J. CHEM. SOC. FARADAY TRANS., 1994, 90(5), 745-750 Hydrogen Evolution Reaction on Electrodes coated with Conducting-polymer Films Krzysztof Maksymiukt and Karl Doblhofer Fritz-Haber-lnstitut der Max-Planck-Gessellscha ft, Faradayweg 4-6,0-14195 Berlin, Germany The cathodic hydrogen evolution reaction has been studied on gold electrodes coated with poly(N-methyl- pyrrole) (PMPy), and with a polymer mixture of PMPy and poly(4-styrenesulfonate) (PMPy-PSS). In all cases, Hf ions were found to permeate the film and react at the metal-electrode surface. In particular, for low concentra- tions of HCI, the PMPy film hinders significantly the movement of the depolarizer to the electrode. On the other hand, in the case of PMPy-PSS the ti+ transport across the film proceeds readily, because the H+ ions consti- tute counter-ions to the immobilized sulfonate groups, present at a concentration of ca.1 mol dm-3. The effect of the partitioning equilibria and the Donnan potential on the observed potential dependence of the charge-transfer rate is discussed in detail. It is shown that the presence of the cation-exchanger coating (PMPy-PSS) enhances the rate of electrochemical H+ reduction (relative to the uncoated electrode) when the concentration of H+ in the electrolyte is small. The diffusion coefficient of H+ ions in the PMPy-PSS matrix was determined: D = 1 x cm2 s-'. Oxidation/reduction reactions of redox species from electro- lytes on electrodes coated with electron-conducting polymer films are a subject of considerable interest from the stand- point of electrocatalysis.'-'o Such reactions may proceed by two fundamentally different mechanisms.First, the polymer may act as a mediator for electron transfer between the elec- trode and the depolarizer in the electrolyte; in this case, the reduction of Ox from solution may be formulated: krcd Ox + Poly -Red + Poly+ (1) where kred is the (bimolecular) rate constant, Poly and Poly+ are the reduced and oxidized polymer sites, and Red is the reduced form of the depolarizer in the electrolyte. This mechanism has been found to be operative in many cases, whereby the electron transfer usually proceeds at the polymer surface;' '-I6 nevertheless, a penetration of the redox species into the matrix is also possible.",'* In this case, the electron- transfer reaction proceeds in a three-dimensional reaction zone; the cathodic current may be described by:'5*19 i-= -nFAk,,, ~cply(x)cox(x)dx (1) where d is the polymer film thickness, A the electrode surface area, cpoly(x)and c,,(x) are the concentrations of polymeric sites and of Ox in the film.In our recent paper,I5 we have studied reduction reactions of selected inorganic redox oncouples: Fe(CN)2-'4-, Ru(NH3)2+I2+ and Eu~+/~+ a rotating-disk electrode (RDE) proceeding on the polymer surface. The limiting reduction currents were found to be lower than in the absence of the polymer, which was caused by a slow rate of the mediated reaction [eqn. (l)], as obtained from Levich-Koutecky plots.Much less attention has been given to the second mecha- nism, in which the depolarizer penetrates the polymer film without reacting with it. The electron-transfer reaction may proceed on the metallic substrate, i.e. at the electrode/ polymer interface. An example is the reduction of H+ ions, as found by Bard and co-~orkers'~and Schultze and co-workers20.21 for typical conducting polymers, such as poly- pyrrole and polyaniline. Clearly, in this case the depolarizer is partitioned from the electrolyte solution into the polymer phase, and transported across the polymer phase. Therefore, t On leave from the Department of Chemistry, Warsaw Uni-versity, Pasteura 1, PL-02-093 Warszawa, Poland. the membrane properties of the polymer will exert a decisive influence on the rate of the charge-transfer reaction.It is the aim of this work to analyse the kinetics of such a charge-transfer reaction as a function of the membrane properties of conducting-polymer coatings. Thus, we present and discuss results concerning