas separation, Macromolecules 44 (2011) 6045-
6056.
[29] K.M. Steel, W.J. Koros, Investigation of porosity of carbon materials
and related effects on gas separation properties, Carbon 41 (2003) 253-
266.
[30] http://www.chromatography-online.org/Chrial-GC/Contemporary-
Chiral-Stationary-Phases/Cyclodextrin. html
[31] F.G. Kerry, Industrial gas handbook: Gas separation and purification,
Taylor and Franc. Group, New York, 2006.
removal of acid gases from natural gas, J. Membr. Sci. 140 (1998) 27-49.
83
[34] R.T. Chern, W.J. Koros, E.S. Sanders, R. Yui, Second component
effects in sorption and permeation of gases in glassy polymers, J.
Membr. Sci. 15 (1983) 157-169.
84
Chapter 5: Permeability, Solubility, Diffusivity and PALS
Data of Cross-linkable 6FDA-based Co-polyimides
5.1 Introduction
Membrane separation has received much attention, and various polymeric
membrane materials have been synthesized and investigated for gas separation
applications. Among them, the polyimide family is one of the most attractive
and favorable materials due to its excellent properties such as high thermal
stability, chemical resistance, mechanical strength, and spinnability [1-9].
Within the polyimide materials, hexafluoroisopropylidene diphthalic
anhydride (6FDA)-based co-polyimides are of particular interest since they
often possess both high gas permeability and selectivity for CO2/CH4 and
olefin/paraffin separation [4-15].
In order to compete with other separation technologies, polymeric membranes

must be carefully designed with consideration of particular applications. A
molecular-level understanding of materials and engineering factors that govern
the permeability, solubility, and diffusivity coefficient of gases across the
membranes must be conducted and comprehended [1-12, 16-21]. Gas
permeation across a polymer membrane usually follows the solution-diffusion
mechanism where the permeability is a product of the solubility coefficient
and diffusivity coefficient [12, 16, 22]. In other words, gas permeation follows
by gas sorption onto the membrane surface, then the gas molecules diffuse
across the polymer matrix, and subsequently desorb from the other surface of
the membrane. Thus, the permeation process involves in a combination of
thermodynamic factors like condensability of the gas and its interaction with
polymer segments plus kinetic factors that are mostly dependent on penetrant
size, segment mobility and free volume [22]. Usually, two out of the
aforementioned three coefficients that characterize a membrane permeability
are sufficient to understand its separation behavior. However, if the polymeric
membrane is in the glassy state, the three coefficients may not be precise since
glassy polymeric membranes are often in a non-equilibrium state. Besides,
structural information on free volume and its size distribution in glassy
polymers is another important factor in determining the transport properties.
85
This is due to the fact that the intrinsic transport properties such as diffusivity
and solubility are strongly related to the chemical structure and the free
volume of membrane materials. The interdependencies have been well
summarized in several articles [3, 7, 8, 13-15, 21-25].
In chapter 4, we have developed novel 6FDA-based co-polyimides which

were structurally grafted by thermally labile groups such as , , and -
Cyclodextrin (CD) and then decomposed with the purpose to enhance gas
separation performance for CO2/CH4, O2/N2 and C3H6/C3H8 separation
[4,9,26-27]. However, no fundamental studies have been conducted to
investigate their sorption behavior and transport mechanism as a function of
thermally labile group. Therefore, the purposes of this chapter are to
comprehend these novel materials by examining the relationship between the
gas permeation properties and grafted CD sizes over a wide range of pressures
and annealing temperatures using the dual-mode sorption model [14, 16-22,
28-29]. In addition to using conventional permeation and sorption cells,
positron annihilation lifetime spectroscopy (PALS) is utilized in this study
[24-26, 30-33]. It is envisioned that this fundamental study may provide useful
insights for better materials, design and explore new materials for industrial
gas separation.

5.2.1 Solubility coefficient of co-polyimide membranes
It is well known that solubility is determined by (1) the inherent
condensability of the penetrant, (2) the polymerpenetrant interactions, and (3)
the amount and distribution of excess free volume in the glassy polymer [13-
22, 28-29, 34-41]. In this study, Figure 5.1 and Figure 5.2 shows the sorption
isotherms as a function of CH4, CO2, C3H8 and C3H6 and their pressures at 35
C. All sorption isotherms are concave to the pressure axis, which were as
expected on the basis of the typical behavior of glassy amorphous polymers
and of the dual mode sorption model. Penetrants sorbed in Henrys sites (CD)
is believed to be same as a gas dissolved in a rubbery polymer, while
penetrants sorbed in Langmuir sites (CH) has a capacity limit due to a hole-
filling process behavior [28-29, 34-42].
86
Figure 5.1: Sorption isotherms of CH4 and CO2 in different co-polyimide
membranes at 35 C
87
Figure 5.2: Sorption isotherms of C3H8 and C3H6 in different co-polyimide
membranes at 35 C
Table 5.1 tabulates the dual sorption parameters of Eq. (2-8) using a nonlinear
least-squares regression to fit the sorption isotherms. The sorption of CH4 is
much less than that of other gases due to the lower condensability and weak
88
interactions with the polymers. As shown in Table 5.1, Figure 5.3 and Figure
5.4, for all the penetrant gases investigated in this study, the original PI-200
has slightly lower solubility coefficients comparing with those of PI-g-CDs-
200, indicating that the sorption capacity of the membranes increases after
grafting various types of CD into the PI.
Table 5.1: Solubility, diffusivity, and dual mode sorption parameters of

different co-polyimide membranes for different gasses at 35 C and 1 atm
-2 -8
Membrane Gas kD C'H b S10 D10
CH4 0.4971 14.196 0.1428 3.07 3.660
CO2 1.9908 41.580 0.6404 23.82 10.310
Original PI-200 C3H8 3.9278 26.427 2.4843 29.26 0.032
C3H6 4.1046 31.535 3.8979 37.53 0.580
CH4 0.6414 20.148 0.1808 4.93 14.910
CO2 2.3087 70.616 0.6260 39.55 45.690
Original PI-425 C3H8 4.5640 39.543 2.3069 42.31 0.111
C3H6 4.0918 56.408 3.9076 63.47 1.570
CH4 0.5078 13.799 0.1531 3.06 4.260
CO2 2.0986 44.784 0.5799 24.37 10.400
PI-g--CD-200 C3H8 3.8223 27.699 2.5880 31.32 0.039
C3H6 3.9091 36.237 3.7180 41.92 0.470
CH4 0.6853 33.133 0.2167 8.68 16.940
CO2 2.2039 84.191 0.9880 58.62 78.420
PPM-g--CD-425 C3H8 5.6030 63.622 3.2611 71.45 0.297
C3H6 4.5863 73.410 3.6354 80.19 5.990
CH4 0.6037 14.223 0.1587 3.39 3.650
CO2 2.2115 45.294 0.5535 24.42 10.680
PI-g--CD-200 C3H8 4.0449 26.580 2.6808 30.79 0.043
C3H6 3.7874 39.277 3.9022 45.24 0.470
CH4 0.7253 35.573 0.2030 8.88 20.380
CO2 2.4210 87.805 1.1002 64.41 94.670
PPM-g--CD-425 C3H8 5.7107 63.226 2.9348 69.59 0.433
C3H6 4.3296 73.883 3.5409 80.12 8.280
CH4 0.5512 14.315 0.1544 3.14 4.040
CO2 2.3337 49.654 0.5622 26.94 10.150
PI-g--CD-200 C3H8 4.0026 28.151 2.7914 32.53 0.048
C3H6 3.9907 40.956 3.8124 47.03 0.580
CH4 0.7866 40.149 0.1998 9.86 24.540
CO2 2.6536 87.672 1.0991 64.40 104.490
PPM-g--CD-425 C3H8 5.6349 66.252 2.8486 71.93 0.504
C3H6 4.3846 78.928 3.5002 84.02 0.290
89
The units of kD is cm3 (STP)/cm3polymer. atm, CH has units of cm3 (STP)/
cm3polymer, b has units of atm-1, S has units of cm3 (STP)/cm3polymer.cm-
Hg, and D has units of cm2/s.
Figure 5.3: Solubility coefficient of CH4 and CO2 in different co-polyimide

membranes at 35 C
90
Figure 5.4: Solubility coefficient of C3H8 and C3H6 in different co-polyimide
membranes at 35 C
Figure 5.5 shows the relative solubility coefficient of different gases for co-
polyimide membranes thermally treated at 425 C using the original PI-200 as
the control. Interestingly, after grafting CD and thermal treatment, CH4 and
CO2 have higher increases in solubility coefficient than C3H8 and C3H6. This
91
phenomenon may be due to the fact the former have smaller molecules and
lower critical temperatures than the latter. One may therefore expect that the
change in solubility coefficient has more effects on CH4 and CO2
permeabilities than C3H8 and C3H6 permeabilities.
Figure 5.5: Relative solubility coefficient of different gasses for co-polyimide

membranes thermally treated at 425 C compared to Original PI-200
Co-polyimide membranes annealed at higher temperatures have higher gas

sorption amounts over the entire pressure range studied. As can be seen in
Figure 5.6, the relative values of kD and b of all membranes thermally treated
at 425 C do not increase as much as C'H when normalizing their values to
those of the original PI-200. This observation indicates that thermal treatment
has minor effects on Henrys solubility coefficient (kD) and Langmuir affinity
constant (b) because they are mainly determined by the polymer and penetrant
chemistry as well as their mutual interactions. However, the C'H values of PI-
g-CDs have more increments than the original PI because the saturated
Langmuir sorption capacity is related to the microvoid size which is increased
with CD size and decomposition temperature. When additional free volume is
created, both C'H and solubility coefficient increase because the spaces
originally occupied by CD are converted to micro-pores. The degrees of
increases in C'H and gas solubility generally follow the order of -CD > -CD
> -CD due to different CD structures and thermal decomposition behaviors.
92
Figure 5.6: Relative kD, b, and CH value of different gasses for co-polyimide
membranes thermally treated at 425 C compared to Original PI-200
93
The PALS results give another evidence for this conclusion. As shown in
Table 5.2, the 3 values of the membranes annealed at 425 C become smaller.
This reduction is due to the cross-linking reaction at this high temperature that
results in a small pore size. On the other hand, the intensity I3 increases when
annealing at 425 C, indicating more free volume cavities are created due to
the decomposition of CDs and the conversion of the CD spaces to micro-
pores. Since the number of micro-pores increases faster than the reduction of
pore sizes, as a result, the fractional free volume (FFV) increases with
annealing temperature and CD size.
Table 5.2: Positron annihilation lifetime spectroscopy (PALS) data of

different co-polyimide membranes
Membranes I3 (%) 3 (ns) R () FFV (%)
Original PI-200 7.51 0.07 3.17 0.03 3.75 0.01 2.98 0.06
Original PI-425 6.46 0.07 2.93 0.03 3.60 0.01 2.26 0.05
PI-g--CD-200 8.04 0.07 3.10 0.03 3.70 0.01 3.08 0.06
PPM-g--CD-425 9.35 0.08 2.97 0.02 3.62 0.01 3.34 0.06
PI-g--CD-200 8.23 0.07 3.07 0.03 3.68 0.01 3.10 0.06
PPM-g--CD-425 9.94 0.15 2.89 0.02 3.56 0.01 3.39 0.08
PI-g--CD-200 8.44 0.07 3.13 0.03 3.72 0.01 3.28 0.06
PPM-g--CD-425 9.98 0.13 2.89 0.02 3.57 0.01 3.42 0.07
5.2.2 Permeability and diffusivity coefficient of co-polyimide

membranes
The permeabilities of CH4, CO2, C3H8 and C3H6 of the newly developed
membranes were measured as a function of upstream pressure at 35 C, and
the results are reported in Figure 5.7 and Figure 5.8. The CH4 permeability
decreases with an increase in upstream pressure for all membranes. This
phenomenon is consistent with the dual-mode model described by Eq. (2-14)
for the gas transport in glassy polymers. As the feed pressure p increases, the
last term of Eq. (2-14), 1 FK decreases and leads to the decline in

(1 bp)
94
permeability when other parameters are constant [20] because of the saturation
of Langmuir sorption sites.
Figure 5.7: Permeabilities of CH4 and CO2 in different co-polyimide

membranes as a function of pressure at 35 C
For highly condensable gases such as CO2, C3H6 and C3H8, a down-and-up
trend is observed for the permeability vs. pressure curves for all membranes
due to the CO2 induced plasticization except PPM-g--CD-425 and PPM-g--
CD-425. As the pressure increases, the penetrant concentration becomes
higher resulting in an increase in segmental mobility of polymer chains that
facilitates the penetrant diffusion. In this case, the increment in diffusion
coefficient surpasses the decrease in the term of 1 FK , leading to the

