Method of forming robust metal, metal oxide, and metal alloy layers on ion-conductive polymer membranes

[0002] This application is a division of U.S. patent application Ser. No. 09/488,402 filed Jan. 20, 2000, now abandoned, a continuation-in-part of U.S. Pat. application Ser. No. 09/003,228 filed Jan. 6, 1998 , now U.S. Pat. No. 6,077,621 which was based on Provisional Application Serial No. 60/035,999 filed Jan. 22, 1997.

[0003] This invention relates to a new method to coat metals, metal oxides, or metal alloys onto solid polymer electrolytes or other ion-conducting polymer surfaces. Materials prepared using this method would gain advantage in electrochemical and membrane-based applications. By employing dual ion-beam assisted deposition (IBAD), one can systematically prepare consistent thin films of these metals on solid polymer electrolyte membranes such as Nafions (DuPont, Wilmington, Del.), a copolymer of tetrafluoroethylene and sulfonyl fluoride vinyl ether and the membranes coated thereby and electrolytic cells containing the same.

[0004] The use of solid polymer electrolytes has greatly expanded the field of electrochemistry. Electrochemical processes depend on the transfer of ionic and electronic charge through the use of an anode, cathode, and an ionic liquid electrolyte. However, with the advent of the solid polymer electrolyte fuel cell, the traditional liquid phase has been replaced with a membrane composed of a polymer electrolyte that transfers ionic charge under typical electrolytic conditions. These solid polymer electrolytes are often ion-conducting membranes that are commercially available. For example, in addition to the previously mentioned Nafion (a cation exchange membrane), Asahi Chemical and Asahi Glass make perfluorinated cation exchange membranes whereby the ion exchange groups(s) are carboxylic acid/sulfonic acid or carboxylic acid. These companies produce cation exchange membranes with only the immobilized sulfonic acid group as well. Non perfluorinated ion exchange membranes are available through Raipore (Hauppauge, N.Y.) and other distributors such as The Electrosynthesis Co., Inc. (Lancaster, N.Y.). Anion exchange membranes typically employ a quaternary amine on a polymeric support and are commercially available as well. Other manufacturers and researchers fill the pores of an inert matrix with immobilized ionomer, creating an effective ion conducting membrane. (For example, see Fedkiw, P. S. and Nouel, K. M. in Electrochemica Acta, 1977).

[0005] Nafion is typically employed in some fuel cells. For the hydrogen/air (O₂) fuel cell, hydrogen and oxygen are fed directly to the anode and cathode. respectively, and electricity is generated. In order for these “gas breathing” electrodes to perform, the electrode structure must be highly porous to allow three phase contact between the solid electrode, the gaseous reactant, and the electrode which can be in the form of a membrane or polymer electrolyte. This class of electrode is called a gas diffusion electrode. In addition to a gaseous hydrogen fuel and gaseous air (O₂) oxidant, others employ a mixed phase system such as the methanol/air (O₂) fuel cell. Here, liquid methanol is oxidized at the anode while oxygen is reduced at the cathode. Another utilization for ion-conducting membranes and gas diffusion electrodes includes the electrochemical generation of pure gases (for example see Fujita et al in Journal of Applied Electrochemistry, vol. 16, page 935, (1986)), electro-organic synthesis (for example see Fedkiw et al in Journal of the Electrochemical Society, vol. 137, no. 5, page 1451 (1990)), or as transducers in gas sensors (for example See Mayo et al in Analytical Chimica Acta, vol. 310, page 139, (1995)).

[0006] Typically, these electrode/ion-conducting membrane systems are constructed by forcing the electrode against the ion conducting membrane. U.S. Pat. Nos. 4,272,353; 3,134,697; and 4,364,813 all disclose mechanical methods of holding electrodes against the conducting membrane. However, the effectiveness of a mechanical method for intimately contacting the electrode to the polymer membrane electrolyte may be limited since the conducting membrane can frequently change dimensions due to alterations in hydration and temperature. Swelling or shrinking can alter the degree of mechanical contact.

[0007] Thus, a preferred method of contacting the electrodes with the polymer membrane electrolyte involves direct deposition of a thin electrode onto one or both sides of the polymer substrate. Nagel and Stucki in U.S. Pat. No. 4,326,930 disclose a method for electrochemically depositing platinum onto Nafion. Others have employed chemical methods whereby the metal salt is reduced within the polymer membrane (for example see Fedkiw et al in Journal of the Electrochemical Society, vol. 139, no. 1, page 15 (1992)). In both the chemical and electrochemical methods, one essentially precipitates the metal onto the ion conducting membrane. This precipitation can be difficult to control due to the nature of the ion-conducting polymer membrane, the form of the metal salt, and the specific method employed to precipitate the metal. As the goal of a thin, porous, and uniform metal layer is often not met via precipitation, practitioners have turned to other deposition methods.

