Solid polymer electrolyte and process for making same

Abstract

A solid polymer electrolyte membrane having a first surface and a second surface opposite the first surface, where the solid polymer electrolyte membrane has a failure force greater than about 115 grams and comprises a composite membrane consisting essentially of (a) at least one expanded PTFE membrane having a porous microstructure of polymeric fibrils, and (b) at least one ion exchange material impregnated throughout the porous microstructure of the expanded PTFE membrane so as to render an interior volume of the expanded PTFE membrane substantially occlusive; (c) at least one substantially occlusive, electronically insulating first composite layer interposed between the expanded PTFE membrane and the first surface, the first composite layer comprising a plurality of first carbon particles supporting a catalyst comprising platinum and an ion exchange material, wherein a plurality of the first carbon particles has a particle size less than about 75 nm, or less than about 50 nm, or less than about 25 nm.

Claims

1. A method of making a solid polymer electrolyte membrane for use interposed between a separate anode and cathode comprising the steps of (a) preparing an ink solution comprising a precious metal catalyst on a supporting particle and an ion exchange material, wherein preparing said ink solution comprises passing said ink solution through a high shear mixer; (b) providing a polymeric support having a first surface, a second surface opposite said first surface, and a microstructure of micropores; (c) applying a solution comprising an ion exchange resin to said first surface, thereby impregnating said microstructure with said ion exchange resin to form a substantially air occlusive, electronically insulating first composite layer, and (d) applying said ink solution to at least one of said first and second surfaces, thereby impregnating said microstructure with said ink solution to form a substantially air occlusive, electronically insulating second composite layer, thereby forming said solid polymer electrolyte membrane.

2. The method of claim 1 wherein the concentration of said precious metal catalyst based on weight percent of dry ion exchange material is between about 0.1% and 10%.

3. The method of claim 2 wherein the concentration of said precious metal catalyst based on weight percent of dry ion exchange material is between about 0.5% and 3%.

4. The method of claim 3 wherein the concentration of said precious metal catalyst based on weight percent of dry ion exchange material is about 1%.

5. The method of claim 3 wherein the concentration of said precious metal catalyst based on weight percent of dry ion exchange material is about 2.5%.

6. The method of claim 1 wherein step (a) further includes (a1) reducing the concentration of large particles in the ink.

7. The method of claim 6 wherein the step of reducing the concentration of large particles in the ink comprises filtering.

8. The method of claim 6 wherein the step of reducing the concentration of large particles in the ink comprises the use of a centrifuge.

9. The method of claim 1 wherein said high shear mixer is a microfluidizer.

10. The method of claim 1 wherein said high shear mixer is a rotor-stator mixer comprising at least one stage.

11. The method of claim 1 wherein said step (d) further includes (d1) applying said ink solution to a thin polymer film and (d2) applying at least one of said first and second surfaces to said ink solution on said thin polymer film.

12. The method of claim 11 wherein said thin polymer film comprises polyethylene, polyethylene terephthalate polypropylene, poly vinylidene chloride, polytetrafluoroethylene, polyesters, or combinations thereof.

13. The method of claim 12 wherein said thin polymer film further comprises a coating capable of enhancing the release characteristics of said polymer film.

14. The method of claim 11 wherein said step (c) further includes (c1) drying said support after impregnation of said ion exchange resin.

15. The method of claim 9 wherein said high shear mixer is a microfluidizer operating at a pressure between about 1,000 and about 25,000 psi.

16. The method of claim 14 wherein step (d) further includes (d1) drying said support after application of said ink solution.

17. The method of claim 14 wherein there is a further step after step (d) of heating said solid polymer electrolyte membrane at an elevated temperature.

18. The method of claim 17 wherein said elevated temperature is between about 100 degrees C. and about 175 degrees C.

19. The method of claim 18 wherein said elevated temperature is between about 120 degrees C. and about 160 degrees C.

20. The method of claim 19 wherein said solid polymer electrolyte membrane is held at said elevated temperature for between about 1 minute and about 10 minutes.

21. The method of claim 20 wherein said solid polymer electrolyte membrane is held at said elevated temperature for between about 3 minutes and about 5 minutes.

22. The method of claim 2 wherein said supporting particle comprises carbon.

23. The method of claim 22 wherein said precious metal catalyst comprises platinum.

24. The method of claim 23, wherein said polymer support comprises expanded polytetrafluoroethylene.

25. The method of claim 1, wherein: the supporting particle in step (a) is a carbon particle; and the polymeric support in step (b) is an expanded polytetrafluoroethylene membrane; and wherein the method further comprises the steps of (c1) drying said expanded polytetrafluoroethylene membrane after impregnation of said ion exchange resin; (d1) applying said ink solution to a thin polymer film; (d2) applying said second surface of said expanded polytetrafluoroethylene membrane to said ink solution on said thin polymer film; (d3) drying said expanded polytetrafluoroethylene membrane after impregnation of said ink solution and removing said thin polymer film.

Description

DESCRIPTION OF THE DRAWINGS

(1) The operation of the present invention should become apparent from the following description when considered in conjunction with the accompanying figure.

(2) FIG. 1 is a drawing illustrating several embodiment of the inventive solid polymer electrolytes.

(3) FIG. 2 is a drawing illustrating additional embodiments of the inventive solid polymer electrolytes.

(4) FIG. 3 is a drawing illustrating further embodiments of the inventive solid polymer electrolytes.

(5) FIG. 4 is a drawing illustrating yet additional embodiments of the inventive solid polymer electrolytes.

(6) FIG. 5 is a schematic of catalyst particles on a supporting particle.

(7) FIG. 6 schematically illustrates an embodiment of the inventive method for preparing the inventive solid polymer electrolytes.

(8) FIG. 7 is a drawing of a fuel cell that uses the inventive solid polymer electrolyte.

(9) FIG. 8 is a transmission electron micrograph of the material prepared in Comparative Example 2.

(10) FIG. 9 is a transmission electron micrograph of a cross-section of a portion the solid polymer electrolyte prepared in Example 8.

(11) FIG. 10 is a higher magnification transmission electron micrograph of the FIG. 9.

(12) FIG. 11 is a plot of the real versus imaginary components of the test described in Example 13 and Comparative Example 7.

DETAILED DESCRIPTION OF THE INVENTION

(13) In order to develop membranes that have a long-life in a fuel cell, the mechanisms of failure need to be understood. Without being held to any particular theory, it is known in the art that there are two major forms of membrane failure, chemical and mechanical. The latter has been addressed by various approaches, for example by the formation of composite membranes described by Bahar et al. in RE 37,707. Approaches to address the former have also been proposed, for example in GB 1,210,794 assigned to E. I. Du Pont de

(14) Nemours, Inc., where a chemical process to stabilize ionomers was described. Degradation, as observed by the concentration of fluoride ions in various ex-situ or in-situ fuel cell tests, can thus be reduced.

(15) The present invention involves a process for making, and a composition of, solid polymer electrolytes that is capable of reducing electrolyte degradation as observed by fluoride release rates from operating fuel cells. Inventors have discovered a composition of solid polymer electrolyte (SPE) that surprisingly reduces membrane degradation as observed by fluoride release rates, and gives a concomitant increase in membrane life. Inventors have discovered that when a plurality of very small particles (for example, less than about 75 nm) that are supporting a catalyst is dispersed in a substantially air occlusive, electronically insulating layer, preferably in an SPE that has high strength, unexpectedly long life is observed when the SPE is tested in a fuel cell.

(16) FIG. 1 shows a schematic of a three different embodiments of the inventive solid polymer electrolyte 10. SPE 10 typically is thin, less than 100 microns, preferably less than 75 microns, and most preferably less than 40 microns thick. It comprises an ion exchange material 11 that is able to conduct hydrogen ions at a high rate in typical fuel cell conditions. The ion exchange materials may include, but are not limited to compositions comprising phenol sulfonic acid; polystyrene sulfonic acid; fluorinated-styrene sulfonic acid; perfluorinated sulfonic acid; sulfonated Poly(aryl ether ketones); polymers comprising phthalazinone and a phenol group, and at least one sulfonated aromatic compound; aromatic ethers, imides, aromatic imides, hydrocarbon, or perfluorinated polymers in which ionic an acid functional group or groups is attached to the polymer backbone. Such ionic acid functional groups may include, but are not limited to, sulfonic, sulfonimide or phosphonic acid groups. Additionally, the ion exchange material 11 may further optionally comprise a reinforcement to form a composite membrane. Preferably, the reinforcement is a polymeric material. The polymer is preferably a microporous membrane having a porous microstructure of polymeric fibrils, and optionally nodes. Such polymer is preferably expanded polytetrafluoroethylene, but may alternatively comprise a polyolefin, including but not limited to polyethylene and polypropylene. An ion exchange material is impregnated throughout the membrane, wherein the ion exchange material substantially impregnates the microporous membrane to render an interior volume of the membrane substantially occlusive, substantially as described in Bahar et al, RE37,307, thereby forming the composite membrane.

