Method of Oxygen Electrode Preparation and Product Thereof

Gas electrodes are well known in and for electrochemical operations. For their use, a gas is passed in contact with the electrode in the presence of a cooperating electrolyte. Most modern gas electrodes are highly porous and, either inherently or by purposive implementation, catalytically active. Quite importantly, the interior interstitial passageway wall surfaces within the electrode body should have good catalytic capability. Maximization of effective, available reaction-promoting surface area for given unit geometric volumes of the electrode configuration may thus be realized.

The indicated manner of electrode construction is advantageous for oxygen gas-bearing (or depolarized) electrodes, particularly cathodes, that are adapted for the electroreduction of oxygen in alkaline media.

Usage of such electrodes involves passage of the oxygen-bearing gas through the porous electrode body for contact with the involved electrolyte. Such practice is very desirable not only for electrolyzing functions but also for galvanic mode operations, as in fuel cells. Oxygen gas-bearing depolarized cathodes are particularly attractive for chlor-alkali and the like manufacturing cell operations.

A very good reason for employing oxygen gas-bearing porous electrodes to electrolyze common salt brine into chlorine and caustic, and for analogous production purposes, is economic. Significant savings in electrochemical power requirements for given workings are anticipatable due to substantial reductions achievable in needs for applied energy consumption when such electrodes are utilized.

Oxygen electrodes have often been catalyzed by various precious or semi-precious metals and compounds thereof. But, such noble metals as gold, osmium, palladium, platinum, silver, their alloys, oxides and other compositions and so forth are extremely expensive, especially for industrial applications. Thus, their consumption as catalysts for electrode preparation must be carefully controlled. Minimization of total quantity usage is usually done by deposition of precious catalyst materials in the form of thin plating or other layers over a suitable substrate, such as a porous nickel plaque.

It would obviously be desirable to have for ready usage an efficacious yet less costly way to provide satisfactory catalytically-active porous electrodes, especially for use in alkaline media as depolarized cathodes for electrolysis (or even for distinctive galvanic mode purposes).

Manganese oxide materials (including mixtures thereof) are known to be excellent electrocatalysts. These, in particular, include: various forms of manganese dioxide (MnO.sub.2); manganic or manganous oxide (Mn.sub.3 O.sub.4 --also known as "hausmannite"); and manganic sesquioxide (Mn.sub.2 O.sub.3). On reasonable evidence, it is fairly supposed that the Mn.sub.3 O.sub.4 form is likely the most chemically stable, especially in association with nickel, of the various indicated forms.

Many patent and literature references relate to the preparation and use, in diverse ways, forms and/or combinations, of electrodes involving manganese oxides or manganese peroxides (oftentimes in combination with other electrochemically active materials). These include: U.S. Pat. Nos. 1,296,188; 1,043,937; 1,143,828; 1,423,071; a,510,172; 3,491,014; 3,535,217; 3,616,302; 3,627,699; 3,775,284; 3,915,837; and 3,948,684; Chemical Abstracts 7:3458; 42:5356c; 50:83b; 74:150359r; 76:62793e; 78:51823p; 79:142363a; 81:57517a; 81:20121r; 83:182053p; and 83:182059v; and Derwent 61865 S/39 and 10379 W/06 (all of which are herein incorporated by reference).

Despite the known art, reliable ways and means to effectively incorporate securely placed deposits of manganese oxide(s) in and for catalyzation of porous electrodes, and the associated desiderata, have heretofore been lacking and wanting; this being especially so in respect of very fine pore, large internal surface area electrodes of newer, modern type and style which are extremely difficult to efficiently internally catalyze when they are not inherently catalytic in nature.

