Analyst, July, 1973, Vol. 98, $$. 475-484 475 Mass and Charge Transfer Kinetics and Coulometric Current Efficiencies Part VI.* The Pre-treatment of Solid Electrodes, and a Review of the Effects of Oxidation of Platinum BY E. BISHIOP AND P. H. HITCH COCK^ (Ckemistvy Department, University of Exeter, Stocker Road, Exeter, EX4 4520) The nature and condition of working electrode surfaces are set in the context of reaction speeds and current efficiencies. The formation of films and their effects are examined for platinum and other metals and alloys. Adsorption and specific adsorption on electrode surfaces are briefly reviewed. An attempt is then made critically to appraise the current state of the art in respect of the activation and deactivation of electrodes. Methods of cleaning electrodes are canvassed.The more significant theories of activation and deactivation are reviewed with specific reference to platinum electrodes. These theories include the impurity theory, the platinisation theory and the various oxygen-containing surface theories. For the last, the formation of oxides, the measurement of film thickness, the film thickness and the nature of the oxide reducible at 0.6 V are discussed, followed by a selective review of the oxygen-bridge theory, the half-reduced oxide theory and the platinum - oxygen alloy theory. Tentative conclusions are reached. Finally, gold electrodes and their behaviour are briefly examined. WITH very few, if any, exceptions, the primary overriding consideration in the rate of the charge-transfer process at a working electrode, and therefore in the current efficiency of the desired process, is the condition of the electrode surface.The coulometric current efficiency is a function of the relative rates of wanted and unwanted electrode processes, and all rates are affected, often to greatly differing degrees, by the condition of the electrode surface. No matter how carefully the rate parameters and solution conditions are determined, or how accurately the total current, I , and working electrode potential, Ewe, are measured at one instant in time, the mere act of using the electrode can change the charge-transfer kinetic parameters by several orders of magnitude. Particularly is this true of solid electrodes, but the assumption that liquid electrodes are self-cleaning is over-optimistic.Nor are the many forms of carbon electrode free from these effects. For any particular reaction at any particular kind of electrode, there is a maximum inherent speed that is governed by the activation energies of the transition states. Commonly, this maximum speed is attained at a perfectly clean electrode, and it is not possible to generalise on how perfect cleanliness is achieved or what it means. From the point of view of current efficiency, it is well to assume this maximum speed for all unwanted reactions, including the background solvent reactions, and to monitor at frequent intervals the real speed of the desired reaction. Unless the reaction is very fast, the anodic and cathodic directions occur a t potentials that are sufficiently different to engender a change in the nature of the electrode surface, and the apparent over-all conditional rate constant, k , will seldom be the same for both anodic and cathodic reactions.Theory states that the charge-transfer coefficient for the cathodic direction is cc, and that for the anodic direction is necessarily (1 - cc), but, while this may hold for fast reactions, the difference in potential between the two directions of reaction again makes it unlikely that the cathodic and anodic charge- transfer coefficients will add up to unity, and so the designation of p for the anodic charge- transfer coefficient retains some practical usefulness. The most generally used solid electrode material is platinum. A highly subjective, and selective, review of the pre-treatment and oxidation of platinum in the restricted context of coulometry is appropriate at this stage as a preliminary to the report of practical results.It cannot be empliasised too often that it is essential in any report fully to specify the electrode pre-treatment and solution purification. The objective of electrode pre-treatment is to give * For Part V of this series, see p. 465. t Present address: Ever Ready Co. (G.B.) Ltd., Central Research Laboratory, St. Ann’s Road, @ SAC and the authors. London, N. 15.476 [Analyst, Vol. 98 a reproducible electrode surface, which can be regained at any time by exact repetition of the pre-treatment. The pre-treatment of electrolyte solutions has not only the objective of discovering and removing any electroactive impurities in them, but also of removing trace amounts of surfactant materials that may be detectable only by their malign influence on electrode behaviour.De-ionised water and surfactant detergents should be avoided at all costs. BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER KINETICS FILMS ON PLATINUM The nature, condition and participation in reactions of the electrode surface is a com- plicated and contentious subject, extremely difficult experimentally to investigate and little productive of unequivocal interpretation, as the extensive literature attests. Noble metals are by no means inert,l and both chemical and anodic oxidation produce similar attacks2 Baker and McNevin3 found that oxidised surfaces slowed the oxidation of arsenic(III), and thereafter adopted a cathodic pre-treatment of platinum as a routine procedure.4 They ascribed the interference of trace amounts of antimony to the formation of a film of the insoluble antimony oxide Sb,O, on the electrode ~urface.