Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Fabrication of silicon carbide sintered supports and silicon carbide membranes
(USC Thesis Other)
Fabrication of silicon carbide sintered supports and silicon carbide membranes
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
FABRICATION OF SILICON CARBIDE SINTERED SUPPORTS AND SILICON
CARBIDE MEMBRANES
By
Wangxue Deng
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMICAL ENGINEERING)
December 2013
Copyright 2013 Wangxue Deng
Table of Contents
Dedication ......................................................................................................................... iv
Acknowledgements ............................................................................................................ v
List of Figures .................................................................................................................. vii
List of Tables ..................................................................................................................... x
Abstract .............................................................................................................................. 1
1. Introduction ................................................................................................................... 3
1.1 Ceramic Membranes .................................................................................................. 3
1.2 Silicon Carbide Membranes ...................................................................................... 7
1.2.1 Pyrolysis of Polymer Precursors ............................................................................ 8
1.2.2 Chemical Vapor Deposition/Infiltration ............................................................... 11
1.3 Porous Silicon Carbide Supports ............................................................................. 13
1.4 Transport Phenomena in Membrane Separation ..................................................... 15
2. Highly Permeable Porous Silicon Carbide Support Tubes ..................................... 19
2.1 Introduction ............................................................................................................. 19
2.2 Experimental Section .............................................................................................. 28
2.3 Results and Discussion ............................................................................................ 35
2.3.1 Effect of Sintering Temperature ....................................................................... 36
2.3.2 Effect of the Concentration of Sintering Aids .................................................. 42
2.3.3 Effect of Different Initial Powder Composition ............................................... 50
2.3.4 Effect of Grinding ............................................................................................. 56
2.3.5 Mechanical Properties....................................................................................... 61
2.3.6 Microporous Membrane Preparation ................................................................ 62
2.4 Conclusion ............................................................................................................... 64
2.5 Acknowledgement ................................................................................................... 65
ii
3. Fabrication of Silicon Carbide Membranes ............................................................. 66
3.1 Introduction ............................................................................................................. 66
3.1.1 Pyrolysis of Polymeric Precursors .................................................................... 67
3.1.2 Chemical Vapor Deposition/Infiltration ........................................................... 69
3.2 Experimental Section .............................................................................................. 71
3.2.1 Fabrication of SiC Membranes by Pyrolysis of AHPCS .................................. 71
3.2.2 Fabrication of SiC Membranes by CVD/CVI................................................... 77
3.3 Results and Discussion ............................................................................................ 79
3.3.1 SiC Membranes Prepared via Pyrolysis............................................................ 79
3.3.2 SiC Membranes Prepared by CVD/CVI ........................................................... 88
3.4 Conclusions ............................................................................................................. 90
4. Suggestions for Future Work ..................................................................................... 91
Nomenclature................................................................................................................... 93
References ........................................................................................................................ 94
iii
Dedication
To my beloved father Junfeng Deng, mother Lixiao Tong, and sister Wangping Deng for giving
me their unconditional love, to my girlfriend for her accompanying,
and
To films and Mount Baldy, dreams of love and hope shall never die.
“ 张骞凿空,未睹昆仑;唐玄奘、元耶律楚材衔人主之命,乃得西游。吾以老布衣,孤筇
双屦,穷河沙,上昆仑,历西域,题名绝国,与三人而为四,死不恨矣。” (徐霞客)
iv
Acknowledgements
I would like to sincerely and deeply thank my advisors, Prof. Theodore Tsotsis and Prof.
Muhammad Sahimi for their tutoring, help, and support in the past four years. Their knowledge,
ideas, and straightforwardness are of great value and cruciality to my life and have always
guided me through my PhD study right from the beginning to the end.
I wish to thank Prof. Grace Lu from department of Physics for being my defense and qualifying
exams committee members. I want to thank Prof. Katherine Shing and Prof. Edward Goo from
Mork family department of chemical engineering and materials science for serving as my
qualifying exam committee members as well. I would like to thank Dr. Sharon Meyers from
department of American language institute for being such a nice language teacher and may she
rest in peace.
I would like to thank Prof. Xinhai Yu, from East China University of Science and Technology,
for helping me with my PhD work with his expertise and experience in Chemical Engineering.
My sincere gratitude also goes to my colleague PhD friends, Dr. Yousef Motamedhashemi,
Xiaojie Yan, Yu Wang, Dr. Jiang Yu, and Dr. Junyi Xu.
I would like to thank my friends in the department: Dr. Ryan Mourhatch, Sasan Dabir, Hamed
Barghi, Sahar Soltani, Basabdatta Roychaudhuri, Dr. Nitin Nair for their help.
I wish to also thank the following staff members in the Mork Family Department of Chemical
Engineering and Materials Science: Ms. Karen Woo, Ms. Tina Silva, Mr. Martin Olekszyk, Ms.
Angeline Fugelso, Mr. Shokry Bastorous, Ms. Heather Alexander, Mr. Andy Chen, Ms. Aimee
Bernard, and Ms. Laura Carlos. I am really grateful for all their help through these years.
v
Above all, I would like to thank my parents, Junfeng Deng and Lixiao Tong, my sister Wangping
Deng for their love, patience, care, encouragement, and relief through all these twenty six years.
I wouldn’t be here without them.
Lastly, financial support by the United States Department of Energy, the National Science
Foundation, and the Mork Family Department of Chemical Engineering and Materials Science is
gratefully acknowledged.
vi
List of Figures
Figure 1-1. Schematic illustration of a typical ceramic membrane (Li, 2007) ............................... 4
Figure 1-2. Application range of different pressure-driven membrane process (Mulder, 1996) .... 5
Figure 1-3. Molecular structures of AHPCS and HPCS ................................................................. 9
Figure 1-4. Pore diffusion effects (Cussler, 2009) ........................................................................ 16
Figure 2-1. Porous SiC tubular supports sintered at 1900
o
C for 3 hr .......................................... 29
Figure 2-2. Schematic diagram of the permeation test apparatus ................................................ 31
Figure 2-3. Permeance vs. Pav of blended powder (50/50) supports sintered at different
temperatures (Permeation data taken at T= 378 K, 4wt% phenolic resin and 0.1wt% B4C) ........ 38
Figure 2-4. Ideal separation factor vs. Pav of blended powder (50/50) supports sintered at
different temperatures (Permeation data taken at T= 378 K, 4wt% phenolic resin and 0.1wt%
B4C)............................................................................................................................................... 39
Figure 2-5. AFM images of surfaces of blended powder (50/50) supports sintered at 1700
o
C,
1800
o
C, and 1900
o
C .................................................................................................................... 40
Figure 2-6. SEM images of blended powder (50/50) supports sintered at 1700
o
C, 1800
o
C, and
1900
o
C.......................................................................................................................................... 41
Figure 2-7. Permeance vs. Pav of blended powder (50/50) supports with different amounts of
sintering aids (Permeation data taken at T= 378 K; Normal means 4wt% phenolic resin and
0.1wt% B4C, Small means 0.4wt% phenolic resin and 0.01wt% B4C, Trace means 0.04wt%
phenolic resin and 0.001wt% B4C, No Add means no sintering aids added) ............................... 44
Figure 2-8. Ideal separation factor vs. Pav of blended powder (50/50) supports with different
amounts of sintering aids (Permeation data taken at T= 378 K; Normal means 4wt% phenolic
resin and 0.1wt% B4C, Small means 0.4wt% phenolic resin and 0.01wt% B4C, Trace means
0.04wt% phenolic resin and 0.001wt% B4C, No Add means no sintering aids added)................ 45
vii
Figure 2-9. AFM images of surfaces of blended powder (50/50) supports with different amounts
of sintering aids (Normal means 4wt% phenolic resin and 0.1wt% B4C, Small means 0.4wt%
phenolic resin and 0.01wt% B4C, Trace means 0.04wt% phenolic resin and 0.001wt% B4C, No
Add means no sintering aids added) ............................................................................................. 48
Figure 2-10. SEM images of blended powder (50/50) supports with different amounts of
sintering aids (Normal means 4wt% phenol resin and 0.1wt% B4C, Small means 0.4wt%
phenolic resin and 0.01wt% B4C, Trace means 0.04wt% phenolic resin and 0.001wt% B4C, No
Add means no sintering aids added) ............................................................................................. 49
Figure 2-11. Permeance vs. composition of large particles in supports (Permeation data taken at
T= 378 K, 4wt% phenolic resin and 0.1wt% B4C, sintering temperature 1900
o
C, Average
measurement pressure Pav=1.43×10
5
Pa) ....................................................................................... 51
Figure 2-12. Porosity and ideal separation factor vs. composition of large particles in supports
(Permeation data taken at T= 378 K, 4wt% phenolic resin and 0.1wt% B4C, sintering
temperature 1900
o
C, Average measurement pressure Pav=1.43×10
5
Pa) ..................................... 52
Figure 2-13. AFM images of surfaces of blended powder supports with different compositions
(4wt/% phenolic resin and 0.1wt% B4C, sintering temperature 1900
o
C, the two numbers indicate
the weight fraction of the 0.6 µm small particles and of the 6 µm large particles) ...................... 54
Figure 2-14. SEM images of blended powder supports with different compositions (4wt/%
phenolic resin and 0.1wt% B4C, sintering temperature 1900
o
C, the two numbers indicate the
weight fraction of the 0.6 µm small particles and of the 6 µm large particles) ............................ 55
Figure 2-15. Effect of surface treatment (grinding) ...................................................................... 57
Figure 2-16. Permeance vs. average pressure of untreated and surface-treated blended powder
(50/50) supports (Permeation data taken at T= 378 K, no sintering aids, sintering temperature
1800
o
C) ........................................................................................................................................ 58
Figure 2-17. Ideal separation factor vs. average pressure of untreated and surface-treated blended
powder (50/50) supports (Permeation data taken at T= 378 K, no sintering aids, sintering
temperature 1800
o
C) .................................................................................................................... 59
viii
Figure 2-18. AFM images of surfaces of untreated and surface treated powder (50/50) supports
(no sintering aids, sintering temperature 1800
o
C) ....................................................................... 60
Figure 3-1. Schematic illustration of the CVD/CVI process (Chen et al., 2008) ......................... 69
Figure 3-2. Schematic of the first type of permeation test set-up for single-gas measurements .. 73
Figure 3-3. Schematic of the first type of permeation test set-up when used for mixed-gas
permeation experiments ................................................................................................................ 75
Figure 3-4. Schematic of the second type of permeance test set-up ............................................. 76
Figure 3-5. Schematic of the first type of CVD apparatus ........................................................... 77
Figure 3-6. Schematic of the second type of CVD apparatus ....................................................... 78
Figure 3-7. Structure of an idealizd SiC membrane ..................................................................... 80
Figure 3-8. Permeation results of SiC membranes made from (+) PS slip-casting solution
(Measurement temperature: 473 K, pressure difference across the membrane 2.41×10
5
Pa; each
individual data point reflects results of three different membranes prepared under identical
conditions, as described in the text) .............................................................................................. 83
Figure 3-9. Permeation results of SiC membranes made from (+) PS slip-casting solution
(Measurement temperature: 473 K, pressure difference across the membrane 2.41×10
5
Pa; each
individual data point reflects results of three different membranes prepared under identical
conditions, as described in the text) .............................................................................................. 84
Figure 3-10. Cross-sectional SEM images of membranes at different stages of preparation ....... 86
Figure 3-11. Top view SEM images of membranes at different stages of preparation ................ 87
Figure 3-12. He and Ar permeances for a membrane prepared via CVD (Deposition temperature:
850
o
C, Argon flow rate: 1.2 ml/s; TPS injection rate: 30 µl/hr) .................................................. 89
ix
List of Tables
Table 1-1. Comparison of different-pressure driven membrane process (Mulder, 1996).............. 6
Table 2-1. SiC supports sintered at different temperatures (Sintering aids: 4wt% phenolic resin
and 0.1wt%B4C; composition: 50wt% 0.6 µm powder and 50wt% 6 µm powder; average
measurement pressure Pav=1.43×10
5
Pa; measurement temperature: 378 K) ............................... 42
Table 2-2. SiC supports sintered with different amounts of sintering aids (Sintering temperature:
1900
o
C; composition: 50wt% 0.6 µm powder and 50wt% 6 µm powder; average measurement
pressure Pav=1.43×10
5
Pa; measurement temperature: 378 K) ...................................................... 46
Table 2-3. SiC supports sintered with different compositions (Sintering temperature: 1900
o
C;
Sintering aids: 4wt% phenolic resin and 0.1wt%B4C; average measurement pressure
Pav=1.43×10
5
Pa; measurement temperature: 378 K, * only one sample tested) .......................... 56
Table 2-4. SiC supports with and without surface treatment (Sintering temperature: 1800
o
C; no
sintering aids used; average measurement pressure Pav=1.43×10
5
Pa; measurement temperature:
378 K) ........................................................................................................................................... 60
Table 2-5. Permeance data of three membranes prepared using 50/50 blend support
(Measurement Temperature: 200
o
C; pressure difference across membranes: 2.07×10
5
Pa) ........ 64
x
Abstract
Efficient separation of hydrogen (H2) under high temperatures and pressures is important to the
development of the clean-energy industry, and has been among the key drivers for research on
inorganic membranes for the last two decades. Although substantial efforts have been devoted to
date to the preparation of nanoporous membranes for H2 separation, the fabrication of high-
temperature and steam-stable inorganic membranes with high hydrogen fluxes and large
separation factors still remains a key challenge. Among all the potential candidates, silicon
carbide (SiC) membranes show potential advantages for use in hydrogen separation processes
under harsh and corrosive conditions such as, for example, the steam reforming and the water
gas shift reactions commonly employed in H2 production; this is because SiC is a material that
has high corrosion resistance, high thermal conductivity, high thermal shock resistance, and
excellent chemical and mechanical stability, making it thus a promising material for application
in industrial processes for clean energy production.
High-quality porous SiC supports are of great importance in the fabrication of hydrogen
permselective SiC nanoporous membranes and their preparation has, thus, been a keen a key
focus in this research. In this Thesis, we report on the preparation, via the pressureless sintering
of β-SiC powders, of SiC tubular supports that are both highly permeable and mechanically
strong. Their transport characteristics were studied via inert-gas permeation tests, while their
structure and surface morphology were characterized by atomic force microscopy (AFM) and
scanning electron microscopy (SEM) analysis. In addition, the effect of varying the composition
of starting powders, the sintering temperature, and the amount of sintering aids utilized on the
transport characteristics and the surface roughness of such sintered SiC porous supports were
1
also systematically investigated. These tubular SiC supports exhibit high fluxes (a He permeance
as high as 5.8×10
-5
mol•m
-2
•s
-1
•Pa
-1
) and are mechanically strong (compressive strength as high
as 106 MPa) to potentially withstand the pressure drops required in their use as membrane
supports.
Utilizing these high-quality SiC sintered supports, nanoporous SiC membranes were prepared by
the pyrolysis of thin ally-hydridopolycarbosilane (AHPCS) films coated on such supports, using
a combination of slip-casting and dip-coating techniques coupled with periodic coatings of
polystyrene sacrificial interlayers. The membranes prepared using these SiC supports are very
permselective (a He/Ar separation factor as high as 2000) and exhibit a He permeance as high as
2.4×10
-7
mol•m
-2
•s
-1
•Pa
-1
.
2
1. Introduction
1.1 Ceramic Membranes
Compared to polymeric membranes, which are rarely used under harsh conditions, ceramic
membranes can withstand high-temperature and high-pressure environments. Although in the
start of the field, their primary gas separation (GS) application was for the enriching of uranium
hexafluoride, their exceptional chemical and mechanical properties make them very suitable for
a variety of other high-temperature applications in the clean energy industry, particularly
involving the production of H2 (such inorganic membranes and applications are the main focus of
this Thesis).
Following the IUPAC classification membranes (both polymeric and inorganic) can be classified
into three different groups: (i) Macroporous membranes that have an average pore diameter >50
nm and are used in microfiltration (MF) and ultrafiltration (UF) applications; (ii) mesoporous
membranes, which have an average pore diameter in the range of (2 - 50 nm) and they are used
in UF and nanofiltration (NF); and (iii) microporous membranes that have a pore diameter <2 nm
and are used mostly in NF, reverse osmosis (RO), and GS applications.
A ceramic membrane has, typically, an asymmetric structure and thus consists of multiple layers,
as shown in Figure 1-1. The very bottom part is a macroporous substrate, usually obtained by
isostatic cold-pressing, extrusion or slip-casting of a dry powder mixed together with binders
and/or sintering aids, followed by high-temperature sintering. These substrates usually have an
average pore diameter in the range of 1 - 15 µm and porosity in the range of 30% - 50%. On the
top of the above substrate one deposits a layer usually obtained by slip-casting and/or suspension
coating followed again by subsequent pyrolysis/sintering. This layer usually has an average pore
3
diameter in the range of 0.05 - 1 µm. Such two-layer membranes can be used either for MF
applications or as substrates for the deposition of additional layers to prepare UF, NF, RO and
GS type membranes. For UF applications, one deposits an additional separation layer (typically
via slip-casting, spin-coating or sol-gel techniques) with an average pore diameter in the range 2
- 50 nm. To prepare NF, RO and/or GS membranes one deposits an additional layer with an
average pore diameter <2 nm on the top of or inside the UF layer either using sol-gel, chemical
vapor deposition (CVD)/chemical vapor infiltration (CVI) techniques or via the dip-coating of
thin polymeric precursor films, followed by controlled-temperature pyrolysis. For GS
membranes, the top layer could be a completely dense layer. For example, hydrogen separation
membranes are made by depositing on porous ceramic or stainless steel substrates of Pd or Pd-
alloy films via electroless plating (Mardilovich et al., 1998; Roa et al., 2003; Kulprathipanja et
al., 2005).
Figure 1-1. Schematic illustration of a typical ceramic membrane (Li, 2007)
Pressure-driven membrane processes (e.g., microfiltration, ultrafiltration, and nanofiltration) are
used to concentrate/purify a dilute solution/gas mixture. Different processes are distinguished
from each other based on the particle or solute size in solution or the molecular size in the gas
4
mixture; different kinds of membranes utilized in these processes are differentiated from each
other based on their average pore size and pore size distribution (PSD). From MF to NF
applications, the pore size of the membranes used decreases from ~10 µm to ~1 nm. The
application range of different pressure-driven membrane process is shown in Figure 1-2, and the
comparison of different pressure-driven membrane processes is shown in Table 1-1.
Figure 1-2. Application range of different pressure-driven membrane process (Mulder, 1996)
5
Table 1-1. Comparison of different-pressure driven membrane process (Mulder, 1996)
Porous inorganic membranes are commonly made from metal oxide ceramic materials (e.g.,
alumina, silica, zirconia, titania) and are either amorphous or polycrystalline (e.g., zeolite
membranes) (Kusakabe, Yoneshige, Murata, and Morooka, 1996; Dong, Lin, Hu, Peascoe, and
Payzant, 2000; Braunbarth, Boudreau, and Tsapatsis, 2000), though recently non-oxide ceramics
(e.g., nitrides or carbides) are also finding applications. As mentioned above, ceramic
6
membranes exhibit potential advantages over polymeric membranes in terms of their superior
thermal, chemical, and mechanical stability. Thermal stability relates to the high melting point of
ceramics. For example, titania has a melting point of 1605
o
C, alumina 2050
o
C, silicon carbide
2500
o
C, and zirconia 2770
o
C. Their chemical stability makes it possible to use ceramic
membranes with almost any organic solvent. And also, because of their chemical stability, many
strong acid and alkali detergents can be used in cleaning such membranes for applications where
fouling is a concern; The superior thermal and mechanical stability of inorganic membranes are
of particular value for their use in reactive separation applications (otherwise known as
membrane reactors) that take place at high temperatures and pressures, or under corrosive
conditions (e.g., the methane steam reforming reaction for the production of hydrogen).
1.2 Silicon Carbide Membranes
As noted above, membrane separation technology has attracted a lot of research attention in the
past 30 years due to its low energy requirements compared to conventional separation
technologies, such as distillation, centrifugation, crystallization, adsorption, and sublimation.
Polymeric membranes have been extensively investigated for many years and are now widely
used in process applications like UF, pervaporation (PV), and desalination. Inorganic membranes,
on the other hand, have received significantly less attention, despite the fact that they show
substantial promise for a broad range of applications. As noted above, one of the drivers for the
further development of inorganic membranes is the efficient production of hydrogen, whose
demand has been increasing recently due its potential use for clean energy generation. Process
intensification during hydrogen production involves a separation step under high-temperature
7
and high-pressure environments, where conventional polymeric membranes cannot be used.
