J. MATER. CHEM., 1994,4(8), 1283-1287 Hydrothermal Modification of Electrocatalytic and Corrosion Properties in Nanosize Particles of Ruthenium Dioxide Hydrate H. Neil McMurrayt Chemistry Department, University of Wales, Swansea, Singleton Park, Swansea, UK SA2 8PP Preformed hydrosols comprising 38 nm diameter particles of amorphous ruthenium dioxide hydrate are subject to hydrothermal modification at temperatures between 100 and 225 "C. Hydrothermally induced changes in sol composition have been characterised by transmission electron microscopy (TEM), UV-VIS spectrophotometry, thermogravimetric (TG) analysis and X-ray powder diffraction (XRD). Hydrothermally modified sols show substantial reductions in oxide hydration but show no evidence of increased crystallinity, Ostwald ripening, particle aggregation or agglomeration.Hydrothermally induced changes in electrocatalytic and corrosion behaviour have been characterized by using sols to catalyse the oxidation of water to oxygen by cerium(iv) ions, and measuring both evolved oxygen and ruthenium tetraoxide produced by oxidative corrosion of the catalyst. Sols hydrothermally modified at 3 200 "C show a 20-fold decrease in corrosion and a >5-fold decrease in electrocatalytic rate; these changes are attributed to increased oxygen bridging between surface Ru atoms resulting from hydrothermally induced condensation reactions between surface hydroxy groups. There is considerable contemporary interest in the technologi- cal potential of ceramic and composite materials prepared from particles in the nanometre size range (1-100 nm).'-' These 'nanophase' and 'nanocomposite' materials have been forecast to exert a major impact in the fields of catalysis, sensing, optics, electroceramics and structural ceramic mate- rial~.~,~,~However, there is significant difficulty in preparing and processing nanosize ceramic particles without the irrevers- ible formation of particle aggregates and agglomerates.' A method has recently been reported for the preparation of stable hydrosols of amorphous ruthenium dioxide hydrate (RuO,.xH,O) comprising substantially monodisperse, spheri- cal and pristine submicrometre particles6 In this paper a hydrothermal treatment is described which dramatically modi- fies hydration, electrocatalytic activity and susceptibility to oxidative corrosion in preformed RuO,.xH,O sols, without altering particle size or the state of particle dispersion.Both Ru02-xH20 and anhydrous Ru02 are technologically important ceramic metal oxides7 which have found extensive application as electronic conductors8 and electrocatalytic anode materials for the electrolytic evolution of chlorine and oxygen from aqueous solution^.^ Ru02.xH20 exhibits superior electrocatalytic properties for anodic oxygen evol- ution in terms of minimum overpotential and Tafel slope," but suffers the serious drawback of undergoing anodic cor- rosion to water soluble higher valence Ru It is known that anodic corrosion in RuO,.xH,O powder samples may be controlled by heating the powders in air (calcination) to limit the extent of hydration, but calcination results in particle coarsening and irreversible particle agglomeration through sintering and crystalline growth, thus reducing the effective surface area for electrocatalysis.10911 Electrocatalytic and corrosion performance in untreated and hydrothermally modified RuO,.xH,O sols were compared by exploiting the observation that individual particles in aqueous RuO,-xH,O dispersions behave as microelectrodes and as such will catalyse the oxidation of water by suitably oxidizing redox couples.'3-16 Sols were used to catalyse the oxidation of water by cerium@) ions and measurements were made of evolved oxygen and of ruthenium tetraoxide (RuO,) produced through anodic corrosion of the catalyst.The