~ Bard5 identified hydrated tin oxide films on cathodes.Lingane and co-workers6 detected oxidation of platinum during the oxidation of iodine to iodate, and later provided chemical evidence of anodic att.ack7 and film formation in chloride media,8 although the precise interpretation of the evidence was ~hallenged.~ DavislO observed oxidation of platinum in the chronopotentiometry of iron( 111) and cerium(IV), and found that repeated reduction of the electrode was necessary in the oxidation of hydroxylamine.ll Anson found that the reduction of iodate was much faster at an oxidised electrode,12 first postulating an oxygen-bridge mechanism and then platinisa- tion.13 Lingane found that an oxide film stopped the oxidation of 0xa1ate.l~ This effect was ascribed to prevention of the adsorption of oxalic acid, which is a necessary preliminary to its 0xidati0n.l~ Organic electrode processes are strongly affected by the condition of the electrode,16 and the rate constant of the hexacyanoferrate( 111) - hexacyanoferrate( 11) process at an oxidised electrode is only one tenth of that at a clean ele~trode.1~ Kozawals ascribed humps in voltammograms for alkaline solutions to the formation of oxides or hydroxides. Laitincii and Enkelg showed that the process of oxidation is slow, and postulated a hydroxyl radical mechanism.Feldberg, Enke and Bricker20 favoured chemisorption followed by a two-step process. Meyell and Langer21 showed that the oxide layer thickened with increasing anodisation potential until it became constant at potentials above 1.8 V. Lingane,, showed that the reduction of oxygen to water at platinum without the intermediate formation of hydrogen peroxide was fast at an electrode that had one tenth of a monolayer of oxide on its surface, but was slow at a reduced electrode. He suggested that a chemical reaction between oxygen and platinum occurred, followed by reduction of the oxide; this suggestion was attacked by Anson and King,13 but was supported by Sawyer and Interrante,23 who ascribed the decreasing speed of reaction with time to ageing of the oxide film, and found the speed at a pre-reduced electrode to be so slow that the electrode retained the characteristics of a reduced surface even in oxygen-saturated solutions.Other metals have been examined less extensively. Palladium and gold are oxidised in the same way chemically and anodically., Gold23-,6 is particularly attacked in media that contain complexing ions such as cyanide and chloride, but is more resistant than platinum to oxidation. Attack of single-crystal silver depends on the crystal face exposed to the solution.27 Alloys of gold and p l a t i n ~ m ~ 8 ~ ~ ~ show interesting properties, as do those of gold with palladium, silver and platinum.30 Iridium migrates preferentially to the surface of platinum - iridium all0ys.~1 ADSORPTION ON ELECTRODES Adsorption in general, or specific adsorption, of reactant, product or other species may be a necessity, a nuisance or a disaster, and again has an extensive literature; some has already been quoted.&l27l5 Anson has made many contributi~ns.~~-~l He ascribed activity to very light platinisation of the electrode on reduction of the oxide,32 found that perchloric acid could not replace sulphuric acid in chemical treatment with iron(II),12 and ascribed the accelerated reduction of iodate1, and vanadate at a pre-oxidised electrode to electrolytic p1atini~ation.l~ The same reason was adduced by Bard42 for the accelerated oxidation of hydrazine at a.c.biassed activated electrodes. Feldberg, Enke and Bricker20 contended thatJuly, 19731 AND COULOMETRIC CURRENT EFFICIENCIES.PART VI 477 the activity resided in a half-reduced oxide. Adsorption effects were detected in studies of the iron(II1) - iron(I1) system,33 but chronopotentiometric evidence of the adsorption of iron ions37 was traced to capillary cracks in the glass-to-metal seal of the electrodeJ36 but adsorption of and prevention of adsorption of oxalate14J5 by oxide films is confirmed. Anson and Schultzf5 postulate that adsorption of certain species, such as oxalate, iodide, hydrogen peroxide , arsenic (111) , thiocyanat e, nitrite , et hylenediaminetetraacet at ocobalt at e (I I) and methanol, is a necessary preliminary to their oxidation, and adsorption and oxidation are inhibited by oxide films. Adsorbed bromide catalyses the anodic oxidation of ethylene- diaminetetraacetatocobaltate(I1) and the product contains no br~mide.~s Adsorption of ethylenediaminetetraacetatocobaltate(II1) ion on a platinum cathode inhibits the reduction of the unadsorbed complex ion.39 Preferentially adsorbed ions, such as iodide, will displace adsorbed complex ions and so eliminate the interference.The cobalt (11) complex reduction product is also adsorbed. Mercury cathodes extensively adsorb the cobalt (11) complex,40 but not the tris(ethy1enediamine)cobalt (111) cationJ41 contrary to previous findings.43344 Inter- face and film inhibition at mercury electrodes have been reviewed by Oeder, Seiler and F i ~ c h e r . ~ ~ Unusual adsorption effects at mercury in the reduction of flavin nucleotide have been observed.46 Murray and co-worker~~~-~~ used several techniques in adsorption studies at mercury electrodes.Reduction of mercury(1) and bismuth( 111) is inhibited by adsorbed lead iodide, bromide and t h i o ~ y a n a t e ~ ~ , ~ ~ ; charge transfer is slowed by a chemical step at an adsorption-blocked surface in the reduction of copper(I1) in tartrate media by brucine. Supporting electrolyte effects in non-aqueous media have been examined ,51 and adsorption has been deliberately used in electrochemical masking with an adsorbed metal complex.52 Adsorbed lead selectively penetrable by different ions was used in the analysis of silver - mercury( 11) mixtures. Interpretative studies by Wopschall and Shai11~~-55 set out to test a theory53 by examination of the reduction of methylene blue,54 wherein the product is strongly adsorbed, and of the azobenzene - hydrazobenzene system,55 wherein there is a succeeding chemical reaction of the adsorbed reactant.Returning to platinum, thin-layer cells have been used in adsorption s t ~ d i e s . ~ ~ - ~ ~ Adsorption of EDTA complexes of cobalt and iron,57 and of iodine - iodide systems58 has been examined. Napp and Bruckenstein59 revealed the risks of unsuspected adsorption in a ring-disc electrode study of 0.5 M hydro- chloric acid at potentials from 0.25 to 1.25 V. They challenged earlier work,8 and ascribed the observed behaviour to a trace amount of copper(I1) in the medium, which is reduced to and adsorbed as copper(1) on the electrode surface. GileadiG0 has reviewed the adsorption of uncharged molecules on solid electrodes.ACTIVATION AND DEACTIVATION OF ELECTRODES A caveat against de-ionised water and detergents has already been entered. Flaming, or dipping in alcohol and flaming, will only poison the electrode. Treatment with chromic acid is strongly to be deprecated: not only does chromium(V1) oxidise the surface, but chrom- ium species are tenaciously adsorbed on platinum and block entirely such reactions as the reduction of vanadium(V).61,62 Reduction does not remove the chromium, and washing so prolonged as to remove it will result in a deactivated electrode. Chemical stripping in fresh, warm aqua regia, followed by washing with quartz-distilled waterJs3 will remove gross con- tamination ; anodisation in hydrochloric acid serves much the same purpose. This stripping stage can be followed by chemical reduction64 or cathodisation.Commonly, however, several cycles of anodisation and cathodisation in purified sulphuric acid, finishing with a protracted cathodisation, are used, and hydrogen adsorbed on or dissolved in the electrode is removed potentiostatically at 0.25 to 0.4 V. It is not possible to generalise further. In use, the electrode will be deactivated and need fresh pre-treatment. Setting aside specific adsorption, discussed above, of the very many theories of electrode behaviour, it is worth examining three, which have commanded considerable attention in the past, or appear to have present value. These theories are the impurity theory, the platinisation theory, and a group that may be termed oxygen-containing surface theories, and all are specifically directed to platinum.THE IMPURITY THEORY- The malign influence of surf ace-active impurities was demonstrated in 1937 by Frumkin’s Ion-exchange resins gave gr0up,~5--~~ and 30 years later68 was used in their determination.478 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER KINETICS [Analyst, l70l. 98 water of low conductance, but M for double-distillation in quartz. Charcoal columns removed most of the surfactant impurities but introduced ionic impurities, which may or may not be of importance. We have found charcoals to be in- effective for sulphuric acid purification, but both and c h a r c ~ a l ~ ~ ~ ~ ~ have been used successfully. We have found adsorption on a large electrode to be successful, but erratic, for sulphuric acid61; Bockris and c o - w ~ r k e r s ~ ~ ~ ~ ~ have used this method.BarkerTd has used a combination of column and electrolytic methods. Trial and error is the only general approach to purification of solvents and electrolytes, and little is known of the nature of surfactants; both organic and inorganic materials can be r e s p o n ~ i b l e . ~ ~ ~ ~ ~ One difficulty may be removed at the cost of introducing another, and particular care is needed in treating sample solutions.70 used solution cleaning methods in a study of the effect of impurities on the iron(II1) - iron(I.1) and hydrogen ion - hydrogen gas processes at platinum. He recorded the current at constant potential of a small, spinning platinum-wire electrode. In untreated iron(II.1) - iron(I1) in perchloric acid, an electrolytically activated electrode showed a current that decreased with time, until it became equal to the current at an untreated electrode and then remained constant.After charcoal “cleaning” of the solution, both treated and untreated electrodes behaved alike, this time in the manner of the initial behaviour of a treated electrode in an untreated solution. He concluded that the electrolytic treatment removed impurities from the surface of the electrode, which, on exposure to a cleaned solution, did not become dirty again. Similar observations were recorded for the hydrogen system, but the charcoal treatment, whether or not followed by electrosorption, did not entirely remove surfactants from the sulphuric acid. Both before and after “cleaning” of the electrolyte, the rate of decay of the current was accelerated by increasing the rotation speed of the electrode. This effect was to be expected when deactivation depended on migration of impurities to the electrode.James concluded that deactivation arose by re-adsorption of solution impurities, and not solution ions, and that it was unlikely that slow dissolution of hydrogen gas in the metal surface contributed to deactivation. Damj anovic, Genshaw and B ~ c k r i s , ~ ~ studying oxygen reduction in acidic media, supported the view that impurities exerted a profound effect. I t is notable that scrupulously clean solutions in equally clean vessels containing perfectly clean electrodes acquire impurities even when the system is completely ~ e a l e d .