Among the inorganic membranes proposed for such application, silicon carbide membranes are a
particularly promising candidate because of their many desirable properties like high corrosion
resistance, high thermal conductivity, high thermal shock resistance and excellent chemical and
mechanical stability.
The preparation methods of SiC membranes consist of two different approaches: One involves
the pyrolysis of polymeric precursors like polycarbosilanes (PCS) and allyl-
hydropolycarbosilanes (AHPCS), and the other is based on the CVD/CVI techniques via the use
of appropriate precursors like tri-isopropylsilane (TPS).
1.2.1 Pyrolysis of Polymer Precursors
One of the most important and widely used polymeric precursors in making SiC materials is
polycarbosilane (PCS). In the preparation of SiC membranes, in order to achieve the pyrolysis
process at a relatively low temperature and to minimize the oxygen content in the final product,
our group has previously utilized a partially allyl-substituted hydropolycarbosilane (AHPCS).
The molecular structures of HPCS and AHPCS are shown in Figure 1-3,
8
Figure 1-3. Molecular structures of AHPCS and HPCS
Hasegawa and coworkers (Yajima et al., 1978; Hasegawa et al., 1980; Hasagawa and Okamura,
1983; Hasegawa and Okamura, 1986; Hasegawa, 1989) systematically investigated the synthesis
of silicon carbide via the pyrolysis of PCS as a precursor. They synthesized a PCS precursor by
thermal deposition of polydimethylsilane in an Argon atmosphere at 450
o
C - 470
o
C for 14 hr.
Afterwards, they converted the PCS into a silicon carbide fiber, via pyrolysis at temperatures of
up to 1350
o
C, in a vacuum environment and characterized various samples at different stages of
pyrolysis. Bouillion et al. (1991) studied the pyrolysis of PCS in detail and found out that the
mechanism of pyrolysis consists of three different stages: first, from 550
o
C to 800
o
C, an
organometallic mineral transition leads into an amorphous hydrogenated solid containing
9
tetrahedral SiC, SiO2 and silicon oxycarbide (SiOC) moieties; then, from 1000
o
C to 1200
o
C, a
nucleation of the SiC moieties results into SiC nuclei surrounded with aromatic carbon layers;
lastly, from 1400
o
C to 1600
o
C, a SiC grain-size coarsening consumes the residual amorphous
phases and simultaneously causes the evolution of SiO and CO. Li et al. (1996) prepared
amorphous Si-C-O membranes on the outer surfaces of α-alumina supports by dip-coating with
p-xylene solutions of polycarbosilane and subsequent pyrolysis. At 773 K this membrane had a
H2 permeance of 1×10
-8
mol•m
-2
•s
-1
•Pa
-1
, and its (H2/N2) separation factor was in the range of
(18 - 63). The addition of polystyrene (PS) as a pore former increased the permeance of
hydrogen by an order of magnitude, but decreased the separation factor to 13. Lee et al. (2001)
fabricated SiC membranes from pyrolysis of poly(dimethylsilane) at 873K under an Argon
atmosphere on SiC porous supports. At 473 K, their optimal membranes exhibit a H2 permeance
of 2.7×10
-9
mol•m
-2
•s
-1
•Pa
-1
and a H2/N2 separation factor of 20. Ciora et al. (2004) prepared
nanoporous SiC membranes by pyrolysis of AHPCS, and these membranes were stable in air at
450
o
C. However, they were unstable in the presence of high-temperature steam. Iwamoto et al.
(2005) prepared amorphous silica membrane by pyrolysis of polysilazane in air on silicon nitride
supports through multiple heat treatment stages with peak temperature of 873 K. Their
membranes exhibit a H2 permeance of 1.3×10
-8
mol•m
-2
•s
-1
•Pa
-1
at 573 K, and a separation factor
of H2/N2 is 141. Suda et al. (2006) investigated the structural evolution during conversion of
PCS into SiC-based microporous membranes. They found that the cross-linking reactions
resulted in a three-dimensional pore network consisting of small micropores having a large
surface area. Nagano et al. (2006) prepared amorphous SiC membranes with molecular sieving
properties on the outer surfaces of γ-alumina-coated α-alumina porous supports through the
pyrolysis of PCS without an oxygen-curing process. At 873 K, this amorphous SiC membrane
10
resulting from PCS pyrolysis at 1073 K had an H2 permeance of 8.1×10
-7
mol•m
-2
•s
-1
•Pa
-1
, with a
(H2/CO2) separation factor of 13. Wach et al. (2007) reported making SiC-based membranes by
pyrolysis at 1123 K in Argon of a polymeric precursor film (a mixture of 80wt% PCS and
20wt% polyvinylsilane) deposited onto porous γ- alumina supports. Their membranes achieve a
10
-10
to 10
-8
mol•m
-2
•s
-1
•Pa
-1
H2 permeance and a (H2/N2) separation factor of over 250. Elyassi
et al. (2007) prepared nanoporous SiC membranes on porous SiC supports by slip-casting of
small-size SiC powder (~100 nm in diameter) in AHPCS, followed by AHPCS dip-coating and
subsequent pyrolysis. At 473 K, this membrane had a He permeance of 8.9×10
-9
mol•m
-2
•s
-1
•Pa
-1
and a (He/Ar) separation factor of 147. After 2 hr of oxidation in air at 450
o
C, the permeance of
He increased to 1.2×10
-8
mol•m
-2
•s
-1
•Pa
-1
, and the (He/Ar) separation factor decreased to 89. In a
later study, Elyassi et al. (2008) prepared better quality SiC membranes by using polystyrene as
the pore former during the dip-coating process; these modified SiC membranes had a He
permeance of 1.8 - 4.3×10
-8
mol•m
-2
•s
-1
•Pa
-1
, and their (He/Ar) separation factor was in the range
of 176 to 465.
1.2.2 Chemical Vapor Deposition/Infiltration
Chemical Vapor Deposition (CVD) is a technique which is widely used in the semiconductor
industry to prepare high-purity thin films. It is a very mature technology and has been used in the
industry for a long time. It has also been utilized to prepare supported SiC membranes in the past
15 years. Sea et al. (1998) prepared SiC membranes by CVD of tri-isopropylsilane (TPS) at 700-
800
o
C inside the macropores of α-alumina support tubes via the forced-flow through the porous
tube. These membranes permeated various gases, other than water, mainly via the Knudsen
11
diffusion mechanism at 50 to 400
o
C. After being heated in Ar at 1000
o
C for 1 hr, these
membranes exhibited a H2 permeance of 5 - 6×10
-7
mol•m
-2
•s
-1
•Pa
-1
at 50 to 400
o
C. For the H2-
H2O-HBr gas mixture, at 400
o
C, the membranes maintained a modest (H2/H2O) selectivity for
more than 100 hr.
Lee et al. (1998) modified asymmetric SiC membranes via the low-pressure chemical vapor
deposition (LPCVD) of SiH4 and C2H2 at a temperature of 800
o
C using Ar as acarrier gas. As
the reaction proceeds, the gas permeance decreases from 10
-5
to 10
-7
mol•m
-2
•s
-1
•Pa
-1
at an
average measurement pressure of 2.0 × 10
5
Pa. They also investigated the variation of flow-
average pore radius, film growth rate, and film thickness change as the reaction takes place.
Takeda et al. (2001) prepared SiC membranes by CVI of SiH2Cl2 and C2H2 at 800
o
C to 900
o
C
on α-alumina supports. The CVI process was carried out by repeatedly alternating the supply of
source gases. They report that the membrane’s permeance decreases as the reaction proceeds and
more pores in the membrane structure are blocked. Their SiC membranes show a H2 permeance
of ~1×10
-8
mol•m
-2
•s
-1
•Pa
-1
and a (H2/N2) separation factor of 3.36 at 350
o
C.
Kleps et al. (2001) studied the preparation and properties of amorphous SiC films made by CVD.
They found that deposition at moderate temperature leads into stoichiometric SiC and deposition
at high temperature leads into Si0.25C0.75 with extra carbon residues. They further reported that
the field emission current density of the a-SiC/Si structures was 2.4 mA/cm
2
at 25 V/µm. Chen et
al. (2004) prepared SiC membranes by CVD of TPS on porous SiC supports. The SiC membrane
had a He permeance of 2.6×10
-7
mol•m
-2
•s
-1
•Pa
-1
and a (He/Ar) separation factor of ~13; the
properties of the resulting membranes depended strongly on the CVD conditions, including the
TPS injection rate, the flow rate of the carrier gas (Ar) and the temperature.
12
Nagano et al. (2006) prepared SiC membranes on the outer surfaces of γ-alumina-coated α-
alumina porous supports by both pyrolysis of polycarbosilanes (PCS) without an oxygen-curing
process and chemical vapor infiltration. It was shown that CVI was a good technique to increase
the membrane selectivity towards gases with smaller kinetic diameters. At 873 K, these
membranes had a He permeance of 7.7×10
-8
mol •m
-2
s
-1
Pa
-1
, a (He/CO2) separation factor of 64,
and a (He/H2) separation factor of 4.4.
1.3 Porous Silicon Carbide Supports
For the preparation of SiC membranes, with the exception of our group, most previous
researchers utilized porous alumina supports. There are a number of problems that result from
the use of such supports. First, alumina supports (particularly mesoporous γ-alumina) have
questionable stability in high-temperature steam. In addition, since alumina and SiC have
different thermal expansion coefficients, this creates challenges both during the preparation of
such membranes (e.g., the pyrolysis and curing steps) and during thermal cycling through their
use. An alternate is the use of porous SiC supports, as noted above, which have exactly the same
thermal expansion coefficient with the top SiC membrane films.
In most industrial processes and applications, the interest is in the preparation of SiC dense
materials. These are made by pressing and sintering of SiC powders and are used as high-
temperature heating elements, brake backing plates for cars and airplanes, parts of mechanical
tools, gas turbine elements, etc. However, dense SiC materials are not appropriate as supports of
SiC membranes, since they do not let gases flow through. Thus, macroporous SiC materials are
needed as supports in making SiC membranes. It is a challenging task to make such materials,
13
since if their porosity is too small, they will have an undesirably small permeance, and if their
porosity is too large, their mechanical properties will be compromised. Generally, the preferred
porosity range for such supports is (30% - 45%).
Generally, without the addition of appropriate sintering aids (e.g., boron carbide, phenolic resin,
alumina or carbon black), it is very difficult to sinter SiC materials and to prepare good quality
supports. However, by adding appropriate sintering aids, either individually or in combination,
significant densification and sintering of the SiC materials can take place. The amount of
sintering aids/additives affects the quality of the final sintered SiC materials. The higher the
quantity of sintering aids that is added to the SiC powder, the more densified and less porous the
materials will be after sintering. This often means that one must choose the optimal
concentration of sintering aids, as too much will result in complete densification, while too little
will prepare materials with inferior mechanical properties.
Our research group has successfully prepared SiC porous supports by pressureless sintering of
SiC powders in an inert Helium atmosphere. We have investigated the effect of using different
starting powders and sintering aids on the final quality of the SiC supports after sintering. It was
found that the porosity, pore size distribution, surface characteristics and structure of the
supports after sintering depended on the type of SiC powder and the amount of sintering aids
utilized. Fukushima et al. (2006) subsequently also prepared SiC porous supports with different
amount of alumina additive by pressureless sintering. They found that pore size and particle size
of supports without alumina additives increased as the sintering temperature increased. However,
for SiC supports with alumina additives, small particles and small size pores were retained after
sintering. In a later study, they fabricated SiC supports via the use of nano-sized SiC powders,
and they report that the compressive strength of the supports increased as the sintering
14
temperature increased. They attributed this phenomenon to the formation of well-developed neck
areas. Most recently, they also utilized Al2O3 and Y2O3 as sintering aids. They reported that
supports prepared with 3 wt% Al2O3 as a sintering aid had a large gas (N2, He and H2)
permeance ranging from 10
-6
- 10
-5
mol•m
-2
•s
-1
•Pa
-1
. Kim et al. (2009) also studied the control of
porosity and microstructure of porous SiC ceramics. By adjusting the template content,
processing parameters, filler content and the sintering additive content, they could effectively
control the porosity of SiC ceramics in the range of 35% - 95%.
1.4 Transport Phenomena in Membrane Separation
Depending on their formation conditions, SiC membranes can have very different pore structures.
In addition, they could be either amorphous or crystalline. As noted above, the technical
literature reports the preparation of nanoporous (pore size <2 nm), mesoporous (2 - 50 nm) and
macroporous membranes (>50 nm) SiC membranes and substrates. In order to illustrate the
transport phenomena that take place in SiC membranes, an idealized case of a membrane
consisting of parallel cylindrical pores is shown in Figure 1-4 to represent the pore structure in
real SiC membranes. The various cases in this figure (from top to bottom) range from situations
where the pore diameter (dp) is much larger than the mean free path (λ) of the diffusing species,
to the opposite end of the spectrum where the dp is much smaller than λ. For the various cases in
Figure 1-4, going from top to bottom, the flux decreases while the selectivity (separation factor)
increases. Other than the last, the so-called solution/diffusion mechanism, all other transport
mechanisms are present during gas separations using SiC membranes. Though we know of no
examples of dense-phase SiC membranes, there are numerous examples of polymeric as well
15
inorganic membranes (Pd and Pd-alloy) where solution/diffusion is the prevailing transport
mechanism, however.
Figure 1-4. Pore diffusion effects (Cussler, 2009)
The membrane permeance is the parameter often used to characterize the permeation
characteristics of a given membrane, and is defined by equation 1-1, as follows:
16
Flux n
Permeance
P PA
= =
∆ ∆×
(1-1)
For some of the transport mechanisms in Figure 1-4, simple equations are available to describe
the permeance as follows (for the definition of the various symbols, see the Nomenclature
Appendix at the end of the Thesis):
Viscous Flow
4
2
8
p
av
r
nP Flux
Permeance
P P A RAl T
υ
π
µ
= = =
∆ ∆×
g
(1-2)
Bulk Diffusion
1
3
1 3
2
2
2
8
p
b
r
n Flux
Permeance T
P P A MRl A
π
−
= = =
∆ ∆×
gg
(1-3)
Knudsen Flow
1
3
1
2
2
2
32
9
p
k
r
n Flux
Permeance T
P P A MRl A
π
−
= = =
∆ ∆×
gg
(1-4)
Surface Flow
2
2
1000
p app
s
Rs
A RT
n Flux x
Permeance dp
P P A PA C S l p
ρ
τ
= = =
∆ ∆× ∆
∫
(1-5)
17
For most SiC membranes, the different transport mechanisms, shown in Figure 1-4, occur
simultaneously. In addition, the pores in real silicon carbide membranes are neither straight nor
uniform in size. Therefore, there are no simple equations available to describe the permeance of
such membranes.
18
2. Highly Permeable Porous Silicon Carbide Support Tubes
2.1 Introduction
In the production of hydrogen as a clean fuel, membrane-based reactive separation technologies
may have an important role to play as they provide for process intensification (Liu et al., 2012),
whereby reaction and separation processes are combined in the same unit, with the synergy
created significantly improving the efficiency and economics of the overall system. Hydrogen
production via either the steam or the autothermal reforming of hydrocarbons or via the
gasification of coal and/or biomass imposes stringent requirements on the properties of the
membrane materials required because of the high temperatures and pressures utilized. Of all the
potential membrane materials for such applications, silicon carbide (SiC) is particularly
promising due to its high temperature and corrosion resistance, high thermal conductivity, high
thermal shock resistance, and excellent chemical and mechanical stability.
One key step in the preparation of quality microporous asymmetric inorganic membranes, in
general (Elyassi et al., 2009; Suwanmethanond et al., 2000), and of SiC nanoporous membranes
(Ciora et al., 2004; Elyassi et al., 2007; Elyassi et al., 2008), in particular, is the preparation of
appropriate porous supports. These supports are called upon to fulfill a very important role,
which is to provide a strong mechanical template on which the real permselective layer will be
deposited. And while mechanical strength is a key need, these supports must also satisfy a
number of other important requirements. They must have, for example, high flux so that
resistance to transport through the support does not negatively impact the overall membrane
throughput. And their surface characteristics (including their surface roughness) must be such so
that they readily permit the deposition of thin nanoporous films; they must be, in addition,
19
resistant to high temperatures and pressures, to the corrosive atmospheres which are encountered
during hydrogen production (e.g., steam), and to thermal and pressure cycling.
Overwhelmingly, the supports of choice for the preparation of microporous inorganic
membranes, in general, and prior to our own (Suwanmethanond et al., 2000) studies for the SiC
nanoporous membranes as well, are either macroporous α-alumina or mesoporous γ-alumina
materials. The main reason for the choice of both these types of supports is their easy
commercial availability. Mesoporous γ-alumina which is, by far, the most frequent choice as a
support material has, in addition, a very smooth surface and is fairly inert; this then makes it
desirable as a substrate for the deposition of high-quality thin films of a variety of materials (e.g.,
Pd, microporous silica, carbon molecular sieves, zeolites, etc.). The same is not true, however,
for other commercial supports such as, for example, porous stainless steel which has a rough
surface and is fairly reactive (e.g., causing intermetallic diffusion into Pd layers), thus requiring
substantial and elaborate surface pre-treatment prior to the deposition of the top metallic films
(Samingprai et al., 2010; Ayturk et al., 2006).
Mesoporous γ-alumina membrane supports, however, face important challenges of their own in
reactive separation applications relevant to hydrogen production because of their questionable
stability under high-temperature and high-pressure steam conditions. In addition, since the
alumina and some of the top layers deposited (e.g., Pd) have quite different thermal expansion
coefficients, repeated thermal cycling presents another key challenge as well. For SiC
nanoporous membranes, in addition to the aforementioned hurdles, the use of mesoporous γ-
alumina supports faces challenges during the preparation step (either via CVD/CVI or via the
deposition and pyrolysis of pre-ceramic precursors) (Chen et al., 2008) because it takes place at
substantially elevated temperatures.
20
To overcome these technical challenges during the preparation of SiC nanoporous membranes,
our team has substituted the mesoporous γ-alumina supports, used in our early efforts (Ciora et
al., 2004), with macroporous SiC supports (Elyassi et al., 2007; Elyassi et al., 2008). The
rationale for using such supports should be obvious considering their thermal and mechanical
stability and, of course, the closeness in the thermal expansion coefficient with the top
microporous SiC layer. These SiC macroporous sintered supports are made via the pressureless
sintering of SiC powders in an inert Ar atmosphere and their preparation method was detailed in
an early publication by our group (Suwanmethanond et al., 2000). There, we presented the
results of our study on the effect of the type of SiC powders and sintering aids utilized on the
porosity, pore size distribution (PSD) and structure of the resulting supports, and on their
transport characteristics determined via single-gas (N2 and Ar) permeation studies. As expected,
the structural and transport characteristics of these supports depend sensitively on the type of SiC
powders and the amount and kind of sintering aids utilized. A key conclusion of this early study
was that with the appropriate choice of starting powders, sintering aids and conditions,
mechanically strong supports could be prepared with a broad range of average pore sizes,
ranging from the mesoporous to the macroporous.
In the ensuing years, our efforts on preparing SiC supports have continued primarily focused on
the ability to use such supports to prepare high-quality nanoporous membranes (Ciora et al.,
2004; Elyassi et al., 2007; Elyassi et al., 2008) and the results of these studies are summarized in
this paper. Before doing so, however, we first briefly review here published efforts by the other
groups during the same period to prepare porous SiC materials. Sung et al. (2002) prepared
macroporous SiC materials with highly ordered pore structures for the potential application as
membrane supports. To do so, they used highly ordered sacrificial colloidal monodisperse silica
21
spheres (500 nm in diameter) as templates, which they infiltrated with low molecular weight SiC
pre-ceramic polymeric precursors, including polymethylsilane (PMS) and polycarbosilane,
followed by pyrolysis at different temperatures in an argon atmosphere. The resulting composite
structures were subsequently etched with HF in order to remove the silica microspheres. The SiC
materials thus produced were highly porous with a total pore volume of 0.24 - 0.26 cm
3
/g, and
had a very narrow pore size distribution which peaks at ~340 nm, somewhat smaller than the size
(500 nm) of the silica microspheres themselves. The approach of using sacrificial silica
microsphere as templates is interesting and quite capable of preparing materials with hierarchical
pore structures. It is unclear, however, how it can be scaled-up (e.g., the use of the highly
corrosive HF is likely to present key challenges) to prepare cylindrical, large-area, inexpensive
membrane supports.