(1 bp)

augment in permeability at high pressures [43-44]. The membranes PPM--

CD-425 and PPM--CD-425 are not governed by this behavior because these
95
two membranes possess higher degrees of cross-linking which confine the
movement of polymer chains [4,27].
Figure 5.8: Permeabilities of C3H8 and C3H6 in different co-polyimide

membranes as a function of pressure at 35 C
Since permeability is a product of diffusivity and solubility, the diffusivity

coefficients of CH4, CO2, C3H8 and C3H6 can be calculated from the
permeability and solubility data using Eq. (2-14) and their average values are
given in Table 5.1. As expected, diffusivity coefficient is strongly dependent
on penetrant size for all samples. Propane, the largest penetrant, has the lowest
diffusivity coefficient while carbon dioxide, the smallest gas, has the highest
one.
The diffusivity coefficients of CO2 and CH4 vary in much broader ranges than
their solubility coefficients. As displayed in Table 5.1, CO2 solubility
96
coefficients of all membranes vary from 23.8 to 64.4 [cm3 (STP)/ cm3 cm-Hg]
with an augment factor of 2.7 while CO2 diffusivity coefficients range from
10.3 to 104.5 [10-8 cm2/s] with an augment factor of about 10. As a result, the
enhancement in CO2 permeability of PPM-g--CD-425 over PI-g--CD-200 is
a product of 2.4-time increase in solubility and 10.3-time increase in
diffusivity. Similar trends are observed for C3H6 and C3H8. Clearly, in addition
to annealing temperature, diffusivity contributes more significantly than
solubility to the change in permeability of PI-g-CDs. The grafting of CDs and
the decomposition of CDs at various temperatures provide us with the tools to
micro-manipulate the amount of free volume and its size as well as the
solubility and diffusivity of the resultant membranes.
Figure 5.9: Plots of permeability coefficients of CH4 and CO2 against (1+bp)-1
for different co-polyimide membranes
97
Figure 5.10: Plots of permeability coefficients of C3H8 and C3H6 against
(1+bp)-1 for different co-polyimide membranes
Eq. (2-14) implies that the plot of permeability vs. (1+bp)-1 should be linear if
the dual sorption model is applicable and all its parameters are constant and
independent of pressure. Then the two parameters DD and DH can be
calculated via the permeability vs. pressure relationship. As demonstrated in
Figure 5.9 and Figure 5.10, fairly good linear relations are observed for all the
membranes. However, the plots for the CO2 case are slightly convex to the x-
axis at high pressures due to the CO2 induced plasticization. Therefore, only
the permeability values before plasticization were used to measure the slope
and intercept.
Table 5.3 summarizes the DD, DH and F values estimated from these plots.
The variation of DH with annealing temperature is much greater than that of
DD because weaker hydrogen bonding (for the original PI) and/or a higher
98
amount of micro-pores and free volume (for the CD grafted PI) are created
when annealing the membranes at high temperatures.
Table 5.3: Diffusivity parameters of different co-polyimide membranes for

different gasses at 35 C
DD10-8 DH10-8 F
Membrane Gas
(cm2/s) (cm2/s)
CH4 19.30 0.98 0.051
CO2 97.50 3.03 0.031
Original PI-200
C3H8 0.20 0.01 0.043
C3H6 3.87 0.24 0.061
CH4 78.30 5.14 0.066
CO2 507.00 24.30 0.048
Original PI-425
C3H8 0.79 0.04 0.048
C3H6 14.60 0.91 0.062
CH4 20.30 1.05 0.051
CO2 104.00 2.56 0.025
PI-g--CD-200
C3H8 0.26 0.01 0.048
C3H6 3.22 0.24 0.076
CH4 130.60 9.74 0.075
CO2 855.00 66.90 0.078
PPM-g--CD-425
C3H8 2.63 0.13 0.051
C3H6 46.20 4.77 0.103
CH4 18.20 1.11 0.061
CO2 93.90 2.88 0.031
PI-g--CD-200
C3H8 0.26 0.01 0.046
C3H6 3.65 0.22 0.061
CH4 151.80 12.10 0.079
CO2 1036.00 78.10 0.075
PPM-g--CD-425
C3H8 3.40 0.22 0.065
C3H6 69.30 6.45 0.093
CH4 19.40 1.13 0.058
CO2 96.50 2.65 0.027
PI-g--CD-200
C3H8 0.32 0.01 0.042
C3H6 4.26 0.32 0.075
CH4 183.80 15.40 0.084
CO2 1317.00 79.20 0.060
PPM-g--CD-425
C3H8 4.30 0.25 0.059
C3H6 82.10 8.32 0.101
A comparison of F values indicates that F values for all gasses in membranes

thermally treated at 425 C are higher than those thermally treated at 200 C.
This phenomenon might be due to the fact that membranes thermally treated at
99
200 C trap gas molecules in Langmuir sites and hence hinder gas transport
across the membranes. The affinity between sorbed gas and Langmuir sites is
reduced for the membranes thermally treated at 425 C because of weaker
hydrogen bonding and/or more porous micro-pores created due to the
decomposition of CDs.
5.2.3 Permselectivity
Both solubility and diffusivity contribute to permselectivity. Table 5.4
tabulates their individual influences to the permselectivity of each membrane
for CO2/CH4 and C3H6/C3H8 separations. The diffusion selectivity fluctuates at
a wider range than the solubility selectivity for these polymers. For the
CO2/CH4 separation, the diffusion selectivity varies from 2.44 to 4.86 while
the solubility selectivity ranges from 6.55 to 8.58 for CO2/CH4.
Table 5.4: Contributions of solubility selectivity and diffusion selectivity to

the permselectivity of each membrane for CO2/CH4 and C3H6/C3H8
separations at 35 C and 1 atm
CO2/CH4 C3H6/C3H8
Membranes Solubility Diffusion Solubility Diffusion

selectivity selectivity selectivity selectivity
Original PI-200 7.76 2.82 21.87 1.28 12.83 16.42
Original PI-425 8.02 3.06 24.58 1.50 14.13 21.20
PI-g--CD-200 7.96 2.44 19.42 1.34 11.91 15.94
PPM-g--CD-425 6.75 4.63 31.27 1.12 20.15 22.62
PI-g--CD-200 7.20 2.93 21.09 1.47 10.94 16.07
PPM-g--CD-425 7.25 4.64 33.69 1.15 19.11 22.01
PI-g--CD-200 8.58 2.51 21.56 1.45 12.09 17.48
PPM-g--CD-425 6.55 4.86 31.59 1.17 20.43 23.68
All of CD grafted membranes thermally treated at 425 C have almost equal or

lower solubility selectivity than the original PI-425 membrane for CO2/CH4
and C3H6/C3H8 separations. In addition, the former has much higher diffusion
100
selectivity than the latter. As a result, diffusion selectivity plays a more
important role than solubility in determining the permselectivity for CD
grafted membranes thermally treated at 425 C.
5.2.4 Comparison of dual sorption behavior, solubility, and diffusivity

coefficient of CO2 in polymeric membranes
According to Tables 5.1, PPM-g--CD-425 (the co-polyimide grafted with -
CD and thermally treated at 425 C) shows the best solubility and diffusivity
with a CO2 solubility of 48.94 cm3 (STP)/cm3polymer atm (64.4 10-2 cm3
(STP)/cm3polymer cm-Hg) and a diffusivity of 104.5 10-8cm2/s.
Since there are many literature data available on dual mode sorption
parameters of CO2 in glassy polymeric membranes, Table 5.5 compares them
and shows that the newly developed membranes (PPM-g--CD-425 and PPM-
g--CD-425) have the higher Langmuir capacity and solubility with
comparable diffusivity. The PIM-1 membrane had the best performance in
previous literatures which has a CO2 solubility of 31.41 cm3
(STP)/cm3polymer atm and a CO2 diffusivity of 128.310-8cm2/s [20].
However, this PIM-1 membrane showed plasticization below 15 atm of CO2
feed pressure [45] and around 1 atm of C3H6 feed pressure [20], while our
membranes not only have comparable solubility and diffusivity coefficient but
also can withstand CO2 plasticization until 30 atm and C3H6 plasticization
until 7atm.
101
Table 5.5: Comparison of dual sorption behavior, solubility, and diffusivity
coefficient of CO2 in polymeric membranes
-8
Polymer Condition T, C kD C'H b S D*10 Ref.
Treat @ 300 C, 1 h 35 1.60 11.2 0.80 6.70 1.1 [2]

Treat @ 350 C, 1 h 35 1.23 30.0 0.50 11.62 1.5 [2]
6FDA-bisAPAF
Treat @ 400 C, 1 h 35 1.60 35.0 0.55 15.64 14.4 [2]
Treat @ 450 C, 1 h 35 1.75 43.0 0.45 16.50 125.0 [2]
Treat @ 180 C, 18 h 35 2.39 44.0 0.72 19.91 3.6 [6]
6FDA-DAM/DABA (3/2)
Treat @ 230 C, 18 h 35 1.95 45.9 0.64 19.75 4.4 [6]
Poly (Phenylene Oxide)
---- 35 0.85 25.9 0.24 4.95 9.9 [16]
(PPO)
PIM-1 ---- 25 2.35 106.8 0.42 31.41 128.3 [20]
Polystyrene ---- 25 0.65 7.7 0.37 2.73 ---- [28]
Poly carbonate ---- 35 0.68 18.8 0.26 4.56 ---- [28]
Cellulose Acetate ---- 30 1.43 34.2 0.20 7.06 ---- [28]
Polyacrylonitrile ---- 25 0.21 3.7 0.19 0.78 ---- [28]
6FDA-HAB Treat @ 450 C, 0.5 h 35 1.60 62.0 0.53 ---- ---- [46]
Ultem ---- 35 1.04 17.3 0.35 6.50 ---- [47]
Matrimid ---- 35 1.44 25.5 0.37 10.50 ---- [47]
Original PI-200 35 1.99 41.6 0.64 18.10 10.3
Original PI-425 35 2.31 70.6 0.63 30.06 45.7
PI-g--CD-200 35 2.10 44.8 0.58 18.52 10.4
6FDA-Durene/DABA PPM-g--CD-425 35 2.20 84.2 0.99 44.55 78.4 This

(9/1) PI-g--CD-200 35 2.21 45.3 0.55 18.56 10.7 Study
PPM-g--CD-425 35 2.42 87.8 1.10 48.95 94.7
PI-g--CD-200 35 2.33 49.7 0.56 20.47 10.2
PPM-g--CD-425 35 2.65 87.7 1.10 48.94 104.5
The units of kD is cm (STP)/cm polymer. atm, CH has units of cm (STP)/ cm3polymer, b has units of
3 3 3
atm-1, S has units of cm3 (STP)/cm3polymer.atm at 1 atm, and D has units of cm2/s.
5.3 Conclusion
We have characterized the intrinsic gas transport properties of cross-linkable
6FDA-based co-polyimides grafted by various CDs for CO2/CH4 and
C3H6/C3H8 separations. The permeability, solubility and diffusion coefficients
over a wide range of pressure have been investigated and interpreted using the
dual-mode sorption and partial immobilization model. The parameters of this
102
model have been calculated and utilized to analyze the effects of pressure and
annealing temperature on the transport behavior of these membranes. The
following conclusions can be drawn:
Sorption isotherms of all gases in these membranes exhibit the typical