~ ~ s ~ ~ s ~ 8 Warner, Schuldiner and Pierama78 found that platinum electrodes kept on open circuit in the solution remained clean for several weeks, but when in continuous use, impurities built up, and they concluded that the impurities originated from the electrodes. This conclusion need not be true. An impurity theory is impossible to disprove because the level needed to interfere with electrode processes is so low that the impuritycannot be detected other than by its interference. Acceptance of a form of impurity theory does not reveal the full picture, and other effects require consideration. THE PLATINISATION THEORY- Smooth platinum gives no zero-current hydrogen-ion response, as it is unable to catalyse the rate-determining combination of hydrogen atoms.When coated with finely divided palladium-black or platinum-black, the effective surface area is multiplied several thousand- fold and contains the necessary energetic sites for the catalysis. The platinisation theory argues for an increase in surface area and catalytic power, and states that anodisation - cathodisation cycling produces light platinisation on reduction of oxide. Certainly initially bright surfaces are quickly dulled and etched by this treatment. Anson and King13 used 60-Hz a.c., probably of a very rough waveform, and produced a grey to black deposit, and Shibata79ss1 observed a similar phenomenon, claiming that the oxide was PtO,. Hoare,82 however, investigated a.c. polarisation of platinum, rhodium, palladium, iridium and gold, and found that only those metals which are capable of dissolving hydrogen, i.e., platinum and palladium, formed roughened surfaces, and ascribed the break-up of the surface to alternate dissolution and removal of hydrogen, One might ask whether roughness increases the charge- transfer rate.Anson and King13 did not purify their reagents, and the activation may simply be removal of impurities.’O Shibatasl repeatedly distilled his hydrochloric acid in a quartz still, but this is still not unequivocal in view of the work of Berezina and Nikolaeva-Fedorovich.68 Shibata cleaned platinum electrodes, aged them for 2 weeks in hydrogen, oxidised them chemically or anodically, and determined the amount of oxide chronopotentiometrically. Oxides formed in either way showed a single halt at 0.6 V and the amount increased with time M in surfactants againstJuly, 19731 AND COULOMETRIC CURRENT EFFICIENCIES.PART VI 479 of oxidation. Prolonged oxidation produced what seemed to be a second oxide, which was reducible at 0.3 to 0.2 V. All electrodes showed the same initial activity in hydrogen produc- tion, and the decay in activity was a function of the oxidation time. At the extreme, anodi- sation for 28 hours at 100 mA cm-2 gave an electrode that showed a negligible loss of activity after 30 hours of continuous use. The dependence of decay rate on oxidation time seems to exclude adsorption of impurities for deactivation, and Shibata opined that the platinised layer produced was unstable, but had a high initial activity that decreased as the platinised layer recrystallised to a stable form.Shibata’s work suggests that impurities alone cannot account for deactivation, but this neither proves the platinisation theory nor disproves the impurity theory. Reports of surface recry~tallis2tion~~-~~ have been challenged by Gilman,86$87 who found the deactivation to be dependent on stirring speed, which argues for impurity deactivation. Warner, Schuldiner and Pierama78 used clean electrodes and electrolyte and showed that well annealed platinum was more active than unannealed or drawn platinum in the reaction but annealing was without effect on simple electrochemical reactions. The work of Khazova, Vasil’ev and Bagotskiis8 would seem effectually to disprove the platinisation theory. They studied the catalytic and electrochemical activities of smooth and platinised platinum over the roughness factor range from 2 to 6500, and found the kinetic principles for chemisorption and electro-oxidation to be the same for all electrodes.The maximum rate of either type o€ reaction in relation to the true surface area decreased by an order of magnitude from smooth to highly platinised electrodes ; the sharpest change occurred in the roughness factor range from 10 to 1000, and further increase in roughness did not affect the rates. Observation of transients absolved mass transfer from the decrease in rate, so platinisation reduces the charge-transfer rate constant. Pt-0 + H2-Pt + H2O OXYGEN-CONTAIKING SURFACE THEORIES- This awkward term attempts to include several theories ranging from the formation of stoicheiometric oxide films to dissolution of oxygen in platinum.The subject is contentious in the extreme, presents enormous experimental difficulties, absorbs a huge research effort, not least in fuel and has produced a formidable literature, of which this review can do no more than skim the surface in the special context of coulometry, as previously adum- brated.64 Dispute continues over the nature of oxidised platinum, originating from the difficulty in distinguishing chernisorption from stoicheiometric phase oxides. Confusion is the more confounded by lack of reproducibility of the results (or inadequate reporting) of the cathodic-stripping determination of the amount of oxygen present. Contentions that films scarcely exceed a m o n ~ l a y e r ~ l - ~ ~ are difficult to prove.Electron