ability -f Current address: Department of Materials Engineering, University of Wales, Swansea, Singleton Park, Swansea. to compare electrocatalysis by RuO,-xH,O samples in a constant state of dispersion makes possible a more exact study of the influence of oxide hydration upon electrocatalytic activity. Experimental RuO, was prepared by a literature method;17 all other chemi- cals were supplied by BDH in AnalaR purity. The formation of RuO,*xH,O sol was initiated by adding 90 cm3 of mol dmA3 aqueous sodium nitrite (NaNO,) to 1.91 dm3 of 7.92 x mol dmP3 aqueous RuO,. Rapid partial reduc- tion of RuO, by NO, ions results in the precipitation of Ru0,-xH,O nuclei which then grow by catalysing the reduction of the remaining RuO, by water over ca.16 h. The kinetics and mechanism of sol development are discussed in detail elsewhere.6 Hydrothermal modification of the as-prepared Ru0,-xH,O sol at temperatures 2 125 "C was carried out using a Linsey and Baskerville high-pressure autoclave. Aliquots of sol (200cm3) were sealed in 300cm3 Pyrex glass bulbs, and the autoclave two-thirds filled with water to ensure equalization of pressures across the bulb wall. Treatments were carried out over 16 h, with ca. 10 h at the specified maximum temperature. Treatments at 75 and 100°C were conducted simply by placing sealed aliquots of sol in a temperature-controlled oven overnight. UV-VIS spectra of as-prepared and hydrothermally modi- fied Ru0,-xH,O sols were recorded using a Perkin-Elmer Lambda 3 spectrophotometer.TEM photographs of colloidal RuO,.xH,O particles were obtained using a JEOL 120C TEM-SCAN instrument.6 Solid samples of RuO,.xH,O for X-ray studies and TG were obtained by the coagulation of untreated and hydrother- mally modified sols using aqueous magnesium sulfate, fol- lowed by washing to remove salt and drying in air at 20°C. TG was performed using a Stanton Redcroft TG-750 instru- ment. Ru02.xH20 samples (typically 6 mg) were heated from ambient temperature to 900 "C at a rate of 20°C min-'in a flow of N, gas (40 cm3 min-I). Powder XRD patterns were recorded with Cu-Ka radiation using a Phillips PW 1720 generator and Debye-Scherrer camera.deb ye-Scherrer films were analysed using a CCD camera and Vilber Lourmat digital microdensitometry system. 1284 The catalysis by Ru02-xH20 sols of the oxidation of water by Ce" ions was studied using the atmospheric pressure gas- line apparatus shown in Fig. 1. The reaction vessel (A) was initially charged with 90 cm3 of sol and the RuO, trap (B) filled with 100 cm3 of 0.1 mol dm-3 sodium hypochlorite (NaOC1) in 1 mol dmP3 NaOH. Nitrogen was passed continu- ously through the gas line at 200 cm3 min-I. The test reaction was initiated by injection, via septum B, of 10 cm3 of 0.5 mol dmP3 (NH,),Ce(NO,), in 5mol dm-, HNO,. Evolved oxygen was measured directly using a Rank Brothers oxygen membrane polarographic detector (0,-MPD) (D).18 The volatile RuO, corrosion product was reduced to involatile perruthenate (RuO,-) in the RuO, trap." RuO,-was subse- quently determined spectrophotometrically (&298=2162 dm3 rno1-I cm-1).20 The detailed operation of the gas-line appar- atus is described elsewhere.12 All experiments were conducted at 20°C.Results The as-prepared RuO,.xH,O sol exhibited a monotonic decrease in UV-VIS absorbance with increasing wavelength which is typical of colloidal RuO~.XH,O.'~*'~ Visualized by TEM, sol particles appeared substantially spherical and monodisperse.6 The number-average mean particle diameter calculated from 200 TEM image diameter measurements was 38 +4 nm. Fig. 2 shows the TG weight loss curve for solid Ru02*xH20 samples derived from the as-prepared sol.The TG weight loss between ambient and 600°C is given in Table 1, no further weight loss was observed at temperatures Fig. 1 Schematic diagram of the atmospheric pressure gas-line apparatus used to evaluated the electrocatalytic performance of RuO,xH,O sols. A, Reactor (250 cm3 dreshel bottle); B, rubber septum; C, RuO, trap (125 cm3 dreshel bottle); D, 0, membrane polarographic detector; E, chart recorder. 