the H+ reduction on elec-trodes coated with poly(N-methylpyrrole) (PMPy), and with poly(N-methylpyrrole) containing immobilized poly(4-styre- nesulfonate) (PMPy-PSS). PMPy is an anion exchanger in the oxidized state. PMPy-PSS constitutes a cation-exchanger matri~,~~,~~where cations from the electrolyte along with the produced PMPy sites compensate the negative charge of + the -SO, groups in the matrix. Experimental Electrochemical measurements have been performed using a two-compartment cell with working gold or platinum elec- trodes of surface area 0.28 cm2, a saturated calomel reference electrode (SCE) and a counter electrode prepared from the same metal as the working electrode.The working electrodes were polished using diamond paste. The platinum electrode surface was additionally activated by cyclic voltammetry in 0.5 mol dmP3 H2S04. The polymer films were deposited on the metal electrodes potentiostatically, at 0.7 V, from aqueous solutions containing 0.05 mol dm-N-methylpyrrole (distilled and kept cool in an argon atmosphere) and 0.1 mol dm-3 NaClO, or 0.1 mol dm-3 sodium poly(4-styrenesul- fonate). The chemicals (p.a. products) were used as received. Water was triply distilled.For other details see previous paper^.^ 3924 The following experiment was conducted to estimate the concentration of sulfonate groups in the films. First, the PMPy-PSS-coated electrode was polarized in HCl electro- lyte for 10 min at -0.45 0s. SCE [step (l)]. At this potential the polymer is completely reduced and the concentration of H+ in the polymer phase is approximately equal to the con- centration of sulfonate groups (see below). Then, the coated electrode was rinsed with distilled water, transferred into a 0.1 mol dm-3 KCl electrolyte, and polarized to -0.8 V us. SCE [step (2)]. In step (2), the H+ ions are reduced to H,, and their place is occupied by K+ ions from the electrolyte. The charge measured in this step, Q(H+@'IY)) was assumed to correspond to hydrogen ion discharge. This value was quite reproducible (f10%)when it was measured not later than 30 s after immersion into the KCl solution.For longer times 746 I I I -6-0 E--.h---0:4-3 u-2-0 50 100 150 200 polymerization charge/mC Fig. 1 Charge required for complete cathodic reduction of the H+ present in polymer coatings, Q(H+(vlY)).Polymer films : PMPy-PSS of thickness defined by the polymerization charge (73 mC correspond to ca. 1 pmZ2). The coating electrodes were equilibrated in 0.1 mol dmV3 HCI; the reduction charge was determined in 0.1 mol dm-3 KCl. a gradual decrease in measured charge was observed, appar- ently because of exchange of H+ ions by K+from the solu- tion.As expected, Q(H+(PlY)) was found to change little when the HCl concentration in solution was varied (in the range 0.02-1 mol dm-3), and it was proportional to the film thick- ness (Fig. 1). From this charge the concentration of bound sulfonate groups in the film was evaluated to be about 1 mol dm-3, assuming that a polymerization charge of 26.2 mC em-’corresponds to the film thickness, 0.1 pm. This value .~~was found by Zhou et ~1 for PMPy produced under the same conditions. Results and Discussion Electron-transfer Reaction proceeding at the Metal/Polymer Interface The cyclic voltammograms obtained with the electrodes coat- ed with the polymers PMPy and PMPy-PSS in 0.1 mol dm-3 KCl are represented in Fig. 2.Note that the anodic oxidation of the polymer (PMPy -+ PMPy’) commences at about 0.0 V us. SCE in the case of PMPy and at about -0.2 V us. SCE with PMPy-PSS. Both these oxidation potentials are more positive than the standard potential of the H+/H2 reaction (-0.242 V us. SCE). Since the maximum concentra- tion of H+ used in this work was 1 mol drnp3, one concludes that even thermodynamically the mediated reaction H + + PMPy +*H2 + PMPy+ is not favoured, i.e. it cannot proceed at a significant rate. To support this conclusion, the cathodic H+ reduction was conducted on a platinum and a gold electrode. Both elec- trodes were (Q) coated