and X-ray diffraction are useless, but ellipsometry assisted by coulometric stripping appear at present to offer a reason- able guide.gi19g5 At potentials cathodic to 0-9s V, there was no ellipsometric evidence of oxide; the sensitivity was about 0.01 nm. At 0.98 & @OP V, a film of 0.02 nm thickness suddenly appeared and grew linearly with increasing potential to 1.6 V, when the thickness was 0-55 nm; higher potentials were not examined. Coulometryg4 indicated that a 0-05-nm thick film was present below 0.98 V, and that the film thickened at about the rate indicated by ellipsometry up to 1.6 V, but the film was always 0.85 nm thicker than was found by ellipsometry. A complete monolayer would be about 0.3 nm thick.Ellipsometry depends on a difference in optical constants of the film and the surrounding medium, and the difference between chemisorbed osygen and water is too small to show up. Below 0.98 V, the coulometrically detectable oxygen was interpreted as being chemisorbed, and above 0-9s V a phase oxide was formed. The coexistence of chemisorbed oxygen and phase oxide had been predicted earlier.96 Film thickness-Chronopotentiometric reduction after conventional anodisation indicates a single oxid? reducible at 0.6 V,7714319978 and coulometry supports the view that no more than a monolayer is formed,7~8~14~20~78~s1~s3~s7~g8 amounting to 0.3 to 0.74 mC cm-2 depending on the roughness factor, but there is some evidence of thicker filims.1,81796 Endeavours by Cooksey and Bishop to establish precise growth-rate laws have been frustrated by poor reproducibility, but have emphasised that the competitive formation of molecular oxygen during oxidation has too often been ignored.Stronger anodisation yields thicker films,21~81~s4~99-101 and in some instances a second reduction h d t at 0.2 to 0.3 V has been noted.829101v102 Meyell and Langer21480 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER KINETICS [Analyst, VOl. 98 have reported many platinum to oxygen ratios, including a “tight” PtO. Their de-oxygenation procedure was inadequate, and Hoarelo3 rightly suggests that unremoved oxygen contributed to the stripping current. Visscher and DevanathanS4 reported a linear relationship between the stripping charge and the potential of oxide formation (Bishop and Cooksey find this to be a gross oversimplification); even with a roughness factor of 2.5,21 oxide formed at 1.7 V required 2 mC cm-, for complete reduction, implying a layer six to seven atoms thick.Kolthoff found that layers 0.3 to 1.4 mC cm-, thick were formed on chemical oxidation, depending on the nature of the oxidant and the time of attack. Prolonged anodisation gave second halts at 0.3 Vsl or 0-2 V,lol and the amount of low-voltage oxide increased with anodisa- tion time. Shibatagl reached 36 mC cm-2, which corresponds, at a roughness factor of 2.5, to forty oxygen atoms per surface platinum atom. The nature of the oxide reducible at 0.6 V-This has been extensively researched, as Young’s104 and Hoare’slo5 books and Hoare’s extended reviewlO3 show.The re-interpretation9 of Anson and Lingane’s chemical stripping results’ has been mentioned; the same objections to other work can be raised.loO Hoarelo3 canvasses a wide variety of likely and unlikely oxides. The most widely held view is that moderate anodisation produces PtO or Pt(OH), in hydrated form, and probably some chemisorbed oxygen as well, but drastic conditions are required in order to produce platinum(1V) oxide, which is the reverse of the situation with sulphide, where PtS, is formed but not PtS.lo6 Decreases, often drastic, in the rate of the charge-transfer process at oxide-filmed electrodes have been rep0rted.13J4J~,4~~~~~~O~ MiillerlO8 demonstrated that the rate of the charge-transfer process is proportional to the surface area of the platinum that remains unoxidised.Reports12y107-109 of reactions accelerated by oxide filming have mostly been d i s c o ~ n t e d ~ ~ , ~ ~ ; what has been observed can be interpreted as reaction at a bare platinum surface after the oxide has been chemically stripped. THE OXYGEN-BRIDGE THEORY- The analogy with the oxygen-bridge mechanism in homogeneous oxidation - reduction reactions is tempting but false. An attemptllO was made to explain an observed “reversibility” of freshly anodised platinum electrodes on a basis of an “oxygen bridge” structure of the oxide layer, claimed to accelerate the iron(II1) - iron(I1) and cerium(1V) - cerium(II1) reac- tions. Anson first lent support to this interpretation,12 but later changed to the platinisation theory.13 Davislo7 found that heavy oxide films suppressed the iron(II1) - iron(I1) reaction as earlier predicted,3 and the vanadium(V) - vanadium(1V) reaction, but claimed that a light film, less than a monolayer, facilitated oxygen-bridge formation and accelerated these reactions.He suggested that oxide formation occurred at grain boundaries, but this suggestion was conclusively challenged.lll James76 demonstrated that a simple oxygenated surface could not account for platinum activation. At a potential of 0-8 V, a t which the oxide should be stable, he found a considerable decay in activity with time ; contrarily, prolonged anodisation produced long-lived activity in hydrogen evolution at 0.0 V, at which the oxide should have been completely destroyed.THE “HALF-REDUCED OXIDE” THEORY- Studying the oxidation of platinum in perchloric acid, Feldberg, Enke and Bricker20 concluded that film formation was a two-step process, the first being slow formation of Pt(OH), and the second fast oxidation to Pt(O),. On cathodic stripping, reduction of Pt(O), to Pt(OH), was fast, but the slow reduction of Pt(OH), to bare platinum was complete only after prolonged potentiostatic reduction, while the process stopped a t Pt (OH), on ampero- static reduction, the current then being used in the generation of hydrogen. If the quantity of electricity required to form the oxide is Qa and for reduction is Qc, then on the first cycle Qa/Qc should be 2, decreasing to unity. This, by and large, is so. The arguments against this theory, that the half-reduced oxide is the active condition, are the same as for the oxygen-bridge the01-y.