2o th8 15-cn -0 E .-10-03 5-I I I I I I 25 100 200 300 400 500 600 TI'C Fig. 2 TG weight-loss curves for RuO,.xH,O samples derived from untreated and hydrothermally modified RuO,.xH,O sols. (a) Weight loss for untreated sol. Other curves show weight loss for sols hydrothermally modified at (b) 125, (c) 150, (d) 175, (e) 200 and (f) 225 "C.J. MATER. CHEM., 1994, VOL. 4 Table 1 TG and catalytic characteristics of as-prepared and hydro- thermally modified colloidal RuO, * xH,O hydrothermal TG wt. loss oxygen yield fractional T/"C 20-600 "C (%) ( mol) corrosion' as prepared 14.5 1.20 0.38 75 15.5 1.21 0.38 100 15.0 1.23 0.30 125 13.5 1.23 0.13 150 12.8 1.23 0.08 175 11.5 1.23 0.05 200 10.2 1.23 0.02 225 9.8 1.25 0.02 "Calculated as (mol RuO, evolved)/(mol RuO, * xH,O initial). >6OO"C. Similar samples were found to be essentially amorphous to X-rays. Two extremely weak and broad diffrac- tion lines were visible at positions consistent with reflections from the [211] and [loll crystal planes of tFtragona1 RuO,, which exhibit d-spacings of 1.685 and 2.55 A, but there wa,s no evidence of the intense [1lo] diffraction line at 3.17 A which would be expected for crystalline RuO,.The observed diffraction lines were insufficiently clear for the measurement of linewidths by microdensitometry. Hydrothermal treatment at temperatures up to 225 "C produced no significant changes in the UV-VIS absorbance-scattering spectrum of Ru02.xH20 sols, nor was there any increase in Tyndal effect, or any other indication of particle coarsening or loss of colloidal stability. Both the as-prepared and hydrothermally modified sols remained without any sign of coagulation or sedimentation for over six months. TEM investigation of the hydrothermally modified sols revealed no measurable changes in the size or shape of the RuO,-xH,O particles.Table 1 lists the total TG weight loss between ambient temperature and 600 "C for solid RuO,-xH,O samples derived from each hydrothermally modified sol, and Fig. 2 shows the corresponding weight-loss curves. Weight-loss curves for sols modified at 75 and 100°C essentially overlie the curve for as prepared sol and have been omitted from Fig. 2 for the sake of clarity. Hydrothermal modification produced no quantifiable changes in the XRD patterns obtained from sol-derived RuO,.xH,O. The diffraction lines observable in the as-prepared sol were unchanged by eye, and remained too indistinct for the estimation of linewidth by microdensitometry.No new diffraction lines appeared. When evaluated using the gas-line apparatus, all the sols catalysed the reduction of the test quantity of Ce" ions within 10 min (determined visually by the disappearance of intensely yellow CeIV), with the simultaneous evolution of ca. 30cm3 of 02.Fig. 3 shows the time-dependent 0,--MPD response curves for various sols, and Table 1 lists the molar 0, yields for each sol calculated from the area under the corresponding response curve. 0,-MPD response curves for sols hydrother- mally modified at 75-150 "C lie between the as-prepared and 175 "C curves shown in Fig. 3 and have been omitted for the sake of clarity. Table 1 also lists values for the molar fraction of RuO,-xH,O oxidatively corroded to RuO, in the course of the test reaction for each sol.Discussion Ru0,-xH,O is thermodynamically unstable with respect to anhydrous RuO,, which is a crystalline solid iso-morphous with r~tile.~ However, precipitation of amorphous RuO,.xH,O is typical when aqueous solutions of upper-valence Ru species are reduced, or when aqueous Ru" species are hydrolysed at room temperat~re.