Zhang et al. (2009) prepared porous SiC ceramics for their potential use in a broad range of
applications, such as grinding media, porous filters, catalytic substrates, thermal insulators, high-
temperature structural materials, and reinforcement media for the production of composites. To
prepare these SiC ceramics, they first mixed mesophase pitch with Si particles (50 µm average
diameter) and heated them at 500
o
C for 1 hr in high-pressure (5 MPa) nitrogen The resulting
templates were then infiltrated by molten Si at 1200
o
C and then treated at 1500
o
C for 4 hr in an
argon atmosphere so as to produce SiC. The resulting material was lastly treated with a (70%
HF/30% HNO3) solution to eliminate any unreacted silicon. The final SiC materials had a
porosity of 55%. As noted previously, though interesting, it is unclear how this technique could
be conveniently adopted to prepare inexpensive, large-area tubular membrane supports.
Schmidt et al. (2001) produced strong porous silicon oxycarbide (SiOC) bulk ceramics by
pyrolysis of PMS resins at 600 - 1200
o
C in flowing nitrogen. They also prepared SiOC ceramics
22
with cellular structures by first preparing (via heating of the various ingredients together) a
mixture consisting of a commercial silicone (SR 350, GE Silicone Products), dichloromethane,
polyurethane, two different kinds of amine catalysts (bis(dimethylamino)ethyl ether dissolved in
di(propylene glycol and triethylenediamine dissolved in di(propylene glycol), and a surfactant
(poly(dimethylsiloxane)copolymer), and then pyrolyzing the mixture under the same conditions
above. The resulting materials had a total pore volume ranging from 0.0013 to 0.3365 cm
3
/g,
and an average pore diameter ranging from 1.77 - 15.1 nm, depending on the different
combinations of starting materials.
Eom et al. (2008) synthesized porous SiC ceramics from carbon-filled polysiloxane using hollow
microspheres as sacrificial templates. To prepare these ceramics, a number of ingredients
including polysiloxane resin (solid silicon resin, density =1.036 g/cm
3
) powder, carbon black,
ultrafine SiC powder (<100 nm in diameter) as fillers, hollow (poly (methyl methacrylate)
microspheres (15 µm to 25 µm in diameter), Al2O3 and Y2O3 (as sintering aids) were mixed
together milled, dried, and pressed into rectangular bars. These were then heated in an argon
atmosphere (following a certain heating protocol) to 1100
o
C so as to convert the polysiloxane
resin into silicon oxycarbide; subsequently, they were further heat-treated to 1450
o
C so that the
SiOC reacted with the carbon black to produce the SiC ceramics; finally, the pyrolyzed
specimens were annealed at 1800 - 2000
o
C for the liquid-phase sintering of SiC using Al2O3 and
Y2O3 as sintering aids. The resulting SiC materials had a ~40% porosity and ~240 MPa of
compressive strength.
The same group (Eom et al., 2008) also reported the preparation of SiC porous ceramics from
SiC powders, using Y3Al5O12, as a sintering aid, poly-(methyl methacrylate-co-ethylene glycol
dimethacrylate) microbeads as templates (three different sizes, 8 µm, 20 µm, and 50 µm in
23
diameter). The above mixture of ingredients was similarly milled, dried and pressed into
rectangular bars, and was subsequently pyrolyzed at 1950
o
C in Argon. Eom and Kim reported
(Eom et al., 2008) that, as expected, the porosity of the resulting ceramics increased as the
template content increased, and the pore and grain size increased as the size of microspheres
increased. Kim and coworkers (Eom et al., 2008; 2009; 2010; Chae et al., 2009) also
investigated the effect of additive composition and inert filler addition on the microstructure and
strength of porous SiC ceramics prepared by a similar method as aforementioned. Their results
showed that the porosity of the sintered SiC materials ranged from 56% to 72%, depending on
the amount and composition of additives. They also reported that the addition of polymer
microbeads as inert fillers led to a finer microstructure in the final product, resulting in higher
strength, at an equivalent porosity, than the specimens without fillers. Chae et al. (2009) report
that they can effectively control the porosity of SiC ceramics prepared by the same method as in
in the range of 35% - 95% by adjusting the template, filler and sintering additive content, and the
other processing parameters.
Gupta et al. (2004) synthesized high surface area and porosity SiC materials by a modified sol-
gel method for potential applications as catalyst and sorbent supports. The preparation of the gel
involved mixing the pre-ceramic precursor (phenyltrimethoxysilane) with water, hydrochloric
acid and methanol for the hydrolysis reaction to take place. Addition of NH4OH or NaOH
coagulated the growing hydrolyzed precursor particles in the solution. Once the gel was formed,
the supernatant solution was drained-off and the gel was then rinsed with water, followed by
vacuum drying. In the step of pyrolysis, dry gels were heat-treated at 1500
o
C in vacuum. The
resulting sintered SiC material had surface areas of 450 - 620 m
2
/g and pore volumes of 0.37 -
0.45 cm
3
/g.
24
For potential applications as ceramic filters and catalytic supports, Kanzaki and coworkers (Yang
et al., 2003) synthesized porous SiC materials by sintering pressed powder compacts of α-Si3N4
(0.5 µm in diameter), and carbon black (13 nm in diameter), at a Si3N4:C = 1:3 molar ratio, and
various sintering aids (4wt% Y2O3 and 6wt% Al2O3). During preparation, they sintered the
materials first at 1600 °C for 2 - 8 hr for the carbothermal reaction to occur between Si3N4 and
carbon black to produce SiC, and then at 1700 °C - 1900 °C for another 2 hr in Ar. The resulting
SiC materials had porosities ranging from 45% to 65%, and sharp pore size distributions, with
peaks ranging from 0.5 µm to 0.8 µm. Unfortunately, these high-porosity materials were not
tested as membrane supports, and no other information was provided about their hydrothermal
stability.
Using cold isostatic pressing of β-SiC powders (0.3 µm in diameter) and the pressureless
sintering method, Fukushima et al. (2006) prepared disk-shaped porous SiC supports with
different amounts of alumina as a sintering additive (0wt% - 4wt%) for the preparation of
membrane supports in hydrogen production. They reported that the pore size and microstructure
of these porous SiC supports could be controlled by varying the sintering temperature, the
pressing pressure, and the amount of alumina additive. The pore size and grain size of SiC
supports prepared without the alumina additive all increased as the sintering temperature
increased from 1500 °C to 1800 °C. Fukushima et al. (2006) also investigated the oxidation
behavior of these supports under water vapor conditions. These tests were performed by
exposing the disks to a flowing mixture of N2/H2O (partial pressure ratio = 1/3) for a period of
100 hr at 600 °C or 1000 °C. Scanning electron microscopy (SEM) and mercury porosimetry
measurements indicated that the porous SiC materials prepared without the alumina additive
were very stable in the presence of high-temperature steam, while the samples prepared with
25
alumina as sintering aid showed an increase of average pore size, and the disappearance of fine
SiC grains as a result of the hydrothermal treatment. Follow-up studies by the same group
(Fukushima et al., 2008) of these supports in the presence of the same N2/H2O mixture showed
that they are suitable for use at 600 °C, but unsuitable for use at 1000 °C where grain coarsening
took place, and their porosity decreased sharply.
The impact of sintering aids on the preparation of SiC supports was further investigated in a
more recent study by the same group (Fukushima et al., 2009) whereby Al2O3 and Y2O3 were
utilized as sintering aids to prepare supports via pressureless sintering of 0.3 µm SiC powders in
the temperature range of 1500 °C - 1800 °C. As our group also previously reported
(Suwanmethanond et al., 2000), they found that the amount of sintering aids determines the pore
size and the microstructure (such as fine grains and small pores) of the sintered supports.
Furthermore, the pore size and grain size of the supports increased with the increase in sintering
temperature.
The same group (Fukushima et al., 2008) also fabricated disk-shaped SiC supports via sintering
of β-SiC powders (0.3 µm in diameter), using nano-sized β-SiC powders (30 nm in diameter) as
additives; they reported that the compressive strength of these supports increased as the sintering
temperature increased from 1500 °C to 1800 °C, and they attributed this phenomenon to the
formation of well-developed neck areas, as observed with SEM. In their most recent study,
Fukushima and coworkers prepared tubular SiC porous supports via sintering of β-SiC powders
(0.3 µm in diameter) and α-Al2O3 (0.17 µm in diameter, as sintering aids) together at three
different temperatures (1800 °C, 1850 °C, and 1900 °C) and tested their transport characteristics
by single-gas (N2, H2, and He) permeation experiments. The gas permeance of these supports
ranged from 10
-6
to 10
-5
mol•m
-2
•s
-1
•Pa
-1
, and the (He/N2) separation factor (ratio of single-gas
26
permeabilities) ranged from 1.11 to 2.29 as the measuring temperature increased from 100 °C to
500 °C. They also reported that these tubular supports showed no strength degradation and
limited oxidation during the water vapor corrosion test at 800 °C for 100 hr.
In summary, since the original publication by our group (Suwanmethanond et al., 2000), a
number of papers have appeared on the preparation and characterization of porous SiC ceramics,
but we are not aware of any of these ceramics being used as supports in the preparation of
nanoporous inorganic membranes. For these materials to be appropriate for such an application,
their surface characteristics and roughness must allow for the deposition of thin nanoporous
films, and their permeation flux must be high enough so that it does not negatively impact on the
permeation characteristics of the resulting membranes. We have previously reported on the use
of our own macroporous SiC supports for the preparation of highly permselective SiC
nanoporous membranes (Elyassi et al., 2007; Elyassi et al., 2008). In this paper, the focus is on
the supports themselves, in particular on the effect of the choice of starting powders, sintering
temperature, the amount of sintering aids and the surface treatment on the surface roughness and
transport characteristics of the resulting SiC porous sintered supports. One key emphasis in our
research has been on improving the permeance of these supports. We currently prepare
membrane supports whose He permeance is almost 2 orders of magnitude larger that the supports
in our original publications (Suwanmethanond et al., 2000; Ciora et al., 2004), and 3 - 10 times
larger than the permeance of the porous SiC supports of Fukushima and coworkers (Zhou et al.,
2011).
27
2.2 Experimental Section
In this study, two different types of commercially available β-SiC powders (HSC-059N and
HSC-1200, from Superior Graphite Co., Chicago, IL) were used in the preparation of the porous
tubular sintered SiC macroporous supports. The manufacturer reports that the HSC-059N powder
has an average particle size of 0.6 µm, while the average particle size for the HSC-1200 powder
is 6 µm. In order to prepare SiC tubular porous supports, these two β-SiC powders were first
mixed together at different weight ratios. Boron carbide (from Alfa AESAR®, Ward Hill, MA)
and phenolic resin (Durez Varcum 29353, available by TLC Ingredients, Inc., Crest Hill, IL –
note that in contrast to prior studies (Prochazka et al., 1974; Hojo et al., 1991; Tanaka et al.,
1991), a liquid phenolic resin rather than a solid powder phenolic resin, is used here, as it appears
to be a significantly more effective sintering aid) were then added to the various mixtures of
powders as sintering aids, using acetone as the dispersing medium. The resulting slurry of
powders and sintering aids was then fully mixed in an ultrasonicator for 1 hr, and subsequently
dried in a fume-hood for three days. Afterwards, the dried slurry was thoroughly ground in a
mortar, and small amounts of oleic acid and toluene were added to it as pressing aids. These
powders were then pressed into the tubular green support samples with a pressure of 6.68 MPa
for 2 min. They were subsequently placed in a high-temperature graphite furnace (Thermal
Technology, Inc., Model 1000-3060-FP20), where they were heated (3 °C/min) in flowing He to
a pre-set sintering temperature, where they stayed for 3 hr, and then cooled down (5 °C/min) to
room temperature. These sintered tubular SiC supports were then taken out of the graphite
furnace and subsequently treated in air in a box-furnace at 450 oC in order to oxidize any
potential carbon residues. Some of these supports were further subjected to centerless grinding of
28
their surface (top 0.0127 cm thick layer removed) by the Peterson Precision Grinding Company
(El Segundo, California) using a Cincinnati Machines XG centerless grinder.
A photograph of several of these tubular supports is shown in Figure 2-1. They have an external
diameter of 1.27 cm, an inner diameter of 0.635 cm, and their length ranged from 5 to 8 cm.
Figure 2-1. Porous SiC tubular supports sintered at 1900
o
C for 3 hr
The permeation characteristics of each support were determined by measuring the flux (mol•m
-2
)
of two inert, non-adsorbing gases, namely He and Ar, and calculating their permeance (mol•m
-2
•s
-1
•Pa
-1
) according to the following equation 1-1.
Flux n
Permeance
P PA
= =
∆ ∆×
(1-1)
29
where n (mol•s
-1
) is the flow-rate of the gas (He or Ar) through the membrane support, ΔP (Pa) is
the transmembrane pressure difference, A (m
2
) is the outside area of the support available for
permeation. The ideal (He/Ar) separation factor is taken to be equal to the ratio of the measured
single-gas permeances.
A schematic of the permeation apparatus utilized is shown in Figure 2-2. For testing its
permeation characteristics, each tubular SiC support was affixed onto a fender-washer using a
high-temperature impermeable glue (J-B WELD 8625-S Cold-Weld Compound). The other end
was hermetically sealed using graphite tape and the same type glue. The membrane support was
then installed in between the two half-cells of the permeation-test unit (via the attached washer)
using O-rings for sealing. During the permeation test, the gas (He or Ar) is fed into the bottom
half-cell, and is forced to permeate through the membrane support into the top half-cell exiting
through the permeate port, with its flow being measured with a bubble-flow meter. The pressure
in the permeate-side was measured by a pressure gauge (OMEGA Engineering, DPG1000B-15G)
and was maintained atmospheric. The pressure in the feed-side was adjusted with the aid of
regulators of the upstream gas cylinders and a needle valve on the reject tube (see Figure 2-2). A
differential pressure transducer (OMEGA Engineering, PX409-050DDUI) was used to measure
the pressure difference between the feed and permeate sides. The test temperature in the
experimental system was kept constant via the use of ceramic heaters and a temperature
controller.
30
Figure 2-2. Schematic diagram of the permeation test apparatus
The porosity of the supports was measured by the Archimedes method, as detailed in our
previous work (Suwanmethanond et al., 2000), using acetone as the wetting liquid. The supports
were also characterized via scanning electron microscopy using a JEOL JSM-7100F thermal
field-emission electron microscope.
The average pore size (dp) of a porous support can be calculated via linear regression of the
experimental data using the following equation 2-1 and equation 2-2 (Neomagus et al., 1998),
av
1 48
Permeance = = +
ln( / ) 3
j
o
o
o oi j
N
B RT
P K
P r r r RT M µπ
∆
(2-1)
31
48
= =
/
oo
p
o
KB
d
K ετ
(2-2)
In the two equations above, Nj is the molar flux of species j (mol•m
-2
•s
-1
, w. r. t the outer surface
of a support), ΔP is the pressure difference across the support (Pa), ro is the outside diameter (m),
ri is the inside diameter (m), R is the gas constant (8.314 J•K
-1
•mol
-1
), T is the temperature (K), ε
is the porosity (dimensionless), τ is the tortuosity (dimensionless), μ is the gas viscosity (mPa•s),
Pav is average pressure across the support (Pa), Mj is the molecular weight of species j (kg•mol
-1
),
Bo is the permeation coefficient (m
2
), and Ko is the Knudsen coefficient (m). The basic
assumption in Equations 2-2 and 2-3 above is that the support pore structure consists of straight
non-intersecting cylindrical large aperture (length/diameter) pores all with the same diameter dp.
Since this is clearly not the case here, the values of dp should be interpreted as a “coarse
measure” of the characteristic pore size to be used only for qualitative comparisons among the
various supports (e.g., see Table 2-1).
Atomic force microscopy (AFM) was utilized to probe the morphology and to measure the
surface roughness of the supports. It was performed using a Digital Instruments’ Dimension
TM
3100 scanning probe microscope. Each measurement frame involved 512 sample lines and the
drive frequency ranged from 200 - 300 kHz. Typically, two parameters, Ra and Rq are used to
characterize the surface roughness. They are defined as,
1
1
=
n
ai
n
R Z
n
=
∑
(2-3)
32
2
1
1
=
n
qi
n
R Z
n
=
∑
(2-4)
In the above equations, Ra is the arithmetic average of the absolute values of the surface height
deviations (Zi) measured from the ideal plane surface, while Rq is the mean square root of the
height deviations (Zi) measured from the same plane surface. In this study, it is mainly Rq that is
used to study the effect of the different preparation parameters on the surface roughness of the
supports, as it appears to be a more sensitive indicator in the resulting changes, though both
parameters showed similar qualitative trends.
There are a number of tests available to characterize the mechanical properties of ceramic
materials, including the Vickers hardness test, the three-point and four-point flexural tests, and
the compressive strength test. For example, in order to study the strength of their porous SiC
supports, Fukushima and coworkers (Fukushima et al., 2011), Kumar et al. (2011) and Zhang et
al. (2009) all performed flexural strength tests. Fukushima coworkers (Fukushima et al., 2011),
for example, directly measured the ultimate load required to crush the tubes in the radial
direction. Kumar et al. (2011) performed four-point bending strength tests on their bar-shaped
supports. Zhang et al. (2009) performed similar three-point bending tests. In this study, we have
investigated the mechanical properties by performing uniaxial compressive tests along the axial
direction on cylindrically-shaped sintered support billets. All the tests were performed using an
Instron® 8800 servo-hydraulic fatigue testing system and were made in the displacement control
mode. For the measurements, the load was applied at a crosshead rate of 0.5 mm min-1 on the
cylindrical SiC support billets with a diameter of ca. 12.7 mm and a height of ca. 13.0 mm
(unless otherwise noted, two samples each prepared under identical conditions were utilized, and
33
the reported values reflect the average). These tests are very similar to the very recent
compressive strength test work on porous ceramic materials reported by Tulianni et al. (2013).
For porous ceramics, in particular, compressive strength, as measured by uniaxial compressive
tests (as in this paper), is one of the important mechanical parameters. The structural integrity of
a membrane subjected to internal or external gas pressure depends both on the compressive and
the tensile strength of the porous support materials. The tensile strength of a ceramic is
substantially more difficult to measure with high accuracy than the compressive strength; this is
because brittle fractures often develop close to the clamps due to stress concentration during the
mechanical strength tests. (Though it is rarely measured directly, most authors assume as a “rule
of thumb” that the tensile strength is ~ 10% of the compressive strength). Unfortunately, no
theory exists today that directly correlates such measurements with the resistance of the supports
to internal and/or external gas pressure. However, to the best of our knowledge, a standardized
test directly reproducing such a mechanism of failure for membrane supports is not currently
available. And, as is also true with the design of steel pressure vessels, most membrane
practitioners agree today that a direct correlation exists between the measured uniaxial
compressive strength (and the estimated tensile strength) and the allowable stresses and the
maximum allowable gas pressure leading to failure for a given membrane The key reason for
carrying out the standard compressive strength test for our supports is so that we have a basis of
comparison with published data (see further discussion to follow), using the same test protocol,
with other competitive membrane supports. Nevertheless, though no theory currently exists
capable to directly translate compressive strength into the exact conditions under which such
supports will fail during operation, it is also difficult to imagine a set of circumstances whereby a
34
material proven significantly more mechanically strong than another, via such a standardized
compressive strength test, will somehow prove inferior when used as membrane support.
2.3 Results and Discussion
The sintering of the SiC macroporous supports is a complex process during which new
connecting interfaces are formed among the various particles that touch each other as they are
being deformed. During sintering (Li, 2007), the “green” samples are reported to go through
different stages as the temperature increases. First, the chemically-bound water on the surface of
the particles vaporizes and the organic sintering aids added during preparation of the “green”
sample burn out; then connecting necks among particles grow rapidly via diffusion, vapor
transport, and plastic and viscous flow. The surfaces of the particles in the sintered supports
become smoother than those of the original supports and 3 - 5% volume shrinkage occurs by the
end of this stage. After the pores of the support tubes reach their equilibrium shapes, grains grow
slowly contributing to a decrease in the average porosity. If the sintering process is allowed to
continue at higher temperatures, the pores of the supports will continue to shrink and may
disappear altogether.
