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 nanoporous silicon carbide membranes for gas separation applications
(USC Thesis Other)
Fabrication of nanoporous silicon carbide membranes for gas separation applications
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
FABRICATION OF NANOPOROUS SILICON CARBIDE MEMBRANES FOR GAS SEPARATION APPLICATIONS by Bahman Elyassi A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (CHEMICAL ENGINEERING) August 2009 Copyright 2009 Bahman Elyassi ii Dedication To my beautiful wife Nafiseh for her true love and support. To my family for their encouragement. iii Acknowledgements I would like to thank my advisors Professor Theodore T. Tsotsis and Professor Muhammad Sahimi, for their support and insightful comments, as well as the freedom to pursue different ideas during the course of this research. My thanks and appreciation go to Professors Massoud Pirbazari, Professor Ted Lee, Jr., and Professor Edward Goo for serving on my qualifying exam committee. I am especially thankful to Professor Pirbazari for spending his valuable time in providing his useful perspectives on professionalism. I would like to give special thanks to my dear wife Nafiseh for her endless support and patience. There are no words that can express my deepest appreciation for all the sacrifices she made for me throughout my research life. To my colleagues; Ryan Mourhatch, Tae Wook Kim, Hyun Tae Wang, and Babak Fayyaz thank you for your assistance and friendship. I would also like to thank Ms. Tina Silva, Ms. Heather Alexander, Mr. Brendan Char, and Ms. Karen Woo of the Mork Family Department of Chemical Engineering & Materials Science for their kindness. Financial support of the National Science Foundation and the US Department of Energy is gratefully acknowledged. iv Table of Contents Dedication ii Acknowledgements iii List of Tables vi List of Figures vii Abstract ix Chapter 1 : Inorganic Membranes 1 1. Introduction 1 2. Membranes in General 5 2.1. Metal Membranes 6 2.2. Zeolite Membranes 12 2.3. Dense Oxygen Transport Membranes 17 2.3.1. Perovskites 19 2.3.2. Fluorite-type oxides 20 2.3.3. Other Non-Perovskite Structures 22 2.4. Proton-Conductive Membranes 23 2.5. Carbon Membranes 25 2.6. Silica Membranes 28 2.7. Other Inorganic Membranes 31 3. Characterization Techniques 32 4. Module Design 33 5. Applications 35 6. Concluding Remarks 36 Chapter 2 : Preparation of Silicon Carbide Membranes and Their Steam Stability 38 1. Introduction 38 2. Experimental 44 3. Results and Discussion 48 3.1. Membrane Characterization and Performance 48 3.2. The Effect of Low Temperature Oxidation on Membrane Performance 55 3.3. Steam Stability of the Membranes 59 4. Conclusions 65 Chapter 3 : A Novel Sacrificial Interlayer-Based Method for the Preparation of Silicon Carbide Membranes 66 1. Introduction 66 v 2. Experimental 68 3. Results and Discussion 75 3.1. Support Characterization 75 3.2. Membrane Characterization and Performance 77 4. Conclusions 85 Chapter 4 : Effect of Polystyrene on the Morphology and Physical Properties of Silicon Carbide Nanofibers 87 1. Introduction 87 2. Experimental 90 3. Results and Discussion 91 4. Conclusions 103 Chapter 5 : Effect of Porous and Nonporous Fillers on the Performance of Silicon Carbide Membranes 104 1. Introduction 104 2. Experimental 104 3. Results and Discussion 109 4. Conclusions 116 References 117 vi List of Tables Table 1-1 Some of the important and well-established applications of zeolites. (Rao et al., 2004). Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. .............................................................. 13 Table 1-2 Membrane characterization techniques. ........................................................... 33 Table 1-3 Typical surface area/volume for different membrane modules (Burggraaf and Cot, 1996; Nunes and Peinemann, 2006). .............................. 34 Table 2-1 Range of single-gas permeances and separation factors, measured at 200 o C, prepared by pyrolysis at 750 o C for 2 h. ............................................. 53 Table 2-2 Single- and mixed-gas permeation properties of two SiC membranes (I, II), measured at 200 o C, prepared by pyrolysis at 750 o C for 2 h. For the mixed gas experiments the membranes were exposed to an equimolar mixture of H 2 /CH 4 or H 2 /CO 2 ......................................................... 54 Table 2-3 Comparison of permeances and ideal separation factors of a SiC membrane before and after oxidation at 450 o C for 2 h. ................................. 57 Table 3-1 Range of single-gas permeances and separation factors, measured at 200 o C, for membranes prepared by pyrolysis at 750 o C for 2 h...................... 82 Table 3-2 Mixed-gas permeation properties of SiC membranes, measured at 200 o C, prepared by pyrolysis at 750 o C for 2 h. For the mixed-gas experiments the membranes were exposed to an equimolar mixture of H 2 /CH 4 or H 2 /CO 2 . ...................................................................................... 84 Table 5-1 Range of single-gas permeances and separation factors, measured at 200 o C, for membranes prepared by nanofiber fillers and pyrolyzed at 750 o C for 2 h. ............................................................................................ 111 Table 5-2 Mixed-gas permeation properties of SiC membranes (with nanofiber and powder fillers), measured at 200 o C, prepared by pyrolysis at 750 o C for 2 h. For the mixed-gas experiments the membranes were exposed to an equimolar mixture of H 2 /CH 4 . ................................................ 111 vii List of Figures Figure 1-1 The structure of inorganic membranes.............................................................. 5 Figure 2-1 XRD of an AHPCS-derived amorphous SiC powder pyrolyzed at 750 o C for 2 h along with Joint Committee on Powder Diffraction Standards (JCPDS) card number 29-1129 for crystalline β-SiC. ................... 50 Figure 2-2 SEM pictures showing the cross-section and the top view of a SiC membrane........................................................................................................ 52 Figure 2-3 H 2 permeance as a function of temperature for a SiC membrane. Solid line, membrane without an air oxidation treatment. Broken line, after the membrane had been subjected to a 2 h air oxidation at 450 o C.............................................................................................................. 55 Figure 2-4 H 2 and Ar permeance and ideal separation factor as a function of temperature for a SiC membrane at 207 kPa (30 psi) transmembrane pressure. .......................................................................................................... 58 Figure 2-5 Single-gas He permeance at a transmembrane pressure of 207 kPa (30 psi) at 200 o C. ........................................................................................... 61 Figure 2-6 Single-gas Ar permeance at a transmembrane pressure of 207 kPa (30 psi) at 200 o C. ........................................................................................... 62 Figure 2-7 He flow rate of permeate gas at 200 o C........................................................... 63 Figure 2-8 Ideal selectivity of He/Ar at a transmembrane pressure of 207 kPa (30 psi) at 200 o C. ........................................................................................... 64 Figure 3-1 SiC tubular supports sintered at 1700 o C for 3 h............................................. 70 Figure 3-2 A schematic of the new preparation technique. .............................................. 73 Figure 3-3 Permeance vs. average pressure across the membrane for He and Ar............ 76 Figure 3-4 SEM picture showing the cross-section of a SiC membrane after three dual (PS+AHPCS) coatings, and one additional coating of PS............. 79 Figure 3-5 He, H 2 , and Ar permeance as a function of temperature for a SiC membrane at 207 kPa (30 psi) transmembrane pressure. ............................... 83 viii Figure 4-1 Two sides of the AAO templates showing a pore diameter of ~100 nm on one side, and ~200 nm on the other side.............................................. 92 Figure 4-2 (A) The TEM, and (B) SEM of SiC nanofibers.............................................. 93 Figure 4-3 (A) TEM, and (B) SEM photographs of SiC nanofibers, prepared by adding 10 wt% of the PS into the AHPCS solution. Hollow domains can be seen as the result of the PS decomposition. The embedded SAED in the TEM picture indicates the amorphous structure of the fibers ............................................................................................................... 95 Figure 4-4 TEM picture of SiC nanofibers, prepared by 10 wt% addition of the PS (AHPCS basis). The nanofibers have a porous structure and hollow ellipsoidal domains can be seen, resulting from the PS decomposition................................................................................................. 97 Figure 4-5 Nitrogen adsorption isotherm of SiC nanofibers formed by the 10 wt% addition of the PS (AHPCS basis), together with the HK pore size distribution for the nanoporous region. ................................................... 98 Figure 4-6 Nitrogen adsorption isotherm of SiC nanofibers, prepared by 30 wt% of the AHPCS in toluene, together with the HK pore size distribution for the nanoporous region.......................................................... 100 Figure 4-7 TEM pictures of SiC nanofibers and hollow fibers, prepared by adding, (A) 5, (B) 15, and (C) 30 wt% PS (AHPCS basis) to 20 wt% AHPCS in toluene solution........................................................................... 101 Figure 5-1 Adsorption isotherm of A) SiC nanofibers and B) SiC powders.................. 110 Figure 5-2 SEM top view picture of a SiC membrane with nanofiber fillers................. 112 Figure 5-3 SEM cross-section of SiC membranes A) with SiC powder fillers and B) with SiC nanofiber fillers.................................................................. 113 Figure 5-4 Adsorption isotherm of nanofibers coated by SiC pre-ceramic polymer (AHPCS) solution and pyrolyzed at 750 o C for 2 h........................ 115 ix Abstract Silicon carbide (SiC) microporous membranes were prepared by the pyrolysis of thin allyl-hydridopolycarbosilane (AHPCS) films coated, using a combination of slip-casting and dip-coating techniques, on tubular SiC macroporous supports. Combining slip- casting with dip-coating significantly improved the reproducibility in preparing high quality membranes. The membranes prepared exhibited an ideal H 2 /CO 2 selectivity in the range of (42-96), and a H 2 /CH 4 ideal selectivity in the range of (29-78). Separation factors measured with the same membranes, using equimolar binary mixtures of H 2 in CO 2 and H 2 in CH 4 , were similar to the ideal selectivity values. Steam stability experiments with the membranes lasting 21 days, using an equimolar flowing mixture of He/H 2 O at 200 o C, indicated some initial decline in the permeance of He, after which the permeance became stable at these conditions. Further, a novel method was presented based on periodic and alternate coatings of polystyrene sacrificial interlayers and SiC pre-ceramic layers on the top of slip-casted tubular SiC supports. Membranes prepared by this technique exhibit single gas ideal separation factors of helium and hydrogen over argon in the ranges (176-465) and (101- 258), respectively, with permeances that are typically two to three times higher than those of SiC membranes prepared previously by the more conventional techniques. Mixed-gas experiments with the same membranes indicate separation factors as high as 117 for an equimolar H 2 /CH 4 mixture. x In the next phase of this research, SiC fibers were fabricated by immersion of anodized aluminum oxide templates in AHPCS solutions and subsequent pyrolysis at 750 o C. The effect of polystyrene (the pore former) on the formation of the nanofibers in the confined channels of the templates was studied. Fibers prepared by adding polystyrene exhibit higher surface area than the ones prepared without adding polystyrene. Finally, the effect of porous (SiC nanofibers) and nonporous (SiC powders) fillers on the performance of SiC membranes was studied. We found membranes prepared with SiC powders in their structure exhibited a higher performance. This was attributed to better packing of SiC powders compared to SiC fibers, as well as, the loss of porosity of the fibers after being implemented into the membrane structure as revealed by BET analysis. 1 Chapter 1 : Inorganic Membranes 1 1. Introduction This chapter provides an overview of the properties of a broad range of inorganic membrane materials, such as Pd and its alloys, various amorphous metals, zeolites, perovskites, carbon, and silica that are all used for the preparation of microporous and dense inorganic membranes. Methods for the preparation of the membranes and techniques for characterizing their properties are described, together with the challenges that they face when being utilized in harsh environments, and the potential solutions to the challenges. This chapter also includes brief discussions of industrial applications of the membranes, and of their modules. Membranes can potentially provide superior performance, when compared to the more traditional separation processes (adsorption/absorption, conventional and cryogenic, distillation, etc.), in terms of the overall energy savings, reduction in the initial capital investment, the required facility foot-print, and the ease in adding additional capacity (1). 1 B. Elyassi, M. Sahimi, T.T. Tsotsis, “Inorganic Membranes”, Encyclopedia of Chemical Processing, edited by S. Lee, 2009. Copyright Taylor & Francis. Reproduced with permission. 2 The 2008 market for membranes was estimated at ~$12 billion, covering such diverse applications as reverse osmosis (RO, with membrane average pore size d p <1 nm), nanofiltration (NF, d p =1-2 nm), ultrafiltration (UF, d p =2-100 nm), microfiltration (MF, d p >100 nm), and gas separation. Of the types of membranes used, about 90% are polymeric (including 4% that are cellulose-based) and the remaining 10% are inorganic (Sutherland, 2004). The preference for polymeric membranes is due to their lower cost, when compared to inorganic membranes, together with their superior processability, but also due to the familiarity that various practitioners in the field have with such membranes. Inorganic membranes, on the other hand, have better chemical, thermal, and mechanical stability than their polymeric counterparts, which makes them better suited for use in corrosive or high-pressure and temperature environments, and in applications which require frequent membrane regeneration and sterilization. To date, the vast majority of commercial uses of inorganic membranes involve liquid-phase separations (Bhave, 1991; Hsieh, 1996; Sutherland, 2004). A variety of mesoporous (2 nm <d p <50 nm) and/or macroporous (d p > 50 nm) inorganic membranes are commercially available, fabricated from materials such as titania, zirconia, and alumina using traditional techniques like powder extrusion, slip-casting, and sol-gel methods. They find liquid-phase MF or UF separation applications, such as in water treatment, in the beverage and dairy industry (for the production of fruit juices, beer, wine, whey, and various other milk products), in biotechnology, and in the petrochemical industries (Bhave, 1991; Hsieh, 1996; Schaefer, 3 Fane and Waite, 2005; Tsuru, 2001). For many of the water treatment, beverage, food, and biotechnology industry applications, fouling (biological, organic, colloidal, and mineral) is a major issue (Plottu, Houssais, Democrate, Gatel and Cavard, 2003; Schaefer et al., 2005), which necessitates regeneration techniques such as back-flushing using high pressures or chemical treatment that polymeric membranes are not suitable for. Additionally, inorganic membranes can be autoclaved or sterilized in situ without adverse effect on their performance. These are, definitely, specialty areas for the application of inorganic membrane that are likely to grow with the related industries, and as industrial practitioners in these areas are becoming more familiar with the value-added features of these membranes. Due to the commercial success and acceptance that mesoporous and macroporous inorganic membranes have enjoyed, there is already a substantial technical literature on the topic (Bhave, 1991; Hsieh, 1996). This chapter focuses, instead, on current efforts for the development of either microporous (d p <2 nm) or completely dense inorganic membranes. There is much current interest in such membranes, due to their potential for a broad range of uses, from liquid-phase RO and NF applications, to pervaporation (PV), vapor permeation (VP) and gas-phase separation (GS), and reactive separation (membrane reactor, MR) applications. Though there are already some commercial applications (e.g., PV for ethanol dehydration (http://www.mes.co.jp/english/press/2007/20070730.html; Caro and Noack, 2008)), these 4 membranes are, as yet, far from realizing their vast commercial potential, as there are still many technical challenges that remain which hinder broader commercial implementation. The key challenge comes from the fact that the aforementioned applications (NF, PV, VP, GS, MR, etc.) require strict control on the membrane selectivity, which in turn dictates tight control on the average pore size, pore size distribution, and surface characteristics. With molecular-size pores comes, of course, the challenge of low permeation rates, which in turn mandates the need for developing asymmetric membrane structures (see the discussion below) consisting of narrow-pore thin layers on the top of meso- or macroporous supports. Thermal expansion mismatch between the support and the thin membrane layer then becomes an issue, particularly for the high-temperature MR applications. Long-term material stability, sealing, and proper module design are among the other challenges one faces at high-temperature conditions (in addition, of course, to the omnipresent challenge of producing cost-effective membranes with reproducible high-level performance). Thus, the goal of this chapter is to present the state-of-the-art developments in the field of microporous and dense inorganic membranes, both from a material as well as from an applications perspective, and to provide a broad view of the challenges that these systems face. It should be noted, however, that in many of the areas of potential application, most notably high-temperature and pressure reactive separations, such membranes face virtually no competition, as polymeric membranes cannot sustain the harsh challenging environments. 5 2. Membranes in General Depending on their structure, membranes can be classified in various ways: dense and porous, and as symmetric (homogeneous) and asymmetric (composite), as shown in Fig. 1-1. In an asymmetric membrane structure, the underlying macroporous support provides the mechanical strength without negatively affecting the mass transfer, while the thin permselective membrane layer on the top provides the separation functionality. The composite structure results in a membrane that is highly selective without sacrificing throughput and flux. Figure 1-1 The structure of inorganic membranes. The performance of a membrane is typically defined by two key properties, its permeability, and separation factor. The permeability is defined as: Q Permeability permeance thickness L Ap =× =× ∆ Symmetric Dense Asymmetric Membrane Symmetric Porous Asymmetric 6 where Q (mol/s) is the flow rate of a given component through the membrane, A (m 2 ) its surface area, p ∆ (Pa) the partial pressure difference across the membrane, and L (m) the membrane thickness. The separation factor is defined in various ways, but typically as the ratio of the permeabilities of the components. The side of the membrane in contact with the feed-stream is known as the feed- or retentate-side, while the side where the various components cross into is called the permeate-side. 2.1. Metal Membranes Pd or Pd-alloy membranes, when fabricated without any defects, allow only pure hydrogen to transfer through by a solution-diffusion mechanism, while excluding all other components. Self-supported Pd or Pd-alloy membranes, with thicknesses of about 100-200 µm, have been commercially available in small-scale units since 1964 (by Johnson–Matthey), in order to generate ultra-pure hydrogen for semiconductor, military, fine-chemical industry, and laboratory applications (Dolan, Dave, Ilyushechkin, Morpeth and McLennan, 2006; Sammells and Mundschau, 2006; Uemiya, 1999). Pd-based membranes have also found use in separating hydrogen isotopes in fission or fusion reactors with high efficiency (Chen, Xing, Zhao and Chen, 2002). An industrial-scale Pd- alloy membrane unit for hydrogen recovery (shown in Fig. 1-2) was installed in an ammonia plant in the former Soviet Union in 1988 by the State Institute of Nitrogen Industry. It was reported that the plant produced hydrogen at a capacity of 2,000 Nm 3 /h (Mordkovich et al., 1993). In a more recent development, catalytic reformers for 7 hydrogen production, featuring Pd-based membrane working at high temperatures, were installed and tested by Tokyo Gas Co. (with a capacity of 40 Nm 3 /hr) and the Mitsubishi Heavy Industries (for mobile hydrogen generation), using natural gas and liquefied petroleum gas (LPG) as feed, respectively (Kikuchi, 2000). However, both units were shut-down after about 3000 h of operation, due to problems with the membranes (Yasuda and Shirasaki, 2006; Yasuda, Shirasaki, Kobayashi, Nakajima, Shimamori, Sasaki, Kabutomori, Uchida, Yoshizawa, Nishida and Shinpo, 2006). Tokyo Gas Co. is continuing the development of the Pd-based catalytic reformers. Hydrogen permeation through a dense Pd membrane is typically expressed by Sievert’s law: 2 2 , , 2 2 () Hpermeate Hfeed nn H H Pp p J t − = where 2 H J is the hydrogen flux (mol/m 2 s), 2 H P the hydrogen permeability (mol.m/m 2 .s.Pa), 2 H p the Figure 1-2 A membrane installation in Rovno, Ukraine, capable of producing hydrogen of up to 2000 Nm 3 /hr with ammonia feed (Mordkovich, Baichtock and Sosna, 1993). Copyright Johnson Matthey PLC. Reproduced with permission. 8 hydrogen partial pressure (Pa), and t (m) the thickness of membrane. n, known as the Sievert’s constant, is an exponent characterizing the flux dependence on pressure, and ranges typically from 0.5-1.0, depending on whether transport is limited by diffusion through the bulk of the membrane or by surface-exchange processes (11). There is typically a critical thickness of the membrane above which the rate-limiting mechanism is bulk-diffusion (for which n=0.5), while for thinner membranes surface processes (H 2 dissociation or associative desorption on the other side of membrane) are controlling (for which 0.5 <n ≤1) (Zhao, Pflanz, Gu, Li, Stroh, Brunner and Xiong, 1998). Pd membranes face a number of technical challenges, which include hydrogen embrittlement (low resistance to thermal cycling), and poisoning from CO and various sulfur components (mercaptans, CS 2 , COS, and H 2 S) which are commonly present in many natural feed-stocks. The embrittlement is caused by the hydrogen dissolving in Pd and forming a hydride phase, which transforms Pd from its α to β form at about 300 o C. The α and β Pd-hydride phases have different mechanical properties (e.g., thermal expansion coefficients), which causes stress and embrittlement of the Pd membranes, when they are thermally cycled in the temperature range of operation. Thus, one needs to be careful when dealing with pure Pd membranes at such conditions (e.g., evacuate the membrane first, prior to lowering the temperature). In case of H 2 S, Pd irreversibly reacts with it to form PdS, the lattice constant of which is twice that of Pd (Gao, Lin, Li and Zhang, 2004). This may result in stress and rupture of 9 the Pd membranes, but also reduces the membrane hydrogen permeability. Alloying with other metals (e.g., Cu and Au) to some extend increases the sulfur tolerance (Iyoha, Enick, Killmeyer, Howard, Ciocco and Morreale, 2007). Adsorbed H 2 O molecules at high temperatures can dissociate and contaminate the surface of Pd membrane by forming adsorbed O and negatively affecting the H 2 transport. At low temperatures CO adsorbs on Pd and restricts the adsorption of H 2 , thus reducing the flux of hydrogen, but this a reversible effect and can be significantly reduced by increasing temperature (Gao et al., 2004). There are reports on adsorption and decomposition of carbon containing compounds on the Pd membrane surface, followed by carbon formation and diffusion through membrane, and subsequent formation of various carbide phases (Hsiung, Christman, Hunter and Homyak, 1999). This is particularly true for reactive applications, where coking from carbon-based components is a major issue. Coke covers the Pd surface, forms Pd carbide or causes carbon diffusion through the membrane (and formation of a solid solution), which adversely impacts the Pd membranes performance, or may even result in mechanical failure, and (for supported membranes, see the discussion below) in delamination of the Pd top film from the support (Gao et al., 2004; Paglieri and Way, 2002). Due to the high cost of noble metals, Pd, unsupported (symmetric) Pd membranes with thicknesses in the range of ~200 µm, or larger, are uneconomical for industrial scale applications. In order to reduce the costs, alloying with cheaper metals (e.g., Cu and Ag) is generally utilized, which has the additional beneficial effect that it typically enhances 10 the properties of the Pd membranes. Alloying is known, for example, to enhance the permeability and mechanical strength, and to also reduce the natural propensity of Pd membranes for embrittlement (Paglieri and Way, 2002). The industrial standard Pd-alloy is Pd77-Ag23 (consisting of 77 wt% Pd and 23 wt% Ag), which has 1.7 times higher permeability than the pure Pd, and also exhibits higher resistance towards CO poisoning (Sammells and Mundschau, 2006). For further reduction in cost, Pd composite membranes are prepared by coating Pd or Pd-alloys on porous stainless steel (PSS), ceramic, or Vycor® glass supports. The PSS is a particularly advantageous choice, as it has a thermal expansion coefficient close to that of Pd, good mechanical durability, and because of the ease in being assembled and sealed in a stainless housing module (a great advantage for the MR application of such membranes). However, atomic inter-diffusion of the stainless steel constituents and Pd can result in hydrogen permeability degradation. In order to eliminate this effect, a hydrogen permeable interlayer (e.g., YSZ (Huang and Dittmeyer, 2007), α-Fe 2 O 3 , γ-alumina (Yepes, Cornaglia, Irusta and Lombardo, 2006), SiO 2 (Nam and Lee, 2001), CeO 2 (Tong, Matsumura, Suda and Haraya, 2005)) can be sandwiched as a barrier in between the membrane and the PSS support. Different preparation techniques have been used to fabricate Pd-based membranes (Uemiya, 1999). They include electroplating and electroless plating, physical vapor deposition (PVD), sputtering, spray pyrolysis, and chemical vapor deposition (CVD). Among these methods, electroless plating is currently the most popular (2008), because of its simplicity, low-cost, uniform deposition characteristics, and its ability to coat films 11 on nonconductive substrates. Nonetheless, each method has its own weaknesses and strengths. For instance, PVD cannot be used to form complex structures (due to shadow effects), but it provides a better control over alloy composition and thickness (Paglieri and Way, 2002; Sammells and Mundschau, 2006). For a detailed review on the various preparation techniques see (Paglieri and Way, 2002). Because of the cost, efforts have been made through the years to replace Pd with less expensive metals. Group V (Nb, Ta, V) and Group IV (Zr, Ti, Hf) metals have high H 2 permeability. They are, however, also susceptible to form hydrides (embrittlement), and are, in addition, less active for H 2 dissociation and reassociation, and subject to surface oxidation which further impacts their hydrogen permeability. Coating them with a very thin film of Pd or Pt (Buxbuam, 2001) enhances their surface exchange rates, and alloying with other metals (e.g., Cu, Ag, Ni, Fe) improves their mechanical stability by reducing their embrittlement propensity (Dolan et al., 2006; Phair and Donelson, 2006). Amorphous alloys of inexpensive elements, such as Zr, Ni, Cu and Al, that exhibit comparable hydrogen permeance with Pd, are other materials that are currently being researched (Dolan et al., 2006). Additionally, being amorphous makes them not only resistant towards forming crystalline hydride, but also mechanically more sound (Schuha, Hufnagel and Ramamurty, 2007). However, in order to prevent crystallization, the use of membranes made of such materials is limited to operating temperatures in the range 300- 500 o C, and the mechanical stability of such membranes is still a concern as well (Dolan et al., 2006). 