(17) The SPE 10 of FIG. 1 also comprises a plurality of particles 14 supporting a catalyst, where the particles are dispersed in a substantially air occlusive, electronically insulating layer 13 adjacent to the surface. A plurality of the particles 14 supporting a catalyst have a size less than about 75 nm, or preferably less than about 50 nm. Such particles 14 may be agglomerated together in groups of two, three or even in larger groupings of many particles, though it is preferable that they are separated in smaller clusters of a few particles, and most preferably, as individual particles. The insulating layer 13 may be only on one side of the ion exchange material 11 (FIGS. 1a and 1b) or on both sides (FIG. 1c). Optionally, a second ion exchange material 12 may also be present (FIG. 1b) on the side opposite the electronically insulating layer 13. The composition of ion exchange material 12 may be the same as ion exchange material 11, or it may be of a different composition.

(18) FIGS. 2-4 schematically illustrate alternative approaches for the inventive solid polymer electrolyte. In FIG. 2, the solid polymer electrolyte 10 has a plurality of particles 14 supporting a catalyst within a composite membrane 21 consisting essentially of at least one expanded PTFE membrane having a porous microstructure of polymeric fibrils, and (b) a solid dispersion comprising a plurality of first carbon particles supporting a catalyst comprising platinum and at least one ion exchange material, wherein a plurality of the carbon particles has a particle size less than about 75 nm, and the solid dispersion is impregnated throughout the porous microstructure of the expanded PTFE membrane so as to render an interior volume of the expanded PTFE membrane substantially occlusive. Additionally, a substantially air occlusive, electronically insulating layer 13 may be adjacent to one (FIG. 2b) or both (FIG. 2c) surfaces. Optionally, a second ion exchange material 12 of the same, or of a different composition than used in 21 may also be present (FIG. 1b) on the side opposite the electronically insulating layer 13. Alternatively, ion exchange material 11, ion exchange material 12, composite membrane 21, and substantially air occlusive, electronically insulating layer 13 may also be present in various alternating arrangements, some examples of which are schematically in FIG. 3a-FIG. 3g and FIG. 4a-FIG. 4e.

(19) A schematic of the cross-section of the particles 14 supporting a catalyst used in the inventive materials is shown in FIG. 5. The particles 14 comprise a support material 51 onto which catalyst 52 has been deposited. Support material may comprise silica; zeolites; carbon; and oxides and carbides of the group IVB, VB, VIB VIIB, and VIII transition metals; and combinations thereof Carbon is a particularly preferable support material. They preferably have high surface area, and so should be small in size, less than 75 nm, or preferably less than 50 nm, or less than 25 nm. They may also optionally be porous. Catalyst 52 comprises metals, oxides or carbides known to be active catalytically for the oxidation of active species. These include, but are not limited to catalysts comprising precious metals, for example platinum, gold, palladium, rhodium, iridium, ruthenium and combinations thereof; and catalytically active oxide and carbides of the group NB, VB, VIB VIIB, and VIII transition metals; and combinations thereof. Particularly preferable catalysts are platinum metal or platinum metal alloys. The catalyst 52 is small in size to maximize its surface area and increase its effectiveness, preferably between about 1 nm and 10 nm in size.

(20) Use of such catalysts on support particles as described herein in any membrane reduces membrane degradation as observed by very low fluoride release rates during fuel cell operation. In order to achieve very long life in a fuel cell, a combination of a high SPE strength and a layer comprising a plurality of catalyst on a supporting particle should be present in the electrolyte. The strength of the membrane can be quantified using several approaches known in the art, but herein, we choose to quantify strength using a tensile test. The details are described more fully below, but four parameters are extracted from this test, the failure force, the tensile strength, the modulus and the stiffness. At least one of these must be above a critical value to achieve the very long electrolyte life described in this invention. The solid polymer electrolyte can achieve the high strength using any of the approaches known in the art to improve strength in polymer films, including but not limited to, adjusting processing to prepare high strength polymer films, for example by extrusion or stretching to orient the polymer film; reinforcing the film with inorganic or polymer particles; or by reinforcing with fabrics, porous or microporous inorganic or polymer films. Particularly preferably methods for preparing a strong solid polymer electrolyte are those taught by Bahar in '707, or by Hobson in '203, which use microporous ePTFE membranes to form composite electrolytes.

(21) An inventive method for preparing an air occlusive integral composite membrane has also been discovered. The method comprises the steps of (a) preparing an ink solution comprising a precious metal catalyst on a supporting particle and an ion exchange material; (b) providing a polymeric support having a microstructure of micropores; (c) applying either the ink solution or a solution comprising an ion exchange resin to the polymeric support; (d) optionally, repeating step (c); wherein at least one application in step (c) or (d) uses the ink solution. In this application, an ink is considered to be a solution containing a catalyst on a supporting particle that is dispersed in a solvent. The ink solution preferably also contains an ion exchange polymer. Solvents used in the ink are those generally known in the art, including but not limited to alcohols, such as ethanol and propanol, or other organic solvents. The preparation of the ink solution preferably uses a high shear mixer, where the high shear mixer may include, but is not limited to, microfluidizers, and rotor-stator mixers comprising at least one stage. Particularly preferable high shear mixers are microfluidizers capable of operating at pressures between 5,000 psi and 25,000 psi. The ink is preferably very well mixed, which may be accomplished by one, two, three or more passes through the high shear mixer. The concentration of the precious metal catalyst in the ink is between about 0.1% and about 20% by dry weight of the ion exchange material, and preferably between about 0.5% and about 3%. This ink may be prepared in one, two or more separate steps if desired. If it is prepared in two or more steps, a more concentrated solution is made in the first step, and subsequent steps are dilutions with ion exchange material to arrive at the final desired concentration. When more than one step of preparing the ink is used, the high shear mixing step described above may be used in one or more of the ink preparation steps. If desired, the first step in a multi-step ink preparation process may be accomplished in advance of the succeeding steps, in which case the ink may be stored for a period of time. If such a concentrated ink is stored for longer than about 30-60 minutes, then the high shear mixing step is preferably repeated at least once, and more preferably two or three times before any subsequent dilution steps needed to arrive at the final ink used for subsequent processing.

(22) Additional steps to remove large agglomerates in the ink solution may also be performed, if desired, at any stage during the ink preparation. Such steps may include, but are not limited to, filtering and using a centrifuge. In either case, the number of large particles removed can be controlled. In the former, by the particular filter chosen; in the latter, by the length of time the sample is centrifuged, and/or the speed of the centrifuge. The centrifuge speed may be varied from between a few hundred rpm, to many thousand rpm, with the higher speeds being preferable. The time to centrifuge may vary from a few minutes to an hour or longer. Shorter times at higher speeds, for example less than 30 minutes at 3000-5000 rpm, are preferable to reduce processing times.

(23) The ion exchange material in the ink may be any known in the art, for example those described above for ion exchange material 11. The precious metal catalyst on a supporting particle may be any of those described above for FIG. 5, 52 and 51, respectively.

(24) The polymeric support having a microstructure of micropores, may be any such material known in the art, including but not limited to microporous polyethylene, polypropylene or polytetrafluoroethylene. A particularly preferable polymeric support is expanded PTFE, such as those described in U.S. Pat. No. 3,953,566 to Gore, in U.S. Pat. No. 6,613,203 Hobson et. al., or in U.S. Pat. No. 5,814,405 to Branca, et. al. Preferably, the polymeric support should be sufficiently strong and/or heavy so that the final solid polymer electrolyte has a failure force (defined more fully below) of greater than 115 g.

(25) The ink solution or a solution comprising an ion exchange resin may be applied to the polymeric support using any process known in the art, including but not limited to the process described in U.S. Pat. No. RE37,707 to Bahar et. al. Another embodiment of the method of the invention for applying the ink to the polymeric support is shown in FIG. 6. In this embodiment, an ink is applied to a thin polymer film 64 using any means known to one of ordinary skill in the art, for example using a pump, syringe 63 or such. The ink is prepared as described above, so may be prepared in a multistep process starting with a concentrated ink that is subsequently diluted, or directly in one step to obtain the desired catalyst concentration in the ink. The applied ink 62 is then spread into a thin layer 66 using any means known in the art for making a thin liquid layer, including but not limited to a draw bar or meyer bar, shown schematically in FIG. 6 as 61. Subsequently, the polymeric support 65 having a microstructure of micropores is placed on the liquid layer 66 and allowed to imbibe. The thin polymer film 64 comprises polyethylene, polyethylene terephthalate polypropylene, poly vinylidene chloride, polytetrafluoroethylene, polyesters, or combinations thereof, and may further comprise a coating of a release material, for example a fluoropolymer compound, to enhance the release of the final product from the polymer film. After the film is completely imbibed, it is allowed to dry, and may optionally be heated to decrease the drying time. Such heating, shown schematically in FIG. 6 as 67, may be accomplished with any means known in the art, including but not limited to forced air heaters, ovens, infrared driers and the like. The process may be repeated if desired, using the same or a different ink, or the same or a different ion exchange resin.