Desirable SiC membrane supports prepared in this study should be highly porous (so that they do
not negatively impact the membrane throughput), be mechanically strong (to withstand the
transmembrane pressure gradients which may be imposed during gas separation), and should
have smooth and crack-free surfaces, so that one is able to deposit on them thin nanoporous
membrane films. Thus, conventional approaches utilized in current industrial processes, which
aim to prepare dense SiC materials, are not necessarily well-suited for preparing high-quality
35
membrane supports. As detailed in our original publication (Suwanmethanond et al., 2000) and
in a number of the studies since (Chae et al., 2009; Fukushima et al., 2006; Fukushima et al.,
2008; Fukuashima et al., 2009; Zhou et al., 2011), preparing membrane supports with the desired
permeation and surface roughness characteristics entails choosing the optimal preparation
conditions, specifically the sintering temperature, the amount of sintering aids, and the
composition of starting powders. This is a challenging undertaking, as supports with large
permeance generally have a large surface roughness, which makes it more difficult to deposit
smooth and crack-free membrane layer on their top surface.
In this work, we systematically investigate the effects of the different preparation conditions,
including the sintering temperature, the type and concentration of sintering aids utilized, and the
composition of starting powders on the properties of SiC membrane supports. The goal is to
prepare supports with substantial throughputs that do not limit membrane performance but with
smooth enough surfaces that allow the preparation of such membranes in the first place. In what
follows, we first describe our studies on the impact of the various preparation conditions on the
permeation properties and surface characteristics of the resulting membrane supports. Then, the
use of these highly permeable supports in the preparation of nanoporous SiC membranes is
discussed and the characteristics of some of these membranes are detailed.
2.3.1 Effect of Sintering Temperature
Figure 2-3 shows the permeance as a function of the average pressure across the membrane for
supports which are made from blends (50/50) of the two SiC powders and using the same
amount of sintering aids (4wt% phenolic resin and 0.1wt% B4C) but sintered at different
36
temperatures (1700
o
C, 1800
o
C, and 1900
o
C). Note that as the sintering temperature increases,
both the Ar and He permeances increase but the effect saturates with increasing temperature.
Figure 2-4 shows the corresponding ideal separation factors which first increase as the
temperature increases, but they saturate out as well. The permeances for both gases have an
approximately linear dependence on the average pressure across the membrane and the ideal
separation factor decreases with average pressure. This is indicative of the fact that transport
through the membrane supports is via combined bulk (non-separatory) and Knudsen (slip) flow.
At much higher sintering temperatures (not studied here because of technical limitations with our
furnace), one expects these permeances to eventually begin decreasing as the pores begin to
shrink or disappear altogether and the porous SiC ceramic densifies.
37
1.08x10
5
1.17x10
5
1.26x10
5
1.35x10
5
1.44x10
5
1.0x10
-6
1.5x10
-6
2.0x10
-6
2.5x10
-6
3.0x10
-6
Ar 1900
o
C
Ar 1800
o
C
Ar 1700
o
C
He 1700
o
C
He 1800
o
C
Average Pressure / Pa
Permeance / mol*m
-2
*s
-1
*Pa
-1
He 1900
o
C
Figure 2-3. Permeance vs. Pav of blended powder (50/50) supports sintered at different
temperatures (Permeation data taken at T= 378 K, 4wt% phenolic resin and 0.1wt% B4C)
38
1.12x10
5
1.20x10
5
1.28x10
5
1.36x10
5
1.44x10
5
2.2
2.4
2.6
2.8
1900
o
C
1800
o
C
Average Pressure / Pa
Ideal Separation Factor
1700
o
C
Figure 2-4. Ideal separation factor vs. Pav of blended powder (50/50) supports sintered at
different temperatures (Permeation data taken at T= 378 K, 4wt% phenolic resin and 0.1wt%
B4C)
Figure 2-5 shows the AFM images of (50/50) blend supports sintered at three different
temperatures (1700
o
C, 1800
o
C, and 1900
o
C). AFM is well-suited as a technique to study the
surface morphology and roughness of these supports. The light spots and tips in the AFM images
in Figure 2-5 correspond to the convex parts on the surface and the dark spots correspond to the
concave ones. As one notes from Figure 2-5, the image of the support sintered at 1700
o
C has
many small distinguishable lighter spots, especially in the lowers parts close to the bottom;
comparatively, the image of the support sintered at 1800
o
C has fewer small distinguishable tips,
while most light tips in the image of the support sintered at 1900
o
C are large and there are not
39
many small distinguishable tips in the bottom. This means then that the small distinguishable tips,
representing the small particles on the surface, interconnect together and become relatively larger
tips as the sintering temperature increases and the support surface becomes coarser. This is also
indicated by the Rq of the three AFM images which increases from a value of 173 nm (for the
sample sintered at 1700
o
C), to 195 nm (for the sample sintered at 1800
o
C), to 225 nm for the
sample sintered at 1900
o
C.
1700
o
C (Rq = 173 nm) 1800
o
C (Rq= 195 nm) 1900
o
C (Rq= 225 nm)
Figure 2-5. AFM images of surfaces of blended powder (50/50) supports sintered at 1700
o
C,
1800
o
C, and 1900
o
C
Similarly with the AFM data, the SEM images shown in Figure 2-6 indicate that there are many
small coarse and distinguishable particles packed together or next to the larger particles in the
support sintered at 1700
o
C. In the SEM images of the supports sintered at the higher
temperatures (1800
o
C and 1900
o
C) these small particles begin to fuse together and to form
relatively smooth larger particles. The images of the supports sintered at the higher temperatures
contain a greater number of larger particles than the image of the sample sintered at 1700
o
C.
These larger particles result from the interparticle necks growth among the smaller particles
40
which eventually either fuse together to form larger particles or become part of a large particle
next to them. The picture that emerges from the SEM and AFM studies is consistent with the
results from the measurements of the porosity, average pore size, and transport characteristics of
samples sintered at various temperatures which are shown in Table 2-1. Though the porosity of
sintered supports decreases somewhat (~10%) indicative of a better packed solid powder,
counter-intuitively perhaps, the permeances for both Ar and He increase as the sintering
temperature increases, consistent with an increasing average pore diameter, dp, (ranging from
227 nm for the support sintered at 1700
o
C to 331 nm for the support sintered at 1900
o
C) but
also, perhaps, implying a much less tortuous (smaller tortuosity factor) pore structure consistent
with the disappearance of the smaller particles and the formation of larger and well intergrown
grains.
1700
o
C 1800
o
C 1900
o
C
Figure 2-6. SEM images of blended powder (50/50) supports sintered at 1700
o
C, 1800
o
C, and
1900
o
C
41
Sample
Sintering
Temperature
(
o
C)
Porosity
Average
Pore Size
(nm)
Compressive
Strength
(MPa)
He Permeance
(×10
-6
mol•m
-
2
•s
-1
•Pa
-1
)
Ar Permeance
(×10
-6
mol•m
-
2
•s
-1
•Pa
-1
)
S1 1700 36.8% 227 43.0 2.36 1.04
S2 1800 35.7% 257 82.9 2.91 1.14
S3 1900 33.1% 331 108.0 3.02 1.23
Table 2-1. SiC supports sintered at different temperatures (Sintering aids: 4wt% phenolic resin
and 0.1wt%B4C; composition: 50wt% 0.6 µm powder and 50wt% 6 µm powder; average
measurement pressure Pav=1.43×10
5
Pa; measurement temperature: 378 K)
2.3.2 Effect of the Concentration of Sintering Aids
Figure 2-7 shows the permeance of various (50/50) blended supports all sintered at 1900
o
C but
prepared with various amounts of sintering aids. Figure 2-8 shows the corresponding ideal
separation factors. As shown in Figure 2-7, as the amount of sintering aids decreases, the
support’s permeance significantly increases, as one transitions from the supports prepared with
the larger amounts of sintering aids to support tubes prepared without any sintering aids at all.
On the other hand, as Figure 2-8 indicates the ideal separation factors for the supports prepared
with smaller amounts of sintering aids are lower than those of samples prepared with the larger
content of sintering aids. This indicates that for the samples prepared with lower concentrations
of sintering aids transport is dominated by convective non-separatory flow. The increase in
permeance and the decrease in the ideal separation factor with decreasing content of sintering
42
aids are also consistent with the porosity and the average pore size data with such samples which
are shown in Table 2-2. As Table 2-2 indicates, the support porosity changes from 33.1% (for the
supports prepared with 4wt% phenolic resin and 0.1wt% B4C) to 37.3% for the supports
prepared without sintering aids so ever. The change in the average pore size is even more
dramatic, with dp, ranging from 331 nm (for the support tube prepared with the higher amount of
sintering aids) to 1660 nm, for the support tube prepared with 0.4wt% phenolic resin and
0.01wt% B4C, to 3340 nm for the support prepared without any sintering aids so ever. It comes
as no surprise, therefore, that the increases in both the porosity and the average pore size are
accompanied by substantial increases in permeance, with the He permeance of the support
prepared without sintering aids being close to 20 times higher than the permeance of the support
prepared with the highest fraction of sintering aids added. More importantly, the mechanical
properties of the supports prepared with smaller amounts of sintering aids are quite adequate for
the proposed application and as good, or even substantially better, than most of the commercial
supports used today for the preparation of inorganic membranes – see further discussion below.
43
1.10x10
5
1.20x10
5
1.30x10
5
1.40x10
5
0.0
8.0x10
-6
1.6x10
-5
2.4x10
-5
3.2x10
-5
4.0x10
-5
Ar (Normal)
He (Normal) Ar (Small)
Ar (Trace)
He (Small)
Ar (No Add)
He (No Add)
He (Trace)
Permeance / mol*m
-2
*s
-1
*Pa
-1
Average Pressure / Pa
Figure 2-7. Permeance vs. Pav of blended powder (50/50) supports with different amounts of
sintering aids (Permeation data taken at T= 378 K; Normal means 4wt% phenolic resin and
0.1wt% B4C, Small means 0.4wt% phenolic resin and 0.01wt% B4C, Trace means 0.04wt%
phenolic resin and 0.001wt% B4C, No Add means no sintering aids added)
44
1.12x10
5
1.20x10
5
1.28x10
5
1.36x10
5
1.44x10
5
1.6
2.0
2.4
2.8
No Add
Trace Add
Small Add
Average Pressure / Pa
Ideal Separation Factor
Normall Add
Figure 2-8. Ideal separation factor vs. Pav of blended powder (50/50) supports with different
amounts of sintering aids (Permeation data taken at T= 378 K; Normal means 4wt% phenolic
resin and 0.1wt% B4C, Small means 0.4wt% phenolic resin and 0.01wt% B4C, Trace means
0.04wt% phenolic resin and 0.001wt% B4C, No Add means no sintering aids added)
45
Sample
Amount of
Sintering Aids
(Weight Fraction)
Porosity
Average
Pore Size
(nm)
Compressive
Strength
(MPa)
He Permeance
(×10
-6
mol•m
-
2
•s
-1
•Pa
-1
)
Ar Permeance
(×10
-6
mol•m
-
2
•s
-1
•Pa
-1
)
S3
4wt% Phenolic
Resin and 0.1wt%
B 4C
33.1% 331 108.0 3.02 1.23
S4
0.4wt% Phenolic
Resin and
0.01wt% B 4C
35.1% 1660 57.2 24.7 13.8
S5
0.04wt% Phenolic
Resin and
0.001wt% B 4C
36.2% 2170 51.2 31.6 18.5
S6 No Additives 37.3% 3340 59.4 58.2 43.9
Table 2-2. SiC supports sintered with different amounts of sintering aids (Sintering temperature:
1900
o
C; composition: 50wt% 0.6 µm powder and 50wt% 6 µm powder; average measurement
pressure Pav=1.43×10
5
Pa; measurement temperature: 378 K)
Figure 2-9 shows the AFM images of (50/50) blended powder supports prepared with various
amounts of sintering aids. The images in Figure 2-9 indicate one of the downsides of using a
smaller amount of sintering aids, as these materials appear to have a significantly greater surface
roughness than the materials prepared with larger amounts of sintering aids. As Figure 2-9 shows,
the Rq=574 nm of the support prepared without sintering aids is almost four times greater in
magnitude than the Rq=145 nm of the support prepared with the highest amount of sintering aids
(despite the high surface roughness of these materials, we have nevertheless succeeded in
preparing high-quality nanoporous SiC membranes using them as supports, see further
discussion to follow). SEM images, shown in Figure 2-10, indicate that the particles in the
supports prepared without sintering aids are large and form pore structures with very large
46
interconnected pores (consistent with the large dp values reported in Table 2-2), while particles
in supports prepared with sintering aids are more densely packed, fused and connected, creating
pore structures with smaller pores (see Table 2-2 for the dp values of the various support
prepared with differing amounts of sintering aids).
These observations are consistent with preliminary observations reported in the earlier study by
our group, though the membrane preparation techniques at that point did not yield high-quality
SiC membranes using large surface roughness substrates. There is, generally, a consensus in the
technical literature that SiC powders are non-sinterable without the addition of large amounts of
sintering aids. However, most of the previous research work focused on the preparation of
completely dense SiC materials. In this work, we have successfully managed to prepare
mechanically strong porous tubular supports from a mixture of two different kinds of SiC
powders (HSC059N with an average particle size of 0.6 µm, the other HSC1200, with an
average pore size of 6 µm) with narrowly-distributed particle size distributions. The use of a
binary mixture of powders with very different sizes (around one order of magnitude difference),
likely results in a more compact packing of the particles, and thus a higher contacting interface
area among them, both helping to enhance the initial sintering process.
47
No Add. (Rq = 574 nm) Trace Add. (Rq = 379 nm)
Small Add. (Rq = 283 nm) Normal Add. (Rq = 145 nm)
Figure 2-9. AFM images of surfaces of blended powder (50/50) supports with different amounts
of sintering aids (Normal means 4wt% phenolic resin and 0.1wt% B4C, Small means 0.4wt%
phenolic resin and 0.01wt% B4C, Trace means 0.04wt% phenolic resin and 0.001wt% B4C, No
Add means no sintering aids added)
48
No Add Trace Add
Small Add Normal Add
Figure 2-10. SEM images of blended powder (50/50) supports with different amounts of
sintering aids (Normal means 4wt% phenol resin and 0.1wt% B4C, Small means 0.4wt%
phenolic resin and 0.01wt% B4C, Trace means 0.04wt% phenolic resin and 0.001wt% B4C, No
Add means no sintering aids added)
49
2.3.3 Effect of Different Initial Powder Composition
The reason behind using a blend of two different powders, one with a small average particle size
and another with a significantly larger particle size, to prepare supports that are both
mechanically strong and also have high permeance was previously discussed (as noted above,
both powders have a fairly narrowly distributed particle size distribution, as reported by the
manufacturer). In this section, the impact of combining the two powders at various ratios on the
transport characteristics and the surface and structural properties of the resulting supports is
systematically investigated (these supports are identified by two numbers, the first referring to
the weight fraction of the smaller 0.6 µm particles and the other to the fraction of the larger 6 µm
particles).
Figure 2-11 shows the He and Ar permeances (measured at an average transmembrane pressure
of 1.43×10
5
Pa) of various supports made from blends of the two powders. These are prepared
using the same amounts of sintering aids (4wt% phenolic resin and 0.1wt% B4C), the same
sintering temperature (1900
o
C) but with different initial powder compositions. The He and Ar
permeances remain relatively fairly invariable as the fraction of the larger size particles in the
initial powder blend increases until it reaches ~50wt%; they substantially increase after that, with
supports prepared with a fraction of larger particles of ~83wt% having a permeance which is
almost five times greater than the materials with a content of large particles of 50wt% (supports
prepared from a pure large particle powder under these sintering conditions have very poor
mechanical strength, and thus their permeances are not included in Figure 2-11). Figure 2-12
shows the ideal separation factor and porosity of sintered supports vs. the weight fraction of the
large particles in the initial blend. The ideal separation factor first decreases slowly up to ~
50wt%, and it decreases more sharply when the fraction of the larger particles becomes greater
50
than 50wt%. The porosity behavior shown in Figure 2-12 is quite interesting, first decreasing
slowly up to a fraction of 50wt%, reaching a minimum and then increasing after that. This is
consistent with the original premise in this study that mixtures of powders with distinct sizes
pack better together. Consistent with prior modeling efforts (Sahimi, 2003) on the packing of
unequal size spherical particles, the experimental porosity data indicate that there is an optimum
such composition. This is because as the fraction of large particles increases the small particles
can no longer completely fill in the space among the larger particles.
0.0 0.2 0.4 0.6 0.8
0.0
3.0x10
-6
6.0x10
-6
9.0x10
-6
1.2x10
-5
1.5x10
-5
He Permeance
Ar Permeance
Weight Fraction of Large Particles
Permeance / mol*m
-2
*s
-1
*Pa
-1
(12 psi)
Figure 2-11. Permeance vs. composition of large particles in supports (Permeation data taken at
T= 378 K, 4wt% phenolic resin and 0.1wt% B4C, sintering temperature 1900
o
C, Average
measurement pressure Pav=1.43×10
5
Pa)
51
0.0 0.2 0.4 0.6 0.8
0.2
0.4
0.6
0.8
1.0
Spn Factor
Porosity
Porosity
Weight Fraction of Large Particles
1.8
2.1
2.4
2.7
3.0
Spn Factor
Figure 2-12. Porosity and ideal separation factor vs. composition of large particles in supports
(Permeation data taken at T= 378 K, 4wt% phenolic resin and 0.1wt% B4C, sintering
temperature 1900
o
C, Average measurement pressure Pav=1.43×10
5
Pa)
Figure 2-13 shows the AFM images of SiC supports made with different initial powder
compositions (100/0, 83/17, 66/34, 50/50, 45/55, 40/60, 34/66, 26/74, and 17/83 – as a reminder,
the first number indicates the weight fraction of the small size particles in the initial blend and
the second number that of the larger particle powder). The support made from pure small
particles has the smoothest surface (Rq = 40.3 nm). As more large particles are added to the mix,
the surface of the blended supports starts to gradually become rougher. The increase in the
surface roughness accelerates beyond the (50/50) composition, with the Rq of the (17/83)
particles being at least an order of magnitude larger than that of the (100/0) support (the supports
52
with a large fraction of large particles have inferior mechanical properties, as Table 2-3 indicates,
and present technical challenges during the AFM measurements as well, so their true Rq could be
significantly higher than what is reported here). Figure 2-14 shows SEM images of the same
supports as Figure 2-13. The information glimpsed from these SEM images is consistent with
what is learned from the AFM studies. The supports made from the powder blends with a
fraction of small particles greater than 50wt% have very similar appearance with a relatively low
surface roughness. In these samples the small particles always completely fit in the void space in
between the larger particles (for real small fractions of large particles in the initial blend one can
distinguish large particles in a “sea” of smaller particles). As the fraction of the large particles in
the initial blend increases, distinguishable small particles in the sintered support gradually
become fewer in numbers and for the samples with the greater fraction of large particles they
cannot completely fill the void space among the large particles (this is clearly visible, for
example, in the SEM images of the (26/74) and the (17/83) supports).