12 2.2. Zeolite Membranes Zeolites are crystalline aluminosilicate materials with uniform pores of molecular dimensions. Tetrahedral building units of TO 4 (T=Si, Al), which are interconnected through oxygen atoms create one-dimensional (1-D), 2-D, and 3-D networks in zeolites (Sommer and Melin, 2005). Pore openings, which are determined by the size of the rings of the T atoms (usually the number of oxygen atoms) vary from about 2 Å in sodalite (SOD) with 6-atom rings, to about 10 Å in UTD-1 with 14-atom rings (Rao, Müller and Cheetham, 2004). In some zeolites, in the crystalline framework, some of Si +4 are replaced by Al +3 , in which case the framework carries a negative charge. Cations (e.g., H + ) loosely sit within the zeolite framework cavities, in order to preserve electroneutrality. The fascinating properties of zeolites, such as their ion-exchange capabilities, sorption capacity, shape and size selectivity, and catalytic activity, all result from their unique structures (Cejka, 2007). As Table 1-1 indicates, due to their interesting properties and the enormous flexibility they provide in tuning such properties, zeolites have been studied extensively and utilized in a wide range of applications. The Structural Commission of the International Zeolite Association (under the IUPAC) assigns each zeolite (based on its framework type) a three-letter code. As of February, 2007, 176 zeolite framework types have been identified (Cejka, 2007). Such codes are accessible at http://www.iza-structure.org/databases. 13 Table 1-1 Some of the important and well-established applications of zeolites. (Rao et al., 2004). Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. Ion-exchange using hydrated zeolites Detergents (e.g., zeolites Na-A and Na-P) Water softener Animal feed Radioactive-waste remediation (e.g., Cs, Sr, with clinoptilolite) Molecular sieving using dehydrated zeolites Air separation (N 2 from O 2 with Li-LSX) Drying agents (e.g., double glazing, a/c) Sulfur removal from natural gas Separation of HFCs (CFC substitutes) Catalysis with dehydrated zeolites Catalytic cracking (gasoline production, zeolite Y-derivatives) Xylene isomerization (for polyester, H-ZSM-5) Butane isomerization (H-FER) Methanol to gasoline (H-ZSM-5) Phenol to hydroquinone (Titanosilicates) Denox reactions (Cu-ZSM-5 and Co-FER) Zeolites have also been utilized for preparing inorganic membranes. The preparation techniques can be divided into two broad classes: (i) in situ one-step hydrothermal synthesis, in which a thin zeolite microporous film is created on an underlying support (e.g., alumina or PSS) by placing it in direct contact with an alkaline zeolite precursor solution, whereby crystals are nucleated and grow on the support, and (ii) secondary (seeded) growth in which crystal seeds are placed, as the first step, on the support, followed by hydrothermal growth of the films on the seeds (Auerbach, Carrado and Dutta, 2003). The type of the Si and Al sources, the Si/Al ratio in the synthesis solution, its 14 alkalinity and water content, the type of other inorganic cations present, and the organic template (structure-directing agent) utilized all affect the final zeolite properties (Cejka, 2007), as does the crystallization temperature and time, as well as the aging period. Using microwave heating, instead of conventional heating, reduces the preparation time from days or hours to minutes or seconds. This is a promising development in terms of the scale-up of production of the materials, though the effect of temperature gradients in the growth solution on the film uniformity remains a concern and must be further investigated (Cundy, 1998; Li and Yang, 2008; Tompsett, Conner and Yngvesson, 2006). If a structure-directing template is employed during membrane preparation, calcination at temperatures above 400 o C with slow heating and cooling rates is necessary, in order to remove the template and not to crack the membrane (Yampol'skii, Pinnau and Freeman, 2006). Organic-template free approaches have also been investigated, in order to reduce the chance of defect formation, which is often the case for template removal by calcination (Hedlund, Noack, Kolsch, Creaser, Caro and Sterte, 1999; Kanezashi, O'Brien and Lin, 2006). Defects (non-zeolite) created during zeolite membrane preparation have a detrimental effect on the membrane performance, which is also generally true for all microporous and dense inorganic membranes. To put this adverse effect into perspective, assuming defects with pore sizes of 1 µm 2 , a defect density of 1 in 10,000 means that the flux through the defects would equal the total flux through the MFI zeolite pores (Noack, Kolsch, Schafer, Toussaint and Caro, 2002). In order to get rid of defects in zeolite membranes, different 15 post-treatments have been suggested, including carbonization (Yan, Davis and Gavalas, 1997a), base (Navajas, Mallada, Tellez, Coronas, Menendez and Santamaria, 2006), and acid (Li, Su and Lin, 2007) treatment, and plugging the defects utilizing a surfactant- templated silica (Xomeritakis, Lai and Tsapatsis, 2001). However, these treatments either result in a reduction in the permeance (Yan et al., 1997a), or have an adverse effect on the separation characteristics of the membranes (Xomeritakis et al., 2001). Because of the crystalline structure of zeolites, it has been shown advantageous to prepare membrane with a specific crystal orientation. For example, an MFI silicalite-1 membrane in a b- orientation was fabricated which exhibited p-xylene/o-xylene separation factors of about 500, compared with the membranes with a c-orientation, which possess separation factors of about 2 (Lai, Bonilla, Diaz, Nery, Sujaoti, Amat, Kokkoli, Terasaki, Thompson, Tsapatsis and Vlachos, 2003). Thus, for a typical zeolite, changing crystal orientation provides an additional means of attaining higher performance, and for overcoming the traditional trade-off between the selectivity and permeance (Snyder and Tsapatsis, 2007). The separation mechanism in zeolites is the result of the interplay between mixture adsorption and the mixture diffusivity, not unlike the solution-diffusion mechanism in polymeric membranes (Caro, Noack, Kolsch and Schafer, 2000). For example, in an MFI zeolite membrane the increase in the separation factor of H 2 /i-butane from about 1.5 (at room temperature) to about 70 (at 500 o C) was explained by the reduction of i-butane adsorption at the higher temperature (Illgena, Schäfera, Noacka, Kölscha, Kühnleb and Caro, 2001). Compared to other types of inorganic membranes, which specifically allow 16 Figure 1-3 The membrane module for the first large scale PV plant installed in Kariya, Japan, by Mitsui Engineering and Shipbuilding Co., Ltd. (Morigami, Kondo, Abe, Kita and Okamoto, 2001). Copyright Elsevier. Reproduced with permission. the transport of one type of molecule (e.g., hydrogen in Pd, hydrogen or oxygen in perovskites), zeolites can separate a wide range of molecules based on molecular shapes, sizes, and the type of surface interaction. Hydrophilic zeolites (zeolite A, X, or Y), for example, can be used in pervaporation for dewatering of alcohols, and for “breaking” various azeotropes. On the other hand, hydrophobic membranes (e.g., MFI silicalite-1) have the potential for alcohol removal from fermentations broths (Weyd, Richter, Puhlfurss, Voigt, Hamel and Seidel- Morgenstern, 2008; Yampol'skii et al., 2006). There have also been efforts to exploit and combine the properties of various zeolites in, for example, a bi-layer structure in which one layer carries out the separation and the other layer the reaction (de la Iglesia, Irusta, Mallada, Menendez, Coronas and Santamaria, 2006; Kiyozumi, Nagase, Hasegawa and Mizukami, 2008; Sterte, Hedlund, Creaser, Ohrman, Zheng, Lassinantti, Li and Jareman, 2001; Zhang, Liu and Yeung, 2005). For instance, ZSM-5 was sandwiched in between two inert silicalite-1 layers, in 17 order to prevent the acidic sites of ZSM-5 to be exposed to nonselective reactions in toluene disproportionation (Sato, Kumagai and Itoh, 2008) in a membrane reactor. Hydrophilic LTA (Na-A) membranes were commercialized for alcohol dehydration and solvent dewatering in the late 1990s. The first large-scale commercial PV plant (Morigami et al., 2001) with a solvent producing capacity of 530 l/h was installed in Japan by Mitsui Engineering and Shipbuilding Co., Ltd. as shown in Fig. 1-3. Using such membranes, Mitsui-BNRI recently installed a plant in India at a steam permeation capacity of 30,000 l/d. The feed was a 93 wt% bio-ethanol mixture, and the product had an ethanol purity of 99.8 wt%. The plant has been in operation since January 2004 (Caro and Noack, 2008). Mitsui Engineering & Shipbuilding Co., Ltd. has announced that they will be building a similar plant in Finland with a pure alcohol production capacity of 130,000 l/d. The plant will have 8 membrane units, each with a 50 m 2 area, composed of 1,700 zeolite membrane tubes with 12 mm (outside diameter), 9 mm (inside diameter), and 800 mm (length) . The success of this process comes from the strong hydrophilic nature of the LTA, and not from any molecular- sieving transport mechanism, which is highly sensitive to defects, as noted previously. 2.3. Dense Oxygen Transport Membranes These materials are of great interest for applications in solid oxide fuel cells (SOFCs), in oxygen separation, partial oxidation MR, and as oxygen sensors. These membranes, when exposed to an oxygen chemical potential gradient, transport only oxygen (in the 18 form of anions) at high temperatures, typically around 500-1000 o C. The material requirements for membranes for use in fuel cells are different from what is required for the other applications. For fuel cell applications, the electronic conductivity of the solid electrolyte membranes should be minimized, in order to improve efficiency, as opposed to oxygen separation membranes for the other applications, where having both electronic and ionic conductivities (mixed conductivity) makes the system design simpler by eliminating the need for electrodes and external circuits. The oxygen flux through a mixed ionic-electronic membrane conductor can be expressed by the Wagner equation: 2 2 2 2 ln 2 2 ln ln (4 ) ( ) po ie o po ie RT jdpo FL σσ σσ ′ ′′ = + ∫ where 2 o j (mol/cm 2 s) is the oxygen flux; i σ (S/cm) the ionic, and e σ (S/cm) the electronic conductivities, 2 po ′ and 2 2 po ′ ′ (atm) are the oxygen partial pressures on the oxygen-rich and oxygen-lean sides of membrane, respectively, L (cm) the membrane thickness, F (96485 C/mol) the Faraday constant, R (8.314 m 3 Pa/mol K) the universal gas constant, and T (K) the temperature (Oleynikov and Ketsko, 2004). For commercial oxygen separation membranes, a minimum ionic conductivity of 1 S/cm (preferably 2-4 S/cm), and electronic conductivity of 5 S/cm (preferably 200-500 S/cm) is desirable (Sammells and Mundschau, 2006). Siemens Westinghouse has installed and successfully tested for over 20000 h a fuel cell unit in a tubular configuration (tubes with a length of 150 cm) with a capacity of 100 kW, and an electrical efficiency of 46%; units of 250 kW and 1 MW are currently under development. Combining the SOFC and gas turbine 19 technology, the same company has built a “proof-of-concept” demonstration unit with a 220 kW capacity (200 kW coming from the SOFC and 20 kW from the turbine) with an electrical efficiency of ~53%, which is impressive for a large-scale system. The unit was successfully tested for 3400 h (http://www.powergeneration.siemens.com/products- solutions-services/products-packages/fuel-cells/demonstrations/). Another successful commercial application of oxygen conductive membrane materials is the use of yttrium- stabilized zirconia (YSZ) solid electrolyte membranes in potentiometric oxygen sensors in automobiles for controlling the air/fuel ratio (Rajabbeigi, Elyassi, Khodadadi, Mohajerzadeh and Sahimi, 2004). These ionic and electronic conductive materials can be divided into three different classes: 2.3.1. Perovskites These materials have the general structural formula ABO 3 , in which A is a rare-earth, alkaline-earth, or alkali metal ion, and B is a transition-group metal ion (Tejuca and Fierro, 1993). By introducing defects (through doping) into the structure of the perovskites, their ionic and electronic conductivities, and their stability can be manipulated. The ability and flexibility to do so, have triggered extensive research on these materials. Their mixed conductivity, which results from oxygen vacancies, and their simultaneous electronic conductivity make them good candidates for oxygen separation membranes. This, as previously noted, eliminates the need for external circuits and electrodes, leading to simplified and economical oxygen separation units. Perovskites, such as SrCoO 3 , SrFeO 3 , LaCoO 3 , LaFeO 3 , LaGaO 3 , along with compounds that result 20 from their single- or double-site substitutions, have been studied extensively (Liu, Tan and Li, 2006b). The goal of the studies was to find the stable perovskites with both high ionic and electronic conductivity at medium temperatures. An oxygen flux of about 0.02- 0.03 mol/m 2 .s is the desirable target for most technological applications (Liu et al., 2006b). A class of perovskites, called mixed perovskites, with the general formula of A 2 B ′ 1+x B ″ 1 −x O 6 − δ ( δ=x) and A 3 B ′ 1+x B ″ 2 −x O 9 − δ ( δ=3x/2) are also attracting recent attention. In these materials, oxygen ion vacancies can be introduced into the structure by changing the stoichiometry, i.e., without the need for any additional doping (Liang, Du and Nowick, 1994). In parallel with the use of mixed conductors, there have been some efforts (Steele, 1998) to achieve high oxygen conductivities by creating a composite structure (called cermet) that combines a metal, such as Pt, Ag, etc., to provide the electronic conductivity, with a conventional ceramic electrolyte membrane (e.g., YSZ). However, the cost of the noble metals and the mechanical stability of the structure due to the thermal mismatch of its components remain key issues. Other researchers are investigating the use of dual-oxide systems such as YSZ-LaMnO 3 or CGO-LSCF (La x Sr 1-x Co y Fe 1-y O 3- δ )(Steele, 1998). 2.3.2. Fluorite-type oxides These solid oxide electrolyte materials (ZrO 2 , CeO 2 , δ-Bi 2 O 3 , ThO 2 , HfO 2 ) exhibit, either intrinsically or through the dissolution of a lower valence metal oxide into their structure, a high oxygen-vacancy concentration (Mallada, 2008). For instance, YSZ (Zr 1-x Y x O 2-x/2 ) 21 was reported to have a maximum ionic conductivity of 0.16 S/cm at 1000 o C, whereas Zr 1-x Sc x O 2-x/2 was reported to have a higher ionic conductivity of 0.38 S/cm at the same temperature. These high ionic conductivities were reported at x=0.08–0.11 and 0.09– 0.11, respectively, for the two materials (Yamamoto, 2000). ZrO 2 -based solid oxide materials, due to their high stability, have been studied extensively. However, they only exhibit high ionic conductivity at relatively high temperatures (~1000 o C), thus, economically, they are still not very attractive. CeO 2 -based solid oxide electrolytes, doped either with Y (YDC), Gd (GDC), or Sm (SDC), exhibit higher ionic conductivity than that of YSZ and, therefore, can operate at lower temperatures (Elyassi, Rajabbeigi, Khodadadi, Mohajerzadeh and Sahimi, 2004; Fergus, 2006). However, they have inferior chemical stability, and at low oxygen partial pressures also exhibit electronic conductivity (due to reduction of Ce +4 to Ce +3 ), which makes them unsuitable for fuel cell applications (Fergus, 2006; Wincewicz and Cooper, 2005). Of all the fluorite-type materials, δ-Bi 2 O 3 has the highest oxygen ion conductivity (~2.3 S/cm at 800 o C). Its use is complicated, however, due to its instability in reducing environments, volatilization of the bismuth oxide at moderate temperatures, low mechanical strength, and a high corrosion activity (Kharton, Marques and Atkinson, 2004). Doping with rare-earth dopants (Y, Dy, Er), or combining with higher valence cations (W, Nb) seems to alleviate some of such problems (Kharton et al., 2004). 22 2.3.3. Other Non-Perovskite Structures Bismuth vanadate (Bi 4 V 2 O 11 ) has several ionic conductive polymorphs, such as α (stable at room temperature) which reversibly changes to the β phase at 447 o C, and then reversibly to the γ phase at 567 o C, and then to γ΄ at 877 o C, which finally becomes liquid at 887 o C (Abraham, Debreuillegresse, Mairesse and Nowogrocki, 1988). Among them, γ-Bi 4 V 2 O 11 exhibits the highest conductivity, about 0.2 S/cm (Vannier, Pernot, Anne, Isnard, Nowogrocki and Mairesse, 2003). Partial substitution of vanadium with a transition metal cation stabilizes the structure and creates a family of oxides with the general formula of Bi 4 V 2-2x Me 2x O 11- γ , also known as BIMEVOX, which have high ionic conductivity at low temperatures (400-700 o C) (Emel'yanova, Shafigina, Buyanova, Zhukovskii, Zainullina and Petrova, 2006). For example, above 400 o C, the ionic conductivity of BIMEVOX is two order of magnitude higher than that of YSZ (Joshi, Nimat and Pawar, 2009). There is also continuing research on a family of compounds called LAMOX, created from La 2 Mo 2 O 9 . This material undergoes an undesirable phase transition from α- to β-La 2 Mo 2 O 9 at 580 o C; however, it can be stabilized by doping (Gao, He and Shen, 2008). The β-phase possesses an almost two order of magnitude higher ionic conductivity of 0.06 S/cm at 800 o C. Apatite (A 10-x (MO 4 ) 6 O 2± δ , M=Si or Ge, and A rare earth and alkaline earth cations) is also looked upon as a low-cost and high oxygen ion conductivity material (Kharton et al., 2004). Rare-earth nickelates with K 2 NiF 4 -type structure (ABO 4+ δ ) -layers of ABO 3 separated with AO layers- which can accommodate a significant oxygen excess leading to high oxygen permeability and ionic conductivity, have received recent attention. For instance, Ln 2 NiO 4+ δ (Ln=La, Pr, Nd) have been 23 studied as SOFCs cathodes or for application in membrane-based oxygen separation (Kharton, Tsipis, Naumovich, Thursfield, Patrakeev, Kolotygin, Waerenborgh and Metcalfe, 2008). Pyrochlore (A 2 B 2 O 7 ) and brownmillerite-type (A 2 B 2 O 5 ) compounds (Liu et al., 2006b; Mallada, 2008) are also receiving attention. Eltron R&D together with Air Products & Chemicals and various other partners, along with the support of the US Department of Energy (DOE), are working to commercialize a brownmillerite-type (La 2 −x Sr x Ga 2 −y Fe y O 5+ δ ,) material for use in a membrane reactor for the synthesis of syngas via partial oxidation of methane (Bouwmeester, 2003; Julbe, Farrusseng and Guizard, 2005). 2.4. Proton-Conductive Membranes A number of perovskites with the general structural formula of AB 1-x M x O 3- δ (A=Ca, Sr, Ba; B=Ce, Zr, Ti, Tb, Tl; M=transition or group III elements) have high mixed protonic- electronic conductivity at high temperatures. They have been of great interest for applications in fuel cells, sensors, hydrogen pumps, and steam electrolyzers since their discovery in 1981 (Islam, Davies and Gale, 2001; Tao and Irvine, 2002), and also have potential as membranes for use in membrane reactors involving hydrogen production. For hydrogen transport at acceptable levels, the electron conductivity should also be high enough to eliminate the need for electrodes or an external circuit, as previously noted. In such materials, the hydrogen transport mechanism is thought to be dominantly controlled by transfer of protons between adjacent OH ־ and O ־ 2 and OH ־ reorientation (Grotthuss- type transport), rather than by OH ־ diffusion (vehicle proton transport) (Kreuer, 1996). 24 SrCeO 3 , and to a lesser extent BaCeO 3 , have been the state-of-the-art of mixed proton- electronic membrane conductors, used for hydrogen transport (Norby, 2007). Their basic character facilitates the formation of protonic charge carriers and, as a result, they both have high protonic conductivity. However, their basic composition also makes them prone to react with acidic or amphoteric gases, such as SO 2 /SO 3 , CO 2 , H 2 S, and H 2 O and to form sulfates, sulfides, carbonates, and hydroxides. Thus, using them in such atmospheres becomes impractical. Additionally, under reducing conditions (i.e., in a H 2 atmosphere) perovskite decomposition or loss of mechanical stability has also been reported (Sammells and Mundschau, 2006). Zirconate-based oxides (SrZrO 3 , CaZrO 3 ) are more stable but, on the other hand, they also have a lower protonic conductivity (Hibino, Mizutani, Yajima and Iwahara, 1992). There are also reports of non-perovskite-type of materials that possess mixed protonic- electronic conductivity. They include, for example, oxides such as Ca-doped Tb 2 O 3 , LaNbO 4 , and Ln 6 WO 12 (Ln=La, Nd, Gd, Er) with the maximum proton conductivity of about 5 · 10 − 3 S/cm at around 850 o C (Haugsrud, 2007; Haugsrud, Larring and Norby, 2005; Haugsrud and Norby, 2006). Fabrication of cermets of perovskites with metals, either to enhance the electronic conductivity or to increase the hydrogen transport, if the metal is hydrogen permeable (Pd, Pd-Ag, Pd-Cu, Nb, Ta, Zr, etc.), is also an area of active research (Balachandran, Lee, Chen, Song, Picciolo and Dorris, 2006), with the challenges here being very similar for the cermets used for oxygen separation. 25 At the lab-scale, dense inorganic membranes for hydrogen and oxygen separation are prepared by pressing powders and through solid-state sintering. However, in order to prepare high-flux, thin film asymmetric membranes for large-scale commercial applications, other techniques, such as sol-gel, spin-coating, gel-casting, phase inversion (hollow fibers), various combustion routes, and most recently a complexation/polymerization route derived from the well-known Pechini’s method, have been utilized (Liu, Tan, Pang, Diniz da Costa, Lu and Liu, 2008; Liu, Wang, Li and Liang, 2007). The major challenge for the commercial use of dense inorganic membranes, either for oxygen- or hydrogen-type applications, remains their stability (i.e., kinetic demixing or decomposition, mechanical stability, etc.). Current research, which focuses on improving material stability and further increasing their conductivities, will hopefully resolve these issues and make the widespread use of these materials a commercial reality. 