dual-mode sorption behavior. The dual-mode sorption parameter C'H
increases with an increase in annealing temperature while the
parameters kD and b do not change significantly. The C'H increment is
mainly attributed to the excess free volume or micro-pores formed by
the decomposition of CDs.
Sorption of CH4 is much less than that of other gases due to its lower
condensability and weak interaction with the polymer. Comparing to
the membranes thermally treated at 200 C, the membranes thermally
treated at 425 C have higher percentages of increase in CH4 and CO2
solubility coefficients than C3H8 and C3H6 ones because the former has
lower critical temperatures and small molecule sizes.
The permeability coefficients of CH4 in all membranes decrease with
an increase in pressure. Whereas, the transport behavior of other gases
is somehow different. Some membranes show significant CO2 and
hydrocarbon-induced plasticization phenomena leading to permeability
increases of CO2, C3H6 and C3H8 with an increase in pressure.
The values of DD, DH, and F are higher for membranes thermally
treated at 425 C than those thermally treated at 200 C due to weaker
hydrogen bonding and/or more porous micro-pores created because of
the CD decomposition.
CD grafted membranes thermally treated at 425 C have almost equal
or lower solubility selectivity than the original PI-425 membrane for
CO2/CH4 and C3H6/C3H8 separations. The former also has much higher
diffusion selectivity than the latter. As a consequence, diffusion
selectivity plays a more important role than solubility in determining
the permselectivity for CD grafted membranes thermally treated at 425
C.
103
5.4 References
[1] Y. Mi, S.A. Stern, S. Trohalaki, Dependence of the gas permeability of
some polyimide isomers on their intrasegmental mobility, J. Membr. Sci.
77 (1993) 41-48.
[2] S. Kim, H.J. Jo, Y.M. Lee, Sorption and transport of small gas molecules
in thermally rearranged (TR) polybenzoxazole membranes based on 2,2-
bis(3-amino-4-hydroxyphenyl)-hexafluoropropane (bisAPAF) and 4,4-
hexafluoroisopropylidene diphthalic anhydride (6FDA), J. Membr. Sci.
441 (2013) 1-8.
[3] Y. Hirayama, T. Yoshinaga, Y. Kusuki, K. Ninomiya, T. Sakakibara, T.
Tamari, Relation of gas permeability with structure of aromatic
polyimides I, J. Membr. Sci. 111 (1996) 169-182.
[4] M. Askari, Y.C. Xiao, P. Li, T.S. Chung, Natural gas purification and
olefin/paraffin separation using cross-linkable 6FDA-Durene/DABA co-
polyimides grafted with , , and -cyclodextrin, J. Membr. Sci. 390-391
(2012) 141-151.
[5] C.H. Jung, Y.M. Lee, Gas permeation properties of hydroxyl-group
containing polyimide membranes, Macromol. Res. 16 (2008) 555-560.
[6] B. Kraftschik, W.J. Koros, J.R. Johnson, O. Karvan, Dense film
polyimide membranes for aggressive sour gas feed separations, J.
Membr. Sci. 428 (2013) 608-619.
Sci. 34 (2009) 561-580.
[8] M.R. Coleman, W.J. Koros, Isomeric polyimides based on fluorinated
dianhydrides and diamines for gas separation applications, J. Membr.
Sci. 50 (1990) 285-297.
[9] M. Askari, T.S. Chung, Natural gas purification and olefin/paraffin
separation using thermal cross-linkable co-polyimide/ZIF-8 mixed
matrix membranes, J. Membr. Sci. 444 (2013) 173-183.
aging on the gas permeability of thin 6FDA-based polyimide membranes
Part 2. with crosslinking, J. Membr. Sci. 282 (2006) 32-43.
104
212 (2003) 13-27.
[13] D.F. Sanders, Z.P. Smith, R. Guo, L.M. Robeson, J.E. McGrath, D.R.
Paul, B.D. Freeman, Energy-efficient polymeric gas separation
membranes for a sustainable future: A review, Polymer 54 (2013) 4729-
4761.
[14] D.R. Paul, Y.P. Yampolskii, Polymeric gas separation membranes. Boca
Raton: CRC, 1994.
[15] B.D. Freeman, Basis of permeability/selectivity tradeoff relations in
polymeric gas separation membranes, Macromolecules 32 (1999) 375-
380.
[16] B.J. Story, W.J. Koros, Sorption and transport of CO2 and CH4 in
chemically modified poly (phenylene oxide) , J. Membr. Sci. 67 (1992)
191-210.
[17] D. R. Paul, W. J. Koros, Effect of partially immobilizing sorption on
permeability and the diffusion time lag. J. Polym. Sci. Polym. Phys. 14
(1976) 675-685.
[18] W.H. Lin, T.S. Chung, Gas permeability, diffusivity, solubility, and
aging characteristics of 6FDA-durene polyimide membranes, J. Membr.
Sci. 186 (2001) 183193.
[20] P. Li, T.S. Chung, D.R. Paul, Gas sorption and permeation in PIM-1, J.
Membr. Sci. 432 (2013) 5057.
[21] Y. Tsujita, Gas sorption and permeation of glassy polymers with
microvoids, Prog. Polym. Sci. 28 (2003) 1377-1401.
[22] C.H. Lau, P. Li, F.Y. Li, T.S. Chung, D.R. Paul, Reverse-selective
polymeric membranes for gas separations, Prog. Polym. Sci. 38 (2013)
740-766.
105
[23] Y.C. Wang, S.H. Huang, C.C. Hu, C.L Li, K.R. Lee, D.J. Liaw, J.Y. Lai,
Sorption and transport properties of gases in aromatic polyimide
[24] F.Y. Li, Y.C. Xiao, Y.K. Ong, T.S. Chung, UV-rearranged PIM-1
polymeric membranes for advanced hydrogen purification and
production, Adv. Energy Mater. 2 (2012) 14561466.
[25] R.C. Ong, T.S. Chung, B.J. Helmer, J.S. de Wit, Characteristics of water
and salt transport, free volume and their relationship with the functional
groups of novel cellulose esters, Polymer 54 (2013) 4560-4569.
saccharide units on the gas separation performance of highly permeable
polyimide membranes, J. Membr. Sci. 415-416 (2012) 375-382.
[28] D.R. Paul, Gas sorption and transport in glassy polymers, Ber.
Bunsenges. Phys. Chem. 83 (1979) 294-302.
[29] W.J. Koros, D.R. Paul, Design considerations for measurement of gas
sorption in polymers by pressure decay, J. Polym. Sci. Polym. Phys. Ed.
14 (1976) 1903-1907.
[30] Y.C. Jean, J-P. Yuan, Q. Deng, H. Yang, Correlations between gas
permeation and free-volume hole properties probed by positron
annihilation spectroscopy, J. Polym. Sci., Polym. Phys. Ed. 33 (1995)
2365-2371.
[31] C.M. McCullagh, Z. Yu, A.M. Jamieson, J. Blackwell, J.D. McGervey,
Positron annihilation lifetime measurements of free volume in wholly
aromatic copolyesters and blends, Macromolecules 28 (1995) 6100-
6107.
[32] Y.C. Jean, Q. Deng, T.T. Nguyen, Free-volume hole properties in
thermosetting plastics probed by positron annihilation spectroscopy:
chain extension chemistry, Macromolecules 28 (1995) 8840-8844.
[33] T.C. Sandreczki, X. Hong, Y.C. Jean, Sub-glass-transition temperature
annealing of polycarbonate studied by positron annihilation
spectroscopy, Macromolecules 29 (1996) 4015-4018.
106
[34] T.A. Barbari, W.J. Koros, D.R. Paul, Gas sorption in polymers based on
bisphenol-A, J. Polym. Sci., Polym. Phys. Ed. 26 (1988) 729-744.
[35] W.J. Koros, D.R. Paul, Observations concerning the temperature
dependence of the Langmuir sorption capacity of glassy polymers, J.
Polym. Sci., Polym. Phys. Ed. 19 (1981) 1655-1656.
[36] S. Kanehashi, K. Nagai, Analysis of dual-mode model parameters for
gas sorption in glassy polymers, J. Membr. Sci. 253 (2005) 117138.
[37] V.I. Bondar, Y. Kamiya, Y.P. Yampolskii, On pressure dependence of
the parameters of the dual-mode sorption model, J. Polym. Sci., Polym.
Phys. Ed. 34 (1996) 369-378.
[38] K. Toi, G. Morel, D.R. Paul, Gas sorption and transport in
poly(phenylene oxide) and comparisons with other glassy polymers, J.
Appl. Polym. Sci. 27 (1982) 2997-3005.
[39] G.R. Ranade, R. Chandler, C.A. Plank, W.L.S. Laukhuf, Evidence of
dual mobility in sulfur dioxidepolyarylate system: an integral sorption
analysis, Polym. Eng. Sci. 25 (1985) 164-169.
[40] W.J. Koros, Sorption and transport of gases in glassy polymers, Ph.D.
dissertation, University of Texas, Austin, TX, 1977.
[41] W.J. Koros, G.K. Fleming, S.M. Jordan, T.H. Kim, H.H. Hoehn,
Polymeric membrane materials for solution-diffusion based permeation
separations, Prog. Polym. Sci. 13 (1988) 339-401.
[42] M. Wessling, I. Huisman, T.V.D. Boomgaard, C.A. Smolders, Time-
dependent permeation of carbon dioxide through a polyimide membrane
above the plasticization pressure, J. Appl. Polym. Sci. 58 (1995) 1959
1966.
[43] J.S. Chiou, D.R. Paul, Effects of carbon dioxide exposure on gas
transport properties of glassy polymers, J. Membr. Sci. 32 (1987) 195
205.
[44] A. Bos, I. Punt, H. Strathmann, M. Wessling, Suppression of gas
separation membrane plasticization by homogeneous polymer blending,
AIChE J. 47 (2001) 10881093.
[45] N. Du, M.M.D. Cin, I. Pinnau, A. Nicalek, G.P. Robertson, M.D. Guiver,
Azide-based cross-linking of Polymers of Intrinsic Microporosity (PIMs)
107
for condensable gas separation, Macromol. Rapid Comm. 32 (2011) 631-
636.
[46] Z.P. Smith, D.F. Sanders, C.P. Riberio, R. Guo, B.D. Freeman, D.R.
Paul, J.E. McGrath, S. Swinnea, Gas sorption and characterization of
thermally rearranged polyimides based on 3,3'-dihydroxy-4,4'-diamino-
biphenyl (HAB) and 2,2'-bis-(3,4-dicarboxyphenyl) hexafluoropropane
dianhydride (6FDA), J. Membr. Sci. 415-416 (2012) 558-567.
[47] T.T. Moore, W.J. Koros, Gas sorption in polymers, molecular sieves,
and mixed matrix membranes, J. Appl. Polym. Sci. 104 (2007) 4053
4059.
108
Chapter 6: Cross-linkable dual-layer hollow fiber membranes
comprising -Cyclodextrin
6.1 Introduction
The current state of gas separation membranes is largely beholden to the
notable breakthroughs in both material science and membrane fabrication
technology. The majority of gas separation membranes, in the early days, were
largely fabricated in the form of asymmetric composite flat or hollow fiber
membranes with a selective layer made from a neat material [1-3].
The asymmetric structure of hollow fibers not only could considerably reduce
the transport resistance of gas species but also offer good mechanical support.
The new features offered by the dual-layer hollow fiber membranes, in
comparison to the traditional single-layer hollow fibers, created great
potentials for material savings through substituting the expensive functional
material with inexpensive alternatives in the support layer. Nonetheless, the
fabrication of dual-layer hollow fiber membranes with desirable characteristics
is still not a trivial task and requires careful considerations from the
physicochemical properties of materials throughout the entire chain of dope
formulation, spinning and phase inversion process. However, the complexity
often arises from the simultaneous co-precipitation of two distinct materials
with different formulations. An ideal dual-layer hollow fiber membrane should
consist of an ultrathin dense functional layer and a porous substructure. In
order to withstand high feed gas pressures, the dual-layer hollow fiber
membrane must be free from any delamination at the interfacial region.
Literatures indicate that the delamination problems can often be minimized or
removed through promoting the integrity of two layers by choosing
compatible materials or using spinnerets with modified designs [2-4].
One of the key benefits of the dual-layer hollow fiber technology lies in the
great opportunities provided for the exploitation of the wide range of high
performance materials for membrane-based separation applications.
Interestingly, only a few studies have been carried out for olefin/paraffin
separation using asymmetric hollow fiber membranes.
109
Lee and Hwang [5] investigated C3H6/C3H8 separation through an asymmetric
hollow fiber membrane made of a 3,3,4,4-biphenyltetracarboxylic
dianhydride (BPDA) based co-polyimide consisting of 4, 4-oxydianiline
(ODA) as the major diamine. They obtained propylene permeance varying
from 0.8 to 1.6 GPU (1 GPU = 1 106 cm3 (STP)/ (cm2.s.cmHg) = 3.35
103 kmol/ (m2.s.kPa)) and C3H6/C3H8 selectivity of about 15 in temperature
and pressure ranges of 3070 C and 26 atm, respectively. They also reported
that the performance of mixed gas data was consistent with the theoretical
model using the single-gas permeation data. This implies no significant
plasticization effect was observed.
Krol et al. [6] reported that the asymmetric Matrimid polyimide hollow fiber
membrane has quite lower C3H6/C3H8 separation performance than the BPDA-
based polyimide membrane mentioned above. They observed significant
plasticization, especially by propylene, whereas the permeance of propane did
not change much and showed a steady-state trend. They also reported that the
plasticization resistance to propylene can also be improved by thermal
treatment of the asymmetric Matrimid membrane at 220 C.
Okamoto et al. [7] investigated C3H6/C3H8 separation through an asymmetric

hollow fiber membrane of a BPDA-based co-polyimide, of which the major
and minor component diamines were dimethyl-3,7-diaminodiphenyl-
thiophene-5,5-dioxide (DDBT) and 3,5-diaminobenzoic acid (DABA),
respectively. However, the membrane showed low separation performance for
mixed gases with propylene permeance of 1.8 GPU and C3H6/C3H8 selectivity
of 3.7 at 100 C and 1 atm. Considering the current performance of polymeric
membranes for olefin/paraffin separation, advances in materials research and
hollow fiber fabrication are urgently needed.
To our best knowledge, this is the first study on the application of cross-
linkable co-polyimide for fabrication of dual-layer hollow fiber membranes
for both CO2/CH4 and olefin/paraffin separation. The main objective is to
develop a novel approach that encompasses the synergistic advantages of the
110
cross-linking and the membrane fabrication process for high performance gas
separation applications. Indeed, this work can be regarded as an extension to
our earlier studies on Chapter 4 that reported the development of high
performance dense flat sheet membranes. Attempts were made in this study to
unravel the underlying mechanisms and engineering to scale up the flat dense
membranes to dual-layer hollow fiber membranes with desirable features.
In this respect, dual-layer hollow fiber membranes consisting of 6FDA-Durene

as an inner layer and 6FDA-Durene/DABA (9/1) grafted with -Cyclodextrin
(PI-g--CD) as an outer layer were prepared. The effects of spinning
parameters such as dope flow rate and take-up velocity and post-treatment
conditions such as annealing temperature and silicone rubber coating on
morphological and gas transport properties of hollow fiber membranes were
investigated. The resultant hollow fiber membranes were analyzed for the
separation of CO2/CH4 and C3H6/C3H8 gases. It is believed that the knowledge
and experience gained throughout this study can create greater opportunities
for further advancements of gas separation membranes.

6.2.1 Morphology of the dual-layer hollow fiber membranes
Preferably, dual-layer hollow fiber membranes shall contain a thin functional
outer layer and a fully porous supporting inner layer [8-10]. The cross-
sectional morphologies of the dual-layer hollow fiber membranes spun at
different conditions are illustrated in Figure 6.1 and the bulk and surface
morphologies of the dual-layer hollow fiber membranes spun at condition B
are demonstrated in Figure 6.2. As can be seen, Figure 6.1 verifies that the
dual-layer hollow fiber is delamination-free and exhibits a high degree of
concentricity.
111
Figure 6.1: Cross-section morphologies of the dual-layer hollow fiber
membranes spun at different conditions
The outer diameter (OD) of the fibers before a thermal treatment is in the
range of 380-710 m and the wall thickness (h) is in the range of 60-150 m.
As a result, except fibers spun from condition C, most fibers have OD/2h
ratios close to 2 and have the required resistance against the high pressure
during gas separation [10]. The bulk of the inner layer is a macrovoid-free
structure and both the outer and inner surfaces of the inner layer are highly
porous. This means that the support layer will not produce additional gas
transport resistance. According to Figure 6.2, before thermal treatment the
structure of the outer layer is asymmetric and macrovoid-free with a dense
skin layer of about 2 m or less at the outer edge. At a magnification of
10,000, the outer surface of the outer layer is apparently dense while the inner
surface is porous.
The relatively thin and dense skin at the outer layer, and the porous interface
and bulk morphologies of the hollow fibers are desirable for achieving
selective gas transport.
112
Figure 6.2: Bulk and surface morphologies of the dual-layer hollow fiber
membranes spun at condition B.
6.2.2 Effect of take-up velocity on gas separation performance

The gas transport properties of the fibers before and after silicone rubber
coating are summarized in Tables 6.1 and 6.2, respectively. To simplify the
sample identification, samples were named according to the spinning
condition, final treated temperature (25 for untreated fibers) and condition.
BSRC denotes before silicon rubber coating, while ASRC means after
silicon rubber coating. For example, A-350-ASRC is for the fiber spun at
condition A, thermally treated temperature of 350 C and then silicon rubber
coating. For samples A, B and C when fibers were spun with increasing take
up speed from 4 to 11.8 m/min, their permeances decrease as the take-up
velocity increases for all the tested gases as shown in the top of Figures 6.3
and 6.4. This reduction phenomenon is most obvious for the fibers before
silicon rubber coating. After silicon rubber coating, the degree of reduction is
113
much reduced, especially for samples spun at a velocity between 7.4 to 11.8
m/min.

