Union College Union | Digital Works
Honors Theses Student Work
6-1966 Platinized asbestos as a fuel cell electrocatalyst Howard Irwin Zeliger Union College - Schenectady, NY
Follow this and additional works at: https://digitalworks.union.edu/theses Part of the Chemistry Commons
Recommended Citation Zeliger, Howard Irwin, "Platinized asbestos as a fuel cell electrocatalyst" (1966). Honors Theses. 2242. https://digitalworks.union.edu/theses/2242
This Open Access is brought to you for free and open access by the Student Work at Union | Digital Works. It has been accepted for inclusion in Honors Theses by an authorized administrator of Union | Digital Works. For more information, please contact [email protected]. UNION COLLEGE - GRADUATE STUDIES Schenectady, New York
PLATINIZED ASBESTOS0AS A FUEL CELL ELECTROCATALYST
A thesis subrrr tted to the Cammi ttee on Graduate
S'tudies and the Department of Chemistry of Union College,
Schenectady, New York, in partial fulfillment of the- requirements for the degree of Master of Science.
by Harold' Irwin Zeliger 11.S /'166 II/
Approved c: e.
TOMY WIFE AND SON •
TABLE''OF CONTENTS
.Page
LIST OF TABLES------111 LIST OF FIGURES------~------iv ACKNOWLEDGEMENTS------vi ABSTRACT------~------vii INTRODUCTION - AIM------1 BACKGROUND------2 EXPERIMENTAL ------·------'------14 RESULTS AND DISCUSSION------30 PART I------30 PART II------64 CONCLUSIONS ------·-- 71 BIBLIOGRAPHY------~--~--~------74 -111-
LI ST OF TABLES. .
Page 1. summary of preparation method and properties of catalysts ------~------..:---· 31 · 2. Compos1t1on of electrodes prepared-~------37 3. Properties of catalysts prepared through stepwise plat1nizat1on ------· 65 4. Composition of electrodes 18, 20, 21, and 22-- 67 .5. Comparison of current densities per unit area of, platinum at constant voltage.------69 -iv-
LIST OF FIGURES Page 1. Waveforms from cells under load containing standard and test electrodes------26 2. Oscilloscope photograph showing the effect of a prief current interruption on a cell under load------29 3. Electron micrograph of catalyst A (hydrazine reduction) 25,000 x magnification------4. Electron micrograph of catalyst B (formaldehyde reduction) 25,000 x magnification ----·------33 5. Electron micrograph of catalyst C (hydrogen reduction, 400 C) 25,000 x magnification --- 34 6. Electron micrograph of catalyst D (sodium borohydride reduction) 25,000 x magnification------35 7. Performance curves for electrodes 1 to 4 on hydrogen showing the effect of cat~~yst preparation------38 8. Performance curves for electrodes l to 4 on oxygen showing the effect of catalyst preparation------39 9. Performance curves for electrodes 4 to 8 on hydrogen showing the effect of diluent----- 41 10. Performance curves for electrodes· 4 to 8 on oxygen showing the effect of diluent------42 11.' Performance curves for electrodes 9 to 12 on hydrogen showing the effect of graphite content------44 12. Performance curves for electrodes 9 to 12 on oxygen showing the effect of graphite · content------~------45 13. Electron micrograph of catalyst E (sodium borohydride reduction) 25,000 x magnification------~-~- . 47 -v-
14; Electron micrograph of catalyst F (wet hydrogen reduction) 25,000 x magnification-- . 48 15. Electron micro§raph of catalyst G (hydrogen reduction, 100 C) 25,000 x magnification --- 49 16. Performance curves for electrodes·13 to 15 on hydrogen showing the effect of catalyst, preparation------51 17. Performance curves for electrodes 13 to 15 on oxygen showing the effect of catalyst preparation------~- 52 18. Electron micrograph of catalyst H (hydrogen reduction, room temperature) 25,000 x magnification------54 19. Electron micro§raph of catalyst I (hydrogen reduction, 200 C) 25,000 x magnification --- 55 20. Performance curves for electrodes 15 to 17 showing the effect of temperature of hydrogen reduction of catalysts on hydrogen performance------56 21. Performance curves for electrodes 15 to 17 showing the effect of temperature of hydrogen reduction of catalysts on oxygen performance------57 22. Electron micrograph of catalyst J6 (6 cycle platinization) 25,000 x magnification------59 23. Electron micrograph of catalyst K (1 cycle platinization) 25,000 x magnification------60 24. Performance curves for electrodes 15, 18, and 19 on hydrogen comparing platinum load- ing with performance.------61 25. Performance curves for electrodes 15, 18, and 19 on oxygen comparing platinum loading with performance~------62 26. IR free performance curves on oxygen for electrodes with varying platinum content 68 -vi-
ACKNOWLEDGEMENTS
I would like first to extend my appreciation to Prof. K. v. Nahabedian of Union College and to Dr. L. w. Niedrach of the General Electric Research and I . Development Center for their guidance in the development of this work. Sincere thanks are extended to Dr. D. w. McKee of the General Electric Research and Development Center for measuring the surface areas of many of the catalysts, and to Mr. E. F. Koch of the General Electric Research and Development Center for taking the electron micrographs of the catalysts. The author is indebted to Dr. H. A. Liebhafsky, manager, Electrochemistry branch, Chemical Systems and Processes Laboratory, General Electric Research and Development Center, for making available the use of the facilities without which this work would not have been possible. -vii-
ABSTRACT
Platinized asbestos electrooatalyete were prepared by a variety of procedures. These catalysts were incorporated into test electrodes which were electrochemically evaluated in hydrogen - oxygen ambient temperature fuel cells. catalysts were further scrutinized by the measurement of their specific platinum surface areas, and by t he examination of electron micrographs of the materials. Fabrication procedure and composition of platinized asbestos containing electrodes we~e optimized, and fuel cell performances of electrodes containing the test catalysts were compared with the specific platinum surface areas of the cata1ysts. -1-
INTRODUCTION - AIM
The work described in this thesis was undertaken to study various methods for preparing platinized asbestos, with the aim of determining the procedure that gives the most active fuel cell electrocatalyst. An additional goal of this work was to compare fuel cell performance with specific surface area of platinum. Catalysts were evaluated by incorporation into electrodes and then monitoring electrode performance in hydrogen - oxygen ambient temperature fuel cells. Catalysts were further scrutinized via surface area measurements and electron micrographs. In order to best accomplish fuel cell evaluation, it was necessary to optimize electrode structure and composition, and find suitable methods for measuring performances of electrodes'. -2-
BACKGROUND*
FUEL CELL
A fuel cell in an electrochemical cell in which
the chemical energy of a fuel oxidation reaction is
directly converted into electrical energy. Fuel and oxidant rrust be continually fed into the cell from
external sources. Fuel cell electrodes serve only as
reaction sites and do not undergo chemical change. In these two respects, fuel cells differ from ordinary
storage batteries which store chemical energy and whose electrodes underso chemical change.
A fuel cell consists of two electrodes, an anode
and a cathode, separated by an electrolyte. An external circuit connects the anode and cathode.
The electrolyte makes up the internal circuit of
the cell. It should have as high a conductivity as
possible in order to minimize ohmic losses within the
cell, and should be impermeable to diffusion of both
fuel and oxidant, or non-electrochemical oxidation would
occur leading to heat production and inefficient operation. Low temperature fuel cells usually have aqueous solutions
of either strong acids such as sulfuric acid or phosphoric acid, or strong bases such as sodium hydroxide.
#~more comprehensive treatment may be found in the literature (1, 2~ 3, 4). -3-
or potassium hydroxide as electrolytes. Fuel cell electrodes serve as catalytic sites for electrochemical reactions. They also serve as current collectors and establish the interfaces between insoluble gases and liquid electrolutes. These functions of electrodes imply that they must be good conductors, resistant to corrosive electrolyte, and have the necessary catalytic activity to promote the electrochemical reaction. More will be said about electrodes later.
PRINCIPLE OF FUEL CELL OPERATION When a fuel is oxidized isothermally at constant pressure, the maximum amount of heat available is given by the enthalpy change of the reaction (6H). The maximum theoretical amount of work, or electrical energy, attainable from the reaction is given by the Gibbs free energy change of the reaction (AG). Free energy and enthalpy are related by the thermodynamic relationship
..6G =~H -TAS ( l) where T4S- is the minimum amount of heat that is- produced when the process occ~rs isothermally and at constant pressure. The theoretical thermodynamic efficiency of a fuel cell is the attainable energy divided by the total energy produced, or -4-
Eff = 4G = 4.T-1-T.18 ( 2 ) L'IH b.H
From equation (2) we see that the thermodynamic requirement for fuel cell efficiency is low temperature of operation, just the opposite of a heat engine whose maximum efficiency is £50Verned by the Car-no t eye le, and given by equation (3) Eff = Thigher - T1ower ( 3) Thigher
HYDROGEN - OXYGEN FUEL CELL In the hydrogen - oxygen fuel cell, fuel (hydrogen) is oxidized at the anode and oxidant (oxygen) is reduced at the cathode. The anode reaction H2 = 2H+ + 2e- produces electrons which travel to the cathode via the external circuit (doin3 electrical work en route) and protons which migrate to the cathode thro~gh the internal circuit (electrolyte). The protons and electrons are consumed by the cathode reaction t o2 + 2~- + 2H+ n20 Thus the overall reaction is
ELECTROCATALYSTS The electrochemical oxidation of fuel and reduction of oxidant in fuel cells are processes that require the -5-
use of catalysts. The term electrocatalyst as used here applies to materials that serve as catalysts in fuel cell oxidation-reduction reactions. The electrocatalyst employed depends upon several parameters including; the fuel· rd oxidant, whether for use in the anode or in the cathode, the electrolyte, and the operating conditions of the cell. It must, of course, posess catalytic activity, but in addition must not be volatile at the temperature of operation, must be insoluble in the electrolyte, must not be subject to electrochemical oxidatio~ or reduction, and ~ust be resistant to sinterin3 under the operatin8 conditions. The catalyst participates in at least three functions of the electrode process; adsorption, electron transfer, and surface reaction. The adsorption of the reactant should be strong enou3h to make it susceptible to oxidation or reduction, and rapid enough to yield high currents. Ener3etically, the adsorption bond should be about 5 kcal/mole and the activation energy of adsorption should be negligible. This type of adsorption is called chemisorption. If the electrons of the adsorbed species are held too strongly by the catalyst, the freedo~ of the adsorbate to further react will be impaired and the rate of reaction is diminished. If the bonding is too weak, it will not permit the molecular orientation -6-
upon which the activated complex formation depends. In attempting to explain catalytic activity, one must consider electronic and geometric factors. The former may influence the catalytic process by int1;;,raction Hith and promotion of the outer orbital electrons of the reactants thereby reducing the free energy of the activated complex and accelerating the reaction. As for the latter, interatomic spacing and lattice arrangements in the catalyst influence the velocity of adsorption and desorntion. For example, Sherman and Eyring (5) have shown that for hydrogen chemisorption on carbon,
interatomic spacing of 3-5 R gives a. cinirnum e ne r-gy , At larger distances, the energy is higher because the hydrogen molecule must be dissociated prior to adsorption. At small spacings, repulsion forces retard adsorption.