~The observation of structure, albeit very broad and indistinct, in the XRD patterns J. MATER. CHEM., 1994, VOL. 4 I I I 1 1 I I -1 0 0 10 20 T/min Fig.3 0,-MPD response curves showing O2 evolution profiles for water oxidation by Ce" ions catalysed by: (a)as-prepared Ru02.xH,0 sols, and sols hydrothermally modified at (b)175,(c) 200 and (d)225 "C obtained from as-prepared sol RuO,.xH,O suggests that small ordered regions do exist in this material.,' For the purposes of discussion these regions of short-range order will be referred to as crystallites.21 Particle Size and Dispersion The as-prepared RuO,-xH,O sols are intrinsically stable with respect to particle aggregation, although coagulation may be induced by relatively low concentrations of salt (ionic strength > mol dm-3).6 This behaviour is characteristic of electro- static stabilization, i.e.colloidal stability arising from repulsive interactions between the similarly charged electrical double layers of sol particles.22 Coagulation of RuO,xH,O hydrosols is known to result in dramatic reductions in optical extinction and absorbance- wavelength slopes in the UV-VIS absorbance-scattering spec- tra.14 Consequently, the observation the UV-VIS spectrum of the as-prepared RuO,*xH,O sol is practically unchanged by hydrothermal modification indicates that hydrothermal treat- ment up to 225 "C does not result in appreciable aggregation or coarsening of sol particles.This conclusion is corroborated by the absence of sedimentation or increased Tyndal effect in the hydrothermally modified sols. The unchanged TEM par- ticle size confirms the absence of Ostwald ripening in the hydrothermally modified sols. These findings are in sharp contradistinction to the case of RuO,.xH,O powders heated to ca. 200 "C in air, where interparticle sintering results in the formation of large (0.1-0.3 pm) agglomerates." Simply drying a fine aqueous metal oxide dispersion in air at room tempera- ture can lead to irreversible agglomeration of particles through 'pressure sintering', where the surface tension of the receding meniscus between particles promotes sintering at points of contact, and 'olation', where condensation reactions between surface hydroxy groups on neighbouring particles result in inter-particle oxygen bridging.23 It is assumed that the immun- ity of hydrothermally modified RuO,.xH,O sols to particle aggregation and agglomeration results from the electrostatic stability of these dispersions.Thus particles are maintained in mutual isolation, individually surrounded by water, through- out the hydrothermal treatment.Hydration and TG TG weight loss from RuO,.xH,O samples is predominantly due to thermally induced dehydration." For samples heated in air, irreversible dehydration commences at ca. 75 "C and becomes complete at ca. 600°C. Samples heated in air to 22200°C give powder XRD patterns Characteristic of the RuO, rutile structure." Exact identification of modes of water loss is difficult in IR-opaque RuO,.XH,O,~ but probable modes may be identified by analogy with IR-transparent amorphous oxide hydrates, such as tin@) oxide hydrate (SnO,.xH,O), which similarly develop the rutile structure on heating.24 Combined TG and IR studies of SnO,.xH,O have shown that weight lost reversibly at lower temperatures is due to physically adsorbed H,O whilst weight lost irrcversibly at higher temperatures is predominantly due to 'chernisorbed water', which is water produced by condensation 1 eactions occurring between hydroxy groups initially presen t at the oxide surface.24 Given the above, the TG weight loss data shown in Table 1 imply that solid RuO,.xH,O samples derived from the as-prepared sol contain ca.14% water by weight and that hydrothermal modification of sols