2.5. Carbon Membranes Microporous carbon membranes are prepared through the carbonization of various polymeric precursors. Precursors utilized for the preparation of such membranes include, among others, polyfurfuryl alcohol (Sedigh, Onstot, Xu, Peng, Tsotsis and Sahimi, 1998), various polyimides (Pei, Chung, Kawi and Guiver, 2004; Su and Lua, 2007), polyetherimide (Sedigh, Jahangiri, Liu, Sahimi and Tsotsis, 2000), cellulose acetate, and various polyaromatic resins (Katsaros, Steriotis, Stubos, Mitropoulos, Kanellopoulos and Tennison, 1997; Morooka, Kusakabe and Kusuki, 2000; Saufi and Ismail, 2004). Carbon membranes are prepared both as supported (flat disks and tubes) and self-supported 26 (hollow fibers and capillaries) structures. For the preparation of supported membranes flat or tubular macroporous supports (alumina and PSS) are first coated by the polymeric precursor (typically by dip-coating, but also by other techniques, such as ultrasound deposition, spray- and spin-coating, and vapor deposition (Ockwig and Nenoff, 2007)), and the composite structure is then pyrolyzed at the appropriate temperature (~500-1000 o C) in inert atmospheres or under vacuum conditions. Unsupported membranes are prepared by the pyrolysis of polymeric hollow fibers and capillaries (Jones and Koros, 1994), typically prepared by phase-inversion techniques. During pyrolysis, small gas molecules (Sedigh et al., 2000) are released and channel their way out, critically contributing in the process to the formation of the membrane structure. Pyrolysis conditions, such as the soaking time, the ramp rate, the pyrolysis atmosphere and temperature, play a critical role in determining the final membrane properties. Generally, higher pyrolysis temperatures and slower heating rates increase the separation factor, but decrease the permeability of the membranes (Park, Lee and Lee, 2005). Sometimes, in order to enhance the separation performance, post treatments (i.e., post-oxidation, CVD, post-pyrolysis, or coating) are applied after pyrolysis (Saufi and Ismail, 2004). Generally speaking, carbon membranes have a hydrophobic nature, and readily chemisorb oxygen to form various oxygen containing surface complexes. Such surface chemistry promotes water (or of various contaminants) adsorption, which adversely affects the membranes’ performance at lower temperatures (for high-temperature applications water or contaminant adsorption is not a serious problem) (Jones and Koros, 27 1995b). Methods that have been proposed to alleviate this problem include coating the membrane with a highly hydrophobic polymer layer (Jones and Koros, 1995a), or passivating the surface by heating the membrane at high temperatures in hydrogen or inert atmospheres to partially remove or modify the surface oxygen complexes (Lagorsse, Magalhaes and Mendes, 2008). Carbon membranes are sensitive to steam and oxygen at high temperatures (>350 o C). Therefore, they should not be utilized under such conditions. Due to selective surface adsorption/diffusion and molecular sieving, carbon membranes can be used for gas separation in a wide range of applications. Media and Process Technology, Inc., of Pittsburgh, PA is currently commercializing supported carbon molecular-sieve membranes for upgrading of various refinery streams and for membrane reactor applications. Their membranes are currently undergoing field-testing at a number of major refinery sites. To further enhance the performance of microporous carbon membranes towards specific applications, a number of groups have incorporated into their structures various inorganic filers, including zeolites (Liu, Wang, Liang, Zhang, Liu, Cao and Qiu, 2006a; Tin, Chung, Jiang and Kulprathipanja, 2005; Titiloye and Islam, 2008; Zhou, Yang, Zhang, Chang, Sun and Wang, 2007), and various metallic (Barsema, van der Vegt, Koops and Wessling, 2005; Suda, Yoda, Hasegawa, Tsuji, Otake and Haraya, 2006c) and oxide nanoparticles (Park et al., 2005). An interesting recent development is the preparation of membranes based on carbon nanotubes. Realization of this concept was made by incorporating parallel carbon 28 nanotubes in supporting dense flat or tubular structures (Hinds, Chopra, Rantell, Andrews, Gavalas and Bachas, 2004; Holt, Park, Wang, Stadermann, Artyukhin, Grigoropoulos, Noy and Bakajin, 2006; Mi, Lin and Li, 2007). Molecular dynamics simulation of transport through carbon nanotubes predicts very large fluxes and better separation (Chen and Sholl, 2006). The carbon nanotube composite membranes were shown to have very interesting properties (Hinds et al., 2004; Holt et al., 2006; Mi et al., 2007; Pol, Pol, Gedanken, Lim, Zhong and Lin, 2006). However, their fabrication process is still rather costly, and future use will likely be limited to unique applications. 2.6. Silica Membranes Among common single-oxide materials, silica is the only one that has found extensive use in the preparation of microporous membranes. The problem with other oxide materials (e.g., titania, alumina, zirconia), which have only been used to prepare meso- or macroporous membranes, is that they contain electropositive metals, which make it difficult to avoid particle formation, unless their alkoxide precursors, which typically have high hydrolysis and condensation rates, are modified by, for instance, chelating them with slowly-hydrolyzing agents. In addition, the sols of these oxides undergo phase transformation or grain growth at low temperatures, that causes a high level of microstructure coarsening (Brinker, Sehgal, Hietala, Deshpande, Smith, Loy and Ashley, 1994). Silica membranes show high hydrogen separation performance, which makes them promising for application in processes involving hydrogen purification and production (Ohta, Akamatsu, Sugawara, Nakao, Miyoshi and Nakao, 2008). However, 29 silica membranes are known not to be hydrothermally stable (Boffa, Blank and ten Elshof, 2008). This is explained to be the result of the hydrolysis of active surface siloxane groups (Si-O-Si) to produce silanol groups (Si-OH), which gradually condense, hence causing gradual nanopore shrinkage and reduction in hydrogen permeance and membrane separation performance (Lu and Zhao, 2004). A number of efforts have been made to improve the hydrothermal stability of silica membranes. One approach is to make the silica surface hydrophobic (by adding methyl or ethyl groups to the structure) at the cost of reducing the membrane’s performance (de Vos, Maier and Verweij, 1999; Wei, Wang, Nie, Yu, Li, Zou and Li, 2008). Another approach is to prepare carbon-templated silica membranes by introducing non-covalently bonded carbon into the membrane structure. Such membranes have promising hydrothermal stability without compromising the membranes’ separation performance (Duke, da Costa, Do, Gray and Lu, 2006). The other strategy is incorporating Al, Ti, or Zr in the structure of silica. However, as mentioned above, these electropositive metals make the control of the pore size difficult (Brinker et al., 1994). It was also reported that doping silica membranes with Ni and Co improved the hydrothermal stability of such membranes (Tsuru, Morita, Shintani, Yoshioka and Asaeda, 2008). Silica membranes are commonly prepared using either sol-gel or CVD techniques. When using the sol-gel method, membranes can be prepared by either the polymeric or the colloidal routes. In the colloidal route, a layer of nano-sized sol particles is deposited on a 30 support using dip-coating, slip-casting, or spin-casting, followed by heat treatment and calcination, which results in densification and membrane formation. This method is generally suitable for preparing mesoporous membranes for UF applications, or as an intermediate layer in membranes for GS applications (Boissiere, Martines, Larbot and Prouzet, 2005; Sørensen, 1999; Xomeritakis, Braunbarth, Smarsly, Liu, Kohn, Klipowicz and Brinker, 2003). In the polymeric route, an alkoxide (Si(OR) 4 , R being an alkyl, commonly tetramethoxysilane (TMOS) or tetraethoxysilae (TEOS), usually dissolved in their respective alcohols are employed as the starting precursors. The solution is deposited on a porous support, where the alkoxide undergoes hydrolysis and condensation reactions and forms a gel, which is subsequently heated to solidify and form membrane. Using an alkoxide (instead of sols) as the molecular building block results in microporous membranes which are suitable for gas separation. Templates, either those covalently attached to the polymer (i.e., methyl groups in methyltriethoxysilane (MTES)) or non-covalently-bonded templates (i.e., C 6 - or C 16 -surfactants), have been used to produce high quality membranes (Lu and Zhao, 2004). The other preparation technique is CVD, in which silicon-based precursors, such as TEOS, TMOS, SiCl 4 , SiH 4 , are thermally decomposed and reacted with an oxidizing agent, such as oxygen, ozone, and water on a support. The CVD is carried out, either with both precursors being fed from the same side of support, or with the silicon-based precursor, and the oxidizing agent being fed from the opposite sides. In the opposing-side feed mode, layer deposition is controlled by the inability of the silicon-based precursor to 31 diffuse through the SiO 2 layer that is formed (Mallada, 2008). Atomic-layer deposition (ALD), also known as atomic-layer epitaxy (ALE), is another technique used to prepare silica-based membranes. The method is actually a modification of the conventional CVD technique, which is based on conducting the reaction in the form of two half-reaction steps in sequence. The first step proceeds until the surface of support is saturated with a monolayer of one of the reactants. Then, as the second step the other reactant is supplied to complete the reaction and to form the SiO 2 nanolayer (McCool and DeSisto, 2004). Repeating the cycle of steps, results in the formation of a thin membrane film. Compared to the traditional CVD, this method provides better control over thickness, pore size, and uniformity of the deposition (Cameron, Gartland, Smith, Diaz and George, 2000). Plasma-enhanced CVD (PECVD) is also used to prepare silica membranes. Compared to the CVD, this method reduces the deposition temperature while keeping a high deposition rate, as the result of creating highly reactive species (ions, free radicals and metastable components) under a plasma atmosphere (Kim, Scibioh, Kwak, Oh and Ha, 2004). 2.7. Other Inorganic Membranes Glass membranes, which are produced by phase separation and then by leaching in basic solutions, are available commercially under the trade-name of Vycor ® (for instance, tubular or flat Vycor ® 7930, Corning). Depending on the conditions of preparation, such membranes have pore sizes in the range 1-300 nm. (A few years back microporous hollow fiber glass membranes were made available by PPG (Bhandarkar, Shelekhin, 32 Dixon and Ma, 1992). However, due to their fragility, they failed to attract any commercial applications). Such mesoporous glass membranes find use as supports for the preparation of microporous membranes. Another strategy is to try to fine-tune the properties of the membranes for gas separation, either by reducing the pore size or changing the interaction between the penetrant and the membrane surface through functionalization (Singh, Way and Dec, 2005). Another area of substantial current interest is the preparation of high-temperature, CO 2 - selective membranes. Membranes prepared by layered double-hydroxide clay materials have attracted the lion’s share of attention, since these materials are highly CO 2 permselective; though progress is being made, preparing high-quality, defect-free microporous membranes still remains an elusive target (Kim, Sahimi and Tsotsis, 2008). 3. Characterization Techniques In order to understand and model the relationship between a membrane’s performance and its chemical or structural properties, various characterization techniques have been used. Table 1-2 presents some of the techniques, which have been utilized for characterizing the morphology, pore structure, surface chemistry, and the overall performance of inorganic membranes. 33 Table 1-2 Membrane characterization techniques. Technique Purpose Permeation test Permeance, separation factor, tortuosity AFM (Atomic Force Microscopy) Surface topography SEM (Scanning Electron Microscopy) Surface morphology, pore size, membrane thickness EDX (Energy Dispersive X-ray Analysis) Elemental analysis TEM (Transmission Electron Microscopy) Structural and morphology analysis AES (Auger Electron Spectroscopy) Depth elemental profiling EPMA (Electron Probe Microanalysis) Elemental analysis XRF (X-ray Fluorescence) Elemental analysis XRD (X-ray Diffraction) Structural analysis Bubble-point, liquid displacement, thermoporometry, permporometry Pore size distribution SE (Spectroscopic Ellipsometry) Pore size distribution, porosity, thickness (for thin films) PALS (Positron Annihilation Lifetime Spectroscopy) Pore size distribution, porosity, interconnectivity, and depth-profiling Nitrogen adsorption Surface area, micro- and mesopore volumes, pore size distribution SAXS (Small-Angle X-ray Scattering) Pore size distribution and porosity (in thin films) Mercury porosimetry Surface area, porosity, pore size distribution Archimedes method Porosity Three- (or four-) point bend test Flexural strength 4. Module Design Modules with high surface area/volume (packing density) result in considerable reduction in cost and process foot-print. Typical surface area/volume numbers for different polymeric membrane modules are given in Table 1-3. Inorganic membranes are fabricated in tubular, flat (in lab-scale) and multi-channel forms, and to a much lesser extent hollow-fiber forms, but not in spiral-wound configuration. 34 Table 1-3 Typical surface area/volume for different membrane modules (Burggraaf and Cot, 1996; Nunes and Peinemann, 2006). Module type Surface area /volume (m 2 /m 3 ) Hollow fine fiber (fiber diameters ~100-250 µm) 10000-2000 Capillary fiber (fiber diameters ~500-2000 µm) 1500-3000 Spiral wound 500-1000 Tubular, plate and frame 150-300 Multi-channel monoliths 130-400 Honeycomb multi-channel monoliths up to 800 As mentioned previously, being rather more expensive and difficult to make, most of the immediate applications envisaged for microporous or dense inorganic membranes are for high-temperature separations, and particularly for integrating them in membrane reactors at high temperatures and pressures, where polymeric membrane, due to the stability issues, cannot be used. However, one issue for using inorganic membranes at such conditions is providing reliable sealing for the membranes in the membrane module housing. The use of the PSS substrates, wherever possible, is the most preferred approach as it allows for the use of conventional sealing methods in the membrane module housing (or reactor body). When this is not possible, one approach is sealing the membranes outside the high-temperature zone with conventional polymeric seals, and cooling the seals with water for additional protection. Another proposed solution is to employ well- established brazing technique for joining the ceramic support tubes to metal end-tubes. One way of further reducing the high thermal stress during brazing is by incorporating the membrane ceramic into the brazing alloy (McLeary, Jansen and Kapteijn, 2006). Another proposed strategy is making the whole membrane module from ceramic 35 materials. However, this may necessitate membrane deposition after module fabrication, which makes the process rather complex (Caro, Noack and Kolsch, 2005). The challenges associated with sealing represent a hurdle to be overcome prior to the broad-scale commercialization of microporous and dense inorganic membranes. Long- term field-tests are also needed to convince the industry to begin trusting the technology. 5. Applications Microporous and dense inorganic membranes have good potential for application in a wide range of processes. They are already making inroads, for example, in separation by pervaporation which is a promising alternative to distillation for heat-sensitive, closely boiling, or azeotropic mixtures (Yampol'skii et al., 2006). Hydrogen separation or hydrogen ratio adjustment in refineries, oxygen production (instead of using cryogenic distillation), and separating carbon dioxide from natural gas are also the other potential applications. Membrane reactors represent a particular promising area for the large-scale application of microporous and dense inorganic membranes. Such reactors have been tested successfully in the laboratory (and in a limited number of field tests) for equilibrium-limited reactions, such as alkane dehydrogenation, steam-reforming of methane, ethane, propane, and ethanol, the water gas-shift reaction, the oxidative coupling of methane, and the partial oxidation of methane (Lu, Dixon, Moser, Ma and Balachandran, 2000; Mallada, 2008; Sanchez Marcano and Tsotsis, 2002). They can also enhance the selectivity and yield of reactions through the uniform distribution of reactants, or by preventing undesirable side-reactions. For instance, in the Fischer- 36 Tropsch reaction, the produced water adversely affects the catalyst activity and reaction rate. Extracting water from the reaction (e.g., through a zeolite microporous membranes) can enhance the process (Rohde, Schaub, Khajavi, Jansen and Kapteijn, 2008). 6. Concluding Remarks Microporous and dense inorganic membranes are very promising for separations in harsh environments (e.g., at high temperatures and pressures, and in the presence of steam and various poisonous components), or for applications in membrane reactors. Key challenges that remain include increasing the reproducibility in the preparation of the membranes, reducing the cost of production, enhancing membrane performance and stability, running long-term stability field-tests, and addressing scaling-up and sealing issues during the preparation of membrane modules. As presented in this chapter a wide range of inorganic materials have been studied for membrane preparation. However, from a material standpoint there is still plenty of work that is needed to be carried out. One approach is to improve the performance of current widely studied materials (silica, palladium, perovskites, zeolites, carbon). In parallel, there is a need to find and investigate new materials to enhance membranes performance and specially to solve the stability issue in realistic conditions. Among different materials we screened, we found silicon carbide due to its many unique properties such as high thermochemical stability, high thermal conductivity, low thermal expansion coefficient, high resistance to abrasion, and high thermal shock resistance to be a potential candidate for such applications. The goal of this thesis is to study the performance and stability of silicon carbide membrane 37 in relatively harsh conditions (e.g., high temperature and presence of steam) close to conditions where low temperature water gas shift reaction is conducted or simply for hydrogen separation from gaseous mixtures in such conditions. Chapter 2 entails our preparation technique for SiC membranes and their long-term hydrothermal stability test. SiC membranes exhibited proming hydrothermal stability within the experiment conditions close to conditions where low temperature water gas shift reaction is typically conducted. In chapter 3, we propose a novel polystyrene sacrificial interlayer-based technique for enhancing the performance of SiC membranes. This promising technique can be seen as a general preparation method for fabricating other type of inorganic membranes. In chapter 4, we describe how we fabricated SiC nanofibers and studied the effect of polystyrene on their morphology and physical properties. In chaper 5, we incorporated such SiC nanofibers in the structure of membranes and investigated their performance. Also, performance of such membranes is compared with those membranes previously prepared using SiC powders as filler. 38 Chapter 2 : Preparation of Silicon Carbide Membranes and Their Steam Stability 1 1. Introduction Growing interest in the hydrogen economy is motivating research on inorganic, hydrogen-permselective membranes, which can be used in processes (related to H 2 production) that take place at high temperatures and pressures. A promising candidate material for a variety of inorganic membrane applications is SiC due to its many unique properties, such as high thermal conductivity (Takeda, Nakamura, Maeda and Matsushita, 1987), thermal shock resistance (Schulz and Durst, 1994), biocompatibility (Rosenbloom, Sipe, Shishkin, Ke, Devaty and Choyke, 2004), resistance in acidic and alkali environments (Li, Kusakabe and Morooka, 1996), chemical inertness, and high mechanical strength (Kenawy and Nour, 2005; Zorman, Fleischman, Dewa, Mehregany, Jacob, Nishino and Pirouz, 1995). There are relatively few reports discussing the preparation of SiC membranes. Two different approaches have been utilized. One involves the use of chemical-vapor deposition (CVD)/chemical-vapor infiltration (CVI) techniques, while the other approach is based on the pyrolysis of polymeric precursors. 1 B. Elyassi, M. Sahimi, T.T. Tsotsis, “Silicon carbide membranes for gas separation applications”, Journal of Membrane Science, 288 (2007) 290–297. Copyright Elsevier. 39 Takeda et al. (Takeda, Shibata and Kubo, 2001) prepared SiC membranes by CVI, using SiH 2 Cl 2 and C 2 H 2 diluted with H 2 , and γ-Al 2 O 3 tubes as supports. Their membrane had a H 2 permeance of 1 × 10 −8 mol m −2 s −1 Pa −1 at 350 o C, with an ideal H 2 /N 2 selectivity of 3.36, lower than the corresponding Knudsen value. Sea et al. (Sea, Ando, Kusakabe and Morooka, 1998) prepared a SiC membrane using an α-alumina tube as support by CVD of (C 3 H 7 ) 3 SiH (tri-isopropylsilane or TPS) at 700–750 o C. After CVD, the membrane was calcined at 1000 o C in Ar, then tested with pure gas permeation, and was shown to follow a Knudsen mechanism, with H 2 permeance in the range (4–6) × 10 −7 mol m −2 s −1 Pa −1 . A permeation test was also carried out with a H 2 –H 2 O–HBr gas mixture (molar ratio of 0.49:0.5:0.01) between 200 and 400 o C. In the beginning of the test, the H 2 /H 2 O separation factor was higher than the Knudsen value (~5). Exposure of the membrane to the mixture at 400 o C resulted in a decline of the H 2 permeance during the first 50 h (by a factor of ~2.5); the permeance stabilized after that, remaining constant for the remainder of the run, which lasted 120 h (Sea et al., 1998). Pages et al. (Pages, Rouessac, Cot, Nabias and Durand, 2001) prepared Si:C:H membranes on alumina supports using plasma enhanced CVD (PECVD) of diethylsilane; the resulting membranes had a permeance in the range of (10 -7 -10 -6 ) mol m −2 s −1 Pa −1 for H 2 , but with an ideal separation factor for H 2 /N 2 in the range (3-4). Our group was the first to report (Ciora, Fayyaz, Liu, Suwanmethanond, Mallada, Sahimi and Tsotsis, 2004) the preparation of microporous SiC membranes using CVD. We utilized CH 3 SiH 2 CH 2 SiH 3 (1,3-disilabutane or DSB) to prepare membranes on γ-Al 2 O 3 40 tubular supports. To prepare the membranes, the substrates were exposed to the DSB vapor at temperatures between 650 and 750 o C. Nanoporous membranes were prepared with a He permeance of 3.5 × 10 −7 mol m −2 s −1 Pa −1 , and a He/N 2 selectivity of ~55 at 550 o C. The DSB membranes were thermally stable, but did not fare well in the presence of high-temperature steam, being unstable. We have also used CVD of TPS at 700-750 o C to prepare SiC membranes. Depending on the preparation conditions, the He permeance ranged from 8.06 × 10 −8 to 1.72 × 10 −6 mol m −2 s −1 Pa −1 with a He/N 2 ideal selectivity ranging from 4 to >100. These membranes are also stable in the presence of high-pressure (1-3 bar) and temperature (<750 o C) steam. However, the preparation procedure involves multiple steps, which makes it costly, and requires a high temperature (1000 o C) treatment, which places a great burden on glass end-seals and membrane housing during the CVD/CVI process. More importantly, the high temperature post- treatment further modifies the nanoporous structure of the membrane prepared at the lower temperature. Therefore, the final product quality is difficult to predict and control; from a membrane manufacturing standpoint, the advantage of the on-line control of the CVD/CVI technique is, therefore, lost. To overcome these hurdles, we have pursued, in tandem with TPS CVD/CVI, a dip- coating technique through the pyrolysis of allyl-hydridopolycarbosilane (AHPCS), a partially allyl-substituted hydridopolycarbosilane (HPCS). The choice of AHPCS is because it is curable in the presence of inert Ar gas, rather than oxygen (as are other PCS pre-ceramic polymers). In fact, curing in oxygen introduces Si-O-C linkages in the resulting PCS-derived ceramics, which have proven thermally and hydrothermally 41 unstable. The use of HPCS was first proposed by Interrante et al. (L.V. Interrante, 1995; L.V. Interrante, 1994). The primary pyrolysis product of both HPCS and AHPCS is a SiC ceramic, having a Si–C bond linkage. Increasing the fractions of allyl groups slightly increases the carbon content of the final ceramic, and also makes the polymer cross- linkable at lower temperatures. The membranes we prepared with conventional dip- coating possessed H 2 permeance in the range (10 -8 -10 -7 ) mol m −2 s −1 Pa −1 , and had an ideal separation factor of about 20 for H 2 /N 2 at 200 o C (Ciora et al., 2004). They proved thermally stable but, however, did not fare well in the presence of high temperature steam, proving highly unstable. Prior to our study, other investigators had reported the pyrolysis of PCS and other polymers for the preparation of SiC membranes. Morooka and coworkers (Kusakabe, Li, Maeda and Morooka, 1995; Li, Kusakabe and Morooka, 1997a; Li et al., 1996), for example, used a PCS that they cured in oxygen to prepare a Si–O–C membranes on γ- and α-Al 2 O 3 tubular substrates with an oxygen content of 13–18 wt%, using polystyrene (PS) as the pore former. A membrane prepared with 1% PS had a H 2 permeance of 4 × 10 −8 mol m −2 s −1 Pa -1 , and an ideal H 2 /N 2 selectivity of 20 at 773 K. The membrane was unstable in Ar at 1223 K, and also hydrothermally unstable when exposed to a 7.8 wt% H 2 O in He mixture at 773 K. Lee and coworkers (Lee and Tsai, 1999; Lee and Tsai, 2001) prepared Si–C–O membranes by the pyrolysis of polydimethylsilane (PMS). The PMS undergoes a thermolytic reaction at 733 K for 14 h under 1 atm of Ar, followed by O 2 curing at 473 K for 1 h. The layer was finally pyrolyzed at various temperatures from 42 523 to 1223 K to produce the membranes. The membrane pyrolyzed at 873 K had the best separation characteristics, exhibiting a He permeance of ~1.4 × 10 −9 mol m −2 s −1 Pa −1 , a H 2 permeance of ~2.7 × 10 −9 mol m −2 s −1 Pa −1 , and an ideal H 2 /N 2 selectivity of 20 at 473 K. Membranes prepared at the higher pyrolysis temperatures were not microporous. Chen et. al. (Chen and Tsai, 2004)prepared PMS-derived SiC membranes in an autoclave under a N 2 atmosphere at low temperatures. Exposure of the membranes to steam increased their permeance, and returned their ideal selectivities to their Knudsen values. Since our study (Ciora et al., 2004), two other groups have prepared membranes by pyrolysis of PCS-type polymers cured in the absence of O 2 . Suda et al. (Suda, Yamauchi, Uchimaru, Fujiwara and Haraya, 2006a) prepared SiC membranes by dip-coating PCS on α-alumina tubes (PS was used as the pore former for some of the membranes). Their membranes exhibited ideal separation factors of (90-150) for H 2 /N 2 , and H 2 permeance in the range of (1-3) × 10 -8 mol m −2 s −1 Pa −1 at 373 K. Nagano et al. (Nagano, Sato, Saitoh and Iwamoto, 2006) prepared SiC membrane by dip-coating of PCS on γ-alumina supports, and reported H 2 permeance ~10 -7 mol m −2 s −1 Pa −1 , and an ideal separation factor of (8-12) for H 2 /N 2 at 873 K. Most recently, Wach et al. (Wach, Sugimoto, Idesaki and Yoshikawa, 2007a; Wach, Sugimoto and Yoshikawa, 2007b) prepared Si-O-C membranes by the pyrolysis of a blend of PCS and polyvinylsilane on porous alumina substrates by radiation curing in the presence of oxygen. Their membrane indicated a H 2 permeance of ~3×10 -9 mol m −2 s −1 Pa −1 and ideal (H 2 /N 2 ) and (He/N 2 ) selectivities of 206 and 241 at 250 o C, respectively (Wach et al., 2007b). 