53
100/0 (Rq = 40.3 nm) 83/17 (Rq = 129 nm) 66/34 (Rq = 147 nm)
50/50 (Rq = 134 nm) 45/55 (Rq = 153 nm) 40/60 (Rq = 205 nm)
34/66 (Rq = 312 nm) 26/74 (Rq = 267 nm) 17/83 (Rq = 439 nm)
Figure 2-13. AFM images of surfaces of blended powder supports with different compositions
(4wt/% phenolic resin and 0.1wt% B4C, sintering temperature 1900
o
C, the two numbers indicate
the weight fraction of the 0.6 µm small particles and of the 6 µm large particles)
54
100/0 Support 83/17 Support 66/34 Support
50/50 Support 45/55 Support 40/60 Support
34/66 Support 26/74 Support 17/83 Support
Figure 2-14. SEM images of blended powder supports with different compositions (4wt/%
phenolic resin and 0.1wt% B4C, sintering temperature 1900
o
C, the two numbers indicate the
weight fraction of the 0.6 µm small particles and of the 6 µm large particles)
55
Sample
Weight Ratio
(Small Particle
/Large Particle)
Porosity
Average
Pore
Size
(nm)
Compressive
Strength
(MPa)
He Permeance
(×10
-6
mol•m
-
2
•s
-1
•Pa
-1
)
Ar Permeance
(×10
-6
mol•m
-
2
•s
-1
•Pa
-1
)
S7 100/0 42.1% 407 68.5 3.06 1.14
S8 83/17 38.9% 361 63.9 3.04 1.11
S9 66/34 33.2% 388 63.9 2.90 1.17
S3 50/50 33.1% 331 108.0 3.01 1.23
S10 45/55 43.9% 1285 83.4 8.86 4.47
S11 40/60 43.9% 1452 48.7 9.53 4.87
S12 34/66 43.7% 1944 44.4* 10.30 5.30
S13 26/74 44.0% 2315 39.7* 12.10 6.38
S14 17/83 42.3% 3092 21.7 15.40 8.35
Table 2-3. SiC supports sintered with different compositions (Sintering temperature: 1900
o
C;
Sintering aids: 4wt% phenolic resin and 0.1wt%B4C; average measurement pressure
Pav=1.43×10
5
Pa; measurement temperature: 378 K, * only one sample tested)
2.3.4 Effect of Grinding
As the discussion above indicates, the preparation conditions that lead into increased permeances
often also lead into supports with much rougher surfaces. One technique to maintain the desired
transport characteristics of such supports while minimizing their surface roughness, as shown in
56
Figure 2-15, is via the surface treatment discussed in the experimental section (the technique is
better suited for small-area, sensor-type applications, however, rather than for the larger area
membrane systems of primary interest in this paper). Figure 2-16 shows the permeance and
Figure 2-17 the ideal separation factor of two SiC supports prepared under identical conditions
but the surface of one of which has been treated by the technique described in the experimental
section. As Figure 2-16 and Figure 2-17 (as well Table 2-4) indicate, the transport characteristics
of the two supports are fairly close to each other (the permeance of the support whose surface
has been ground is a bit higher because after grinding its thickness is a bit smaller as well).
Figure 2-18 shows the AFM images of both supports indicating that surface grinding is very
effective in reducing the surface roughness of the SiC supports, e.g., decreasing the Rq by more
than a factor of ten.
Figure 2-15. Effect of surface treatment (grinding)
57
1.12x10
5
1.20x10
5
1.28x10
5
1.36x10
5
1.44x10
5
1.8x10
-5
2.4x10
-5
3.0x10
-5
3.6x10
-5
4.2x10
-5
4.8x10
-5
5.4x10
-5
Ar (Regular)
He (Regular)
Ar (Surface Treated)
Permeance / mol*m
-2
*s
-1
*Pa
-1
Average Pressure / Pa
He (Surface Treated)
Figure 2-16. Permeance vs. average pressure of untreated and surface-treated blended powder
(50/50) supports (Permeation data taken at T= 378 K, no sintering aids, sintering temperature
1800
o
C)
58
1.12x10
5
1.20x10
5
1.28x10
5
1.36x10
5
1.44x10
5
1.05
1.40
1.75
2.10
2.45
Ideal Separation Factor
Average Pressure / Pa
Surface Treated
Regular
Figure 2-17. Ideal separation factor vs. average pressure of untreated and surface-treated blended
powder (50/50) supports (Permeation data taken at T= 378 K, no sintering aids, sintering
temperature 1800
o
C)
59
Untreated Support (Rq = 570 nm) Surface Treated Support (Rq = 48 nm)
Figure 2-18. AFM images of surfaces of untreated and surface treated powder (50/50) supports
(no sintering aids, sintering temperature 1800
o
C)
Sample
Surface
Treatment
Porosity
Average
Porosity
(nm)
He Permeance
(×10
-6
mol•m
-
2
•s
-1
•Pa
-1
)
Ar Permeance
(×10
-6
mol•m
-
2
•s
-1
•Pa
-1
)
S15 No 35.3% 3340 46.7 26.2
S16 Yes 35.6% 3430 52.1 29.4
Table 2-4. SiC supports with and without surface treatment (Sintering temperature: 1800
o
C; no
sintering aids used; average measurement pressure Pav=1.43×10
5
Pa; measurement temperature:
378 K)
60
2.3.5 Mechanical Properties
Porous ceramics can be thought of as bi-phasic materials, whereby the second phase is a gas
residing inside the pores of the solid. Pores, however, act as stress concentrators and the origin of
potential cracks and fractures (Vales et al., 1999). When the porous ceramics are subjected to
various loads, as is the case with the membrane tubes prepared in this study which are expected
to withstand transmembrane pressure gradients ranging from a few bar to tens of bar, this
system of cracks of various orientations may spread and lead into embrittlement and eventual
macro-failure. Throughout this study, therefore, we have paid great attention in testing the
mechanical strength of the resulting supports using the experimental approach described in the
Experimental Section. Support tubes prepared with the blended powders show excellent
mechanical properties (see Tables 1-3) with the compressive strengths of support prepared with
blends containing more than 50wt% of the smaller particles being greater than 60 MPa. These
are comparable or significantly better than the compressive strengths of both currently available
commercial ceramic products and reported results from the published literature. For example,
commercially available cordierite, mullite, and corundum-mullite ceramics are reported by their
manufacturers to have a compressive strength which is larger than 12 MPa; alumina foam filters
have a compressive strength of 1 - 3 MPa, commercial α-alumina porous supports (19 - 43 MPa)
alkaline pottery porous supports (8 - 11 MPa), and kaolinite porous supports (13 - 29 MPa). Itoh
et al. (1998), for example, prepared strengthened porous alumina membrane tube, via internal
anodic oxidation, which increased their compressive strength from 10.6 MPa to 32.6 MPa; Li
and Zhong (2002) prepared kieselguhr-mullite ceramic membrane substrates which showed a
compressive strength in the range of 1-6 MPa; Liu et al. (2013) prepared low-cost multi-channel
α-alumina ceramic membrane tubes and reported their compressive strength to be up to 8 MPa.
61
Mesoporous silica-based antibiotic releasing scaffolds were prepared by Shi et al. (2009), whose
compressive strength was as high as 10 MPa, though these porous materials have not been used
in membrane application; Drisko et al. (2011) prepared strong silica monoliths using agarose gel
templates and reported their compressive strength to be as high as 25 MPa. Fukushima and
coworkers fabricated highly porous SiC by gelation-freezing method and reported a compressive
strength of up to 16.6 MPa.
As expected, conditions leading into better sintered substrates also result into more mechanically
strong support tubes. For example, Table 2-1 clearly indicates that increasing the temperatures
results in a significant increase in the compressive strength of the tubular supports. The same can
be observed when increasing the amount of sintering aids (see Table 2-2), though the effect
saturates rather quickly with supports without sintering aids, prepared under the conditions
indicated in Table 2-2, still showing quite impressive compressive strengths. For the supports
prepared from blends of powders, like in this study, a key determinant of the mechanical strength
of the resulting supports is the ability of the starting powders to pack well together. For example,
as Table 2-3 indicates, supports made with a higher fraction of the large (6 µm) particles have
poor mechanical properties.
2.3.6 Microporous Membrane Preparation
The tubular substrates prepared from the sintering of blended SiC powders are being used to
prepare microporous membranes via the deposition on the tubular substrates of thin SiC
nanoporous films. These are made via the dip-coating of polymeric ceramic precursors films, in
particular for the membranes prepared here using allyl-hydridopolycarbosilane (AHPCS), a
62
partially allyl-substituted hydridopolycarbosilane (HPCS). The technique utilized involves the
use of polystyrene (PS) sacrificial interlayers, and has been described in detail elsewhere. In
short, the ceramic tubes are dip-coated in a solution of 1wt% of polystyrene (GPC grade,
Mw=2500, Scientific Polymers Products, Inc.) in toluene (for this, a dip-coating time of 12 s and
a drawing rate of 0.6 mm/s were utilized). The tubes with the PS layer on the top were first dried
at 100 oC for 1 hr, and then dip-coated in a solution of 10wt% of AHPCS in hexane, with a dip-
coating time of 12 s and a drawing rate of 2 mm/s. Following the coating of the top layer, the
membranes were pyrolyzed in Ar. For that, the coated tubes were heated in flowing Ar in a tube
furnace (Lindberg/Blue, Model STF55433C) at a rate of 2
o
C/min, first to 200
o
C, where they
were kept for 1 hr, then to 400
o
C, where they were also kept for 1 hr, and finally to 750
o
C,
where they were kept for an additional 2 hr. Subsequently, they were cooled down to room
temperature in flowing Ar with a cooling rate of 3
o
C/min. Generally, a number of top layers
must be deposited before the resulting membrane exhibits a molecular sieving behavior. After
the final pyrolysis step the composite membranes were treated in flowing synthetic air for 2 hr at
450
o
C to remove carbon-containing residues that may remain.
The membranes prepared are characterized for their morphology by SEM and AFM, their pore
structure via BET, and their transport characteristics via permeation measurements. Table 5
shows the He and Ar single gas measurements for three different membranes (the substrates used
in the preparation of these membranes are prepared by cold isostatic pressing of the mixture of
50wt% 0.6 µm and 50wt% 6 µm SiC powders sintered at 1900
o
C without the use of any
sintering aids). For all the measurements the temperature was kept at 200
o
C, the pressure on the
membrane permeate side was kept atmospheric, while the transmembrane pressure gradient was
kept constant at 207 kPa (the estimated relative experimental error for the measured permeances
63
is <5%). As it is clear from Table 2-5, using the blended substrates we are able to prepare highly
permselective membranes. The task of preparing high-quality membranes using substrates with
such varying degrees of surface roughness is made easier via the use of a novel technique that
involves depositing on the substrates (prior to the deposition of the SiC nanoporous layers) of an
interlayer scaffold made of either highly uniform SiC nano-powders or SiC nanotubes, for which
further details can be found elsewhere (Elyassi et al., 2013).
Membrane 1 Membrane 2 Membrane 3
He Permeance at
200
o
C (×10
-8
mol•m
-2
•s
-1
•Pa
-1
)
3.8×10
-8
1.2×10
-8
2.1×10
-8
Spn. Factor
(He/Ar)
>2000
Table 2-5. Permeance data of three membranes prepared using 50/50 blend support
(Measurement Temperature: 200
o
C; pressure difference across membranes: 2.07×10
5
Pa)
2.4 Conclusions
In this paper, porous SiC tubular supports with high permeances (of up to 5.8×10
-5
mol•m
-2
•s
-
1
•Pa
-1
) were prepared via the use of blended SiC powders and their structure and permeation
characteristics were studied. The effect of various preparation parameters were investigated
including the sintering temperature, the type and the amount of sintering aids utilized, and the
composition of the starting powders. Mechanically strong and highly permeable supports have
been prepared and have been used to prepare highly permselective SiC membranes. Using blends
64
of powders rather than pure powders alone appears to offer an advantage in terms of preparing
supports with very high permeances and good mechanical stability. This is because combining
the two powders with distinctly different sizes allows them to pack better together and to create
better sintered and more highly permeable structures.
2.5 Acknowledgement
The support of the U.S. Department of Energy (DE-SC0003586) and the National Science
Foundation (CBET-0854427) are gratefully acknowledged. The Superior Graphite Co. is also
acknowledged for kindly providing the SiC powders. The help of Dr. Suzhou Zhang (from East
China University of Science and Technology) in performing the compressive strength tests, and
the assistance of Mr. Mingyuan Ge, Ms. Jia Liu, and Ms. Xin Fang (from the University of
Southern California) in carrying out the AFM and/or SEM characterization work are also
gratefully acknowledged.
65
3. Fabrication of Silicon Carbide Membranes
3.1 Introduction
As noted in this Thesis previously, membrane separation technology has attracted a lot of
research work in the past 30 years due to its low energy cost compared to the more conventional
separation technologies, such as distillation, centrifugation, crystallization, adsorption, and
sublimation. Polymeric membranes have been extensively investigated for many years and are
now widely used in process applications like ultrafiltration, pervaporation, and desalination.
Inorganic membranes, on the other hand, have received much less attention, despite the fact that
they show substantial promise for a broad range of applications. One of the drivers for the further
development of inorganic membranes is the efficient production of hydrogen, whose demand in
the chemical industry has been increasing recently due its potential use for clean energy
generation. Process intensification during hydrogen production involves a separation step under
high-temperature and pressure environments, where conventional polymeric membranes cannot
be used. Among the inorganic membranes proposed for such application, silicon carbide (SiC)
membranes are a particularly promising candidate because of their many desirable unique
properties like high corrosion resistance (Li et al.,1996), high thermal conductivity (Takeda et al.,
1987), high thermal shock resistance (Schulz et al., 1994) and excellent chemical and mechanical
stability (Kenawy et al., 2005).
As also noted in the Introduction Section, the preparation methods of SiC membranes mainly
consist of two different approaches. One involves the pyrolysis of polymeric precursors like
polycarbosilanes (PCS) and allyl-hydropolycarbosilanes (AHPCS), and the other is based on the
66
technique of Chemical Vapor Deposition (CVD) or Chemical Vapor Infiltration (CVI) via the
use of appropriate precursors like tri-isopropylsilane (TPS).
3.1.1 Pyrolysis of Polymeric Precursors
One of the most important and widely used polymeric precursors in making SiC materials is
polycarbosilane (PCS). In the preparation of SiC membranes, in order to achieve the pyrolysis
process at a relatively low temperature and to minimize the oxygen content in the final product,
our group has previously utilized a partially allyl-substituted hydropolycarbosilane (AHPCS).
The molecular structures of HPCS and AHPCS were previously shown in Figure 1-3, in Chapter
1 of this Thesis.
Several research efforts on preparing SiC membrane by the pyrolysis of PCS/AHPCS were
carried out in the last 15 years. Li et al. (1996) prepared amorphous Si-C-O membranes on the
outer surfaces of alpha-alumina supports by dip-coating with p-xylene solutions of
polycarbosilane and subsequent pyrolysis. At 773 K this membrane had a H2 permeance of 1×10
-
8
mol•m
-2
•s
-1
•Pa
-1
, and its (H2/N2) separation factor was in the range of (18 - 63). The addition of
polystyrene (PS) as a pore former increased the permeance of hydrogen by an order of
magnitude, but decreased the separation factor to 13. Our group (Ciora et al., 2004) prepared
nanoporous SiC membranes by pyrolysis of AHPCS, and these membranes were stable in air at
450
o
C. However, they were unstable in the presence of high-temperature steam. Suda et al.
(2006) investigated the structural evolution during conversion of PCS into SiC-based
microporous membranes. They found that the cross-linking reactions resulted in a three-
dimensional pore network consisting of small micropores, having a large surface area. In another
67
study, the same group (Suda et al., 2005) prepared SiC membranes by pyrolysis of non-
crosslinked PCS, crosslinked PCS, and polystyrene (PS) dispersed PCS in Argon atmosphere of
up to 973 K. They reported a H2 permeance of up to 3×10
-8
mol•m
-2
•s
-1
•Pa
-1
, and a (H2/N2)
separation factor of up to 200. Nagano et al. (2006) prepared amorphous SiC membranes with
molecular sieving properties on the outer surfaces of γ-alumina-coated α-alumina porous
supports through the pyrolysis of PCS without an oxygen-curing process. At 873 K, this
amorphous SiC membrane resulting from PCS pyrolysis at 1073 K had an H2 permeance of
8.1×10
-7
mol•m
-2
•s
-1
•Pa
-1
, with a (H2/CO2) separation factor of 13. Wach et al. (2007) reported
that they developed SiC ceramic membranes via coating on porous alumina support of a
polymeric precursor mixture consists of 80wt% polycarbosilane (PCS) and 20wt%
polyvinylsilane (PVS) followed by pyrolysis in an Ar atmosphere at 450
o
C. Their membranes
showed a H2 permeance of 0.8 - 1×10
-7
mol•m
-2
•s
-1
•Pa
-1
and a N2 permeance of 2.5 - 4×10
-8
mol•m
-2
•s
-1
•Pa
-1
. At 250
o
C, their membrane has a (H2/N2) separation factor of 206 and a (He/N2)
separation factor of 241. Our group (Elyassi et al., 2007) prepared nanoporous SiC membranes
on porous SiC supports by slip-casting of small SiC powders (~100 nm diameter) in AHPCS,
followed by AHPCS dip-coating and subsequent pyrolysis. At 473 K, this membrane had a He
permeance of 8.9×10
-9
mol•m
-2
•s
-1
•Pa
-1
and a (He/Ar) separation factor of 147. After 2 hr of
oxidation in air at 450
o
C, the permeance of He increased to 1.2×10
-8
mol•m
-2
•s
-1
•Pa
-1
, and the
(He/Ar) separation factor decreased to 89. In a later study, Elyassi et al. (2008) prepared better
quality SiC membranes by using polystyrene as the pore former during the dip-coating process;
this modified SiC membrane had a He permeance of 1.8 - 4.3×10
-8
mol•m
-2
•s
-1
•Pa
-1
, and its
(He/Ar) separation factor was in the range of (176 - 465). Most recently, Elyassi et al. (2013)
prepared SiC membranes with porous nanofiber fillers and compared such membranes with those
68
made from nonporous fillers. Membranes made from nanofibers fillers show similar permeation
results to those made from nonporous fillers.
3.1.2 Chemical Vapor Deposition/Infiltration
Chemical Vapor Deposition (CVD) is a technique which is widely used in the semiconductor
industry to prepare high-purity thin films. It is a very mature technology and has been used in
industry for a long time. It has also been utilized to prepare supported SiC membranes in the past
15 years. An idealized schematic illustration of this deposition process can be shown as
following Figure 3-1.
Figure 3-1. Schematic illustration of the CVD/CVI process (Chen et al., 2008)
In the past 15 years, several research efforts has been made in this particular area. Sea et al.
(1998), for example, prepared SiC membranes by CVD of tri-isopropylsilane at 700-800
o
C
69
inside the macropores of α-alumina support tubes via forced flow through the porous tube. These
membranes permeated various gases, other than water, mainly via the Knudsen diffusion
mechanism at 50 - 400
o
C. After being heated in Ar at 1000
o
C for 1 hr, these membranes
exhibited a H2 permeance of 5 - 6×10
-7
mol•m
-2
•s
-1
•Pa
-1
at 50 - 400
o
C. For the (H2-H2O-HBr)
gas mixture, at 400
o
C, the membranes maintained a modest (H2/H2O) selectivity for more than
100 hr. Takeda, et al. (2001) prepared SiC membranes on γ-Al2O3-coated porous α-Al2O3 tubes
by alternatively using SiH2Cl2 and C2H2 (as source gases) diluted with hydrogen. The reaction
temperature was 800 - 900
o
C. By using repeated cycles of alternating gas supply, they could
control the SiC infiltration process. The SiC membranes made by this method possessed a H2
permeance of 1×10
-8
mol•m
-2
•s
-1
•Pa
-1
and a (H2/N2) separation factor of 3.36 at 350
o
C. Lee et al.
(2005) used (SiH4/C2H2/Ar) low-pressure chemical vapor deposition (LPCVD) to modify
asymmetric SiC membranes. By doing so, they were able to prepare asymmetric mesoporous SiC
membranes on the surfaces of Al2O3-doped SiC macroporous supports. Not surprisingly, after
modification, the permeance of SiC membranes decreased while the selectivity increased.
Nagano et al. (2007) prepared SiC membranes on the outer surfaces of γ-alumina-coated α-
alumina porous supports by both pyrolysis of polycarbosilanes (PCS) without an oxygen-curing
process and chemical vapor infiltration. It was shown that CVI was a good technique to increase
the membrane selectivity towards gases with smaller kinetic diameters. At 873 K, these
membranes had a He permeance of 7.7×10
-8
mol•m
-2
•s
-1
•Pa
-1
, a (He/CO2) separation factor of 64,
and a (He/H2) separation factor of 4.4. Chen et al. (2008) prepared SiC membranes by CVD of
TPS on porous SiC supports. The SiC membrane had a He permeance of 2.6×10
-7
mol•m
-2
•s
-
1
•Pa
-1
and a (He/Ar) separation factor of ~13; the properties of the resulting membranes
70
depended strongly on the CVD conditions including the TPS injection rate, the flow rate of the
carrier gas (Ar) and the temperature.