This invention concerns an improved and efficient procedure for catalytically-activating porous electrodes with effective internal placements of manganese oxide deposits (to particularly nicely suit them for electroreduction of oxygen in alkaline media) and the resulting catalyzed electrode products, the technique for which is neither expensive nor complicated and, with high levels of confidence thereabout, gives good and reliable results in practice; the furnishment of same to readily and easily get electrocatalytically-active electrodes characterizably by their having good operability in and good various electrochemical applications being amongst the principal aims and objectives of the present invention.

The instant contribution to the art is directed to an efficient method of providing an electrocatalytically-active manganese oxide deposit on the interior pore wall surfaces of a porous electrode body comprising: applying a liquid reagent, preferably a dispersion and/or solution o a stable, pyrolyzable (or, with equivalent meaning, "pyrolytic"), inorganic permanganate salt to and on the interior pore walls of the electrode body; then heating the reagent-bearing electrode in a gaseous oxidizing atmosphere at a body temperature between one that is at least at about the decomposition temperature of the reagent and about 700.degree. C. until the reagent decomposes to provide layer deposit formations of desirable manganese oxide catalyst materials on said pore walls.

The method step sequence may be repeated until a necessary or desired loading level of laid catalyst deposit is attained. Most advantageously, the method of the invention is operated with an additional treatment of a refining nature by a peroxide rinse or bath which is done at least as the final step in the procedure.

The invention also pertains to the advantageous manganese oxide(s) catalyzed porous electrode articles of manufacture obtained as products of the delineated method.

The working proportional details and other significant and more exact specifics of the invention are also ingeniously set forth in the ensuing Specification and description.

Further features and details hereof are brought forth in the ensuing Specification and description, taken in conjunction with the accompanying drawing, wherein as therein appears all of the several involved Figures are either presentations of data in graphical form or fanciful or actual reproductions of microphotographs taken of actual catalyzed electrode structures.

Particularly good and securely fixed placements of desirable manganese oxide forms of deposits for enhancing the electrocatalytic capability of porous electrode bodies (especially those of nickel) are achieved by the following of the indicated practice according to the present invention by: making a good and thoroughly wetting or interior surface covering application of a liquid reagent, for example a dispersion and/or solution, of a pyrolysable (or, "pyrolytic") permanganate salt or mixture of such salts (which, preferably, is water-soluble and applied from aqueous solution) to and upon the interior wall surfaces of the porous electrode body structure to be catalyzed; then, in a controlled gaseous oxidizing atmosphere, subjecting the reagent-laden structure to heat at an elevated temperature not in excess of about 700.degree. C. but at least at or above the decomposition temperature of the involved pemanganate salt(s) so as to cause deposit formation of electrocatalytically-active manganese oxide(s) on the interior wall surfaces of the structure being catalytically beneficiated. Very beneficially, the deposited manganese oxide(s) formation is subsequently reduced with, preferably, a chemically-active peroxide rinse. Then, optionally and/or if necessary to get adequate or desired catalyst deposit loading in the structure, the procedure may be once or plurally repeated.

The conditions of heat in which the reagent pyrolytic permanganate salt is converted in an oxidizing atmosphere upon and within the electrode structure to a catalytic manganese oxide deposit depends on the particular thermal decomposition point of the involved permanganate salt. This, in the case of the advantageously utilized potassium permanganate (KMnO.sub.4), is at least about 250.degree. C. up to the indicated 700.degree. C. maximum. More often and especially with the natural variations in the decomposition points of the involved permanganate salts, a temperature between about 400.degree. C. and about 600.degree. C. is utilized. Usually, higher heating temperatures tend to result in greater proportions of the desirable Mn.sub.3 0.sub.4 in the manganese oxide(s) deposition made. In all cases, the designated temperature is that of the actual electrode body being treated.

Heating of the reagent-carrying substrate may be done by any desired and appropriately effective means. It may thus be in an air or other oxidizing atmosphere over a flame, such as a gas flame propogated by appropriate burning apparatus. Alternatively, especially when larger production manufacturing operations are called for, directly- or indirectly-fired ovens or the like fueled by gas, electricity or other suitable heating means may be employed for the heating. The desired oxidizing atmosphere is, of course, directly controllable during the heating and permanganate reagent decomposition-conversion step. An oxidizing atmosphere is readily achieved by keeping the reagent-treated electrode in an ambience of air, oxygen-enriched air or oxygen over the flame or other heating source or means, as may be created in a heating oven.