[0008] Scientists and engineers have long realized that the specialized coating methods of ultra-high vacuum (UHV) evaporation, chemical vapor deposition (CVD), and sputter deposition (also called sputtering), may offer a better method to create thin metal electrode surfaces. A successful surface treatment via UHV starts with creating a clean substrate. Insuring the substrate surface is atomically clean before deposition can be of vital importance for adhesion. A source metal is contained in a water cooled copper hearth. This metal is vaporized through resistive, eddy-current, electron bombardment, or laser heating. The resulting vaporized metal diffuses to the substrate and condenses to form a film. Evaporation rates depend exponentially on temperature and thus means for precise temperature control is needed.

[0009] One method for minimizing these control problems is to heat the substrate to a temperature that still allows for condensation of the vapor. UHV is often used when freedom from contamination in both the film and the interface between the film is of the utmost importance—for example in the electronics industry. While UHV is appropriate for modifying electrode surfaces, using this technique for the direct deposit of a thin electrode layer onto an ion-conducting polymer substrate may be limited due to the temperature of deposition and constraints of an atomically clean substrate.

[0010] The chemical vapor deposition process occurs at atmospheric pressure and typically employs temperatures lower than UHV or sputtering. In CVD, the constituents of a vapor phase are often diluted with an inert carrier gas. The carrier and vapor phase are reacted at a hot surface and subjected to ions created by circular magnetrons. The substrate target is not part of the ion-creating circuit so only neutral vapors are deposited onto the target. However, unlike the condensation of UHV and sputtering, the surface reaction for CVD is considered a chemical reaction. For example, to deposit tungsten, one would mix hydrogen with tungsten hexafluoride at 800° C. Tungsten metal then deposits on the substrate via diffusion. While CVD may offer a potential method to coat an ion-conductive polymer membrane, once again the temperature restraints may allow this technique only limited application.

[0011] The most common metal deposition method is sputtering. The process begins by mounting a sample to a water-cooled support. The sample is next subjected to a vacuum, although not as high as in the UHV technique. Once vacuum is achieved, a source of metal is heated until the metal vaporizes. These metal atoms are further bombarded by positive ions of a carrier gas. The now ionized metal atoms diffuse to the substrate. Since the sample is cooled, the resulting metal vapor condenses on the sample. However, continuous ion bombardment imparts enough energy to re-evaporate the deposited metal on the substrate.

[0012] This annealing process of bombardment, condensation, and additional evaporation from the substrate eventually forms a thin metallic film. The pressure and substrate temperature employed controls the morphology of film formation. Often, substantial heating of the substrate occurs as the sputtered ions cool and evaporative energy is transferred to the substrate. Substrate heating is especially problematic when metal with high ejection energies are employed (platinum, tungsten, tantalum, rhenium, and uranium). Excessive heating can cause distortion in the film through differential expansion between the film and substrate. For this reason, sputtered films of platinum on Nafion are not stable: the difference in thermal capacity between the two materials is too great.

[0013] Regardless, the necessity to construct thin layers of electrodes has led researchers to coat anodes or cathodes. Sputtering has been employed to make fuel cell type electrodes by depositing a thin layer of metal onto a carbon support and not directly on the ion-conducting polymer membrane. Weber et al show in The Journal of the Electrochemical Society(vol 134, no. 6, page 1416, 1987) that platinum can be successfully sputtered onto carbon composite electrodes, and the performance of these electrodes in an alkaline (liquid) fuel cell is comparable, if not better than traditional methods of preparation. Aside from performance, there is an advantage in being able to use lower amounts of platinum without extensive loss in performance, thus reducing the cost of the composite.

[0014] Ion-conducting polymer membranes are often used as selective separators in electrolytic processes. For example, Nafion is a cation exchange separator employed in brine electrolysis. Sodium ions from brine migrate through the separator to combine with hydroxide formed at the cathode. Chloride ions remain in the anolyte compartment due to charge repulsion with the immobilized sulfonic acid groups of the Nafion. Other uses of charged membranes for separation include electrodialysis and synthesis. For most of these applications, it is desired to transport one species at the exclusion of others. However, ion-exchange separators do not always perform as desired. For example, the previously cited methanol/oxygen fuel cell suffers from unwanted transport of methanol through the ion-conducting separator. In addition to protons, the separator allows methanol to cross over to the oxygen-reduction side, eventually parasitically reacting at the cathode, thereby lowering the cathode potential.

