This invention is related to the inventions described in co-pending U.S. patent application Ser. No. 09/822,927, filed Mar. 30, 2001 and Ser. No. 09/952,186, filed Sep. 13, 2001.
1. Field of the Invention
The present invention relates generally to gas separation membranes and, more particularly, to a metal membrane for the separation of hydrogen gas from a gas stream.
2. Related Art
Membranes and membrane modules for the separation of hydrogen from other gases are generally known. In particular, useful membranes for hydrogen separation typically can be categorized as being of four general types: (i) polymeric, (ii) porous inorganic, (iii) self-supporting non-porous metal, and (iv) non-porous metal supported on a porous rigid matrix such as metal or ceramic.
Polymeric membranes are commonly used in the form of extended flat sheets of small diameter hollow fibers. Flat sheet polymeric membranes are most often incorporated into spiral-wound modules. Hollow fiber membranes are incorporated into hollow fiber modules, which are very similar in design to shell-and-tube heat exchangers.
Polymeric membranes and membrane modules for hydrogen separation suffer from a lack of high selectivity toward hydrogen over other gases, which results in a relatively impure product gas. Such membranes also suffer from a lack of stability at operating temperatures above 480° F. (250° C.) and chemical incompatibility with many chemicals such as hydrocarbons that are present in an impure hydrogen feed stream.
Porous inorganic-based membranes are typically fabricated from titania, zirconia, alumina, glass, molecular sieving, carbon, silica and or zeolites. All are fabricated with a narrow pore-sized distribution, with the porous inorganic membranes exhibiting high hydrogen permeability, but low selectivity due to relatively large mean pore diameters. Such materials are brittle and thus susceptible to failure due to cracking, and the sealing and fixturing of such porous inorganic-based membranes limit their use to relatively low temperature applications.
Development of supported metal membranes has focused on the utilization of ceramic tubes coated with a thin film of foil of non-porous or dense palladium (Pd) or palladium alloys. The ceramic support tube typically is of a graded porosity from one surface thereof to a second opposite surface. More specifically, the porosity of the ceramic support tube typically is densest at the surface upon which the palladium or palladium alloy is disposed, and the porosity of the tube increases from this surface to a maximum porosity on the surface opposite the layer of palladium. The layer of palladium or palladium alloy is selectively permeable to hydrogen gas and is typically capable of withstanding temperatures of 1500–1600° F. (815–870° C.).
Such ceramic-supported metal membranes are typically housed in shell and tube modules and are fitted with compression gaskets to seal the membrane tube into the module to prevent leakage of the feed gas stream into the permeate gas stream. Potential leak paths between the feed and permeate gas streams can exist due to differences in the coefficients of thermal expansion of the ceramic tube and the metal compression fittings. Additionally, the ceramic support-tubes are inherently brittle and can experience long term thermal fatigue due to repetitive process or system startup and shutdown cycles.
The mechanical adherence of the thin palladium or palladium alloy layer upon the surface of the ceramic support tube requires secure attachment of the film or foil onto the surface of the ceramic, as well as the absence of pinholes or other mechanical rupturing that can occur during manufacture or use of the ceramic tube membrane.
For porous metal membranes such as porous stainless steel and microporous noble metals, Knudsen diffusion or combined Knudsen diffusion-surface diffusion are the primary mechanisms by which gas transport occurs across the membrane. For dense metal membranes such as palladium or palladium alloy foil or film, however, the primary mechanism of gas transport through the metal layer is traditional chemisorption-dissociation-diffusion. Broadly stated, chemisorption-dissociation-diffusion transport involves chemisorption of hydrogen molecules onto the membrane surface, dissociation of hydrogen into atomic hydrogen, transportation of atomic hydrogen through the dense metal, reassociation of atomic hydrogen into hydrogen molecules, and desorption of hydrogen molecules from the media. While Knudsen diffusion typically offers greater flow rates across a membrane than chemisorption-dissociation-diffusion, Knudsen diffusion suffers from reduced hydrogen selectivity as compared with chemisorption-dissociation-diffusion. It is also known that the interaction of a gas stream with catalytic materials can increase the concentration of hydrogen within the reactant or process gas stream. Such catalytic materials enhance the water/gas-shift reaction whereby carbon monoxide is reacted with water to form carbon dioxide and hydrogen gas. Catalytic materials also promote the decomposition of ammonia, which also increases the concentration of hydrogen.