at different conditions before silicon rubber coating
Heat treatment Hollow fiber Permeance (GPU) Selectivity
Condition
temperature ( C) code CO2 CH4 C3H6 C3H8 CO2/CH4 C3H6/C3H8
Precursor A-25 -BSRC 283 186 125 72 1.53 1.74
A 350 A-350-BSRC 202 69 81 24 2.93 3.3
400 A-400-BSRC 168 19 66 8.3 8.96 7.95
Precursor B-25 -BSRC 138 22 75 11 6.22 6.52
B 350 B-350-BSRC 118 12 56 7.0 9.53 8.03
400 B-400-BSRC 103 6.3 49 4.3 16.3 11.5
Precursor C-25 -BSRC 107 27 38 8.2 3.97 4.62
C 350 C-350-BSRC 104 15 33 6.1 6.96 5.54
400 C-400-BSRC 99 6.4 27 2.5 15.4 10.8
Precursor D-25 -BSRC 194 36 93 20 5.39 4.72
D 350 D-350-BSRC 161 21 83 13 7.55 6.27
400 D-400-BSRC 134 8.6 73 7.0 15.5 10.5
Precursor E-25 -BSRC 324 191 130 58 1.69 2.24
E 350 E-350-BSRC 225 92 103 40 2.45 2.55
400 E-400-BSRC 205 27 86 17 7.54 5.12

at different conditions after silicon rubber coating
Heat treatment Hollow fiber Permeance (GPU) Selectivity
Condition
temperature ( C) code CO2 CH4 C3H6 C3H8 CO2/CH4 C3H6/C3H8
Precursor A-25 -ASRC 155 18 77 9.2 8.67 8.43
A 350 A-350-ASRC 120 11 57 5.9 10.5 9.56
400 A-400-ASRC 99 9.0 48 3.9 11.1 12.3
Precursor B-25 -ASRC 95 6.6 45 3.5 14.3 12.9
B 350 B-350-ASRC 90 5.3 32 2.4 16.9 13.3
400 B-400-ASRC 84 4.5 29 1.9 18.7 15.3
Precursor C-25 -ASRC 89 5.4 22 1.7 16.6 13.2
C 350 C-350-ASRC 86 4.6 19 1.3 18.7 14.3
400 C-400-ASRC 81 4.1 17 1.1 19.9 15.6
Precursor D-25 -ASRC 129 9.7 66 5.8 13.3 11.4
D 350 D-350-ASRC 103 6.4 56 4.6 16.1 12.1
400 D-400-ASRC 96 5.5 44 3.3 17.6 13.1
Precursor E-25 -ASRC 148 36 89 14 4.12 6.21
E 350 E-350-ASRC 126 17 74 9.5 7.26 7.84
400 E-400-ASRC 112 13 51 5.2 8.64 9.96
114
Interestingly, before silicon rubber coating, as the take-up velocity is varied
from 4 to 7.4 m/min, higher selectivities are obtained, but a further increase in
take-up velocity to 11.8 m/min results in lower selectivities. However, after
silicon rubber coating, both CO2/CH4 and C3H6/C3H8 selectivities increase as
the take-up velocity increases.
a) b)
c) d)
Figure 6.3: CO2/CH4 separation performance of dual-layer hollow fiber
membranes at different conditions.
Two hypotheses are proposed to account for the influence of various take-up
velocities on gas permeances and selectivities. Firstly, the elongational draw
ratios corresponding to the take-up velocity of 4, 7.4, and 11.8 m/min are 3.3,
6.7, and 16.7, respectively. This means that spinning at a higher take-up
115
velocity induces a higher elongational stress on the extruded fibers before
entering and precipitation in the external coagulant bath. Even though the
outer surface of the fibers is quite dense and gas transport phenomenon is
controlled by the solution-diffusion mechanism, Knudsen diffusion via the
larger pores is commonly present. The higher elongational stress may decrease
the amount of Knudsen pores in the dense outer skin. To validate this claim,
the gas transport properties of hollow fiber membranes were quantified after
silicon rubber coating. According to Table 6.2, the gas permeances of the
hollow fibers after silicone rubber coating decrease as a result of the effective
sealing of Knudsen pores and large pores. The effect of shear stress on hollow
fiber membranes during the spinning process has been investigated by Wang
and Chung [11]. They have reported that increasing shear stress during
spinning changes the surface porosity of the membranes. The results presented
here suggest that the change in elongational stress may introduce a similar
change in the surface porosity of the membranes.
Another hypothesis to explain the observed phenomenon is the change in

polymer chain packing which is induced by the elongational stress. In an
earlier work by Peng and Chung [12], the influence of elongational draw ratio
on the O2/N2 transport properties of Torlon single-layer hollow fibers was
investigated. They found that the O2/N2 selectivity shows an up-and-down
trend with increasing draw ratio. The first increase in O2/N2 selectivity is
attributed to the improved elongation- induced chain packing which makes
more mono disperse space. The reverse trends are observed after an optimal
draw ratio of about 10, and this can be according to the creation of minor
defects from the overstretching of polymer chains. Compared to our present
work, a similar up-and-down relationship between CO2/CH4 and C3H6/C3H8
selectivities and draw ratio (or take-up velocity) is observed for the fibers
before silicon rubber coating. After silicon rubber coating, the selectivities
increase with increasing the take-up velocity. This phenomenon implies that
silicone rubber may seal the Knudsen pores and minor defects and recover the
selectivities. This improvement is more obvious for samples spun with a
higher take-up velocity that may have more defects because of the
overstretching of polymer chains.
116
a) b)
c) d)
Figure 6.4: C3H6/C3H8 separation performance of dual-layer hollow fiber
membranes at different conditions.
6.2.3 Effect of outer-layer dope flow rate on gas separation performance

A comparison of gas separation performance for samples B, D and F which
were spun with a decrease in outer-layer dope flow rate from 0.5 to 0.2 m/min,
the permeances of all gases increase with a decrease in outer-layer dope flow
rate as summarized in Tables 6.1 and 6.2 and illustrated in Figures 6.3 and 6.4
for CO2/CH4 and C3H6/C3H8 separation, respectively. This is due to a thinner
skin layer that leads to reduce mass transfer resistance of the outer layer. It is
also obvious that the selectivities decrease with a reduction in outer-layer dope
flow rate. This can be caused by the formation of defects at the outer surface
117
of the outer layer during the spinning process. Since there is a velocity
difference between the inner-layer and the outer-layer dopes and the inner-
layer dope velocity at the exit of the spinneret is higher than the outer-layer
dope, it may suddenly accelerate the outer-layer dope velocity and apply
additional stresses to the thin-skinned outer layer and hence induce defects [8].
Therefore, a lower outer-layer dope flow rate does not necessarily lead to the
formation of an ultrathin dense-selective layer and in this work the optimal
outer-layer dope flow rate is 0.5 ml/min.
6.2.4 Effect of thermal treatment and silicon rubber coating on gas

separation performance
In previous Chapters using dense flat membranes made of the same co-
polyimide (PI-g--CD), cross-linking was found to commence at about 350
C. The partially cross-linked dense films treated at 400 or 425 C for 2 h
show reasonably flexible and impressively high permeabilities. The same
conditions were employed to thermally treat these dual-layer hollow fiber
membranes but the resultant hollow fibers were quite dense. For polymers
containing carboxylic groups, the thermal treatment can result in two possible
phenomena: thermally induced cross-linking and densification. Although there
have been some studies reported on the effect of thermal treatment
temperature and duration on the degree of cross-linking [13-17], relatively less
studies have been devoted to understand the influence of membrane thickness
on the degree of thermal cross-linking and densification especially for the
hollow fiber membranes. Compared to conventional dense flat-sheet
membranes, hollow fiber membranes have a much thinner selective skin layer
and asymmetric structure, which may facilitate the higher heat transfer during
thermal treatment and result in a higher degree of cross-linking and excessive
densification of the dense selective layer, thus reducing the permeance of the
hollow fibers [18]. Therefore, 350 or 400 C for 1 h were employed in this
work and the effects of thermal treatment on fiber morphology and separation
performance were investigated.
As shown in Figure 6.5, thermal treatment temperature has a significant effect

on fiber morphology. The diameter of the fibers shrinks around 15 and 21%
118
for fibers treated at 350 and 400 C, respectively. In comparison with 400 C
treated fibers as illustrated in Figures 6.3 and 6.4, 350 C treated fibers show
lower selectivities and higher permeances close to those of pristine fibers,
suggesting a lower degree of cross-linking. A supportive evidence of this
claim can be verified by the fiber dissolution test [18] in hot NMP at 110 C;
the 350 C treated fibers were dissolved much faster than 400 C treated fibers.
A higher selectivity observed in the 400 C treated fibers suggests that the
decrease in permeance can be attributed to skin layer densification occurring
between 350 and 400 C thermal treatment. This densification may mitigate
defects in the skin layer of the outer layer and decrease the Knudsen pores. As
a result, the selectivities of the 400 C treated hollow fibers after silicon rubber
coating are approaching the selectivities of the 400 C treated dense film as
listed in Table 6.3.
Figure 6.5: Comparison of cross-sectional morphologies of the fiber

membranes spun at condition B at different thermal treatment temperature
It is well known that silicone rubber coating can decrease the effect of defects
on separation performance. A comparison of permeance before and after
silicon rubber coating in Figure 6.3 and Figure 6.4 indicates that the effect of
119
coating is more pronounced for precursor and low-temperature thermally
treated membranes than high-temperature thermally treated membranes. After
the coating, the selectivities increase significantly for precursor fibers and the
fibers treated at 350 C. However, for the fibers thermally treated at 400 C,
the increase in selectivities is small but the reduction in permeance still
remains large. This indicates that fibers treated at 400 C have a less amount of
defects and Knudsen pores in the selective skin, but the silicon rubber may
create additional transport resistance. In addition to depositing it on top of the
outer surface, the silicone rubber solution could also penetrate into the top
surface region and fill out part of micropores [19, 20]. This intrusion may
significantly reduce permeance and this effect is more significant for hollow
fibers thermally treated at lower temperatures than at high temperatures.
Table 6.3: Intrinsic gas transport properties of flat dense PI-g--CD co-
polyimide membranes heat treated at different temperatures a [16]
Permeability (Barrer b) Selectivity
Membranes
CO2 CH4 C3H6 C3H8 CO2/CH4 C3H6/C3H8
PI-g--CD-200 C 238 12 10 0.67 20.67 15.05
PI-g--CD-350 C 593 37 18 1.4 15.91 12.15
PI-g--CD-400 C 772 41 81 5.7 18.65 13.88
a
. Pure gas permeation tests done at 35 C and 1.01 106 Pa (10 atm)
b
. Barrer = 1 10-10 (cm3 (STP) cm)/(cm2 s cmHg)
According to Tables 6.1 and 6.2, without silicon rubber coating, sample D-
400-BSRC (spun from condition D and thermally treated at 400 C) show the
best CO2/CH4 separation with a CO2 permeance of 130 GPU and a selectivity
of 15.5. After silicon rubber coating, sample C-400-ASRC (spun from
condition C and thermally treated at 400 C) shows the best selectivity of 20
and a CO2 permeance of 82 GPU.
6.2.5 CO2 plasticization and mixed gas permeation experiments

As mentioned earlier, plasticization is unpleasant in gas separation processes.
The newly developed hollow fibers before and after thermal treatments were
tested under various CO2 pressures in order to explore their potential in
aggressive and harsh environments. Figure 6.6 illustrates the effect of thermal
treatment temperature and silicon rubber coating on CO2-plasticization
120
behavior for fibers spun from condition B. Without thermal treatment, fibers
with and without silicon rubber coating, exhibit plasticization at a CO2
pressure of around 7 atm. By increasing the thermal treatment temperature, the
plasticization pressure increases. Its obvious that cross-linking reactions at
high temperatures enhance anti-plasticization properties. The 400 C thermally
treated fibers display anti-plasticization characteristics up to 25 atm due to a
higher degree of cross-linking reaction.
Figure 6.6: Plasticization behavior of fiber membranes spun at condition B at

different thermal treatment temperature before and after silicon rubber coating
To explore potential applications of natural gas purification, it is important to

achieve a high selectivity for CO2/CH4 mixed gas experiments. In this study,
mixed gas experiments were conducted using a 50:50 feed mixture of
CO2/CH4 with a total feed pressure of 20 atm at room temperature. Table 6.4
summarizes the gas permeance and selectivity of dual layer hollow fiber
membranes spun at condition B, thermally treated at 400 C and then coated
with the silicon rubber (B-400-ASRC). The pure gas data listed in the table are
different from those reported in Table 6.2 due to the different upstream and
downstream pressures used in the tests (the downstream is connected to the
vacuum and the selectivities are higher). As can be seen in Table 6.4, the
permeances of B-400-ASRC membranes in mixed gas tests slightly decrease
121
while the selectivity is almost the same as those in pure gas tests. The mixed
gas results signify that the developed fibers display a good CO2 anti-
plasticization behavior under a high pressure which can be a good candidate
for industrial applications.
Table 6.4: Mixed gas permeance and selectivity of the dual-layer hollow fiber
membranes spun at condition B, thermally treated at 400 C after silicon
rubber coating
Permeance Permeance
Membranes Selectivity Selectivity
(GPU) (GPU)
B-400-ASRC 97 3.4 28.63 77 2.7 28.26
a
Measured at 10 atm and room temperature
b
Measured at 20 atm and room temperature
6.2.6 Comparison of C3H6/C3H8 separation performance of hollow fiber