[0015] It is known that thin coatings of certain metals, metal oxides, or alloys selectively allow the transport of ions. For example, thin layers of palladium, tungsten oxide, molybdenum oxide or hydrogen uranyl phosphate (HUO₂PO₄) have been shown to selectively transport protons. If a thin layer of these or other metals could be securely fixed to an ion-conducting membrane, then one would expect enhanced selectivity for ion transport.

[0016] Thus, if one were able to coat ion-conducting membranes with a thin, robust metal, metal oxide, or alloy layer, one would accrue performance and cost advantages in both electrochemical and separation applications. As individual techniques, the current metal deposition technologies (UHD, CVD, sputtering) are limited in their ability to generally apply a thin metal film to an ion-conductive membrane due to either operating temperatures or process conditions.

[0017] The novel dual beam assisted deposition process of the invention for the preparation of an ion-conducting membrane provided with a thin film of a metal or metal oxide deposited thereon comprises subjecting under vacuum the ion-conductive membrane to an electron beam having an energy less than 500 eV which cleans the membrane surface and then subjecting under vacuum the cleaned membrane to a high energy electron beam having an energy between 500 and 2000 eV and containing ions of the metal to be deposited to form a thin metal or metal oxide film on the membrane useful as an electrode-membrane structure. The resulting metallized membrane structures and fuel cells containing the same are also novel and part of the invention.

[0018] The metal-membrane electrodes can be used in fuel cells such as described in U.S. Pat. No. 4,044,193 or hydrogen-reduction metal recovery cells described in U.S. Pat. No. 4,331,520 or gas diffusion cells as described in U.S. Pat. No. 5,047,133 as well as in the electrolysis of gaseous hydrogen halides such as anhydrous hydrogen chloride.

[0019] Dual ion-beam assisted deposition (Dual IBAD) is a hybrid method that combines some of the best features of the above metal deposition technologies. Using both a vacuum and temperature control, IBAD combines vapor deposition with simultaneous ion-beam bombardment. The vapor deposition is initiated via electron-beam evaporation of the source. Along with the evaporated species, two ion-beams converge on the substrate. A relatively low energy source of ions is initially focused on the substrate. Typical low energy ion beams are Ar⁺. This first beam both cleans the surface through ion-sputtering and in some cases imparts a unique atomic-scale texture to which the coating will be applied.

[0020] The second, higher energy beam, e.g. O₂⁺ or N₂⁺, and the electron-evaporated species (e.g. platinum, iridium, gold, rhenium, rhodium, tantalum, tungsten, silver, zinc, iron, copper, nickel, etc.) are aimed at the surface. It is believed that the concurrent ion stitching densifies the now forming film and improves the adherence between the film and the substrate.

[0021] In some cases the energy of the two energy beams is identical, and the ion carrier is a single gas. This variation is simply called IBAD. The use of Dual IBAD versus IBAD is determined by the peculiarities of the alloy, metal, or metal oxide source target and the polymer substrate to be coated.

[0022] The use of the concurrent ion beams allows one to construct adherent films of highly controlled porosity, depth, and ion composition on a variety of materials. For example, IBAD has been used to deposit metal, metal oxide, metal alloy, or metalloid films on metals, ceramics, or insulating polymers such as silicone rubber and polyimide. By adjusting deposition parameters, a film's porosity can be controlled and range from being highly dense and impervious (for example as a thin coating to create ion-selective membranes) to highly porous (a desirable quality for gas diffusion or fuel cell electrodes). Also, different metals, metal oxides, metalloid, or alloys can be constructed and deposited on the substrate at the atomic level. Typical coating times are 5 minutes per square meter: the process can be adapted to a semi-continuous drum feed whereby large quantities of material can be coated at once.

[0023] One key to the IBAD process as a general coating procedure is that the method functions over a wide range of temperature and pressure conditions. The essential attributes of the IBAD process are as follows:

[0024] It is a low temperature process which can be used on any substrate material.

[0025] It results in excellent adhesion of metallic or ceramic films on metal, polymer membranes, or ceramic substrates.

[0026] One can form dense, low-stress metal films with a porous or non-porous structure.

[0027] Superb control over film microstructure and chemical composition.

[0028] Scaleable, economic process.