43 Past studies presented only single gas permeation data; and other than the membranes prepared by TPS CVD (Ciora et al., 2004), the rest were shown to be unstable in steam. Membranes prepared by AHPCS pyrolysis (Ciora et al., 2004) showed good separation factors and thermal stability, but also proved unstable in steam. This chapter describes the progress we made to fabricate high performance SiC membranes and investigating their hydrothermal stability in harsh conditions. SiC microporous membranes were prepared by the pyrolysis of thin allyl- hydridopolycarbosilane (AHPCS) films coated, using a combination of slip-casting and dip-coating techniques, on tubular silicon carbide macroporous supports. Combining slip- casting with dip-coating significantly improved the reproducibility in preparing high quality membranes. The membranes prepared, so far, exhibited an ideal H 2 /CO 2 selectivity in the range of (42-96), and a H 2 /CH 4 ideal selectivity in the range of (29-78). Separation factors measured with the same membranes, using equimolar binary mixtures of H 2 in CO 2 and H 2 in CH 4 , were similar to the ideal selectivity values. Steam stability experiments with the membranes lasting 21 days, using an equimolar flowing mixture of He/H 2 O at 200 o C, indicated some initial decline in the permeance of He, after which the permeance became stable at these conditions. The principal goal of the present study, therefore, was to improve membrane hydrothermal stability. Another important goal was to improve the reproducibility of the fabrication method, since one of the factors that have, so far, hampered the broad application of inorganic membranes is their cost (de Vos and Verweij, 1998). In this 44 chapter, we report significant improvements on the fabrication of high performance SiC membranes using a combination of slip-casting and dip-coating techniques. Membranes are prepared that are stable under conditions that are relevant to the WGS environment. The effect of low temperature oxidation on the properties of these membranes is also reported here. 2. Experimental Ultra-high purity gases (He, H 2 , Ar, CO 2 , from Glimore Liquid Air Company, and CH 4 from Specialty Air Technologies, Inc.) were used in the experiments. They were further purified using standard moisture traps. For the CO 2 , in addition, we utilized a Hi-EFF ® organic trap (Alltech). Porous SiC support tubes were prepared using uniaxial cold- pressing of β-SiC powder (HSC059, by Superior Graphite Co., with an average particle size of 0.6 µm as reported by the manufacturer), together with the appropriate sintering aids. For sintering, the green samples were heated at a temperature of 1800 o C, where they were kept in a furnace (Thermal Technology Inc., Model 1000-4560-FP20) for 3 h in flowing He. After sintering, the samples were cooled down to room temperature (cooling rate of 3 o C/min). The tubes, after sintering, had dimensions of 40 mm × 12 mm, with a thickness of 3 mm. Further details about the sintering characteristics of various SiC powders and the preparation of porous supports were presented elsewhere (Suwanmethanond, Goo, Liu, Johnston, Sahimi and Tsotsis, 2000). 45 The support tubes used in the membrane preparation were treated in flowing synthetic air at 450 o C, with the purpose to oxidize any residual carbon that may be introduced during preparation, sonicated several times in acetone, and then dried prior to membrane film deposition. In order to prepare the slip-casting solution, the 0.6 µm SiC powder was mixed with acetone, and the lighter particles floating on the solution top were separated and dried. Scanning electron microscopy (SEM, Philips, XL3) observations of these particles indicated their size was mostly ~(100-200) nm. The dip-coating solution was prepared by dissolving AHPCS (SMP-10, Starfire Systems, Inc.) in hexane (10 wt%). A solution containing 10 wt% of these SiC particles in the dip-coating solution constitutes the slip-casting solution. Slip-casting was used to coat the first layer on the outer surface of the support tubes. To prepare this layer, the tube was placed in the slip-casting solution for 12 s and, then, drawn out of the solution at a speed of 2 mm/s. 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 h, then to 400 o C, where they were also kept for 1 h, and finally to 750 o C, where they were kept for an additional 2 h. Subsequently, they were cooled down to room temperature in flowing Ar with a cooling rate of 3 o C/min. The reason for the relatively slow heating (and holding at 200 o C) is that we have found (as did others (L.V. Interrante, 1994; Suda et al., 2006a)), that using such heating rates, and treatment at lower temperatures result, generally, in better cross-linked amorphous 46 SiC materials (L.V. Interrante, 1994), and membranes with higher hydrogen permeabilities (Suda et al., 2006a). For membrane preparation, additional layers were deposited on the support tubes, with the slip-casted SiC layer on the top, by dip-coating with the 10 wt% solution of AHPCS in hexane using the same procedure (i.e., 12 s of dipping time, drawing rate of 2 mm/s). The membranes, reported here, were dip-coated four times. After each coating, the new layer was pyrolyzed following the same protocol with the slip-casted layer. Layer-by- layer dip-coating and pyrolysis result in membranes with decreased permeance and higher separation factors. Depositing several thin layers, instead of a single thick layer, provides the advantage of facilitating the release of the gases from the pyrolysis process, and also decreases the chance of defect formation due to film shrinkage. After coating the final layer, the membranes were typically treated in air for 2 h at 450 o C with the purpose to oxidize potential residual carbon. This, in general, results in higher permeances and lower separation factors (the effect of low temperature air treatment is discussed further later). Although AHPCS has been reported to result in a SiC ceramic with a near stoichiometric Si:C ratio (L.V. Interrante, 1994), one cannot exclude the possibility that after pyrolysis there may remain trivial amounts of carbon, which may influence membrane performance. Low-temperature air oxidation has been shown effective in removing minute amounts of carbon, for example as a common way for carbon nanotube purification (Li and Zhang, 2005; Osswald, Flahaut and Gogotsi, 2006; 47 Park, Banerjee, Hemraj-Benny and Wong, 2006). By oxidizing in air the SiC membranes, prior to their use, one ensures that no further variation in membrane properties will occur due to accidental exposure to air. Surface and cross-section morphologies of the membranes were characterized by SEM. AHPCS-derived powders, pyrolyzed with the same protocol as the membranes, were analyzed by X-ray diffraction analysis (XRD, Rigaku X-ray diffractometer with CuK α radiation). Permeation experiments were carried out using a Wicke-Kallenbach-type permeation apparatus previously utilized to measure permeation through flat-disk SiC membranes (Ciora et al., 2004). In order to use this apparatus for tubular membranes, one end of the membranes is completely sealed using graphite tape and high temperature non- porous glue (J-B WELD); the other end of the membrane is attached to a flat metal ring using the same glue. The metal ring bearing the membrane is then installed in between the two half-cells of the permeation test-unit using O-rings. Single-gas permeation experiments were carried out by flowing a given gas through the apparatus half-cell facing the membrane dead-end, under constant pressure and temperature, and by measuring the amount of gas that permeates through the membrane to the permeate side which was maintained at atmospheric pressure. In the experiments reported here, the pressure drop across the membranes was kept at 207 kPa (30 psi), and the temperature was varied from room temperature to 200 o C. We report the ideal selectivity for these membranes, which is defined as the ratio of the permeances of the different gases. 48 Mixed-gas permeation tests were also carried out with select membranes using equimolar binary (H 2 /CH 4 , and H 2 /CO 2 ) gas mixtures. For the mixed-gas experiments, in addition to the gas flow to the permeate side of the membrane, the composition was also analyzed using a mass spectrometer. During the experiments, the flow of gas permeating through the membrane was less than 0.5 percent of the gas flowing in the feed-side. Therefore, it was assumed that the concentration in the feed-side remained invariant. Steam stability tests, using an equimolar mixture of He and water, were also conducted with the membranes at 200 o C. Water was delivered by a metering micro-pump into a stainless steel steam vessel (vaporizer) in flowing He. The vaporizer was filled with glass beads for better mixing, and its temperature was kept constant at 200 o C. 3. Results and Discussion 3.1. Membrane Characterization and Performance The quality of the membrane support is known to be important in determining the quality of the membranes that are prepared using these supports. Large defects in the supports are thought to lead to defects in the membranes, resulting in low reproducibility in preparing membranes with high separation factors. The ability to strongly anchor the thin films on the underlying supports is also of key importance. In our prior study (Ciora et al., 2004) we directly dip-coated the membrane films; the resulting membranes had good separation characteristics, but proved to be hydrothermally unstable. In this study we have used slip-casting in order to first condition the surface of supports. Slip-casting 49 results in the formation of a film which strongly adheres onto the underlying support, as manifested by the superior hydrothermal stability of these membranes (see further discussion below), and on which we deposit the subsequent layers by dip-coating. In addition, the success rate in preparing “good” microporous membranes using the porous SiC tubes modified by slip-casting was much higher than when preparing membranes using the original support tubes (a “good” membrane is one with an ideal H 2 /CH 4 separation factor ~ 30 or larger, and with a hydrogen permeance >10 -9 mol m −2 s −1 Pa −1 at 200 o C). The type of powder used during slip-casting has also a significant effect on the ability to prepare appropriate top layers, which in turn can be used for further membrane deposition in the preparation of high performance membranes. For example, using a different β-SiC powder with an average size of 30 nm (provided to us by Sumitomo Osaka Cement Co.) did not result in smooth uniform top layers. Instead, after pyrolysis the particles agglomerated on the support tube surface. The difference in behavior could be attributed to the chemical or physical differences between these two SiC powders, which came from two different manufacturers (Shinoda, Nagano and Wakai, 1999; Shojai, Pettersson, Mantyla and Rosenholm, 2000; Zawrah and Shaw, 2004). In fact, we found that the Sumitomo powder was more difficult to uniformly disperse in solution and, therefore, to use for producing a homogeneous slip-casting solution. The thickness of the slip-casted film also influences membrane performance. Too thick of a film (prepared, for example, 50 20 30 40 50 60 70 80 2 Ө Intensity (111) (200) (311) (220) (222) by increasing the residence time in the slip-coating solution) resulted in membranes with poor performance. In our group previous paper, X-ray photoelectron spectroscopy (XPS) analysis and in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) were utilized to follow the various processes that occur during the pyrolysis of AHPCS to produce SiC (Ciora et al., 2004). We noted that increasing the pyrolysis temperature increases material crystallinity, as indicated both by the XRD and DRIFTS spectra (Ciora et al., 2004). Fig. 2-1, for example, shows the XRD of the AHPCS-derived SiC. The broad peaks show that the SiC is still primarily amorphous. Figure 2-1 XRD of an AHPCS-derived amorphous SiC powder pyrolyzed at 750 o C for 2 h along with Joint Committee on Powder Diffraction Standards (JCPDS) card number 29-1129 for crystalline β-SiC. 51 Fig. 2-2 is the SEM picture of the cross-section and the top surface of one of the membranes. The thickness of the membrane layer is about 2 µm, and sits on top of the SiC support. Energy Dispersive X-ray (EDX) analysis of the membranes revealed the presence of some oxygen in the structure of the membrane. Since the curing and cross- linking of the AHCPS takes place in the presence of Ar, it is unlikely that the oxygen will be introduced during the pyrolysis stage. There are reports, however, indicating the presence of minor amounts of oxygen in the structure of AHPCS-derived SiC (ranging from 3 to 6.7 wt%), and which claimed that the oxygen is introduced into the structure during the polymer preparation or handling stages (L.V. Interrante, 1995; Puerta, Remsen, Bradley, Sherwood and Sneddon, 2003; Zheng, Kramer and Akinc, 2000) (however, the wt% of oxygen in these studies is calculated by difference during elemental analysis, and is not likely to be highly accurate). What is more likely is that the oxygen detected by EDX is the result of membrane exposure to air prior to the analysis; however, elemental analysis of SiC powders before and after low-temperature (450 o C) oxidation in flowing air indicated that the Si:C ratio remains practically invariant, so that any oxygen incorporation, if present, is likely to be limited to the surface region. 52 2 µ m Figure 2-2 SEM pictures showing the cross-section and the top view of a SiC membrane. Before coating the membranes, the He and Ar permeances (and the ideal selectivity) of the SiC supports were measured at room temperature. The permeance of He was of the order ~10 -7 mol m −2 s −1 Pa −1 , corresponding to an ideal separation factor of 2.6 for He/Ar (below the Knudsen value of 3.16). After deposition of the membrane film on the supports, the permeance of the resulting membranes was again measured at room temperature. Membranes, which exhibited separation factors higher than 6 at room temperature, were selected for further testing at higher temperatures. Single-gas (H 2 , CH 4 , and CO 2 ) membrane permeances, and the corresponding ideal separation factors at 200 o C (for a number of membranes that were previously treated in air at 450 o C) are presented in Table 2-1. As can be seen, the membranes show relatively high ideal separation factors, ranging between (42-96) for H 2 over CO 2 , and (29-78) for 53 H 2 over CH 4 . The permeance of H 2 is in the range (10 -9 -10 -8 ) mol m −2 s −1 Pa −1 . As expected, membranes with higher separation factors had, typically, lower permeances. The relatively low permeances of our asymmetric membranes are partly due to the relatively low permeance of the support itself, and also of the thick membrane films that are formed. Efforts are currently under way to improve the permeance of these membranes (e.g., through the aid of pore-formers). Extrapolated to a temperature of 500 o C (see Section 2.2., below), the expected permeance for H 2 is of the order ~10 -7 mol m −2 s −1 Pa −1 . By comparison, Pd and Pd-alloy membranes exhibited, around 500 o C, permeances in the range (10 -7 -10 -6 ) mol m −2 s −1 Pa −1 (Rothenberger, Cugini, Howard, Killmeyer, Ciocco, Morreale, Enick, Bustamante, Mardilovich and Ma, 2004). Table 2-1 Range of single-gas permeances and separation factors, measured at 200 o C, prepared by pyrolysis at 750 o C for 2 h. Single gas H 2 CH 4 CO 2 Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) 0.54-1.18 0.007-0.041 0.006-0.028 S.F. (H 2 /gas) 29-78 42-96 However, Pd membranes are known to be sensitive to the presence of H 2 S and hydrocarbon impurities; use of Pd-alloy membranes (e.g., Pd-Ag) also faces challenges for long-term usage at temperatures significantly higher than 500 o C. 54 Table 2-2 summarizes the single- and mixed-gas experiments with two different membranes using the experimental procedure previously outlined. The mixed-gas H 2 permeances (calculated on the basis of its partial pressure gradient across the membrane) for the H 2 /CH 4 mixture are close to the single gas values, while the CH 4 permeance decreases, resulting in mixed-gas separation factor which is ~15-25% higher. For the H 2 /CO 2 mixtures both permeances are lower than that of the single-gas values. However, the CO 2 permeance decreases more than H 2 permeance, resulting in mixed-gas separation factors that are higher by ~10-20% than their single-gas counterparts. Table 2-2 Single- and mixed-gas permeation properties of two SiC membranes (I, II), measured at 200 o C, prepared by pyrolysis at 750 o C for 2 h. For the mixed gas experiments the membranes were exposed to an equimolar mixture of H 2 /CH 4 or H 2 /CO 2 . Single gas Mixed gas Mixed gas Sample I H 2 CH 4 CO 2 H 2 CH 4 H 2 CO 2 Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) 1.18 0.041 0.028 1.01 0.028 0.89 0.018 S.F. (H 2 /gas) 29 42 36 50 Sample II H 2 CH 4 CO 2 H 2 CH 4 H 2 CO 2 Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) 0.86 0.026 0.019 0.84 0.023 0.79 0.016 S.F. (H 2 /gas) 33 46 37 50 55 y = -2.1238x - 14.246 R 2 = 0.9944 y = -1.4536x - 16.15 R 2 = 0.9992 -20.2 -20 -19.8 -19.6 -19.4 -19.2 -19 -18.8 -18.6 2 2.2 2.4 2.6 2.8 1000/T(K) ln (Permeance) (mol m -2 s -1 Pa -1 ) After Oxidation Before Oxidation 3.2. The Effect of Low Temperature Oxidation on Membrane Performance To understand the effect that low temperature oxidation has on membrane performance, a membrane was prepared without using the final air oxidation step, as previously described in the membrane preparation procedure, and its H 2 , He and Ar permeances were measured (see Table 2-3). Fig. 2-3 shows H 2 permeances measured at different temperatures with the same membrane (solid line). Hydrogen transport is by an activated diffusion process, as shown in Fig. 2-3, namely, The calculated activation energy for hydrogen transport is 12 kJ/mol. Figure 2-3 H 2 permeance as a function of temperature for a SiC membrane. Solid line, membrane without an air oxidation treatment. Broken line, the membrane had been subjected to a 2 h air oxidation at 450 o C. ) ( exp 0 RT E P P H H H − = 56 After the permeation experiments were completed, the membrane was subjected to air oxidation at 450 o C and its permeation characteristics were again measured. The H 2 , He and Ar permeances of the membrane, after it had been subjected to the oxidation treatment, are also shown in Table 2-3. The H 2 and He permeances increase substantially; however, the Ar permeance more than doubles, with the net result that the ideal separation (He/Ar) decreases from 147 prior to oxidation to 89 after oxidation. The hydrogen diffusional process is again activated, but the activation energy is somewhat higher, about 17.6 kJ/mol. Though the transport of both He and H 2 is activated, the permeation of Ar is not, as shown in Fig. 2-4. This indicates, as noted in our prior publication [10], that H 2 (and He) transport through a different set of pores in the membrane than Ar and the other gases with larger kinetic diameter (CH 4 and CO 2 ). In fact, the permeance of Ar follows a (1/T) 0.5 dependence on temperature, indicative of Knudsen-type of transport. That air oxidation improves the Ar transport, in our view, means that it removes impurities from the mesoporous region through which Ar transport takes place. That He and H 2 permeances also increase may mean that the same impurities also occupy the pore mouths of the microporous regions. This is based on the fact that oxygen, with a kinetic diameter of 3.46 A o , close to that of CH 4 (3.8 A o ) and Ar (3.4 A o ), is unlikely to be able to have access to the microporous regions through which He and H 2 transport; however, no other direct experimental evidence exists to support this conjecture, at this point and time. 57 It is not clear, furthermore, where the likely (CH) x impurities come from. The AHPCS we utilize has an allyl mole fraction of ~10% (as reported by the manufacturer), which translates into C:Si ratio of 1.3; it is reported to result in a ceramic with a Si:C ratio very close to 1:1 (L.V. Interrante, 1995). Elemental analysis (at Galbraith Laboratories, Inc.) of AHPCS-derived SiC powders that we prepared using the same pyrolysis protocols with the SiC membranes shows a C:Si ratio of 1.2. The air oxidation does very little to change this ratio (in fact, after oxidation the ratio slightly increases to 1.25, but the small increase in the carbon content after oxidation is within the experimental error of the measurement technique). This indicates that air oxidation does not result in bulk oxidation of the material, and that any oxygen incorporation must be limited to the surface region. Table 2-3 Comparison of permeances and ideal separation factors of a SiC membrane before and after oxidation at 450 o C for 2 h. Before oxidation After oxidation He H 2 Ar He H 2 Ar Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) 0.89 0.44 0.006 1.21 0.71 0.014 S.F. (He/gas) 2 147 1.7 89 The elemental analysis also indicates that very little Al or Li is present in the SiC ceramic, a concern since LiAlH 4 is used as a catalyst during the preparation of AHPCS. Despite the fact that the air treatment decreases the separation factor, it is considered essential in 58 0.00E+00 1.00E-09 2.00E-09 3.00E-09 4.00E-09 5.00E-09 6.00E-09 7.00E-09 8.00E-09 0 10 20 30 40 50 60 70 80 0.00E+00 1.00E-10 2.00E-10 100 150 200 T ( o C) H2 SF Ar Separation Factor Permeance (mol m -2 s -1 Pa -1 ) order to prevent any further changes in membrane performance in case it is exposed accidentally to trace amounts of oxygen during its chemical reactor application. Figure 2-4 H 2 and Ar permeance and ideal separation factor as a function of temperature for a SiC membrane at 207 kPa (30 psi) transmembrane pressure. 