(26) When the imbibing steps are completed, an additional heating step at an elevated temperature may optionally be applied using an oven, infrared heater, forced air heater or the like. The temperature of this heating step is between about 100° C. and about 175° C. and preferably between about 120 degrees C. and about 160° C. The solid polymer electrolyte is held at the elevated temperature for between about 1 minute and about 10 minutes, and preferably for between about 1 minutes and about 3 minutes. Finally, the completed solid polymer electrolyte membrane is cooled, and removed from the thin polymer film before use. The removal may be accomplished by simply pulling the SPE off the thin polymer film, either in air or in water.

(27) As is well understood by one of ordinary skill in the art, the process described above and in FIG. 6 can by automated using roll goods, and automated pay-off and collection systems so that each step is accomplished in a continuous fashion, and the final product is a roll of solid polymer electrolyte supported on a thin polymer film.

(28) The solid polymer electrolyte of the instant invention may also be used to form a catalyst coated membrane (CCM) using any methods known in the art. In FIG. 7, the CCM 70 comprises an anode 71 of a catalyst for oxidizing fuel, a cathode 72 for reducing an oxidant, and the solid polymer electrolyte 10 described above interposed between the anode and cathode. The anode and cathode may be prepared using any of the procedures known in the art including but not limited to physical or chemical deposition, either on a supporting particle, or directly on the SPE, or from a catalyst-containing ink solution containing the catalysts that is deposited either directly on the SPE, or on a film that is subsequently used in a lamination step to transfer the electrode to the SPE.

(29) The anode and cathode electrodes comprise appropriate catalysts that promote the oxidation of fuel (e.g., hydrogen) and the reduction of the oxidant (e.g., oxygen or air), respectively. For example, for PEM fuel cells, anode and cathode catalysts may include, but are not limited to, pure noble metals, for example Pt, Pd or Au; as well as binary, ternary or more complex alloys comprising the noble metals and one or more transition metals selected from the group Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Ru, Rh, Ag, Cd, In, Sn, Sb, La, Hf, Ta, W, Re, Os, Ir, Tl, Pb and Bi. Pure Pt is particularly preferred for the anode when using pure hydrogen as the fuel. Pt-Ru alloys are preferred catalysts when using reformed gases as the fuel. Pure Pt is a preferred catalyst for the cathode in PEMFCs. The anode and cathode may also, optionally, include additional components that enhance the fuel cell operation. These include, but are not limited to, an electronic conductor, for example carbon, and an ionic conductor, for example a perfluorosulfonic acid based polymer or other appropriate ion exchange resin. Additionally, the electrodes are typically porous as well, to allow gas access to the catalyst present in the structure.

(30) A fuel cell 73 can also be formed from the instant invention. As shown in FIG. 7, such PEM fuel cells 73 comprise the CCM 70 and may optionally also include gas diffusion layers 74 and 75 on the cathode 72 and anode 71 sides, respectively. These GDM function to more efficiently disperse the fuel and oxidant. The fuel cell may optionally comprise plates (not shown in FIG. 7) containing grooves or other means to more efficiently distribute the gases to the gas diffusion layers. As is known in the art, the gas diffusion layers 74 and 75 may optionally comprise a macroporous diffusion layer as well as a microporous diffusion layer. Microporous diffusion layers known in the art include coatings comprising carbon and optionally PTFE, as well as free standing microporous layers comprising carbon and ePTFE, for example CARBEL® MP gas diffusion media available from W. L. Gore & Associates. The fluids used as fuel and oxidant may comprise either a gas or liquid. Gaseous fuel and oxidant are preferable, and a particularly preferable fuel comprises hydrogen. A particularly preferable oxidant comprises oxygen.

(31) The following test procedures were employed on samples which were prepared in accordance with the teachings of the present invention.

TEST PROCEDURES

(32) Cell Hardware and Assembly

(33) For all examples, standard hardware with a 23.04 cm.sup.2 active area was used for membrane electrode assembly (MEA) performance evaluation. This hardware is henceforth referred to as “standard hardware” in the rest of this application. The standard hardware consisted of graphite blocks with triple channel serpentine flow fields on both the anode and cathode sides. The path length is 5 cm and the groove dimensions are 0.70 mm wide by 0.84 mm deep.

(34) Two different cell assembly procedures were used. In the first procedure, designated as procedure No. 1, the gas diffusion media (GDM) used was a microporous layer of Carbel® MP 30Z placed on top of a Carbel® CL gas diffusion layer (GDM), both available from W. L. Gore & Associates, Elkton, MD. Cells were assembled with two 10 mil UNIVERSAL® ePTFE gaskets from W. L. Gore & Associates, having a square window of 5.0 cm×5.0 cm, two 2.0 mil polyethylene naphthalate (PEN) films (available from Tekra Corp., Charlotte, N.C.) gaskets hereafter referred to as the spacer, and two 1.0 mil polyethylene naphthalate (PEN) films hereafter referred to as the sub-gasket. The sub-gasket had an open window of 4.8×4.8 cm on both the anode and cathode sides, resulting in a MEA active area of 23.04 cm.sup.2.

(35) In the second procedure, designated as procedure No. 2, assembly materials were the same as procedure No. 1, with the exceptions that the GDM used was Carbel® CL GDM alone, and no spacers were incorporated.

(36) All the cells were built using spring-washers on the tightened bolts to maintain a fixed load on the cell during operation. They are referred to as spring-loaded cells. The assembly procedure for the cells was as follows: 1. The 25 cm.sup.2 triple serpentine channel design flow field (provided by Fuel Cell Technologies, Inc, Albuquerque, N. Mex.) was placed on a workbench; 2. One piece of 10 mil ePTFE gasket with a 2.0 mil PEN spacer was placed on anode side of the flow field; 3. One set of the GDM was placed inside the gasket so that the MP-30Z layer was facing up; 4. The window-shaped sub-gasket of PEN sub-gasket sized so it slightly overlapped the GDM on all sides was placed on top of the GDM; 5. The anode/membrane/cathode system was placed on top of the sub-gasket with anode-side down; 6. Steps (2) through (4) were repeated in reverse order to form the cathode compartment. The gasket used on the cathode side was the same as that used on the anode side. 7. There are total of eight bolts used in each cell, all bolts had spring washers, Belleville disc springs, purchased from MSC Industrial Supply Co. (Cat# 8777849). The bolts were then tightened to a fixed distance that previously had been established to provide a compressive pressure of 100-120 psi in the active area. Compression pressure was measured by using Pressurex® Super Low Film pressure paper from Sensor Products, Inc., East Hanover, N.J.
Fuel Cell Life Testing

(37) Because the inventive membranes typically last a very long time (thousands of hours) under normal fuel cell operating conditions, two different types of accelerated test protocols were developed to establish membrane lifetimes. These protocols, identified as Test Protocol 1 and Test Protocol 2, are described more fully below.

(38) Test Protocol 1

(39) Materials to be tested were prepared as outlined below in the examples, and then assembled into a cell using the procedure outlined above. The cell was connected to a test station, conditioned, and then the test was started under test temperature and pressure as outlined below. The assembled cells were tested in fuel cell test stations with GlobeTech gas units 3-1-5-INJ-PT-EWM (GlobeTech, Inc., Albuquerque, N. Mex.), and Scribner load units 890B (Scribner Associates, Southern Pines, N.C.). The humidification bottles in these stations were replaced by bottles purchased from Electrochem Corporation (Woburn, Mass.). The humidity during testing was carefully controlled by maintaining the bottle temperatures, and by heating all inlet lines between the station and the cell to four degrees higher than the bottle temperatures to prevent any condensation in the lines. In all cases the inlet and/or outlet relative humidity of the anode and/or cathode was measured independently using dew point probes from Vaisala (Vantaa, Finland) to ensure the input hydrogen and air were humidified to desired relative humidity (RH) at the testing temperatures.

(40) The cells were first conditioned at a cell temperature 80° C. with 100% relative humidity inlet gases on both the anode and cathode. The outlet gas pressure of both sides was controlled to be 15 psig. The gas applied to the anode was laboratory grade hydrogen supplied at a flow rate of 1.3 times greater than what is needed to maintain the rate of hydrogen conversion in the cell as determined by the current in the cell (i.e., 1.3 times stoichiometry). Filtered, compressed and dried air was supplied to the cathode humidification bottle at a flow rate of 2.0 times stoichiometry.

(41) The cells were conditioned for 4 hours. The conditioning process involved cycling the cell at 80° C. between a set potential of 600 mV for 30 seconds, 300 mV for 30 seconds and 950 mV for 5 seconds for 4 hours. Then a polarization curve was taken by controlling the applied potential beginning at 600 mV and then stepping the potential in 50 mV increments downwards to 400 mV, then back upward to 900 mV in 50 mV increments, recording the steady state current at every step. The open circuit voltage was recorded between the potential steps of 600 mV and 650 mV.