[0029] IBAD conditions are dependent on the chemistry of the substrate to be coated and the ionization of the target metal and ion beam gas. For example, if ultra-high purity “zero defect” depositions are desired, then very high vacuums are employed, on the order of less than or around 10⁻¹⁰-10⁻¹²torr.

[0030] The temperature employed is both dependent on the substrate and the chemistry of deposition. However, for IBAD, the temperature does not affect the process as much as the material characteristics of the polymer substrate. For example, for Nafion® type polymer substrates, one could employ a range equal or less than 80-160° C., but preferably less than 80° C. If one is coating a Teflon type polymer, a temperature range of 50-250° C. is acceptable, but in all cases below 300° C. One prefers to operate below the glass-transition point of the polymers.

[0031] The energy of the high and low power ion beam is dictated by both the gas employed, and ultimately (for the high power) by the power available to the instrument. Typical low power beams range from 100-500 eV while high power beams range from 500-42000 eV.

[0032] The IBAD methodology has been exploited to coat ion-conducting membranes and create electrode and membrane structures of far greater film adhesion and interface control than previously possible. Porous electrode-membrane structures created via IBAD have a primary use as gas diffusion and find immediate implementation in fuel cell, electrosynthetic, sensor, and electrochemical separation fields. However, since thin, dense, and impervious films can also be formed on ion-conducting membranes, metal membrane structures find use as highly selective separators. For example, the problem of methanol cross-over and reactions at the cathode in a methanol/oxygen fuel cell may be alleviated by depositing a thin impervious layer of tungsten oxide on the ion-conducting membrane. The tungsten oxide would allow the passage of protons while isolating methanol from the cathode reaction.

[0033] In another facet of the invention, the process may be used for coating a gas diffusion electrode that has had a layer of ionomer applied to its surface prior to the IBAD treatment. This alternate ionomer layer would be like an in situ membrane.

[0034] Typically these liquid ionomers are obtained as solutions and can be thermally or chemically treated to form a thin film. For example, Nafion solution can be purchased as a 5% (wt) solution in isopropyl alcohol and other solvents whereby the equivalent weight of the polymer is 1100. This solution is sprayed or painted onto an uncatalyzed ELAT™, and heated at 80° C. to form a polymeric layer on the electrode. IBAD could then be employed to coat metals, alloys, or metal oxides on the ion conducting polymer layer. Structures formed in this manner can still be mated to a standard ion conducting membrane as well. Thus, a complete electrode pair would consist of the following layers:

[0035] ELAT/Nafion (from sln)/IBAD-metal/Nafion(memb.)/IBAD-metal/Nafion (from sln.)/ELAT

[0036] “IBAD-metal” designates an ion beam assisted deposition of a layer of metal, alloy, or metal oxide.

[0037] As a new process for constructing a metal, metal oxide, or alloy interface on an ion-conducting membrane, many benefits are obvious. The cost for creating electrodes or membranes is expected to be significantly reduced as all of the deposition process is automated. The throughput of structures created through IBAD is expected to be as fast or faster than the current labor-intensive techniques. Finally, the ability to tailor the surface and interface of the metal ion-conducting structure to the demands of a specific application without significantly altering the actual manufacturing process allows for short development times of custom fabricated structures.

[0038] The electrolysis of HCl is one of the most important industrial electrochemical processes, as it provides at the same time for disposal of an abundant by-product in the industry of chlorine while recovering value-added chlorine gas. It is known in the art that chlorine may be recovered from aqueous solutions of hydrochloric acid, both by a conventional electrolysis process (with hydrogen evolution at the cathode and chloride evolution at the anode) or by a depolarized process (wherein an oxygen consuming cathode is coupled to a chlorine evolving anode). While aqueous hydrochloric acid may be advantageous when the source of raw acid is already a solution, dissolution in water of a pure gaseous hydrogen chloride by-product would be a useless waste of energy and of pure water, involving the treatment of large volumes of solutions. Furthermore, there are some process limitations that force the electrolysis of aqueous hydrochloric acid to be carried out at a current density not exceeding 5 kA/m², and more typically around 3 kA/m². Operation at higher current density would be preferred to minimize the investment costs on the plants and enhance productivity.

[0039] An alternative to aqueous HCl electrolysis has thus been developed, wherein pure gaseous hydrogen chloride fed at the anode side of a cation exchange membrane electrolyzer is electrolyzed, while in the same compartment, pure chlorine is evolved and protons are generated. Said protons are forced to migrate through the cation exchange membrane to the cathodic compartment, which is fed with an acidic aqueous solution (having the function of withdrawing the generated heat and also of keeping the membrane wet to ensure its electric conductivity), and wherein hydrogen is evolved. Unfortunately, the issue of lifetime of state-of-the-art gas diffusion electrodes in the case of anhydrous hydrogen chloride is quite critical, and this single issue has so far prevented this technology to be successful in an industrial scale.