at temperatures 3 125 "C results in a progressive reduction in water content, with Ru0,-xH,O derived from sols hydrothermally modified at 225 "C containing only 9.8% water. Studies of polycrystalline RuO, have shown that loss of physically adsorbed water is mostly complete at TG temperatures of llO"C.ll If the same is assumed to be true for RuO,-xH,O, Fig.2 implies that the ratio of chemisorbed to physically adsorbed water decreases with increasing hydrothermal modification temperature. Comparison of curves (a) and (f)in Fig. 2 show that hydro- thermal modification at 225 "C results in a 40% reduction in weight lost between 100 and 600°C (mostly chemisorbed water) but only a 25% reduction in weight lost between ambient temperature and 110 "C (mostly physically adsorbed water). One implication of this observation is that, if the quantity of physically adsorbed water is assumed to be proportional to surface area, hydrothermal treatment does not appear to reduce the specific surface area of sol derived Ru02-xH20 by more than 25%. XRD Behaviour Even when hydrothermally modified at 225"C, the Ru02-xH,0 sols fail to show any changes in their powder XRD pattern indicative of significant development of the RuO, rutile structure.This finding is in contradistinction to the case of calcined Ru02.xH,0 powders, for which clear and measurable diffraction lines corresponding to the [1101, [1011 and [211] RuO, crystal planes appear after calcination at temperatures >200 "C.ll The hydrothermal inhibition of crystallization was not anticipated, as it is that known hydro- thermal conditions frequently allow phase changes and crys- tallization in solid materials to take place at relatively low temperature^.,^ However, hydrothermally induced phase changes often result from a process of dissolution and repre- ~ipitation,,~and the solubility of RuO,-xH,O in neutral water is known to be very small.For freshly precipitated RuO ,*xH,O measurement of the equilibrium K. 4RuO2*2H2O,,, 6Ru4(OH);',+(,,, +4 OH;, , (3) gives K,M 5 x mo15 dm-5, implying a room-temperature solubility for Ru,(OH),,~+ of 5 x mol dm-3 at pH 7.7 Aged Ru02-xH20 precipitates show reduced solubility and none of the Ru0,-xH20 sols described herein showed any sign of room-temperature solubility even at pH 0. The insolu- bility of Ru0,-xH,O also explains the absence of Ostwald ripening in the hydrothermally modified sols. The loss of chemisorbed water from heated metal oxide hydrates occurs either as a result of condensation reactions between adjacent hydroxy groups at the surface of a single crystallite (two-dimensional condensation) or condensation reactions between hydroxy groups present at the surfaces of J.MATER. CHEM., 1994, VOL. 4 least five times as fast as that for the sol hydrothermally modified at 225 "C. The above observations closely parallel the findings of both macroelectrode" and dispersed microelectrode' studies using calcined RuO2-xH,O powders, which have shown that pro- gressive reduction of chemisorbed water content by calcination increases resistance to anodic corrosion but eventually neighbouring crystallites (three-dimensional c~ndensation).~~ produces Condensation in three dimensions implies an effective coalesc- ence of crystallites, and it is this process which eventually leads to long-range order and the appearance of XRD patterns characteristic of the anhydrous oxide.Given the observation that crystallinity does not appear in the hydrothermally modified Ru02.xH20, it would seem that hydrothermally induced dehydration is largely restricted to two-dimensional OH condensation processes. It is possible that the inhibition of three-dimensional condensation results from condensation reactions, generally, being thermodynamically disfavoured under hydrothermal conditions, in accordance with