Patent Application
20070240566
This invention relates to a durable palladium-based alloy, especially for use in a hydrogen generation membrane. More particularly, the invention relates to a durable palladium-based alloy having stability as a hydrogen separation membrane in water gas shift (WGS) and/or reforming applications in the production and purification of hydrogen. More particularly still, the invention relates to such alloy and membrane being relatively tolerant to sulfur.
The trend towards “clean energy” and a “hydrogen economy” will be enabled by the efficient supply (e.g. from biomass) and distribution of low cost hydrogen. Fossil fuels can also be converted to hydrogen. The resulting hydrogen is then available for a number of applications, with the fueling of the electrochemical reaction in a fuel cell being one prominent use. Today most commercial hydrogen is produced through catalytic steam reforming of natural gas followed by one or more water gas shift reactors. In the future the gasification of coal and other forms of carbon like lignite probably coupled with some form of carbon dioxide sequestration may also become an important source of hydrogen. However, the low cost production of hydrogen from these resources requires significant technological advances in both hydrogen separation/purification and the efficiency of the Water Gas Shift (WGS) reaction. The WGS reaction:
CO+H2OH2+CO2,(ΔG=−41 kJ/mole)
is critical to most hydrogen production routes. However, as is expected for exothermic reactions, thermodynamic equilibrium dictates the lower the temperature, the further the reaction can proceed toward the H2 production side. Unfortunately, catalytic activity decreases with temperature and catalysts become more easily inhibited or poisoned, e.g. by sulfur in particular.
Recently, noble metal-based WGS shift catalysts with sufficient volumetric activity and thermal stability have been demonstrated. These easily operate in the 300 to 400° C. range. Such catalysts are well suited to work in concert with a hydrogen separation membrane to both purify the hydrogen produced and to help drive the Water Gas Shift reaction by removing the hydrogen product as it is formed. Thin palladium (Pd), or Pd alloy, membranes, supported on a porous substrate, are suitable for high temperature hydrogen separation and membrane reactor applications, and their application has been discussed in the prior art. Use of such separation membranes avoids the difficulties with present methods of hydrogen separation like Pressure Swing Absorption (PSA). PSA is capital intensive, has a large footprint and high operating costs.
There are important situations where improved sulfur tolerance of both the Water Gas Shift catalyst and the Pd alloy membrane would greatly facilitate the production of fuel cell-grade hydrogen. One is in the small scale production of hydrogen from reformate produced from logistic fuels like gasoline, kerosene, diesel or jet fuel. Reformate is produced by reacting these hydrocarbons with oxygen and/or water in catalytic partial oxidation, autothermal reforming or catalytic steam reforming processes. The resulting “reformate” is a mixture of carbon monoxide and hydrogen, also with some carbon dioxide, water, methane and often nitrogen, etc. For this application, the membrane has be sulfur tolerant enough to handle 20 ppm-wt S-containing fuel that produces on a dry basis about 3 ppmV S reformate. The sulfur content or partial pressure may be lower depending on the amount of N2 from the air introduced into the reforming process. Membranes for hydrogen recovery from such reformates typically require a hydrogen permeance of about 10 m3/(m2-hr-atm0.5) at 350° C., though a higher permeance of 50 m3/(m2-hr-atm0.5) at 350° C. in the absence of sulfur is very desirable, as the higher the permeance, the smaller the membrane area needed. In the presence of ≦6 ppmV sulfur at a partial pressure (as H2S or equivalent) of about 7×10−5 atm or less, slightly lower permeance can be accepted because the cost of sulfur removal can be avoided. Usually for hydrogen recovery, the reformate undergoes a Water Gas Shift Step to adjust the H2/COx ratio, and it is desirable for the Pd-based hydrogen membrane to be compatible with the Water Gas Shift system and especially desirable for the membrane to be suitable for use in a membrane water gas shift reactor.
In portable or vehicular applications, the membrane life should be on the order of 5000 hours, but it should be able to withstand several hundreds of start-up and shutdown cycles with concomitant changes in gas pressure, composition and temperature. It must be tolerant of cycling from −40° C. to 500° C. and operating from ˜200° C. to ˜400° C.