The heat-applying treatment is continued until thorough decomposition of the applied permanganate salt is achieved. For most normally sized electrode bodies (depending precise temperature involved), this is most frequently within a time period of 5-10 to 20-30 minutes. Longer heating times can be employed, but are generally not necessary to achieve the desired decmposition of the permanganate.

The manganese oxide phases obtained in the deposited catalyst formations, notwithstanding the varying oxidative states of the therein included manganese, are generally found to be distributed throughout the porous electrode body rather than being confined to or concentrated on its external surfaces.

After each deposit-forming heat treatment, the electrode is carefully cooled. After each cooling it is, advantageously, thoroughly leached or washed with water (best hot) for general cleaning purposes. This directly removes easily soluble and loose materials remaining after the catalyst deposition procedure. It may then be dried and used as an electrode. Sometimes, however, intermediate water washing steps may be eliminated, especially when the treated piece is subjected to a subsequent, finishing peroxide treatment.

Preferably and to generally secure utmost benefit from practice of the invention, the treated piece being fabricated (even in water-washed condition) is finally subjected to the reducing (with for example peroxide) post-treatment. This vastly beneficiates the applied deposits of manganese oxide(s) in at least two ways. For one and by the therein involved physical rinsing and scrubbing action, it tends to positively leave a more solid and substantial and better-adhering layer of the catalyst on the interior pore surfaces of the electrode. For another, it tends to chemically reduce the deposited oxide(s) (which although somewhat arcane in nature are almost invariably composed in varying respective proportions of: usually the .alpha.-, but sometimes .delta.-form of Mn0.sub.2 ; Mn.sub.2 O.sub.3 ; and Mn.sub.3 O.sub.4) to desirably greater occurences of the Mn.sub.2 O.sub.3 and Mn.sub.3 O.sub.4 forms.

The peroxide treatment is simply done by rinsing, immersing (or otherwise spraying or passing) the deposit-laden electrode in a suitable peroxide solution. Preferably, aqueous hydrogen peroxide is employed for the purpose and is used at practically any desired concentration (although 10 or so weight percent aqueous H.sub.2 O.sub.2 is usually convenient). The peroxide treatment is quite exothermic and substantially heat generating. It may thus be ordinarily commenced with the peroxide bath at normal room or other handling temperature. Contact of the manganese oxide(s) deposit bearing electrode with the peroxide solution generally gives rise to quite vigorous physical bubbling and scrubbing action. This, as indicated, serves to remove during the peroxide treatment a great deal--if not all--of the deposited catalyst material which may have been only loosely laid over and not tightly adherent on the electrode after deposit in the permanganate decomposing heating step. Beneficially, this also tends to minimize pore-plugging propensities in the treated electrode.

After final peroxide treatment (again taking into account individual procedure repetitions to secure desired level of quantity placement of the deposited catalyst), the electrode is again washed and dried. It is then ready for electrochemical usage.

Along this line and if desired, the peroxide treatment may be applied after each reagent application and permanganate decomposing heat treatment; even though it frequently suffices for the treatment to be done only once and at the conclusion of the entire procedure.