with identical PMPy-PSS films, and (b) uncoated. The results are presented in Fig. 3. Clearly, if the H+ reduction were mediated by PMPy sites, the reaction would proceed on both coated electrodes at the same elec- trode potential, but this is not the case.In fact, the larger overvoltage of the coated gold electrode is an experimental proof that the charge-transfer reaction, at least in the case of gold, proceeds indeed at the metal surface. Hydrogen Reaction on PMPy-coated Electrodes In Fig. 4, H+ reduction currents on Au/PMPy electrodes are compared with the corresponding results on uncoated Au J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 T 10.5 mA IIIIIIIII -0.2 0 0.2 0.4 E/Vvs. (SCE) Fig. 2 Cyclic voltammetric curves of gold electrodes (area, 0.28 cm2)covered with the ‘conducting polymers’ PMPy (a) and PMPy- PSS (b), in 0.1 mol dm-3 KCI.Scan rate, 0.1 V s-’; charge used for polymerization, 90 mC. electrodes. The measured currents do not depend on the rota- tion speed, both at uncoated and coated electrodes, indicat- ing that diffusion/convection in solution is not the rate- determining step. The reduction of H+ proceeds in the poten- tial range where the polymer is completely reduced and constitutes a neutral organic phase. Electroneutrality coup- ling requires that both H+ and C1-, i.e. HC1, enter into the polymer. The partitioning equilibrium may be described by : where cr$ and c&, are the‘ equilibrium concentrations of HCl in the polymer and in the electrolyte, respectively, and I I0-NI5 -0.2 -aE2 >r C.--$ -0.4 -0 c. 2 3 t J.CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 NI5 -0.2 I-1.0 I -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 electrode potential, €/V vs. SCE Fig. 4 H+ reduction current observed with Au electrodes coated with PMPy (---) in (a)1, (b) 0.1 and (c) 0.02 mol dm-j HCl. Rota- tion speed, 900 rpm; scan rate, 1 mV s-’; charge used for poly- merization, 90 mC for the electrode of area 0.28 cm2. For comparison, the corresponding currents obtained with two HCl con- centrations on the uncoated electrodes are included (-). is the difference in standard chemical potential in the two phases, i.e. the partial molar Gibbs energy of tran~fer.’~ In aqueous electrolytes the polymer (in the reduced state) is unswollen and constitutes a phase of low relative permit- tivity.Considering that the HCl molecule is rather polar, one would expect Apicl to have a large positive value. Further- more, the degree of dissociation of the HCl partitioned in the polymer phase should be small. Although it is difficult to make a quantitative prediction, one expects the concentration of free H+ in the polymer to be significantly smaller than in the electrolyte. The situation is represented schematically in Fig. 5. Comparing the theoretical considerations with the results of Fig. 4, it appears that the result obtained with 0.02 mol dmP3 HCl is plausible. The stationary flux of H+ across the polymer film is defined by the diffusion coefficient and the concentration gradient of H (cpjY/d),which are both small + and lead to a reduction current which is not measurable on the current scale of Fig.4. C 4‘ I 0 d Fig. 5 Schematic representation of the partitioning equilibrium of H+ (HCl) between the electrolyte and polymer (PMPy) 747 Unexpectedly, at the larger HCl concentrations the H+ reduction currents rise by factors much larger than the rise in HCl concentration in the electrolyte (x 5, x 10). The current plateau at 0.1 mol dmP3 HCl indicates that the current is still limited by the rate of H+ transport across the film. One con- cludes that the equilibrium concentration of H+ in the polymer rises more than proportional to the HCl concentra- tion in the electrolyte, i.e. the value of A&,-, decreases with increasing HCl concentration.The reason for this may be the beginning of protonation of the N-methylpyrrole units (the pK, value of N-methylpyrrole is -2.926). This would have an autocatalytic effect on the H transport, because of swelling + of the polymer associated with the solvation of the produced ions. Hydrogen Reaction on PMPy-PSS-coated electrodes The membrane state of a polymer containing a significant concentration of fixed charges is readily discussed in a quan- titative way. The -SO, groups in the considered polymer PMPy-PSS require cations as counter-ions. As a conse-quence of the large ion concentration, the polymer is highly swollen, and it constitutes a polyelectrolyte gel. Across the interface between this solvated polymer and a liquid electro- lyte, the well known ‘Donnan equilibrium’ defines the par- titioning of the exchangeable ions, in particular of H + : where A#D = -#’ is the ‘Donnan potential’, i.e.the interfacial potential difference between the polymer and the electrolyte. Consider the situation in which the electrolyte contains only HCl. The electroneutrality condition for the polymer phase may then be formulated: ct;.,Iy = [x-]+ cg!y (4) where c are the concentrations of the species indicated as subscripts, and [X-] is the fixed-anion concentration. Com- bining eqn. (3) and (4) under the assumption that the concen- trations equal the activities, u, and [X-] = 1 mol dm-3, one obtains the cp?, czly and A&, results summarized in Fig. 6. Note that, over a considerable electrolyte concentration range, the concentration of the considered depolarizer H + in > E..aU -0 --50 --100 -2 -1 0 1 log[cS/moldw3] Fig. 6 Equilibrium concentrations of mobile cations (cEjy)(a) and anions (CE!~)(b) in an anion exchanger of fixed-charge density, [X-]= I rnol dm-3, as a function of the electrolyte (HCl) concentra- tion, cs. (c) A&,, the Donnan potential, as a function of the electro- lyte concentration. the polymer does not depend in a significant way on its con- centration in the electrolyte, i.e. crjyx constant x [X-]. This has a remarkable consequence for the dependence of the CT rate on the H+ concentration in the electrolyte. Consider the distribution of the electric potential across the uncoated and coated electrode at a certain value of -4’.This corresponds to a particular value of the elec- trode potential, E, at which cathodic H reduction proceeds + (Fig. 7).Two concentrations of H+ in the electrolyte are con- sidered, ‘1’ and ‘2’. In the case of the uncoated electrode [Fig. 7(a)] the dependence of the rate of H, production on 4, is a consequence of the linear dependence of the rate of the electron-transfer process on the depolarizer concentra- tion. It may be formulated, disregarding double-layer effects : i-= -nFAk-cL+ exp( -(54g) i-= -nFAk-c;, exp[ -RT where i-is the reduction current (at the considered over- voltages, the back reaction may be disregarded); k-is the cathodic CT rate constant; a is the cathodic CT coefficient; and 4 is the electric potential of the phase defined by the superscript.The logarithm of the current depends linearly on E, with a slope of (59/a) mV decade-’. The situation is rep- resented schematically in Fig. 8, where c1 is a reference con- centration and a = *. In the presence of the film [Fig. 7(b)]the CT rate is defined by the concentration of H+ in the polymer and the potential drop across the metal/polymer interface. Assuming the same electrochemical rate constant as in the case of the uncoated electrode (see Section 4.3.2 of ref. 25 for a discussion of the validity of this assumption), one may write: J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 3 Fig. 8 Illustration of the difference between a fixed-charge polymer coated and an uncoated electrode, with respect to the charge-transfer current density, I, as a function of the electrode potential, E (relative scale), and the depolarizer concentration in the electrolyte (relative to a reference concentration cl): (a) cl, (b) 10cl, (c) lOOc,.The arrows point to the change in I corresponding to the reduction of the depo- larizer concentration by a factor of 10 (lOOcl -+lOc,). In Fig. 9 experimental results are summarized, which support the above analysis. At a given current density, the