Hydrogen separation membranes could play a key role in the large scale production of hydrogen from “cleaned” coal gas. A representative gas composition for an advanced dry-feed bituminous coal gasifier is: 34% H2, 61%CO, 2%CO2, 3% “other” (dry basis), along with about 200 ppmV sulfur-containing gas, typically an equilibrium mixture of H2S and COS (carbonyl sulfide). A slurry-fed coal gasifier will have proportionally more H2 and CO2 and less CO. For water gas shift, this gas would be blended with about 3 H2O for every CO. Because, for the Water Gas Shift reaction, the ratio of H2O to CO is typically about 3H2O per CO in the feed, a Pd alloy for coal gas water gas shift service should be stable to at least ˜420° C. in the presence of feed gas with partial pressures of, for example, in Atm.: ˜6 H2, 10.8 CO, 0.35 CO2, 0.5 inerts and CH4, 32.3 H2O and about 0.004 atm H2S. As the reaction proceeds, CO+H2O react, providing CO2 and H2. Thus, the membrane also has to be stable to gas with higher CO2 partial pressure. H2 from the water gas shift reaction that is passed through the membrane and is termed “permeate”. Assuming perhaps the permeate side of the membrane involves a counter flow steam sweep gas, then as much as 95% or more of the H2 could diffuse through the membrane, lowering the H2 partial pressure at the gas exit end of the membrane at which the reacted reformate, or “retentate”, appears. The optimum amount of CO conversion achieved and the size of the membrane, and hence the degree of H2 recovery, is typically a matter of site-specific economic calculations. Thus for illustrative purposes, the membrane has to remain stable and permeable to H2 in the presence of (in atmospheres) about 1.22 H2, 0.158 CO, 16.13 CO2, 0.78 inerts and CH4, 31.71 H2O, and about 0.0052 H2S (in equilibrium with COS) in the retentate gas. Such a Pd membrane has to be stable for thousands of hours, preferably 8,000 hrs and more preferably >25,000 hrs, under these harsh operating conditions of >300° C., typically about 400° C. to 420° C., with a peak temperature of about 500° C. or more depending on the system in the presence of 0.005 atm partial pressure or more of sulfur as H2S and carbonyl sulfide. The membrane should have a hydrogen permeance of at least 20 m3/(m2-hr-atm0.5). These operating conditions and the temporal stability required would be a severe challenge to all known Pd-based membranes. Furthermore, any useful hydrogen membrane has to endure, without cracking, system upsets that can result in sudden loss of temperature and pressure. While pure Pd membranes cannot withstand such system upsets, Pd alloys such as described below can.
Pd-atomic hydrogen interactions give rise to a β-phase Pd-hydride, unstable above 295° C. and an α-phase hydride, stable at high temperatures and characterized by markedly lower hydrogen content. At these higher temperatures, the miscibility gap between the α- and β-hydride phases narrows and they coalesce at a critical point around 295° C. and ˜20 atm. The α-β-transition, occurring for pure Pd near room temperature, causes serious alteration in the atom spacing of the metal lattice. The consequent dimensional changes can distort the membrane, making it less mechanically resistant, more brittle and prone to rupture. As a result, the resistance to repeated start-up/shut-down cyclic stresses is low.
Binary addition elements, having in general Face-Centered-Cubic (FCC) structures such as Ag, stabilize the α-hydride phase against the β-hydride phase, reducing the problem of embrittlement, and yield a hydrogen permeability that is greater than that of pure Pd. However, Pd—Ag is rapidly and irreversibly poisoned by sulfur and even with the best sulfur clean-up technologies, there is a reasonable likelihood that a process upset or a change in feedstock will expose the membrane to sulfur. Clearly, neither Pd nor Pd—Ag membranes, in and of themselves, are suitable for use with sulfur-containing feed gas.
A somewhat sulfur-tolerant Body-Centered-Cubic (BCC) phase Pd-40 wt. %Cu alloy with a higher hydrogen permeance than pure Pd has been described in a 2002 DOE report by J. D. Way entitled “Palladium/Copper Alloy Composite Membranes for High Temperature hydrogen Separation from Coal-Derived Gas Streams”. This alloy also avoids the alpha-hydride/beta hydride transition problem that plagues pure Pd. However, the optimum Pd—Cu composition is a BCC β-PdCu phase and is perilously close to the β-PdCu phase/α(FCC)+β(BCC) mixed phase stability boundary. This means that, with an increase in temperature, it can rapidly change structure, from the desirable BCC to the less desirable FCC phase, thus losing its hydrogen permeance and structural integrity. This structural change can be caused by a system upset that increases the temperature beyond the stability of the β phase or because of Pd—Cu segregation occurring over time. Sulfur-containing reformate, rich in CO, at elevated pressures such as described earlier is an ideal environment for this segregation.