Examples of such catalytic materials include platinum (Pt), palladium (Pd), rhodium (Rh) and the like. While it has been known to apply such catalytic materials to ceramic support substrates to form composite membranes, such composite membranes still suffer from the aforementioned problems associated with the application of palladium and palladium alloy foils and films to ceramic support tubes.
Application Ser. No. 09/822,927, filed Mar. 30, 2001, advances the art in the foregoing respect by providing a porous graded metal substrate on which the palladium or palladium alloy could be mounted either directly or through the interface of a ceramic washcoat. While this teaching significantly advances the art, there is a further need for improving the bond between the substrate and the palladium layer for very high temperature applications such as are found in integrated gasification combined cycle (IGCC), pressurized-fluidized bed combustion (PFBC) or pressurized-circulating fluidized bed combustion (PCFBC) applications. In addition, to promote hydrogen separation, and reduce the cost and use of palladium, the inclusion and application of a mixed metal alloy layer is identified.
An aspect of the present invention is to provide a gas separation membrane for separating hydrogen from a gas stream that is capable of operating at very high temperatures in an integrated gasification combined cycle or in other pressurized fluidized bed combustion or pressurized circulating fluidized bed combustion applications for extended periods of time. This aspect is achieved by employing a gas separation membrane having a support structure made up of a porous fibrous or particle metal media base layer, preferably followed by a layer or series of metal media layers that contain a higher fiber or particle density than the underlying metal support substrate. A barrier layer composed of metal particles is applied to the surface of the porous metal media substrate. The metal particles in the barrier layer, which include aluminum, are bonded together employing a chemical bond phase. When fired to provide strength, a portion of the aluminum is oxidized, forming alumina which combines with iron or nickel in the substrate metals, establishing an iron aluminide or nickel aluminide spinel within the barrier layer. Physical/mechanical interlocking of the metal/ceramic barrier layer with the underlying fibrous or particle metal support substrate results, providing improved adherence of the coating. The thickness of the barrier layer is less than 100 microns and preferably between 20 and 50 microns in thickness. The barrier layer is followed by a dense palladium chemisorption, dissociation, transport, reassociation layer.
In an alternate preferred embodiment, a tantalum/niobium structural layer is interposed between the palladium layer and the barrier layer, which allows for a reduced thickness of the palladium layer and lower's the cost of the membrane.
A further understanding of the invention can be gained from the following description of the preferred embodiments when read in conjunction with the accompanying drawings in which:
This invention is an improvement to the porous metal membranes described in patent application Ser. No. 09/822,927, filed Mar. 30, 2001, one embodiment of which is illustrated in
The metal gas separation membrane 4 includes a transmission member 6 that can be of numerous shapes, such as plates, tubes, honeycomb configurations and other such shapes. The transmission member 6 is depicted schematically in cross-section in
The metal particles 12 that make up the porous body 10 can be of numerous physical configurations such as metal fibers, metal powder and other shapes, and the porous body 10 can be made up of one type of particle or combinations of these different shaped metal particles 12. The metal particles 12 can be fabricated from known superalloys and/or intermetallic materials to permit the porous body 10 to withstand high temperatures, meaning not only that the material does not melt at the elevated temperatures, but also that the material resists corrosion in the potentially oxidizing or reducing environment in which the membrane 4 is used. It is understood, however, that other materials suited to different applications can be employed to manufacture the metal particles 12 without departing from the concept of the present invention.
The metal particles 12 are compacted and sinter-bonded to form a thin, dense layer of material that makes up the porous body 10. The resultant porous body 10 has a thickness that is generally in the range of approximately 100 microns to 5 mm, although other thicknesses may be appropriate depending upon the specific needs of the particular application. The aforementioned “thickness” thus refers generally to the distance between the first and second surfaces 20 and 24.
As can be seen in
As will be set forth more fully below, the porous body 10 is advantageously configured to provide relatively high hydrogen selectivity at the first surface 20 due to its relatively low porosity. The regions of the porous body extending from the first surface 20 to the second surface 24, being relatively more porous than the first surface 20, have a relatively lower hydrogen selectivity, but correspondingly permit the relatively free flow of gases therethrough. These relatively porous regions of the thicker porous body 10 can thus be said to provide support to the low porosity region of the porous body 10 at the surface 20, which generally is relatively thin and of low strength, without meaningfully impeding the flow of hydrogen therethrough from the second surface 24 to the first surface 20. Preferably, the thicker to thinner regions are continuously graded. It can be seen that the graded porosity feature of the porous body 10 advantageously provides a low porosity region (at the first surface 20) that has high hydrogen selectivity and is relatively thin to promote high hydrogen permeability therethrough, with this low porosity region being supported by a relatively high porosity, thicker region of the porous body 10, which does not impede the flow of hydrogen therethrough. As such, the graded porosity feature of the porous body 10 itself provides both high hydrogen selectivity and high hydrogen flow rates therethrough without the high pressure drop that would be experienced by the hydrogen if the porosity of the porous body 10 was that of the first surface 20 throughout the porous body 10.