It warrants repetition that the peroxide treatment is instrumental in enriching the proportion of Mn.sub.2 O.sub.3 and even Mn.sub.3 O.sub.4 in the manganese oxide(s) electrode catalyst deposits formed in practice of the present invention. While available evidence displays the activity as oxygen (and other) catalysts of the specific MnO.sub.2, Mn.sub.2 O.sub.3 and Mn.sub.3 O.sub.4 oxide forms of manganese, the last-mentioned form seems, as has been noted, to lend optimum chemical stability in and with the formed catalyst. The particular conditions of permanganate-decomposing, oxidizing heat treatment of the applied pyrolizable permanganate salt plus the subsequent peroxide treatment allows relatively specifically different proportions of the three desirable oxides to be present in the resulting deposition. It is, by the way, exceptionally difficult to accurately quantify the relative proportion of individual manganese oxides in a given mixed deposit thereof; this being all due to the different, yet closely similar, specific compositional forms and altering crystal variations in which they usually occur.

Peculiar as it may seem, analyses made after electrodes have been used in alkaline media in cells show that the oxide remaining in preponderant proportion is Mn.sub.3 O.sub.4. This is borne out in such authoritative literature sources as "Electrochim. Acta", 20, 221-225 (1975) and the 1966 Edition of the "Atlas Of Electrochemical Equilibria In Aqueous Solution". These sources indicate that when a nickel electrode substrate is involved the interaction of the catalyst with the nickel is responsible for the Mn.sub.3 O.sub.4 build-up. The mechanism believed responsible for this is the reduction by nickel of MnO.sub.2 (and Mn.sub.2 O.sub.3) to Mn.sub.3 O.sub.4 causing oxidation and resulting in formation of nickel hydroxide (Ni(OH).sub.2).

In this connection, microscopic examination of utilized electrodes shows that many manganese oxide catalyzed structures which initially perform quite well (e.g., with voltage savings in a chlor-alkali cell as great as those on the order of 0.7-0.9 volt against the theoretically +1.229 volts savings difference for depolarized electrodes, with which E.degree. is +0.40l volt, as compared to the -0.828 volt E.degree. value with traditionally common chlor-alkali cells) and then unexpectedly fail develop pore-blocking particle growths within the structure. Analysis of the growth, as by X-ray diffraction, identifies the largest proportion of the material as Ni (OH).sub.2.

This effect is well illustrated in fanciful representation in FIGS. 1 and 2 of the accompanying Drawing. These, more or less illustrate schematically "before" and "after" illustrations of the potentially blocking action of Ni(OH).sub.2 formation on a metallic nickel substrate as a result of the effect of a caustic (i.e., NaOH) solution on the metallic nickel. In FIGS. 1 and 2, the nickel electrode body is generally represented by reference numeral 10. The porous nickel body, per se, 11 has an idealized pore there-through represented generally by reference numeral 12. In the "before" representation of the unused electrode of FIG. 1, the manganese oxide catalyst deposit 13 is illustrated as the compact, initially-formed placements made by practice of the present invention. In FIG. 2, the more or less flocculated hydroxide (i.e., Ni(OH).sub.2) formation growth is represented by number 14.

Ni(OH).sub.2 can also be formed simply by the action of a caustic (NaOH) solution on metallic nickel. Such hydroxide formation growth is known to be a reason for failure, due to pore blockage, in silver-catalyzed nickel electrodes. The formation of Ni(OH).sub.2 through reaction with manganese oxdes, of course, aggravates the pore-blocking possibility.

The same effect is photographically represented by and easily observable in the "before" and "after" micro photographic pictures in FIGS. 3 and 4 of the accompanying Drawing.

Similar and analogous phenomena also tend to occur when the electrode is composed of other useful and more or less popular metal substrates in place of nickel, such as iron and cobalt. Hence the desirability of having as much of the Mn.sub.3 O.sub.4 form as possible in the permanganate-decomposing, catalyst-forming heating step.

Advantageously, the pyrolytic permanganate salt reagent employed in practice of the present invention is potassium permanganate, KMnO.sub.4. Nonetheless, other stable, advantageously water-soluble permanganates can also be utilized. These include: other alkali metal permanganates, such as lithium permanganate, LiMnO.sub.4, and sodium permanganate, NaMnO.sub.4 ; as well as such other permanganate salt materials as those of calcium, Ca(MnO.sub.4).sub.2, barium Ba(MnO.sub.4).sub.2, silver, AgMnO.sub.4 ; and so forth.