Thus, there remains a need for a sulfur tolerant, long life, relatively low cost and high permeance Pd alloy, and a hydrogen generation membrane of such alloy, that does not suffer from either the alpha-hydride/beta-hydride transition problem, having a phase boundary close to the membrane operating point, or metal segregation under sulfur containing reformate with time, under conditions like the production of hydrogen from “cleaned” coal gas via the water gas shift reaction over a sulfur tolerant catalyst such as described above.
Accordingly, a primary advantage of the present invention is the provision of an improved Pd-based alloy, and a membrane of such alloy, that is durable under operating conditions of extended temperature ranges and/or sulfur presence. A further advantage is the provision of such alloy and/or membrane thereof having high H2 permeance. Yet a further advantage is the provision of such alloy and/or membrane thereof being relatively cost effective. A still further advantage is the provision of an alloy and/or membrane thereof, having one or more of the foregoing characteristics and including Cu in the alloy. An even further advantage is the provision of a membrane having such characteristics in the context of incorporation/use with reforming and/or WGS reactors.
The present invention provides a Pd-based alloy for use in and/or as, an H2 separation membrane, especially in WGS applications, which alloy has appropriate alloy phase control at the service conditions to achieve one or more, and typically all, of high permeance, sulfidation resistance, structural stability and long life. The Pd-based alloy of the invention is particularly suited for the production of cost effective H2 separation membranes for use under challenging operating conditions in the production/separation of H2. Further, the Pd is normally alloyed with at least Cu as a binary constituent of the alloy, for the increased sulfur tolerance afforded by such inclusion. Beyond these two basic constituents of the alloy, the invention provides for at least a one (ternary) constituent from a group “M” of metals that are BCC and act to stabilize the β BCC phase of the alloy such that any non-BCC type phases present in the alloy have negligible effect on temperature-related phase changes of the alloy, thereby to retain hydrogen permeability during operating temperature and pressure cycles.
The invention, in one embodiment, provides, in addition to the at least one constituent from group M, for at least one constituent from a further group of metals “N” that are non-BCC (e. g., are FCC) and act to enhance a further property of the membrane other than thermal stability. For convenience of discussion, the alloy of the invention may be identified as PdCuX, where “X” includes at least one or more metal constituents from the group “M”, and optionally, also includes one or more metal constituents from the group “N”. Indeed, in the extreme, “X” may include multiple constituents from one, the other, or both of the groups M and N. The PdCuX alloy, where X=M+N, may change from one phase to the other as a function of temperature according to Gibb's Phase Rule (which determines the possible number of degrees of freedom in a closed system at equilibrium, in terms of separate phases and the number of chemical constituents in the system). Specifically, one aspect of the present invention is the stabilization of the PdCuX alloy in the β (BCC) phase during the service, or operating, temperature range. A further aspect is to do so with acceptable hydrogen permeability, as well as CO and sulfur tolerance.
The metals constituting the group “M” consist of Fe, Cr, Nb, Ta, V, Mo, and W, with perhaps Nb and Ta being the most preferred because their lattice sizes are closest to Pd and Cu. The metals constituting the group “N” consist of Ag, Au, Re, Ru, Rh, Y, Ni, Ce, Pt, Ir, Co, La and In, with perhaps Au being the most preferred because of its likelihood of better sulfur tolerance than most other members of the group, and its positive effect on improved ductility.
In the context of the parameters above, the atomic % of Pd present in the alloy is in the general range of 35-55 at. %, the Cu is present in the general range of 30-64 at. %, and the remaining constituent(s) X of the alloy is/are present in the general range of 1-15 at. %. In a more preferred formulation of the alloy, the Pd is present in a range of 38-50 at. %, the Cu is present in a range of 50-62 at. %, and the remaining constituent(s) X is/are present in the range of 4-10 at. %. Of two different specific formulations, each particularly preferred, one is preferred for minimum phase change and has 40 at. % Pd, 53 at. % Cu and 7 at. % X, whereas the other is preferred for maximum permeability and has 47 at. % Pd, 46 at. % Cu and 7 at. % X.