ANODE CATALYSTS'
Transition metals such as nickel, iron, cobalt, platinum, palladium, rhodium, ruthenium, and iridium are active in fuel cells as electrocatalysts for hyorogen oxidation, and of these, platinum has been found to be the best. No satisfactory explanation has been given for this, but proposals have been made (6, 7). CATHODE CATALYSTS
Very little study has been ar ~d to understanding -7-
the oxygen cathode, but ~m9irical studies have de-onstrated that platinum is a superior catalyst, arid that 3roup lb metals, particularly silver (in basic electrolyte systeQs) are also effective catalysts (8). One theory proposed as to what occurs when oxy3en is reduced at a cathode is that of Berl (9). He investigated t~e reduction of oxygen at carbon cathodes and concluded that the products of the reduction are perhydroxyl and hydroxyl ions. Perhydroxyl may react with hydronium ions in the electrolyte to yield hydrosen peroxide, which in turn is unstable and is decomposed to oxygen and water. o2 + H20 + 2e- =OH-+ 02H- 02H- +OH- +2H+ = H20 + H202
If we assume that this reaction sequence represents the cathode reaction, we would expect an oxygen catalyst to meet two requirements: (a) oxygen should be readily adsorbed, and (b) hydrogen peroxide should be decomposed. The most active oxygen catalysts adsorb oxygen and decompose peroxide. FUEL CELL ELECTRODES Practical fuel cells produce high current densities and have minimum voltage losses when current is drawn. In considering the performances obtained from fuel cells, -8-
one must consider electrode structure and the individual steps that occur at the electrodes. Electrodes may be thought of as consisting of three regions. The first region consists of gas filled pores. By keeping this region as thin as possible, gas diffusion barriers are kept to a minimum and performance is· not impaired. The second region bridges the gas and electrolyte. Small menisci form within the pores and the surface is covered with a thin film of electrolyte that permits ready transport of dissolved reactant while offering little resistance to the transport of ionic species. In region three the surface is flooded by electrolyte and le inactive. It is desirable to keep the third region as thin as possible, so as to minimize the internal resistance of the electrodes. With this· in mind, let us now consider the individual processes that occur at fuel cell electrodes. Six specific steps may be listed: 1. Mass transport of fuel or oxidant molecules to the electrodes. 2. Absorption of reactant molecules into the electrodes (into region one). 3. Adsorption of molecules on reaction sites (in region two). 4. Electron transfer and generation of ionic species (in -9-
region two). 5. Desorption of products (into region three). 6. Mass transfer of. products away from the electrodes (out of region three). In addition to these steps, there may be side reactions such as surface reactions between adsorbed molecules, discharged ions or radicals. Each of the above steps may have associated with it energy losses that give rise to polarization.
FUEL' CELL POLARIZATIONS· From thermodynamic considerations, one would expect a hydrogen - oxygen fuel cell to produce 1.23 volts regardless of any current being produced by the cell. There are, however, voltage losses associated with various irreversible processes occuring within the cell while current is flowing. These voltage losses are generally refered to as polarizations or overvoltages. To consider these polarizations let us briefly reexamine what occurs at fuel cell electrode~. Fuel and oxidant are delivered to the anode and cathode respectively where the molecules are absorbed and reacted, and products are desorbed and carried away from the electrodes. Electrons are circulated from anode to cathode via the external circuit. All the steps involving mass transport, adsorption, surface chemical reaction, or charge transfer, -10-
have energy barriers associated with them which give rise to polarizations. Three types of polarization are characterized; activa~ion polarization, concentration polarization, and ohmic polarizati~n.
Activation polarization is the voltage loss associated with the activation energies of adsorption, electron transfer, and surface reactions, and is most evident at low current densities. It may be reduced by operating the cell at higher temperatures, employing active catalysts (with higher surface areas) and use of effective electrode design.
Concentration polarization is the overvoltage due to mass transport limitations. It is most evident at high current densities where rapid transport of reactant molecules to the catalytic sites becomes important.
This type of polarization is associated with limiting current densities. Limitation on the rate of mass transport is associated with electrode structure, and as such may be minimized by effective electrode structure.
(A thin porous electrode will allow for rapid mass transport of reactants to catalytic sites). Ohmic polarization is due to internal resistances of electrodes and .electrolyte. It generally varies lin nrly with inoreaeing current deneity. Ohmic polarization may be minimized by the use of highly conducting -11-
electrodes and electrolyte, thin electrolyte gap, and higher temperature. Ohmic voltage losses may be lowered but never eliminated. All three types of polarization may be reduced through effective electrode design, however, in designing an electrode one must consider other parameters, including leakage of electrolyte, gas diffusion, and structural integrity. It is for this reason that one may sometimes be compelled to live with some voltage losses which seemingly could be eliminated.
CATALYST SUPPORTS* Catalysts need· not always be homogeneous phases of one material. In some processes, heterogeneous· mixtures of two or more materials make up ''the catalyst" , while for other applications it is profitable to support the catalytic material. Catalysts are supported for any one or combination of the following reasons: 1. Supporting a catalyst may give !ise to larger exposed surface of the catalyst, and thereby increase the surface area of the catalyst. Since it is'commonly believed that the interior of metal catalyst particles do not contribute to the catalytic process it is desirable to •
*A more detailed summary may be found in the literature. See for example (10, 11). -12-
keep it to a minimum, and supporting the catalyst is one way of accomplishing this objective. 2. Supporting a catalyst may increase stability by separating its fine crystals· so sintering can't occur. 3. A supported catalyst may have improved properties. 4. The support 1 tself may aid in the catalytic process·. 5. The support may impart poison resistance to the catalyst. 6. The support may prevent sintering by dissipating the heat of a reaction. 1. The support may increase the accessibility of active surface. Among the factors· to be considered in selecting a support are: 1. Possible catalytic inhibition of the reaction.
2. Surface area, particle size and density of t he support. 3. Porosity, or lack of such according to requirements. 4. Structure of the support (is it fibrous, or does it consist of spherical particles etc.). 5. Heat and electrical conductivity requirements. 6. Stability under reaction conditions.
ASBESTOS-AS A SUPPORT The use of asbestos· as a catalyst support is well known, however, no mention of asbestos as a support for a fuel cell electrocatalyst was found in the literature. -13-
The use of supports for fuel cell catalysts has mainly been with the intention of increasine; the surface area of the platinum catalyst and preventing its sintering.
As a support, asbestos meets these requirements. Its fibrous structure offers many sites for the deposition of the platinum and also increases the accessibility of active surface to reacting gases and electrolyte. In addition, asbestos has been shown to improve the
structural integrity of fuel cell electrodes (12).
Anthophyllite variety asbestos is attractive for use as a support in acid electrolyte fuel cells, because
of it~ resistance to acids, resistance to electrochemical
oxidation and reduction, very porous structure, small
fiber size and high surface area (about 11 m2/gram) (12). -14-
EXPERI~i:ENTAL
PLATINIZED ASBESTOS Anthophyllite variety asbestos was platinized to contain from one to twenty percent platinum. In each oaae the source of platinum was chloroplatinic acid (H2Ptc16). The procedures used in these preparations are described below. FORMALDEHYDE REDUC~ION (13) A solution of chloroplatinic acid in 6N hydrochloric acid is cooled to below 5°0 and treated with a one and one half fold excess of 40% formaldehyde. The desired amount of asbestos is added to the solution and the mixture ma5netically stirred and kept below 5°c while a 50% solution of potassium hydroxide is added slowly until the reduction is complete (evolution of hydrogen ceases).
4HCOH + H2PtCl6 + lOOH- =Pt+ 4Hcoo- + 2H2 + 601- + 6H20 When the reaction is complete, the temperature is raised to so?c and held there for fifteen rm.nut e a , The cat.a Lys-t is then transfered to a fine fritted glass filter and washed with water and 6N acetic acid to remove excess base, and finally with hot then cold distilled water until the rinse water is chloride ion free (silver nitrate test). The catalyst is dried in air at 150°0 and is ready for use. -15-
HYDRAZINE REDUCTION (14) A mixture of chloroplatinic acid solution (1n water) hydrazine hydrate (one and one half fold excess) and asbestos is evaporated to dryness. The residue la heated to 110-120°0 with occasional stirring, and then cooled to -s0c. The reduction is carried out by slowly adding 20% sodium hydroxide solution to the dried mass with the
' temperature kep t at -s0c. The reduction is complete when nitrogen evolution ceases.
The washing procedure is the same one used in the formaldehyde reduction procedure.