(42) After the above procedure, the cells were set to the life-test conditions. This time was considered to be the start of the life test, i.e., time equal to zero for all life determinations. Specific test conditions in this protocol were (Table 2): cell temperature of 95° C., 50% RH for both hydrogen and air, with a stoichiometry of 1.3 and 2.0, respectively. Outlet pressure was 25 psig in all cases. The current density of the cells in Protocol No. 1A and Protocol No. 1B was controlled to be 100, and 800 mA/cm.sup.2, respectively.

(43) TABLE-US-00001 TABLE 2 Operation Conditions for Accelerated Chemical Degradation Tests Cell Gas Current Outlet Pressure Protocol Temp. Gas Type Inlet RH (%) Stoichiometry Density (anode/cathode) No. (° C.) (anode/cathode) (anode/cathode) (anode/cathode) (mA/cm.sup.2) (psig) 1A 95 H.sub.2/Air 50/50 1.3/2.0 100 25/25 1B 95 H.sub.2/Air 50/50 1.3/2.0 800 25/25
Test Protocol 2

(44) In test Protocol 2, the materials were prepared as described below in the examples, and assembled into cells as described above. The cells were then conditioned, and subsequently tested using the procedure outlined more fully below. Life of the membrane was determined using the physical pin-hole test described below.

(45) The test stations used for this protocol were fuel cell test stations with Teledyne Medusa gas units Medusa RD-890B-1050/500125 (Teledyne Energy Systems, Hunt Valley, Md.), and Scribner load units 890B. The gas units were modified with additions of solenoid valves from Parker outside of the humidification bottles. These valves control directions of gas flow so that the cells can be tested in wet and dry cycles.

(46) The conditioning procedure used in this protocol was as follows: the cells were first conditioned at a cell temperature 70° C. with fully humidified (100% RH) inlet gases. The gas applied to the anode was laboratory grade hydrogen supplied at a flow rate of the greater of 150 cc/min or 1.2 times greater than what is needed to maintain the rate of hydrogen conversion in the cell as determined by the current in the cell (i.e., 1.2 times stoichiometry). Filtered, compressed and dried air was supplied to the cathode at a flow rate of the greater of 650 cc/min or two times stoichiometry. Then, the cells were continuously cycled at 70° C. by fixing a set potential of 600 mV for 45 seconds, followed by open circuit voltage (OCV) for 30 seconds, 300 mV for 60 seconds, and finally OCV for 30 seconds. This cycling was repeated continuously for 10 hours. Then a polarization curve was taken by controlling the applied potential beginning at 600 mV for 8 minutes and then stepping through the following potentials and times intervals: 500 mV for 8 minutes, 400 mV for 8 minutes, 450 mV for 8 minutes, 550 mV for 8 minutes, 650 mV for 8 minutes, 750 mV for 8 minutes, 850 mV for 6 minutes, 900 mV for 4 minutes, 800 mV for 6 minutes, 700 mV for 8 minutes, 600 mV for 8 minutes, recording the steady state current at every step. Then the following current densities were applied in steps: 100 mA/cm.sup.2 for 3 minutes, 500 mA/cm.sup.2 for 3 minutes, 800 mA/cm.sup.2 for 3 minutes, and finally the cell was left at open circuit potential for 2 minutes, recording the steady state potential at every step.

(47) After the above procedure, the cells remain at 700 mV for between 0 and 24 hours. Then the cell was pressured to 25 psig. The cells were further conditioned at a cell temperature of 80° C. with dry relative humidity inlet gases on both the anode and cathode. The hydrogen gas applied to the anode was at a utilization of 0.83 with a minimum flow rate of 50 cm.sup.3/min. Filtered, compressed and dried air was supplied to the cathode at a flow rate of the greater of 100 cm.sup.3/min or 4.0 times stoichiometry. The current was set to 200 mA/cm.sup.2 for 30 minutes and the potential was recorded. Then the cell was changed to 100% RH inlet gases for 90 seconds. An open circuit voltage decay measurement was then initiated by stopping the gas flow on the cathode and removing the load. The resulting voltage was measured every 3 seconds for 1 minute as the oxygen on the cathode is consumed by hydrogen crossing over from the anode to cathode. This is a measure of membrane health at beginning of life, under pressure. The flow on the anode was then set to 150 cc/min and 650 cc/min on the cathode for 10 seconds under no load. Then the load is applied at 800 mV for 20 seconds. Finally the anode flow was set to the greater of 50 cc/min or 1.2 times stoichiometry and cathode flow is set to the greater of 100 cc/min or four times stoichiometry. The current was set to 200 mA/cm.sup.2 for 30 seconds and the potential is recorded. This ends the initial conditioning.

(48) After initial conditioning and diagnostics, the MEA was tested under the following test conditions. The cell temperature remained at 80° C. The cell was pressurized on the anode with hydrogen and on the cathode with air to 25 psig. The hydrogen flow rate on the anode was at 1.2 times stoichiometry with a minimum flow rate of 50 cm.sup.3/min. Air was supplied to the cathode at a flow rate of the greater of 100 cm.sup.3/min or 4.0 times stoichiometry. The current was set to 200 mA/cm.sup.2 and potential was recorded. The inlet gas was cycled from by pass of the humidification bottles to flow through the humidification bottles. This cycling is controlled by the solenoid valves that switch every 45 seconds. The result was an inlet humidification that rises and falls every 45 seconds. The inlet gases reach the following maximum and minimum humidification:

(49) Anode wet condition 61 ° C. dew point=44% RH

(50) Anode dry condition 31° C. dew point=10% RH

(51) Cathode wet condition 75° C. dew point=80% RH

(52) Cathode dry condition 14-20° C. dew point=3-5% RH

(53) During the test, the open circuit voltage (OCV) decay was measured two times every hour. The first measurement was done after a 45 second wet cycle, and the second 30 minutes later after a 45 second dry cycle. These measurements were made under pressure and automatically by the test station, as described above except that the air flow to the cathode was shut off for 3 minutes, instead of 1 minute.

(54) Chemical Degradation Rate

(55) For all the tests the amount of fluoride ions released into the product water was monitored as a means to evaluate chemical degradation rate. This is a well-known technique to establish degradation of fuel cell materials that contain perfluorosulfonic acid ionomers. Product water of fuel cell reactions was collected at the exhaust ports throughout the tests using PTFE coated stainless steel containers. The collected water was then concentrated about 20 fold (for example, 2000 ml to 100 ml ) in PTFE beakers heated on hot plates. Before concentration, 1 ml of 1M KOH was added into the beaker to prevent evaporation of HF. Fluoride concentration in the concentrated water was determined using an F.sup.−-specific electrode (ORION® 960900 by Orion Research, Inc.). Fluoride release rate in terms grams F.sup.−/cm.sup.2-hr) was then calculated.

(56) Membrane Life Measurement

(57) The life of the membrane was established by determining the presence of flaws in the membrane that allow hydrogen to cross through it. In this application, this so-called hydrogen cross-over measurement was made using a flow test that measures hydrogen flow across the membrane. Because this test is somewhat tedious, and may itself weaken the membrane, it was only performed when there was an indication that the integrity of the membrane was questionable. The membrane integrity was thus first evaluated during testing using an OCV decay measurement performed at ambient pressures. In Test Protocol 1, this measurement was carried out while the cell remained as close as possible to the actual life test condition. In Test Protocol 2 this measurement was performed under 100% RH conditions. This ambient OCV decay test was performed periodically as indicated by the performance of the cell. Typically, it was performed less frequently near the beginning of cell life (e.g., once a week), and more frequently the longer the cell operated (e.g., as often as once per day toward the end of life). Details of the measurement were as follows: 1. The cell was set at 0.6V, anode and cathode minimum flow rate to be 800 cc/min. and 0 cc/min, respectively; 2. The outlet pressure of anode and cathode side was reduced to 2.0 and 0 psig, respectively; 3. The cell was then taken off load while remaining at the test temperature; meanwhile, outlet flow of the cathode side was shut off by a valve; 4. The OCV value was recorded every second for 180 seconds; 5. The decay in the OCV during this measurement was examined. If this decay was significantly higher than previously observed, e.g., when the open circuit voltage value decayed to less than 250 mV in less than 30 seconds, a physical flow check was initiated to determine if the membrane had failed; 6. If the decay was close to that of the previous measurement, the life testing was resumed. When a physical flow check was indicated, it was performed as follows: 7. The cell was taken off load, and set at open circuit condition while maintaining the cell temperature and RH conditions at the inlets. The gas pressure of the cell was then reduced to ambient pressure on both anode and cathode sides. 8. The gas inlet on the cathode was disconnected from its gas supply and capped tightly. The cathode outlet was then connected to a flow meter (Agilent® Optiflow 420 by Shimadzu Scientific Instruments, Inc., Columbia, Md.). The anode inlet remained connected to the H.sub.2 supply and anode outlet remained connected to the vent. 9. The anode gas flow was increased to 800 cc/min, and the anode outlet pressure was increased to 2 psi above ambient pressure. 10. In Test Protocol 2, the H.sub.2 gas is supplied at 0% RH for 30 minutes. 11. The amount of gas flow through the cathode outlet was measured using the flow meter. 12. A failure criterion of 2.5 cc/min was established, so that when the measured gas flow of H.sub.2 was greater than this value, the membrane was identified as having failed. 13. If the criterion for failure was met the test was stopped, and the membrane life was recorded as the number of hours the cell had been under actual test conditions when it failed the physical flow check (>2.5 cc/min). If the criterion for failure was not met, the cell was returned to testing.
Mechanical Property Measurement