As shown in
The metal coating 16 is a dense layer or foil of precious metals such palladium, palladium alloys, and the like that enhance gas phase chemisorption-diffusion of hydrogen therethrough. As is known in the relative art, the expression gas phase “chemisorption-dissociation-diffusion” of hydrogen refers to molecular chemisorption and dissociation of hydrogen along the high pressure side of the metal coating 16, proton and electron diffusion through the lattice of the metal coating 16, and proton and electron re-association and recombination and desorption of molecular species along the opposite side of the metal coating 16. The metal coating 16 thus can be referred to as a chemisorption-dissociation-diffusion coating. The metal coating 16 can be applied to the porous body 10 in any of a variety of fashions such as via electrolysis plating, electroplating, sputtering, electrodeposition, and the like, or other appropriate method if the coating is non-metallic.
In the embodiment depicted in
Hydrogen that passes through porous body 10 and reaches the first surface 20 is then permitted to flow through the metal coating 16 primarily by chemisorption-dissociation-diffusion transport. Hydrogen that has been transported fully through the metal coating 16 forms the permeate stream 26.
As is known in the relevant art, as the thickness of the metal coating 16 increases, the flow rate of hydrogen therethrough by chemisorption-dissociation-diffusion transport decreases. The metal coating 16 is thus preferably configured to be as thin as possible to enhance the flow rate of hydrogen therethrough while maintaining the mechanical integrity of the metal coating 16. The thickness of the metal coating 16 in the prior art applications is thus preferably in approximately the range of 0.1 to 10 microns, although other thicknesses may be appropriate depending upon the specific needs of the particular application.
It can thus be seen that the porous body 10 serves to mechanically support the metal coating 16 thereon. In this regard, it can be seen that the metal particles 12 provide numerous points of contact between the porous body 10 and the metal coating 16, which helps the metal coating 16 to adhere on the porous body 10 during operation of the metal gas separation membrane 4 and during the thermal expansion and contraction of the metal gas separation membrane 4 during startup and shutdown operations. Further, in this regard, the metal particles 12 and metal coating 16 preferably are configured to have compatible coefficients of thermal expansion in order to limit mechanical stresses between the porous body 10 and the metal coating 16, although this is not strictly necessary to achieve the beneficial aspects of the present invention.
As indicated herein before, the metal coating 16 is depicted in
From the foregoing, it can be seen that the metal gas separation membrane, in its simplest form, could comprise solely a porous body being manufactured out of metal particles or fibers and having a constant density and porosity throughout, with the porous body serving as the transmission member that separates hydrogen from a gas stream. In addition thereto, the porous body may be of a graded density to reduce the pressure drop across the porous body 10 and/or can additionally include the metal coating 16 to further increase hydrogen selectivity. As such, the configuration of the metal gas separation membrane 4 can be varied to have differing characteristics such as hydrogen selectivity hydrogen flow rate, suitability to high temperatures, and the like, as well as other characteristics, that are suited to the particular environment in which the metal gas separation membrane 4 is used by varying one or more of the physical characteristics of the transmission member 6.
The metal gas separation membrane 4 thus provides a device that can be specifically configured to have desirable characteristics as to hydrogen selectivity, hydrogen flow rate, and suitability for high temperatures, as well as other characteristics. The metal gas separation membrane 4 overcomes many of the problems associated with previous gas separation asymmetric ceramic membranes by providing relatively close correspondence in the coefficients of thermal expansion of the porous body 10 and the metal coating 16, and by promoting the adherence of the metal coating 16 onto the porous body 10. The metal gas separation membrane 4 thus provides increased functionality and versatility.