Le Chatalier's principle. Alternatively, it is possible that the intrusion of water between crystallites interferes with the approximation of surfaces necessary for three-dimensional condensation. Electrocatalytic Behaviour When RuO2-xH20 dispersions are mixed with aqueous solu- tions of redox couples, such as Cerv/Ce"', which are thermo- dynamically but not kinetically capable of oxidizing water, catalytic oxygen evolution may proceed as a result of RuO,.xH,O particles adopting a mixture potential such that the currents at the particle surface, due to electrochemical water oxidation and electrochemical CeIV reduction, are equal and opposite.13 The anodic corrosion of dispersed Ru02.xH20 has similarly been studied using the microelectrode approach.26 When exposed to CerV ions in 1mol dm-, HNO, the RuO,.xH,O sols may either catalyse the oxidation of water to 02 or undergo anodic corrosion to Ru04 The formal potential of the Ce'v/Ce''' couple in 1mol dm-, HNO, is 1.61 V us.NHE.27 Comparing this value with E' for the H,O/O, couple (1.229 V us. NHE at pH )28 and E' of the RuO,/RuO, couple (1.401 V us. NHE at pH 0)29shows that ample overpotential exists for both reactions (1) and (2) under the specified experimental conditions. It may be seen from Table 1that the fraction of RuO,.xH,O consumed through reaction (2) is dramatically influenced by hydrothermal modification of the Ru02-xH20 sols. Table 1 also shows that the yield of evolved O2 is substantially independent of hydrothermal sol modification, remaining practically constant at 96-100% of the maximum theoretical yield of reaction (1) (1.25 x lo-, mol 0,).However, Fig. 3 shows that the rate at which 0, is evolved in the test reaction decreases markedly for sols hydrothermally modified at tem- peratures > 175 "C.The dead volume of the gas-line apparatus (ca. 200cm3) makes the shape of the 0,-MPD response curves an imperfect measure of the rate of reaction (l),but measurement of the initial slopes of curves (a) and (d) in Fig. 3 suggests that initial 0, evolution for the unmodified sol is at loss of electrocatalytic activity. Calcination of RuO,.xH,O powders at 140°C has been shown to produce an X-ray amorphous material which will catalyse the oxi- dation of water by CeIv ions without measurable corrosion, but powders calcined at > 140 "C give reduced rates of 0, evolution." Formulation of an exact explanation for the increase in corrosion resistance observed in Ru02.xH20 hydrothermally modified at 2125°C is made difficult by the amorphous nature of the material.However, the formation of oxygen bridges between surface Ru atoms due to two-dimensional condensation between surface OH groups would certainly be expected to increase the energy with which individual Ru atoms are bound into crystallites, with the result that such atoms become more difficult to volatilize through anodic corrosion. Increased corrosion resistance in calcined Ru02.xH20 powders has similarly been attributed to a reduction in surface defects.', Decreased electrocatalytic activity in RuO,.xH,O powders calcined at temperatures >14O"C has been attributed to a decrease in effective surface area due to the crystallisation and sintering of particles." This seems entirely reasonable in the case of calcined powders but fails to explain the observed decrease in 0, evolution rates in RuO,.xH,O sols hydrother- mally modified at >175"C which do not exhibit crystalliz- ation, sintering or any change in their state of dispersion.One possible explanation is that hydrothermal treatment influences the surface concentration of electrocatalytically active sites. Electrocatalytic O2 evolution at Ru0,-xH,O anodes is believed to proceed uia the formation of covalent bonds between surface Ru atoms and water-derived 0 atom^.