(58) Certain membranes were subjected to mechanical testing at room conditions of 21° C. and 60% RH. A dynamic mechanical analyzer (DMA) (TA Instruments, Wilmington, Del.) mode RSA3 was used. Each membrane type tested in machine as well as transverse directions. The membrane was die cut to a rectangular shape with a width of 4.8 mm, and a length of 50 mm. The grip gap was set to be 15 mm, and the membrane sample was pulled at the rate of 0.5 mm/s until failure. Tension force during the sample elongation was recorded, and the maximum value before sample failure was regarded as the failure force. During data analysis, values of force were plotted against elongation. The slope of the linear portion of the curve, specifically, from 0 to 0.04 elongation portion of the curve was calculated as the stiffness. Cross sectional area for each sample was calculated using sample width, i.e. 4.8 mm, times sample thickness. Values of failure force and stiffness were divided by sample's cross section area to obtain strength and modulus values, respectively. The lesser values of the mechanical properties from the transverse and machine direction are those reported here as Failure Force, Stiffness, Strength and Modulus.

(59) Platinum Loading Measurement

(60) To confirm that the platinum used to prepare the inventive solid polymer electrolytes had not been lost in processing, the amount of platinum in the membranes was measured using a bench-top x-ray fluorescence unit (XRF from SPECTRO TITAN, Kleve, Germany) pre-calibrated to display Pt content in units of mg Pt per cm.sup.2 surface area. Three separate measurements of the concentration of sections of the as-prepared solid polymer electrolytes were taken by placing the as-prepared inventive solid polymer electrolyte in the unit and recording the displayed values. The values reported in the examples below are the average values of the three measurements taken for each material. In all cases, the measured amounts were equal to the expected amounts within experimental error of the measurement.

(61) Transmission Electron Microscopy and Interparticle Spacing Measurement

(62) In order to observe distribution of supported catalysts inside the membrane, transmission electron microscopy (TEM) was performed on cross sections of selected inventive solid polymer electrolytes. A section of the solid polymer electrolyte was embedded in Spun® epoxy resin and cured at 60° C. for eight hours. The embedded sample was first trimmed with a razor blade and then thin sectioned at room temperature using a Diatome diamond knife on a Leica Ultracut UCT ultramicrotome. The microtome was set to cut 75 nm thick sections which were collected on 300 mesh copper grids. TEM was performed using a JEM 2010 Field Emission TEM, at 200 KV at various magnifications. The interparticle spacing between support particles was determined as follows: a micrograph representative of the observed microstructure was obtained at a magnification where a large number of the plurality of support particles could be seen, at least 20, and preferably at least 50. The distance between 15 different pairs of surrounding neighbors of the plurality of support particles chosen at random was measured. The interparticle spacing was calculated as the average of the 15 measurements. To determine the interparticle spacing between catalyst particles, the following procedure was used: a micrograph representative of the catalyst particles on the support particle was obtained at a magnification where at least one support particle could be observed, and a plurality of catalyst particles on the support could be seen, at least 4, and preferably 6 or more. The distance between 6 to 10 different pairs of the plurality of support particles chosen at random was measured, and the interparticle spacing was calculated as the average of the measurements.

(63) Without intending to limit the scope of the present invention, the solid polymer electrolytes and method of production of the present invention may be better understood by referring to the following examples

EXAMPLES

(64) In the examples below, three different ion exchange materials were used to prepare solid polymer electrolytes. The first material, identified herein as Type 1, was prepared according to the teachings of Wu, et. al in U.S. Patent Application 20030146148, Example 5 except the reactants were adjusted to produce a product with equivalent weight of about 920.

(65) This polymer had a melt flow index (MFI) that was typically 6±2 g/10 min with a range between 2 and 12. The MFI was measured by placing a 2160 gram weight onto a piston on a 0.8 cm long die with a 0.20955 cm orifice, into which 3-5 grams of as-produced polymer had been placed. Three separate measurements of the weight of polymer that flowed through the orifice in 10 minutes at 150° C. was recorded. The MFI in g/10 min was calculated as the average weight from the three measurements times 3.333. To make the ion exchange material more stable, this product was treated with 500 kPa fluorine gas at 60° C. in one five-hour cycle and three four-hour cycles, each one separated by an evacuation step, essentially according to the teachings in GB 1,210,794. The polymer was subsequently extruded, pelletized and acidified using procedures standard in the art. Then it was made into a dispersion by forming a solution of 20%-30% of the acid form of the Type 1 polymer, 10-20% deionized water, and balance alcohol in a glass-lined pressure vessel. The vessel was sealed, and the temperature was raised to 140° C. at a rate slow enough to maintain the pressure at less than 125 psi. It was held at 140° C. and about 125 psi for 2.5 hours. Then, a final solution was obtained by adding sufficient water to produce a solution consisting of approximately 20% solids, 20% water and 60% alcohol.

(66) A second ion exchange material, Type 2, was prepared in the same way as Type 1, but the fluorine gas treatment of the ion exchange polymer was effected at 135° C. in 500 kPa of 20% fluorine/80% nitrogen for two 4 hour periods and two 6 hour periods. The acid form of this polymer was formed into a dispersion as described above for the Type 1 polymer except the temperature and pressure during the solution preparation process was 160° C. and 210 psi. The MFI of this polymer was typically 4.4 g/10 min with a range between 2 and 12.

(67) The final polymer, Type 3, was prepared as described for Type 1 but it had an MFI of about 0.9 g/10 min. It was treated with fluorine gas in the same fashion as Type 2. The polymer was then made into a dispersion by forming a solution of 10% of the acid form of the polymer, and the balance ethanol in a glass-lined pressure vessel. The vessel was sealed, and the temperature was raised to 140° C. at a rate slow enough to maintain the pressure at less than 125 psi, and held at 140° C./125 psi for 2.5 hours. Then, a final solution was obtained by adding a weight of water approximately equal to the polymer weight and then concentrating the solution by evaporating the solvent at room temperature. The final solution then consisted of approximately 20% solids, 20% water and 60% alcohol.

Example 1

(68) In Example 1, a solid polymer electrolyte membrane was prepared as follows:

(69) A concentrated catalyst ink consisting of platinum on a carbon support (type V11-D50 catalyst, Englehard Corporation, Iselin, N.J.) at a 1:1 weight ratio (35% water by weight), Type 1 ion exchange material, and normal propanol in the following approximate ratios, respectively, 8.54%, 4.27%, and 87.19% was prepared. This was accomplished as follows. A slurry was prepared by mixing a portion of the n-propanol with the catalyst powder in an 30 liter glass reactor

(70) (H. S. Martin, Inc., Vineland, N.J.) after evacuating it, and refilling with nitrogen. Subsequently, the slurry was pumped into a 50 liter vessel where agitation was supplied for 20 minutes by a rotor/stator agitator (Model AX200 by Silverson Machines Inc., Longmeadow, Mass.) while the solution was recirculated through a ISG motionless static mixer (Charles Ross & Sons, Hauppauge, N.Y.). To this slurry, the ionomer was added continuously over about 45 minutes. The solution of ionomer, solvent and catalyst was further mixed in the same container for an additional 30 minutes. Then, the solution was recirculated through a Model M-700 Microfluidizer (Microfluidics, Newton, Mass.) at 10,000 psig for 45 minutes. Finally, the solution was further mixed using the Silverson mixer with recirculation for an additional 20 minutes. The final concentrated ink solution was pumped into a holding tank, the system flushed with rinse solvent that was subsequently also pumped into the holding tank. The solution was stirred continuously for a five day period with a low shear propeller agitation system and stored in a plastic container for a period of time ranging from a few days to a few weeks. Immediately before use, this ink was passed through a Microfluidizer at 19,000 psig three times. It was then stirred with a magnetic stir bar until use, generally within about 30 minutes.

(71) An inventive solid polymer electrolyte membrane was prepared as follows. First, an expanded polytetrafluoroethylene (ePTFE) membrane was prepared with mass per area of 7.0 g/m.sup.2, thickness of 20 microns, and porosity of at least 85%, and a longitudinal matrix tensile strength of about 67 MPa, and a transverse matrix tensile strength of about 76 MPa using the teachings of U.S. Pat. No. 3,953,566 to Gore. The ink prepared above was then diluted with Type 1 ion exchange material to give a concentration of 0.8% platinum based on weight percent of dry ionomer solids. This ion exchange material solution was coated on a polyethylene naphthalate (PEN) film stretched onto a glass plate using a drawdown blade on which the coating gap can be adjusted between 1 and 10 mil. For this first coating, the gap was adjusted to 0.0038 inches (0.00965 cm). The ePTFE membrane was then stretched over the wet coating and allowed to infiltrate. After infiltration, it was dried for 20-60s with a hair drier. Then, a second coating of the same ion exchange material solution was applied with a 0.0019 inch (0.00483 cm) gap set on the draw bar. The second coating was then also dried with a hair drier for 20-60 s. This membrane was placed in a 160° C. air furnace for three minutes and then removed to cool. The membrane was then removed from the PEN backer being careful not to stretch it severely. The measured platinum loading of this membrane was 0.015 mg/cm.sup.2, and the final thickness of the solid polymer electrolyte was 18 microns.