The further improvement of this invention over U.S. patent application Ser. No. 09/822,927 is disclosed in
The improvements to the high temperature, hydrogen gas separation membrane disclosed herein include a physically interlocking barrier layer 28 formed on the metal substrate 10 in a manner to create a spinel or intermetallic architecture bonding of the metal barrier layer 28 onto the surface of the fibrous, metal media, structural support substrate 10. Both physical/mechanical interlocking and spinel or intermetallic formation improve adherence between the fibrous metal substrate and barrier coating during use of the metal media membrane in high temperature, gas separation applications.
Additionally, in the preferred embodiment, a tantalum/niobium layer is formed between the outer transport layer 32 and the barrier layer 28 to facilitate enhanced hydrogen flux through the proton transport layer by enabling the formation of a thinner transport layer made out of a precious metal such as palladium, palladium alloys, and the like and serves to initiate catalysis/chemisorption and dissociation of hydrogen in the process gas stream. The use of the tantalum/niobium interface additionally reduces the thickness of the proton transport layer 32, and thus the reduced cost of the gas separation membrane and improves its efficiency.
As a result of the application of the barrier layer 28, the porosity of the matrix is reduced, but still permits passage of air through the structure. Part of the uniqueness of applying a barrier layer 28 onto the surface of the fibrous metal substrate 10, is the capability of the barrier layer to form spinel phases with the underlying fibers, forming a more adherent bond between the adjoining layers. Experimental results have shown that a nickel-aluminum spinel phase/interface region results when an “Alseal-500” series coating, available from Coatings for Industry, Souderton, Pa., is applied to the surface of a dense nickel-based Haynes® 214 catalytic combustion substrate tube. The Haynes® 214 substrate alloy is available from Haynes International, Inc., Kokomo, Ind. It is known from the literature that thin metal wire (tungsten or phoretic stainless steel) can be aluminized via application of metallic aluminum onto the surface of the wire. When heated, the outermost layer of aluminum is converted to alumina, and encourages diffusion of the inner layer of aluminum into the body of the metal wire, forming a thin metal/aluminum alloy layer.
The outer surface of the barrier layer 28 is the surface onto which the precious metal 32, such as palladium, can be deposited. Palladium serves to chemisorb hydrogen molecules, with subsequent dissociation, proton transport through the thickness of the palladium coating and finally, reassociation of hydrogen in the permeate gas stream. The thickness of the applied dense palladium layer 32 should be less than 100 microns and, more preferably due to the cost of the noble metal, less than 15 microns thick. To further reduce the thickness of the palladium layer, a tantalum/niobium (Ta/Nb) structural layer 30 is applied onto the barrier layer 28. The Ta/Nb layer should be understood throughout this application to comprise either Ta or Nb or a combination of both elements. The final palladium layer 32 is used in the disclosed hydrogen gas separation concept of this invention to serve as the catalytic activator, initiating chemisorption and dissociation of gas phase hydrogen. Limited mixing of the Pd and Ta/Nb can be expected to result in a very narrow diffusion zone.
It is known that onsite purification of hydrogen using a non-porous metal membrane can be used to effect the separation of ultra-pure hydrogen. Specifically, the techniques which have been used involve the use of vanadium or tantalum foil coated along both sides with palladium. In operation, molecular hydrogen (H2) dissociates into monatomic hydrogen (H) on palladium and diffuses through it. Diffusion then carries the monatomic hydrogen through the vanadium and the second palladium layer, where it recombines into molecular hydrogen. With a higher solubility of hydrogen in (flux through) the Ta/Nb layer in comparison to Pd and the ease at which Ta/Nb can be deposited as a dense coating in comparison to Pd, an effective and reduced overall cost of the external hydrogen or proton transport coating can be realized.
While specific embodiments of the invention have been described in detail, it will be appreciated by those skilled in the art that various modifications and alternatives to those details could be developed in light of the overall teachings of the disclosure. Accordingly, the particular embodiments disclosed are meant to be illustrative only and not limiting as to the scope of the invention, which is to be given the full breadth of the appended claims and any and all equivalents thereof.