~~ The observations can be interpreted in several ways.Vetter and Berndt112 adduced different reactions, saying that the oxide was formed from water by a four-electron step, and reduced to hydrogen peroxide by a two-electron step. Breiter,l13 using low-frequency cyclic voltammetry, found that Qa/Qc started at greater than 2 but did not fall below 1.18, which agrees with our findings,64 and considered that the observations were contrary to previous t h e ~ r i e s . ~ O , ~ ~ ~ He suggested that Qa included oxidation of organic i m p ~ r i t i e s , l l ~ , ~ ~ ~ while Qc was little affected because reduction of the oxide occurred atJuly, 19731 AND COULOMETRIC CURRENT EFFICIENCIES. PART VI 481 potentials negative to 0.9 V, when such reactions became very slow.He also claimed that whereas oxygen atoms dissolve readily in platinum, their desorption is slow. Preferring the first explanation, he regarded Qc as a measure of the “electrochemical cleanliness” of the system. We s u g g e ~ t ~ l ? ~ ~ that the formation of molecular oxygen occurs during anodisation, and that this oxygen in part diffuses away from the electrode and is therefore not reduced during cathodisation. Gilmans6ys7 reached similar conclusions. THE PLATINUM - OXYGEN ALLOY THEORY- Breiter’s suggestion about oxidisable i m p u r i t i e ~ l l ~ J ~ ~ may well be right, but his suggestion of dissolution of oxygen in platinum is supported by other work116 in which a pulse and decay technique was used.Perhaps the most significant work at the time of writing is that of Hoare,92J03~105~117-127 who ascribes the activity of pre-anodised electrodes to oxygen dis- solved in the surface layers of the bulk Oxygen has been reported to dissolve in massive platinum to a depth of three to four layers before saturation is rea~hed.10~~11~1128-130 Schuldiner and Warner116 found a direct connection between oxygen dissolved in platinum and the catalytic activity of the metal. Hoare based his hypothesis126 on earlier work,92s119y122 and made solutions of reagent grade chemicals in triply distilled water and pre-electrolysed the solutions for a t least 24 hours before use,117 and was thus reasonably assured of impurity- free media.Further, he found that oxidation in concentrated nitric acid produced an electrode that displayed a steady potential of 1.225 V for over 24 hours when immersed in 1.0 M sulphuric acid saturated with pure oxygen, in accord with prediction.121 Such an electrode also cata- lysed the reduction of oxygen and hydrogen peroxide in acidic media.122J26 A minimum of 50 hours’ soaking was needed in order to produce an electrode of stable activity, which is much longer than is necessary in order to remove impurities or to coat the metal with oxide. This result is compatible with the postulate that oxygen dissolved in the metal to give a “platinum - oxygen alloy” electrode.92 A cathodic chronopotentiogram of one of Hoare’s activated electrodes has been quoted76 as showing a single arrest a t 0.7 V similar to that of conventionally anodised electrodes.Hoare reports that when his activated electrodes were immersed in hydrogen-saturated acidic solution, they took 40 minutes to reach the same potential as a reduced electrode, compared with 4 minutes for conventionally anodised elec- trodes.1191127 Breiter131 reported that nitrogenous species were tenaciously adsorbed on platinum soaked in nitric acid, but Hoare could not detect such species.127 Hoare’s prolonged treatment with nitric acid seems to produce electrodes with unique properties, which he attributes to the platinum - oxygen alloy, but it can equally be postulated that prolonged anodisation produces not only a surface-phase oxide but also some of this alloy or solid solution, which suffices to activate the electrodes towards oxidation - reduction systems.The oxide layer would be stripped chemically or cathodically, but the oxygen in solid solution would be removed only very slowly.126 Hoare does not offer a mechanism; this is probably not yet possible. Much of the earlier work can be re-interpreted in the light of Hoare’s suggestion. Particularly, Shibata’s resultss1 concerning prolonged anodisation could be explained by the extensive formation of the platinum - oxygen alloy, with consequent long-lived activity, rather than by platinisation. CONCLUSIONS There can be no finality yet: hot debate continues. No sooner has one read one paper that appears to be convincing and conclusive, than another of contrary view that is equally convincing and conclusive appears.Tentatively, however, desorption and adsorption of impurities can be said to contribute substantially to activation and deactivation processes, respectively, and a t the moment Hoare’s platinum - oxygen alloy theory seems to be the most attractive. Chemical oxidation1,2 seems to produce the sameoxide, which can be stripped a t 0-6 V, as anodisation, but electrochemical activation is to be preferred. There are many report~,103~10*y1~~~13~ some of considerable