(72) The mechanical properties of a section of this solid polymer electrolyte were tested using the procedures described above. The results for the Failure Force, Stiffness, Strength and Modulus are shown in Table 6.

(73) Another section of the completed solid polymer electrolyte was placed between two PRIMEA® 5510 electrodes (available from Japan Gore-Tex, Inc., Tokyo, Japan) with 0.4 mg Pt/cm.sup.2 loading in the each electrode. This sandwich was placed between platens of a hydraulic press (PHI Inc, Model B-257H-3-MI-X20) with heated platens. The top platen was heated to 180 degrees C. A piece of 0.25″ thick GR® sheet (available from W. L. Gore & Associates, Elkton, Md.) was placed between each platen and the electrode. 15 tons of pressure was applied for 3 minutes to the system to bond the electrodes to the membrane. This MEA was assembled into a fuel cell as described above, and tested under Test Condition 1A. The Lifetime and Fluoride Release Rate were measured, and results are shown in Table 2.

Example 2

(74) To illustrate the importance of the mechanical properties of the solid polymer electrolytes to the inventive materials herein, a material was made that is otherwise identical to an embodiment of the inventive materials, but has weaker mechanical properties. It thus has low fluoride release rates indicative of a chemically stable membrane, but its life is not as long as Example 1 because the polymer electrolyte membrane is not as strong as that formed in Example 1. The solid polymer electrolyte of this example was prepared as follows. First, an ePTFE membrane was prepared according to the teachings of Gore in '566 with an average mass per area of about 3.3 g/m.sup.2, a thickness of about 7.8 microns, an average ball burst strength of about 1.18 lbs, and an average Frazier number of about 42 ft3/min/ft2 at 0.5 inches of water. The ball burst is a standard test (see for example, U.S. Pat. No. 5,814,405 to Branca, et. al.) performed on porous membranes that measures the relative strength of a sample of membrane by determining the maximum load at break. A single layer of membrane is challenged with 1 inch diameter ball while being clamped and restrained in a ring of 3 inch inside diameter. The membrane is placed taut in the ring and pressure applied against it by the steel ball of the ball burst probe. Maximum load is recorded as “Ball Burst” in pounds.

(75) An ink was prepared as described in Example 1 using ion exchange material Type 3 to give a concentration of 0.8% platinum based on weight percent of dry ionomer solids. The as-prepared ink was passed through the Microfluidizer three consecutive times with a pressure setting of 19,000 psi. Then the solid polymer electrolyte was prepared as follows: for the first coating, a #44 Meyer Bar was used to coat onto a PEN film stretched tight over a glass plate using the prepared ink. The ePTFE membrane was then stretched over the wet coating and allowed to infiltrate. After infiltration, it was dried for 20-60s with a hair drier. Then, a second coating using the ink solution prepared above was applied with a #22 Meyer Bar. The second coating was then also dried with a hair drier for 20-60 s. This membrane was placed in a 160° C. air furnace for three minutes and then removed to cool. The membrane was then removed from the backer in room temperature deionized water. The measured platinum loading of this membrane was 0.017 mg/cm.sup.2, and its final thickness was 20 microns.

(76) A section of the same material was made into an MEA as described above, and assembled in a fuel cell. Testing using Test Protocol 1A showed Lifetime of 78 h and Fluoride Release Rate of 4.5×10.sup.−8. This is to be compared, for example, with the Lifetimes from Example 1 of nearly an order of magnitude higher (718 h).

(77) The mechanical properties were measured on a different section of this solid polymer electrolyte. The results (Table 6) showed that it was weaker than Examples 1, 5 and 7, having for example, a Failure Force of 106 g. This illustrates that although all the inventive materials have low fluoride release rates, the combination of a layer containing a catalyst of a supporting particle, and a strong solid polymer electrolyte are essential for long Lifetimes.

Example 3

(78) To confirm the importance of the mechanical properties together with the composite layer comprising a catalyst on a supporting particle, another solid polymer electrolyte was prepared that had a composite layer of catalyst on supporting particle but had low mechanical properties. This was done as follows. A Type 1 ion exchange material with 0.8% Pt to dry ionomer weight was prepared using the procedures of Example 1. The drawdown bar was set to 0.025 cm(0.010 inches) and only one coating was done directly onto a glass substrate (no polymer film). No ePTFE was used. After drying, the solid polymer electrolyte was removed in room temperature water. The resulting solid polymer electrolyte was 20 to 26 microns thick.

(79) The mechanical properties measured on a separate piece of this same material show that it is significantly weaker than those measured on the material of Ex. 1 (Table 6)

(80) A different section of this solid polymer electrolyte was prepared into an MEA using the procedures above, assembled into a fuel cell, and tested using Test Protocol 1A. The results show that the fluoride release rates are very low.

Comparative Example 1

(81) A PRIMEA® series 5700 MEA with 0.4 mg Pt/cm.sup.2 loading (W. L. Gore & Associates, Elkton, Md.) in each electrode was assembled into a fuel cell as described above and tested in Test Condition 1A. This MEA is reinforced with ePTFE and is the latest commercial offering (as of the date of filing) from W. L. Gore & Associates, so provides an indication of state-of-the-art performance for durable, composite membranes. There is no catalyst present in the solid polymer electrolyte of this catalyst coated membrane. The results shown in Table 2 indicate that the MEA in Example 1 using the inventive solid polymer electrolyte has nearly three times the life of the MEA of this Comparative Example, and has 1-2 orders of magnitude higher fluoride release rate than the inventive Examples 1-3.

(82) TABLE-US-00002 TABLE 2 Fluoride Lifetime Release Rate Example No. (hr) (g/hr .Math. cm.sup.2) Ex. 1 718 1.10E−07 Ex. 2. 78 4.50E−08 Ex. 3 90 7.10E−08 Comp. Ex. 1 243 1.30E−06

Example 4

(83) Another section of the solid polymer electrolyte that was prepared in Example 1 was used to prepare an MEA in the same fashion as Example 1. It was assembled into a fuel cell using the procedures described above, and tested using Test Protocol 1B. The Lifetime and Fluoride Release Rate results are shown in Table 3.

Comparative Example 2

(84) In order to compare the inventive solid polymer electrolyte materials to those prepared previously in the prior art, a solid polymer electrolyte was prepared using a procedure similar to that used in U.S. Pat. No. 5,472,799 to Watanabe et. al. Specifically, a dispersion of unsupported platinum particles was formed in a solid polymer electrolyte by the following procedure: 1) 0.219 grams of hydrogen hexachloroplatinate (IV) hydrate salt (H.sub.2Cl.sub.6Pt.H.sub.2O) (available from Sigma-Aldrich, St. Louis, Mo.) was dissolved in 10 grams of FLEMION® dispersion with equivalent weight of 950 (Asahi Glass Co. Ltd, Chemicals, Tokyo, JAPAN) in a 9% solid ionomer water/alcohol solution using a magnetic stir bar and a stir plate; 2) 25 cm.sup.3 of water solution of sodium boron hydride (NaBH.sub.4 from Sigma-Aldrich, St. Louis, Mo.) with concentration of 0.05 M was then prepared; 3) The NaBH.sub.4 solution was titrated into the H.sub.2Cl.sub.6Pt containing ionomer solution slowly. During the titration, platinum ions (Pt.sup.4+) are reduced to colloidal platinum metal (Pt) particles through the reduction effect of NaBH.sub.4. As this reaction proceeded further, the ion exchange solution turned to a dark color due to increasing amount of colloidal Pt particles. The relative concentration of Pt.sup.4+ and BH.sub.4.sup.− was monitored by measuring the electrochemical potential difference between a Pt wire working electrode and a Hg/HgSO.sup.4 reference electrode emerged in the ion exchange solution. The end-point of the titration was marked by a sudden drop of the electrochemical potential; 4) After reaching the end-point, the ionomer mixture was poured out into a shallow glass dish and dried at room temperature under dry nitrogen flow; 5) The dried ionomer mixture was washed using 0.05 M, high purity sulfuric acid (H.sub.2SO.sub.4) solution to eliminate ions such as chlorine (CF) and sodium (Na.sup.+) ions; 6) After washing with acid three times, the ionomer was washed with de-ionized water three times; 7) The cleaned ionomer was dissolved into water/alcohol at room temperature using a magnetic stir bar on a stir plate to obtain a solution consisting of approximately 20% solids, 20% water and 60% alcohol. Within the ionomer solids, pure Pt colloidal particles accounted for approximately 1% by weight. 8) A solid polymer electrolyte was then prepared from this solution using a process similar to that used in Example 4. Specifically a roll of expanded polytetrafluoroethylene (ePTFE) membrane with mass per area of 7.0 g/m.sup.2, a thickness of 20 microns, a porosity of at least 85%, a longitudinal matrix tensile strength of about 67 MPa, and a transverse matrix tensile strength of about 76 MPa, was prepared using the teachings of U.S. Pat. No. 3,953,566 to Gore. Then, the solution of prepared in (7) was coated onto a film paid off of a roll of PEN film at 3 feet/min using a #44 Meyer bars. The ePTFE was then applied to the wet solution, and passed through a 3 ft oven held at 150° C. in air. A second coating of the solution was made onto this material by running it through the same line at 3 feet/min using the same Meyer Bar. The final solid polymer electrolyte was 25 microns thick, and had a platinum loading in the solid polymer electrolyte of 0.016 mg/cm.sup.2.