Number | Name | Date | Kind |
---|---|---|---|
2824620 | DeRosset | Feb 1958 | A |
2848620 | Backus | Aug 1958 | A |
2958391 | DeRosset | Nov 1960 | A |
3022187 | Eyraud et al. | Feb 1962 | A |
3208198 | Rubin | Sep 1965 | A |
3241298 | Pierce | Mar 1966 | A |
3245206 | Bonnet | Apr 1966 | A |
3251173 | Ehlers et al. | May 1966 | A |
3392510 | Koch, Jr. | Jul 1968 | A |
3413777 | Langley et al. | Dec 1968 | A |
3477288 | Krcal et al. | Nov 1969 | A |
3713270 | Farr et al. | Jan 1973 | A |
4039703 | Kamijo et al. | Aug 1977 | A |
4132668 | Gryaznov et al. | Jan 1979 | A |
4472175 | Malon et al. | Sep 1984 | A |
4496373 | Behr et al. | Jan 1985 | A |
4589891 | Iniotakis et al. | May 1986 | A |
4689150 | Abe et al. | Aug 1987 | A |
4699637 | Iniotakis et al. | Oct 1987 | A |
4711719 | Leenaars et al. | Dec 1987 | A |
4857080 | Baker et al. | Aug 1989 | A |
4865630 | Abe | Sep 1989 | A |
4880441 | Kesting et al. | Nov 1989 | A |
4971696 | Abe et al. | Nov 1990 | A |
5015269 | Garrett et al. | May 1991 | A |
5139540 | Najjar et al. | Aug 1992 | A |
5139541 | Edlund | Aug 1992 | A |
5205841 | Vaiman | Apr 1993 | A |
5217506 | Edlund et al. | Jun 1993 | A |
5259870 | Edlund | Nov 1993 | A |
5310414 | Najjar et al. | May 1994 | A |
5342431 | Anderson et al. | Aug 1994 | A |
5358553 | Najjar et al. | Oct 1994 | A |
5376167 | Broutin et al. | Dec 1994 | A |
5384101 | Rockenfeller | Jan 1995 | A |
5393325 | Edlund | Feb 1995 | A |
5415891 | Liu et al. | May 1995 | A |
5451386 | Collins et al. | Sep 1995 | A |
5456740 | Snow et al. | Oct 1995 | A |
5468283 | French et al. | Nov 1995 | A |
5487774 | Peterson et al. | Jan 1996 | A |
5498278 | Edlund | Mar 1996 | A |
5518530 | Sakai et al. | May 1996 | A |
5599383 | Dyer et al. | Feb 1997 | A |
5738708 | Peachey et al. | Apr 1998 | A |
5782959 | Yang et al. | Jul 1998 | A |
5782960 | Ogawa et al. | Jul 1998 | A |
5810912 | Akiyama et al. | Sep 1998 | A |
5827569 | Akiyama et al. | Oct 1998 | A |
5891222 | Hilgendorff et al. | Apr 1999 | A |
5904754 | Juda et al. | May 1999 | A |
5931987 | Buxbaum | Aug 1999 | A |
5938822 | Chen et al. | Aug 1999 | A |
5955044 | Lucht et al. | Sep 1999 | A |
5980989 | Takahashi et al. | Nov 1999 | A |
5989319 | Kawae et al. | Nov 1999 | A |
6010560 | Witzko et al. | Jan 2000 | A |
6066592 | Kawae et al. | May 2000 | A |
6083297 | Valus et al. | Jul 2000 | A |
6083390 | Hartmann | Jul 2000 | A |
6139810 | Gottzmann et al. | Oct 2000 | A |
6152987 | Ma et al. | Nov 2000 | A |
6171574 | Juda et al. | Jan 2001 | B1 |
6171712 | Thornton | Jan 2001 | B1 |
6183542 | Bossard | Feb 2001 | B1 |
6183543 | Buxbuam | Feb 2001 | B1 |
6913736 | Alvin et al. | Jul 2005 | B1 |
6916454 | Alvin | Jul 2005 | B1 |
20020062738 | Yoshida et al. | May 2002 | A1 |
20020141919 | Alvin | Oct 2002 | A1 |
20020141920 | Alvin et al. | Oct 2002 | A1 |
20030061937 | Ito et al. | Apr 2003 | A1 |
20030183080 | Mundschau | Oct 2003 | A1 |
20030213365 | Jantsch et al. | Nov 2003 | A1 |
20030233940 | Takatani et al. | Dec 2003 | A1 |
20040182242 | Mitani et al. | Sep 2004 | A1 |
20040237779 | Ma et al. | Dec 2004 | A1 |
20040237780 | Ma et al. | Dec 2004 | A1 |
20040244583 | Ma et al. | Dec 2004 | A1 |
20040244589 | Bossard et al. | Dec 2004 | A1 |
20040244590 | Ma et al. | Dec 2004 | A1 |
20050011359 | Pan et al. | Jan 2005 | A1 |
Number | Date | Country | |
---|---|---|---|
20050061145 A1 | Mar 2005 | US |