[0040] Gas diffusion electrodes are in fact activated with noble metals, alloys and/or oxides thereof, for instance with ruthenium dioxide, which offers a particularly desirable combination of activity and stability. However, for economical reasons, such noble metal species cannot be used as bulk components, having instead to be supported in a highly dispersed form on a current conductive and chemically inert material. The catalyst supports known in the art which have the most advantageous characteristics are almost exclusively based on carbon or graphite materials, and in any case they undergo, to some extent, corrosion phenomena in an environment as aggressive as the one relative to gaseous hydrogen chloride electrolysis. This is especially true as far as the anodic compartment is concerned, wherein chlorine is evolved in the presence of moisture which diffuses from the cathodic compartment through the water permeable membrane. The corrosion of the support eventually leads to deactivation of the gas diffusion electrode, which must be replaced, thus adding unacceptable direct and indirect costs to the process. On the other hand, the IBAD technique offers a cheap and effective way to get rid of catalyst supports, by dispersing the active ruthenium—ruthenium dioxide mixtures or ruthenium dioxide in a highly subdivided form directly on the membrane. Ruthenium metal per se is not useful for anhydrous hydrogen halide electrolysis due to the corrosive conditions of the process.

[0041] In the following examples, there are described several preferred embodiments to illustrate the invention. However, it should understood that the invention is not intended to be limited to the specific embodiments.

[0042] The following example demonstrates the use of an IBAD-prepared gas diffusion electrode in a hydrogen/air (O₂) fuel cell. A sample of Nafion 115 (8″×8″) is used as received except for wiping the surface clean with a soft tissue before placing it on a drum. The drum was rotated at 2.5 rpm throughout the process and the system was evacuated to 10⁻⁶torr. The metal source target was platinum while argon was ionized in both the energy beams. The energy of the beams was between 200 and 1000 eV. A source of platinum was the target for an evaporation electron beam powered with a 14K watt supply. The incipient platinum film's progress was monitored via quartz crystallometer. The sample temperature was monitored and ranged from 35 to 65° C. For this trial, one side of the Nafion was coated with platinum. Four samples were prepared whereby the average film thickness was 3000, 1000, 574, and 241 angstroms. The assemblies with 574 and 241 angstrom platinum films were judged sufficiently porous to run as gas diffusion electrodes.

[0043] The hydrogen/air (O₂) fuel cell test stand consisted of a single anode/cathode pair of 16 cm²exposed electrode area. Appropriate hardware fed oxygen or air (O₂) to the cathode at 4 bar while hydrogen was supplied to the anode at 3.5 bar. In this test, air provided a source of oxygen and the cell temperature was maintained to 70° C. A current load ranging from 0-10 kA/m²was applied to the cell and a resulting steady-state voltage was recorded.

[0044] The evaluation consisted of testing three configurations of electrode. As a control, the cell was assembled with two standard commercially-available gas diffusion electrodes pressure-fit against an ion-conducting membrane (Nafion 115). An ELAT™ platinum gas diffusion electrode (E-TEK, Inc., Natick, Mass.) was employed as both the anode and cathode in this configuration. The second configuration replaced the Nafion membrane and ELAT anode with a single platinum-Nafion assembly whereby the platinum layer was 241 angstroms. The last configuration consisted of a similar anode assembly except the platinum layer was 574 angstroms. In all these configurations, no additional Nafion paint was applied to the metal coated membrane or ELAT as is sometimes cited in the literature. For all of these configurations, air (O₂) was fed to the cathode.

[0045] This example illustrates the difference between the structures formed when typical sputtering methodology is employed to coat ion exchange membranes such as Nafion. A sample of Nafion 115 (8″×8″) was used as received except for wiping the surface clean with a soft tissue before placing on a drum. The sample was subjected to a vacuum applied to a sputtering chamber and a target of platinum was evaporated at high energy, and vaporized platinum condensed on the Nafion. The sputtering was halted once an approximately 200 angstrom film was formed. Samples of Nafion thus prepared were tested as anodes and cathodes in a fuel cell set-up previously described in Example 1.