^ It follows that the surface Ru atom must lie in a lattice, edge or defect,,' position which permits the necessary changes in oxygen c~ordination.~~,~~ It is possible that extensive oxygen bridging of surface Ru atoms through two-dimensional con- densation of surface OH groups reduces the number of Ru atoms for which facile changes in oxygen coordination are feasible.Conclusions Hydrothermal treatment of electrostatically stabilized disper- sions of amorphous nanosize RuO,.xH,O particles at pro- gressively increasing temperatures between 100 and 225 "C results in progressive reduction in oxide hydration without significant changes in the crystallinity, size or dispersity of RuO,.xH,O particles. The predominant mode of oxide dehy- dration appears to be hydrothermally induced condensa- tion reactions occurring between hydroxy groups initially present at the oxide surface.Hydrothermal modification of RuO,.xH,O dispersions at temperatures 3 175 "C sharply reduces susceptibility of particles to oxidative (anodic) cor- rosion but simultaneously reduces electrocatalytic activity for anodic water oxidation. It is proposed that both of these changes in electrochemical properties result from increased oxygen bridging of surface Ru atoms. References 1 S. Komarneni, J. Mater. Chem., 1992,2, 1219. 2 G. W. Kriechbaum and P. Kleinshmit, Adv. Mater., 1989,28,1416. J. MATER. CHEM., 1994, VOL. 4 1287 3 4 5 E. Matijevic, Pure Appl. Chem., 1988,60, 1479. A. L. Robinson, Science, 1986,233,25. J. Karch, R. Birringer and H. Gleiter, Nature (London), 1987, 21 22 B. D. Cullity, in Elements of X-ray DifSraction, Addison Wesley, Mass., 2nd.edn., 1978, p. 321. D. J. Shaw, in Introduction to Colloid and Surface Chc.mistry, 330, 556. Butterworths, London, 3rd. edn., 1980, p. 183. 6 7 8 9 H. N. McMurray, J. Phys. Chem., 1993,97,8039. J. A. Rard, Chem. Rev., 1985,85, 1, and references therein. H. C. Angus and P. E. Gainsburg, Electronic Components, 1968, 84. S. Trasatti and G. Lodi, in Electrodes of Conducting Metal Oxides, 23 24 R. D. Nelson, in Handbook of Powder Technology, Dispersing Powders in Liquids, ed. J. C. Williams and T. Allen, Elsevier, Amsterdam, 1988, vol. 7, pp. 1-20. P. G. Harrison and A. Guest, J. Chem. SOC., Faraday Trans. I, 1987,83,3383. 10 ed. S. Trasatti, Elsevier, Amsterdam, 1981, pp. 521-626. C. Iwakura, K. Hirao and H. Tamura, Electrochim.Acta, 1977, 24 A. R. West, in Solid State Chemistry and its Applications, J. Wiley, Chichester, 1984, pp. 41-43. 22, 329; 335. 25 A. Mills and H. Davies, J. Chem. SOC., Faraday Trans. I, 1991, 11 A. Mills, S. Giddings, I. Patel and C. Lawrence, J. Chem. SOC., 87, 473. Faraday Trans. I, 1987,83,2331. 26 A. A. Noyes and C. S. Gardener, J. Am. Chem. SOC., 1936,58,1264. 12 A. Mills, S. Giddings and I. Patel, J. Chem. SOC., Faraday Trans. I, 1987,83,2317. 27 M. Pourbaix and N. de Zoubov, in Atlas of Electrochemical Equilibria in Aqueous Solutions, ed. M. Pourbaix, Pergamon 13 A. Mills and N. McMurray, J. Chem. SOC., Faraday Trans. 1,1989, 85,2047; 2055. 28 Press, Oxford, 1966, p. 343. M. Pourbaix and N. de Zoubov, in Atlas of Electrochemical 14 A. Mills and H. N. McMurray, J. Chem. SOC., Faraday Trans. I, 1988,84, 379. Equilibria in Aqueous Solutions, ed. M. Pourbaix, Pergamon Press, Oxford, 1966, p. 97. 15 17 18 19 20 A. Harriman and M.-C. Richoux, J. Chem. SOC., Faraday Trans. 1, 1987,83, 3001. R. E. Connick and C. R. Hurley, J. Am. Chem. Soc., 1952,74,5012. A. Mills and C. Lawrence, Analyst (London), 1984,109, 1549. R. P. Larson and L. E. Ross, Anal. Chem., 1959,31, 176. J. L. Woodhead and J. M. Fletcher, J. Chem. SOC., 1961,5039. 29 30 31 G. Lodi, E. Silvieri, A. De Battisti and S. Trasatti, J Appl.Electrochem., 1978,8, 1011. G. C. A. Schuit, Int. J. Quantum Chem., 1977,12,43. M. Boudart, Am. Sci., 1969,57,97. Paper 4/02267J;Received 18th Aprir',1994