membranes
According to Tables 6.1 and 6.2, without silicon rubber coating, sample D-
400-BSRC (spun from condition D and thermally treated at 400 C) show the
best C3H6/C3H8 separation with a C3H6 permeance of 73 GPU and a selectivity
of 10.5. After silicon rubber coating, sample B-400-ASRC (spun from
condition B and thermally treated at 400 C) shows the best selectivity of 15.3
and a C3H6 permeance of 29 GPU.
There is a considerable amount of literature data presented in Table 6.5 for the
application of various single layer hollow fiber polyimide membranes for
olefin/paraffin separation. Compared to them, the newly developed
membranes have a higher permeance with moderate selectivities. The best
performance of the previous studies in Table 6.5 has a C3H6 permeance of 51
GPU and a C3H6/C3H8 selectivity of 12. However, this BPDA-DDBT/DABA
polyimide was pyrolyzed at 600 C and it is a carbon membrane [7], while our
membranes have comparable separation performance but thermally treated at
lower temperatures.
122
Table 6.5: C3H6/C3H8 separation performance of polyimide and co-polyimide
hollow fiber membranes
Polymer Condition T/P (C/atm) PC3H6 (GPU) C3H6/C3H8 Ref.
6FDA/BPDA (1/1) -DDBT ---- 100/1.0 (3.6) (15) [21]
---- 50/1.0 (1) (20) [21]
---- 35/1.0 (0.51) (20) [21]
---- 35/1.0 0.6 24 [21]
Silicon Rubber Coating 100/1.0 (2.55) (17) [21]
Silicon Rubber Coating 35/1.0 0.5 33 [21]
Silicon Rubber Coating 50/1.0 (0.8) (26.6) [21]
BPDA-DDBT/DABA ---- 100/1.0 (2.2) (5.7) [21]
6FDA-BAAF ---- 20/4.0 0.3 24 [21]
6FDA/BPDA (1/1) -DDBT ---- 100/1.0 (2.9) (12) [22]
Pyrolyzed @ 500 C 100/1.0 (11) (12) [22]
Pyrolyzed @ 520 C 100/1.0 (18) (14) [22]
Pyrolyzed @ 550 C 100/1.0 (27) (14) [22]
Pyrolyzed @ 600 C 100/1.0 (21) (14) [22]
Pyrolyzed @ 650 C 100/1.0 (12) (19) [22]
Pyrolyzed @ 700 C 100/1.0 (0.74) (15) [22]
Polyimide (UBE, Japan) ---- 30/2.0 0.843 14.5 [5]
---- 30/3.0 0.870 14.7 [5]
---- 30/4.0 0.902 14.5 [5]
---- 30/5.0 0.925 15.1 [5]
---- 70/2.0 1.558 16.6 [5]
---- 70/3.0 1.585 15.8 [5]
---- 70/4.0 1.622 15.9 [5]
---- 70/5.0 1.654 16.1 [5]
Matrimid ---- 27/1.0 0.10 8 [6]
---- 27/2.0 0.14 14 [6]
---- 27/3.0 0.18 24 [6]
---- 27/3.5 0.22 29 [6]
Treatment @ 150 C, 5 min 27/1.0 0.10 ---- [6]
Treatment @ 150 C, 5 min 27/2.0 0.11 ---- [6]
Treatment @ 150 C, 5 min 27/3.0 0.12 ---- [6]
Treatment @ 150 C, 5 min 27/3.5 0.17 ---- [6]
Treatment @ 250 C, 5 min 27/1.0 0.08 ---- [6]
Treatment @ 250 C, 5 min 27/2.0 0.07 ---- [6]
Treatment @ 250 C, 5 min 27/3.0 0.07 ---- [6]
Treatment @ 250 C, 5 min 27/3.5 0.07 ---- [6]
BPDA-DDBT/DABA ---- 100/1.0 (1.8) (3.7) [7]
Pyrolyzed @ 600 C 100/1.0 51 12 [7]
PI-g--CD/6FDA-Durene A-400-ASRC 25/3.5 48 12.3
B-400-BSRC 25/3.5 49 11.5
B-400-ASRC 25/3.5 29 15.3 This
C-400-ASRC 25/3.5 17 15.6 Study
D-400-BSRC 25/3.5 73 10.5
D-400-ASRC 25/3.5 44 13.1
The data in parentheses were obtained from mixed-component systems with a feed composition of
50/50 mol% and 1 GPU=10-6 cm3/ (cm2.s.cmHg)
123
6.3 Conclusion
In this work cross-linkable co-polyimide dual layer hollow fiber membranes
with high gas separation performance for CO2/CH4 and C3H6/C3H8 have been
fabricated. The effects of spinning and thermal treatment conditions on the gas
separation performance of these membranes were investigated.
Experimental results show that permeances decrease with increasing take-up

velocity. Before silicon rubber coating, higher selectivities can be obtained by
increasing take-up velocity from 4 to 7.4 m/min. A further increase in take-up
velocity to 11.8 m/min results in lower selectivities due to defect formation.
On the other hand, after silicon rubber coating, the CO2/CH4 and C3H6/C3H8
selectivities increase as the take-up velocity increases. In addition, permeances
increase with a reduction in outer-layer dope flow rate due to the thinner skin
layer and lower transport resistance. However, selectivities decrease when the
outer-layer dope flow rate decreases owing to defect formation.
For as-spun (i.e., pristine and without heat treatment) fibers, the optimal
spinning condition was found to have a take-up velocity of 7.4 m/min with an
outer layer dope flow rate of 0.5 ml/min. The resultant fiber shows a CO2/CH4
selectivity of 6.22 and 14.3 before and after silicon rubber coating,
respectively and also C3H6/C3H8 selectivity of 6.52 and 12.9 before and after
silicon rubber coating, respectively.
Thermal treatment temperature plays a significant role on fiber morphology.

Fiber diameters shrunk after treatment. The 350 C treated fibers show higher
permeances and lower selectivities close to those of untreated fibers, while the
400 C treated fibers show higher selectivities but lower permeances due to the
skin densification and a higher degree of cross-linking. As a result, the silicon
rubber coating has a less effect on the 400 C treated fibers than the 350 C
treated fibers.
For CO2/CH4 separation, without silicon rubber coating, sample D-400-BSRC

show the best CO2/CH4 separation with a CO2 permeance of 130 GPU and a
selectivity of 15.5. After silicon rubber coating, sample C-400-ASRC shows
124
the best selectivity of 20 and a CO2 permeance of 82 GPU. For C3H6/C3H8
separation, without silicon rubber coating, sample D-400-BSRC show the best
C3H6/C3H8 separation with a C3H6 permeance of 73 GPU and a selectivity of
10.5. After silicon rubber coating, sample B-400-ASRC shows the best
selectivity of 15.3 and a C3H6 permeance of 29 GPU.
125
6.4 References
[1] S.S. Hosseini, N. Peng, T.S. Chung, Gas separation membranes
developed through integration of polymer blending and dual-layer
hollow fiber spinning process for hydrogen and natural gas enrichment,
J. Membr. Sci. 349 (2010) 156-166.
[2] T. Kawai, Y.M. Lee, A. Higuchi, K. Kamide, Separation of mixed gases
through porous polymeric membranes, J. Membr. Sci. 126 (1997) 67-76.
[3] D.F. Li, T.S. Chung, R. Wang, Morphological aspects and structure
control of dual-layer asymmetric hollow fiber membranes formed by a
simultaneous co-extrusion approach, J. Membr. Sci. 243 (2004) 155-175.
[4] N. Widjojo, T.S. Chung, W.B. Krantz, A morphological and structural
study of Ultem/P84 co-polyimide dual-layer hollow fiber membranes
with delamination-free morphology, J. Membr. Sci. 294 (2007) 132-146.
[5] K.-R. Lee, S.-T. Hwang, Separation of propylene and propane by
polyimide hollow fiber membrane module, J. Membr. Sci. 73 (1992) 37-
45.
[6] J.J. Krol, M. Boerrigter, G.H. Koops, Polyimide hollow fiber gas
separation membranes: preparation and suppression of plasticization in
propane/propylene environments, J. Membr. Sci. 184 (2001) 275-286.
[7] K. Okamoto, S. Kawamura, M. Yoshino, H. Kita, Olefin/paraffin
separation through carbonized membranes derived from an asymmetric
polyimide hollow fiber membrane, Ind. Eng. Chem. Res. 38 (1999)
4424-4432.
[8] F.Y. Li, Y. Li, T.S. Chung, H. Chen, Y.C. Jean, S. Kawi, Development
and positron annihilation spectroscopy (PAS) characterization of
polyamide imide (PAI)-polyethersulfone (PES) based defect-free dual-
layer hollow fiber membranes with an ultrathin dense-selective layer for
gas separation, J. Membr. Sci. 378 (2011) 541-550.
[9] S.A. McKelvey, D.T. Clausi, W.J. Koros, A guide to establishing hollow
fiber macroscopic properties for membrane applications, J. Membr. Sci.
124 (1997) 223-232.
[10] O.M. Ekiner, G. Vassilatos, Polyaramide hollow fibers for
hydrogen/methane separation-Spinning and properties, J. Membr. Sci. 53
(1990) 259-273.
126
[11] R. Wang, T.S. Chung, Determination of pore sizes and surface porosity
and the effect of shear stress within a spinneret on asymmetric hollow
fiber membranes. J. Membr. Sci. 188 (2001) 29-37.
[12] N. Peng, T.S. Chung, The effects of spinneret dimension and hollow
fiber dimension on gas separation performance of ultra-thin defect free
Torlon hollow fiber membranes, J. Membr. Sci. 310 (2008) 455-465.
6056.
[14] J.D. Wind, D.R. Paul, W.J. Koros, Natural gas permeation in polyimide
diaminobutane (DAB) dendrimers, J. Membr. Sci. 238 (2004) 153163.
(2012) 141-151.
[18] C.C. Chen, S.J. Miller, W.J. Koros, Characterization of thermally cross-
linkable hollow fiber membranes for natural gas separation, Ind. Eng.
Chem. Res. 2012, No. 10.1021/ie2020729.
[19] J. Peter, K.-V. Peinemann, Multilayer composite membranes for gas
separation based on crosslinked PTMSP gutter layer and partially
crosslinked Matrimid 5218 selective layer, J. Membr. Sci. 340 (2009)
62-72.
[20] T.S. Chung, Calculation of the intrusion depth and its effects on
microporous composite membranes, Sep. Purif. Technol. 12 (1997) 17
23.
[21] M. Yoshino, S. Nakamura, H. Kita, K.-I. Okamoto, N. Tanihara, Y.
Kusuki, Olefin/paraffin separation performance of asymmetric hollow
127
fiber membrane of 6FDA/BPDADDBT copolyimide, J. Membr. Sci.
212 (2003) 13-27.
[22] M. Yoshino, S. Nakamura, H. Kita, K.-I. Okamoto, N. Tanihara, Y.
Kusuki, Olefin/paraffin separation performance of carbonized
membranes derived from an asymmetric hollow fiber membrane of
6FDA/BPDADDBT copolyimide, J. Membr. Sci. 215 (2003) 169-183.
128
Chapter 7: Thermal cross-linkable co-polyimide/ZIF-8 mixed
matrix membranes
7.1 Introduction
In this Chapter, we aimed to synergistically combine the strengths of ZIF-8
and cross-linkable polyimide and to molecularly design MMMs for natural gas
purification and olefin/paraffin separation. 6FDA-Durene/DABA co-
polyimides with different molar ratios of Durene to DABA diamine and nano-
size ZIF-8 particles were synthesized for the fabrication of MMMs. The
6FDA-Durene polyimide, 6FDA-Durene/DABA (9/1) co-polyimide, and
6FDA-Durene/DABA (7/3) co-polyimide (referred as PI, CPI (9/1), and CPI
(7/3), respectively) were chosen because of their impressive performance for
CO2/CH4 and C3H6/C3H8 separation [1-8]. The effects of (1) diamine ratio
(i.e., the ratio of Durene to DABA monomers in the co-polyimide structure),
(2) annealing temperature (i.e., different degrees of cross-linking) and (3)
plasticization phenomenon on membrane separation performance would be
systemically investigated. Permeation properties of thermally treated 6FDA-
Durene/DABA-ZIF-8 MMMs with different ZIF-8 loadings would be
conducted using pure CO2, CH4, C3H6, C3H8, and mixed CO2/CH4. To our best
knowledge, this is the first study on the application of thermal cross-linkable
co-polyimide/ZIF-8 mixed matrix membranes for CO2/CH4 and olefin/paraffin
separation.

7.2.1 Characterizations
The weight losses of pure ZIF-8 and CPI (9/1) with different ZIF-8 loadings as
a function of temperature in air atmosphere are shown in Figure 7.1. From 300
C to 450 C, there is a weight loss of less than 5% in the pure ZIF-8
spectrum, which may be attributed to residual solvents or reactant molecules
trapped in the nano-crystals during synthesis and post-treatment, as reported
by Song et al. [9]. As seen from the TGA curves, 6FDA-Durene/DABA (9/1)
(CPI (9/1)), ZIF-8 and their MMMs all exhibit excellent thermal stability up to
450 C in air atmosphere. The CPI (9/1) shows two stages of decomposition.
129
Figure 7.1: Thermo gravimetric analyses of pure CPI (9/1), ZIF-8 and CPI
(9/1) - ZIF-8 MMMs under air atmosphere
The first stage of weight loss of around 10 % is detected at 350-450 C, which

is caused by the decarboxylation of the DABA moiety in the co-polyimide [3-
5]. In the second stage, additional 90 % weight loss occurs from 450 to 600 C
mainly due to the decomposition of CPI (9/1). The TGA spectra of CPI (9/1)
ZIF-8 MMMs also have these two stages of decomposition. Since CPI (9/1) is
completely decomposed in the second stage and only zinc oxide remains [10],
the exact loading of ZIF-8 in the membrane can be calculated according to the
130
stoichiometric amount of remaining zinc oxide. Table 7.1 tabulates the exact
ZIF-8 loadings in various MMMs and show that there are no much differences
between the pre-weight and TGA methods.
Table 7.1: Single gas separation performance of CPI (9/1) ZIF-8 MMMs
with different ZIF-8 loading (Annealed at 200 C during sample preparation)
Tg Permeability (Barrer) Selectivity