[0046] The resulting metal film adhered tightly to the Nafion membrane and, even upon exposure to water, the metal film remained intact. However, as shown in FIG. 1, the electrochemical performance of such a structure was sub-standard. FIG. 1 also plots the results of a sputtered platinum anode on the same axis as the standard electrode and IBAD prepared electrodes. It is readily apparent from this plot that the current and voltage obtained from the sputtered metal film was far below the IBAD or standard electrodes. This shows that the sputtering process leads to metal film deposits with undesirable morphological characteristics.

[0047] FIG. 1: A current potential plot of four fuel cell trials. The single cell is run with hydrogen (3.5 bar) and air (O₂) (4 bar) while the temperature is maintained at 70° C. In all cases Nafion 115 is employed. The cathode-anode pair consisted of either an ELAT-ELAT, ELAT-241 Å IBAD platinum, ELAT-574 Å IBAD platinum, or 200 Å sputtered platinum. All experimental electrodes are run as anodes while the cathode is the same material.

[0048] This example illustrates that gas diffusion electrodes made with IBAD can perform as well as assemblies currently commercially available. An ELAT is a carbon-cloth based support onto which carbon and hydrophilic wet-proofing layers are applied to both sides and whereby the platinum catalyst is similarly applied by hand. In contrast, the IBAD electrode is made with a minimum of labor, and as shown employs far greater utilization of the applied platinum by using less platinum. Moreover, the metal-membrane electrodes of the invention have much improved cell voltages.

[0049] This example demonstrates the use of IBAD—modified ion conducting membrane as a highly selective separator. The problem of methanol crossover in a direct methanol fuel cell (DMFC) can be alleviated if a thin, dense, and impervious film is created on an ion-exchange membrane whereby protons are selectively transported through the structure. Such an assembly thus allows methanol oxidation at the anode, migration of the resulting protons through the structure, and oxygen reduction at the cathode whereby the protons re-combine with the oxygen reduction production to form water.

[0050] Tungsten oxide is a known selective proton conductor. However, previously known methods to form thin, impervious films of this oxide on an ion conductive membrane have precluded its use in the DMFC, or other applications.

[0051] Thin films of tungsten trioxide were formed on 8″×8″ pieces of Nafion. As in the previous example, the Nafion was mounted on a drum after a simple tissue wiping. The drum was rotated from 2 to 5 rpm throughout the process. The sample chamber was evacuated to 10⁻⁶torr. and the metal source was electronic grade WO₃. A single beam was employed to assist in the ion deposition (i.e. “IBAD”). Argon was the ion carrier for the ion beam and the energy for the deposition beam ranged from approximately 100 to 1000 eV. The tungsten oxide source was vaporized with an evaporation electron beam powered by a 14K watt supply. Using this supply, the source metal was evaporated at a coating rate of 3-10 Å/s. Formation of the metal oxide film was monitored via a quartz crystallometer and the sample temperature was below 100° C. at all times. Since the objective was to form a selective barrier, only one side of the ion-conductive membrane was coated. Two thicknesses were prepared: 0.45 and 1.05 microns. (4,500 and 10,500 Å).

[0052] Prior to testing the modified Nafion, a membrane electrode assembly (MEA) was constructed. The Los Alamos “decal” method was employed to affix platinum or platinum-ruthenium oxide electrodes to both the side with the tungsten oxide film and the unmodified side. More details in the method are described in Proton Conducting Membrane Fuel Cells1, S. Gottesfeld et al, The Electrochemical Society, Pennington, N.J., Oct. 95, p. 252.

[0053] The primary assembly steps are described here. Anode and cathode inks were prepared by dispersing catalyst powders in an alcoholic solution of Nafion ionomer (5% wt. Solution, EW=1100). The anode ink catalyst was created from platinum-ruthenium oxide while the cathode was platinum black. The inks were painted onto a 5 cm²Teflon blank and oven dried. In this example, the tungsten oxide film was on the membrane side that ultimately served as the anode. A Teflon blank was pressed against the cathode (uncoated) side of the Nafion at a pressure of 2000 psi for 2 minutes while maintaining a temperature of 125° C. The anode ink consisting of platinum ruthenium oxide was painted onto carbon cloth gas diffusion media (ELAT™ from E-TEK Inc., Natick, Mass.) and this carbon blank is then pressed to the tungsten oxide at 125° C. and 200 psi for 2 minutes. The resulting electrodes contained an anode metal loading of 2-3 mg/cm²while the cathodes contained a metal loading of 2-3 mg/cm²of platinum black.