The reagent is generally best utilized in as great a concentration as possible. Saturated solutions (or relatively highly concentrated dispersions) thereof are therefore de sirable. In this, there is not much point in using weaker strength materials (even though they can be employed). Low strength reagents unavoidably become rapidly concentrated and dried out during the electrode heating in which the deposit is formed. In the cases when suspensions are utilized, solids concentrations of between about 0.015 and 0.2 mg/cc of suspensoid in the involved liquid are found suitable for the purpose.

The wetting or other surface covering application of the permanganate salt reagent may be done in any effective way. This includes immersion, spraying or misting application, and so forth. It is generally desirable for the pores of the electrode body to be as full as possible or substantially saturated with the liquid-dispersed reagent before commencing the heat treatment.

The relative quantity of manganese oxide(s) deposit ultimately provided in the electrode body can vary widely in accordance with various factors involved. These include: the activity capability of the particular forms of manganese oxide produced in the laid deposit; the number of cycle repetitions, if any, undertaken; and the pore size(s) and actual character of the electrode body being catalyzed Obviously, deposit loading should not be so heavy as to reach a point beyond which electrode performance may be diminished by pore blockage and consequent prevention of fluid movement into or through the pores. Also, it is not always necessary or possible to effect continuous deposition formations throughout the pore wall surfaces being catalyzed; it oftentimes being sufficient to have spotty or patched areas of deposit or preferential (or unavoidable) situs areas on which the deposit(s) form(s) without consequent disadvantage or impediment.

Along the line last above mentioned, it is frequently noted that maximum catalyst deposit formations appear to be laid and remain in dual porosity electrode structures at or near the neck-down or bore size change points where the nominal pore dimensions change between associated wall layers respectively containing the differently sized fine and coarse pored interstitial passageways.

Without inflexible limitation to the recited parameters, it is generally desirable for the catalyst deposit load level that is laid to be at least about 0.005 milligram of catalyst per cubic centimeter of porous electrode body volume. Maximum deposit loadings (almost unavoidably necessitating a considerable number of repetitive treatments in the operation) may be, on the weight of deposited manganese oxide(s) formation per unit volume of electrode body basis, as high as about 7.5 mg/cc. For most situations, however, suitable catalyst loadings are in a range between about 1 and about 2.5 to as much as 4.0 mg/cc.

The base material for the porous electrode bodies is almost invariably metallic. Metals such as nickel and its alloys (including the well known Monel metals and "Inconel"), tantalum or titanium, copper, nickel plated copper, and various ferrous alloys (particularly the so-called stainless or corrosion-resisting steels), are adapted to employment. Nickel is an oftentimes desirable material of construction. The electrode body material should be at least substantially, if not completely, resistant to chemical attack--at least during cell operation--by the contacting electrolyte material that is utilized and oxygen when the electrode is employed as a depolarized cathode.

Appropriate pore size in the porous electrode bodies to be catalyzed depends, as will be appreciated by those skilled in the art, upon the particular use application to be made thereof. The pores may sometimes be as large as 25-50 microns in diameter. Usually, however, those of nominal pore size diameters in the range of from about 0.1 and about 12 or so microns are adaptable to a wide variety of electrochemical utilizations. It is often more advantageous for this range to be between about 1 and 11 mircrons; it sometimes being preferable for it to fall in an approximate 3-8 micron nominal diameter span.

Correlative to pore size and multiplicity in any given electrode is its porosity factor. This is usually taken as the percentage by volume of void space in any given unit volume of electrode body. The porosity, as is evident, is the difference upon subtraction from 100 volumetric percent of the percentage of the actual (or apparent) density of any specific electrode mass from the theoretical solid density of the material of construction of the electrode. Porous electrode bodies adapted to be catalyzed by practice of the present invention can ordinarily range in porosity at values of from about 30 to about 90 percent. More commonly, the porosity is on an order between about 60 and about 85 percent.