(85) A TEM micrograph of this membrane shows the presence of platinum particles 80 in FIG. 8.

(86) This solid polymer electrolyte was used to prepare an MEA and then placed into a fuel cell using the procedures described above. It was then tested using Test Protocol 1B. The results (Table 3) indicate that the lifetimes are shorter, and the fluoride release rates are higher than the inventive membranes (Ex. 4).

Comparative Example 3

(87) A PRIMEA® series 5700 MEA with 0.4 mg Pt/cm.sup.2 loading (W. L. Gore & Associates, Elkton, Md.) in each electrode was assembled into a fuel cell as described above and tested in Test Condition 1B. This MEA is reinforced with ePTFE and is the latest commercial offering (as of the date of filing) from W. L. Gore & Associates, so provides an indication of state-of-the-art performance for durable, composite membranes. There is no catalyst present in the solid polymer electrolyte of this catalyst coated membrane. The results shown in Table 3 indicate that the MEA in Example 4 using the inventive solid polymer electrolyte has nearly twice the life of the MEA of this Comparative Example, and over seven times lower fluoride release rate.

(88) TABLE-US-00003 TABLE 3 Fluoride Lifetime Release Rate Example No. (hr) (g/hr .Math. cm.sup.2) Ex. 4 1365 4.30E−08 Comp. Ex. 2 527  2.3E−07 Comp. Ex. 3 700 1.90E−07

Example 5

(89) In this example an inventive solid polymer electrolyte was prepared with platinum in a layer on only one side of the final solid polymer electrolyte. This was done as follows. First, an expanded polytetrafluoroethylene (ePTFE) membrane was prepared using the teachings of Hobson et. al. in U.S. Pat. No. 6,613,203, incorporated herein in its entirety. A membrane similar to the Type 2 ePTFE in Hobson was prepared except the processing parameters were adjusted so the mass per area was about 7.5 g/m.sup.2, the thickness was 25 microns, the longitudinal matrix tensile strength was about 267 MPa (38,725 psi), the transverse matrix tensile strength was about 282 MPa (40,900 psi), the Gurley number was about 8.5 seconds, and the aspect ratio was about 29. An ink prepared as described in Example 1 was mixed with ion exchange material Type 3 to give a concentration of 2.4% platinum based on weight percent of dry ionomer solids. This solution was passed through the Microfluidizer three consecutive times with a pressure setting of 19,000 psi. Then the solid polymer electrolyte was prepared as follows: for the first coating, a #44 Meyer Bar was used to coat onto a PEN film stretched tight over a glass plate. Pure Type 3 (with no platinum in it) was used for this first coating. The ePTFE membrane was then stretched over the wet coating and allowed to infiltrate. After infiltration, it was dried for 20-60 s with a hair drier. Then, a second coating using the ink solution prepared above was applied with a #22 Meyer Bar. The second coating was then also dried with a hair drier for 20-60 s. This membrane was placed in a 160° C. air furnace for three minutes and then removed to cool. The membrane was then removed from the backer in room temperature deionized water. The measured platinum loading of this membrane was 0.022 mg/cm.sup.2, and its final thickness was 18 microns. This materials was tested in a Gurley Densometer Model 4110 (Gurley Precision Instruments, Troy, N.Y.) and found to have a Gurley number greater than 10,000 s.

(90) In this example, an MEA was prepared with a section of the solid polymer electrolyte using the procedure described in Example 1. This MEA was placed in a fuel cell using the procedures described above, so that the side with the layer containing carbon particles supporting platinum was facing the anode compartment. It was then tested using Test Protocol 1A. The Lifetime and Fluoride Release Rate results are shown in Table 4. The mechanical properties of a separate section of the solid polymer electrolyte were also obtained, with the results also shown in Table 6.

Example 6

(91) A different section of the solid polymer electrolyte prepared in Example 5 was made into an MEA as described in Example 5 and tested in a fuel cell using Test Protocol 1A. In this Example, though, the side with the layer containing carbon particles supporting platinum was facing the cathode compartment. The lifetime and fluoride release rate results are shown in Table 4.

Comparative Example 4

(92) In order to obtain an indication of the improvement in properties of the inventive materials of Example 5 and 6, a material made with the same reinforcement and same ionomer used in Examples 5 and 6 was prepared. A solid polymer electrolyte was prepared using the same methods described above, except only pure Type 3 ionomer was used so that no catalyst supported on a carbon layer was present. The results from testing in Test protocol 1A (Table 4) surprisingly show that the lifetime of the inventive materials was about two (Ex. 5) to over seven (Ex 6) times higher than this Comparative Example. The fluoride release rates were half (Ex. 5) to more than ten times (Ex. 6) lower than those observed in this Comparative Example.

(93) TABLE-US-00004 TABLE 4 Fluoride Lifetime Release Rate Example No. (hr) (g/hr .Math. cm.sup.2) Ex. 5 523 3.90E−07 Ex. 6 1677 3.40E−08 Comp. Ex. 4 283 8.60E−07

Example 7

(94) In order to show that a microporous reinforcement is not required for the inventive materials to achieve improved lifetimes, a NAFION® N101 membrane was purchased from Ion Power, Inc. (Bear, Del.). Unlike some NAFION® membranes, this material is processed in a way to make the membrane relatively strong. A layer of ion exchange materials comprising carbon supporting platinum catalyst was then laminated onto this membrane to prepare an inventive solid polymer electrolyte. The procedure was as follows: first, a solid polymer electrolyte of ion exchange material Type 3 containing ink with a platinum concentration of 2.4% was cast onto a fluoropolymer treated polyethylene terepthalate (PET) film using a #22 meyer bar. This membrane was dried at 80° C. for 5 minute and removed from the PET film at room temperature in air. It had a thickness of five microns. This layer was then laminated to the N101 membrane at 180° C. for 1 minute. The final membrane had a measured platinum content of 0.019 mg/cm.sup.2 and a thickness of 30 microns. The mechanical properties of a section of this membrane were tested, and the results are shown in Table 6.

(95) An MEA was prepared with this membrane and it was assembled into a fuel cell with the layer of carbon supporting platinum catalyst facing the cathode as described above. The results from testing in Test Protocol 1A are shown in Table 5.

Comparative Example 5

(96) A solid polymer electrolyte was prepared using the same procedures outlined in Example 7, except the cast Type 3 membrane had no ink in it, i.e., it was pure Type 3 ion exchange material with no platinum supported on carbon in it. This cast membrane had a thickness of 5 microns after drying. After lamination to the N101 membrane, the resulting membrane had a thickness of 30 microns. This solid polymer electrolyte was tested using Test Protocal 1A and the same procedures as Example 7. The results (Table 5) show that the inventive solid polymer exchange material, Example 7, has nearly twice the lifetime, and more than three times lower fluoride release rate than Comparative Example 5.

(97) TABLE-US-00005 TABLE 5 Fluoride Lifetime Release Rate Example No. (hr) (g/hr .Math. cm.sup.2) Ex. 7 138 1.00E−06 Comp. Ex. 5 74 3.40E−06

(98) TABLE-US-00006 TABLE 6 Failure Force Stiffness Strength Modulus Example No. (g) (g) (g/cm.sup.2) (g/cm.sup.2) Ex. 1 237 1911 2.63E+05 2.12E+06 Ex. 3 106 1547 1.18E+05 1.72E+06 Ex. 5 409 3310 4.54E+05 3.68E+06 Ex. 7 181 2724 1.21E+05 1.83E+06 Comp. Ex. 2 110 1772 9.15E+04 1.48E+06

Example 8

(99) In order to demonstrate the utility of the inventive solid polymer electrolytes under conditions that might occur in real applications, a sample was prepared and tested in Test Protocol 2. The sample was prepared with platinum supported on carbon in layers on two sides using the general procedures of Example 1. Here, though, the expanded polytetrafluoroethylene (ePTFE) membrane was prepared using the teachings of Hobson et. al. in U.S. Pat. No. 6,613,203. A membrane similar to the Type 2 ePTFE in Hobson was prepared except the processing parameters were adjusted so the mass per area was about 7.5 g/m.sup.2, the thickness was 25 microns, the longitudinal matrix tensile strength was about 267 MPa (38,725 psi), the transverse matrix tensile strength was about 282 MPa (40,900 psi), the Gurley number was between 10 and 12 seconds, and the aspect ration was about 29. The ink was prepared as described in Example 1 using ion exchange material Type 1 to give a concentration of 0.8% platinum based on weight percent of dry ionomer solids. In this example, the ink was used for both coating steps described in Example 1. In the first coating step the drawdown bar was set to 0.0965 cm (0.038 inches), while in the second, it was set to 0.0483 cm (0.019 inches). After drying the second coating with a hair dryer, the membrane was placed in a 160° C. air furnace for three minutes and then removed to cool. The membrane was then removed from the PEN film in room temperature deionized water. The measured platinum loading of this membrane was 0.015 mg/cm.sup.2, and its final thickness was between 19 and 21 microns.