Still further, the alloy of the invention finds particular use as, or in, an H2 separation membrane in a water gas shift membrane reactor, a hydrocarbon or biomass reforming reactor, and/or a reforming reactor.
The foregoing features and advantages of the present invention will become more apparent in light of the following detailed description of exemplary embodiments thereof as illustrated in the accompanying drawings.
A description of the invention initially includes further reference to the Background Art discussed previously, followed by a detailed description of the invention and one or more preferred modes for carrying out the invention, with reference to the accompanying drawings.
Alloy development plays an important role in overcoming the problem of Pd embrittlement and in developing S-tolerant Pd-based alloy membranes. As a function of temperature and hydrogen partial pressure, Pd-hydrogen interactions give rise to a β hydride-phase, unstable above 295° C. and an a-phase, stable at high temperatures and characterized by markedly lower hydrogen content. The α-β hydride transition causes serious alteration in the atom spacing of the metal lattice. The consequent dimensional changes can distort the membrane, making it less mechanically resistant, more brittle and prone to rupture. As a result, the resistance to repeated start-up/shut-down cyclic stresses could be low.
As mentioned earlier, the binary addition elements such as Ag, Ru, Rh, Y, Ni, Ce, Cu and Au may stabilize the α-hydride phase against the β-hydride phase, reducing the problem of embrittlement. For these binaries and certain ternary Pd alloys such as Pd—Ru—In, Pd—Ag—Ru and Pd—Ag—Rh, the hydrogen permeability is greater than that of pure Pd, as indicated in Table I below, taken from a paper by H. Y. Ma, et al., entitled “Composite Palladium and Palladium-alloy Membranes for High Temperature Hydrogen Separations”, ACS Fuel Chemistry preprints, 2003, 48(1), 333.
Specifically, 30 wt. % Ag stabilizes the α-phase even at room temperature. The equilibrium adsorption of hydrogen in the Pd—Ag alloy increases as the relative amount of Ag in the alloy is increased. However, the diffusion coefficient of hydrogen into the alloy decreases with increasing Ag content. As a result of these two opposing factors, the permeability (which is the product of solubility coefficient and the diffusion coefficient), for a Ag—Pd alloy reaches a maximum value of 1.7 times the permeability of pure Pd at 23 wt. % Ag and a temperature of 350° C.
Regarding the relative costs of the above materials, Cu and Ag are the least expensive. Moreover, the 60wt. % Pd-40 wt. % Cu composition is sulfur-tolerant, as indicated by the preliminary results described in the earlier-mentioned DOE report by Way that show that exposure to 100 ppm H2S in a 50%/50% H2/N2 mixture inhibits or reduces the H2 permeation, but this effect is reversible. In the presence of steam, the reduction of the H2 permeation due to H2S exposure is less severe.
In view of the foregoing, Cu becomes a preferred binary constituent with Pd, for the reasons of cost and relative sulfur tolerance. Within the Pd—Cu binary system, the 60 wt. % Pd-40 wt. % Cu composition has previously been found, in U.S. Pat. No. 3,439,474 to McKinley, to exhibit the highest H2 permeation because of the formation of an ordered, BCC crystal structure (β-phase). However, there are long-term stability implications with this composition that need to be addressed. Thermal stability is a key requirement for the long-term (>10,000 h) durability of the H2-separation membrane, in addition to permeability, cost and S-tolerance.
That 60 wt. % Pd-40 wt. % Cu alloy is somewhat of an anomaly. It has an ordered BCC structure, while both Pd and Cu have FCC crystal structures. The optimum bulk structure, exposed crystallographic orientation at the surface, and reactivity of surface binding sites for a given alloy composition are determined by the most stable arrangement and bonding hybridization of the local directional transition metal electronic states.