It is also usually beneficial for the degree of porosity built into the electrode body to be such that the total hole or open area on any given section of exposed surface of the structure provides an adequate proportion of the total area of that exposed surface to ensure satisfactory mass transfer capabilities and effects for fluid ingress and egress relative to the electrode body.

Literally any desired configuration and/or body shape can be present in the porous electrodes catalyzed by means of the present invention. Wall section face area size can be as desired; and thickness can be suitable for given intended usages. It is a frequent practice to make the electrode body in a generally thin and flat rectangular or circular plaque form; although various composite structures (such as dual porosity arrangements) and other built-up forms may also be provided. Body thickness in such constructions may be varied in the overall between about 5 and 500 or more mils (ca. 0.13 millimeters and 1.3 or so centimeters). More often it is in the range, frequently occurring as a matter of choice or design, between about 10 and about 250 mils (ca. between about 0.25 millimeters and 0.6 centimeter). In many cases, an appropriate body thickness is found to be between about 15 and about 100 mils (ca. 0.38 and 2.54 millimiters).

To demonstrate the advantageous practice of the present invention, a number of experiments were performed. The testings done and results obtained were as in the following.

In all cases, the electrode bodies were commercially obtained and made of sintered nickel electrode stock. They were in the form of flat, disc-like bodies, each of which had a 21/2 inch (ca. 6.35 cm) diameter. The average thickness was 70 mils (ca. 0.178 cm). Unless otherwise specified, each of the electrode bodies was of the dual porosity variety, with a fine pore layer of about 20 mils (ca. 0.051 cm) of the total thickness in which the pores had an average nominal pore diameter size of about 2 microns and a 50 mil (ca. 0.126 cm) coarse pore layer with about 10 micron nominal pore diameters.

In the experimentations run to demonstrate manganese oxide catalyzed electrode efficacy in usage, the catalyzed electrode products were tested in a standard experimental cell. For this, each specimen was separately used as an electrode in which it was mounted for evaluation as a depolarized cathode in the test cell which had expanded titanium mesh anode coated with an oxide of ruthenium and titanium. Anode-to-cathode locations had an intermediate "Nafion" (DuPONT TM) ion exchange membrane separator in the cell. The aqueous anolyte contained 300 g/l NaCl and the catholyte was an aqueous solution containing 100 g/l NaOH; with the cell operated at a temperature of about 70.degree.C. and positive gas pressure on the back side of the cathode. The applied current density was 0.5 amp/in.sup.2.

Such testing allows determination upon performance of the measure over increasing time periods of voltage savings realized in the comparisons between operation with both nitrogen and oxygen gases applied to the electrode. The difference--by subtraction--between voltage values obtained from nitrogen (i.e., inert gas) operation at any given point of measure and those from oxygen (i.e., active gas) operation at the same point of measure provides reliable indication of voltage savings obtainable with and corresponding depolarization effect experienced upon use of the electrode as a cathode.

More particularly, the below-described Third through Eighth Examples report the results of the several involved cell testings.

To illustrate typically attainable catalyst loading levels, different electrode specimens were, for each single cycle of operation, first dipped until thoroughly wet in an aqueous solution of KMnO.sub.4 at 0.40 Molar (saturated) concentration; then held at about 250.degree.C. until decomposition of the permanganate reagent was completed in an air or oxygen oxidizing atmosphere; then washed in water. The cycle was repeated several times with each electrode body tested. One of the electrodes involved was reinforced with an "Exmet" expanded nickel lath screen. The loading achieved (measured in milligrams per square centimeter of one side face area of the electrode) after each of the consecutive catalyst-depositing cycles conducted are as set forth in FIGS. 5, 6 and 7 of the accompanying Drawing, which also contain other pertinent data relevant to the experiments.