59 3.3. Steam Stability of the Membranes To test their hydrothermal stability, the membranes were exposed to an equimolar mixture of steam and He at 310 kPa (45 psi) for a number of days at 200 o C (due to the limitations of our experimental system, this is the highest temperature we could expose the membranes to during the hydrothermal test; this temperature is of relevance, however, to the potential future use of such membranes for the water-gas shift reaction). The permeances of He and Ar were measured at the beginning of the experiment, shown as day 0 in Figs. 2-5 and 2-6. Subsequently, the membrane was exposed to the He/steam mixture, and the flow rate of He to the permeate side, after condensing the water, was measured on line, see Fig. 2-7 (the first point in Fig. 2-7 was measured 2 h after exposure to steam). After 48 h from the point when the membranes were first exposed to steam, the steam flow was turned off. Subsequently, the membrane was exposed to 12 h of He flow under the same pressure and temperature conditions; at the end of this period the He permeance was measured (the 2 nd point in Fig. 2-5). Its value had decreased by 23% from the value measured prior to the exposure to steam. The gas was then switched to Ar for an additional 12 h, and at the end of the period the Ar permeance was measured. Then, the He/steam flow was turned on for an additional 48 h, while monitoring on line the dry He flow. Then, the steam was again turned off, and the cycle was repeated, with the entire experiment lasting 21 days, out of which the membrane was exposed to steam for a period of 14 days. It is clear from Figs. 2-5, 2-7 that exposure to steam has an impact on He permeance. However, the effect saturates out after a period of a few days. The effect, furthermore, is indicative of structural changes rather than reflecting water condensation 60 in the microporous structure, as indicated by measuring the He permeance (Fig. 2-5) after the membrane had dried out for a period of 12 h. The Ar permeance, on the other hand, remains totally unaffected; indicating that whatever effect steam has on the membrane is limited only to the surface region, rather than resulting in bulk transformations of the SiC material. The membrane, after 21 days of exposure, still remains microporous as shown in Fig. 2-8, which shows the ideal He/Ar separation factor. It is not clear what caused the original decline in He permeation. Since the membrane used in the experiments described in Figs. 2-5, 2-8 had been exposed to the high-temperature air treatment, previously described, it may contain some surface oxygen. Si-O groups are known to be sensitive to the presence of steam (Park, Nishiyama, Egashira and Ueyama, 2001; Wu, Sabol, Smith, Flowers and Liu, 1994). Prior research on SiC ceramics indicated that exposure to water-containing air or steam at high temperatures enhanced the oxidation rates for both crystalline (More, Tortorelli, Ferber and Keiser, 2000; Opila, 1999) and amorphous (Tsui and Fang, 2005) SiC. These studies were, however, performed at much higher temperatures, and it is not clear whether they relate to any phenomena observed in this study. That He permeation stabilized after a few days on stream is significant for the use of these materials in reactive applications. Results from these studies will be presented in future publications. 61 0.00E+00 2.00E-09 4.00E-09 6.00E-09 8.00E-09 1.00E-08 1.20E-08 0 2 4 6 8 10 12141618 2022 Days Permeance (mol m -2 s -1 Pa -1 ) Figure 2-5 Single-gas He permeance at a transmembrane pressure of 207 kPa (30 psi) at 200 o C. 62 0.00E+00 1.00E-10 2.00E-10 3.00E-10 4.00E-10 5.00E-10 6.00E-10 7.00E-10 0 2 4 6 8 10 12 14 16 18 20 22 Days Permeance (mol m -2 s -1 Pa -1 ) Figure 2-6 Single-gas Ar permeance at a transmembrane pressure of 207 kPa (30 psi) at 200 o C. 63 0.00E+00 2.00E-03 4.00E-03 6.00E-03 8.00E-03 1.00E-02 1.20E-02 1.40E-02 0 2 4 6 8 10 12 14 16 18 20 22 Days Flow rate (Cm 3 /s) Figure 2-7 He flow rate of permeate gas at 200 o C. 64 0 5 10 15 20 25 30 35 40 0 2 4 6 8 10 121416182022 Days Separation Factor Figure 2-8 Ideal selectivity of He/Ar at a transmembrane pressure of 207 kPa (30 psi) at 200 o C. 65 4. Conclusions A new membrane preparation procedure was developed which involves coating, first, the membrane supports by a thin layer using slip-casting, prior to dip-coating additional layers. The resulting membranes exhibited enhanced performance, when compared to membranes prepared using the same supports without slip-casting a top surface layer. The reproducibility in preparing high quality membranes also improved significantly. The membranes reported here are microporous and have good hydrothermal stability, in contrast to membranes prepared previously using the same pre-ceramic polymer. The enhanced hydrothermal stability may be attributed to a different heat-treatment protocol used for AHPCS pyrolysis that leads to better cross-linked SiC membrane. Small size molecules like H 2 and He permeate through these membranes by activated transport, while Ar and molecules with larger kinetic diameters permeate by Knudsen diffusion. As a result, the separation characteristics of these membranes improve with temperature. SiC membranes that we prepared were hydrothermally stable for conditions close to water-gas shift reaction. Next phase of this research aims at increasing the SiC membranes’ permeability and selctivity which is of vital importance for more efficient hydrogen separation. In the next chapter, we will propose a new preparation technique with promisng results and of general interest for producing other type of inorganic membranes. 66 Chapter 3 : A Novel Sacrificial Interlayer-Based Method for the Preparation of Silicon Carbide Membranes 1 1. Introduction In this chapter we propose a novel technique based on the use of sacrificial interlayers for the preparation of nanoporous silicon carbide membranes. It involves periodic and alternate coatings of polystyrene sacrificial interlayers and silicon carbide pre-ceramic layers on the top of slip-casted tubular silicon carbide supports. Membranes prepared by this technique exhibit single gas ideal separation factors of helium and hydrogen over argon in the ranges (176-465) and (101-258), respectively, with permeances that are typically two to three times higher than those of silicon carbide membranes prepared previously by the more conventional techniques. Mixed-gas experiments with the same membranes indicate separation factors as high as 117 for an equimolar H 2 /CH 4 mixture. We speculate that the improved membrane characteristics are due to the sacrificial interlayers filling the pores in the underlying structure and preventing their blockage by the pre-ceramic polymer. The new method has good promise for application to the preparation of a variety of other inorganic microporous membranes. 1 B. Elyassi, M. Sahimi, T.T. Tsotsis, “A novel sacrificial interlayer-based method for the preparation of silicon carbide membranes”, Journal of Membrane Science, 316 (2008) 73-79. Copyright Elsevier. 67 In this research, PS sacrificial interlayers were employed with the goal of increasing the membrane permeance, but without negatively impacting the membrane separation characteristics. Previously, other researchers employed PS as a pore-forming agent (mixed in the PCS prior to layer formation) to increase the hydrogen permeance (Li et al., 1996; Li, Kusakabe and Morooka, 1997b; Suda et al., 2006a). This was attributed to phase separation and formation of PS domains during pyrolysis (Li et al., 1997b; Suda, Yamauchi, Uchimaru, Fujiwara and Haraya, 2006b; Suda et al., 2006c) to form mesoporous and macroporous regions, and potentially membrane defects. The resulting membranes had, as a result, a lower separation factor. On the other hand, the membranes reported in the present study have both an improved membrane permeance and also a higher separation factor. In a broader concept, sacrificial layers to prepare microporous inorganic membranes have been previously utilized to prepare silica and zeolite membranes. Gavalas and coworkers, for example, formed (Jiang, Yan and Gavalas, 1995; Yan, Davis and Gavalas, 1997b) a carbon barrier layer (by polymerization and carbonization of furfuryl alcohol) inside the support pores during the preparation of silica and zeolite ZSM-5 membranes. Hedlund and coworkers used (Hedlund, Jareman, Bons and Anthonis, 2003; Hedlund, Sterte, Anthonis, Bons, Carstensen, Corcoran, Cox, Deckman, De Gijnst, de Moor, Lai, McHenry, Mortier and Reinoso, 2002) wax as a masking layer against infiltration of precursors into the support for preparing MFI membranes. In these studies, the stated role of the sacrificial layer was to prevent infiltration of the membrane layer into the 68 underlying porous support in order to reduce the membrane thickness, and also to prevent crack formation. Though preventing film penetration into the underlying porous structure is also one of the roles that the PS sacrificial layers potentially play during SiC membrane preparation, their role during membrane formation is likely more complex, as PS has been shown previously (Suda et al., 2006b) to impact not only the cross-linking process, but also the formation of the three-dimensional (3D) pore structure during PCS pyrolysis. The key goal of this chapter is to introduce this new preparation technique, which could potentially be used to prepare other asymmetric inorganic membranes. 2. Experimental Ultra-high purity gases (He, H 2 , Ar, CO 2 , from Glimore Liquid Air Company, and CH 4 from Specialty Air Technologies, Inc.) were used in the experiments. Porous SiC support tubes were prepared using uniaxial cold-pressing of β-SiC powder (HSC059, provided by Superior Graphite Co., with an average particle size of 0.6 µm), together with appropriate sintering aids. For sintering, the green samples were heated at a temperature of 1700 o C, where they were kept in a furnace (Thermal Technology Inc., Model 1000-3060-FP20) for 3 h in flowing He. After sintering, the samples were cooled down to room temperature (with a cooling rate of 3 o C/min). The resulting tubes, shown in Fig. 3-1, had dimensions of 40 mm × 12 mm, with a thickness of 3 mm. The Archimedes method was used to measure the porosities of the supports, and their average pore size was estimated through permeation measurements. Further details about the sintering characteristics of various SiC powders, and the preparation and characterization of porous membrane 69 supports were presented elsewhere (Elyassi, Sahimi and Tsotsis, 2007; Suwanmethanond et al., 2000). The support tubes used in the membrane preparation were treated in flowing synthetic air at 450 o C (with the purpose of oxidizing any potential contaminants present), sonicated several times in acetone, and then dried prior to membrane film deposition. The dip- coating solution was prepared by dissolving 10 wt% of AHPCS (SMP-10, Starfire Systems, Inc.) in hexane. In order to prepare the slip-casting solution, the 0.6 µm SiC powder was mixed with acetone, and the lighter particles floating on the solution top were separated and dried. Scanning electron microscopy (SEM, Cambridge 360) observations of the particles indicated that their size was mostly in the range of (100-200) nm, which is close to the mean pore diameter of the support, as obtained from the permeation data (see section 2.1). The slip-casting solution consisted of 5 wt% of the SiC particles intermixed in the dip-coating solution. In our prior paper (Elyassi et al., 2007) the slip-casting solution contained a larger concentration of particles (10% vs. 5%). We found, however, that decreasing the concentration of the particles in the slip-casting solution down to this lower level did not impact adversely the quality of the final membrane layers using the sacrificial layer approach. Slip-casting was used to coat the first layer on the outer surface of the support tubes. To prepare this layer, the tubes were placed in the slip-casting solution for 12 s and, then, drawn out of the solution at a speed of 0.25 mm/s. The coated tubes were heated in 70 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 h, then to 400 o C, where they were also kept for 1 h, and finally to 750 o C, where they were kept for an additional 2 h. Subsequently, they were cooled down to room temperature in flowing Ar with a cooling rate of 3 o C/min. Figure 3-1 SiC tubular supports sintered at 1700 o C for 3 h. 71 The reason for the relatively slow heating (and holding at 200 o C) is that we have found (Elyassi et al., 2007), as have others (L.V. Interrante, 1994; Suda et al., 2006c), that using such heating rates and treatment at lower temperatures result, generally, in better cross- linked amorphous SiC materials (L.V. Interrante, 1994), and membranes with higher hydrogen permeabilities (Suda et al., 2006a). Subsequently, the membranes were dip-coated in a solution of 1 wt% of polystyrene (GPC grade, M w =2500, Scientific Polymers Products, Inc.) in toluene. For the PS coating, a dip-coating time of 12 s and a drawing rate of 0.58 mm/s were applied. After drying the PS layer at 100 o C for 1 h, the membranes were dip-coated in a 10 wt% of AHPCS in hexane solution, with a dip-coating time of 12 s and a drawing rate of 2 mm/s. In order to prevent significant dissolution of the formed polystyrene barrier layer into the AHPCS dip-coating solution, the solvent utilized for the solution must not dissolve the PS layer neither during dip-coating nor afterwards. The choice of hexane as the solvent for AHPCS was, therefore, due to polystyrene showing no substantial solubility in hexane. Following the coating of the AHPCS solution, the membranes were pyrolyzed at 750 o C for 2 h in Ar, following the same heat treatment protocol used for the preparation of the slip-casted tubes. A schematic of the preparation process is shown in Fig. 3-2. After the pyrolysis, three additional layers of PS and AHPCS were coated on the support (first the PS layer, followed by the AHPCS layer, etc.) with the same dipping protocol. After each 72 dual layer (PS+AHPCS) deposition, the membrane was pyrolyzed under the same conditions used for the pyrolysis of the first dual-layer. After coating and pyrolysis of the final dual-layer, the membranes were treated in flowing synthetic air for 2 h at 450 o C, with the purpose of oxidizing any potential carbon-based residues resulting from the preparation process (low-temperature air oxidation has been shown to be effective in removing minute amounts of carbon, for example as a common way for carbon nanotube purification (Li and Zhang, 2005; Osswald et al., 2006; Park et al., 2006). Though AHPCS has been shown to result in a SiC ceramic with a near stoichiometric Si:C ratio (L.V. Interrante, 1994), one cannot exclude the possibility that after pyrolysis there may still remain trivial amounts of carbon-based residues, which may influence overall membrane performance. We showed previously (Elyassi et al., 2007) with AHPCS-derived SiC membranes that the mild temperature air oxidation results in higher permeances but lower separation factors. This is an undesirable outcome, of course, but by oxidizing the SiC membranes in air, prior to their use, one ensures that no further variation in the membrane properties will occur due to accidental exposure to air. Moreover, in the potential use of these materials in membrane reactors (e.g., in steam reforming of hydrocarbons), if carbon deposits form during the reaction and affect the membranes, one possible means of recovering their performance would be by mild temperature air oxidation. Using such an oxidation treatment during the membrane preparation phase, one can hopefully ensure that no dramatic changes in the membrane properties will occur during regeneration. 73 Support Slip-casting and pyrolysis Dip-coating in polystyrene as sacrificial barrier Pyrolysis and decomposition of sacrificial interlayer Dip-coating in pre-ceramic polymer Figure 3-2 A schematic of the new preparation technique. 74 The morphology of the membranes (top surface and cross sections) were characterized by SEM. Permeation experiments were carried out using a Wicke-Kallenbach type permeation apparatus previously utilized to measure permeation through flat-disk and tubular SiC membranes (Ciora et al., 2004; Elyassi et al., 2007; Suwanmethanond et al., 2000). In the apparatus for tubular membranes, one end of the membrane is completely sealed using graphite tape and high temperature non-porous glue (J-B WELD); the other end of the membrane is attached to a flat metal ring using the same glue. The metal ring bearing the membrane is then installed in between the two half-cells of the permeation test-unit using O-rings. Single-gas permeation experiments were carried out by flowing a given gas through the apparatus half-cell facing the membrane dead-end, under constant pressure and temperature, and by measuring the amount of gas that permeates through the membrane to the permeate side, which was maintained at atmospheric pressure. In the experiments reported here, the pressure drop across the membranes was kept at 207 kPa (30 psi), while the temperature was varied from room temperature to 200 o C. We report the ideal selectivity for the membranes, which is defined as the ratio of the permeances of the different gases. Mixed-gas permeation tests were also carried out with select membranes using equimolar binary (H 2 /CH 4 , and H 2 /CO 2 ) gas mixtures. For the mixed- gas experiments, in addition to the gas flow to the permeate side of the membrane, the composition was also measured using a mass spectrometer. 75 3. Results and Discussion 3.1. Support Characterization As noted above, the porosity of the membrane supports was obtained by the Archimedes method (details about the technique and its use for the characterization of SiC macroporous substrates can be found in (Suwanmethanond et al., 2000)). The porosities of the various supports were 26.5 (±1%). For membrane applications, higher porosity for supports is more desirable, and we are currently working to increase the porosity of the supports without unduly impacting their mechanical properties. Fig. 3-3 shows the measured He and Ar single gas permeances for one of the supports. Clearly, the support is macroporous and the two permeances are a linear function of P av . One can express the permeance of the support for species j as (see, e.g., (Neomagus, van Swaaij and Versteeg, 1998)) . 148 ln( / ) 3 j o av o oo i j N B RT Permeance P K Pr r rRT M µπ ⎡ ⎤ == + ⎢ ⎥ ∆ ⎢ ⎥ ⎣ ⎦ (1) where 2 2 o o K B ε τ = (2) 4 o p K d ε τ = (3) In the above equations N j is the molar flux of species j, ∆P the pressure gradient across the support, r o the support outside diameter, r i the support inside diameter, R the universal 76 gas constant, T the temperature, ε the porosity, τ the tortuosity, µ the gas viscosity, P av. the arithmetic average pressure across the support, M j the molecular weight of species j, B o the viscous flow parameter, and K o the Knudsen flow parameter. Figure 3-3 Permeance vs. average pressure across the membrane for He and Ar. 200 250 300 350 400 450 0.0 2.0x10 -7 4.0x10 -7 6.0x10 -7 8.0x10 -7 1.0x10 -6 1.2x10 -6 1.4x10 -6 Permeance (mol m -2 Pa -1 s -1 ) P av. (kPa) He Ar 77 Using the permeation data in Fig. 3-3 and equations (1-3), we calculate an average pore size of ~130 nm for the support (and a tortuosity of 2); this is within the range of average particle size measured (from SEM images) of the powder used in our slip-casting solution (100-200 nm). We expect, therefore, most of these particles to sit on the top of the support, rather than propagate and be lodged deep into the membrane structure during slip-casting. 3.2. Membrane Characterization and Performance The quality of the membrane support is known to be important in determining the quality of the membranes that are prepared using these supports. Large defects in the supports are thought to lead to defects in the membranes themselves, thus resulting in low reproducibility in preparing membranes with high separation factors. Previously, we showed that putting a first layer on the support by slip-casting improved not only the performance of the membranes, but also increased the success rate in preparing high quality membranes (Elyassi et al., 2007). Therefore, in this study we continued placing a first layer by slip-casting as a means of conditioning the membrane support surface. In order to gain some insight into what role the sacrificial PS layer plays in the membrane preparation process, SEM pictures were taken of the membrane’s cross-section after coating the last PS layer and before coating the final layer of the pre-ceramic polymer. A SEM picture of such a cross-section is shown in Fig. 3-4, where within the SEM’s resolution one cannot discern a distinguishable uniform PS top layer on the membrane, 78 meaning that the PS layer is either very thin or that it has potentially infiltrated inside the porous structure, or both, thus forming a barrier prior to the deposition of the final AHPCS layer. That the PS polymer chains would potentially infiltrate into the underlying pyrolyzed AHPCS layer is not totally unexpected since we selected a PS with a low average molecular weight of 2500 with a corresponding radius of gyration of ~1 nm (Huber, Bantle, Lutz and Burchard, 1985). The fact that the PS sacrificial interlayer acts as an effective barrier is also supported by other experimental evidence. We observed, for example, that when using the sacrificial interlayer technique, the amount of dip-coating solution that is being uptaken is less than what is used when using the conventional technique without sacrificial layers coated on the supports, implying, likely, that the presence of the sacrificial interlayer allows for less AHPCS polymer to infiltrate within the underlying pores during coating. Using the sacrificial interlayer technique, the formation of a visible top SiC layer can be observed from the beginning of the membrane formation process, typically after a couple of coatings. Thicker membrane films also form, despite the fact that less AHPCS is being utilized. As seen in Fig. 3-4, the thickness of the membrane layer, even before the final SiC layer coating, is ~7 µm, which is larger than the thickness of membranes formed (typically ~2 µm (Elyassi et al., 2007)), without the aid of PS interlayers. 79 Figure 3-4 SEM picture showing the cross-section of a SiC membrane after three dual (PS+AHPCS) coatings, and one additional coating of PS. 7µm 80 Despite the fact that the new technique prepares thicker membranes, their permeances are significantly higher than those of the membranes prepared by the conventional technique, as discussed below. Why this happens is not entirely clear at this point, since the presence of the PS sacrificial layers implies the onset of a set of complex phenomena during the membrane preparation process. For instance, the PS layer has a melting point of ~240 o C (Brandrup, Immergut and Grulke, 1999) and its decomposition starts at temperatures higher than 350 o C (Marcilla and Beltran, 1995); we expect cross-linking of AHPCS to be taking place within the same temperature range during pyrolysis. Also relevant here is the recent investigation by Suda et al. (Suda et al., 2006b), in whose studies PS was used as a pore former (rather than as a sacrificial barrier layer used in our studies) mixed with regular polycarbosilane which, unlike AHPCS, needs to be cross-linked by the addition of chemical cross-linking agents. PS appeared to actively participate in the formation of the 3D pore structure. How the AHPCS layer interacts with the PS sacrificial layer during pyrolysis, and the impact on the membrane characteristics definitely need more in-depth research. The permeation properties of three different single gases (He, H 2 , and Ar) through the membranes were measured at 200 o C. The range of permeances and the corresponding ideal separation factors are presented in Table 3-1. The H 2 permeance of the membranes, varying in the range (1.05-2.0)×10 -8 mol m −2 s −1 Pa −1 , is significantly higher than the H 2 permeances reported previously with membranes prepared with the same exact protocol (number of AHPCS layers, deposition conditions, pyrolysis temperature, etc.) but without 81 the use of sacrificial PS layers (Elyassi et al., 2007). This is so, despite the fact, as previously noted, that the thickness of the membranes prepared by the use of sacrificial layers, is ~3 times larger than the corresponding thickness of the membranes prepared by the technique without the use of sacrificial layers. The corresponding ideal separation factors are also significantly higher. For example, for the He/Ar pair the ideal separation factors varied between 176 and 465. The He/Ar ideal separation factor reported previously was 89 (Elyassi et al., 2007). The permeances of He, H 2 , and Ar, as a function of temperature, for one of the membranes prepared by the sacrificial layer method, are shown in Fig. 3-5. Small molecules, such as H 2 and He, permeate through the SiC membranes by activated transport, while Ar and other molecules with larger kinetic diameters permeate following a Knudsen diffusion mechanism. For He the activation energy for transport is 8.9 kJ/mol, while for H 2 the corresponding value is 11.7 kJ/mol, which is close to the values measured previously (Elyassi et al., 2007; Wach et al., 2007a). As a result, the separation characteristics of the membranes improve with temperature. The permeance of H 2 at 200 o C is 2×10 -8 mol m −2 s −1 Pa −1 . Extrapolated to a temperature of 500 o C, the expected permeance for H 2 should be 6.5×10 -8 mol m −2 s −1 Pa −1 . By comparison, Pd and Pd-alloy membranes, around 500 o C, typically exhibit permeances in the range (10 -7 -10 -6 ) mol m −2 s −1 Pa −1 (Rothenberger et al., 2004). However, Pd membranes are known to be sensitive to the presence of H 2 S and hydrocarbon impurities; use of Pd-alloy membranes (e.g., Pd- 82 Ag) also faces challenges for long-term usage at temperatures significantly higher than 500 o C. Table 3-1 Range of single-gas permeances and separation factors, measured at 200 o C, for membranes prepared by pyrolysis at 750 o C for 2 h. He H 2 Ar Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) 1.8-4.3 1.05-2 0.01 S.F. (gas/Ar) 176-465 101-258 83 Figure 3-5 He, H 2 , and Ar permeance as a function of temperature for a SiC membrane at 207 kPa (30 psi) transmembrane pressure. 