Because it is difficult to observe the influence of alloy local atomic scale ionic and electronic structure on key desired performance properties directly through experiment, the present invention arose from the use of first principles modeling to design alloy properties from the ground up. Modeling was used to test compositions chosen by 1) the belief that to stabilize a BCC phase formed from two elements, Pd and Cu, that are normally FCC, a third element that is normally BCC should be added, and 2), that the choice of this should be guided in part by some empirical metallurgy guidance such as Hume-Rothery rules. As step 3), VASP (Vienna ab initio simulation package) is used to choose the possible elements selected following 1) and 2) above that when substituted for either Pd or Cu, preferably will lower the energy of the system. That is, the substitution is exothermic, or at most energy neutral or only very slightly endothermic. VASP is density functional theory atomic scale simulations, implemented with the plane wave, projector-augmented wave potentials. The VASP code was used to survey alloy bulk and surface structures formed by substituting one or more elements in the Pd—Cu ordered BCC structure. Atomic scale and thermodynamic modeling were combined to virtually tune the sulfur-tolerance, phase stability, and hydrogen permeability of advanced membrane compositions. Referring to
Analyzing the Pd—Cu system from an atomistic basis for clarification of lattice structure, it is noted that the maximum transformation temperature involving the ordered PdCu structure has been reported variously to be for a content of ˜37 to ˜50 at. % Pd. However, referring to
Regardless of these attributes, the thermal cycles associated with start-up and shutdown of a typical WGS reactor would cause a Pd-alloy, such as 47 at. % Pd-53 at. % Cu, to undergo phase changes as it crosses the β/α+β solvus line. For example, the lattice parameter of an alloy with 45.5 at. % Pd changes from 0.3752 nm in the disordered state (α-phase, FCC), to 0.2973 nm in the ordered state (β-phase, BCC). The consequent dimensional changes can distort the membrane, leading to mechanical instability over time.
An important aspect of the present invention is the development of an alloy system wherein the permeability (which appears to be associated with the maximum degree of ordering in the α-phase) of the Pd—Cu system, is retained during operating cycles. Such operating cycles typically involve cycling the temperature on start-up from ambient to >400° C., and the reverse on cool down. This is achieved by stabilizing the β-phase within the range of service temperatures.
The 40 at. % Pd composition, alloy “B” depicted in
Selected elemental atom sites may be substituted in the intrinsic crystallographic structure of a given metallurgical phase with other elements that tend to form the same or different crystal structures as a means to stabilize or destabilize respectively, the parent phase structure to achieve different property changes. An example includes the alloying of steels to stabilize the austenitic (FCC) or ferritic (BCC) phase structures with corresponding FCC or BCC elements, respectively. Similarly, in complex superalloys, corresponding elements may be incorporated to stabilize the gamma or gamma-prime phases respectively, to meet structure/property goals.
The present invention uses the novel substitution of one or more other elements, “X”, into selected elemental atom sites in the intrinsic crystallographic structure of Pd alloys, where “X” is at least an element that tends to form the same or different crystal structures as a means to stabilize or destabilize, respectively, the parent phase structure to achieve different property changes. In the context of usage herein, the one or more elements “X” include at least one element from group “M”, and preferably also at least one element from group “N”. As noted earlier, it is also possible to have more than one element from each of the groups “M” and “N”. The relative stability of this substitution is established by atomistic modeling. In addition, the stability of the resulting ternary, quaternary, or higher order alloy to elemental segregation in the presence of sulfur-containing reformate is established using atomistic modeling. Reformate is defined broadly as a mixture of gases containing at least H2, and CO, and for the purposes of conducting the water gas shift reaction, steam, H2O. Reformate may also contain C02, N2, Ar, CH4, H2S and non-methane hydrocarbons as well as COS and other impurities.
As indicated above, the alloy “C” composition (47 at. % Pd-53 at. % Cu), shown in
It has been determined that elements that tend to stabilize the BCC structure are typically also BCC, such as Fe, Cr, Nb, Ta, V, Mo, and W. These potentially suitable ternary or higher order elements and their atomic size and lattice relationships are given in Table II below. The ability of one element to substitute into the crystal lattice of another depends on several factors. The difference, among other requirements, between the unit cell lattice parameters of the parent Pd—Cu matrix and a particular constituent addition from group “M” is preferably small. Partial guidance could be obtained by considering the alloy system in terms of one or more of the Hume-Rothery empirical rules. Briefly, these rules indicate that for favorable substitution to occur, the atomic size difference should be less than 15%, the electronegativities should be comparable, valences should be similar, and the crystal structures should be the same.