To demonstrate the tenacious and tightly adhering character of the manganese oxide(s) catalyst depositions prepared according to practice of the present invention, a number of catalyzed electrodes were prepared and run as cathodes in cells of the above-described test style. The cell effluents were collected for one (1) week and analyzed in comparison with the same test conducted on effluent from a conventional, chlor-alkali operating cell. The following tabulation identifies the various cathodes tested, how they were prepared and the amount of manganese and nickel seen in the material filtered out of the effluent (i.e., "Solids") and in the filtered solution (i.e., "Solution").

The electrode disc substrate used had a single 10 micron nominal diameter pore size throughout.

The catalyst was deposited by dipping once in an aqueous solution containing 0.45 g/l KMnO.sub.4 followed by heating at 125.degree. C. After this, the cathode was water washed, then washed with 1 percent NH.sub.4 OH and redried at 125.degree. C. The final loading was 3.4 mg/cm.sup.2.

The performance of this cathode was relatively stable both on air and on oxygen over 21 days. The depolarization averaged 0.53 V on oxygen and 0.40 V on air.

The procedure of the Fourth Example was essentially duplicated excepting to terminate it with a 30 minute soak of the catalyzed electrode in a 10% H.sub.2 O.sub.2 aqueous solution. The cell test results are graphically depicted in FIG. 8 of the accompanying Drawing.

The catalyst was deposited by dipping in an aqueous solution saturated with KMnO.sub.4, heating at 250.degree. in air and leaching in boiling water. This cycle was done four times. The final loading was 17.6 mg/cm.sup.2. This was Cathode "A" in the Table of the Second Example.

An X-ray analysis of the catalyzed electrode showed .alpha.-MnO.sub.2 to be the predominant manganese oxide in the catalyst deposit. The specimen, upon testing, gave a cell voltage on oxygen that was very steady at .about.2.0 V for 47 days.

The catalyst was deposited by dipping in a saturated KMnO.sub.4 aqueous solution, heating at 250.degree. C. in air and leaching in boiling water. This cycle was carried out four times. The cathode (No. "B" in the Table) was then soaked in 10% H.sub.2 O.sub.2. The final loading was 14.6 mg/cm.sup.2. X-ray analysis indicated Mn.sub.2 O.sub.3 to be the predominant manganese oxide in the catalyst deposit.

In testing, the cell voltage on oxygen started out at .about.1.95 V and gradually increased to 2.15 V at 34 days and 2.48 V at 41 days. The cathode was removed after 45 days with an oxygen cell voltage of 2.65 V.

The catalyst was deposited by dipping in a saturated KMnO.sub.4 aqueous solution, heating to 250.degree. C. in an oxygen atmosphere and leaching in boiling water. This cycle was carried out four times. The final loading was 18.7 mg/cm.sup.2. The product cathode was No. "C" in the Table. X-ray analysis of the catalyzed electrode identified .alpha.-MnO.sub.2 as the predominant manganese oxide in the deposited catalyst layer. The performance was reasonably steady near about 2 volts through 35 days.

The catalyst was deposited by dipping in a saturated KMnO.sub.4 aqueous solution, heating at 250.degree. C. in air and leaching in boiling water. This cycle was carried out four times. The cathode was then soaked in 10% H.sub.2 O.sub.2. The final loading was 10.3 mg/cm.sup.2. The initial depolarization was good at 0.84 V.

Analogous excellent results are obtainable when the foregoing procedures are repeated excepting: to use other permanganates as reagents for the catalyst deposition, including lithium permanganate, silver permanganate, barium permanganate, calcium permanganate and so forth; and to make the catalyst deposits on porous electrodes of other substrate material, such as cobalt and iron.

Many changes and modifications can readily be made and adapted in embodiments in accordance with the present invention without substantially departing from its apparent and intended spirit and scope, all in pursuance and accordance with same as it is set forth and defined in the hereto appended claims.