SODIUiv~ BOROHYDRIDE REDUCTION ( 15) Chloroplatinic acid solution (in water) and asbestos· are slurried up and constantly mixed by means of a magnetic stirrer. Five percent sodium borohydride solutlon is added dropwise until its addition no longer causes Violent evolution of hydr.ogen gas. 18H20 + H2PtC16 + 6NaBH4 =Pt+ 6H3Bo3 + 6NaCl + 22H2 (16) Excessive amounts of sodium borohydride are required because acid hydrolyzes sodium borohydride and platinum ~s a hydrolysis catalyst here. The platinized asbestos is washed with distilled water, first hot then cold, until it is chloride ion free, and dried in air at 150°c before use. -16-
HYDROQEN REDUCTION (17) Asbestos and chloroplatinic acid are slurried up in as· little water as is required and constantly mixed by means of a magnetic stirrer while the material is' slowly evaporated down to dryness over the course of at least three hours. This assures- impregnation of asbestos' with chloroplatinic acid. The dried material ia then reduced at the desired temperature under a stream of hydrogen for two hours or until hydrochloric acid evolution ceases. The reduction is- carried out in a closed vessel (a tube furnace is convenient) which is thoroughly purged with argon before and after the reduction to insure the exclusion of oxygen while hydrogen is in contact with platinum, or a fire and sintering of platinum will result.
H2PtC16 + 2H2 =Pt+ 6HC1 The catalyst is·washed in the same manner as that prepared via sodium borohydride reduction. STEPWISE PLATINIZATION BY HYDROGEN REDUCTION (18} Thia method is essentially 'the· same as the previous one except that the asbestos- is impregnated with a dilute solution of chloroplatinic acid and the excess solution filtered off rather than evaporated to dryness. The reduction is carried out as previously described at 100°c. This cycle is repeated as many times as desired. This· procedure allows stepwise platinization of asbestos and • is thought to prevent the -formation of large metal nuclei thus- giving high surface area. Using a 3% solution of -17-
chloroplatinic acid, one obtains samples of platinized asbestos whose platinum content increases by increments of one percent with.each plat1n1zat1on cycle. This information has been determined via x-ray emission and wet analysis.
\·lET HYDROGEN REDUCTION An aqueous· solution of chloroplatinic acid is slurried up with asbestos in a cbnical centrifuge tube. A thin Teflon tube (eo mil o.d.) is inserted all the way down to the bottom and a stream of hydrogen is passed through it at a rate great enough to vigorously stir the mixture. The hydrogen reduces the platinum onto the support. From all appearances, platinum is completely deposited on the asbestos. This point will be examined in more detail later on. The chemical reaction is the same as the one written for hydrogen reduction. The catalyst prepared via this route is washed as described in the sodium borohydride reduction procedure. The reduction procedures outlined above are general. The quantities of reagents used varied with the composition of catalyst desired. As an example of an actual catalyst preparation, the following procedure is given. To prepare five grams of 9% platinum on asbestos via sodium borohydride reduction, 1.125 g H2PtCl6 (40% Pt) -18-
w~s dissolved in 50 ml. of water. 4.55 6 asbestos was slurried up in this solution and 200 ml. of a 5% (by weight) solution of sodium borohydride was added dropwise to the mixture(which was constantly stirred with a magnetic stirrer) to accomplish the reduction. The freshly prepared catalyst was transfered to a fine fritted slass filter and washed first with 3 liters of
hot distilled water and then with 3 liters of cold distilled water. The last wash water was tested for chloride ion with silver nitrate solution. The test was ne3ative. Had it been positive, the washing procedure would have been repeated until a negative test was obtained. The catalyst was dried in an air oven at 150°c.
ELECTRON MICROGRAPHS' Electron micrographs were taken of all the samples of platinized asbestos prepared. The following ls the procedure used in obtaining the micrographsr (19). The platinized asbestos was·mixed into a solution of 1% colloidion in amyl acetate and worked between two_ ~lass· , microscope slides in order to dispurse the powder. A thin film of this suspension was then transfered to a 200 mesh electron microscope grid, and for added strength a thin film of carbon was vacuum evaporated onto the specimen. The samples were magnified to 25,000x magnification, produced on the electron microscope screen and the -19-
photographs taken with Kodak Tri-X film using a 35 mm " camera.
SURFACE AREA DETER1-1INATIONS The specific surface area of platinum was measured for each of the catalysts prepared. Following is a brief description of the principles and procedure (20). Hydrogen chemisorbs strongly and rapidly on clean platinum metal surfaces over a wide range of temperatures. It has been shown (21) that this process is dissociative in nature, chemisorption proceeding until each metallic site is occupied by one chemisorbed hydrogen atom. The chemis·orption of hydrogen is the most widely used method of measuring the surface area of a dispersed metal phase. The volume of hydrogen taken up does not vary appreciably with temperature, but 200-30o°C ier usually used for the measurement. The assuDption is made that the hydrogen uptake results solely from the dissociative chemisorption on the base metal surface and not as a result of the reduction of any oxide which may be present. As oxide films are ,generally present on metal surfaces after exposure to air, it is necessary to give the metal a preliminary reduction treatment usually at 200-300°c followed by a long evacuation often at 400-500°0 to remove hydrogen. The catalyst is then -20-
cooled. to the adsorption temperature, and the chemi£orption of hydrogen measured. In practice a volumetric adsorption apparatus is used to measure hydrogen upt ake , A weighed catalyst
sample is sealed into an adsorption bulb and re-duced by hydrogen. The system is then evacuated and the sample brought to the adsorption temperature •. Small quantities of adsorbate (hydrogen) are admitted and. the adsorption calculated from the changes in pressure at constant Volume.
ELECTRODE FABRICATION The electrodes used in this study are modified Niedrach-Alford electrodes (22). They consist of a mix of catalyst, Teflon binder*, and conducting diluentiHt pressed into a supporting platinum screen which also serves as a current collector. A porous Teflon film on the gas side of the electrode provides for proper electrode wetting so that both the electrolyte and reacting gas have satisfactory access to the catalyst incorporated into the structure. This gives a good electrode structure
* T-30 Teflon suspension (from DuPont) was used as the binder. It contains 59.6% solids and has a density of 1.5, giving 0.900 g Teflon/cc of suspension. ~~An electrically conducting diluent was required in these electrodes because of the small quantities of Platinum present. -21-
tha. t enables the achievement of high effective area on Which·the electrochemical reaction can take place. The electrode structure is similar to that described in the background section. It provides for high electronic conductivity for electron flow either in or out and rapid input and removal of the appropriate participating ions• In the fabrication of test electrodes, the following procedure was used. Platinized asbestos and conducting diluent were thoroughly milled together with a mortar and Pestle to form the catalyst mix. The required amounts of catalyst mix (0.09-0.15 g) T-30.Teflon suspension (0.03-0.06 cc) and Rohm & Haas Triton X-100 wetting agent solution (2 or 3 drops of a 5% solution) were mixed together to form a cream-like slurry which was uniformly spread to cover the entire area of a 17.7 cm2 circular disk of aluminum foil. Two such similar spreads were Prepared for each electrode. The spreads were dried at 50°c and the temperature then slowly raised to 350°c to expell the volatile wetting agent (which is also present in the Teflon suspension) and sinter the Teflon. The two spreads are pressed one on each side of a 45 mesh Platinum screen which is woven from 8 mil _wire at 350°c and 2230 lbs. per square inch pre·ssure. (Other pressing Pressures were used, and these will be discussed later). -22-
After pressing, the aluminum f0il was dissolved frbm ihe electrodes with warm 20% sodium hydroxide solution, and the electrodes were thoroughly washed with distilled water for one hour. Afte'r washing, the electrodes were dried in an air oven for one hour at l50°c. A wetproofing Teflon film was then applied to one side of each electrode by spraying a T-30 Teflon suspension diluted with seven volumes water. Films corresponding to 1.6 mg Teflon per cm2 were applied by spraying 2.25 cc of the mix onto a 5"x5" area. A Paasche Type V artists' air brush was used to apply the film in the form of a fine spray so that several tracings could be made in order to obtain a uniform thickness. The spraying was done on a hot plate at 120-150°C to cause evaporation of the water as the spray hit. This prevented the development of wet areas which could run after spraying and form a non-uniform film. Followin3 the spraying, the electrodes were placed between the platens of the press which was at 350°c. The jaws of the press were closed to within 1/8 inch for about one minute. This procedure evaporated the wetting agent and sintered the Teflon to give the desired film. The electrodes were Pressed as previously with aluminum fmil disks placed on either side of the electrodes. A final dissolvins of the aluminum foil follo~ed by a one hour wash with distilled -23-
water completed the electrode fabrication process.
TEST CELL All the electrodes prepared were evaluated as hydrogen anodes and as-oxygen cathodes in test fuel cells. The test cells contained the test electrode, a counter electrode (which was a standard Niedrach• Alford electrode (22), 1.e., an electrode prepared similarly to the test electrode, but with neither support nor diluent, and containing 34 mg platinum black per cm2) and 5N sulfuric acid electrolyte. The test cells were so constructed that the two electrodes were separated by a 3/8 inch thick-It inch wide inside diameter Teflon spacer which forms the electrolyte compartment. Appropriate F.E.P. Teflon gaskets, 1/8 inch thick Teflon spacers, and 1/8 inch thick stainless' steel face plates were arranged in the cell in such a manner as to provide 1/8 inch thick gas compartments on the outside of each electrode. Gases and electrolyte were circulated into and out of the cell via 80 mil (o.d.) platinum tubing and #13 Teflon tubing. Gases were circulated at the rate of
20-30 cc per minute, and electrolyte was circulated by gravity feed at the rate of one CC' per minute. All the electrodes-were 17.7 cm2 in area, but the gasketing in the test cells was such that the working -24-
area of the electrodes was 11.4 cm2• Current densities
(m111iamps per cm2) were therefore calculated by dividing the currents read on the· ammeter by 11.4.