3.2 Experimental Section
As mentioned in the introduction, there are two distinct ways of making SiC membranes. One
involves the pyrolysis of a polymeric precursor (PCS/AHPCS), while the other uses the chemical
vapor deposition (CVD) of the TPS. Here, we describe both methods. There are two kinds of
tubular SiC sintered supports used in the fabrication of SiC membranes. One kind of supports
were made from pure 0.6 µm SiC powders sintered at 1900
o
C without any sintering aids. And
the other kind of supports were made from a SiC powders mixture (50% 0.6 µm SiC powder and
50% 6 µm SiC powders) at 1900
o
C without any sintering aids
3.2.1 Fabrication of SiC Membranes by Pyrolysis of AHPCS
In the pyrolysis method, the SiC supports were sonicated several times in acetone and dried in air,
before the slip-casting and dip-coating steps. The dip-coating solution consists of 10wt%
AHPCS (SMP-10, Starfire Systems, Inc.) and 90wt% hexane (HX0299-5, EMD Chemicals).
There are two kinds of slip-casting solutions utilized: the first one (referred to as (-) PS slip-
casting solution) consists of the dip-coating solution above containing 5wt% of slip-casting
particles (see below), and the second one (referred to as (+) PS slip-casting solution) consists of
the slip-casting solution above containing, in addition, 1.25wt% polystyrene (PS, GPC grade,
Mw=2500, Scientific Polymers Products, Inc.). In order to prepare the fine SiC slip-casting
71
particles used in the aforementioned slip-casting solutions, 7 g of the 0.6 µm SiC powder was
mixed with 200 ml of acetone in a beaker. The top 50 ml of the 200 ml solution were then
separated, the acetone was evaporated and the dry particles were collected and used to prepare
the slip-casting solutions.
For the slip-casting step, the SiC support was immersed in the slip-casting solution for a 12 s,
and then drawn out of the solution at a speed of 0.28 mm/s. The coated SiC support was then
placed in a tube furnace (Linberg/Blue, Model: STF55433C) in flowing Argon, heated at a rate
of 2
o
C/min first to 200
o
C, where it was kept for 1 hr, then to 400
o
C where it was kept for
another 1 hr, and finally to 750
o
C where it was kept for an additional 2 hr. Afterwards, the
coated SiC support was cooled down to room temperature at a rate of 3
o
C/min. This type of
heating profile has been shown by our team and others (Suda et al., 2006; Interrante, et al., 1994)
to generally yield better cross-linked amorphous SiC materials during the pyrolysis of polymeric
precursors. Moreover, our group, by using such a heating protocol, has been successful in
fabricating membranes with high hydrogen permeance (Elyassi et al., 2007; 2008, 2013).
Following the above step, the membranes were immersed in a solution of 3.12wt% polystyrene
in toluene for 12 s, and drawn out of the solution at a speed of 0.58 mm/s. After drying the
membranes at 100
o
C in air for 1 hr, they were then immersed in a dip-coating solution for 12 s,
drawn out of the solution at a rate of 1.52 mm/s, and were then subjected to the same heating
protocol as that followed during the slip-casting step. This dual-step process (first a coating of PS
followed by a coating of AHPCS – herein after referred to as “dual”) was repeated three times.
Compared to single AHPCS layer deposition alone (without the coating with PS) it has been
shown that it can yield better permeation properties, and cure the cracks in the SiC membranes
that are formed during the prior pyrolysis step. Layer-by-layer dip-coating and pyrolysis result in
72
lower permeances and higher separation factors. Thus, by properly selecting the number of dip-
coated layers, we can take advantage of the trade-off between the gas permeance and the desired
separation properties.
The SiC membranes were treated in air for 2 hr at 450
o
C after the dip-coating and pyrolysis of
the final layer in order to oxidize any potential carbon residue that may remain. Low-temperature
oxidation is a good way of removing the small amount of carbon. It is widely used in the
purification of carbon nanotubes (Li and Zhang, 1994; Osswald et al., 2006; Park et al., 2006). In
this work, low-temperature oxidation has been shown to increase the gas permeance and to
decrease the separation, but to also lead to SiC membranes which are stable to air and steam.
In order to measure the permeation characteristics of SiC supports and membranes, we have used
two different types of permeation test cells. The first type of test cell and the overall permeation
apparatus are shown in Figure 3-2.
Figure 3-2. Schematic of the first type of permeation test set-up for single-gas measurements
73
In order to test the membrane permeation in this apparatus, the cylindrical SiC membrane tube is
affixed on one end to a fender washer using a high-temperature impermeable glue (J-B WELD
8625-S Cold-Weld Compound). The other end is hermetically sealed using graphite tape and the
same type glue. The membrane-washer set is then installed in between the two half-cells of the
permeation-test unit using O-rings for sealing.
For single-gas permeation tests, the gas is fed into the bottom half-cell from the feed port then
permeates through the membrane into the top half-cell and exits through the permeate port, with
its flow being measured with a bubble-flow meter. The pressure in the permeate side is measured
by a pressure gauge (OMEGA Engineering, DPG1000B-15G) and is maintained close to
atmospheric. The pressure in the feed-side is adjusted with the aid of regulators of the upstream
gas cylinders and a needle valve on the reject tube. A differential pressure transducer (OMEGA
Engineering, PX409-050DDUI) is used to measure the pressure difference between the feed and
permeate sides. The range of temperatures in the experimental system is from room temperature
to 200
o
C.
For mixed-gas permeation tests, we follow a similar procedure as the single-gas testing above.
The difference is that for the mixed-gas permeation tests, we provide the feed gas of an
appropriate composition through the use of mass flow controllers (MFC). In addition, the
composition of the gas mixture in the feed-side, reject-side, and permeate-side are analyzed by
gas chromatography (GC). The experimental system utilized (shown in Figure 3.3) is very
similar to the one shown in Figure 3-2, other than the fact that MFC are used to prepare the feed
composition, and gas chromatographs are used to measure the gas compositions.
74
Figure 3-3. Schematic of the first type of permeation test set-up when used for mixed-gas
permeation experiments
The second type of test cell and the overall apparatus is shown in Figure 3-4,
75
Figure 3-4. Schematic of the second type of permeance test set-up
For measurements in this testing apparatus, the two ends of the SiC membrane are glazed, so as
to make them impermeable. Then the SiC membrane is inserted into the test cell which is made
of a stainless steel (SS) tee (Swagelok). The feed side and reject side tubes are connected onto
the tee with Swagelok fittings. The permeate side exit tube is connected to the third leg of the SS
tee. The experimental procedure for single-gas and mixed-gas experiments is similar to the one
described above.
76
3.2.2 Fabrication of SiC Membranes by CVD/CVI
We utilize two different types of CVD apparatus to prepare SiC membranes, whose schematics
are shown in Figure 3-5 and Figure 3-6; these two set-ups are mostly similar to each other, with
only minor differences on the direction of vapor flow and in the tubing connections.
Figure 3-5. Schematic of the first type of CVD apparatus
77
Figure 3-6. Schematic of the second type of CVD apparatus
In the CVD experiments the two ends of the macroporous SiC support tubes are connected to
dense (non-porous) mullite tubes. To accomplish this connection, graphite tapes are inserted to
fill the space in between the outer surface of the mullite tube and the inner surface of the SiC
support tube. A high-temperature impermeable putty (7551A23 ceramic putty, from McMaster-
Carr) is used to further seal the two parts. Then, the whole assembled part (SiC support tube and
the two mullite tubes) is installed inside a quartz tube which functions as the CVD reactor.
Argon is used as the carrier gas, and tri-isopropylsilane as the CVD precursor. During operation,
the CVD reactor is heated first to the desired temperature in Ar and kept at that temperature for
12 hr, before the TPS flow is initiated. In most experiments, because of the relatively slow Ar
78
flow rates, the pressures of both permeate and the reject sides are kept near atmospheric.
However, by adjusting the needle-valve on the reject tube and the regulator valve of the carrier
gas cylinder, we are also able to pressurize the feed side, if so desired. For the first type of CVD
apparatus, shown in Figure 3-5, one of the end mullite tubes is plugged and the Ar and the TPS
flow from the outside of the tube to the inside where they exit from the permeate port. In the
second kind of CVD apparatus, shown in Figure 3-6, one of the end mullite tubes is connected to
the feed line while the other mullite tube is connected to the exit (reject) line. The Ar and TPS
flow in the inside of the SiC tube. Some of the TPS (and the Ar) flow through the membrane and
exit through the permeate side, as shown in Figure 3-6.
One advantage that these two CVD set-ups offer is that we are able to measure the permeance of
the membranes on line during the CVD process. In order to accomplish that, we carry out the
CVD process for a certain period time; we then stop injecting TPS while continuing the flow of
Ar for a certain amount of time. We then measure the permeance of Ar and He (or of any other
gases we choose). After the permeation tests are completed, we can switch back from the test
gases to the carrier gas and begin injecting TPS in order to initiate the CVD process. The
(CVD/permeation measurement) cycle can be repeated as many times as one desires until the
membrane reaches the desired permeation characteristics.
3.3 Results and Discussion
3.3.1 SiC Membranes Prepared via Pyrolysis
The structure of an idealized SiC membrane is shown in Figure 3-7.
79
Figure 3-7. Structure of an idealizd SiC membrane
As shown in Figure 3-7, the SiC membrane consists of three parts: from the bottom to the top, a
macroporous support, a mesoporous layer, and the top nanoporous membrane film (n real
membranes, unlike the idealized situation shown in Figure 3-7, the interfaces between the three
parts may not be as clear and distinct, however). Of the three parts, the macroporous support is
typically highly porous, and gas molecules can easily permeate through it without much
resistance. The mesoporous intermediate layer connects the macroporous support and the
nanoporous membrane layer, and acts as a bridge in between the top part and the bottom part. Its
thickness (incluidng the part of the layer that infiltrates the underlying support) strongly impacts
the permeance of the resulting membrane. The nanoporous top membrane layer provides the
nanosize pores for gas separation. The SiC membranes described in this Chapter consists of a
80
macroporous SiC support, an intermediate slip-casted layer after pyrolysis, and the final top layer
resulting from the pyrolysis of a number of dip-coated layers.
As noted above, a desirable membrane structure should consist of a very porous support for
gases to permeate through without much resistance, a thin and smooth intermediate layer that
causes a relatively small permeance decrease, and a thin crack-free nanoporous top layer layer
that possesses good separation properties. One way to reduce the impact on peremeance of the
inetrmediate layer is to reduce to its thickness, as noted above; however, if the interlayer is too
thin and has a lot of defects, then, in the subsequent dip-coating steps, a large fraction of the
polymeric precursor molecules (AHPCS in our case) may penetrate through and deposit into the
support’s macropore structure, which results in a great decrease in the permeance after pyrolysis.
In this study thin and reletively defect-free intermediate layers are prepared by varying the
preparion conditions and the type of powders and other additives utilized. The permeance and
the separation factor of the nanoporous top layer is controlled by varying the number of deposied
layers and by the conditions of pyrolysis.
Our group has previously prepared two different kinds of SiC membranes with good peremation
characteristcs (Elyassi et al., 2007, 2008, 2013). They were both made using supports prepared
with pure 0.6 µm SiC powder. The difference between the two is that for the second kind of
membrane, a polystyrene solution was used as a pore former following the approach we also use
in this Thesis. The first type of SiC membranes the best perfomance attained was a Helium
permeance of 0.89 - 1.2×10
-8
mol•m
-2
•s
-1
•Pa
-1
at 473 K, with a (He/Ar) separation factor of 89 -
147. For the second type of SiC membranes, with the use of PS as a pore former, the permeance
of Helium was in the range (1.8 - 4.3×10
-8
mol•m
-2
•s
-1
•Pa
-1
) and the (He/Ar) separation factor
was in the range of 176 - 465.
81
In Chapter 2, we reported the preparation of supports using a mixture of two powders with
substantially differing sizes. The permeance and surface roughness of these supports depends on
the initial composition of powders, the concentration of sintering aids used, and the sintering
temperature and other conditions. In the preliminary results on the prepaartion of SiC
nanoporous membranes presented here we have utilized supports made from pure 0.6 µm SiC
powders sintered at 1900
o
C without the use of any sintering aids. The focus of our efforts has
been to systematically investigate how the deposition of the single slip-casted layer and each
additional dip-coated layer affects the permeance and separation factor of the resulting
membranes sample.
The permeation results of membranes made using the (+) PS slip-casted solution are shown in
Figure 3-8 and those using the (-) PS slip-casted solution are shown in Figure 3-9. (Each
experimental point represents the average permeance and separation factor of three different
membranes prepared under identical conditions, while the arrow represents the range of
experimental values measured),
82
Support Slip-casting 1 Dual 2 Dual 3 Dual
10
-8
10
-7
10
-6
10
-5
Spn Factor of He / Ar
He Permeance / mol*m
-2
*s
-1
*Pa
-1
Permeance
0
100
200
300
400
500
600
Spn Factor
Figure 3-8. Permeation results of SiC membranes made from (+) PS slip-casting solution
(Measurement temperature: 473 K, pressure difference across the membrane 2.41×10
5
Pa; each
individual data point reflects results of three different membranes prepared under identical
conditions, as described in the text)
83
Support Slip-casting 1 Dual 2 Dual 3 Dual
10
-8
10
-7
10
-6
10
-5
Spn Factor of He / Ar
He Permeance / mol*m
-2
*s
-1
*Pa
-1
(35 psi)
0
200
400
600
800
1000
1200
Figure 3-9. Permeation results of SiC membranes made from (+) PS slip-casting solution
(Measurement temperature: 473 K, pressure difference across the membrane 2.41×10
5
Pa; each
individual data point reflects results of three different membranes prepared under identical
conditions, as described in the text)
It should be clear from the data in both figures, that as we deposit more layers on a given SiC
support, the He permeance keeps decreasing and the (He/Ar) separation factor keeps increasing.
In Figure 3-8, for example, after the slip-casted layer is deposited and pyrolyzed, the He
permeance decreases by an order of magnitude, and the separation factor increases somewhat;
then after the first dual-layer (PS + AHPCS) deposition and pyrolysis, the He permeance
84
decreases to (1.8 - 2.5×10
-7
mol•m
-2
•s
-1
•Pa
-1
) while the separation factor increases to (60 - 131).
Compared to the previous membranes (Elyassi et al., 2007, 2008, 2013) prepared by our group,
the permeance of these membranes is signifcantly higher (an order of magnitude) while the
separation factor is still quite high. The membranes prepared with two and three dip-coated
layers have properties which are in line with membranes previously prepared by our group. For
the membranes prepared without polystyrene being present in the slipcasted solution (see Figure
3-9), the trends are qualitatively very similar with those observed in Figure 3-8. However,
membranes with only a single dip-coated layer are not microporous. It takes at least two such
layers being deposited before these membranes become nanoporous, as Figure 3-8 shows. This
then indicates a beneficial effect of using PS in the preparation of the slip-casting solution and
layer.
85
Slip-casting layer 1dual layer
2 dual-layers 3 dual-layers
Figure 3-10. Cross-sectional SEM images of membranes at different stages of preparation
86
Slip-casting layer 1dual-layer
2 dual-layers 3 dual-layers
Figure 3-11. Top view SEM images of membranes at different stages of preparation
87
Figures 3.10 and 3.11 show SEM images of membranes made from the (+) PS dip-coating
solution, at various stages of preparation as we deposit successively more layers on the support.
The impact of adding the various layers is clear from these images. For example, in Figure 3-11
that shows the top views of the membranes in the image of the membrane with only the slip-
casting layer on the top, we can see a lot of distinguishable and coarse small particles filling the
void space among larger particles. However, as additional layers are deposited on the top, many
of the small particles are no longer distinguishable and some of them are even “blurred out” by a
black amorphous SiC layer formed by the pyrolysis of AHPCS (this is especially true in the 3
dual-layer image). These observations in Figures 3-10 and Figure 3-11 are consistent with the
permeation data for the same membranes discussed above.
3.3.2 SiC Membranes Prepared by CVD/CVI
Our group had previously made SiC membrane by the CVD method (Chen et al., 2008) on pure
0.6 µm SiC supports. The best membranes prepared had a He permeance of ~2.6×10
-7
mol•m
-2
•s
-
1
•Pa
-1
and a (He/Ar) separation factor of 13, which are not as good as the membranes prepared
via the pyrolysis of polymeric precursors (AHPCS/PCS).
In our present work, we also carried out a considerable number of CVD experiments using both
pure and blended SiC supports or supports with a single slip-casted layer on the top. The results
of membranes made by this method were not, generally, good compared with the membranes
made following the pyrolysis procedure. CVD seemed to decrease the permeance for both Ar
and He (see for example, the data in Figure 3-12 using a 50/50 blended support with a single
slip-casted layer on the top) but never to the point where the membranes became nanoporous.
88
Since our group previously prepared high-quality SiC nanoporous using γ-alumina substrates
(Ciora et al., 2004) our inability to prepare similar quality membranes may relate to the fact that
SiC supports (even those with a slip-casted top layer) contain large pores and cracks and
imperfections which are not appropriate for the CVD approach
0.0 0.5 1.0 1.5
4.0x10
-6
8.0x10
-6
1.2x10
-5
1.6x10
-5
2.0x10
-5
Time / hr
Permeance / mol*m
-2
*s
-1
*Pa
-1
Argon
Helium
Figure 3-12. He and Ar permeances for a membrane prepared via CVD (Deposition temperature:
850
o
C, Argon flow rate: 1.2 ml/s; TPS injection rate: 30 µl/hr)
89
3.4 Conclusions
Utilizing tubular SiC macroporous sintered supports, SiC membranes were fabricated using both
pre-ceramic pyrolysis and CVD/CVI techniques. SiC membranes with a He permeance in the
range of 1.8 - 2.5×10
-7
mol•m
-2
•s
-1
•Pa
-1
and a (He/Ar) separation factor of 60 - 131 were
fabricated using the pyrolysis method. Using the CVD/CVI technique and the same supports, we
were unable to prepare membarnes with competitive characteristics.
90
4. Suggestions for Future Work
As described previously, in this research different kinds of SiC sintered supports were prepared
with very different permeation characteristics and surface roughness. How various process
parameters (sintering temperature, amount of sintering aids, and composition of initial powders)
affect the permeation characteristics, surface roughness, and porosity of these sintered SiC
supports was also systematically investigated. After trying a number of different supports, those
made from pure 0.6 µm SiC powders, sintered at 1900
o
C without any sintering aids, proved to
be a good choice, and we fabricated SiC membranes with high permeances and relatively good
separation factors. However, the factors that determine why a given support prepares a good
membrane vs. another support that it does not are still not entirely clear, and must be further
investigated. In particular, any future work needs to address the impact of surface roughness and
the pore size distribution of these supports, as from our own experience they appear to be crucial
factors in determining whether one is able to prepare crack-free, high-quality SiC nanoporous
membranes.
The characteristics of the slip-casting layer and its impact on the subsequent deposition steps
must be better understood as well. The impact of the quantity and particle size distribution of the
fine SiC powders utilized, as well as the effect of pore formers (like PS) and other additives
employed must be further investigated and better understood. In addition, the interplay between
the characteristics of the support and intermediate (slip-casting) layer need to be studied in
greater depth and detail. We know for example, from our own studies, so far, that supports with a
large average pore size and/or wide PSD require thick intermediate layers to make them
appropriate for further top-layer deposition and high-quality membrane preparation. This then
91
means that the higher permeance of such supports is more negatively impacted (from the
deposition of the intermediate layer) than the permeance of other supports with a smaller average
pore size and a narrower PSD (and a correspondingly lower starting permeance).
In terms of the deposition of the top nanoporous and highly permselective SiC layers, an area
where substantial improvements are needed is in the reproducibility in making these membranes.
Though we are able to make membranes with excellent permeation characteristics, the fraction of
such membranes is still quite small among a batch of membranes made and which show inferior
characteristics, hovering typically 20% - 30%. This is an unacceptably low success rate, and if
the SiC membranes proposed here are to eventually find commercial application, reproducibility
in their preparation must be significantly improved. A good starting point here would be a better
fundamental understanding of the reasons that cause so many of the membranes prepared under
identical conditions with others to fail. This should be the key focus of any future studies in this
area, in our opinion.