a n t i q ~ i t y , l ~ ~ - l ~ ~ of yellow or red films formed on platinum by means of a periodic current of 50 to 60 Hz, with a d.c. bias. Such films could well consist of oxide. In our laboratory, extremely hard, scratch-proof gold films have been produced by precisely symmetrical waveforms, but the rough waveforms from such an iron-cored inductance as a transformer gave black films that could be wiped off with a moist paper tissue.482 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER KINETICS [A%dySi!, Vol.98 GOLD ELECTRODES The behaviour of gold is simpler because neither hydrogen nor oxygen is soluble, and has raised less controversy. An oxidised surface renders gold a poor electrode for oxidation - reduction reaction~,~5,~~ but as soon as the oxide has been reduced, reactions proceed at normal speed. Hoare has examined gold as a possible oxygen electr0de,1~~J~~-1~5 but it seems118J40 that gold gives a mixed potential in oxygen-saturated solutions. Potentiostatic in acidic oxygen solutions indicate a partial layer of oxygen to be adsorbed between 0-9 and 1-3 V, which is a good electrical conductor. Above 1.36 V, Au,03, probably hydrated, is formed and is not a good conductor,l12 so that passivation occurs between 1.36 and 1.60 V.I R e p o r t ~ ~ , ~ ~ ~ of the formation of Au20 and AuO are not held in much regard. A trans-passive region between 1-60 and 2.0 V has been attributed to the migration of gold atoms through the loosely packed Au20, film.f42 Such migration produces an oxide film that is much thicker than a Reduction of an electrode anodised at high potentials gives considerable roughening, attesting to a thick porous oxide film.140 Baumann and recommend oxidation - reduction pre-treatment, and ascribe the improved activity to the removal of surface impurities. This improvement is probably the main consequence of electrolytic pre-treatment, although roughening may give minor benefit.No attempts have been made to explain the activity of gold electrodes on a basis of an oxygen-containing surface. Charging curves for go1d20,23J45 indicate that Qa = Qc, which accords with the insolubility of oxygen in gold. The proposal of oxygen diffusion along grain boundaries23 has been challengedlll on the grounds of lack of evidence. Gold is more resistant25 than platinum to chemical oxidation in non-complexing media, but gold dissolves readily in media that contain cyanide or halide ions. This attack starts at 0.6 V in chloride media, and so severely restricts the use of gold as an anode. We express our sincere gratitude to Imperial Chemical Industries Limited for a research grant covering 3 years.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. REFERENCES Kolthoff, I. M., and Tanaka, N., Analyt. Chem., 1954, 26, 632. Lee, J. K., Adams, R. N., and Bricker, C. E., Analytica Chim. Acta, 1957, 17, 321. Baker, B. B., and McNevin, W. M., J . Awer. Chem. SOC., 1953, 75, 1476. ___- , Ibid., 1953, 75, 1473. Bard, A. J., J . Electroanalyt. Chem., 1962, 3, 117. Anson, F. C., and Lingane, J. J., J . Amer. Chem. SOG., 1957, 79, 1015. , Ibid., 1957, 79, 4901. Peters, D. G., and Lingane, J. J., J . Electroanalyt. Chem., 1962, 4, 193. Breiter, M. W., and Weininger, J . L., J , Electvochena. SOG., 1962, 109, 1135. Davis, D. G., Analyt. Chem., 1961, 33, 1839.Anson, F. C., J . Amer. Chem. Soc., 1959, 81, 1554. Anson, F. C., and King, D. M., Analyt. Chem., 1962, 34, 362. Lingane, J. J . , J . Electroanalyt. Chem., 1960, 1, 379. Anson, F. C., and Schultz, F. A., Analyt. Chem., 1963, 35, 1114. Bogdanovskii, G. A., and Shlygin, A. I., Zh. Fiz. Khim., 1960, 34, 26. Daum, I?. H., and Enke, C. G., Analyt. Chem., 1969, 41, 653. Kozawa, A., Bull. Chem. SOC. Japan, 1962, 35, 1051. Laitinen, H. A., and Enke, C. G., J . Electrochem. Soc., 1960, 107, 773. Feldberg, S. W., Enke, C. G., and Bricker, C. E., Ibid., 1963, 110, 826. Meyell, J. S., and Langer, S. H., Ibid., 1964, 111, 438. Lingane, J. J., J . Electroaizalyt. Chem., 1961, 2, 296. Sawyer, D. T., and Interrante, L. V., Ibid., 1961, 2, 310. Laitinen, H. A., and Chao, M. S., J .Electrochem. SOC., 1961, 108, 726. Baumann, F., and Shain, I., Analyt. Chem., 1957, 29, 303. Genshaw, M. A,, Damjanovic, A., and Bockris, J. O’M., J . Electroanalyt. Chem., 1967, 15, 163. Nikulin, V. N., Russ. J . Phys. Chem., 1961, 35, 40. Breiter, M. W., J . Electroanalyt. Chem., 1966, 11, 157. Damjanovic, A., and BrusjC, V., Electrochim. Acta, 1967, 12, 1171. Giner, G., Parry, J. M., and Swette, L., Abstv. Pap. 154th Meet, Amer. Chew. Soc., September 1967, L15. Chodos, A. A., and Meites, I>., Analyt. Chem., 1969, 41, 846. Anson, F. C., Ibid., 1961, 33, 934. -, Ibid., 1961, 33, 939. -, Ibid., 1961, 33, 1123. -, Ibid., 1961, 33, 1438. I -- -, Ibid., 1963, 35, 764. -, Ibid., 1961, 33, 1498.July, 19731 AND COULOMETRIC CURRENT EFFICIENCIES.PART VI 483 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. -, J . Amer. Chem. Soc., 1961, 83, 2387. -, J. Electrochem. Soc., 1963, 110, 436. -, Analyt. Chem., 1964, 36, 520. -, Ibid., 1964, 36, 932. Bard, A. J., Ibid., 1963, 35, 1602. Laitinen, H. A., and Randles, J. E. B., Trans. Faraday Soc., 1955, 51, 54. Laitinen, H. A., and Chambers, L. &!I., Analyt. Chem., 1964, 36, 5. Oeder, D., Seiler, W., and Fischer, H., Z . analyt. Chem., 1966, 216, 256. Hartley, A. M., and Wilson, G. S., Analyt. Chem., 1966, 38, 681. Gross, D.J., and Murray, R. W., Ibid., 1966, 38, 405. Kodama, M., and Murray, R. W., Ibid., 1065, 37, 1638. Murray, R. W., and Gross, D. J . , Ibid., 1966, 38, 392. Murray, R. W., and Kodama, M., Ibid., 1965, 37, 1759. Murray, R. W., and