!v'.EASURING PERFOR!·~ANCES~ Fuel cell performance is conveniently analyzed by
.taking polarization curves, 1. e., measuring the vol t age s at different current densities and plotting these parameters against eachother. c·urrent densities- (currents) were measured with a General Electric,Type OP-H D.C. ammeter and voltages were measured with a Rubicon Model 2730 potentiometer. In the first part of this work, polarization curves were taken with the aid of a low frequency (60 cyclea- per second) interrupter Kordesch-Marko bridge (23). Its Principle of operation and use are briefly described here. If the current flow of a fuel cell on load is suddenly interrupted, the cell ~oltage will increase toward its value at open circuit. The purely resistive component of .the polarization decays more rapidly than the activation and concentration components- of the overvoltage. By interrupting the circuit for short time intervals, measurements can be made of voltages without IR lossas. The Kordesch-Marko bridge employs an interrupter circuit whereby current flows through the cell during only half of the cycle. Measuring the· voltage -25-
during this period (current flowing) gives potentials corre~onding to direct D.C. voltage measurements. During the other half of the cycle no current flows, but the interruption is so brief that the measurement can be made before decay of activation polarization and concentration polarization can occur. Voltage measurements during this part of the cycle give IR free data. What was just about the Ko r-de ac h-Mar-ko brid' ge holds· when measuring performances of high platinum content electrodes, however, with low platinum content electrodes the· time constants for the non-ohmic polarizations are less than the bridge cycle time and as a result, polarization transients can occur throughout the cycle and non-ideal voltage fluctuations can occur (24). Fi5ure 1 shows behavior of high platinum content electrodes (a) and behavior of low content (platinum) electrodes (b), where decay is evident, as photographed from an oscilloscope screen with a Polaroid camera. The electrodes tested in this work were all of low
Platinum content (no 3reater than 2 mg per cm2) and as a result, the data obtained with the Kordesch-Marko bridge cannot be used quantitatively. In the first part of thia work, optimum catalyst preparation and electrode content and fabrication procedure were elucidated, and quantitative data are not essential for the interpretation
-- -26-
,/
·------~ ------la
lb
Figure 1. Waveforms from cells under load via a Kordesch- Marko 60 c/s bridge. la shows waveform from a cell with high platinum content electrodes, and lb shows waveform obtained from a cell with a low platinum content electrode. -27- '
of·res~lts. The object of the first part of the research was to determine relative performances so that conclusions could be drawn as to which is the beet way to prepare a catalyst and fabricate an electrode. With this in mind, IR included performances are reported using data obtained. with a Kordesch-Marko bridge. In this part of the work, electrodes were compared with eachother, and IR included data are helpful for this comparison. 'For the second part of the work, in which performance • per unit area of platinum was analyzed, it was necessary. to determine quantitative IR free 'data. To accomplish this goal, it was necessary to determine performance under steady D.C~ drain. To obtain performance data under the condition of steady D.C. drain, the method of Niedrach and Tochner
was used (24). The celle0tested were discharged through a resistive Load , IR inc·luded vol ta.gee were measured directly across the "termi'nals or\ the fuel cells with the potentiometero To obtain the IR drop, and thereby permit the calculation of IR free voltages- the ~urrent was interrupted arid the voltage change was photographed with a Polaroid camera off the oscilloscope on which the Voltage was being monitored. At sweep speeds of ' 100 microseconds per division, sharp steps corresponding to the Ir drop could be readily recorded and measured.
t • -28-
Throug,P the-use of a calibrated oscilloscope, the voltages could be determined. Figure 2 shows a typical oscillos-cope photograph. The sweep speed is- 100 microseconds per division, the current density 8.9 millarnps per cm2, the vertical scale 40 millivoltw per division, the IR Included voltage o.630 v, and the IR free voltaue o.649 v. In the measurementff, a Clare mercury wetted relay was used for the interrupter. The reproducible delay in the• opening of the relay after the coil is energized, permits ready synchronization with the oscilloscope sweep. The oscilloscope was triggered at- the appropriate time by Using a Tektronix Type 162 Waveform Generator and a type 161 Pulse Generator. The oscilloscope used was a Tektronix Type 536 with a Type T Time Base Generator and a Type D High Gain Differential c-alibrated Preamplifier plug-in. A Polaroid camera w~s mounted over the screen of the oscilloscope and Type 47 Polaroid film {3000 speed) was employed to obtain the photographs. -29-
Figure 2. Oscilloscope photograph showing the effect of a brief current interruption on a cell under steady D.C. drain. The sweep speed-is 100,i(sec/div, with 4o mv/div on the·vertical. Current density is 8.9 ma/cm2, IR included voltage is 0.630 v, and IR free voltage is 0.649 v.
I. -30-
RESULTS- AND DISCUSSION
PART. I. Initially four samples of platinized asbestos were prepared. These catalysts {A to D) contained twenty percent platinum on asbestos,and were prepared via hydrazine reduction, formaldehyde reduction, hydrogen reduction at 4oo0c, and sodium borohydride reduction respectively. {Table one gives a summary of all the catalysts prepared, the specific surface areas of the platinum, and reference to electron micrographs of the catalysts). As ie evident from table one, the surface_ areas of the platinum in these catalysts are not very high, the largest being 10~0 m2 per gram for number D. Figures· 3 to 6 are electron micrographs,of catalysts A to D respectively, and the following may be noted from these photographs-. The hydrazine and formaldehyde reduced.catalysts (figures· 3 ard 4 respectively) show very little adhesion of platinum to asbestos, the sodium borohydride red~ced catalyst (figure 6) shows intermediate adhesion~ and the hydrogen reduced catalyst (figure 5) shows the best adhesion of metal to sµpport. The relative crystal sizes of the platinum in these fc;>ur catalysts are in agreement with the measured platinum surface areas. SOd1um borohydride reduction gives the smallest platinum
I ·' -31-
TABI.E 1
Catalyst Method of % Pt Specific Pt Electron reduction eurf~ce area micrograph in m /gram · figure no.
A Hydrazine 20 6.4 3
B Formaldehyde 20 8.5 4
C' Hydrogen, 20 4.4 5 4oooa
D sod i.um 20 10.0 6 Borohydride
E"' Sodium 9 11.0 13 Borohydride
F Wet Hydrogen '9 10.3 14 G Hydrogen, 9 12~5 15 l00°C
H Hydrogen, . 9 12.3 18 Room Temp.
I Hydrogen9 9 11.3 19 '200°0
J6 Hydrogenj 6 ' 23.0 22 l00°C, 6 cycles \ K Hydrogen, 6 17.8 23 100°0
Table 1. A summary of preparation method and properties of the.catalysts. Reference is given·to figures of electron micro~raphs of the catalrsts. '
I ·' -32-'
...'Ill
I I I I... . _ ...
Figure 3. Electron micrograph of catalyst A (hydrazine reduction) 25,000 x magnification. -33-
i j -. \.,. , I . j (\ '
(
I' i I I
----~--~v , ·., - ..
Figure 4. Electron micrograph of catalyst B (formaldehyde reduction) 25,000 x magnification.
I ' -34-
, , , . I
j j I
I ;i
I
. J
Figure 5. Electron micrograph of catalyst C (hydrogen reduction at 4oo0c) 25,000 x magnification. -35-
Figure 6. Electron micrograph of catalyst D (sodium borohydride reduction) 25,000 x magnification. -36-
Pa!'tic~les and the highest surface area, while hydrogen
reduction at 4oo0c gives the largest platinum crystal si~e and the lowest surface area. These results are not surprising.