92
Nomenclature
[ ]
[ ]
[ ]
2
2
Flux of a Gas , ,, ,
s
Permeance of a Gas Permeance
s
Pressure Difference
Outside
kb s
mol
nn n n n
m
mol
m Pa
P Pa
υ
⇒ =
×
⇒=
××
⇒∆ =
[ ]
[ ]
[ ]
[ ]
0
2
Diamter
Inside Diamter
Membrane Area
Molecular Weight
i
rm
rm
A m
Mg
⇒=
⇒=
⇒ =
⇒=
[ ]
[ ]
/
Gas Constant 8.317 / ( )
Membrane Thickness
Pore Radius
Porosity
p
mol
R J mol K
l m
rm
⇒ = ⋅
⇒=
⇒ =
[ ]
[ ]
[ ]
[ ]
2
0
0
dimensionless
Permeation Coefficient
Kudsen Coefficient
Mean Molecular Speed /
T
Bm
Km
ms
ε
µ
⇒=
⇒=
⇒=
⇒ =
[ ]
[ ]
2
emperature
Average Pressure
Pore's Cross Sectional Area
Apparent Density of the Membrane
av
p
a
TK
P Pa
Am
ρ
⇒ =
⇒=
⇒ =
⇒
[ ]
[ ]
[ ]
3
21
Tortuosity of the Membrane dimensionless
Coefficient of Resistance
The Pore Length
Specific Surface of
pp
R
kg m
C kg m s
l m
τ
−
−−
=
⇒=
⇒ =⋅
⇒=
g
[ ]
2
the Solid over
which the adsorbed gas are mobile
s
S m ⇒=
93
References
Ayturk, A. E., Mardilovich, I. P., Engwall, E. E., Ma, Y. H., Synthesis of composite Pd-porous
stainless steel (PSS) membranes with a Pd/Ag intermetallic diffusion barrier. Journal of
Membrane Science 2006, 285, (1-2), 385-394.
Barnes, A. C., Zorman, C. A., Feng, P. X. L., Amorphous silicon carbide (a-SiC) thin square
membranes for resonant micromechanical devices. In Silicon Carbide and Related materials
2011, Pts 1 and 2, Devaty, R. P., Dudley, M., Chow, T. P., Neudeck, P. G., Eds. Trans Tech
Publications Ltd: Stafa-Zurich, 2012; Vol. 717-720, pp 533-536.
Bautista, M. A., Cancapa, J. Q., Fernandez, J. M., Rodriguez, M. A., Singh, M., Microstructural
and mechanical evaluation of porous biomorphic silicon carbide for high temperature filtering
applications. Journal of the European Ceramic Society 2011, 31, (7), 1325-1332.
Bouillon, E., Langlais, F., Pailler, R., Naslain, R., Cruege, F., Huong, P. V., Sarthou, J. C.,
Delpuech, A., Laffon, C., Lagarde, P., Monthioux, M., Oberlin, A., Conversion mechanisms of a
polycarbosilane precursor into an SiC-based ceramic material. Journal of Materials Science 1991,
26, (5), 1333-1345.
Braunbarth, C. M., Boudreau, L. C., Tsapatsis, M., Synthesis of ETS-4/TiO2 composite
membranes and their pervaporation performance. Journal of Membrane Science 2000, 174, (1),
31-42.
Chae, S. H., Kim, Y. W., Effect of inert filler addition on microstructure and strength of porous
SiC ceramics. Journal of Materials Science 2009, 44, (5), 1404-1406.
94
Chae, S. H., Kim, Y. W., Song, I. H., Kim, H. D., Narisawa, M., Porosity control of porous
silicon carbide ceramics. Journal of the European Ceramic Society 2009, 29, (13), 2867-2872.
Chen, C. L., Tsai, D. S., Permeation properties of microporous membranes prepared via coating
of evaporated polydimethylsilane. Journal of Membrane Science 2004, 237, (1-2), 163-165.
Chen, F., Mourhatch, R., Tsotsis, T. T., Sahimi, M., Experimental studies and computer
simulation of the preparation of nanoporous silicon-carbide membranes by chemical-vapor
infiltration/chemical-vapor deposition techniques. Chemical Engineering Science 2008, 63, (6),
1460-1470.
Choi, W. Y., Park, N. S., Wiesner, M. R., Kim, J. O., Preparation and characterization of fouling-
resistant photocatalytic metal membrane embedded with self-organized TiO2 nanotube. Water
Science and Technology 2010, 62, (4), 963-971.
Ciora, R. J., Fayyaz, B., Liu, P. K. T., Suwanmethanond, V., Mallada, R., Sahimi, M., Tsotsis, T.
T., Preparation and reactive applications of nanoporous silicon carbide membranes. Chemical
Engineering Science 2004, 59, (22-23), 4957-4965.
Colombo, P., Paulson, T. E., Pantano, C. G., Synthesis of silicon carbide thin films with
polycarbosilane (PCS). Journal of the American Ceramic Society 1997, 80, (9), 2333-2340.
Cussler, E. L., Diffusion Mass Transfer in Fluid Systems. Third ed.; Cambridge University Press:
New York, 2009; pp 190.
Delavari, H., Kokabi, M., Silicon carbide nanowires from polyvinyl alcohol/silica electrospun
nanofibers. Nano 2011, 6, (1), 41-45.
95
Dong, J. H., Lin, Y. S., Hu, M. Z. C., Peascoe, R. A., Payzant, E. A., Template-removal-
associated microstructural development of porous-ceramic-supported MFI zeolite membranes.
Microporous and Mesoporous Materials 2000, 34, (3), 241-253.
Drisko, G. L., Wang, X., Caruso, R. A., Strong silica monoliths with large mesopores prepared
using agarose gel templates. Langmuir 2011, 27, (6), 2124-2127.
Elyassi, B., Deng, W., Sahimi, M., Tsotsis, T. T., On the use of porous and nonporous fillers in
the fabrication of silicon carbide membranes. Industrial & Engineering Chemistry Research 2013,
52, (30), 10269-10275.
Elyassi, B., Kim, T. W., Sahimi, M., Tsotsis, T. T., Effect of polystyrene on the morphology and
physical properties of silicon carbide nanofibers. Materials Chemistry and Physics 2009, 118, (1),
259-263.
Elyassi, B., Sahimi, M., Tsotsis, T. T., Silicon carbide membranes for gas separation applications.
Journal of Membrane Science 2007, 288, (1-2), 290-297.
Elyassi, B., Sahimi, M., Tsotsis, T. T., A novel sacrificial interlayer-based method for the
preparation of silicon carbide membranes. Journal of Membrane Science 2008, 316, (1-2), 73-79.
Elyassi, B., Sahimi, M., Tsotsis, T. T., Inorganic Membranes. In Encyclopedia of Chemical
Processing, Lee, S., Ed. Taylor & Francis Group, LLC: 2009; Vol. 1, pp 1-16.
Eom, J. H., Kim, Y. W., Effect of template size on microstructure and strength of porous silicon
carbide ceramics. Journal of the Ceramic Society of Japan 2008, 116, (1358), 1159-1163.
Eom, J. H., Kim, Y. W., Effect of additive composition on microstructure and strength of porous
silicon carbide ceramics. Journal of Materials Science 2009, 44, (16), 4482-4486.
96
Eom, J. H., Kim, Y. W., Effect of additives on mechanical properties of macroporous silicon
carbide ceramics. Metals and Materials International 2010, 16, (3), 399-405.
Eom, J. H., Kim, Y. W., Song, I. H., Kim, H. D., Processing and properties of polysiloxane-
derived porous silicon carbide ceramics using hollow microspheres as templates. Journal of the
European Ceramic Society 2008, 28, (5), 1029-1035.
Fernando, J. A., Chung, D. D. L., Improving an alumina fiber filter membrane for hot gas
filtration using an acid phosphate binder. Journal of Materials Science 2001, 36, (21), 5079-5085.
Fernando, J. A., Chung, D. D. L., Thermomechanical properties of alumina fiber membrane.
Ceramics International 2005, 31, (3), 453-460.
Fukushima, M., Zhou, Y., Miyazaki, H., Yoshizawa, Y. I., Hirao, K., Iwamoto, Y., Yamazaki, S.,
Nagano, T., Microstructural characterization of porous silicon carbide membrane support with
and without alumina additive. Journal of the American Ceramic Society 2006, 89, (5), 1523-
1529.
Fukushima, M., Zhou, Y., Yoshizawa, Y. C., Hirao, K., Water vapor corrosion behavior of
porous silicon carbide membrane support. Journal of the European Ceramic Society 2008, 28, (5),
1043-1048.
Fukushima, M., Zhou, Y., Yoshizawa, Y. I., Fabrication and microstructural characterization of
porous silicon carbide with nano-sized powders. Materials Science and Engineering B-Advanced
Functional Solid-State Materials 2008, 148, (1-3), 211-214.
97
Fukushima, M., Zhou, Y., Yoshizawa, Y. I., Fabrication and microstructural characterization of
porous SiC membrane supports with Al2O3-Y2O3 additives. Journal of Membrane Science 2009,
339, (1-2), 78-84.
Fukushima, M., Zhou, Y., Yoshizawa, Y. I., Hirao, K., Oxidation Behavior of porous silicon
carbide ceramics under water vapor below 1000
o
C and their microstructural characterization.
Journal of the Ceramic Society of Japan 2006, 114, (1336), 1155-1159.
Gallaher, G. R., Liu, P. K. T., Characterization of ceramic membranes, thermal and hydrothermal
stabilities of commercial 40 angstrom membranes. Journal of Membrane Science 1994, 92, (1),
29-44.
Gallucci, F., Fernandez, E., Corengia, P., van Sint Annaland, M., Recent advances on
membranes and membrane reactors for hydrogen production. Chemical Engineering Science
2013, 92, (0), 40-66.
Gavalas, G. R., Megiris, C. E., Nam, S. W., Deposition of H2 permselective SiO2. Chemical
Engineering Science 1989, 44, (9), 1829-1835.
Gopalakrishnan, S., Costa, J. C. D., Hydrogen gas mixture separation by CVD silica membrane.
Journal of Membrane Science 2008, 323, (1), 144-147.
Guo, P., Martin, C. R., Zhao, Y. P., Ge, J., Zare, R. N., General method for producing organic
nanoparticles using nanoporous membranes. Nano Letters 2010, 10, (6), 2202-2206.
Gupta, P., Wang, W., Fan, L. S., Synthesis of high-surface-area SiC through a modified sol-gel
route: Control of the pore structure. Industrial & Engineering Chemistry Research 2004, 43, (16),
4732-4739.
98
Hasegawa, Y., Synthesis of continuous silicon-carbide fiber .6. Pyrolysis process of cured
polycarbosilane fiber and structure of SiC fiber. Journal of Materials Science 1989, 24, (4),
1177-1190.
Hasegawa, Y., Iimura, M., Yajima, S., Synthesis of continuous silicon-carbide fiber 2.
Conversion of polycarbosilane fiber into silicon-carbide fibers. Journal of Materials Science
1980, 15, (3), 720-728.
Hasegawa, Y., Okamura, K., Synthesis of continuous silicon-carbide fiber 3. Pyrolysis process of
polycarbosilane and structure of the products. Journal of Materials Science 1983, 18, (12), 3633-
3648.
Hasegawa, Y., Okamura, K., Synthesis of continuous silicon-carbide fiber 4. The structure of
polycarbosilane as the precursor. Journal of Materials Science 1986, 21, (1), 321-328.
Hilder, T. A., Yang, R., Gordon, D., Rendell, A. P., Chung, S. H., Silicon carbide nanotube as a
chloride-selective channel. Journal of Physical Chemistry C 2012, 116, (7), 4465-4470.
Hojo, J., Sintering behavior of ultrafine silicon carbide powder. In Silicon Carbide Ceramics-1,
Somiya, S., Inomata, Y., Eds. Elsevier Applied Science: 1991; pp 149.
Hosseini, S. M., Madaeni, S. S., Khodabakhshi, A. R., Preparation and characterization of
PC/SBR heterogeneous cation exchange membrane filled with carbon nano-tubes. Journal of
Membrane Science 2010, 362, (1-2), 550-559.
Ichikawa, H., Machino, F., Mitsuno, S., Ishikawa, T., Okamura, K., Hasegawa, Y., Synthesis of
continuous silicon-carbide fiber 5. Factors affecting stability of polycarbosilane to oxidation.
Journal of Materials Science 1986, 21, (12), 4352-4358.
99
Ikoma, Y., Yahaya, H., Kuriyama, K., Sakita, H., Nishino, Y., Motooka, T., Semiconductor
nanopores formed by chemical vapor deposition of heteroepitaxial SiC films on SOI(100)
substrates. Journal of Vacuum Science & Technology B 2011, 29, (6).
Iliescu, C., Chen, B. T., Poenar, D. P., Lee, Y. Y., PECVD amorphous silicon carbide
membranes for cell culturing. Sensors and Actuators B-Chemical 2008, 129, (1), 404-411.
Interrante, L. V., Whitmarsh, C. W., Sherwood, W., Fabrication of sic matrix composites by
liquid phase infiltration with a polymeric precursor. MRS Online Proceedings Library 1994, pp
365.
Interrante, L. V., Whitmarsh, C. W., Sherwood, W., Wu, H. J., Lewis, R., Maciel, G., High yield
polycarbosilane precursors to stoichiometric SiC: synthesis, pyrolysis and application. MRS
Online Proceedings Library 1994, pp 346.
Itoh, N., Tomura, N., Tsuji, T., Hongo, M., Strengthened porous alumina membrane tube
prepared by means of internal anodic oxidation. Microporous and Mesoporous Materials 1998,
20, (4-6), 333-337.
Itoh, N., Tomura, N., Tsuji, T., Hongo, M., Strengthened porous alumina membrane tube
prepared by means of internal anodic oxidation. Microporous and Mesoporous Materials 1998,
20, (4-6), 333-337.
Iwamoto, Y., Sato, K., Kato, T., Inada, T., Kubo, Y., A hydrogen-permselective amorphous
silica membrane derived from polysilazane. Journal of the European Ceramic Society 2005, 25,
(2-3), 257-264.
100
Janowska, I., Hajiesmaili, S., Begin, D., Keller, V., Keller, N., Ledoux, M. J., Pham-Huu, C.,
Macronized aligned carbon nanotubes for use as catalyst support and ceramic nanoporous
membrane template. Catalysis Today 2009, 145, (1-2), 76-84.
Kafrouni, W., Rouessac, V., Julbe, A., Durand, J., Synthesis of PECVD α-SiCxNy:H membranes
as molecular sieves for small gas separation. Journal of Membrane Science 2009, 329, (1-2),
130-137.
Kenawy, S. H., Nour, W. M. N., Microstructural evaluation of thermally fatigued SiC-reinforced
Al 2O 3/ZrO 2 matrix composites. Journal of Materials Science 2005, 40, (14), 3789-3793.
Kim, J. O., Choi, W. Y., Fabrication and application of photocatalytic TiO2 metal membrane
with self-organized nano tubes for water treatment. In Eco-Materials Processing and Design XI,
Kim, H. S., Yang, J. F., Sekino, T., Anpo, M., Lee, S. W., Eds. 2010; Vol. 658, pp 483-486.
Kim, Y., Kim, E., Kim, S. R., Suh, M. H., Choi, D. J., Kwon, W. T., Hydrogen separation
characteristics of SiC nanoporous membrane at high temperature. In Advanced Materials and
Processing, Chang, Y. W., Kim, N. J., Lee, C. S., Eds. Trans Tech Publications Ltd: Stafa-Zurich,
2007; Vol. 26-28, pp 271.
Kim, Y. H., Kim, S. R., Song, B. G., Dabhade, V. V., Sea, B. K., Kwon, W. T., Nanoporous SiC
membrane derived from preceramic polymer. In Advances in Nanomaterials and Processing, Pts
1 and 2, Ahn, B. T., Jeon, H., Hur, B. Y., Kim, K., Park, J. W., Eds. Trans Tech Publications Ltd:
Stafa-Zurich, 2007; Vol. 124-126, pp 1733-1736.
101
Kim, J. O., Choi, W. Y., Fabrication and application of photocatalytic TiO2 metal membrane
with self-organized nano tubes for water treatment. In Eco-Materials Processing and Design Xi,
Kim, H. S., Yang, J. F., Sekino, T., Anpo, M., Lee, S. W., Eds. 2010; Vol. 658, pp 483-486.
Kim, Y. W., Jin, Y. J., Eom, J. H., Song, I. H., Kim, H. D., Engineering porosity in silicon
carbide ceramics. Journal of Materials Science 2010, 45, (10), 2808-2815.
Kita, K., Kondo, N., Hyuga, H., Izutsu, Y., Kita, H., Study of modification on alumina surface
by using of organosilicon polymer. Journal of the Ceramic Society of Japan 2011, 119, (1389),
378-381.
Kita, K., Kondo, N., Izutsu, Y., Kita, H., Investigation of the properties of SiC membrane on
alumina by using polycarbosilane. Materials Letters 2012, 75, 134-136.
Kleps, I., Angelescu, A., Correlations between properties and applications of the CVD
amorphous silicon carbide films. Applied Surface Science 2001, 184, (1-4), 107-112.
Kulprathipanja, A., Alptekin, G. O., Falconer, J. L., Way, J. D., Pd and Pd-Cu membranes:
inhibition of H2 permeation by H2S. Journal of Membrane Science 2005, 254, (1-2), 49-62.
Kumar, B. V. M., Lim, K. Y., Kim, Y. W., Influence of submicron SiC particle addition on
porosity and flexural strength of porous self-bonded silicon carbide. Metals and Materials
International 2011, 17, (3), 435-440.
Kusakabe, K., Li, Z. Y., Maeda, H., Morooka, S., Preparation of supported composite membrane
by pyrolysis of polycarbosilane for gas separation at high-temperature. Journal of Membrane
Science 1995, 103, (1-2), 175-180.
102
Kusakabe, K., Yokoyama, S., Morooka, S., Hayashi, J. I., Nagata, H., Development of supported
thin palladium membrane and application to enhancement of propane aromatization on Ga-
silicate catalyst. Chemical Engineering Science 1996, 51, (11), 3027-3032.
Kusakabe, K., Yoneshige, S., Murata, A., Morooka, S., Morphology and gas permeance of ZSM-
5-type zeolite membrane formed on a porous alpha-alumina support tube. Journal of Membrane
Science 1996, 116, (1), 39-46.
Kwon, W. T., Kim, S. R., Kim, J., Lee, Y. J., Bhave, R. R., Kim, Y., Effect of heat treatment
temperature and atmosphere on microstructure of polyphenylcarbosilane. In Eco-Materials
Processing and Design XII, Kim, H. S., Yang, J. F., Han, C. H., Thongtem, S. C., Lee, S. W.,
Eds. Trans Tech Publications Ltd: Stafa-Zurich, 2011; Vol. 695, pp 239-242.
Lee, C., Baik, S., Vertically-aligned carbon nano-tube membrane filters with
superhydrophobicity and superoleophilicity. Carbon 2010, 48, (8), 2192-2197.
Lee, L. L., Tsai, D. S., A hydrogen-permselective silicon oxycarbide membrane derived from
polydimethylsilane. Journal of the American Ceramic Society 1999, 82, (10), 2796-2800.
Lee, L. L., Tsai, D. S., Synthesis and permeation properties of silicon-carbon-based inorganic
membrane for gas separation. Industrial & Engineering Chemistry Research 2001, 40, (2), 612-
616.
Lee, L. L., Tsai, D.-S., Silicon carbide membranes modified by chemical vapor deposition using
species of low sticking coefficients in a silane/acetylene reaction system. Journal of the
American Ceramic Society 1998, 81, (1), 159-165.
Li, C. F., Zhong, S. H., Kieselguhr-mullite ceramic membrane substrate fabricated by line
compounding technique. Journal of Membrane Science 2002, 204, (1-2), 89-95.
Li, J., Zhang, Y., A simple purification for single-walled carbon nanotubes. Physica E: Low-
dimensional Systems and Nanostructures 2005, 28, (3), 309-312.
103
Li, K., Ceramic Membranes for Separation and Reaction. John Wiley & Sons, Ltd: Chichester,
England, 2007, pp 2, 41, 104.
Li, Z. Y., Kusakabe, K., Morooka, S., Pore structure and permeance of amorphous Si-C-O
membranes with high durability at elevated temperature. Separation Science and Technology
1997, 32, (7), 1233-1254.
Lin, P. K., Tsai, D. S., Preparation and analysis of a silicon carbide composite membrane.
Journal of the American Ceramic Society 1997, 80, (2), 365-372.