Interestingly, Table II shows that the atomic size difference between Cu and the alloying elements ultimately preferred herein is generally large and greater than ˜15%, thus seeming to be nominally unsuitable under that Hume-Rothery criterion. Application of this Hume-Rothery empirical rule with regard to atomic size difference between Pd and the several elements of interest herein indicates that the size factor (12.3%) and some other criteria are met for Pd—Cu, which shows complete solid solubility at elevated temperatures. However, while Fe, Cr, V, Mo, and W also have low atomic size differences, they do not form complete solid solutions with Pd, in part because of variance in the other factors. Thus, while some of the Hume-Rothery empirical rules for complete solid solubility provide some guidance in the determination of metals that make up the group “M”, they do not guide or even make obvious the ultimate selection because, in part, it is inconsistent with some of those rules. For instance, while the first Hume-Rothery criterion, size, has to be noted, the third criterion, that concerning the alloying element having the same crystal structure, is inconsistent with the results of the invention. Both Pd and Cu are FCC metals, while the metals of group “M”, to wit, V, Nb, Ta, alpha-Cr, Mo, W, and alpha-Fe are all BCC. Similarly, V, Nb, Ta, alpha-Cr, Mo, and W could not be said to have the same valence as either Pd or Cu.
The determination of the alloys of the present invention is not readily apparent from a consideration of conventional alloying guidelines such as the Hume-Rothery rules. Instead, with respect to identifying alloying elements to solve the present needs, a different methodology was undertaken. Specifically, considering that the alloying elements were all essentially cubic lattices (FCC, BCC), the difference in lattice sizes could also reflect the differences in atomic sizes, but be more discerning in the alloying characteristics in this instance. The elements are compared largely on the basis of alpha- or beta-phase stabilizers so that alloying beta-stabilizers (typically BCC with more open lattices than FCC regarding hydrogen diffusivity etc.) could influence both the atomic size averages and the type of lattice formation to suit the alloy design. The elements in Table II are, therefore, also compared on the basis of lattice size differences. By utilizing this approach, Table II indicates, in particular, Ta (8.6%) and Nb (8.7%) to be the most preferred elements, with the difference in lattice size of the metal from group “M” with respect to each of Pd and of Cu being less than about 16%. Neither element would be selected based on the atomic size differences, also shown in Table II, which are ˜33% each for both Ta and Nb. It remains unclear whether the lattice size difference adheres to a specific number around 15%, but the smaller the difference the more favorable it would appear for positive effects.
The alloys of the present invention are designed to meet the criteria of maximizing the H2 permeance and thermal stability as well as stability in the presence of carbon oxides, steam and sulfur containing gases, rather than complete solid solubility alone. Indeed, it is notable that preferred compositions are not contingent on meeting all the Hume-Rothery criteria. In order for BCC elements to substitute into Alloy C, the lattice size factor gives an initial guide indicating Ta and Nb as preferred additions from group “M”. Retaining the H2 permeance while minimizing the level of phase change, is the goal of these alloy designs. In general, the approximate compositions of selected ternary alloys, each including an element from group “M” and derived by atomic substitution, are indicated in Table III.
The philosophy, or logic, that guided the formulation of Table III is as follows. The difference in the compositions denoted by “B” and “C” lines in
As an alternative to the foregoing approach of retaining or enhancing the single phase stability of alloy C, it is also possible to add suitable BCC elements, such as the elements of group “M”, with alloy B to retain β-phase thermal stability while improving the permeance and degree of ordering by shifting the composition toward these characteristics. Potentially suitable ternary compositions derived from this approach are shown in Table IV.
Between the two aforementioned approaches, it is possible to identify compositions that optimize both permanence and stability. Such compositions, as indicated above, may be ternary or higher order, e.g. Pd—Cu—Ta—Au, etc, where the quaternary constituent is selected from group “N” for the primary purpose of favorably balancing the composition, as needed, for other desirable membrane characteristics, such as ductility or coefficient of thermal expansion. The component from group “N”, here selected to be Au for its malleability and positive effect on permeability, serves to preserve or enhance the ductility of the alloy system under the WGS operating conditions. Using the “B” alloy technique, at least the ternary element is selected from group “M” of BCC elements that tend to stabilize the β-phase. More specifically, for reasons of smaller atomic misfit and overall alloy stability (i. e., highly negative values of ΔGo; the heat of formation), Ta and Nb are preferred. Hence a Pd—Cu—X alloy, e. g., Pd—Cu—Ta (or Nb)-alloy, wherein X=M+N (and N may be zero) and which is a robust product, is the primary goal achieved by the present invention.