Hiller, L. K., Ibid., 1967, 39, 1221. Murray, R. W., and McNeeley, R. L., Ibid., 1967, 39, 1661. Wopschall, R. H., and Shain, I., Ibid., 1967, 39, 1514. , Ibid., 1967, 39, 1527. , Ibid., 1967, 39, 1535. Hubbard, A. T., and Anson, F. C., J . Electroanalyt. Chem., 1965, 9, 163. -__ , Analyt. Chem., 1966, 38, 1601. Hubiard, A. T., Osteryoung, R. A., and Anson, F. C., Ibid., 1966, 38, 692. Napp, D. T., and Bruckenstein, S., Ibid, 1968, 40, 1030. Gileadi, E., J. Electroanalyt. Chem., 1966, 11, 137. Hitchcock, P. H., Ph.D. Thesis, University of Exeter, 1969.Bishop, E., and Hitchcock, P. H., Analyst, 1973, 98, in the press (Part VII of this series). Bishop, E., and Sutton, J . R. B., Analytica Chim. Acta, 1960, 22, 590. Bishop, E , and Riley, M., Analyst, 1973, 98, 416. Levina, S., and Sarinsky, V., Acta Phys-Lhim. URSS, 1937, 6, 475. -- , Ibid., 1937, 6, 491. Erschler, B., and Proskurnin, M., Ibid., 1937, 6, 195. Berezina, N. P., and Nikolaeva-Fedorovich, N. V., Soviet Electrochem., 1967, 3, 1. Hillson, P. J., Trans. Faraday Soc., 1954, 50, 385. James, S. D., Electrochim. Acta, 1967, 12, 939. Joshi, K. M., Mehl, W., and Parsons, R., in Yeager, E., Editor, “Transactions of the Symposium on Electrode Processes, 1959,” John Wiley, New York, 1960, p. 249. Bockris, J . O’M., and Conway, B. E., Trans.Faraday Soc., 1949, 45, 989. Damjanovic, A., Genshaw, M. A., and Bockris, J . O’M., J . Electrochem. SOC., 1967, 114, 466. Barker, G. C., Rep. U.K. Atom. Energy Auth., AERE-C/R 1563, 1957. -, J. Electrochem. SOG., 1966, 113, 1024. James, S. D., Ibid., 1967, 114, 1113. Bockris, J . O’M., and Huq, A. K. M. S., Proc. R. SOC., A , 1956, 237, 277. Warner, T. B., Schuldiner, S., and Pierama, B. J., J . Electrochem. SOC., 1967, 114, 1120. Shibata, S., Nippon Kagaku Zasshi, 1958, 79, 239. -, Bull. Chem. SOC. Japan, 1963, 36, 525. Hoare, J. P., Nature, Lond., 1964, 204, 71. Breiter, M. W., and Kennel, B., 2. Elektrochem., 1960, 64, 1180. Will, F. G., and Knorr, C. A., Ibid., 1959, 63, 1008. , Ibid., 1960, 64, 258. Gilman, S., J. Phys. Chem., 1963, 67, 78. -, J. Electroanalyt.Chem., 1965, 9, 276. Khazova, 0. A., Vasil’ev, Yu. B., and Bagotskii, V. S., Soviet Electrochem., 1967, 3, 915. Bagotskii, V. S., Nekrasov, L. N., and Shumolova, N. A. Buss. Chem. Revs, 1965, 34, 717. Kronenberg, M. L., J . Electroanalyt. Chem., 1066, 12, 122. Boeld, W., and Breiter, M., Electrochim. Acta, 1961, 5, 145. Hoare, J. P., Nature, Lond., 1964, 204, 71. Vetter, K. J., “Electrochemical Kinetics,” Academic Press, New York, 1967, p. 619. Visscher, W., and Dcvanathan, M. A. V., J . Electroanalyt. Chem. 1964, 8, 127. Reddy, A. K. N., Genshaw, M., and Bockris J. O’M., Ibid., 1964, 8, 406. Dietz, H., and Gohr, H., Electrochim. Acta, 1963, 8, 343. Schuldiner, S., and Roe, R. M., J . Electrochem. Soc., 1963, 110, 332. Schuldiner, S., Warner, T. B., Anson, F.C., and Osteryoung, R. A. Analyt. Chew., 1964, 36, 2510. Every, R. L., J . Electrochem. SOC., 1965, 112, 524. Every, R. L., and Grimsley, R. L., J . Electroanalyt. Chem., 1965, 9, 165. Kozawa, A., Ibid., 1964, 8, 20. Obrucheva, A. D., Z h . Fiz. Khim., 1952, 26, 1448. Hoare, J. P., “The Oxygen Electrode on Noble Metals,” in Delahay, P., Editor, “Advances in Electrochemistry and Electrochemical Engineering,” Volume IV, John Wiley, New York, 1967. Young, L., “Anodic Oxide Films,” Academic Press, New York, 1961. Hoare, J. P., “The Electrochemistry of Oxygen,” Interscience Publishers, New York, 1968. Najdeker, E., and Bishop, E., J . Electroanalyt. Chew., 1973, 41, 79. -, Ibid., 1966, 38, 54. , __- > __- -, Ibid., 1959, 80, 453. * --484 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. BISHOP AND HITCHCOCK Davis, D. G., Talanta, 1960, 3, 335. Muller, L., EZectrochim. Acta, 1967, 12, 557. Karp, S., andMeites, L., J , Amer. Chem. Soc., 1962, 84, 906. Kolthoff, I. M., and Nightingale, E. R., Analytica Chim. Acta, 1957, 17, 329. Mohilner, D. M., Argersinger, W. J., and Adams, R. N., Ibid., 1962, 27, 194. Vetter, K. J., and Berndt, D., 2. Elektrochem., 1958, 62, 378. Breiter, M. W., Electrochim. Acta, 1966, 11, 905. -, Ibid., 1962, 7 , 533. -, J. Electrochem. Soc., 1962, 109, 42. Schuldiner, S., and Warner, T. B., Ibid., 1965, 112, 212. Hoare, J. P., Ibid., 1962, 109, 858. -, Ibid., 1963, 110, 245. -, Ibid., 1963, 110, 1019. -, Electrochim. Acta, 1964, 9, 599. -, J. Electrochem. Soc., 1965, 112, 602. -, Electvochim. Acta, 1966, 11, 203. -, Ibid., 1966, 11, 311. -, Ibid., 1966, 11, 549. -, J. Electroanalyt. Chem., 1966, 12, 260. Hoare, J. P., Miebuhr, S. G., and Thacker, R., J . Electrochem. Soc., 1966, 113, 1078. Hickling, A., Trans. Faraday Soc., 1945, 41, 333. Kalish, T. V., and Burshtein, R. Kh., Dokl. Akad. hTauk SSSR, 1951, 81, 1093. -- , Ibid., 1953, 88, 863. Breiter, M. W., J . EZectroanalyt. Chem., 1964, 8, 230. Nagel, K., and Dietz, H., Electrochim. Acta, 1961, 4, 1. Margules, M., Wiedemanns AnnZn, 1899, 65, 629. Marie, C., C. R. Hebd. Se'anc. Acad. Sci., Paris, 1907, 145, 117. -, J . Chim. Phys., 1908, 6, 596. Ruer, R., 2. Elektrochem., 1908, 14, 309. -, Ibid., 1908, 14, 633. Watanabe, N., and Devanathan, M. A. V., J . Electrochem. SOL, 1964, 111, 615. El Wakkad, S. E. S., and El Din, A. M. S., J . Chem. Soc., 1954, 3098. Barnartt, S., J. Electrochem. Soc., 1959, 106, 722. Armstrong, G., and Butler, J. A. V., Trans. Faraday SOG., 1934, 30, 1173. Armstrong, G., Himsworth, F. R,, and Butler, J. A. V., PYOG. R. Soc., A , 1933, 143, 89. Hickling, A., Trans. Faraday Soc., 1946, 42, 518. -, Ibzd., 1965, 112, 849. -, Ibid., 1899, 66, 540. -, Ibid., 1908, 146, 475. Received January 17th, 1973 Accepted March ZOth, 1973