Electrodes 1 to 4 were prepared incorporating catalysts A through D resp~ctively. These electrodes contained 2.0 mg Pt per cm2, and 1.0 gram of finely divided (surface area of 7.6 m2 per gram) tantalum powder. A summary of the content of these electrodes as well as· • all others prepared during the first part of the work is given in table 2. * Figures 7 and 8 show performances of electrodes 1 to 4 as hydrogen anodes~ and oxygen cathodes respectively. Ct>mpared with performanc.es subsequently obtained, all these performances were very poor, however, electrode 4 (sodium borohydride reduced catalyst) showed the best Performance on both hydrogen and.oxygen. This is in line With the higher surface area o~ this catalyst. The poor performances initially obtained, even with r an electrode containing a catalyst with moderate platinum surface area (10.0 m2 per gram in the sodium borohydride
* All the electrodes' tested' in this study contained 0.12 cc· of T-30 Teflon binder, 45 mesh platinum screen, a ~etproofing film on the gas side of 1.6 mg Teflon per om2, and were pressed at a pressure of 2230 lbs. per square inch, unless otherwise noted. -37-
TABLE 2
Electrode Catalyst Amount of mg Pt/ Diluent Amount of Number Catalyst cm2 diluent (in grams) (in grams)
1 A 0.18 2.0 Tantalum 1.0 2 B 0.18 2.0 Tantalum 1.0 3 c 0.18 2.0 Tantalum 1.0 4 n· 0.18 2.0 Tantalum 1.0 5 D 0.18 2.0 Graphite 0.3. 6 D 0.18 2.0 Boron 0.3 Carbide
7 D 0.18 2.0 Shawinigan 0.15 Black 8 D 0.18 2.0 Vulcan Black 0.15 9 J6 0.18 0.6 Graphite 0.2 10 J6 0.18 o.6 Graphite 0.3 11 J6 0.18 o.6 ·Graphite .0.4 I 12 I 0.18 \ o.6 J6 I Graphite 0.6 13 E 0.18 \ 1.0 Graphite· 0.3 14 F 0.18 1.0 Graphite Os3 15 G 0.18 1.0 Graphite~ 0.3 16 H 0.18 1.0 Graphite 0.3 17 I 0.18 1.0 Graphite 0.3 18 ,.., J6 0.30 1.0 Graphite 0.3 19 K o·.30 1.0 Graphite 0.3
Table 2. Composition of electrodes prepared. All electrodes contained 0.12 cc T-30 Teflon binder and were pressed at 2230 lbs. per square inch. ~38-
:-~.:...- ,_...._ ._,_ ,~ 0) --·-~ ...... ,_,_ .!-l-i- , - CJ) ~-- -- I- - •--- __ (!) . A s:: I-'-'- ,_,___ (!) 'D 'OC\J cQ 0 -·- 't:l 0 H SW ...... --i 0 .H cd 0 0 +l .-L' ~ .µ•+:> 'O'-..,. (!) 0 0 ...._ ,__ +:> er, 0 S::"S:: +l +l ..--..o ;:s __ .__ 0 (1) p.., 0 C\l Pi s:: 0 '(j 1--i- ·-- Q) rl ..-! .µ i::: rl o sr (l) ' rl (l) (l) +l 0) 0()0 --n H Q) ti) C\l ~ H s:: +:> .. 0 +l H I +l .0 0 s:: (l) ~ ~ Cl) C\l 0 ..--i ;:s ~ 0 'O i-..-- 0 'O (l) p. Cl) I"\ (1) - +' ro or-i .,., ;. ~ »+> (!) Q) rl 0 Q) +:> H ..c H'O.£! (l) ;:s H 0 Cfl -o l P,o.µ 'O ;:s >. '- '- I->-· Q) s::- .£! -n: Q) (1) rel _c: :>o C1.l +l +:> ;;,;Q) H -o (l) 0 H Q) s:: Cl) <, >. H H I ;:s ..::t rel ·o O>O OJ I"\ (l) .£! 0 I 0 H Q) s:: (!) s:: ,0 I 'O (l) • 'O (!) s:: ...... --l 'O ' (l) s:: I 0 • V)- ' I I ' I I I , ___ I ,_ 0 0 ~ ~ I'<· I ~ ,L I 1, ,, I I Vi ~ I ~ v ,_ ~- ,_,_ ., v ,_ ,_,__ ,_ '-" ~ .... J .r ... ,.... J ,_..- ' ·~- - . .- ..:. ,_ .,. ·- ~- ·- -- ~- 0 "'I"" 'O 00 J> q 0 ----- 0 0 -40- . reduced catalyst) suggested that the electrode composition and/or structure needed modification. The first variable studied was the conducting diluent. Electrodes 5 to 8 were prepared containing identical quantities (0.18 g) of the beet catalyst from among the four prepared initially, D (sodium borohydride reduction). These electrodes contained 0.3 g graphite (electrode 5), 0.3 g boron carbide(# 6), d.15 g Shawinigan carbon black (# 7), and 0.15 g Vulcan carbon black(#· 8). The performa.nces of these electrodes are compared with that of electrode 4 (tantalum diluent) on hydrogen and oxygen in figures 9 and 10 re,spective~y. The polarization curves show that all four test diluents· are superior to tantalum. We see that boron carbide is best for hydrogen oxidation, and that graphite is superior for oxygen reduction. Both these diluents are better than the two carbon blacks. In general, oxygen performance is a better guideline in picking the beat electrode, because hydroge~ is such a good fuel that it is probably insensitive to the electrode employed for its oxidation. Since the graphite containing electrode Performed best on oxygen, it was chosen as the diluent to be employed in subsequent electrodes. Having decided on graphite as the diluent of choice, the optimum amount of graphite per electrode was next determined. Electrodes 9 to 12 were prepared containing -41- ... ~ _,_ , __ -- eo --1------j-- -- -ii~ '.~~n~ !U.:: s:::: co ,_ ·1-''- - co I .,; Q) __ ()) +:> +:> S::::.-0 '!r:r 1-=ii=.= a:l 0 .,; 0 I 0 4-; (!) Cll s, .q"' I/ I/ I/ .c... s:: .o rl +:> +:> 0 ...... 1-!-<-· (!) I/ I II /_I- (!) ..c 0 co ~~ ,__ 0 bO +:> 0 (I) (\j 0 L, -·-- (!) 0,f-) I/ I/ I I/ 0 Orl:I! ct! _,_ +_,_ ,_, _ _ ,_ rl s, Cl) OJ (1) ,-j ...... ~- -·- •-11i ~'- (!) -- ,_ 'O <'+-! (!) I- - .c: z ,D ...!:!! I H-""" >.G-i +:> 0 '0 l.[\ 0 ·-·-- s, _e --_,_ _,_ I I (!) cd .c... s:: cd -- I I ! 0 I Q) cd I Q.--j • (l) I I s:: 'O .o .o ---·- I I OJ (!) ...... 0 .c: .. s:: .C... J I I 0) +:> (]) ti) cd +' (l) cd s::: t-1- -1- ._. (!)CD 'CJ -1..:> +:> >. 'O 0 0 ~ WO s::: O'l rl .,..; ,D I I s, 'O s:: s::: c 0 ,D s:: .C... 1-~. l.1 "- ;:::i I I ,... """V) I I I .- ~... ~ I I I I A I I I I I I I/ I II _/ I I I I I j I I v I i I I I/ 0 I I ~ 0 ~I ... I I J I " 'f 'ti I I I I ! >-.. i I I/ I I I / I I / I 77 J / I - / -~- V I .. ~. ,_, i '·~ 'I I -: I ' / n7 / / / L / / /[_ / / I 1/ -·- / /I I -- --, I ,. / /1 I I I ~··1 \) - _ ..... I I /LJ' ,. 'II' I I / ~·- ..... - - _., , _ "· -[;::: ~- 7!:'.'e ·- - ..,..,. 0 0 ...:.. -43- o.2J 0.3, o.4, and o.6 grams of graphite respectively. These electrodes were otherwise identical (see table 2). Figures 11 and 12 show the hydroe;en and oxygen performances· of these four electrodes respectively. Examination of figures 11and12 shows that 0.3 grams of graphite is the optimum amount of thia,diluent per electrode. More or less graphite results in poorer performance. Halving the T-30 Teflon binder content had no effect .on performance. Less than 0.06 cc T-30 could not be used,. since this was the minimum amount that could be employed and still obtain an electrode with at r-uc t.ur-a.L' integrity. ,Reducing.the pressing pressure from 2230 to 638, 1590, or 159 lbs per square inch showed very slightly improving performance with decreasing pressing pressure, but at all the Lower- pressures, the electrodes prepared delaminated. Having optimized electrode fabrication and composition, the next step wa~ to 1nvestigat~ ways of preparing more highly active catalysts (higher platinum surface area) than those initially made. The hydrazine and formaldehyde reductions.were abandoned because they gave catalysts with low surface area and showed very little adhesion of Platinum to support. The sodium borohydride reduced catalyst also showed little adhesion of the platinum to asbestos, but the relatively high surface area obtained I ' -44- I I I I -l-+-+-f-~-+-1- -·- 0 -~l-l--i'-11--H-l--f-+-j.....j..-A-+->-+-+-l-!1·- -~ ·-1- 0 ' -1--1-l-!.-l-"-J.-'J--1-4-f-++-++-1-+~l-++-+-++-t-+-+-+'-~""l ...... ,., ...J...~-+~-+-IH-+-H-*-+~;-HH-r+--r+-t-t-t-t-t--r-t-to .E -LlJ_j..4µ-A-µ...+-¥+-l-+-H-++-HH-+-+~~-+-H--H~~ I I II /I/ I , T I/ I I , I/ I I I I +H I I I t-+-+-+...{/~1/~/q..,,~1~4-f-++-J.-+-+-t-+-!-+-f-+1-l-t+-f-++-t-t-J-i-t-i-t-1·-tti-:i""tlrt-rlt"l-- I 1t1-e• ~~~1--J...,~~~'~,~:J... I J I I/ I I I I ·' _,.:. I I I I I/ I I ti- ll- I- ·;; I I I I/ I I ~!---'- "..J ..... It' .. ~ ' I J I J I ,, J I/ I. , , I/ - . I I J , , " __ t-·J~_ l ,_ I/ / - I I ! I I I I/ I I I I/ - I '. ~ , .. ;\' N '-!° ,,. .. - v v ...... - ~- ...... lD ... -- -- .. , .J> 'iJCrJt./011 ~ ·o _____ 6_ _ _,_11.