Actual ZIF-8
Membrane ID
Loading (wt %) (C) CO2 CH4 C3H6 C3H8 CO2/CH4 C3H6/C3H8
CPI(9/1) - 0 wt% ZIF-8 0 411 256 13.1 13.2 1.13 19.51 11.68
CPI(9/1) - 5 wt% ZIF-8 4.6 261 13.2 14.9 1.14 19.72 13.12
CPI(9/1) - 10 wt% ZIF-8 8.4 412 301 15.7 17.1 1.15 19.15 14.86
CPI(9/1) - 15 wt% ZIF-8 13.9 351 17.7 18.2 1.14 19.85 15.97
CPI(9/1) - 20 wt% ZIF-8 20.4 414 392 19.2 20.5 1.14 20.45 18.08
CPI(9/1) - 30 wt% ZIF-8 31.2 602 30.2 29.7 1.28 19.94 23.18
CPI(9/1) - 40 wt% ZIF-8 41.9 416 779 37.4 47.3 1.73 20.85 27.38
The glass transition temperatures (Tg) of all membranes were determined by

differential scanning calorimetry (DSC). The PI shows a Tg of 417 C which is
consistent with the value in the literature [11]. Tg of the membranes decreases
with increasing DABA moiety in the polymer structure. The glass transition
temperatures of 6FDA-Durene/DABA (9/1) (CPI (9/1)) and 6FDA-
Durene/DABA (7/3) (CPI (7/3)) are around 411 and 392 C, respectively. As
can be seen in Table 7.1, Tg also increases with ZIF-8 loading. A similar
phenomenon on Tg rise have been reported on MMMs consisting of zeolite
[12] and MgO [13] particles due to chain rigidification. Clearly, there are
changes in polymer packing structure after adding DABA and ZIF-8 nano-
particles into the matrix polymer.
Dynamic light scattering (DLS) was employed to measure the apparent

particle size distribution in CPI (9/1)/ZIF-8 suspensions. DLS can be a direct
evidence of particle agglomeration if it exists [14]. Three suspensions were
prepared in different conditions and kept in room temperature for 24 hrs prior
to characterization. The first solution consisted of as-synthesized ZIF-8
131
particles suspending in methanol. The as synthesized ZIF-8 particles were
collected before centrifugation and washing. The second solution was made of
ZIF-8 particles suspending in NMP. The ZIF-8 particles were obtained after
centrifugation, washed by NMP, and re-suspended in NMP. The third one was
a solution consisting of ZIF-8 and CPI (9/1) in NMP. As shown in Figure 7.2,
the particle size distribution shifts slightly toward big particles after
redistributing ZIF-8 in NMP and then CPI (9/1)/NMP solutions. However, the
shifts are not big for dense membranes. The aggregation of particles after
washing with NMP can be due to changing the media. After changing the
media, particles have more tendencies to aggregate in NMP than Methanol due
to the different energy of interaction between the particles and media.
Figure 7.2: ZIF-8 particle size distribution
Figure 7.3 shows the XRD patterns of pure CPI (9/1), ZIF-8, and CPI (9/1) -
ZIF-8 MMMs between 5 and 35 in order to determine their crystalline
structures. The XRD pattern of pure CPI (9/1) shows a broad peak from 8 to
22, which is a characteristic of amorphous structure and is consistent with the
reported XRD pattern for CPI (9/1) in the literature [4]. The diffraction
patterns of CPI (9/1) - ZIF-8 MMMs display intense, characteristic ZIF-8
crystalline peaks matching extremely well with pure ZIF-8 and CPI (9/1)
132
patterns, proving that ZIF-8 and CPI (9/1) structures are present in the
membranes. In addition, the intensity of characteristic peaks of ZIF-8
increases with increasing ZIF-8 loading in MMMs. Figure 7.3 also shows the
XRD patterns of CPI (9/1) - 20 wt% ZIF-8 MMMs annealed at different
temperatures. The annealing temperature does not affect too much on the ZIF-
8 structure and the characteristic peaks of ZIF-8 nano-particles are still
detectable in the XRD spectra.
Figure 7.3: XRD pattern of pure CPI (9/1), ZIF-8, and CPI (9/1) - ZIF-8
MMMs with various ZIF-8 loading and CPI (9/1) 20 wt% ZIF-8 MMMs at
different annealing temperatures
Figure 7.4 displays the FESEM images of ZIF-8 nano-particles and the cross-
section morphology of MMMs with different ZIF-8 loadings. The ZIF-8 nano-
particles are dispersed uniformly in the polymer matrix and there is no
evidence of phase separation even at the highest loading of 40 wt%. As the
ZIF-8 loading increases to around 40 wt%, the continuous phase gradually
shifts from CPI (9/1) to ZIF-8. No interfacial gaps are observed between the
polymeric phase and the nano-particle phase, indicating good compatibility
between CPI (9/1) and ZIF-8. The good compatibility may be resulted from
two factors; namely, (1) the small size of ZIF-8 nano-particles of less than 80
133
nm and (2) the use of the same solvent (i.e., NMP) for both ZIF-8 and CPI
(9/1) solutions. As a result, ZIF-8 nano-particles remain in a suspension state
before being casted to form membranes.
Figure 7.4: FESEM morphologies of CPI (9/1) ZIF-8 MMMs with various
ZIF-8 loadings without annealing
Figure 7.5: SEM-EDX mapping for Zn from the cross-section of CPI (9/1)
ZIF-8 MMMs with various ZIF-8 loadings annealing @ 200 C
134
Figure 7.5 reconfirms the uniform dispersion of ZIF-8 nano-particles in
MMMs with different ZIF-8 loadings by mapping the Zn element with the
assistance of the energy-dispersive X-ray spectroscopy (EDX). Since zinc
element exists only in ZIF-8 nano-particles, the uniform distribution of the Zn
element noticeably confirms the uniform dispersion of ZIF-8 nano-particles in
the polymer matrix.
Figure 7.6 shows FESEM cross-section images of CPI (9/1) 20 wt% ZIF-8
MMMs after annealing at different temperatures. The MMM structure
apparently transforms to a denser structure with an increase in annealing
temperature due to chain densification and cross-linking reaction.
Figure 7.6: FESEM morphologies of CPI (9/1) 20 wt% ZIF-8 MMMs at

different annealing temperatures
135
7.2.2 Pure gas permeation experiments
7.2.2.1 Effect of ZIF-8 loading on gas separation performance
Figure 7.7 shows the gas permeability and ideal selectivity of CPI (9/1) - ZIF-
8 MMMs for CO2/CH4 and C3H6/C3H8 separation as a function of ZIF-8
loading. At a low loading of 5% (w/w) ZIF-8, the gas separation performance
of both pure CPI (9/1) and MMM are comparable for CO2/CH4 separation,
while the MMM shows a higher C3H6/C3H8 selectivity than the pure CPI (9/1).
At a high loading of 40% (w/w) ZIF-8, a significant higher permeability is
observed for the CPI (9/1) 40 wt% ZIF-8. Compared to the pure CPI (9/1)
membrane, an 185% increase in CH4 permeability from 13.1 to 37.4 Barrers,
204% for CO2 from 256 to 779 Barrers, 258% for C3H6 from 13.2 to 47.3
Barrer and 53.1% for C3H8 from 1.13 to 1.73 Barrers are observed.
Figure 7.7: Gas permeability and ideal selectivity trend of CPI (9/1) - ZIF-8
MMMs with various ZIF-8 loadings for a) CO2/CH4 and b) C3H6/C3H8
In terms of gas pair selectivity, there is a minor increment of 6.87% for the
CO2/CH4 selectivity from 19.51 to 20.85 and a big increment of 134% for the
C3H6/C3H8 selectivity from 11.68 to 27.38. These different degrees of
enhancements in permeability and selectivity for CO2/CH4 and C3H6/C3H8
separation are consistent with previous literatures [9, 10, 15-18]. The addition
of ZIF nano-particles into the polymer matrix tends to increase the
permeability not selectivity for CO2 and CH4 separation because of their small
gas diameters and good interactions with ZIF-8 [9, 15]. Since the kinetic
diameters (4.3 vs. 4.5 for C3H6 and C3H8) and collision diameters (4.68 vs.
136
5.06 for C3H6 and C3H8) of C3H6 and C3H8 are bigger than the pore size (3.4
) of ZIF-8 [19], ZIF-8 may have a sharp discrimination for the transport of
this gas pair. In addition, C3H6 tends to have better sorption selectivity than
C3H8 even though they have very close critical temperatures (364.9 vs. 369.8
K for C3H6 and C3H8) [20]. As a consequence, there is no surprise to observe a
big increment in the C3H6/C3H8 selectivity. Since the pure ZIF-8 membrane
has a C3H6/C3H8 selectivity of 122 and a permeability of 277 Barrer at 30 C
[15], one may be able to estimate the separation performance of the MMMs
made of CPI (9/1) with different ZIF-8 content by substituting these intrinsic
values into the Maxwell equation [21-23]. Surprisingly, the calculated C3H6
permeability and C3H6/C3H8 selectivity are close to our experimental data, as
shown in Table 7.2. The permeability and selectivity of MMM made of CPI
(7/3) 20 wt% ZIF-8 are lower than calculated C3H6 permeability and
C3H6/C3H8 selectivity. This can be due to the higher amount of DABA moiety
in the structure of the CPI (7/3) that can increase the chance of the reaction
between ZIF-8 nano-particles and the COOH group of the DABA moiety. This
reaction may increase the selectivity of the membrane compared to calculated
data.
Table 7.2: Comparison between experimental data and calculated data by

using the Maxwell equation
Experimental Results Calculated results*
Perm. Perm.
Selectivity Selectivity
(Barrer) (Barrer)
C3H6 C3H6/C3H8 C3H6 C3H6/C3H8
CPI (9/1) 10 wt% ZIF-8 @ 200 C 17.1 14.86 17.9 14.46
CPI (9/1) 20 wt% ZIF-8 @ 200 C 20.5 18.08 24.3 17.87
CPI (9/1) 30 wt% ZIF-8 @ 200 C 29.7 23.18 32.5 21.90
CPI (9/1) 40 wt% ZIF-8 @ 200 C 47.3 27.38 42.7 26.55
CPI (7/3) 20 wt% ZIF-8 @ 200 C 19.1 25.47 18.6 20.03
*.
The pure ZIF-8 membrane has a C3H6/C3H8 selectivity of 122 and a C3H6 permeability of 277
Barrer at 30 C [15]
137
7.2.2.2 Effect of Durene to DABA ratio on gas separation performance
As mentioned earlier, three 6FDA-based polyimides were synthesized in this
work. The Durene moiety may result in membranes with a high permeability
due to the introduction of steric hindrance [24], while the DABA moiety
provides a reactive acid site useful for grafting or cross-linking [3-6, 25, 26].
Since the cross-linking reaction occurred at elevated temperatures and
produced phenyl radicals via decarboxylation of the COOH group, the phenyl
radicals further attack other portions of the polyimide for cross-linking. The
thermal induced cross-linking reaction is sensitive to the chemical structure of
the polyimide and annealing temperature. Increasing DABA moiety in the
polyimide structure can increase the chance of cross-linking.
Table 7.3: Single gas separation performance of pure polyimides and

polyimides with 20 wt% ZIF-8 at different annealing temperatures
Permeability (Barrer) Selectivity
Annealing
Membrane ID
Temp. CO2 CH4 C3H6 C3H8 CO2/CH4 C3H6/C3H8
PI 352 21 15 1.4 16.59 10.58

CPI (9/1) 256 13 13 1.1 19. 51 11.68
CPI (7/3) 158 6 10 0.7 25.48 14.17
200 C
PI 20 wt% ZIF-8 487 27 26 1.5 17.91 16.93
CPI (9/1) 20 wt% ZIF-8 392 19 21 1.1 20.45 18.08
CPI (7/3) 20 wt% ZIF-8 276 11 19 0.8 26.29 25.47
PI 432 31 18 1.9 13.80 9.53
CPI (9/1) 392 23 16 1.4 17.42 11.67
CPI (7/3) 369 15 15 0.9 24.44 15.91
350 C
PI 20 wt% ZIF-8 857 65 43 3.5 13.12 12.41
CPI (9/1) 20 wt% ZIF-8 604 35 37 2.1 17.24 17.43
CPI (7/3) 20 wt% ZIF-8 551 22 32 1.2 25.26 26.53
PI 541 41 30 3.2 13.11 9.31
CPI (9/1) 519 27 28 1.9 19.01 14.42
CPI (7/3) 429 17 25 1.4 25.96 17.26
400 C
PI 20 wt% ZIF-8 1090 84 55 4.6 12.96 11.94
CPI (9/1) 20 wt% ZIF-8 892 47 49 2.4 18.84 20.18
CPI (7/3) 20 wt% ZIF-8 698 27 43 1.6 25.84 27.47
138
Table 7.3 shows the gas separation performance as a function of Durene to
DABA ratio and confirms our hypothesis. Both pure membranes and MMMs
made of CPI (7/3) have higher selectivity but lower permeability than those
made of CPI (9/1) regardless of annealing temperature.
7.2.2.3 Effect of thermal treatment temperature on gas separation

performance
Table 7.3 also compares the gas separation performance of the pure membrane
and MMMs comprising a 20 wt% ZIF-8 loading as a function of annealing
temperature, while Figure 7.8 compares these MMMs with the tradeoff lines
for CO2/CH4 and C3H6/C3H8 separation.
Figure 7.8: Trade off lines of CO2/CH4 and C3H6/C3H8 separation
Interestingly, the pure membrane (i.e., 6FDA-Durene or PI), the MMM made
of CPI (9/1), and the MMM made of CPI (7/3) show large enhancements in
permeability but display different responses in selectivity with an increase in
annealing temperature. Generally, permeability may increase with annealing
temperature due to the breaking of hydrogen bonding [27] and thermal
expansion [11]. The former is especially true for CPI (9/1) and CPI (7/3)
polymers because they have strong hydrogen bonds among COOH groups of
DABA [27], while the latter is true for 6FDA-Durene (i.e., PI) because its
chain mobility and d-space increase with increasing annealing temperature due
to the thermal expansion [11]. As a result, the ideal selectivity of PI and the
139
MMM made of CPI (9/1) decreases with increasing annealing temperature.
However, due to the thermally induced cross-linking reaction, the ideal
selectivity of the MMM made of CPI (9/1) turns upward at a high annealing
temperature of 400 C. Since CPI (7/3) has a high amount of cross-linkable
COOH, the ideal selectivity of the MMM made of CPI (7/3) always increases
with an increase in annealing temperature. Comparing to the trade-off lines,
both MMMs made of CPI (9/1) and CPI (7/3) have reasonably high CO2/CH4
separation performance but their C3H6/C3H8 separation performance is
superior to the trade off line.
7.2.3 Plasticization behavior and mixed gas experiments