[0054] A single cell consisting of a 5 cm²active electrode area was assembled and placed in the test apparatus. A standard hydrogen/air and hydrogen/oxygen gas mix was fed to condition and obtain baseline data for the cell. Next, a 1 M MeOH solution was fed to the cell at 2 ml/min while oxygen was fed to the cathode at 400 ml/min and 60 psi while the system was held at 80° C.

[0055] Monitoring the cathode outlet for carbon dioxide with an infrared spectrometer allowed one to quantitate methanol crossover. The amount of CO₂was then used to calculate a methanol flux value, which was then used to calculate an effective methanol crossover current. The methanol crossover current was the loss of cell current due to the undesirable transport of methanol through the membrane to the cathode. Large values of mA/cm²indicated poor cell performance. In addition to taking polarization curves, the cell was subjected to high frequency AC-impedance experiments whereby the effective resistance of the membrane was determined. The results for the 0.45 micron tungsten oxide are summarized in Table 1 below.

[0056] Table 1: Summary of DMFC with Tungsten Oxide Selective Barrier

[0057] Examination of Table 1 shows that the transport of methanol had decreased approximately 31% by using the tungsten oxide modified Nafion. Furthermore, and just as important, the conductivity of the Nafion did not decrease. The functional nature of tungsten oxide was preserved, as the suboxides were considered both ionically and electronically conductive. This example demonstrates that IBAD can be employed to create selective barriers. In this case, proton transport was maintained while unwanted methanol transport was inhibited.

[0058] While the current and prior examples illustrate the key idea of ion assisted deposition, they serve as illustrations and do not intend to limit the application to only IBAD. Dual IBAD is applicable in certain circumstances. Similarly, the use of two or more separate metal or metal oxide targets is possible, allowing the formation of binary or trinary valve or platinum groups, metal alloys such as Pt:Ru, Pt:Sn, Pt:Mo, Pt:Rh, Pt:Ir, Pt:Pd, Rh:Mo, Pt:Co:Cr, Pt:Co:Ni, etc. directly on the ion conducting membrane.

[0059] This example illustrates the advantages of IBAD-modified membranes as compared to conventional gas diffusion electrodes for electrolysis of anhydrous hydrogen chloride. The same is true for other gaseous species at a gas diffusion chloride-ion exchange membrane interface.

[0060] A first set of 6″×6″ samples of Nafion 115 was used as received from the manufacturer except for wiping the surface clean with a soft tissue and isopropyl alcohol before placing on a drum. The drum was rotated at 2.5 rpm throughout the process. The system was evacuated to 10-6 torr. The metal source target was ruthenium metal while argon was ionized in both the low and high energy beams. The energy of these beams was approximately 300-400 and 1000-1500 eV respectively. In addition to the source of ruthenium and argon ions, oxygen gas was fed to a plasma, and the ionized vapor was subsequently stripped of positive radicals via an extraction grid. These positive oxygen ions were focused along with the ionized ruthenium onto the spinning Nafion samples. The stoichiometry of metal to oxygen was controlled via the amount of oxygen fed to the plasma chamber. The ratio of oxygen to metal was less than two to one to restrict formation of higher oxides of ruthenium. The incipient ruthenium oxide film's progress was monitored via quartz crystallometer and it was found that these conditions deposited a film at approximately 2 Å/second. Films of approximately 500 Å were fabricated on either one or both sides of the Nafion substrate and the sample temperature was monitored and ranged from 35 to 65° C.

[0061] A second set of 6″×6″ Nafion 115 samples was coated by the same procedure outlined above, except that the target for the high energy ionization beam was a pellet obtained by subjecting a powder of ruthenium dioxide to high pressure. Instead of introducing ionized oxygen as above, the make-up ions were Ar+ or N+.

[0062] A preliminary evaluation of both sets of samples consisted of testing the film's stability and adhesion. Adhesion was tested via the “tape test” whereby adhesive tape was pressed onto the coated surface and pulled away. No residual metal oxide was removed in this test. The film stability was tested in a solution of concentrated HCl (150 g/l) heated at 80° C. and saturated with chlorine gas. The absence of any color in this solution after one hour of exposure indicated that no ruthenium or ruthenium containing species dissolved from the film.