~'l-0 0 -46- looked encouraging and warrented further investigation. Perhaps a lower platinum loading would give higher area. The hydrogen reduction procedure too was chosen for further investigation because of the excellant adhesion of platinum to support. It was· felt that lower reducing temperature would give higher platinum surface area. The wet hydrogen reduction method was also tried because it . was felt that this method would give a catalyst 1n which all the platinum would be deposited on the a abe a t o a, leaving no free metal. Catalysts E, F, and G, all with 9% platinum on asbesto~ were. prepared via sodium borohydride, wet hydrogen, and hydrogen at 100°0 reductions respectively. Their platinum surface areas were measured as 11.0, 10.3, and 12.5 m2 per gram respectively. Figures 13, 14P and 15 are electron micrographs of these catalysts. Figure 13, showing the sodium borohydride reduced catalyst, is very similar to figure 6 showing the 20% plati~um catalyst prepared the . same way. The crystal sizes of the platinum are about the same despite the differing platinum quantities. This- agrees with surface area compariso~s of the two catalysts (11.0 for catalyst E compared with 10.0 for catalyst D). Both catalysts contain a considerable amount of unsupported Platinum. Sodium borohydride reduction does not give a ( highly supported catalyst, and apparently the surface -47- \ • ·~ : ", ~ .. ). ~·..· • • .. .·: I.· •... '• . .. ·' •' .. .. · ..1~~ .: -i. ·1. ·-1 "' ~;> • 91•t ., '1 ·'· .. ,, • ·~· l~ .,.,;, : ~~....!~ ~...: , .:.~ ~..--. ~- ;..: :A. l,..~. , ~. ~ ...... ::~ .. ··:·:~· . 1 ...... J ; ...... ' . l ~, ''1~ '· ;."'. ·' f, " Figure 13. Electron micrograph of catalyst E (sodium borohydride reduction) 25,000 x magnification. 1 •• -48- , . \ ...· . . : ' . . . .... Figure 14. Electron micrograph of catalyst F (wet hydrogen reduation) 25,000 x magnification. -49- ,,...... :. ' . .' I, . .. ,· ...... t,.·· Figure 15. Electron·micrograph of catalyst G (hydrogen reduction ~t 100°c) 25,000 x magnification. ' .• -50- area of the platinum produced does not vary much with the quantity of platinum used. Figure 14, an electron micrograph of catalyst F (wet hydrogen reduction) shows that the platinum is well bonded to the support (no free platinum can be seen) but seeding occurs, 1.e., platinum preferentially deposits on platinum allready reduced rather than on asbestos. Thie gives rise to larger crystalline size and lower surface area (10.3 m2 per gram for catalyst F). Figure 15 shows small platinum cryst. als deposited on asbestos, with very little free platinum (catalyst G'). The higher platinum surface area of this catalyst~ . . (12.3). bares this out. Hydrogen reduction at 100°0 appears to give the best catalyst so far. The electrochemical data support this belief. Electrodesrl3, 14P and 15 were made inco~porating catalysts E, F, and G respectively. These electrode~ were R similar in·every respect {containing 1.0 mg Pt per cm) except the catalyst (see table~). Figures 16 and 17 show comparison polarization curves for these electrodes performing on hydrogen and oxygen respectively. The order of performance for these electrodes is related to the platinum surface areas of the catalysts. Electrode 15 (100°0 hydrogen reduction) is superior to the otqers on both hydrogen and oxygen,# 13 (sodium borohydride reduction) intermediate, and# 14 (we:b hydrogen reduction) poorest. • I •' -51- 0 ...... ,... V) +> s::: '-'-- IV) 0) 0) . 0 .- ++ 1-·- ·+-~ (!) I - (!) Cf.I t.{) •rl ,...... -1'-'- II FF ,_, J :fl: 'O (!) __ I/ ! +> s +> 0 -17 I I 0 G-1 H 0 0 I I +· •H +:> 0 I 0) ;:J ..:;t ..c ::::l,...... O + +> m" +>'01"\0 'Cl I r-'-HI - - . s:::o I 0 +> 0) CJJ QJ s::: (l) Or-1 (l) s::: O©:»O...C:·rl Hori 11+ -H+-· rl Q) Q) 'O rl 0 +:> +> .. - I -t+,__· (!) C(j I ,__ t04-t 0 H-riCO (!) 0 f.:: 1)- ,_ - ·~ I OG-1 S:::+> P. ;:;-:: '-..... 'O ;:s 0 rr .H H f cS. 1 I v """' 1;(° • .. ~" ,...... ~ I j / \...... ) ) J '>-... J I I/ ~ ) I J () ~· I / ln s: + / - ' Q) I } A I / I I I '1-' ·/ :s; J J (!) I i.• J I/ I s... .:S I , I 00 J I/ ;;;; j 0 "f I I - :!:.) !-.. /II.' I ""' ,_ I ' ' I I I I/ I I ' I I '~ I "fr: ~ - Io I J j / , _ _J ·+ I ' 1/ 71'/ I t I 1•, .I .-., ' 'll v_, .... r,,-,,. ...-~ - I ·- 0 O· 6 -52- l/ J ~l----i.- II I , I r--t-r--1-1-+--+-f-+-J-1---+--1--1--1-1,--!-~.._.-!_,_,_~T - _ t-+-H-t-H-++-H-+-+-f"~-1-H-+-++++•"L..... J I/ 0 t-H-t-t--t-t--t-H-++++-+-Hl-+-++1-A-l-+!!'!-~ 1- t<\4 m r+ rl rl Q) Q) Q) I II ro ro re 0 0 0 I/ H H l'... I +.) +' +:> o o 0 Q) Q) Q) : ) r-l rl rl I i:x.1 ll-1 o ...... rr~ t-+1~H-+-t~-+-+-+-f-l--l.1-1--/l-+-~~--1----l--l~...I. 0 ~ II II ll t-+-Hl~-+-+-+--~,lf~-+-~)~-!--1----~~---+-l-l------l--l-l~ ~ ~ o I/ J ; I J , l-t--t-t-t-t--H-t-+-l--+~i--t-+H+H+H-Y-1 --J--ll-'H-l---~--l--l--_i- ._ -Y--7-1-+--+---1--1--1--J-1--+-1---+-l--1---+-+--1 II, I I I/ I/ j , -, I / I J IJ -53- Since the 100°c hydrogen reduction gave the catalyst with the highest platinum surface area and the best performing electrode, it was decided to further investigate hydrogen reduction as a method of preparing platinized asbestos. It was previously shown that 4oo0c iss too high a reducing temperature, and that 100°0 appears to be satisfactory as a reduction temperature. The next effort was applied to trying intermediate and lower temperatures. Catalysts Hand I were prepared at room .• temperature and 200°0 respeciively. Figures 18 and 19, electron micrographs of these two catalysts, show that they ,are very.similar in appearance to eachother and to catalyst G (loo0c reduction). Their surface areas of 12.3 and 11.5 m2 per gram respectively are close to that of catalyst G (12.5 m2 per gram). Electrodes 16 and 17 were prepared lncorporating catalysts Hand I respectively with platinum loadings of 1.0 mg per cm2• These electrode~ were e.valuated on hydrogen and oxygen and compared with electrode 15 •. Figures 20 and 21 show the performance curves for all three electrodes on hydrogen and oxygen respectively. The performances obtained fron all three electrodes were about the same~ with' 15 (catalyst reduced at 100°0) being slightly better. than the other two. Thie effect is in line with the higher surface area of catalyst G used in electrode 15. 1 ' -54- \ • .. ... ,, ' .. . . ·"' \ . ·'tt~ - ... .• ...... ' Figure 1$. Electron micrograph of catalyst H (hydrogen reduction at room temperature) 25,0000 x magn1f1oe.t1on. -55- j· "" .. \. .r.» ... ,.'<' \ .1 ~::, '""' I ~· f t· I \,\ Figure 19. Electron micrograph of catalyst I (hydrogen reduction at 200°0) 25,000 x magnification. t ·•. 56 - ,., 0 - ' Cl.) +> >-- 1--"- ru++-- · CJ-J=±- ~ -1- -=r-·-3----1- ·-"-,_--~m-~= ~- -~~-- I ~-I ,- Q) ...---.. c (/) - .. -·- ;- 1- 1->-- 'O (1) (1) 1-1- s:: Ll.L I- - 0 ro ..C H 'U .. 0 111 I-'-._ -r1- G-; ..-; Q) I ' ::... +' +> 0 -::t -rl -- 1- . mo ..-;cH HO +> +> --· - - f- ---r () s:: ;;: G-; .;..J Cl) o -17 1 ~I L (l )1=:+>0 ..-; () (\J ;::I If__ ,_ i--1- rl Q) 0 -rl 11) 'U Q) ::c 'O -- Q) VJ Q) +' Q) rl (!) - 1--1-1 I ,_,._ OG-; 0 'O +' Q) ,_,_ 1+ z H __ I±,._ H H4-i;::IO,l1 Lf\ ,_ - ,....,._. 0 'D (1) 'O H ,.....,. (1) ...---.. f-i= -o 'O ' ' n ., i--.O G-; ?-. G>+''d::... II s:: H i;: t+ i /I/ II I ..c Q) ::... o (l) Cll 0 ;::I 0 - - - .c (1) C) 'O (1) ..-; +' ..-; 7rtt _,_ --- i::+:>S::ri:::IS::+' +' m +' f-- '-- -~- 0 Q) (!) 'O ci:l ~ C) H o ti) tt) Q) +' rl ;::I Q) ;::I t:-S::o+:>::...roo -o 0.'d - I J /I/ ,_ ,_,_ - rl ..-i H ro ::... . Q) El (l) /II -r ~'Ci (!) (/) I +> p, ::... (l) H -o 0 ~+> +> 0 cd +> ~ ..c ..c O'.l O'.l Q) b:) 0 D .n1 I m (/) I ~(!) rl 0 so tJ -''1 .c:! rl 'l'.:Hi! a> 0 00 .. Q) oromo +:> 000 I I I \() ::.....-c ©+>..C. ~ rl ::... C\J ,._,._ I . rl ;::1 ..c 'd m +'C\J rl ...... - .µ;;::oomso ,_,._ ,._ . /I/ I - -1- .. m ~ 0 o H UI \0 t:- I UI::...+:> ...... +> rl rl rl ,_ t- ,_ ,._ rl (l) cd (\J • +:i 0 I p, . s {/) ~ QJ (l) QJ (l) S©+>OQJ rl 'O 'O 'O Q) ::... ;j...... _::... !iJ (1) 0 0 0 I I +:> ::> 0 +> ;::I s H::...::... /I/ +> ~+> ..c +> +> +:> - _,__ I +' cd 'd cd ..:;:.t 0 0 0 () I I I I s:: ::... Q) b()):...f'l\S:: (!) Q) Q) ->-- +--I I (!) Q) ..-; s Q) ..-I rl rl rl - ·rl p,::... P.~C (:£!~µ) I I .o !:J ::...o s +>OJ I l.:: (!) c\) . (!) ·rl'-... II II II m +> o rl +:i ;;:: r<\ I C> ' I I I " ' I I 'I,/ I I I// I I 1- I I ""''- I J ' /J I J I I J 7 II ~ , __ .. -1·- t-t-·.!' 'I" ·i.i -- /1 I • ~ , I I I IJ If ~ I/ ' ~ ~ ., ' I I ~/I/ I •// I 'f:>.. v<:: I '""'~ ,_,.. "'J 00 C?. c:;> - 0 0 -57- ,_,_ ,_ -+-+-1-+--t--+-+-1-+--i-lH--1->- lj_ __ l/_'.- .....rTo-r - ..... --r- -T---,-1-r-,-+-tT-t-+4-+-+-l-+-_, T~t~t 1 f I .