As aforementioned, plasticization is not desirable in gas separation processes.
Plasticization is a pressure dependent phenomenon attributable to the
dissolution of certain feed components within the polymer matrix. As feed
pressure increases, gas permeability of glassy polymers usually decreases due
to the saturation of Langmuir sites. However, gas permeability may start to
increase at the plasticization pressure when a large amount of feed
components dissolves into the polymeric membrane that facilitates inter-chain
mobility [24, 25, 28, 29]. Figure 7.9 demonstrates the CO2 plasticization
behavior of the PI 20 wt% ZIF-8, CPI (9/1) 20 wt% ZIF-8 and CPI (7/3)
20 wt% ZIF-8 MMMs after annealing at different temperatures.
Three conclusions can be drawn. Firstly, all membranes annealed at 200 C

show some degrees of plasticization phenomenon at a CO2 feed pressure of
around 10 atm. Secondly, the MMMs made of PI (i.e., uncross-linkable
6FDA-Durene) and then annealed at 300 and 400 C still undergo
plasticization even though their plasticization pressures increase slightly to
around 12-14 atm. Lastly, the MMMs made of cross-linkable CPI and then
annealed at elevated temperatures display enhanced plasticization resistance
against CO2. The degree of enhancement is dependent on the amount of cross-
linkable moiety (i.e., COOH) and annealing temperature. The MMM made of
CPI (9/1) and 20 wt% ZIF-8 shows plasticization suppression characteristics
up to 30 atm only when the annealing is conducted at 400 C, while the MMM
140
made of CPI (7/3) and 20 wt% ZIF-8 and annealed at 350 C already shows
plasticization suppression characteristics up to about 30 atm. Clearly, the
higher content of cross-linkable moiety and the higher annealing temperature
would result in MMMs with a higher degree of cross-linking reaction and
higher resistance against plasticization.
a) b)
c)
Figure 7.9: CO2 plasticization behaviour of a) PI 20 wt% ZIF-8, b) CPI

(9/1) 20 wt% ZIF-8, and c) CPI (7/3) 20 wt% ZIF-8 at different annealing
temperatures.
Table 7.4 compares the gas permeability and selectivity of CPI (9/1), CPI (9/1)
20 wt% ZIF-8, and CPI (9/1) 40 wt% ZIF-8 membranes thermally treated
at different temperatures under pure and mixed gas tests. For mixed gas tests,
the feed contains a mixture of CO2/CH4 (50/50) with a total feed pressure of
20 atm at 35 C.
141
Perm. Perm.
Selectivity Selectivity
(Barrer) (Barrer)
CPI (9/1) @ 200 C 256 13 19.51 305 24 13.81

CPI (9/1) 20 wt% ZIF-8 @ 200 C 392 19 20.45 368 23 17.18
CPI (9/1) 20 wt% ZIF-8 @ 400 C 892 47 18.84 728 39 19.61
CPI (9/1) 40 wt% ZIF-8 @ 200 C 779 34 20.85 712 39 18.05
a
Measured at 10 atm and 35 C, b Measured at 20 atm and 35 C
The mixed gas selectivity of the CPI (9/1) membrane annealed at 200 C drops
tremendously when comparing with its pure gas data due to the effect of
plasticization and sorption competition. For the MMMs made of CPI (9/1) and
20 wt% ZIF-8 and 40 wt% ZIF-8 and then thermally treated at 200 C, their
CO2 permeability and CO2/CH4 selectivity become lower while CH4
permeability becomes higher in mixed gas tests. Clearly, plasticization and
sorption competition continuously influence on their gas transport properties.
Interestingly, the MMM made of CPI (9/1) and 20 wt% ZIF-8 and then
thermally treated at 400 C overcomes the aforementioned trends. Not only
does it improve the mixed gas selectivity from 18.84 to 19.61 but also
decrease the CO2 permeability from 892 to 728 Barrer when comparing with
pure gas data. This positive result is mainly due to the fact that thermal
treatment at 400 C induces a high degree of cross-linking and thus enhances
membranes plasticization suppression properties.
Figure 7.10 illustrates the mixed gas permeation of CPI (9/1) 20 wt% ZIF-8
membranes as a function of CO2 pressure, thermally treated at 200 C and 400
C. The mixed gas of CO2/CH4 (50/50) feed mixture was used to show the
resistance of the membranes to a more aggressive condition.
142
a) b)
Figure 7.10: Mixed gas CO2 permeability as a function of feed pressure for
CPI (9/1) 20 wt% ZIF-8 membranes. a) thermally treated at 200 C, and b) at
400 C. The permeability curves obtained with CO2/CH4 (50/50) at 35 C
As shown in Figure 7.10a, the membrane without cross-linking (i.e., thermally

treated at 200 C) appears to undergo plasticization at around 10 atm feed
pressure. In contrast, Figure 7.10b shows that CO2 permeability of the cross-
linked membrane (i.e., thermally treated at 400 C) decreases with an
increasing in feed pressure. Figure 7.10 also shows that the mixed gas
CO2/CH4 selectivity of the cross-linked membrane decreases slightly with the
increase of feed pressure, though the membrane without cross-linking dropped
tremendously. It is obvious that the membrane without cross-linking was
swelled or plasticized in a CO2/CH4 (50/50) mixture and the cross-linked
membranes did not show plasticization behavior even up to a feed pressure of
36 atm (500 psia).
7.3 Conclusion
In this work three 6FDA-based cross-linkable co-polyimide/ZIF-8 mixed
matrix membranes with high gas separation performance for CO2/CH4 and
C3H6/C3H8 were fabricated. The effects of ZIF-8 nano-particles and thermal
treatment conditions on gas separation performance of these membranes have
been investigated. The following conclusions can be made.
Experimental results show that permeability for all gases increases

rapidly with increasing ZIF-8 loading due to the additional free volume
143
provided by ZIF-8 particles and increased distance among polymer
chains. Ideal C3H6/C3H8 selectivity improves visibly with increasing
ZIF-8 loading due to the inherently high selectivity of ZIF-8 for
C3H6/C3H8 separation. Similar to other works, the addition of ZIF
nano-particles into the polymer matrix mainly increases the
permeability (i.e., not selectivity) for CO2/CH4 separation because of
their small gas diameters and good interactions with ZIF-8.
Annealing and annealing temperature play significant roles on gas

separation performance. For 6FDA-Durene (PI), 6FDA-Durene/DABA
(9/1) (CPI (9/1)), 6FDA-Durene/DABA (7/3) (CPI (7/3)) polymers,
permeability of all gases increases with an increase in annealing
temperature up to 400 C due to the breakage of hydrogen bonds or
thermal expansion, while ideal selectivity follows different trends
depending on the existence of cross-linkable moiety and the degree of
cross-linking. Generally, a higher ratio of Durene to DABA moiety
may produce MMMS with a high permeability, while a higher ratio of
DABA to Durene moiety may result in MMMs with a higher
selectivity and plasticization resistance because DABA provides a
reactive acid site for thermal cross-linking.
All membranes annealed at 200 C or below show severe plasticization

phenomena at a CO2 pressure of about 10 atm, while MMMs made of
uncross-linkable 6FDA-Durene and annealed at 350 and 400 C still
undergo plasticization at around 12-14 atm. The enhancement in
plasticization resistance is strongly dependent on the amount of cross-
linkable moiety (i.e., DABA or COOH) and annealing temperature.
The MMM made of CPI (9/1) and 20 wt% ZIF-8 shows plasticization
suppression characteristics up to 30 atm only when the annealing is
conducted at 400 C. On the other hand, the MMM made of CPI (7/3)
and 20 wt% ZIF-8 and annealed at 350 C already shows plasticization
suppression characteristics up to about 30 atm. In summary, the higher
144
content of cross-linkable moiety and the higher annealing temperature
would result in MMMs with higher resistance against plasticization.
The MMM made of CPI (9/1) and 40 wt% ZIF-8 and thermally
annealed at 400 C possess impressive performance for C3H6/C3H8
separation with an ideal C3H6/C3H8 selectivity of 27.2 and a C3H6
permeability of 47.3, while the thermally annealed MMM made of CPI
(9/1) and 20 wt% ZIF-8 at 400 C show a CO2/CH4 selectivity of 19.61
and a notable CO2 permeability 728 Barrer in mixed gas tests. The
newly developed MMMs may have great potential for industrial nature
gas purification and C3H6/C3H8 separation.
145
7.4 References
[1] T.S. Chung, J.Z. Ren, R. Wang, D.F. Li, Y. Liu, K.P. Pramoda, C. Cao,
W.W. Loh, Development of asymmetric 6FDA-2,6DAT hollow fiber
membranes for CO2/CH4 separation: Part 2. Suppression of
plasticization, J. Membr. Sci. 214 (2003) 57-69.
[2] A.M. Kratochvil, W.J. Koros, Decarboxylation-induced cross-
linking of a polyimide for enhanced CO2 plasticization resistance,
(2012) 141-151.
saccharide units on the gas separation performance of highly permeable
polyimide membranes, J. Membr. Sci. 415-416 (2012) 375-382.
[6] M. Askari, T. Yang, T.S. Chung, Natural gas purification and
olefin/paraffin separation using cross-linkable dual-layer hollow fiber
membranes comprising -cyclodextrin, J. Membr. Sci. 423-424 (2012)
392-403.
[7] J.K. Ward, W.J. Koros, Crosslinkable mixed matrix membranes with
surface modified molecular sieves for natural gas purification: I.
Preparation and experimental results, J. Membr. Sci. 377 (2011) 75-81.
[8] W.H. Lin, T.S. Chung, Gas permeability, diffusivity, solubility, and
aging characteristics of 6FDA-durene polyimide membranes, J. Membr.
Sci. 186 (2001) 183193.
[9] Q. Song, S.K. Nataraj, M.V. Roussenova, J.C. Tan, D.J. Hughes, W. Li,
P. Bourgoin, M.A. Alam, A.K. Cheetham, S.A. Al-Muhtaseb, E.
Sivaniah, Zeolitic imidazolate framework (ZIF-8) based polymer
nanocomposite membranes for gas separation, Energy Environ. Sci. 5
(2012) 8359-8369.
146
229-239.
[11] L. Shao, T.S. Chung, K.P. Pramoda, The evolution of physicochemical
and transport properties of 6FDA-durene toward carbon membranes;
from polymer, intermediate to carbon, Microporous Mesoporous Mater.
84 (2005) 59-68.
[12] Y. Li, H.M. Guan, T.S. Chung, S. Kulprathipanja, Effects of novel silane
modification of zeolite surface on polymer chain rigidification and
partial pore blockage in polyethersulfone (PES)zeolite A mixed matrix
[13] L.Y. Jiang, T.S. Chung, R. Rajagopalan, Matrimid/MgO mixed matrix
membranes for iso-propanol dehydration by pervaporation, AIChE J. 53
(2007) 17451757.
[14] C. Courteille, Ch. Hollenstein, J.L. Dorier, P. Gay, W. Schwarzenbach,
A.A. Howling, E. Bertran, G. Viera, R. Martins, A. Macarico, Particle
agglomeration study in rf silane plasmas: In situ study by polarization-
sensitive laser light scattering, J. Appl. Phys. 80 (1996) 2069-2078.
[15] C. Zhang, Y. Dai, J.R. Johnson, O. Karvan, W.J. Koros, High
performance ZIF-8/6FDA-DAM mixed matrix membrane for
composite membranes for hydrogen purification, Energy Environ. Sci. 4
(2011) 4171-4180.
[17] Y.C. Pan, Z.P. Lai, Sharp separation of C2/C3 hydrocarbon mixtures by
zeolitic imidazolate framework-8 (ZIF-8) membranes synthesized in
aqueous solutions Chem. Commun. 47 (2011) 1027510277.
[18] Y.C. Pan, T. Li, G. Lestari, Z.P. Lai, Effective separation of
propylene/propane binary mixtures by ZIF-8 membranes, J. Membr. Sci.
390391 (2012) 9398.
[19] S.S. Chan, R. Wang, T.S. Chung, Y. Liu. C2 and C3 hydrocarbon
separations in poly (1,5-naphthalene-2,2-bis(3,4-phthalic)
147
Membr. Sci. 210 (2002) 55-64.
separation using polymeric membranes, J. Membr. Sci. 211 (2003) 299
309.
[21] R. Mahajan, R. Burns, M. Schaeffer, W.J. Koros, Challenges in forming
successful mixed matrix membranes with rigid polymeric materials, J.
Appl. Polym. Sci. 86 (2002) 881-890.
[22] L.Y. Jiang, T.S. Chung, S. Kulprathipanja, Fabrication of mixed matrix
hollow fibers with intimate polymer-zeolite interface for gas separation,
AIChE J. 52 (2006) 2898 - 2908.
[23] T.S. Chung, L.Y. Jiang, Y. Li, S. Kulprathipanja, Mixed matrix
membranes (MMMs) comprising organic polymers with dispersed
inorganic fillers for gas separation, Prog. Polym. Sci. 32 (2007) 483-507.
Sci. 34 (2009) 561-580.
aging on the gas permeability of thin 6FDA-based polyimide
membranes. Part 2. With crosslinking, J. Membr. Sci. 282 (2006) 3243.
[26] J.D. Wind, C. Staudt-Bickel, D.R. Paul, W.J. Koros, Solid-state covalent
cross- linking of polyimide membranes for carbon dioxide plasticization
reduction, Macromolecules 36 (2003) 18821888.
6056.
[28] A. Bos, I.G.M. Pnt, M. Wessling, H. Strathmann, Plasticization-
[29] S. Neyertz, D. Brown, S. Pandian, N.F.A. van der Vegt, Carbon dioxide
diffusion and plasticization in fluorinated polyimides, Macromolecules
43 (2010) 7813-7827.
148
Chapter 8: Conclusion and Recommendation
8.1 Conclusion
With the considerate of the limitation on current available polymeric
membranes, especially for glassy polymer that is described by Robesons
upper bound trade-off limit, the exploration of high performance polymeric
membrane materials had been carried out in this study. It is confirmed that
polymeric membranes with attractive gas separation performance can be
obtained by designing novel materials, or modifications on the current
polymeric membranes. This study was first performed on the modification of
one of the common 6FDA-based co-polyimide material. In this modification,
three kinds of Cyclodextrins (CDs) were chemically grafted to the cross-
linkable polyimide matrix and then decomposed at elevated temperatures.
Considering the advantageous values of hollow fiber over flat sheet membrane
in the industry, a cross-linkable co-polyimide dual layer hollow fiber
membrane has been fabricated. After that, cross-linkable mixed matrix
membranes were developed using 6FDA-based co-polyimide embedded with
different wt% of ZIF-8 nano-particle. As introduced in Chapter 1, the cross-
linkable mixed matrix membrane concept is an attractive technology that
promises both increased efficiency and plasticization resistance for natural gas
purification and olefin/paraffin separation in comparison to traditional
polymer membranes. The developed high performance membranes are
specifically for gas separation applications, e.g., CO2/CH4, and C3H6/C3H8. In
summary, three aspects had been studied:
1. Cross-linkable 6FDA-Durene/DABA co-polyimides grafted with , ,
and -Cyclodextrin.
2. Cross-linkable dual-layer hollow fiber membranes comprising -
Cyclodextrin.
3. Thermal cross-linkable co-polyimide/ZIF-8 mixed matrix membranes.
The above mentioned studies have shown significant enhancement in gas
separation capability. As a whole, the best membranes for natural gas
separation are PPM--CD-425, and PPM--CD-425, and the best membranes
for C3 separation regarding to C3H6 permeability are PPM--CD-425, and
PPM--CD-425, and regarding to C3H6 /C3H8 selectivity are CPI (9/1) 40
149
wt% ZIF-8 thermally treated @ 200 C and CPI (7/3) 20 wt% ZIF-8
thermally treated @ 400 C. These membranes have very high permeability
with moderate selectivity that can decrease the required area of the
membranes. It is clear that the most expensive part of the membrane process is
material cost and by decreasing membrane area, the material cost of this
process decrease. Then, permeability and selectivity improvement of these
membranes can decrease the cost of the membrane process. The detailed
conclusions of each study have been derived and summarized as follows:
8.1.1 Cross-linkable 6FDA-Durene/DABA co-polyimides grafted with