Liu, P. K. T., Sahimi, M., Tsotsis, T. T., Process intensification in hydrogen production from
coal and biomass via the use of membrane-based reactive separations. Current Opinion in
Chemical Engineering 2012, 1, (3), 342-351.
Liu, R. P., Wang, C. A., Wang, Y. B., Preparation, characterization and properties of new low
cost multi-channel ceramic membrane tube for water treatment. Energy and Environment
Materials, Eds. 2013; Vol. 743-744, pp 629-635.
Maddocks, A. R., Cassidy, D. J., Jones, A. S., Harris, A. T., Synthesis of nanoporous silicon
carbide via the preceramic polymer route. Materials Chemistry and Physics 2009, 113, (2-3),
861-867.
Mardilovich, P. P., She, Y., Ma, Y. H., Rei, M. H., Defect-free palladium membranes on porous
stainless-steel support. AIChE Journal 1998, 44, (2), 310-322.
Miyajima, K., Eda, T., Ohta, H., Ando, Y., Nagaya, S., Ohba, T., Iwamoto, Y., Development of
Si-N based hydrogen separation membrane. In Advances in Polymer Derived Ceramics and
104
Composites, Colombo, P., Raj, R., Eds. Amer Ceramic Soc: Westerville, 2010; Vol. 213, pp 87-
94.
Morooka, S., Yan, S., Kusakabe, K., Akiyama, Y., Formation of hydrogen-permselective SiO2
membrane in macropores of alpha-alumina support tube by thermal-decomposition of TEOS.
Journal of Membrane Science 1995, 101, (1-2), 89-98.
Mulder, M., Basic Principles of Membrane Technology. Second ed.; Kluwer Academic
Publishers: Dordrecht, The Netherlands, 1996, pp 60, 286.
Nagano, T., Sato, K., Fujisaki, S., Saitoh, T., Iwamoto, Y., Helium-permselective amorphous
sicmembrane modified by chemical vapor infiltration. Soft Materials 2006, 4, (2-4), 109-122.
Nagano, T., Sato, K., Saitoh, T., Iwamoto, Y., Gas permeation properties of amorphous SiC
membranes synthesized from polycarbosilane without oxygen-curing process. Journal of the
Ceramic Society of Japan 2006, 114, (1330), 533-538.
Neomagus, H. W. J. P., van Swaaij, W. P. M., Versteeg, G. F., The catalytic oxidation of H2S in
a stainless steel membrane reactor with separate feed of reactants. Journal of Membrane Science
1998, 148, (2), 147-160.
Nhu-Ngoc, B., Lind, M. L., Hoek, E. M. V., McCutcheon, J. R., Electrospun nanofiber supported
thin film composite membranes for engineered osmosis. Journal of Membrane Science 2011, 385,
(1-2), 10-19.
Nomura, M., Aida, H., Gopatakrishnan, S., Sugawara, T., Nakao, S., Yamazaki, S., Inada, T.,
Iwamoto, Y., Steam stability of a silica membrane prepared by counter-diffusion chemical vapor
deposition. Desalination 2006, 193, (1-3), 1-7.
105
Nomura, M., Ono, K., Gopalakrishnan, S., Sugawara, T., Nakao, S. I., Preparation of a stable
silica membrane by a counter diffusion chemical vapor deposition method. Journal of Membrane
Science 2005, 251, (1-2), 151-158.
Omiya, T., Tanaka, A., Shimomura, M., Morphological study on porous silicon carbide
membrane fabricated by double-step electrochemical etching. Japanese Journal of Applied
Physics 2012, 51, (7).
Osswald, S., Flahaut, E., Gogotsi, Y., In situ Raman spectroscopy study of oxidation of double-
and single-wall carbon nanotubes. Chemistry of Materials 2006, 18, (6), 1525-1533.
Pagès, X., Rouessac, V., Cot, D., Nabias, G., Durand, J., Gas permeation of PECVD membranes
inside alumina substrate tubes. Separation and Purification Technology 2001, 25, (1-3), 399-406.
Paglieri, S. N., Way, J. D., Innovations in palladium membrane research. Separation and
Purification Methods 2002, 31, (1), 1-169.
Park, T. J., Banerjee, S., Hemraj-Benny, T., Wong, S. S., Purification strategies and purity
visualization techniques for single-walled carbon nanotubes. Journal of Materials Chemistry
2006, 16, (2), 141-154.
Pashchanka, M., Engstler, J., Schneider, J. J., Siozios, V., Fasel, C., Hauser, R., Kinski, I., Riedel,
R., Lauterbach, S., Kleebe, H. J., Flege, S., Ensinger, W., Polymer-derived SiOC nanotubes and
nanorods via a template approach. European Journal of Inorganic Chemistry 2009, (23), 3496-
3506.
Prochazka, S., Sintering of Silicon Carbide. In Ceramics For High Performance Applications, 1st
ed.; Burke, J. J., Gorum, A. E., Katz, R. N., Eds. Brook Hill Publishing Company: Ohio, 1974.
106
Ravanchi, M. T., Kaghazchi, T., Kargari, A., Application of membrane separation processes in
petrochemical industry: a review. Desalination 2009, 235, (1-3), 199-244.
Roa, F., Way, J. D., McCormick, R. L., Paglieri, S. N., Preparation and characterization of Pd-Cu
composite membranes for hydrogen separation. Chemical Engineering Journal 2003, 93, (1), 11-
22.
Rosenbloom, A. J., Nie, S., Ke, Y., Devaty, R. P., Choyke, W. J., Columnar morphology of
porous silicon carbide as a protein-permeable membrane for biosensors and other applications. In
Silicon Carbide and Related Materials 2005, Pts 1 and 2, Devaty, R. P., Ed. Trans Tech
Publications Ltd: Stafa-Zurich, 2006; Vol. 527-529, pp 751-754.
Sahimi, M., Heterogeneous Materials I. Springer: New York, 2003; Vol. I, pp 691.
Sahimi, M., Radom close packing versus maximally random jamming. In Heterogeneous
Materials I, Springer: New York, 2003; Vol. I, pp 66, 80.
Samingprai, S., Tantayanon, S., Ma, Y. H., Chromium oxide intermetallic diffusion barrier for
palladium membrane supported on porous stainless steel. Journal of Membrane Science 2010,
347, (1-2), 8-16.
Schmidt, H., Koch, D., Grathwohl, G., Colombo, P., Micro-/Macroporous ceramics from
preceramic precursors. Journal of the American Ceramic Society 2001, 84, (10), 2252-2255.
Schulz, K., Durst, M., Advantages of an integrated system for hot gas filtration using rigid
ceramic elements. Filtration & Separation 1994, 31, (1), 25-28.
107
Sea, B. K., Ando, K., Kusakabe, K., Morooka, S., Separation of hydrogen from steam using a
SiC-based membrane formed by chemical vapor deposition of triisopropylsilane. Journal of
Membrane Science 1998, 146, (1), 73-82.
Sharif, S. M., Ahmad, Z. A., Othman, M. R., Preparation and characterization of tubular ceramic
support for use in transesterification process. Science of Sintering 2012, 44, (2), 187-195.
Shi, X., Wang, Y., Ren, L., Zhao, N., Gong, Y., Wang, D. A., Novel mesoporous silica-based
antibiotic releasing scaffold for bone repair. Acta Biomaterialia 2009, 5, (5), 1697-1707.
Song, Y. Y., Schmidt-Stein, F., Berger, S., Schmuki, P., TiO2 nano test tubes as a self-cleaning
platform for high-sensitivity immunoassays. Small 2010, 6, (11), 1180-1184.
Soraru, G. D., Babonneau, F., Mackenzie, J. D., Structural evolutions from polycarbosilane to
SiC ceramic. Journal of Materials Science 1990, 25, (9), 3886-3893.
Suda, H., Yamauchi, H., Uchimaru, Y., Fujiwara, I., Haraya, K., Preparation and gas permeation
properties of silicon carbide-based inorganic membranes for hydrogen separation. Desalination
2006, 193, (1-3), 252-255.
Suda, H., Yamauchi, H., Uchimaru, Y., Fujiwara, I., Haraya, K., Structural evolution during
conversion of polycarbosilane precursor into silicon carbide-based microporous membranes.
Journal of the Ceramic Society of Japan 2006, 114, (1330), 539-544.
Suh, M. H., Kwon, W. T., Kim, E. B., Kim, S.-R., Bae, S. Y., Choi, D. J., Kim, Y., H2
permeable nanoporous SiC membrane for an IGCC application. Journal of Ceramic Processing
Research 2009, 10, (3), 359-363.
108
Sung, I. K., Yoon, S. B., Yu, J. S., Kim, D. P., Fabrication of macroporous SiC from templated
preceramic polymers. Chemical Communications 2002, (14), 1480-1481.
Suwanmethanond, V., Goo, E., Liu, P. K. T., Johnston, G., Sahimi, M., Tsotsis, T. T., Porous
silicon carbide sintered substrates for high-temperature membranes. Industrial & Engineering
Chemistry Research 2000, 39, (9), 3264-3271.
Takeda, Y., Nakamura, K., Maeda, K., Matsushita, Y., Effects of elemental additives on
electrical resistivity of silicon carbide ceramics. Journal of the American Ceramic Society 1987,
70, (10), 266-267.
Takeda, Y., Shibata, N., Kubo, Y., SiC coating on porous gamma-Al2O3 using alternative-supply
CVI method. Journal of the Ceramic Society of Japan 2001, 109, (4), 305-309.
Takeyama, A., Sugimoto, M., Yoshikawa, M., Gas permeation property of SiC membrane using
curing of polymer precursor film by electron beam irradiation in helium atmosphere. Materials
Transactions 2011, 52, (6), 1276-1280.
Tanaka, H., Sintering of silicon carbide. In Silicon Carbide Ceramics-1, Somiya, S., Inomata, Y.,
Eds. Elsevier Applied Science: 1991; p 213.
Tang, Z., Qiu, C., McCutcheon, J. R., Yoon, K., Ma, H., Fang, D., Lee, E., Kopp, C., Hsiao, B.
S., Chu, B., Design and fabrication of electrospun polyethersulfone nanofibrous scaffold for
high-flux nanofiltration membranes. Journal of Polymer Science Part B-Polymer Physics 2009,
47, (22), 2288-2300.
109
Tulliani, J. M., Lombardi, M., Palmero, P., Fornabaio, M., Gibson, L. J., Development and
mechanical characterization of novel ceramic foams fabricated by gel-casting. Journal of the
European Ceramic Society 2013, 33, (9), 1567-1576.
Vales, F., Rezakhanlou, R., Olagnon, C., Determination of the fracture mechanical parameters of
porous ceramics from microstructure parameters measured by quantitative image analysis.
Journal of Materials Science 1999, 34, (16), 4081-4088.
Vecitis, C. D., Schnoor, M. H., Rahaman, M. S., Schiffman, J. D., Elimelech, M.,
Electrochemical multiwalled carbon nanotube filter for viral and bacterial removal and
inactivation. Environmental Science & Technology 2011, 45, (8), 3672-3679.
Wach, R. A., Sugimoto, M., Idesaki, A., Yoshikawa, M., Molecular sieve SiC-based membrane
for hydrogen separation produced by radiation curing of preceramic polymers. Materials Science
and Engineering B-Solid State Materials for Advanced Technology 2007, 140, (1-2), 81-89.
Wach, R. A., Sugimoto, M., Yoshikawa, M., Formation of silicone carbide membrane by
radiation curing of polycarbosilane and polyvinylsilane and its gas separation up to 250
o
C.
Journal of the American Ceramic Society 2007, 90, (1), 275-278.
Wandera, D., Wickramasinghe, S. R., Husson, S. M., Modification and characterization of
ultrafiltration membranes for treatment of produced water. Journal of Membrane Science 2011,
373, (1-2), 178-188.
Wei, J. A., Synthesis of nanoporous silicon carbide ceramics by thermal evaporation process.
Applied Surface Science 2010, 256, (22), 6626-6629.
110
Wickramasinghe, S. R., Stump, E. D., Grzenia, D. L., Husson, S. M., Pellegrino, J.,
Understanding virus filtration membrane performance. Journal of Membrane Science 2010, 365,
(1-2), 160-169.
Wu, J. C. S., Sabol, H., Smith, G. W., Flowers, D. L., Liu, P. K. T., Characterization of
hydrogen-permselective microporous ceramic membranes. Journal of Membrane Science 1994,
96, (3), 275-287.
Yajima, S., Hasegawa, Y., Hayashi, J., Iimura, M., Synthesis of continuous silicon-carbide fiber
with high-tensile strength and high Youngs modulus 1. Synthesis of polycarbosilane as precursor.
Journal of Materials Science 1978, 13, (12), 2569-2576.
Yan, S. C., Maeda, H., Kusakabe, K., Morooka, S., Akiyama, Y., Hydrogen-permselective SiO2
membrane formed in pores of alumina support tube by chemical-vapor-deposition with tetraethyl
orthosilicate. Industrial & Engineering Chemistry Research 1994, 33, (9), 2096-2101.
Yang, G. C. C., Tsai, C. M., Effects of starch addition on characteristics of tubular porous
ceramic membrane substrates. Desalination 2008, 233, (1-3), 129-136.
Yang, H., Zhao, H. S., Li, Z. Q., Zhang, K. H., Liu, X. X., Tang, C. H., Microstructure evolution
process of porous silicon carbide ceramics prepared through coat-mix method. Ceramics
International 2012, 38, (3), 2213-2218.
Yang, J. F., Zhang, G. J., Kondo, N., She, J. H., Jin, Z. H., Ohji, T., Kanzaki, S., Porous 2H-
silicon carbide ceramics fabricated by carbothermal reaction between silicon nitride and carbon.
Journal of the American Ceramic Society 2003, 86, (6), 910-914.
111
Yang, W., Su, Y. J., Araki, H., Abe, E., Suzuki, H., Noda, T., Synthesis of nanotube array
composed of an amorphous matrix embedded with NaCl-type SiC crystallites by chemical vapor
infiltration techniques. Journal of the American Ceramic Society 2010, 93, (6), 1557-1560.
Yip, N. Y., Tiraferri, A., Phillip, W. A., Schiffman, J. D., Elimelech, M., High performance thin-
film composite forward osmosis membrane. Environmental Science & Technology 2010, 44,
(10), 3812-3818.
Zahir, M. H., Sato, K., Mori, H., Iwamoto, Y., Preparation and properties of hydrothermally
stable gamma-alumina-based composite mesoporous membranes. Journal of the American
Ceramic Society 2006, 89, (9), 2874-2880.
Zhang, H., Li, S., White, C. J. B., Ning, X., Nie, H., Zhu, L., Studies on electrospun nylon-
6/chitosan complex nanofiber interactions. Electrochimica Acta 2009, 54, (24), 5739-5745.
Zhang, X. W., Du, A. J., Lee, P. F., Sun, D. D., Leckie, J. O., TiO2 nanowire membrane for
concurrent filtration and photocatalytic oxidation of humic acid in water. Journal of Membrane
Science 2008, 313, (1-2), 44-51.
Zhang, X. W., Zhang, T., Ng, J., Sun, D. D., High-performance multifunctional TiO2 nanowire
ultrafiltration membrane with a hierarchical layer structure for water treatment. Advanced
Functional Materials 2009, 19, (23), 3731-3736.
Zhang, Z., Wang, F., Yu, X., Wang, Y., Yan, Y., Li, K., Luan, Z., Porous silicon carbide
ceramics produced by a carbon foam derived from mixtures of mesophase pitch and Si particles.
Journal of the American Ceramic Society 2009, 92, (1), 260-263.
112
Zhong, P. S., Widjojo, N., Chung, T.-S., Weber, M., Maletzko, C., Positively charged
nanofiltration (NF) membranes via UV grafting on sulfonated polyphenylenesulfone (sPPSU) for
effective removal of textile dyes from wastewater. Journal of Membrane Science 2012, 417-418,
(0), 52-60.
Zhong, Y., Kong, Y., Shen, X., Cui, S., Yi, X., Zhang, J., Synthesis of a novel porous material
comprising carbon/alumina composite aerogels monoliths with high compressive strength.
Microporous and Mesoporous Materials 2013, 172, (0), 182-189.
Zhongyang, L., Kusakabe, K., Morooka, S., Preparation of thermostable amorphous Si-C-O
membrane and its application to gas separation at elevated temperature. Journal of Membrane
Science 1996, 118, (2), 159-168.
Zhou, W., Yang, J. L., Sun, G. S., Liu, X. F., Yang, F. H., Li, J. M., Fracture properties of silicon
carbide thin films by bulge test of long rectangular membrane. Journal of
Microelectromechanical Systems 2008, 17, (2), 453-461.
Zhou, Y., Fukushima, M., Miyazaki, H., Yoshizawa, Y. I., Hirao, K., Iwamoto, Y., Sato, K.,
Preparation and characterization of tubular porous silicon carbide membrane supports. Journal of
Membrane Science 2011, 369, (1-2), 112-118.
Zorman, C. A., Fleischman, A. J., Dewa, A. S., Mehregany, M., Jacob, C., Nishino, S., Pirouz, P.,
Epitaxial-growth of 3C-SiC films on 4 inch diam (100) silicon-wafers by atmospheric-pressure
chemical-vapor-deposition. Journal of Applied Physics 1995, 78, (8), 5136-5138.
113
Abstract (if available)
Abstract
Efficient separation of hydrogen (H₂) under high temperatures and pressures is important to the development of the clean-energy industry, and has been among the key drivers for research on inorganic membranes for the last two decades. Although substantial efforts have been devoted to date to the preparation of nanoporous membranes for H₂ separation, the fabrication of high-temperature and steam-stable inorganic membranes with high hydrogen fluxes and large separation factors still remains a key challenge. Among all the potential candidates, silicon carbide (SiC) membranes show potential advantages for use in hydrogen separation processes under harsh and corrosive conditions such as, for example, the steam reforming and the water gas shift reactions commonly employed in H₂ production
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Fabrication of nanoporous silicon carbide membranes for gas separation applications
PDF
Experimental studies and computer simulation of the preparation of nanoporous silicon-carbide membranes by chemical-vapor infiltration/chemical-vapor deposition techniques
PDF
A process-based molecular model of nano-porous silicon carbide membranes
PDF
Continuum and pore netwok modeling of preparation of silicon-carbide membranes by chemical-vapor deposition and chemical-vapor infiltration
PDF
Silicon carbide ceramic membranes
PDF
Molecular modeling of silicon carbide nanoporous membranes and transport and adsorption of gaseous mixtures therein
PDF
The use of carbon molecule sieve and Pd membranes for conventional and reactive applications
PDF
Exploring properties of silicon-carbide nanotubes and their composites with polymers
PDF
Fabrication of nanoporous silicon oxycarbide materials via a sacrificial template technique
PDF
Methanol synthesis in a membrane reactor
PDF
Hydrogen storage in carbon and silicon carbide nanotubes
PDF
On the use of membrane reactors in biomass utilization
PDF
A hybrid adsorbent-membrane reactor (HAMR) system for hydrogen production
PDF
Development of carbon molecular-sieve membranes with tunable properties: modification of the pore size and surface affinity
PDF
Biogas reforming: conventional and reactive separation processes and the preparation and characterization of related materials
PDF
Fabrication of silicon-based membranes via vapor-phase deposition and pyrolysis of organosilicon polymers
PDF
Studies of transport phenomena in hydrotalcite membranes, and their use in direct methanol fuel cells
PDF
A flow-through membrane reactor for destruction of a chemical warfare simulant
PDF
Thermal properties of silicon carbide and combustion mechanisms of aluminum nanoparticle
PDF
A study of the application of membrane-based reactive separation to the carbon dioxide methanation
Asset Metadata
Creator
Deng, Wangxue
(author)
Core Title
Fabrication of silicon carbide sintered supports and silicon carbide membranes
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Chemical Engineering
Publication Date
12/04/2013
Defense Date
09/27/2013
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,silicon carbide membranes,silicon carbide supports
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Tsotsis, Theodore T. (
committee chair
), Lu, Grace (
committee member
), Sahimi, Muhammad (
committee member
)
Creator Email
davekozg@gmail.com,wangxued@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-355292
Unique identifier
UC11297580
Identifier
etd-DengWangxu-2207.pdf (filename),usctheses-c3-355292 (legacy record id)
Legacy Identifier
etd-DengWangxu-2207.pdf
Dmrecord
355292
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Deng, Wangxue
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
silicon carbide membranes
silicon carbide supports