[0063] All of the experiments outlined above were repeated after performing a macroscopic surface roughening of the Nafion 115 membrane. It was found that a degree of surface roughening could be formed on Nafion by using templates of fibrous character. Enhancements to geometric surface area for metal or metal-oxide coated substrates are expected to demonstrate increases in effective current density. In this case, each piece of Nafion 115 to be coated was cut to 8×8 cm. Two pieces of a highly fibrous template, such as Toray Carbon Paper THPH-120, were cut to similar dimensions and placed on both sides around the Nafion. The platens of a press were heated to 85° C. After coming to temperature, the assembly was pressed between two sheets of stainless steel at 15-20 kg/cm²for 20 minutes. The assembly was removed and allowed to cool to room temperature. The carbon paper backing was peeled from the Nafion leaving a highly textured surface. One key to this process is selection of pressing conditions that do not allow the template to form a good bond with the Nafion, but allow the deformation of the Nafion surface.

[0064] Samples prepared by the above-described procedures were subsequently characterized in a 50 cm²active area electrolysis cell, wherein each membrane-electrode assembly (MEA) to be tested was placed between two shells made of PVDF-blended graphite, provided with openings for feeding pure hydrogen chloride and for discharging the product gases. In each experiment, the anodic compartment was fed with 85 Nl/h of pure hydrogen chloride at ambient temperature and at a gauge pressure of 3 bars. The cathodic compartment was fed with 7.5 l/h of a 3% w/o aqueous solution of HCl at the same pressure and the temperature of the catholyte was monitored and always kept under 85° C.

[0065] A blank test was carried out with state-of-the-art components, and the cell was equipped with a MEA consisting of a Nafion 115 membrane, coupled to two gas diffusion electrodes made of carbon cloth, activated with 4 g/m²of 30% ruthenium dioxide on Vulcan XC-72 carbon catalyst. These gas diffusion electrodes are commercially sold under the trademark ELAT by De Nora North America in the U.S.A. The cell was first put in operation at 0.5 kA/m², and a cell voltage of 1.1 V was recorded under such conditions; the current density was subsequently raised to 5, then to 10 kA/m², and cell voltages of respectively 1.49 and 1.85 V were recorded. The cell was then left in operation at 10 kA/m²for 150 hours; at the end of this period, the cell voltage had progressively increased up to 2.25 V, and a significant amount of ruthenium was detected in the anodic drain. Post mortem analysis of the electrodes revealed corrosion of the carbon support, that likely caused the observed release of a part of the supported ruthenium.

[0066] Further tests were carried out equipping the cell with a MEA consisting in a single side RuO2 coated Nafion 115 membrane, obtained with the above-described procedures, coupled to an ELAT gas diffusion cathode. The IBAD coated side of the membrane was employed as, anode for chlorine evolution, while hydrogen was evolved on the ELAT electrode. One test was carried out with a Nafion membrane not subjected to pre-roughening, with a RuO2 coating obtained from a Ru metal target, as above disclosed. The cell was operated in the same way as during the blank experiment and, although the initial cell voltage at 10 kA/m²was slightly above 2V, it remained almost stable during 500 hours, reaching a final value of 2.10 V.

[0067] Another test was carried out with a sample of Nafion 115 membrane not subjected to pre-roughening, but wherein the active coating was obtained from a pellet of RuO2 as the target. The difference in the recorded cell voltages with respect to the sample obtained from the ruthenium metal target was within the experimental error, showing that the two ruthenium sources are practically equivalent. This assumption was confirmed by testing two subsequent one sided RuO2 IBAD coated membranes, wherein the Nafion membrane was subjected to the above-disclosed pre-roughening treatment prior to carrying out the IBAD procedure. In this case, however, when the cell was equipped with such two samples, always coupled to ELAT type cathodes, the cell voltage displayed at 10 kA/m²was identical, within the experimental error, to the initial voltage recorded in the blank test (1.85 V), and it remained almost stable during 500 hours, after which it reached a value of 1.92 V.

[0068] A final test was carried out with a two side coated pre-roughened Nafion 115 membrane, produced with the ruthenium metal target; in this case, no ELAT electrode was used, and hydrogen was evolved on one activated surf ace of the membrane used as the cathode, while chlorine was evolved at the anode side. A cell voltage of 1.87 V was detected at 10 kA/m², and it remained absolutely stable during 3000 hours, after which the test was discontinued.

[0069] It has thus been demonstrated that the initial performances of state-of-the-art ELAT electrodes in the membrane electrolysis of anhydrous HCl can be almost reached by a much cheaper technique, involving the direct deposition of catalyst on the membrane surface through the IBAD technique; moreover, the absence of a corrodible catalyst support renders such initial performance very stable in time, solving an issue which appears very critical when ELAT electrodes are employed, especially as chlorine evolving anodes.

[0070] Various modifications of the process and products of the invention may be made without departing from the spirit or scope thereof and it is to be understood that the invention is intended to be limited only as defined in the appended claims.