Cyclodextrins
Cross-linkable co-polyimide (6FDA-Durene/DABA (9/1)) membranes grafted
with various sizes of Cyclodextrin (CD) were successfully fabricated and then
thermally treated at elevated temperatures. The study showed that the
modified co-polyimide membranes with high-performance gas separation
could enhance both membrane permeability and resistance to plasticization
simultaneously. The results indicate that separation performance of thermally
treated polyimides that were originally grafted with the CD is much better than
that without the CD. For CO2/CH4 separation, all the co-polyimide
membranes grafted by CDs surpass the trade-off line while the thermally
treated original polyimide cannot. For C3H6/C3H8 separation, all membranes
thermally treated at 425 C can surpass the trade-off line with almost the same
permselectivity. However, their permeability follows the order of PPM--CD-
425 > PPM--CD-425 > PPM--CD-425 > Original PI-425. The enhancement
in permeability can be attributed to the decomposition of CD that may convert
the spaces originally occupied by CD to micro-pores and thus increase
fractional free volume and gas permeability. The results suggest this
permeability increase is strongly related to the cavity size of CD; the bigger
CD size, the higher permeability jumps after thermal treatments at elevated
temperatures. On the other hand, permselectivity increases when the thermal
treatment is conducted over 400 C. A possible reason for permselectivity
increment can be due to the occurrence of cross-linking reaction in the
polyimide matrix that tightens the d-space. The resultant thermally treated
150
polyimides that were originally grafted with the CD show much better anti-
plasticization characteristics than that without the CD and PPM--CD-425
membrane can resist over 30 atm while the original polyimide is less than 15
atm.
8.1.2 Cross-linkable dual-layer hollow fiber membranes comprising -

Cyclodextrin
Thermally cross-linkable co-polyimide dual-layer hollow fiber membranes
grafted with -Cyclodextrin for separation of CO2/CH4 and propylene/propane
were fabricated. In order to find the best spinning condition, the performance
of hollow fiber membranes at various take-up velocities and outer-layer dope
flow rates was investigated. The fiber membranes were thermally cross-linked
at different temperatures, and the performance of the fibers before and after
the silicon rubber coating was studied using CH4, CO2, propane and
propylene. Experimental results show that permeances decrease with
increasing take-up velocity. Before silicon rubber coating, higher selectivities
can be obtained by increasing take-up velocity from 4 to 7.4 m/min. A further
increase in take-up velocity to 11.8 m/min results in lower selectivities due to
defect formation. On the other hand, after silicon rubber coating, the CO2/CH4
and C3H6/C3H8 selectivities increase as the take-up velocity increases. In
addition, permeances increase with a reduction in outer-layer dope flow rate
due to the thinner skin layer and lower transport resistance. However,
selectivities decrease when the outer-layer dope flow rate decreases owing to
defect formation.
Thermal treatment temperature plays a significant role in fiber morphology.
Fiber diameters shrunk after treatment. The 350 C treated fibers show higher
permeances and lower selectivities close to those of untreated fibers while the
400 C treated fibers show higher selectivities but lower permeances due to the
skin densification and a higher degree of cross-linking. As a result, the silicon
rubber coating has a less effect on the 400 C treated fibers than the 350 C
treated fibers. Selectivities of thermally treated fiber membranes at 350 C
were slightly higher than those of the precursor fibers, and this improvement
was more significant for membranes treated at 400 C. This enhancement
151
demonstrates that cross-linking is more severe at 400 C than other thermal
treatment temperatures.
8.1.3 Thermal cross-linkable co-polyimide/ZIF-8 mixed matrix

membranes
Using three 6FDA-based polyimides (6FDA-Durene, 6FDA-Durene/DABA
(9/1), 6FDA-Durene/DABA (7/3)) and nano-size zeolitic imidazolate
framework-8 (ZIF-8), mixed matrix membranes (MMMs) with uniform
morphology comprising ZIF-8 as high as 40 wt% loading by directly mixing
as-synthesized ZIF-8 suspension into the polymer solution were fabricated.
Experimental results show that permeability for all gases increases rapidly
with increasing ZIF-8 loading due to the additional free volume provided by
ZIF-8 particles and increased distance among polymer chains. Ideal
C3H6/C3H8 selectivity improves visibly with increasing ZIF-8 loading due to
the inherently high selectivity of ZIF-8 for C3H6/C3H8 separation. Similar to
other works, the addition of ZIF nano-particles into the polymer matrix mainly
increases the permeability (i.e., not selectivity) for CO2/CH4 separation
because of their small gas diameters and good interactions with ZIF-8.
Annealing and annealing temperature plays significant roles in gas separation
performance. For 6FDA-Durene (PI), 6FDA-Durene/DABA (9/1) (CPI (9/1)),
6FDA-Durene/DABA (7/3) (CPI (7/3)) polymers, permeability of all gases
increases with an increase in annealing temperature up to 400 C due to the
breakage of hydrogen bonds or thermal expansion while ideal selectivity
follows different trends depending on the existence of cross-linkable moiety
and the degree of cross-linking. Generally, a higher ratio of Durene to DABA
moiety may produce MMMs with a high permeability while a higher ratio of
DABA to Durene moiety may result in MMMs with a higher selectivity and
plasticization resistance because DABA provides a reactive acid site for
thermal cross-linking. All membranes annealed at 200 C or below show
severe plasticization phenomena at a CO2 pressure of about 10 atm while
MMMs made of uncross-linkable 6FDA-Durene and annealed at 350 and 400
C still undergo plasticization at around 12-14 atm. The enhancement in
plasticization resistance is strongly dependent on the amount of cross-linkable
moiety (i.e., DABA or COOH) and annealing temperature. The MMM made
152
by CPI (9/1) and 20 wt% ZIF-8 shows plasticization suppression
characteristics up to 30 atm only when the annealing is conducted at 400 C.
On the other hand, the MMM made by CPI (7/3) and 20 wt% ZIF-8 and
annealed at 350 C already shows plasticization suppression characteristics up
to about 30 atm. In summary, the higher content of cross-linkable moiety and
the higher annealing temperature would result in MMMs with higher
resistance against plasticization.
8.2 Recommendation for future work

Although the above mentioned research works have exhibited a significant
enhancement in gas separation applications, the following recommendations
for future work may provide great insights for the further development of
membrane materials and fabrication technology for the commercialization of
those advanced functional materials.
8.2.1 Potential future project directions

Polymeric glassy polymers are inherently non-equilibrium materials. They
undergo constant molecular rearrangements to attain an equilibrium state. This
process is termed physical aging. Generally, the physical aging of polymeric
films would result in a continuous decrease in gas permeability with the time.
This, on the other hand, has constrained the applicability of polymeric
membranes for industrial use. Therefore, the long term physical aging study
(i.e., Stability test) is one of the crucial steps prior to determining the industrial
applicability of a membrane. Previously, both cross-linkable co-polyimide
membranes grafted with Cyclodextrin and mixed matrix membranes with ZIF-
8 nano-particle have exhibited superior gas separation performance and anti-
plasticization behavior that far exceeding the recent upper-bounds for the
state-of-art polymeric membranes (refer to the Chapter Four and Seven). The
long term stability test (e.g., At least 6 months) would be another key
parameter in determining the feasibility of those membranes for industrial use.
Additionally, the membrane thickness should be very thin to achieve a high
flux of the permeating component, preferably in the range of less than 1 m.
Another reason to conduct the aging test with a thin polymer film is due to the
153
totally different permeation and aging behavior of submicron and nano-sized
glassy polymer membranes.
Other thermally labile groups can be grafted in the structure of the polymers to
investigate the effect of these materials on the gas separation performance of
the membranes. Generally, thermally labile groups could be desired for
polymer structure modification if they have below properties:
1. Commercially available
2. Low cost material
3. High molecular weight
4. High decomposition temperature (near to Tg of polymer)
5. High OH or COOH group in their structure
These materials, with above characterizations can be change transport
properties of the membranes upon thermal treatment and could be a good
candidate to improve gas separation properties of the membranes. The OH and
COOH groups in the structure of these materials can increase the chance of
cross-linking in the structure of the membranes after thermal treatment and
this point is important for polymer architecture to design and generate a good
cross-linking membrane with high separation performance.
8.2.2 Hollow fiber spinning of the cross-linkable mixed matrix

membranes
Since a proof of concept has been established for the use of cross-linked
mixed matrix membranes for gas separation, and if nano-sized ZIF-8 can be
achieved, the next step in membrane development is scale-up from dense film
membranes to hollow fiber membrane modules. A few technical issues will
have to be overcome in the scale-up process. The most important one is that
the outer selective layers of hollow fiber membranes, which are similar to
dense films, should be prepared in more than 20 wt% polymer in NMP
solution and before spinning it will be needed to remove the bubbles from the
dope solution. This bubble removing needs degassing procedure with ultra-
sonic and ultra-sonic may lead to gel formation due to the reaction of ZIF-8
with COOH group of DABA moiety. Then, further study is needed to find the
new method for degassing the outer layer dope solution.
154

Askari Mohammad-Thesis-23 July 2014

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Askari Mohammad-Thesis-23 July 2014

Diunggah oleh

Hak Cipta:

Format Tersedia

CO2/CH4 AND PROPYLENE/PROPANE SEPARATION

USING CROSS-LINKABLE POLYMERIC

A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF

DEPARTMENT OF CHEMICAL AND BIOMOLECULAR

NATIONAL UNIVERSITY OF SINGAPORE

I wish to first express my sincere appreciation to my academic advisor,

I gratefully acknowledge the research scholarship by the National University

Perhaps most influential in my graduate school success has been my best

DIFFUSIVITY COEFFICIENT OF CO2 IN POLYMERIC MEMBRANES ....................101

5.3 CONCLUSION ......................................................................................102

6.2.3 EFFECT OF OUTER-LAYER DOPE FLOW RATE ON GAS SEPARATION

6.2.4 EFFECT OF THERMAL TREATMENT AND SILICON RUBBER COATING

6.2.5 CO2 PLASTICIZATION AND MIXED GAS PERMEATION EXPERIMENTS 120

HOLLOW FIBER MEMBRANES .........................................................................122

6.3 CONCLUSION ......................................................................................124

7.2.2.2 EFFECT OF DURENE TO DABA RATIO ON GAS SEPARATION

7.2.2.3 EFFECT OF THERMAL TREATMENT TEMPERATURE ON GAS

SEPARATION PERFORMANCE ......................................................................139

7.2.3 PLASTICIZATION BEHAVIOR AND MIXED GAS EXPERIMENTS ............140

GRAFTED WITH CYCLODEXTRINS ..................................................................150

8.1.2 CROSS-LINKABLE DUAL-LAYER HOLLOW FIBER MEMBRANES

8.1.3 THERMAL CROSS-LINKABLE CO-POLYIMIDE/ZIF-8 MIXED MATRIX

Membrane is an emerging technology that holds great promises and displays

Secondly, with the purpose of better fundamental understanding of this class

Thirdly, considering the importance of hollow fiber for industrial use,

Table 1.1: Thermodynamic properties of propane and propylene .................... 6

Figure 1.1: Sources of carbon dioxide emissions globally ............................... 3

A Effective area of membrane (cm2)

P Average permeability coefficient (Barrer)

Global warming, by means of an increase in the average temperature of earth

Besides, natural gas plays an important role in todays energy production. It

The other important separation process in industries is olefin/paraffin

Separation of gases is one of practical but very important unit operations in

Membrane separation, compared to amine absorption, pressure swing

Figure 1.1: Sources of carbon dioxide emissions globally [4]

Figure 1.2: Global propane and propylene consumption [13,14]

1.1 Membrane Technology for Gas Separation

Gas separation membranes currently comprise a market of over 150 million

1.1.1 Natural Gas Purification

1.1.2 Carbone Dioxide Capture

1.1.3 Olefin/Paraffin Separation

Table 1.1: Thermodynamic properties of propane and propylene [23]

Propylene C3H6 42 226 365.2

Propane C3H8 44 230.4 369.9

The membrane has a symmetric structure always named dense membrane. It

The asymmetric membrane consists of a number of layers, each with different

Depending on the process applications, different types of membrane modules

1.3 Research Objective and Organization of Dissertation

Chapter 2 summarizes background information, a review of significant

Chapter 3 documents the experimental approaches and methodologies along

Chapter 4 reports the preparation of cross-linkable co-polyimide (6FDA-

Chapter 5 investigates the intrinsic gas permeation properties of 6FDA-

Chapter 6 discusses the preparation of cross-linkable dual-layer hollow fiber

2.1 Polymeric Membranes for Gas Separation

Depending on the glass transition temperature of the membrane material, the

Among the polymeric membranes for gas separation application, polyimides

2.1.1 General Transport Theory

Understanding the glassy state of amorphous polymers is very important in

Permeability, solubility, and diffusivity coefficient through the glassy

2.1.1.1 Permeability Coefficient

At steady-state, we assume that the diffusion coefficient is independent of

where Di is the local diffusion concentration of the gas in the membrane, Mi is

With this equation, the permeability of a gas can be expressed as:

where the SAv. is the equilibrium solubility coefficient of a gas penetrant A in