0 50 100 150 200 0.0 5.0x10 -11 1.0x10 -10 1.5x10 -10 1.0x10 -8 2.0x10 -8 3.0x10 -8 4.0x10 -8 Permeance (mol m -2 Pa -1 s -1 ) T ( o C) He H 2 Ar 84 As noted in the introduction, a key potential application for these SiC-based membranes is in hydrogen production, through methane steam reforming where such membranes are integrated in membrane reactors. Another key potential application is in modern IGCC power generation systems, where such membranes may potentially be useful in producing a CO 2 -free H 2 stream for further use in turbines or fuel cells. The permeation characteristics of two of these membranes towards an equimolar H 2 /CH 4 mixture were tested using the experimental procedure previously outlined, and also described in greater detail in (Elyassi et al., 2007) (the permeation characteristics of one of the membranes were also tested with an equimolar H 2 /CO 2 mixture). The membranes exhibited good selectivity towards hydrogen (see Table 3-2) with H 2 /CH 4 separation factors that are significantly higher than those reported previously with membranes that were prepared without the use of sacrificial PS layers (Elyassi et al., 2007). Table 3-2 Mixed-gas permeation properties of SiC membranes, measured at 200 o C, prepared by pyrolysis at 750 o C for 2 h. For the mixed-gas experiments the membranes were exposed to an equimolar mixture of H 2 /CH 4 or H 2 /CO 2 . H 2 Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) S.F. (H 2 /CH 4 ) H 2 Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) S.F. (H 2 /CO 2 ) Sample I 1.54 104 1.47 93 Sample II 1.55 117 --- --- 85 4. Conclusions A new procedure was developed for the preparation of SiC-based membranes which involves the periodic coating of sacrificial polystyrene interlayers in between pre-ceramic AHPCS layers. Applying this technique, we have been able to prepare membranes with significantly enhanced performance when compared to the membranes we prepared previously without using sacrificial interlayers. The mechanism by which the sacrificial PS layers participate in the membrane formation process is likely to be very complex. We have selected to form the sacrificial interlayers using a low molecular weight (~2500) PS with a corresponding small radius of gyration of ~1 nm (Huber et al., 1985). The idea here was that the low molecular weight PS chains would be able to penetrate the underlying -[Si-C]- type pores, thus preventing the AHPCS chains, during the recoating, from re-entering these pores. The phenomena are likely to be more complex, however. As previously noted, the PS layer has a melting point of ~240 o C (Brandrup et al., 1999), and begins to decompose at temperatures higher than 350 o C (Marcilla and Beltran, 1995), where we expect cross-linking of AHPCS to be also taking place. As noted in the recent study by Suda et al. (Suda et al., 2006b), PS plays an important role as a pore-former through the release of various low molecular weight gases but also, in addition, actively participates during the cross-linking and 3D pore-structure formation process. It is the ability of PS to directly influence the 3D structure of the resulting SiC-based ceramic that makes the technique different from previous efforts using sacrificial layers (Hedlund et al., 2003; Hedlund et al., 2002; Jiang et al., 1995; Yan et al., 1997b), where the barrier layer did not directly participate in the zeolite membrane formation process. 86 There is still, of course, a lot that is not understood about the preparation technique, and additional studies are currently in progress. For example, the thickness of the PS barrier appears to be a key factor; too thick barrier layers, for example, are undesirable, as they result in peeling of the SiC layer after pyrolysis. The methodology employing sacrificial barrier layers, such as polystyrene, which decompose in inert or oxidizing environments is potentially of general applicability for the formation of other microporous inorganic thin films and membranes. The next phase of this reseach is an investigation on preparing highly porous SiC nanofibers. Also, the effect of PS on the morphology and physical properties of such nanofiber is investigated. The goal is to apply such highly porous nonfibers as filler embedded into the structure of membranes. 87 Chapter 4 : Effect of Polystyrene on the Morphology and Physical Properties of Silicon Carbide Nanofibers 1 1. Introduction Due to its many unique properties, such as high thermal conductivity, stability, chemical inertness, and high oxidation resistance, silicon carbide (SiC) is a promising candidate for a wide range of applications that involve harsh environments. It has potential for application as a high-temperature, high-power, and high-frequency semiconductor in the aerospace industry (Casady and Johnson, 1996), as a gas sensor (Spetz, Tobias, Baranzahi, Martensson and Lundstrom, 1999) and hydrogen storage medium for the automotive industry (Mpourmpakis, Froudakis, Lithoxoos and Samios, 2006; Yampol'skii et al., 2006), as a material from which hydrogen permselective (Elyassi et al., 2007; Elyassi, Sahimi and Tsotsis, 2008) or even biocompatible membranes for the bioMEMS industry (Rosenbloom et al., 2004) are made, as a catalyst support (Nhut, Vieira, Pesant, Tessonnier, Keller, Ehret, Pham-Huu and Ledoux, 2002; Pham-Huu, Keller, Ehret and Ledoux, 2001; Vannice, Chao and Friedman, 1986), as well as a 1 B. Elyassi, T.W. Kim, M. Sahimi, T.T. Tsotsis, “Effect of polystyrene on the morphology and physical properties of silicon carbide nanofibers”, Materials Chemistry and Physics, Submitted. 88 reinforcing material (Silenko, Shlapak, Tomila, Bykov, Kuz'menko, Okun and Ragulya, 2008). Due to its many promising applications, there has been increasing interest in producing various SiC nanostructures. In the production of such nanostructures carbon nanotubes (CNT) have been commonly employed as the templates, in order to react with Si or SiO vapors at high temperatures and under ultra-low pressures (Keller, Pham-Huu, Ehret, Keller and Ledoux, 2003; Sun, Li, Wong, Wong, Lee, Lee and Teo, 2002; Taguchi, Igawa, Yamamoto and Jitsukawa, 2005; Wu, Yang, Pan and Chen, 2007). Such a synthesis route has resulted in crystalline SiC nanotubes and nanowires. However, it is difficult to use the technique and maintain product uniformity and control on the final product. The approach is also costly because one needs to operate at high temperatures and low pressures for long periods of time and, generally, only a small fraction of the CNTs is converted. Another method for producing SiC nanostructures involves catalyst- assisted chemical-vapor deposition (CVD), during which silicon -and carbon- containing compounds are reacted on metallic catalysts (Kang, Lee and Boo, 2004; Leonhardt, Liepack, Biedermann and Thomas, 2005; Silenko et al., 2008; Xie, Tao and Wang, 2007). A third method involves the use of templates of anodic aluminum oxide (AAO) and utilizing the CVD technique, or a solution deposition of polymeric precursors to produce SiC nanotubes and nanowires inside the substrates. Li et al. (Li, Zhang, Meng and Guo, 2006), for example, prepared SiC nanowires inside the AAO channels by reacting SiO vapors with C 3 H 6 at 1230 o C. Xu et al. (Xu, He, Guo and Wang, 2006) reported the 89 preparation of amorphous SiC nanorods inside the AAO templates by RF sputtering. Yen et al. (Yen, Jou and Chu, 2005) reported the preparation of silicon oxycarbide nanotubes by infiltration of polycarbosilane (PCS) into the AAO templates, followed by pyrolysis at 1100 o C in a vacuum furnace. Cheng et al. (Cheng, Interrante, Lienhard, Shen and Wu, 2005) prepared SiC nanostructures by the CVD of SiC polymeric precursor with the trade-name SP-4000 (Starfire® Systems, Inc., with nominal structure [SiH 2 CH 2 ] n , n=2-8, including branched and cyclic isomers (Gallis, Futschik, Castracane, Kaloyeros, Efstathiadis, Sherwood, Hayes and Fountzoulas, 2003)) and the solution infiltration of polysilaethylene ([SiH 2 CH 2 ] n or PSE) into the AAO at 1000 o C. In their work, the CVD technique produced uniform SiC nanotubes, but the solution deposition technique produced SiC nanofibers with a bamboo-like structure. The bamboo-like nanofibers produced by the liquid phase technique were explained to be the result of the Rayleigh instability (Chen, Zhang and Russell, 2007). In this chapter we report on the synthesis of uniform porous SiC nanofibers by infiltration into the AAO templates of allyl-hydridopolycarbosilane (AHPCS), which is a polymeric precursor that can be converted into SiC with a near stoichiometric ratio at relatively low temperatures in inert atmospheres, without the need for any catalyst or oxygen for cross- linking (Elyassi et al., 2007; Interrante, Whitmarsh, Sherwood, Wu, Lewis and Maciel, 1994). Aside from the work of Cheng et al. (Cheng et al., 2005), we are not aware of any other reports on the solution infiltration method used for producing SiC nanofibers. The study of Cheng et al. produced nanofibers with a bamboo-like structure. The main theme 90 of this report, however, is studying the effect that adding polystyrene (PS), as a pore former, into the AHPCS precursor solution has on the structure and property of the nanofibers. These nanofibers are being used by our group in the preparation of nanoporous hydrogen perm-selective membranes (Elyassi et al., 2008). Due to their very high surface areas (see below), they could potentially also be used as catalysts supports. 2. Experimental The AAO disks (Anodisc ® , Whatman Inc.) were used as templates. They are asymmetric alumina disks consisting of a top layer with a thickness of 2 µm and an average pore size of ~100 nm, sitting on the top of a support layer with a thickness of 58 µm and average pore size of ~200 nm (Vermisogiou, Pilatos, Romanos, Karanikolos, Boukos, Mertis, Kakizis and Kanellopoulos, 2008). The solution that contained the SiC precursor was prepared by dissolving a concentration of either 20 or 30 wt% of AHPCS (SMP-10, Starfire® Systems, Inc.) in hexane. In order to study the effect that adding polystyrene to the AHPCS solution has on the nanofiber morphology, polystyrene was added to the AHPCS solutions in toluene. 5, 15, and 30 wt% of the PS (on a per AHPCS basis, i.e., g of the PS per g of the AHPCS in the solution) were added to a 20 wt% solution of the AHPCS in toluene. Prior to immersing the AAO disk templates into the solutions, they were dried at 70 o C in air for 1 h. The templates, after being infiltrated by the solutions, were heated in flowing Ar in a tube furnace (Lindberg/Blue, Model STF55433C) at a rate of 2 °C/min, first to 200 °C, at which temperature they were kept for 1 h, then to 400 °C, where they were kept for 1 h, and finally to 750 °C, where they were kept for an 91 additional 2 h for pyrolysis. Subsequently, they were cooled down to the room temperature in flowing Ar with a cooling rate of 3 °C/min. After pyrolysis, the disks were immersed in a (6 M) NaOH solution for 48 h, in order for them to dissolve. The SiC nanofibers were separated by centrifuging the solution at 2600 rpm for 20 min, and were then washed several times with deionized water. The process of washing with water and re-centrifuging was continued until the pH of the solution was at the neutral value. The morphology of the resulting nanofibers was then studied using the SEM (Cambridge 360) and TEM (Philips EM420). The nitrogen adsorption isotherm of the nanofibers was measured at 77 K with a Micromeritics ASAP 2010 apparatus. 3. Results and Discussion The photographs of the two sides of the AAO templates, shown in Fig. 4-1, indicate that the templates have a top layer with an average pore size of ~100 nm, sitting on a bottom layer with an average pore size of ~200 nm. Using low concentrations of the AHPCS in hexane (2-5 wt%) did not produce SiC nanofibers. This may be due to the fact that at such concentrations the polymer matrix is below the percolation threshold (Sahimi, 2003) (i.e., the minimum volume fraction for the formation of stable and connected structures), or that weak structures form which break down during or after template removal. Using higher AHPCS concentrations, such as 20 wt% and 30 wt%, made it possible to form SiC nanofibers. Although 20 wt% of polymer translates to 14 vol% which is slightly below the percolation threshold for random continuous materials (i.e., the minimum volume fraction at which a continuous macroscale 92 matrix is formed), one should, however, recognize that the polymer does not necessarily have a random structure. It is more likely that there are some positive correlations in the polymer structure, which result in a lower percolation threshold (Sahimi, 2003). Figure 4-2 shows the SEM and TEM photographs of SiC nanofibers, obtained using 30 wt% solutions of AHPCS in hexane. The diameter of the nanofibers is, as expected, within the range of the template pore sizes which is ~200 nm. Figure 4-1 Two sides of the AAO templates showing a pore diameter of ~100 nm on one side, and ~200 nm on the other side. 93 Figure 4-2 (A) The TEM, and (B) SEM of SiC nanofibers. 94 Polystyrene has been used by several investigators as a pore former, in order to improve the separation characteristics of PCS-derived SiC membranes (Li et al., 1996; Li et al., 1997b; Suda et al., 2006b; Yasuda et al., 2006). Such membranes exhibit high permeances, which are the result of the contribution of the gaseous products of polystyrene pyrolysis to the formation of the 3D material structure. On the other hand, one should also be cautious in indiscriminately adding polystyrene, as it may segregate and form distinct PS ellipsoidal domains within the PCS bulk after solvent removal, which would then result in defects during the PS decomposition, leading to adverse impact on the membrane selectivity (Li et al., 1997a; Suda et al., 2006b), if the material is used as a membrane. Here, the idea behind adding the PS to the AHPCS precursor during the preparation of the various SiC nanostructures was to investigate whether it impacts the structural characteristics of these materials similar to that observed during SiC nanoporous membrane formation. One expects that, by increasing the amount of the PS added to the AHPCS precursor solution, one can produce a broad range of materials, ranging form the uniform porous nanofibers that are produced in the absence of the PS, to porous nanofibers containing large ellipsoidal porous domains, and nanotubes formed when the polystyrene domains segregate and create a central sacrificial PS core. Figure 4-3 shows the nanofibers that are formed by adding the PS, at a concentration of 10 wt% (g of the PS/g of AHPCS), to 30 wt% AHPCS solution in toluene. The selected area electron diffraction (SAED) pattern in this figure reveals that the nanofibers have an amorphous structure. Increasing the temperature causes evaporation of the solvents and, 95 thus, leading the stable solution to pass the binodal or spinodal curves of the ternary (AHPCS-toluene-PS) phase diagram and formation of polymer-rich and polymer-lean domains seen on phase inversion. Figure 4-3 (A) TEM, and (B) SEM photographs of SiC nanofibers, prepared by adding 10 wt% of the PS into the AHPCS solution. Hollow domains can be seen as the result of the PS decomposition. The embedded SAED in the TEM picture indicates the amorphous structure of the fibers. 96 Phase segregation of the PS inside the nanofibers into ellipsoidal domains, and their complete decomposition during pyrolysis, have resulted in the formation of hollow ellipsoidal regions. This is indicated more clearly in Fig. 4-4, which shows high resolution TEM images of two nanofibers that were produced from the pyrolysis of the 10 wt% PS in the AHPCS precursor solution. Visible in these photographs is the presence of hollow ellipsoidal domains. Further careful examination of these nanofibers indicates that, the region in the vicinity of the hollow domains is darker than the other parts of nanofiber in the bright field images. This indicates that, as a result of the interaction between the PS and AHPCS during pyrolysis, the boundary regions have a denser structure. The high resolution TEM photograph of the nanofibers in Fig. 4-4 reveals a porous structure, which was expected from our previous experience with the AHPCS- derived SiC membranes (Elyassi et al., 2007; Elyassi et al., 2008). Within the resolution of this figure, pores as small as 5 nm can be discerned. XRD investigations of the nanofibers reveal an amorphous structure, which is expected for SiC materials generated by pyrolysis of the AHPCS at the relatively low pyrolysis temperature at which we prepared the nanofibers (Sterte et al., 2001). Further increase of the PS wt% fraction to 15wt% (g of the PS/g of AHPCS) in a 30 wt% AHPCS toluene solution leads to an opaque solution, which indicates that it is in the two- phase region of the ternary phase (AHPCS-toluene-PS) diagram. Thus, further addition of PS was not possible. 97 Figure 4-4 TEM picture of SiC nanofibers, prepared by 10 wt% addition of the PS (AHPCS basis). The nanofibers have a porous structure and hollow ellipsoidal domains can be seen, resulting from the PS decomposition. 98 Figure 4-5 presents the nitrogen adsorption isotherm of the SiC nanofibers, prepared using the 10 wt% PS/30 wt% AHPCS in the toluene precursor solution. The fibers have a high BET surface area of 457 m 2 /g with a total pore volume of 0.79 cm 3 /g. The Horvath Kawazoe (HK) pore size distribution of the nanofibers is characterized by a sharp peak at a pore size of ~4.5 Å, with a corresponding pore volume of 0.16 cm 3 /g in the nanoporous region. Analysis of the BJH pore size distribution indicates an average pore size of ~7 nm for the mesoporous region of the material. Figure 4-6 shows the nitrogen adsorption isotherm of the SiC nanofibers prepared using the 30 wt% AHPCS in toluene solution, without adding the PS. These fibers have a lower BET surface area of 366 m 2 /g with a total pore volume of 0.78 cm 3 /g. The HK pore size distribution of these nanofibers is again characterized by a sharp peak at a pore size of ~4.5 Å, with a corresponding pore volume of 0.12 cm 3 /g in the nanooporous region, and a slightly larger BJH average pore size of ~8 nm. Higher surface areas for the SiC samples prepared with the addition of the PS to the PCS solutions were also reported by Suda et al (Suda et al., 2006b), who studied the effect of the PS addition during SiC membrane formation. The higher surface areas were explained to be due to the release of gases during the PS decomposition, and their contribution to the formation of the 3D structure. 99 Figure 4-5 Nitrogen adsorption isotherm of SiC nanofibers formed by the 10 wt% addition of the PS (AHPCS basis), together with the HK pore size distribution for the nanoporous region. 100 Figure 4-6 Nitrogen adsorption isotherm of SiC nanofibers, prepared by 30 wt% of the AHPCS in toluene, together with the HK pore size distribution for the nanoporous region. 101 A number of precursor solutions were also prepared by adding the PS to a 20 wt% AHPCS in toluene solution. Lowering the concentration of the AHPCS allows one to prepare stable precursor solutions that contain larger fraction of the PS in the AHPCS. Precursor solutions that were transparent and stable were prepared by dissolving 5, 15, and 30 wt% of the PS (g of the PS/g of AHPCS) into a 20 wt% AHPCS in toluene solution. As can be seen in Fig. 4-7, increasing the concentration of the PS increases the number of hollow domains in the fibers, and at a fraction of 30 wt% of the PS in the AHPCS, one begins to see hollow fiber formation. The hollow fibers are probably formed as the result of the coalescence of the hollow domains. Further increase of the PS fraction (to 65 wt% per AHPCS) resulted in opaque solutions, which were not studied further. 102 Figure 4-7 TEM pictures of SiC nanofibers and hollow fibers, prepared by adding, (A) 5, (B) 15, and (C) 30 wt% PS (AHPCS basis) to 20 wt% AHPCS in toluene solution. 103 4. Conclusions Uniform porous SiC nanostructures with high surface area were fabricated using a template technique. Adding polystyrene, as a pore former, to the precursor AHPCS solution resulted in SiC nanofibers with ellipsoidal hollow domains inside the fiber structure, as the result of phase segregation during the drying process. Increasing the concentration of the PS in the starting AHPCS precursor solution resulted in nanofibers with an increased fraction of hollow domains, and even in the formation of some completely hollow fibers. Furthermore, formation of hollow domains in the structure is both interesting and important, and opens the way to encapsulation of other materials inside the nanofibers if used instead of the PS. In the next chapter, we present the effect incorporating such highly porous nanofibers into the structure of SiC membranes has on their performance. In addition to these fillers, in order to get some insight into possible roles that different fillers play in the final performance of membranes; nonporous SiC powders fillers were investigated. Performances of such membranes were measured and compared. Some undersanding on the interaction of filler with membrane matrix and the effect that shape or size of the filler has on membrane performance is created. 104 Chapter 5 : Effect of Porous and Nonporous Fillers on the Performance of Silicon Carbide Membranes 1. Introduction In this chapter we study the effect that incorpation of highly porous SiC nanofibers and nonporous fine SiC powders fillers has on the performance of SiC membranes. SiC membranes were prepared by incorporating porous and nonporous fillers in the membrane structure using a novel sacrificial interlayer-based technique that was described in chapter 3. Membranes prepared using nanofiber fillers exhibit single gas ideal separation factors of helium and hydrogen over argon in the ranges (71-385) and (45-146), respectively. Such membranes exhibited mixed gas H 2 /CH 4 separation factor as high as 70 which is a little lower than those fabricated with SiC powder fillers exhibiting separation factor of 90. BET analysis of fibers coated with SiC -in the same condition that membranes were prepared- revealed a drastic reduction in accessible pore volume and surface area. We speculate packing degree of fillers and their interaction with membrane matrix influence the final performance of these membranes. 2. Experimental Ultra-high purity gases (He, H 2 , Ar, from Glimore Liquid Air Company, and CH 4 from Specialty Air Technologies, Inc.) were used in the experiments. Porous SiC support tubes 105 were prepared using uniaxial cold-pressing of β-SiC powder (HSC059, provided by Superior Graphite Co., with an average particle size of 0.6 µm), together with appropriate sintering aids and sintering at 1700 o C. Further details about the sintering characteristics of various SiC powders, and the preparation and characterization of porous membrane supports were presented elsewhere (Ciora et al., 2004; Elyassi et al., 2007; Elyassi et al., 2008). The support tubes used in the membrane preparation were treated in flowing synthetic air at 450 o C (with the purpose of oxidizing any potential contaminants present), sonicated several times in acetone, and then dried prior to membrane film deposition. The dip- coating solution was prepared by dissolving 10 wt% of AHPCS (SMP-10, Starfire Systems, Inc.) in toluene. Two types of fillers were incorporated into the structure of membranes; SiC nanofibers (with length of couple of microns and diameter of ~250 nm) vs. fine SiC powders. SiC nanofibers were prepared by immersing anodized aluminum oxide (AAO) templates in a solution of AHPCS and PS and then pyrolyzing; details will be published later. SiC fine powders were obtained by mixing the 0.6 µm SiC powder with acetone, and separating and drying the lighter particles. Scanning electron microscopy (SEM, Cambridge 360) observations of the particles indicated that their size was in the range of (100-200) nm, which is close to the mean pore diameter of the support, as obtained from the permeation data (Elyassi et al., 2008). BET analysis was conducted on both fillers to get insight into their structural properties. Slip-coating solutions with the same volume percentage (1%) of fibers and powders in dip-coating 106 solution (10 wt% of AHPCS in toluene) were prepared. The slip-coating solution consisted of each fillers separately intermixed in the dip-coating solution. Slip-coating solution was sonicated and the support tubes were placed in the slip-coating solution for 12 s and, then, drawn out of the solution at a speed of 0.25 mm/s. 