-~--L-.------~ -1--- r+-+-r-r-1--- - 1/ ..... _,_ r+-+-t-+-t-+-+-l-+-f-l-+-ll-1--1- r- .j -~ - -.- 0 I/ ·-,-1-•-1.; ~ r- i-- r- --!- ·-t-+-t--t-+-t-f-+-t-,f-1-,.,,_.,,.1.~•-+--i--<- , I !/ r+l-+-~-+-+-+-l-+-+-1'-i-+-J'~,~c:..:.V-''i-f·~-"1-~-t-+-f 1-+-t-T4-t--t-l-t-+-1'-f--t-~J :_ --t--t-,H--+-l-+-f <) ~ :-t-+-t-+-t-t-+-t-1-t-T--1f1.. 1fIf+ H' --+-1-+-+-+-+,.-(of I ri I t l!l \0 l:'• 1-+-+- +:> ..µ 0 0 0 (I) (1) (I) rl rl rl o.-....., f.:r l r=1 µ.') 11 o ... I 11 l l 11 II " ~ I ~ 1 ~ ~ <:» I "')- 'd ~ 0 <') II J - l.ll ~ fl 'f 'A I I Ii ..µ !-!-++-l-+-t-t-+-t-t-+-t-t-t-t-+-t-t--l-i-f--l-1f-t--t-t-++-1-+-+J··~l+-1'--t--t-1-r-+-t-+-+-+-+-t-+-l-+--t-1-++-1 s: t-t-1-~-1-+-t-t-+-t-t-t-t-t-t-t--t-f-+-,-i-+-t- I R 1- I ~ I I I ' I/ I I '/J I , ,, l I /. t:tli:~I ,~~:::r:t_."""...- tt1:1:ttttj:Jjt::tr:tt1::t!:tt1:1:tttt:tt1:ttttt:ttt:tl!:1o 0. rl 0 () 0 - .: -58- The catalysts with 9% platinum on asbestos, with only one exception, had higher platinum surface areas than the catalysts containing 20% platinum. Thie seemed to indicate that in general, the lower the platinum content, the higher the specific surface·area of the platinum. Since 100°0 hydrogen reduction was- the best method of_ preparing platinized asbestos (g:l.ving highest platinum surface area and best fuel celi performance)' catalysts were prepared via this route with lower _platinum loadings. Catalyst :J6 was prepared via a stepwise 6 cycle platinization depositing 1% per cycle, and catalyst K was prepared in a similar fashion to catalyst G (loo0c·hydrogen redu~tion) but with 6% platinum on asbestos. The platinum surface areas of . I these two catalysts were 23 and 17.8 m2 per gram respectively. In figures 22 and 23 (electron micrographs of catalysts J6 and K respectively) we see that catalyst _ J6 has the finer platinum part~cle, size of the two, but that both have smaller platinum crystals than a~y of the_ previous catalysts prepared. These observations are in agreement with the measured platinum surface ar~as of these catalyst.a. Electrodes 18 and 19 were prepared incorporating catalysts J6 and K respectively, and containing 1 mg of platinum per cm2• The performances of these electrodes \ ,• -59- ' . . -~ •. . ' ·• 'l:\ ' ...... • ••. ,•,• ' I .. , : ... •' (' Figure 22. Electron micrograph of catalyst J (six cycle platinization at 100°0) 25,000x m2gn1f1cation. ' I,•• -60- ____ 1.,... J • .... ·I \ , Figure 23. Electron micrograph of catalyst K (hydrogen reduction at 100°0) 25,000 x magnification. ' .. -61- ~o~s:: s:: ..-; s:: ..-;0 +='m ..,0 +=' N +l m ..-; ct:! NS:: N •rl ..-; •rl s:: +-' s:: ..-; m ·r1 +:> r-l +=' --·h ·- --1-':-1--i-r-r-r-t-t·-r--i ctl P. cd "j-i--j-,...:-i- -, .. - - 111 -- - t-t--t-i-l"--t-+-t--t--111, j t- -- --i- -1- ·-r- :- - ~ rl rl P. Q) P. ,-,-,.,1.,1. ~ J ··r- _,_,_c-:;~. i OJ o, ..-; P. r r-- -r- -·;1r-r-r-;'-r-11LLL - ~ - r- ·-. "") Q) ;:,;: Q) -f-r-1-'-/ +:> P. .,_:> I ii I (I) Q) 0) I I +:> rl (I) rl I ...... I I .I I I t-1-+--HH-"ilttt11+-t-t-+-!~1t-t-t-+-t-+-+-!-+-Hr-t--+-1 1-+~-+0 t-1-+--r-1H--i1-~11-t-r-t-ii-r-+-1-r-ri-r-i--t-r-t-+-t-+-+-t1-+_V, 111--r-11-r-~r-r-1-rt,n-r-t-r-T-1·-j-+-t-t--r-1!-t--;t-t-t--t=t" Lf\CO 0\ I ' rl rl rl 11 J Q) Q) Q) 'O 'O 'O J j 0 0 0 I I I H H H +:> +:> +:> r-r.--r-1·-r-r-r--11n-r-r-i-t-t--r-1-t-t-·t-t-t--t-1-r-t-1--t--t-1 0 C) 0 t-1-t-T~,1~/'t-t-t-11+-t-t-+-!!-l--t-1-t--t-!-+-t-t-t-t-t +-t--t-+-I Q) Cl) Q) rl rl rl I r.:l 111 111 ,.-, ~ ··r11--tr1.r+-t-+--H1-r-;r-t-t--t-t-t-i-t-t-t-1-t-+-+-t C> j -0~ 11 II II - fl "" " I ~ Q I J J -jJ'>-- •-H-, -rr-t"i-j-+-rt-r-t-t--Ht-r+-t-t-rT-'t-I ~71/~-t-t-+-t-t--+-t-t-t-t-,--t-t-+-+-t-t-t--H-t-+-+-l-t-+\n0 .-~ lt,_ltJ-il-r-Ti-r-r-ir-r-rt-t-i-t-t-t-ttttt-i-,f'-j-i-j-j--t--t-+-t-t-+-t-t-t-t-+-t-t-+-1-t-+-1H-~-f-+~~' I I QJ L I ,c::) I I L I J ~ I L Q) Ill ~ I L.. J ~ -n--r-t-r-t""M-nr-r-r-r-n-rr"HJ"I -rrI 'iH--t-t--t-t-t-t-t-t-H~r-t-H-H-+-+-++++++~-+-+-H-o'--'I , ' "t-H-4t-t-r-t-t-T-t-+-t-t-+-t-t--H'ftiL 1't-t-t+-II t-t-+~-t-t~-+-t-+-+-tH-+-t-+-t-t-~-+-+-+-t-t-t-+-+-t-++-!H- ~ ·~ l'+t'r-t-t-r1-t-t-t-t-t-t-+-t-t-+-+-1-+-+-t-+-H1-t-+-!-++-1-++-1-+-+-t -t-t-t-t-t-t--t-+-t-t-i-t-+--t-11-t-~'~f-l~-J-r-t-,-t-+-t-t-t-t-+-t-+--t-iH--i-'.l-++-t-+-!-+-t-+-+-t-t-+-f-+--l-1f-J J .L I I I I J J I I v -~6 2- i "'~ .c:: ,_ - - r.o ...... o..-.. __ ._ ,_,_._ -~ !ZJ -r1· · '-'- ._M,_ -: - 1--··· Q) .c:: •rl .c:: I/ --''- -o Q) .d .. 0 +' 0 I/ ,__,__ 0 +' .£! +:> 4 ori m •rl ,_h =t- ~~= __ L I s, +' ctl +' ...-; 0 +' N +' '-'- -f-- +' P-t .£! ;;: en w crl ...-; (.1j -+ ~+~ C) Q) (\J N N +' +' r.o .c:: >- '--'- 1-1-· -, a> m .c.. ori en -o ::r:: orl ..-t •rl ' L,_ J± •ri Q) 'O (}) 0 s:: +' s:: 'I I I Q) .c::-;:;; .c:: Q) 'O HZ ori ctl ori - -I ·-·- Q) 0 crl S... 0 +' Lf\ +' rl +' ·-- --- I S... Wrl ::s s, 0 m o, ctl I j -- 0 0 :>-. --ru +' Q) I r-1 r-1 I ,_ -· ..: 0 1\-4 >< +' Q) Q) Ori o, 0) p, Y) om 0 0 Q) 0) Q) (I} I r.o .£! >=! 0 rl +' p. orl P~ I I Q) .c:: +' crl H Q) +' :>-. (I) ~ Q) I ~·-·- I> 0 s p.. Cl) r-1 +' p,+:J I i:... tO s... 'O 0) 0 [I) Q) 0) I ::s 0\ s:: 0 >=! S... +> S... +' ,_,_ ...... _ I ,_ .... I 0 " •rl 1\-4 0 m +.:> . r-1 Cl) r-1 I/ ' II . --r ~ .... ··-1'0 I 0 p.. I/ I Q) 'O 0 Q) +' s:: Q) td ...... I I 0 s:: -~,..., P. crl ctl r.o rl w +> +' +' ~· Ir s:: m 0) N +' ·(!) r~ P-t P-t P-t <;; ro S....ori 0) 'L1 Q) ''" s .. (!) Q) >=! 0 • +' '\';-~~~ I ,_ Ha) s... +' ....-! I ..("'...(\) :>-. 0\\0 \() .._,_ 0 r-1 ::s +:> +' +' Sri I 1/ ,I/ 1\-4 +' GJ m CJ) m 0 0 --- I I S... .. cd..Orl (!) O'-.,. S... lf\a) 0\ (!) Lf\ s... P..'O +> +' r-1 r-1 r-1 J I P-t r-1 0) 0) 0 • P-t 0 '-'--· I P< a> (l) ~ Q) Q) Q) Q) ~I- s I> W I I I I 7 1 ·- J I I I -j=- I I I II I I IJ 11 ,, I I J , 1I J I 0 I~ 0 I I/ I I - I •\ .J , " I I 1 , I I I J I I J ->-- I J I I ,_,_._ J I I , , .., . - I. -.f-·- I/ I J IJ H-: I I 1 Ii I/ I ! i , ' II I ·, ""' , U; ,'iV / '~"':;.. "- ~ 0 ...... - -63- on hydrogen and oxygen are compared with those of electrode 15 in figures 24 and 25 respectively. Electrode 18 can be seen to be superior to# 19 which in turn is superior to# 15. These results are in agreement with the surface areas of the three catalysts incorporated into these electrodes, the higher the platinum surface area, the better the performance obtained. \ ,, . -64- FART.,. II In order to study the relationship between platinum surface area and fuel cell performance, it is necessary th have electrodes which are prepared similarly, and which contain catalysts th~t are similar but of different surface· area. Ideally, if performance is· directly , I proportional to surface area (only the surface of the platinum serves as catalytic' material with the 1.nterior bulk contributing. nothing) the, number of m1111amps of current obtained per unit platinum area at a given voltage should be a constant. ,The above hypothesis was tested as follows: In the course of preparation of catalyst J6, samples were rerr.oved at the end of each cycle. These catalysts were·labeled J1, J2, J3, J4, and J5,respeotively. These catalysts contained 1, 2, 3, 4~ and 5 % platinum, and had surface areas of 46, 42p 35,·32, and 28 m2 pe~ gram respectively, compared with 23 m2 per gram for cataly~t J6 which contained 6% platinum. Electron micrographs of these catalysts were unrevealing and are therefore not presented. It can.be said that all five had small platinum particle sizes that appeared to be well adhered to the asbestos support. Table 3 summarizes the preparation and properties of these catalysts. Electrodes 20~ 21, and 22 were prepared incorporating I,' -65- TABLK'3 Catalyst Method of % Pt Specific Pt Preparation Surface Area in m2/gram J• 1 cycle hydrogen 1.0 46 1 reduction at 100°c J2 2 cycle hydrogen 2.0 42 reduction at 100°c . J3 3 cycle hydrogen 3.0 37 reduction at 100°c J4 ·4 cycle hydrogen 4.o 32 reduction at 100°0· J 5 cycle hydrogen 5.0 28 5 reduction at 100°c ' ·J6 6 cycle hydrogen 6.o 23 reduction at 100°0 Table 3. Properties of catalysts prepared through stepwise platinization • . \ --, I ,• -66- ~1,.J2, and J4, with each containing 0.30 g of platinized asbestos. Because