(100) An MEA was prepared with a section of the solid polymer electrolyte using the procedure described in Example 1. This MEA was placed in a fuel cell using the procedures described above and tested using Test Protocol 2. The lifetime and fluoride release rate results are shown in Table 7.

(101) In order to observe distribution of supported catalysts inside the membrane, transmission electron microscopy was performed on cross sections of the ion exchange membrane used in this example. A section of the solid polymer electrolyte of this example was embedded in Spurr® epoxy resin and cured at 60° C. for eight hours. The embedded sample was first trimmed with a razor blade and then thin sectioned at room temperature using a Diatome diamond knife on a Leica Ultracut UCT ultramicrotome. The microtome was set to cut 75 nm thick sections which were collected on 300 mesh copper grids. Transmission Electron Microscopy was performed using a JEM 2010 Field Emission TEM, at 200 KV at various magnifications. The results indicated that there was a plurality of very fine Pt/C particles of size less than 75 nm (FIG. 9). The interparticle spacing between these Pt/C particles was measured between 15 different pairs of particles and found to be about 115 nm on average.

(102) The presence of Pt and C in these particles was confirmed at higher magnifications (FIG. 10) both by contrast, and the presence of lattice images consistent with C and Pt, 100 and 101, respectively in FIG. 9. As an aid to the eye, the dotted line in FIG. 10 has been added to show the approximate extent of a carbon particle supporting platinum. The interparticle spacing between these Pt particles was measured between 10 different pairs of Pt particles and found to be about 10 nm on average.

Examples 9-10

(103) An additional solid polymer electrolyte was prepared using the same procedure as Example 5 except that the final step of passing the diluted ink solution through the Microfluidizer was omitted. The measured platinum loading of this SPE was 0.016 mg/cm.sup.2 and its final thickness was 15-18 microns. Two MEAs were prepared with sections of the solid polymer electrolyte using the procedure described in Example 1. These MEAs were placed in a fuel cell using the procedures described above and tested using Test Protocol 2. The lifetime and fluoride release rate results are shown in Table 7.

Example 11

(104) To confirm the surprisingly low fluoride release rates of the inventive materials, a different section of the solid polymer electrolyte prepared in Example 2 was made into an MEA using the procedures above, assembled into a fuel cell, and tested using Test Protocol 2. The fluoride release rate was very low, comparable to that observed in Example 2, which was tested in a different test protocol.

Example 12

(105) To further confirm the surprisingly low fluoride release rate of the inventive materials, a different section of the solid polymer electrolyte prepared in Example 3 was made into an MEA using the procedures above, assembled into a fuel cell, and tested using Test Protocol 2. The fluoride release rate was again very low, comparable to that observed in Example 3, which was tested under a different test protocol.

Comparative Example 6

(106) A sample to compare to Example 8 was prepared in this Comparative Example. The preparation procedure was similar to that used in Example 8 except that no ink was used in the preparation of the solution so there was no platinum supported on carbon in the final SPE. A #28 meyer bar was used for the first coating (instead of the drawdown bar), a #22 meyer bar was used for the second coating (instead of the drawdown bar), and the heat treatment after drying the second coating took place at 150° C. for 1 minute. After removing the final solid polymer electrolyte from the PEN film at room temperature in air, the thickness was measured to be 18 microns.

(107) The resulting solid polymer electrolyte was made into an MEA as described above, and tested in a fuel cell using Test Protocol 2. The results (Table 7) show that the inventive solid polymer electrolyte of Example 11-12 have about an order of magnitude lower fluoride release rates than this Comparative Example. Further, when a strong solid polymer electrolyte is used such as in Example 8-10, the observed Lifetime was over three times longer, and the fluoride release rate an order of magnitude lower than observed for Comparative Example 6.

(108) TABLE-US-00007 TABLE 7 Fluoride Lifetime Release Rate Example No. (hr) (g/hr .Math. cm.sup.2) Ex. 8 582 1.30E−08 Ex. 9 637 8.38E−09 Ex. 10 586 2.46E−08 Ex. 11 <20 1.80E−08 Ex. 12 <31 1.90E−08 Comp. Ex. 6 184 1.30E−07

Example 13 and Comparative Example 7

(109) The inventive materials contain a composite layer of a solid dispersion comprising a plurality of support particles supporting a catalyst comprising a precious metal catalyst and an ion exchange material. In this example, this composite layer is shown to be substantially occlusive and electronically insulating. Two samples were prepared, a composite layer of a solid dispersion of a plurality of carbon particles supporting a platinum catalyst in an ion exchange material, and the same ion exchange material without the platinum supported on carbon. These two samples were prepared using ion exchange material Type 3 using the general procedure outlined in Example 1. Here, the drawdown bar was set to 0.0254 cm (0.010 inches), the concentration of platinum in the ink was 2.4% for Example 13, and no ink was used for Comparative Example 7. Example 13 was cast on a polyethylene terepthalate film whose surface that had been treated with a fluoropolymer to enhance release, while Comparative Example 7 was cast onto a glass plate. Only one pass was made, and no microporous film was applied. Both samples were heat treated at 160° C. for 3 min. Example 13 was removed from the film at room temperature, while Comparative Example 7 was removed under room temperature water.

(110) The catalyst-containing membrane layer, Example 13, is a physical model of a composite layer of the invention, while Comparative Example 7 is a layer without catalyst. The latter is used herein to show that the properties of the inventive composite layer are the same as a homogenous layer without the platinum supported on carbon. To characterize the electrical properties of the membrane-catalyst layer, electrochemical impedance measurements were conducted on the two membrane layers, Example 13 and Comparative Example 7: Fuel cell electrodes with a loading of 0.4 mg-Pt/cm.sup.2 coated on a release layer were attached to both sides of the membranes using 15 tons of pressure at 160° C. for 3 minutes. The test was performed using the experimental procedures described by Johnson and Liu (ECS Proceedings Volume 2002 -5, pages 132-141). The impedance spectra were measured at a temperature of 80° C. and a relative humidity of 88% in an atmosphere of nitrogen gas. The impedance data for a frequency range of 20.0 kHz to 2.0 Hz are shown as a Nyquist plot (the imaginary vs. the real component of the impedance) in FIG. 11.

(111) The impedance spectra for the two membranes are nearly identical, indicating the membranes have essentially the same electrical properties. Furthermore, these spectra are characteristic of ionically conductive membranes that are electronic insulators, i.e., there are not adequate pathways for electrons to pass through the membrane. Therefore, the composite layer of a solid dispersion comprising a plurality of support particles supporting a catalyst comprising a precious metal catalyst and an ion exchange material layer of the invention is an electronic insulator.

(112) A separate cast material prepared identically to that described above for Example 13 was tested using a standard Gurley air flow test. It had a Gurley value of greater than 10,000 s indicative of a substantially occlusive material. These tests taken together thus demonstrate that the composite layer comprising a plurality of carbon particles supporting a catalyst comprising platinum and an ion exchange material used in the inventive solid polymer electrolyte is both substantially occlusive and electronically insulating.

Example 14

(113) An additional inventive solid polymer electrolyte was prepared using the procedures of Example 5 except the Type 3 ion exchange material was mixed with ink to produce a solution that was 11.5% platinum weight to dry ionomer (instead of 2.4% used in Example 5). This solution was not passed through the Microfluidizer, but instead, a portion of it was placed in a 25 ml centrifuge tube, and subsequently centrifuged in an Adams Compact II Centrifuge (Beckton-Dickenson Inc., Franklin Lakes, N.J.) for about 20 min at 3200 rpm. After centrifuging, the supernatant was used to prepare a solid polymer electrolyte as described in Example 5. The final membrane had a very light grey color, a thickness of 15-18 microns, and a measured platinum loading below the detection limit of the XRF (<0.001 mg/cm.sup.2).

(114) Although several exemplary embodiments of the present invention have been described in detail above, those skilled in the art readily appreciate that many modifications are possible without materially departing from the novel teachings and advantages which are described herein. Accordingly, all such modifications are intended to be included within the scope of the present invention, as defined by the following claims.