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 h, then to 400 o C, where they were also kept for 1 h, and finally to 750 o C, where they were kept for an additional 2 h. Subsequently, they were cooled down to room temperature in flowing Ar with a cooling rate of 3 o C/min. The reason for the relatively slow heating (and holding at 200 o C) is that we have found (Elyassi et al., 2007), as have others (L.V. Interrante, 1994; Suda et al., 2006a), that using such heating rates and treatment at lower temperatures result, generally, in better cross-linked amorphous SiC materials (L.V. Interrante, 1994), and membranes with higher hydrogen permeabilities (Suda et al., 2006a). We applied polystyrene sacrificial interlayer-based technique for preparing all of the membranes as explained here. The membranes were dip-coated in a solution of 1 wt% of polystyrene (GPC grade, M w =2500, Scientific Polymers Products, Inc.) in toluene. For the PS coating, a dip-coating time of 12 s and a drawing rate of 0.58 mm/s were applied. After drying the PS layer at 100 o C for 1 h, the membranes were dip-coated in a 10 wt% of AHPCS in hexane solution, with a dip-coating time of 12 s and a drawing rate of 2 mm/s. In order to prevent significant dissolution of the formed polystyrene barrier layer into the AHPCS dip-coating solution, the solvent utilized for the solution must not 107 dissolve the PS layer neither during dip-coating nor afterwards. The choice of hexane as the solvent for AHPCS was, therefore, due to polystyrene showing no substantial solubility in hexane. Following the coating of the AHPCS solution, the membranes were pyrolyzed at 750 o C for 2 h in Ar, following the same heat treatment protocol used for the preparation of the slip-casted tubes. After the pyrolysis, five additional layers of PS and AHPCS were coated on the support (first the PS layer, followed by the AHPCS layer, etc.) with the same dipping protocol. After each dual layer (PS+AHPCS) deposition, the membrane was pyrolyzed under the same conditions used for the pyrolysis of the first dual-layer. After coating and pyrolysis of the final dual-layer, the membranes were treated in flowing synthetic air for 2 h at 450 o C, with the purpose of oxidizing any potential carbon-based residues resulting from the preparation process (low-temperature air oxidation has been shown to be effective in removing minute amounts of carbon, for example as a common way for carbon nanotube purification (Li and Zhang, 2005; Osswald et al., 2006; Park et al., 2006). Though AHPCS has been shown to result in a SiC ceramic with a near stoichiometric Si:C ratio (L.V. Interrante, 1994), one cannot exclude the possibility that after pyrolysis there may still remain trivial amounts of carbon-based residues, which may influence overall membrane performance. We showed previously (Elyassi et al., 2007) with AHPCS-derived SiC membranes that the mild temperature air oxidation results in higher permeances but lower separation factors. Although this is an undesirable 108 outcome, of course, but by employing such treatment, prior to membrane use, one ensures that no further variation in their properties will occur due to accidental exposure to air. Moreover, in the potential use of these materials in membrane reactors, if carbon deposits form during the reaction and affect the membranes, one possible means of recovering their performance would be by mild temperature air oxidation. Using such an oxidation treatment during the membrane preparation phase, one can ensure that no dramatic changes in the membrane properties will occur during regeneration. The morphology of the membranes were characterized by SEM. Permeation experiments were carried out using a Wicke-Kallenbach type permeation apparatus previously utilized to measure permeation through flat-disk and tubular SiC membranes (Ciora et al., 2004; Elyassi et al., 2007; Elyassi et al., 2008). In the apparatus for tubular membranes, one end of the membrane is completely sealed using graphite tape and high temperature non- porous glue (J-B WELD); the other end of the membrane is attached to a flat metal ring using the same glue. The metal ring bearing the membrane is then installed in between the two half-cells of the permeation test-unit using O-rings. Single-gas permeation experiments were carried out by flowing a given gas through the apparatus half-cell facing the membrane dead-end, under constant pressure and temperature, and by measuring the amount of gas that permeates through the membrane to the permeate side, which was maintained at atmospheric pressure. In our experiments, the pressure drop across the membranes was kept at 207 kPa (30 psi) and temperature at 200 o C. We report the ideal selectivity for the membranes, which is defined as the ratio of the permeances of 109 the different gases. Mixed-gas permeation tests were also carried out with select membranes using equimolar binary (H 2 /CH 4 ) gas mixtures. For the mixed-gas experiments, in addition to the gas flow to the permeate side of the membrane, the composition was also measured using a HP 5890 series II GC. 3. Results and Discussion Previously, we showed that putting a first layer of SiC fine powders on the support by slip-coating improved not only the performance of the membranes, but also increased the success rate in preparing high quality membranes (Elyassi et al., 2007; Elyassi et al., 2008). Therefore, in this study we continued placing a first layer by slip-coating as a means of conditioning the membrane support surface. However, as mentioned before in this study two different fillers were employed to investigate their effects on the membrane performance. One is highly porous nanofibers and the other impermeable SiC fine powders. Fig. 5-1(A, B) show the adsorption isotherms of these two fillers. Powders exhibited a low BET surface area of 11.6 m 2 /gr with a total pore volume of 0.07 cm 3 /gr while fibers exhibited a high surface area of 457 m 2 /gr with a total pore volume of 0.79 cm 3 /gr. Assuming spherical powder and based on measured BET surface area one can calculate diameter of about 200 nm for powders which is in close agreement with SEM pictures (not shown here). This proves nonporous structure of powders versus highly porous structure of fibers. 110 0 100 200 300 400 500 0 0.2 0.4 0.6 0.8 1 P/P o V o lu m e A d so rb ed (cm 3 /g r ) ST P B 0 0.4 0.8 1.2 1.6 2 2.4 0 0.2 0.4 0.6 0.8 1 1.2 Pore Diameter (nm) Po r e V o lu m e (cm 3 /g .n m ) 0 10 20 30 40 50 0 0.2 0.4 0.6 0.8 1 P/P o V o lu m e A d so rb ed (cm 3 / g r ) ST P A BET surface area: 12 m 2 /gr BJH adsorption pore volume: 0.07 cm 3 /gr HK pore volume: 0.003 cm 3 /gr BET surface area: 457 m 2 /gr BJH adsorption pore volume: 0.63 cm 3 /gr HK pore volume: 0.16 cm 3 /gr HK pore size distribution Figure 5-1 Adsorption isotherm of A) SiC nanofibers and B) SiC powders. 111 The permeation properties of three different single gases (He, H 2 , and Ar) through the membranes were measured at 200 o C. The range of permeances and the corresponding ideal separation factors for membranes with nanofiber fillers are presented in Table 5-1. Table 5-1 Range of single-gas permeances and separation factors, measured at 200 o C, for membranes prepared by nanofiber fillers and pyrolyzed at 750 o C for 2 h. He H 2 Ar Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) 0.8-2 0.5-0.9 0.005-0.01 S.F. (gas/Ar) 71-385 45-146 Table 5-2 exhibits mixed gas H 2 permeance and H 2 /CH 4 separation factors of the two select membranes with powder and nanofiber fillers. As can be seen, these membranes show similar performance. Table 5-2 Mixed-gas permeation properties of SiC membranes (with nanofiber and powder fillers), measured at 200 o C, prepared by pyrolysis at 750 o C for 2 h. For the mixed-gas experiments the membranes were exposed to an equimolar mixture of H 2 /CH 4 . H 2 Permeance×10 8 (mol m − 2 s − 1 Pa − 1 ) S.F. (H 2 /CH 4 ) Membrane with nanofiber filler 1.2 70 Membrane with SiC powder filler 1.1 90 112 2 µm Figure 5-2 shows top view of membranes prepared by incorporating nanofiber fillers revealing a good coverage of supports with nanofibers. Figure 5-3 (A, B) shows membranes with powder and nanofiber fillers revealing approximately a thickness of 15 µm for these membranes. Figure 5-2 SEM top view picture of a SiC membrane with nanofiber fillers. 113 20 µm 50 µm A B Figure 5-3 SEM cross-section of SiC membranes A) with SiC powder fillers and B) with SiC nanofiber fillers. 114 Contrary to our expectation that highly porous fibers (~70%) could potentially create less resistance to the gas transport compared to nonporous SiC powders; as seen in Table 5-2 they are not contributing much in improving the H 2 permeance. Also, typically such membranes exhibit lower separation factors for H 2 /CH 4 . In order to investigate this effect, an equiweight mixture of nanofibers and AHPCS coating solution was pyrolyzed at the same condition membranes were fabricated. Figure 5-4 exhibits the adsorption isotherm of such coated nanofibers. BET analysis exhibits a surface area of 66 m 2 /gr and a total pore volume 0.1 cm 3 /gr for such fibers (compared to BET surface area of 457 m 2 /gr and a total pore volume of 0.79 cm 3 /gr for original nanofibers). Also it is seen increase of HK average pore size from 4.5 Å to 8.5 Å. Such drastic decrease in the accessible pore volume can partly explain why membranes permeances do not increase noticeably for porous fibrous fillers. It implies that after applying nanofibers into the membrane structure considerable amount of pore volume of nanofibers becomes inaccessible. This observation indicates how important the interaction between fillers and the membrane matrix is. On the other hand, one should also consider that powders are in spherical shape and at lower dimensions (dp=100-200 nm) than fibers (dp=250 nm and length of couple of microns) which results in lower packing degree for fibers compared to powders. This could explain lower separation factors for fibrious type fillers. In fact, better packing degree potentially could better cover the defects of underlying supports and lead to higher separation factors. 115 0 10 20 30 40 50 60 70 0 0.2 0.4 0.6 0.8 1 P/P o Volume Adsorbed (cm 3 /gr) STP BET surface area: 65 m 2 /gr BJH adsorption pore volume: 0.08 cm 3 /gr HK pore volume: 0.02 cm 3 /gr 0 0.1 0.2 0.3 0.4 0 0.2 0.4 0.6 0.8 1 1.2 Pore Diameter (nm) Pore Volume (cm 3 /g.nm) HK pore size distribution Figure 5-4 Adsorption isotherm of nanofibers coated by SiC pre-ceramic polymer (AHPCS) solution and pyrolyzed at 750 o C for 2 h. 116 4. Conclusions The effect of spherical impermeable SiC powder and highly porous SiC fiber incorporated into the SiC membrane structure were investigated on the membranes performance. Membranes prepared with fine SiC powder fillers showed similar performance as of those prepared by fibrous fillers. It is found that shape and size of fillers and also interaction of the filler and membrane matrix are of significant importance. Fibrous fillers are not as effective as of spherical particles in covering the supports defects. Interaction-wise, it is also observed highly porous nanofibers when incorporated into the membrane structure lost their porosity (accessible pore volume). Thus, we conclude filler interaction with membrane matrix and packing degree of fillers (as the result of shape of fillers) should be considered in designing membranes. 117 References: . Abraham, F., Debreuillegresse, M.F., Mairesse, G. and Nowogrocki, G., 1988. Phase- transitions and ionic-conductivity in Bi 4 V 2 O 11 an oxide with a layered structure. Solid State Ionics, 28: 529-532. Auerbach, S.M., Carrado, K.A. and Dutta, P.K., 2003. Handbook of Zeolite Science and Technology. Dekker, New York. Balachandran, U., Lee, T.H., Chen, L., Song, S.J., Picciolo, J.J. and Dorris, S.E., 2006. Hydrogen separation by dense cermet membranes. Fuel, 85(2): 150-155. Barsema, J.N., van der Vegt, N.F.A., Koops, G.H. and Wessling, M., 2005. Ag- functionalized carbon molecular-sieve membranes based on polyelectrolyte/polyimide blend precursors. Advanced Functional Materials, 15(1): 69-75. Bhandarkar, M., Shelekhin, A.B., Dixon, A.G. and Ma, Y.H., 1992. Adsorption, permeation, and diffusion of gases in microporous membranes .1. Adsorption of gases on microporous glass membranes. Journal of Membrane Science, 75(3): 221-231. Bhave, R.R., 1991. Inorganic Membranes Synthesis, Characteristics, and Applications. Van Nostrand Reinhold, New York. Boffa, V., Blank, D.H.A. and ten Elshof, J.E., 2008. Hydrothermal stability of microporous silica and niobia-silica membranes. Journal of Membrane Science, 319(1-2): 256-263. Boissiere, C., Martines, M.U., Larbot, A. and Prouzet, E., 2005. On the specific filtration mechanism of a mesoporous silica membrane, prepared with non-connecting parallel pores. Journal of Membrane Science, 251(1-2): 17-28. Bouwmeester, H.J.M., 2003. Dense ceramic membranes for methane conversion. Catalysis Today, 82(1-4): 141-150. Brandrup, J., Immergut, E.H. and Grulke, E.A., 1999. Polymer Handbook. Wiley, New York. Brinker, C.J., Sehgal, R., Hietala, S.L., Deshpande, R., Smith, D.M., Loy, D. and Ashley, C.S., 1994. Sol-gel strategies for controlled porosity inorganic materials. Journal of Membrane Science, 94: 85-102. 118 Burggraaf, A.J. and Cot, L., 1996. Fundamentals of Inorganic Membrane Science and Technology. Membrane science and technology series ; 4. Elsevier, Amsterdam ; New York. Buxbuam, R.E., 2001. Apparatus and methods for gas extraction. US Patent 6,183,543. Cameron, M.A., Gartland, I.P., Smith, J.A., Diaz, S.F. and George, S.M., 2000. Atomic layer deposition of SiO 2 and TiO 2 in alumina tubular membranes: Pore reduction and effect of surface species on gas transport. Langmuir, 16(19): 7435-7444. Caro, J. and Noack, M., 2008. Zeolite membranes – Recent developments and progress. Microporous and Mesoporous Materials, 115(3): 215-233. Caro, J., Noack, M. and Kolsch, P., 2005. Zeolite membranes: From the laboratory scale to technical applications. Adsorption Journal of the International Adsorption Society, 11(3-4): 215-227. Caro, J., Noack, M., Kolsch, P. and Schafer, R., 2000. Zeolite membranes - state of their development and perspective. Microporous and Mesoporous Materials, 38(1): 3-24. Casady, J.B. and Johnson, R.W., 1996. Status of silicon carbide (SiC) as a wide-bandgap semiconductor for high-temperature applications: A review. Solid State Electronics, 39(10): 1409-1422. Cejka, J., 2007. Introduction to Zeolite Science and Practice. Studies in Surface Science and Catalysis ; Elsevier, Amsterdam ; Boston. Chen, C.L. and Tsai, D.S., 2004. Permeation properties of microporous membranes prepared via coating of evaporated polydimethylsilane. Journal of Membrane Science, 237(1-2): 163-165. Chen, H.B. and Sholl, D.S., 2006. Predictions of selectivity and flux for CH 4 /H 2 separations using single walled carbon nanotubes as membranes. Journal of Membrane Science, 269(1-2): 152-160. Chen, J.T., Zhang, M.F. and Russell, T.P., 2007. Instabilities in nanoporous media. Nano Letters, 7(1): 183-187. Chen, S.H., Xing, P.F., Zhao, P.J. and Chen, W.M., 2002. A palladium alloy composite membrane for the purification of hydrogen isotopes. Separation Science and Technology, 37(11): 2701-2716. Cheng, Q.M., Interrante, L.V., Lienhard, M., Shen, Q.H. and Wu, Z.Z., 2005. Methylene- bridged carbosilanes and polycarbosilanes as precursors to silicon carbide-from ceramic 119 composites to SiC nanomaterials. Journal of the European Ceramic Society, 25(2-3): 233- 241. Ciora, R.J., Fayyaz, B., Liu, P.K.T., Suwanmethanond, V., Mallada, R., Sahimi, M. and Tsotsis, T.T., 2004. Preparation and reactive applications of nanoporous silicon carbide membranes. Chemical Engineering Science, 59(22-23): 4957-4965. Cundy, C.S., 1998. Microwave techniques in the synthesis and modification of zeolite catalysts. A review. Collection of Czechoslovak Chemical Communications, 63(11): 1699-1723. de la Iglesia, O., Irusta, S., Mallada, R., Menendez, M., Coronas, J. and Santamaria, J., 2006. Preparation and characterization of two-layered mordenite-ZSM-5 bi-functional membranes. Microporous and Mesoporous Materials, 93(1-3): 318-324. de Vos, R.M., Maier, W.F. and Verweij, H., 1999. Hydrophobic silica membranes for gas separation. Journal of Membrane Science, 158(1-2): 277-288. de Vos, R.M. and Verweij, H., 1998. High-selectivity, high-flux silica membranes for gas separation. Science, 279(5357): 1710-1711. Dolan, M.D., Dave, N.C., Ilyushechkin, A.Y., Morpeth, L.D. and McLennan, K.G., 2006. Composition and operation of hydrogen-selective amorphous alloy membranes. Journal of Membrane Science, 285(1-2): 30-55. Duke, M.C., da Costa, J.C.D., Do, D.D., Gray, P.G. and Lu, G.Q., 2006. Hydrothermally robust molecular sieve silica for wet gas separation. Advanced Functional Materials, 16(9): 1215-1220. Elyassi, B., Rajabbeigi, N., Khodadadi, A., Mohajerzadeh, S.S. and Sahimi, M., 2004. An yttria-doped ceria-based oxygen sensor with solid-state reference. Sensors and Actuators B: Chemical, 103(1-2): 178-183. Elyassi, B., Sahimi, M. and Tsotsis, T.T., 2007. Silicon carbide membranes for gas separation applications. Journal of Membrane Science, 288(1-2): 290-297. Elyassi, B., Sahimi, M. and Tsotsis, T.T., 2008. A novel sacrificial interlayer-based method for the preparation of silicon carbide membranes. Journal of Membrane Science, 316(1-2): 73-79. Emel'yanova, Y.V., Shafigina, R.R., Buyanova, E.S., Zhukovskii, V.M., Zainullina, V.M. and Petrova, S.A., 2006. Oxide ion conductors of the BIMEVOX family: Synthesis, structure, and conductivity. Russian Journal of Physical Chemistry, 80(11): 1725-1730. 120 Fergus, J.W., 2006. Electrolytes for solid oxide fuel cells. Journal of Power Sources, 162(1): 30-40. Gallis, S., Futschik, U., Castracane, J., Kaloyeros, A.E., Efstathiadis, H., Sherwood, W., Hayes, S. and Fountzoulas, C.G., 2003. Thermal chemical vapor deposition of silicon carbide films as protective coatings for microfluidic structures., Materials Research Society Symposium Proceedings, pp. 85-93. Gao, H.Y., Lin, Y.S., Li, Y.D. and Zhang, B.Q., 2004. Chemical stability and its improvement of palladium-based metallic membranes. Industrial & Engineering Chemistry Research, 43(22): 6920-6930. Gao, X.M., He, T.M. and Shen, Y., 2008. Structures, electrical and thermal expansion properties of Sr-doped La 2 Mo 2 O 9 oxide-ion conductors. Journal of Alloys and Compounds, 464(1-2): 461-466. Haugsrud, R., 2007. Defects and transport properties in Ln 6 WO 12 (Ln = La, Nd, Gd, Er). Solid State Ionics, 178(7-10): 555-560. Haugsrud, R., Larring, Y. and Norby, T., 2005. Proton conductivity of Ca-doped Tb 2 O 3 . Solid State Ionics, 176(39-40): 2957-2961. Haugsrud, R. and Norby, T., 2006. High-temperature proton conductivity in acceptor- doped LaNbO 4 . Solid State Ionics, 177(13-14): 1129-1135. Hedlund, J., Jareman, F., Bons, A.J. and Anthonis, M., 2003. A masking technique for high quality MFI membranes. Journal of Membrane Science, 222(1-2): 163-179. Hedlund, J., Noack, M., Kolsch, P., Creaser, D., Caro, J. and Sterte, J., 1999. ZSM-5 membranes synthesized without organic templates using a seeding technique. Journal of Membrane Science, 159(1-2): 263-273. Hedlund, J., Sterte, J., Anthonis, M., Bons, A.J., Carstensen, B., Corcoran, N., Cox, D., Deckman, H., De Gijnst, W., de Moor, P.P., Lai, F., McHenry, J., Mortier, W. and Reinoso, J., 2002. High-flux MFI membranes. Microporous and Mesoporous Materials, 52(3): 179-189. Hibino, T., Mizutani, K., Yajima, T. and Iwahara, H., 1992. Evaluation of proton conductivity in SrCeO 3 , BaCeO 3 , CaZrO 3 and SrZrO 3 by temperature programmed desorption method. Solid State Ionics, 57(3-4): 303-306. Hinds, B.J., Chopra, N., Rantell, T., Andrews, R., Gavalas, V. and Bachas, L.G., 2004. Aligned multiwalled carbon nanotube membranes. Science, 303(5654): 62-65. 121 Holt, J.K., Park, H.G., Wang, Y.M., Stadermann, M., Artyukhin, A.B., Grigoropoulos, C.P., Noy, A. and Bakajin, O., 2006. Fast mass transport through sub-2-nanometer carbon nanotubes. Science, 312(5776): 1034-1037. Hsieh, H.P., 1996. Inorganic Membranes for Separation and Reaction. Membrane science and technology series ; Elsevier, Amsterdam ; New York. Hsiung, T.H., Christman, D.D., Hunter, E.J. and Homyak, A.R., 1999. Methane formation on H 2 purification using a commercial Pd-Ag membrane. AIChE Journal, 45(1): 204-208. Huang, Y. and Dittmeyer, R., 2007. Preparation of thin palladium membranes on a porous support with rough surface. Journal of Membrane Science, 302(1-2): 160-170. Huber, K., Bantle, S., Lutz, P. and Burchard, W., 1985. Hydrodynamic and thermodynamic behavior of short-chain polystyrene in toluene and cyclohexane at 34.5 o C. Macromolecules, 18(7): 1461-1467. Illgena, U., Schäfera, R., Noacka, M., Kölscha, P., Kühnleb, A. and Caro, J., 2001. Membrane supported catalytic dehydrogenation of iso-butane using an MFI zeolite membrane reactor. Catalysis Communications, 2(11-12): 339-345. Interrante, L.V., Whitmarsh, C.W., Sherwood, W., Wu, H.-J., Lewis, R. and Maciel, G., 1994. High yield polycarbosilane precursors to stoichiometric SiC. Synthesis, pyrolysis and application, Mater. Res. Soc. Symp. Proc., pp. 593-603. Islam, M.S., Davies, R.A. and Gale, J.D., 2001. Hop, skip or jump? Proton transport in the CaZrO 3 perovskite oxide. Chemical Communications(7): 661-662. Iyoha, O., Enick, R., Killmeyer, R., Howard, B., Ciocco, M. and Morreale, B., 2007. H 2 production from simulated coal syngas containing H 2 S in multi-tubular Pd and 80 wt% Pd-20 wt% Cu membrane reactors at 1173 K. Journal of Membrane Science, 306(1-2): 103-115. Jiang, S., Yan, Y.S. and Gavalas, G.R., 1995. Temporary carbon barriers in the preparation of H 2 permselective silica membranes. Journal of Membrane Science, 103(3): 211-218. Jones, C.W. and Koros, W.J., 1994. Carbon molecular-sieve gas separation membranes .1. Preparation and characterization based on polyimide precursors. Carbon, 32(8): 1419- 1425. Jones, C.W. and Koros, W.J., 1995a. Carbon composite membranes - a solution to adverse humidity effects. Industrial & Engineering Chemistry Research, 34(1): 164-167. 122 Jones, C.W. and Koros, W.J., 1995b. Characterization of ultramicroporous carbon membranes with humidified Feeds. Industrial & Engineering Chemistry Research, 34(1): 158-163. Joshi, R.S., Nimat, R.K. and Pawar, S.H., 2009. Synthesis of fuel cell grade Bi 2 Co 0.1 V 0.9 O 5.35 solid electrolyte thin films. Journal of Alloys and Compounds, 471(1-2): 461-465. Julbe, A., Farrusseng, D. and Guizard, C., 2005. Limitations and potentials of oxygen transport dense and porous ceramic membranes for oxidation reactions. Catalysis Today, 104(2-4): 102-113. Kanezashi, M., O'Brien, J. and Lin, Y.S., 2006. Template-free synthesis of MFI-type zeolite membranes: Permeation characteristics and thermal stability improvement of membrane structure. Journal of Membrane Science, 286(1-2): 213-222. Kang, B.C., Lee, S.B. and Boo, J.H., 2004. Growth of beta-SiC nanowires on Si(100) substrates by MOCVD using nickel as a catalyst. Thin Solid Films, 464-65: 215-219. Katsaros, F.K., Steriotis, T.A., Stubos, A.K., Mitropoulos, A., Kanellopoulos, N.K. and Tennison, S., 1997. High pressure gas permeability of microporous carbon membranes. Microporous Materials, 8(3-4): 171-176. Keller, N., Pham-Huu, C., Ehret, G., Keller, V. and Ledoux, M.J., 2003. Synthesis and characterisation of medium surface area silicon carbide nanotubes. Carbon, 41(11): 2131- 2139. Kenawy, S.H. and Nour, W.M.N., 2005. Microstructural evaluation of thermally fatigued SiC-reinforced Al 2 O 3 /ZrO 2 matrix composites. Journal of Materials Science, 40(14): 3789-3793. Kharton, V.V., Marques, F.M.B. and Atkinson, A., 2004. Transport properties of solid oxide electrolyte ceramics: a brief review. Solid State Ionics, 174(1-4): 135-149. Kharton, V.V., Tsipis, E.V., Naumovich, E.N., Thursfield, A., Patrakeev, M.V., Kolotygin, V.A., Waerenborgh, J.C. and Metcalfe, I.S., 2008. Mixed conductivity, oxygen permeability and redox behavior of K 2 NiF 4 -type La 2 Ni 0.9 Fe 0.1 O 4+ δ . Journal of Solid State Chemistry, 181(6): 1425-1433. Kikuchi, E., 2000. Membrane reactor application to hydrogen production. Catalysis Today, 56(1-3): 97-101. 123 Kim, D.J., Scibioh, M.A., Kwak, S., Oh, I.H. and Ha, H.Y., 2004. Nano-silica layered composite membranes prepared by PECVD for direct methanol fuel cells. Electrochemistry Communications, 6(10): 1069-1074. Kim, T.W., Sahimi, M. and Tsotsis, T.T., 2008. Preparation of hydrotalcite thin films using an electrophoretic technique. Industrial & Engineering Chemistry Research, 47(23): 9127-9132. Kiyozumi, Y., Nagase, T., Hasegawa, Y. and Mizukami, F., 2008. A process for synthesizing bilayer zeolite membranes. Materials Letters, 62(3): 436-439. Kreuer, K.D., 1996. Proton conductivity: materials and applications. Chemistry of Materials, 8(3): 610-641. Kusakabe, K., Li, Z.Y., Maeda, H. and Morooka, S., 1995. Preparation of supported composite membrane by pyrolysis of polycarbosilane for gas separation at high- temperature. Journal of Membrane Science, 103(1-2): 175-180. L.V. Interrante, C.W.W., W. Sherwood, 1995. Fabrication of SiC matrix composites by liquid phase infiltration with a polymeric precursor, Mater. Res. Soc. Symp. Proc., pp. 139–146. L.V. Interrante, C.W.W., W. Sherwood, H.-J. Wu, R. Lewis and G. Maciel, 1994. High yield polycarbosilane precursors to stoichiometric SiC. Synthesis, pyrolysis and application, Mater. Res. Soc. Symp. Proc., 593–603. Lagorsse, S., Magalhaes, F.D. and Mendes, A., 2008. Aging study of carbon molecular sieve membranes. Journal of Membrane Science, 310(1-2): 494-502. Lai, Z.P., Bonilla, G., Diaz, I., Nery, J.G., Sujaoti, K., Amat, M.A., Kokkoli, E., Terasaki, O., Thompson, R.W., Tsapatsis, M. and Vlachos, D.G., 2003. Microstructural optimization of a zeolite membrane for organic vapor separation. Science, 300(5618): 456-460. Lee, L.L. and Tsai, D.S., 1999. A hydrogen-permselective silicon oxycarbide membrane derived from polydimethylsilane. Journal of the American Ceramic Society, 82(10): 2796-2800. Lee, L.L. and Tsai, D.S., 2001. Synthesis and permeation properties of silicon-carbon- based inorganic membrane for gas separation. Industrial & Engineering Chemistry Research, 40(2): 612-616. 124 Leonhardt, A., Liepack, H., Biedermann, K. and Thomas, J., 2005. Synthesis of SiC nanorods by chemical vapor deposition. Fullerenes Nanotubes and Carbon Nanostructures, 13: 91-97. Li, G., Su, X.H. and Lin, R.S., 2007. Preparation of highly water selective mordenite membranes via post-synthetic treatment with oxalic acid. Materials Letters, 61(23-24): 4576-4578. Li, J.Y. and Zhang, J.F., 2005. A simple purification for single-walled carbon nanotubes. Physica E-Low-Dimensional Systems & Nanostructures, 28(3): 309-312. Li, Y.S. and Yang, W.S., 2008. Microwave synthesis of zeolite membranes: A review. Journal of Membrane Science, 316(1-2): 3-17. Li, Z., Kusakabe, K. and Morooka, S., 1997a. Pore structure and permeance of amorphous Si-C-O membranes with high durability at elevated temperature. Separation Science and Technology, 32: 1233-1254. Li, Z.J., Zhang, J.L., Meng, A. and Guo, J.Z., 2006. Large-area highly-oriented SiC nanowire arrays: Synthesis, Raman, and photoluminescence properties. Journal of Physical Chemistry B, 110(45): 22382-22386. Li, Z.Y., Kusakabe, K. and Morooka, S., 1996. Preparation of thermostable amorphous Si-C-O membrane and its application to gas separation at elevated temperature. Journal of Membrane Science, 118(2): 159-168. Li, Z.Y., Kusakabe, K. and Morooka, S., 1997b. Pore structure and permeance of amorphous Si-C-O membranes with high durability at elevated temperature. Separation Science and Technology, 32(7): 1233-1254. Liang, K.C., Du, Y. and Nowick, A.S., 1994. Fast High-Temperature Proton Transport in Nonstoichiometric Mixed Perovskites. Solid State Ionics, 69(2): 117-120. Liu, H., Tan, X., Pang, Z., Diniz da Costa, J.C., Lu, G.Q. and Liu, S., 2008. Novel dual structured mixed conducting ceramic hollow fibre membranes. Separation and Purification Technology, 63(1): 243-247. Liu, M., Wang, R., Li, D.F. and Liang, D.T., 2007. A novel combustion route for the preparation of perovskite-type oxygen permeable materials. Materials Chemistry and Physics, 102(2-3): 132-139. Liu, Q.L., Wang, T.H., Liang, C.H., Zhang, B., Liu, S.L., Cao, Y.M. and Qiu, J.S., 2006a. Zeolite married to carbon: A new family of membrane materials with excellent gas separation performance. Chemistry of Materials, 18(26): 6283-6288. 