of the different platinum content of the· catalysts, these electrodes contained 0.17, 0.33, and 0.66 mg Pt per cm2 respectively, compared with 1.0 mg Pt per cm2 · I for electrode 18 with which these electrodes were compared. All four electrodes were otherwise similar. (See table 4 for a summary of the fabrication and content of these electrodes). Figure 26 shows IR free performances of electrodes 18, 20, 21, and 22 on oxygen. These data were obtained via the direct D.a. discharge method described in the experimental section. IR free data are desirable here .because in these experiments it is the performance of the platinum that is being scrutinized. Internal resistances of the electrodes are different and these differences .would effect the apparent performance~ of the cat~lysts if IR included data were used. ·Also, IR free data may be compared with dat~,taken in other cells with different electrolytes, o~ different electrolyte gaps etc. ) Figure 26 shows the expected result, namely, that performances are related to the amounts of platinum in the electrodes. The relationship of platinum surface area to performance may be seen as follows: The current densities of the four electrodes at 0.7 volt are obtained from figure 26, and the total platinum surface area per -67- TABLE 4 Electrode Catalyst Amount of Diluent Amount of Number Catalyst Diluent (in' grams) (in grams) 18 J6 0.30 1.0 Graphite 0.3 20 Jl 0.30 o.66 Graphite . 0.3 21J J2 0$30 0.33 Graphite 0.3 ' 22 J4 0~30 0.17 Graphite 0.3 Table 4. Summary of content of electrodes 18, 20, 21, and 22. All electrode~ contained ~.12 cc T-30 Teflon binder and were pressed at 2230 lbs. I?er square inch. t \ \ ,• -±J--1--- ..... : I~ -~H-f:H-H-- ;++- j' I _,_.__! I I 1-H- --+--1-1-,_LH I I -+·- - - -~ ~j- - .!_t- l- ·-L - -<--+-+--+--+--!·-<- _...._ --1_ ,+-,-+-+-!-<- ..... ,_ --l-+-+-+-+-!-+-+-!-1--t-;-r-J = _ ..... _J_ ·.1 . 0 _,_,_ - :•++-+-+-t-+-t-·-l-j...... 1 .....1-i•-1 ·- -l•--'·-1-l-+-l-l·-+-lt ~ --1--- _I_>- -1--1--1-+-+-l-++-l-+ L>--+ +t+ -LC - I L - _..._,__ ~~-+--H-+-i-'-+-+-+-+--H l-+-H-t--!-1-+-+-+-+' -+-+ - ~-L--:-- -'-+-1-i-+-!1-1--+-i-+-+-f-+-r-+-•-+-+-t--t-+-t-t-+-+ . . . . -+---l-H+B-H·--+-+-+-+-+-t-+--"-J-t--HH--t-!-H_J._._,.... - rl 0 0 0 :- ~ I (X) 0 rl C\J rl C\J (\j (\) ,__.__,__ -1--!-+-+-+-l-+--HH-+-il-+-+--t-+-!-H·-t-+-i--t-+-t-! (1) Cl) (1) (1) -o 'CPc:J 'd 0 0 0 0 -o H H H H C\J +> +> +> +> ..--.. 0 0 0 0 (!) (!) (!) (!) 1-1-..w.,_:~~w·-+- I/ r.;. I I/ I ! I I/ I j v I, / I/ ' I ,._ If I/ :11/ 17 ,. -69- TABlE 5 Electrode mg Pt per Specific Total current G'Urrent Number and electrode Pt Surfac~ Pt A~ea Density Density Catalyst Area in m in m at 0.7 v per Unit per gram. IR Free Area at in ma/cm2' 0.7 v IR free,2in2 ma/cm /m 20 ·- Jl 3.0 46 0.139 0.0079 0.057 21 - J2 6.0 42 0.252 '0.0149 0.059 22 - J4 12.0 32 0.392 0.0228 0.058 18 - J6 18.0 23 o.414 0.0250 0.061 Table 5. Table showing data for comparison of current density obtained per unit area of platinum. ... , \ .. -70- ·1s-~btained by multiplying the specific surface area by the weight of platinum in the electrode. Dividing the current densities at 0.7 volt by the total platinum area gives the current density per unit area at constant voltage. These calculations are summarized in table 5. As can be seen from the data in table 5, there is an excellant correlation between surface area and performance. All four electrodes, each with a different platinum content, gave between 0.057 and 0.061 milliampe per cm2 per m2 of platinum area .a t o. 7 volt .• \ ' ·' -71- CONCLUSIONS'~ The objectives of the work described in this thesis were to find the best method of preparing high surface area platinized asbestos, find an optimun electrode composition and structure for evaluating platinized asbestos catalysts, and correlate specific surface area of platinum with fuel cell performance. These objectives have been attained and the.conclusion~ reached are here presented. ELECTRODE STRUCTURE··- AND, COMPO SIT ION The. structure and composition of platinized asbestos containing electrodes were optimized. The optimum electrode s·tructure is· a modefied Niedrach-Alford electrode. Of the diluenta tested, graphite is the beet, and 0.3 g is the optimum quantity of graphite. Boron carbide is a good diluent, and the two carbon blacks tested are fair, while tantalum powder is a poor dilueht. 0.12 cc of T-30 Teflon binder holds electrodes tightly together and while half this amount is adequate·for test purposes, the lesser amount of binder does not imp~ove performance. 2230 lbs per square inch is the minimum pressing pressure that gives good electrode ' structure. Lower pressure slightly improves performance, but electrodes prepared with leas pressure leak and delaminate. ' .. -72- CATALYST PREPARATION Platinized asbestos prepared via reduction of chloroplatin1c acid with hydrazine, formaldehyde, and sodium borohydr1de 81ve catalysts in which there is little adhesion of platinum to the support. The wet hydrogen reduction gives excellant adhesion of platinum to asbestos, but platinum preferentially deposits on platinum rather than on asbestos' .. giving rise to larger platinum particle size and lower specific surface area. Direct hydrogen reduction at 100°c gives the best catalysts. Catalysts prepared· via this procedure give the best performances and highest surface areas. The amount of platinum in the catalyst also has a bearing on the surface area. The greater the amount of platinum deposited on asbestos, the' lower the specific surface area of the platinum. Aleo, stepwise platinization gives a higher platinum specific surface area than one step plat1nizat1on. \ RELATION OF FUEL CELL PERFORMANCE T.0 SURFACE AREA OF PLATINUM Electrocatalytic activity of platinized asbestos as expressed by fuel cell performance is quantitatively related to specific platinum surface area. The 6urrent density obtained per unit area of platinum at a given IR free. voltage is essentially constant even though the tot.al -73- amount of platinum may vary widely in the electrodes tested. This shows that the interior bulk of the platinum plays no electrocatalytio role, only the surface of the catalyst is active. • \ \ .... ~ -74- BIBLIOGRAPHY 1. Grubb, W. T., and Niedrach,· L. w., Advanced Energy Conversion Systems, Sutton, G. w., ed., McGraw Hill, New York, in press, chapter on fuel cells. 2. Liebhafsky, H. A., and Cairns, E. J., Fuel Cells and Fuel Batteries, J. Wiley and Sona Inc., New York, to be published. 3~ Young, G. J., ed., Fuel Cells, Reinhold Publishing Corp., New York, 1960. 4. Mitchell, W. Jr., ed., Fuel Cells, Academic Pre~a, New York, 1963. 5. Sherman, A., and Eyring, H., J. Am. Chem. Soc., ~.· 2661 (1932). 6. Young, G. J., and Rozelle, R. B., Fuel Cells, Young, G. J., ed., Reinhold Publishing Corp., New York, 1960, chapter 3. 7. Parsons, R., Trans~ Faraday Soc.,~' 1053 (1958). 8. Ref. ( 6) , p , 30. 9. Berl, w. G., Trans. Electrochem. Soc., §2., 253 (1943). 10. Berkman, s., Morrell, J. c., and Egloff, G., Catalysis - Inorganic and Organic, Reinhold Publishing. Corp., New York, 1940, chapter 7. 11. Innes, w. B~, Catalysis, Vol. I, Emmett, P.H., ed., Reinhold Publishing Corp., New York,1954, chapter.6. 12. Niedrach, L. W., Toohner, M. , and Zeliger, H. I. , to be published. 13. Willstatter, R., and Waldschmidt-Leitz, E., Ber., ~' 122 (1921). 14. Paal, c., Ber., "21_, 1217 (1904). 15. Brown, H. c., and Brown, c. A., J. Am. Chem. Soc., 84, 1493 (1962). 16. Newkirk, A. E., personal communication. ------·--- -75- 17. Zel1nsk11, N. D., and Borissow, P. P., Ber., 2,l, 105 (1924). . - 18. Richards, D. A., Phil. Mag., 16, 778 (1933) • . 19. Koch, E. F., personal communication. 20. McKee, D. w., personal communication. 21. Bond, G. c., Catalysis by Metals, Academic Press", London and New York, 1962, chapter 8. 22. Niedrach, L. w., and Alford, H. R., J. Electrochem. Soc., 112, 117 (1965). 23. Kordesch, K., and Marko, A., J. Electrochem. S'oc., 107, 480 (1960). . 24. Niedrac.h,L. w., and Tochner, M., to be published.