An example of a preferred alloy results from VASP atomic modeling of 6.25 at. % Nb or Ta in substitution for Cu in the Cu sublattice of a 50 atomic % Pd-50 atomic % Cu ordered alloy. For the resulting alloy, 50 at. % Pd, 43.75 at. % Cu, and 6.25% at. % Nb, the heat of formation at 0° K is 6.84 kJ/mole atom lower than the unsubstituted PdCu alloy. The corresponding Ta alloy has a heat of formation at 0° K. that is 9.35 kJ/mole atom lower than the unsubstituted PdCu alloy. In other words if Pd8Cu8 has a heat of formation of −12.35 kJ/mole atom, the Pd8Cu7Nb heat of formation is −19.19 kJ/mole atom and that of Pd8Cu7Ta is −21.70 kJ/mole atom. In other words, the substitution is exothermic. It appears that the substitution of these transition metals for Cu is consistently more favorable than their substitution on the Pd sublattice. Thus the results are construed to indicate that the ternary Pd—Cu—Nb or Pd—Cu—Ta should be spontaneous and stable with respect to segregation to the binary 50 atomic % Pd-50 atomic % Cu ordered alloy and the pure substituents phase. Moreover, the substitution of these transition metals increases the average atomic volume and bond distances, increased average atomic volume and bond distances correlate well with improved hydrogen mobility. Thus, the ternary alloy formation is expected to stabilize the BCC ordered structure, while offering the likelihood of improved hydrogen permeance.
Additionally, there are alloys within the scope of this invention that will stabilize the BCC structure and yield sufficient hydrogen permeance under harsh operating conditions while having a slightly endothermic heat of formation. For example, the Pd8Cu7W alloy that has a heat of formation of −10.47 kJ/mole atom. It thus has a 15.22% higher heat of formation than Pd8Cu8 at −12.35 kJ/mole atom.
Referring to
It may thus be appreciated from the above detailed description of the preferred embodiment that the present invention relates generally to the development of an inorganic composite Pd—Cu-based membrane reactor module that is structurally stable under the fuel reforming Water Gas Shift (WGS) reaction conditions with a high selectivity and flux rate for hydrogen permeation through the membrane and acceptable tolerance to sulfur contamination. More particularly, the present invention relates to the development of an alloy system used as the membrane, such as the Pd—Cu—X, wherein the measured, maximum permeability of the β BCC phase is retained without undergoing a significant phase change during operating temperature cycles. As used here, a significant phase change is one that changes or decreases hydrogen permeability, life, thermal cycling tolerance, etc. more than a few percent. The life of an alloy is normally a function of the use environment and the number of stop-start cycles. A resistance to sulfidation is understood to mean resistance to forming a sulfide or oxysulfide phase under the normally-expected operating conditions.
Further, the present invention relates to and embraces the general use of the herein described alloys in a Water Gas Shift Membrane reactor, a hydrocarbon or biomass reforming reactor producing hydrogen or a membrane section between Water Gas Shift Catalyst or Reforming catalyst segments, or between Water Gas Shift Reactors or reforming reactors. As will be readily understood by those skilled in the art, these membranes may be made through a wide variety of methods ranging from electroless plating through chemical vapor deposition to various vacuum sputtering techniques. These membranes may be self supporting or on a porous support to provide physical strength. The porous support may be a composite of an oxide or refractory compound or compounds or metal on a porous tubular material or plate of stainless steel or other material. Due consideration should be given to the impact of the various coefficients of thermal expansion and for the Pd alloy membrane lattice expansion in the presence of hydrogen when designing such a composite system for use in a hydrogen plant, especially in conjunction with Water Gas Shift reactors.
Although the foregoing description of the present invention has been shown and described with reference to particular embodiments and applications thereof, it has been presented for purposes of illustration and description and is not intended to be exhaustive or to limit the invention to the particular embodiments and applications disclosed.
This application is a continuation-in-part of International Application PCT/US2005/047045, with an international filing date of Dec. 23, 2005 and from which priority is claimed.
The U.S. Government has a paid-up license in this invention and the right in limited circumstances to require the patent owner to license others on reasonable terms as provided for by the terms of (contract No. DE-FC26-05NT42453) awarded by the Department of Energy.
| Number | Date | Country | |
|---|---|---|---|
| Parent | PCT/US05/47045 | Dec 2005 | US |
| Child | 11712200 | Feb 2007 | US |