125 Liu, Y.Y., Tan, X.Y. and Li, K., 2006b. Mixed conducting ceramics for catalytic membrane processing. Catalysis Reviews-Science and Engineering, 48(2): 145-198. Lu, G.Q. and Zhao, X.S., 2004. Nanoporous Materials: Science and Engineering. Series on chemical engineering ; v. 4. Imperial College Press, London. Lu, Y.P., Dixon, A.G., Moser, W.R., Ma, Y.H. and Balachandran, U., 2000. Oxygen- permeable dense membrane reactor for the oxidative coupling of methane. Journal of Membrane Science, 170(1): 27-34. Mallada, R. (Editor), 2008. Inorganic Membranes: Synthesis, Characterization and Applications. ELSEVIER, Amsterdam. Marcilla, A. and Beltran, M., 1995. Kinetic study of the thermal decomposition of polystyrene and polyethylene-vinyl acetate graft-copolymers by thermogravimetric analysis. Polymer Degradation and Stability, 50(1): 117-124. McCool, B.A. and DeSisto, W.J., 2004. Synthesis and characterization of silica membranes prepared by pyridine-catalyzed atomic layer deposition. Industrial & Engineering Chemistry Research, 43(10): 2478-2484. McLeary, E.E., Jansen, J.C. and Kapteijn, F., 2006. Zeolite based films, membranes and membrane reactors: Progress and prospects. Microporous and Mesoporous Materials, 90(1-3): 198-220. Mi, W.L., Lin, Y.S. and Li, Y.D., 2007. Vertically aligned carbon nanotube membranes on macroporous alumina supports. Journal of Membrane Science, 304(1-2): 1-7. Mordkovich, V.Z., Baichtock, Y.K. and Sosna, M.H., 1993. The large-scale production of hydrogen from gas mixtures - A use for ultra-thin palladium alloy membranes (Reprinted from Platinum Metal, Vol 36, Pg 90-97, 1992). International Journal of Hydrogen Energy, 18(7): 539-544. More, K.L., Tortorelli, P.F., Ferber, M.K. and Keiser, J.R., 2000. Observations of accelerated silicon carbide recession by oxidation at high water-vapor pressures. Journal of the American Ceramic Society, 83(1): 211-213. Morigami, Y., Kondo, M., Abe, J., Kita, H. and Okamoto, K., 2001. The first large-scale pervaporation plant using tubular-type module with zeolite NaA membrane. Separation and Purification Technology, 25(1-3): 251-260. Morooka, S., Kusakabe, K. and Kusuki, Y., 2000. Microporous carbon membranes. In: K. N.K. (Editor), Membrane Science and Technology, 323-334. 126 Mpourmpakis, G., Froudakis, G.E., Lithoxoos, G.P. and Samios, J., 2006. SiC nanotubes: A novel material for hydrogen storage. Nano Letters, 6(8): 1581-1583. Nagano, T., Sato, K., Saitoh, T. and Iwamoto, Y., 2006. Gas permeation properties of amorphous SiC membranes synthesized from polycarbosilane without oxygen-curing process. Journal of the Ceramic Society of Japan, 114(1330): 533-538. Nam, S.E. and Lee, K.H., 2001. Hydrogen separation by Pd alloy composite membranes: introduction of diffusion barrier. Journal of Membrane Science, 192(1-2): 177-185. Navajas, A., Mallada, R., Tellez, C., Coronas, J., Menendez, M. and Santamaria, J., 2006. The use of post-synthetic treatments to improve the pervaporation performance of mordenite membranes. Journal of Membrane Science, 270(1-2): 32-41. Neomagus, H.W.J.P., van Swaaij, W.P.M. and Versteeg, G.F., 1998. The catalytic oxidation of H2S in a stainless steel membrane reactor with separate feed of reactants. Journal of Membrane Science, 148(2): 147-160. Nhut, J.M., Vieira, R., Pesant, L., Tessonnier, J.P., Keller, N., Ehret, G., Pham-Huu, C. and Ledoux, M.J., 2002. Synthesis and catalytic uses of carbon and silicon carbide nanostructures. Catalysis Today, 76(1): 11-32. Noack, M., Kolsch, P., Schafer, R., Toussaint, P. and Caro, J., 2002. Molecular sieve membranes for industrial application: Problems, progress, solutions (Reprinted from Chem. Ing. Tech., vol 73, pg 958-967, 2001). Chemical Engineering & Technology, 25(3): 221-230. Norby, T., 2007. Ceramic proton and mixed proton-electron conductors in membranes for energy conversion applications. Journal of Chemical Engineering of Japan, 40(13): 1166- 1171. Nunes, S.P. and Peinemann, K.V., 2006. Membrane Technology in the Chemical Industry. Wiley-VCH, Weinheim, Germany. Ockwig, N.W. and Nenoff, T.M., 2007. Membranes for hydrogen separation. Chemical Reviews, 107(10): 4078-4110. Ohta, Y., Akamatsu, K., Sugawara, T., Nakao, A., Miyoshi, A. and Nakao, S., 2008. Development of pore size-controlled silica membranes for gas separation by chemical vapor deposition. Journal of Membrane Science, 315(1-2): 93-99. Oleynikov, N.N. and Ketsko, V.A., 2004. Design of oxygen-conducting membrane materials: A review. Russian Journal of Inorganic Chemistry, 49: S1-S21. 127 Opila, E.J., 1999. Variation of the oxidation rate of silicon carbide with water-vapor pressure. Journal of the American Ceramic Society, 82(3): 625-636. Osswald, S., Flahaut, E. and Gogotsi, Y., 2006. In situ Raman spectroscopy study of oxidation of double- and single-wall carbon nanotubes. Chemistry of Materials, 18(6): 1525-1533. Pages, X., Rouessac, V., Cot, D., Nabias, G. and Durand, J., 2001. Gas permeation of PECVD membranes inside alumina substrate tubes. Separation and Purification Technology, 25(1-3): 399-406. Paglieri, S.N. and Way, J.D., 2002. Innovations in palladium membrane research. Separation and Purification Methods, 31(1): 1-169. Park, D.H., Nishiyama, N., Egashira, Y. and Ueyama, K., 2001. Enhancement of hydrothermal stability and hydrophobicity of a silica MCM-48 membrane by silylation. Industrial & Engineering Chemistry Research, 40(26): 6105-6110. Park, H.B., Lee, S.Y. and Lee, Y.M., 2005. Pyrolytic carbon membranes containing silica: morphological approach on gas transport behavior. Journal of Molecular Structure, 739(1-3): 179-190. Park, T.J., Banerjee, S., Hemraj-Benny, T. and Wong, S.S., 2006. Purification strategies and purity visualization techniques for single-walled carbon nanotubes. Journal of Materials Chemistry, 16(2): 141-154. Pei, S.T., Chung, T.S., Kawi, S. and Guiver, M.D., 2004. Novel approaches to fabricate carbon molecular sieve membranes based on chemical modified and solvent treated polyimides. Microporous and Mesoporous Materials, 73(3): 151-160. Phair, J.W. and Donelson, R., 2006. Developments and design of novel (non-palladium- based) metal membranes for hydrogen separation. Industrial & Engineering Chemistry Research, 45(16): 5657-5674. Pham-Huu, C., Keller, N., Ehret, G. and Ledoux, M.J., 2001. The first preparation of silicon carbide nanotubes by shape memory synthesis and their catalytic potential. Journal of Catalysis, 200(2): 400-410. Plottu, A., Houssais, B., Democrate, C., Gatel, D. and Cavard, J., 2003. Autopsies of membranes fouled on Mery-sur-Oise pilot units: many lessons for the behaviour of the water treatment plant. Desalination, 157(1-3): 367-367. 128 Pol, V.G., Pol, S.V., Gedanken, A., Lim, S.H., Zhong, Z. and Lin, J., 2006. Thermal decomposition of commercial silicone oil to produce high yield high surface area SiC nanorods. Journal of Physical Chemistry B, 110(23): 11237-11240. Puerta, A.R., Remsen, E.E., Bradley, M.G., Sherwood, W. and Sneddon, L.G., 2003. Synthesis and ceramic conversion reactions of 9-BBN-modified allylhydridopolycarbosilane: A new single-source precursor to boron-modified silicon carbide. Chemistry of Materials, 15(2): 478-485. Rajabbeigi, N., Elyassi, B., Khodadadi, A., Mohajerzadeh, S.S. and Sahimi, M., 2004. A novel miniaturized oxygen sensor with solid-state ceria-zirconia reference. Sensors and Actuators B: Chemical, 100(1-2): 139-142. Rao, C.N.R., Müller, A. and Cheetham, A.K., 2004. The Chemistry of Nanomaterials : Synthesis, Properties and Applications in 2 volumes. Wiley-VCH ;John Wiley, Weinheim Chichester. Rohde, M.P., Schaub, G., Khajavi, S., Jansen, J.C. and Kapteijn, F., 2008. Fischer- Tropsch synthesis with in situ H 2 O removal - Directions of membrane development. Microporous and Mesoporous Materials, 115(1-2): 123-136. Rosenbloom, A.J., Sipe, D.M., Shishkin, Y., Ke, Y., Devaty, R.P. and Choyke, W.J., 2004. Nanoporous SiC: A candidate semi-permeable material for biomedical applications. Biomedical Microdevices, 6(4): 261-267. Rothenberger, K.S., Cugini, A.V., Howard, B.H., Killmeyer, R.P., Ciocco, M.V., Morreale, B.D., Enick, R.M., Bustamante, F., Mardilovich, I.P. and Ma, Y.H., 2004. High pressure hydrogen permeance of porous stainless steel coated with a thin palladium film via electroless plating. Journal of Membrane Science, 244(1-2): 55-68. Sahimi, M., 2003. Heterogeneous Materials. Springer, New York. Sammells, A.F. and Mundschau, M.V. (Editors), 2006. Nonporous Inorganic Membranes: for Chemical Processing. Wiley-VCH ; John Wiley [distributor], Weinheim Chichester. Sanchez Marcano, J.G. and Tsotsis, T.T., 2002. Catalytic Membranes and Membrane Reactors. Wiley-VCH, Weinheim. Sato, T., Kumagai, A. and Itoh, N., 2008. A catalytic ZSM-5 membrane sandwiched with silicalite-1 layers for highly selective toluene disproportionation, ICIM10, Tokyo, Japan. Saufi, S.M. and Ismail, A.F., 2004. Fabrication of carbon membranes for gas separation - a review. Carbon, 42(2): 241-259. 129 Schaefer, A.I., Fane, A.G. and Waite, T.D., 2005. Nanofiltration: Principles and Applications. Elsevier Advanced Technology, Oxford ; New York. Schuha, C.A., Hufnagel, T.C. and Ramamurty, U., 2007. Mechanical behavior of amorphous alloys. Acta Materialia, 55(12): 4067-4109. Schulz, K. and Durst, M., 1994. Advantages of an integrated system for hot gas filtration using rigid ceramic elements. Filtration & Separation, 31(1): 25-28. Sea, B.K., Ando, K., Kusakabe, K. and Morooka, S., 1998. Separation of hydrogen from steam using a SiC-based membrane formed by chemical vapor deposition of triisopropylsilane. Journal of Membrane Science, 146(1): 73-82. Sedigh, M.G., Jahangiri, M., Liu, P.K.T., Sahimi, M. and Tsotsis, T.T., 2000. Structural characterization of polyetherimide-based carbon molecular sieve membranes. AIChE Journal, 46(11): 2245-2255. Sedigh, M.G., Onstot, W.J., Xu, L.F., Peng, W.L., Tsotsis, T.T. and Sahimi, M., 1998. Experiments and simulation of transport and separation of gas mixtures in carbon molecular sieve membranes. Journal of Physical Chemistry A, 102(44): 8580-8589. Shinoda, Y., Nagano, T. and Wakai, F., 1999. Fabrication of nanograined silicon carbide by ultrahigh-pressure hot isostatic pressing. Journal of the American Ceramic Society, 82(3): 771-773. Shojai, F., Pettersson, A.B.A., Mantyla, T. and Rosenholm, J.B., 2000. Electrostatic and electrosteric stabilization of aqueous slips of 3Y-ZrO2 powder. Journal of the European Ceramic Society, 20(3): 277-283. Silenko, P.M., Shlapak, A.N., Tomila, T.V., Bykov, A.I., Kuz'menko, L.N., Okun, I.Y. and Ragulya, A.V., 2008. Fe-catalyzed synthesis of SiC nanofibers from methyltrichlorosilane. Inorganic Materials, 44(4): 388-394. Singh, R.P., Way, J.D. and Dec, S.F., 2005. Silane modified inorganic membranes: Effects of silane surface structure. Journal of Membrane Science, 259(1-2): 34-46. Snyder, M.A. and Tsapatsis, M., 2007. Hierarchical nanomanufacturing: From shaped zeolite nanoparticles to high-performance separation membranes. Angewandte Chemie- International Edition, 46(40): 7560-7573. Sommer, S. and Melin, T., 2005. Performance evaluation of microporous inorganic membranes in the dehydration of industrial solvents. Chemical Engineering and Processing, 44(10): 1138-1156. 130 Sørensen, T.S., 1999. Surface Chemistry and Electrochemistry of Membranes. Surfactant science series ; v. 79. M. Dekker, New York. Spetz, A.L., Tobias, P., Baranzahi, A., Martensson, P. and Lundstrom, I., 1999. Current status of silicon carbide based high-temperature gas sensors. IEEE Transactions on Electron Devices, 46(3): 561-566. Steele, B., 1998. Ceramic ion conducting membranes and their technological applications. Comptes Rendus De L Academie Des Sciences Serie Ii Fascicule C-Chimie, 1(9): 533- 543. Sterte, J., Hedlund, J., Creaser, D., Ohrman, O., Zheng, W., Lassinantti, M., Li, Q.H. and Jareman, F., 2001. Application of the seed-film method for the preparation of structured molecular sieve catalysts. Catalysis Today, 69(1-4): 323-329. Su, J.C. and Lua, A.C., 2007. Effects of carbonisation atmosphere on the structural characteristics and transport properties of carbon membranes prepared from Kapton ® polyimide. Journal of Membrane Science, 305(1-2): 263-270. Suda, H., Yamauchi, H., Uchimaru, Y., Fujiwara, I. and Haraya, K., 2006a. Preparation and gas permeation properties of silicon carbide-based inorganic membranes for hydrogen separation. Desalination, 193(1-3): 252-255. Suda, H., Yamauchi, H., Uchimaru, Y., Fujiwara, I. and Haraya, K., 2006b. Structural evolution during conversion of polycarbosilane precursor into silicon carbide-based microporous membranes. Journal of the Ceramic Society of Japan, 114(1330): 539-544. Suda, H., Yoda, S., Hasegawa, A., Tsuji, T., Otake, K. and Haraya, K., 2006c. Gas permeation properties of carbon molecular sieve membranes dispersed with palladium nano particles via supercritical CO 2 impregnation. Desalination, 193(1-3): 211-214. Sun, X.H., Li, C.P., Wong, W.K., Wong, N.B., Lee, C.S., Lee, S.T. and Teo, B.K., 2002. Formation of silicon carbide nanotubes and nanowires via reaction of silicon (from disproportionation of silicon monoxide) with carbon nanotubes. Journal of the American Chemical Society, 124(48): 14464-14471. Sutherland, K., 2004. Profile of the International Membrane Industry: Market Prospects to 2008. Elsevier, Oxford, UK ; New York, NY. Suwanmethanond, V., Goo, E., Liu, P.K.T., Johnston, G., Sahimi, M. and Tsotsis, T.T., 2000. Porous silicon carbide sintered substrates for high-temperature membranes. Industrial & Engineering Chemistry Research, 39(9): 3264-3271. 131 Taguchi, T., Igawa, N., Yamamoto, H. and Jitsukawa, S., 2005. Synthesis of silicon carbide nanotubes. Journal of the American Ceramic Society, 88(2): 459-461. Takeda, Y., Nakamura, K., Maeda, K. and Matsushita, Y., 1987. Effects of elemental additives on electrical resistivity of silicon carbide ceramics. Journal of the American Ceramic Society, 70(10): C266-C267. Takeda, Y., Shibata, N. and Kubo, Y., 2001. SiC coating on porous gamma-Al 2 O 3 using alternative-supply CVI method. Journal of the Ceramic Society of Japan, 109(4): 305-309. Tao, S.W. and Irvine, J.T.S., 2002. Structure and properties of nonstoichiometric mixed perovskites A 3 B ' 1+x B'' 2-x O 9- δ . Solid State Ionics, 154: 659-667. Tejuca, L.G. and Fierro, J.L.G., 1993. Properties and Applications of Perovskite-Type Oxides. Chemical industries ; Marcel Dekker, New York. Tin, P.S., Chung, T.S., Jiang, L.Y. and Kulprathipanja, S., 2005. Carbon-zeolite composite membranes for gas separation. Carbon, 43(9): 2025-2027. Titiloye, J.O. and Islam, H., 2008. Synthesis and characterization of silicalite-1/carbon- graphite membranes. Journal of colloid and interface science 318: 50-58. Tompsett, G.A., Conner, W.C. and Yngvesson, K.S., 2006. Microwave synthesis of nanoporous materials. Chemphyschem, 7(2): 296-319. Tong, J.H., Matsumura, Y., Suda, H. and Haraya, K., 2005. Thin and dense Pd/CeO 2 /MPSS composite membrane for hydrogen separation and steam reforming of methane. Separation and Purification Technology, 46(1-2): 1-10. Tsui, B.-Y. and Fang, K.-L., 2005. A novel wafer reclaim method for amorphous SiC and carbon coped oxide films. IEEE Trans. Semicond. Manuf., 18: 716-721. Tsuru, T., 2001. Inorganic porous membranes for liquid phase separation. Separation and Purification Methods, 30(2): 191-220. Tsuru, T., Morita, T., Shintani, H., Yoshioka, T. and Asaeda, M., 2008. Membrane reactor performance of steam reforming of methane using hydrogen-permselective catalytic SiO 2 membranes. Journal of Membrane Science, 316(1-2): 53-62. Uemiya, S., 1999. State-of-the-art of supported metal membranes for gas separation. Separation and Purification Methods, 28(1): 51-85. Vannice, M.A., Chao, Y.L. and Friedman, R.M., 1986. The preparation and use of high surface-area silicon-carbide catalyst supports. Applied Catalysis, 20(1-2): 91-107. 132 Vannier, R.N., Pernot, E., Anne, M., Isnard, O., Nowogrocki, G. and Mairesse, G., 2003. Bi 4 V 2 O 11 polymorph crystal structures related to their electrical properties. Solid State Ionics, 157(1-4): 147-153. Vermisogiou, E.C., Pilatos, G., Romanos, G.E., Karanikolos, G.N., Boukos, N., Mertis, K., Kakizis, N. and Kanellopoulos, N.K., 2008. Synthesis and characterisation of carbon nanotube modified anodised alumina membranes. Microporous and Mesoporous Materials, 110(1): 25-36. Wach, R.A., Sugimoto, M., Idesaki, A. and Yoshikawa, M., 2007a. 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, 140(1-2): 81-89. Wach, R.A., Sugimoto, M. and Yoshikawa, M., 2007b. Formation of silicone carbide membrane by radiation curing of polycarbosilane and polyvinylsilane and its gas separation up to 250 degrees C. Journal of the American Ceramic Society, 90(1): 275-278. Wei, Q., Wang, Y.L., Nie, Z.R., Yu, C.X., Li, Q.Y., Zou, J.X. and Li, C.J., 2008. Facile synthesis of hydrophobic microporous silica membranes and their resistance to humid atmosphere. Microporous and Mesoporous Materials, 111(1-3): 97-103. Weyd, M., Richter, H., Puhlfurss, P., Voigt, I., Hamel, C. and Seidel-Morgenstern, A., 2008. Transport of binary water-ethanol mixtures through a multilayer hydrophobic zeolite membrane. Journal of Membrane Science, 307(2): 239-248. Wincewicz, K.C. and Cooper, J.S., 2005. Taxonomies of SOFC material and manufacturing alternatives. Journal of Power Sources, 140(2): 280-296. Wu, J.C.S., Sabol, H., Smith, G.W., Flowers, D.L. and Liu, P.K.T., 1994. Characterization of hydrogen-permselective microporous ceramic membranes. Journal of Membrane Science, 96(3): 275-287. Wu, R.B.B., Yang, G.Y., Pan, Y. and Chen, J.J., 2007. Synthesis of silicon carbide nanorods without defects by direct heating method. Journal of Materials Science, 42(11): 3800-3804. Xie, Z.F., Tao, D.L. and Wang, J.Q., 2007. Synthesis of silicon carbide nanotubes by chemical vapor deposition. Journal of Nanoscience and Nanotechnology, 7(2): 647-652. Xomeritakis, G., Braunbarth, C.M., Smarsly, B., Liu, N., Kohn, R., Klipowicz, Z. and Brinker, C.J., 2003. Aerosol-assisted deposition of surfactant-templated mesoporous silica membranes on porous ceramic supports. Microporous and Mesoporous Materials, 66(1): 91-101. 133 Xomeritakis, G., Lai, Z.P. and Tsapatsis, M., 2001. Separation of xylene isomer vapors with oriented MFI membranes made by seeded growth. Industrial & Engineering Chemistry Research, 40(2): 544-552. Xu, D.Y., He, Z.W., Guo, Y.P. and Wang, Y.Y., 2006. Fabrication of quasi-one dimension silicon carbide nanorods prepared by RF sputtering. Microelectronic Engineering, 83(1): 89-91. Yamamoto, O., 2000. Solid oxide fuel cells: fundamental aspects and prospects. Electrochimica Acta, 45(15-16): 2423-2435. Yampol'skii, Y.P., Pinnau, I. and Freeman, B.D., 2006. Materials Science of Membranes for Gas and Vapor Separation. Wiley, Chichester, England ; Hoboken, NJ. Yan, Y.S., Davis, M.E. and Gavalas, G.R., 1997a. Preparation of highly selective zeolite ZSM-5 membranes by a post-synthetic coking treatment. Journal of Membrane Science, 123(1): 95-103. Yan, Y.S., Davis, M.E. and Gavalas, G.R., 1997b. Use of diffusion barriers in the preparation of supported zeolite ZSM-5 membranes. Journal of Membrane Science, 126(1): 53-65. Yasuda, I. and Shirasaki, Y., 2006. Development and demonstration of membrane reformer system for highly-efficient hydrogen production from natural gas. Materials Science Forum, 539-543: 1403-1408. Yasuda, I., Shirasaki, Y., Kobayashi, K., Nakajima, H., Shimamori, T., Sasaki, T., Kabutomori, T., Uchida, M., Yoshizawa, H., Nishida, M. and Shinpo, Y., 2006. Development of membrane technology for highly-efficient hydrogen production, 16th World Hydrogen Energy Conference. Yen, H.M., Jou, S. and Chu, C.J., 2005. Si-O-C nanotubes from pyrolyzing polycarbosilane in a mesoporous template. Materials Science and Engineering B-Solid State Materials for Advanced Technology, 122(3): 240-245. Yepes, D., Cornaglia, L.M., Irusta, S. and Lombardo, E.A., 2006. Different oxides used as diffusion barriers in composite hydrogen permeable membranes. Journal of Membrane Science, 274(1-2): 92-101. Zawrah, M.F. and Shaw, L., 2004. Liquid-phase sintering of SiC in presence of CaO. Ceramics International, 30(5): 721-725. Zhang, X.F., Liu, H. and Yeung, K.L., 2005. Novel two-layered zeolite NaA-silicalite-1 membranes. Journal of Physics and Chemistry of Solids, 66(6): 1034-1038. 134 Zhao, H.B., Pflanz, K., Gu, J.H., Li, A.W., Stroh, N., Brunner, H. and Xiong, G.X., 1998. Preparation of palladium composite membranes by modified electroless plating procedure. Journal of Membrane Science, 142(2): 147-157. Zheng, J., Kramer, M.J. and Akinc, M., 2000. In situ growth of SiC whisker in pyrolyzed monolithic mixture of AHPCS and SiC. Journal of the American Ceramic Society, 83(12): 2961-2966. Zhou, Z.H., Yang, H.H., Zhang, Y., Chang, L.F., Sun, W.G. and Wang, J.Q., 2007. NaA zeolite/carbon nanocomposite thin films with high permeance for CO 2 /N 2 separation. Separation and Purification Technology, 55(3): 392-395. Zorman, C.A., Fleischman, A.J., Dewa, A.S., Mehregany, M., Jacob, C., Nishino, S. and Pirouz, P., 1995. Epitaxial- growth of 3C-Sic films on 4 inch diam (100) silicon-wafers by atmospheric-pressure chemical vapor deposition. Journal of Applied Physics, 78(8): 5136-5138.
Abstract (if available)
Abstract
Silicon carbide (SiC) microporous membranes were prepared by the pyrolysis of thin allyl-hydridopolycarbosilane (AHPCS) films coated, using a combination of slip-casting and dip-coating techniques, on tubular SiC macroporous supports. Combining slip-casting with dip-coating significantly improved the reproducibility in preparing high quality membranes. The membranes prepared exhibited an ideal H2/CO2 selectivity in the range of (42-96), and a H2/CH4 ideal selectivity in the range of (29-78). Separation factors measured with the same membranes, using equimolar binary mixtures of H2 in CO2 and H2 in CH4, were similar to the ideal selectivity values. Steam stability experiments with the membranes lasting 21 days, using an equimolar flowing mixture of He/H2O at 200 oC, indicated some initial decline in the permeance of He, after which the permeance became stable at these conditions.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Silicon carbide ceramic membranes
PDF
Fabrication of silicon carbide sintered supports and silicon carbide membranes
PDF
Molecular modeling of silicon carbide nanoporous membranes and transport and adsorption of gaseous mixtures therein
PDF
Development of carbon molecular-sieve membranes with tunable properties: modification of the pore size and surface affinity
PDF
Continuum and pore netwok modeling of preparation of silicon-carbide membranes by chemical-vapor deposition and chemical-vapor infiltration
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
A study of the application of membrane-based reactive separation to the carbon dioxide methanation
PDF
An integrated 'one-box' process for hydrogen production
PDF
The use of carbon molecule sieve and Pd membranes for conventional and reactive applications
PDF
A hybrid adsorbent-membrane reactor (HAMR) system for hydrogen production
PDF
Biogas reforming: conventional and reactive separation processes and the preparation and characterization of related materials
PDF
On the use of membrane reactors in biomass utilization
PDF
Fabrication of nanoporous silicon oxycarbide materials via a sacrificial template technique
PDF
Hydrogen storage in carbon and silicon carbide nanotubes
PDF
Exploring properties of silicon-carbide nanotubes and their composites with polymers
PDF
Studies of transport phenomena in hydrotalcite membranes, and their use in direct methanol fuel cells
PDF
Methanol synthesis in a membrane reactor
PDF
Process intensification in hydrogen production via membrane-based reactive separations
PDF
Preparation of polyetherimide nanoparticles by electrospray drying, and their use in the preparation of mixed-matrix carbon molecular-sieve (CMS) membranes
Asset Metadata
Creator
Elyassi, Bahman
(author)
Core Title
Fabrication of nanoporous silicon carbide membranes for gas separation applications
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Chemical Engineering
Publication Date
07/28/2009
Defense Date
06/06/2009
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
AHPCS,gas separation,high temperature,hydrogen,membrane,OAI-PMH Harvest,silicon carbide
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Tsotsis, Theodore T. (
committee chair
), Pirbazari, Massoud M. (
committee member
), Sahimi, Muhammad (
committee member
)
Creator Email
bahmanelyassi@yahoo.com,elyassi@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m2394
Unique identifier
UC1294352
Identifier
etd-Elyassi-3099 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-561549 (legacy record id),usctheses-m2394 (legacy record id)
Legacy Identifier
etd-Elyassi-3099.pdf
Dmrecord
561549
Document Type
Dissertation
Rights
Elyassi, Bahman
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
AHPCS
gas separation
high temperature
hydrogen
membrane
silicon carbide