cathodic overvoltage increases by ca. 120 mV decade-’ reduction of H+ concentration in the case of the uncoated gold electrode. On the other hand, in the presence of the PMPy-PSS coating the overvoltage increases only by about 59 mV decade-l, as expected (Fig.8). It was found that the experimental results obtained with the coated electrodes are much more reproducible than the ‘clean’-gold results. The strong dependence of the rate of hydrogen evolution on the slightest variations of the surface state of gold electrodes is well known, see for instance ref. 27. The above result has a remarkable consequence. In the case of dilute ionic depolarizers the rate of the electrode reac- i-= -nFAk-cg!Yexp (6b) tion may be enhanced, at constant overvoltage, by coating [-aF(ERiA4D)]the electrode with an ion exchanger for which the depolarizer Since the effective depolarizer concentration, ~fi’!~, ions constitute counter-ions. remains largely constant, a change of the electrolyte concentration will affect the charge-transfer rate via the Donnan potential, eqn.qb). For example, when at constant applied E the elec- trolyte concentration is reduced from lOOc, to lOc, (see Fig. 8), the CT rate will be reduced only by a factor of about three, compared with a factor of 10 in absence of the fixed- charge polymer film. (a1 (b 1 Me ;I electrolyte Me ’/I poly [electrolyte Fig. 7 Distribution of the electric potential, 4, across the interfaces between electrolyte (S) and an uncoated (a) and polymer-coated (b) electrode of the same metal, Me, at a certain value of -4’. Two concentrations of H+ in the electrolyte are considered, ‘1’ and ‘2’. Ad,, is the (negative) Donnan potential. 0-NI E” -0.2 -a E.a -.-5 4.4In Q)U -E -0.6 g3 -0.8 --1.0 -I I I I I -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 electrode potential, €/V vs.SCE Fig. 9 H+ reduction currents on rotated uncoated (-) and PMPy-PSS-coated (---) gold electrodes, as a function of electrolyte (HCl) concentration: (a)0.02, (b) 0.1 and (c) 1 mol dm-’ HCl. Rota- tion speed, 900 rpm; scan rate, 1 mV s-’; charge used for poly- merization, 90 mC for an electrode of area 0.28 an2. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 Note that the above consideration relates to the kinetics of the CT reaction. The equilibrium potential of the redox system is the same in the presence and absence of the polymer film:25 When equilibrium prevails across the inter- face between metal and solvated polymer, one may formulate Using the definition of the electrochemical potential, ii, one obtains : where K is a constant.The equilibrium condition across the polymer/electrolyte interface may be formulated as : Combining eqn. (7) and (9), one sees immediately that in the presence and absence of the coating the same electrochemical equilibrium prevails : Thus, the same equilibrium potential is expected in the pres- ence and absence of the coating. Diffusion Coefficient of H+ Ions in the PMPy-PSS Film At larger overpotentials, the rate-determining step in the H+-reduction reaction is eventually the transport of hydrogen ions towards the electrode surface. Fig. 10 presents two typical current-potential curves on rotated disk electrodes obtained for reduction of 4 mmol dmA3 H+ (HCl).The sup- porting electrolyte (1 mol dm-3 KCl) should minimize the effect of migration. The limiting current obtained for the Au/ PMPy-PSS electrode is distinctly lower than the current on the bare gold electrode. The Levich-Koutecky analysis of I I ‘I 0.6 1 cy 0.4 r I a E.-. t I I 1 0 0.02 0.04 0.06 1/w’/2 Fig. 11 Levich-Koutecky plots of limiting currents, obtained from experiments as represented in Fig. 10. w is the electrode rotation rate in rpm. (0)Au/PMb-PSS, (V) Au. such results is given in Fig. 11. It demonstrates that the process at the bare metal electrode is controlled by diffusion in the solution, as opposed to the situation in the presence of the polymer film.The results of such analyses show that i, (currents from Levich-Koutecky plots, extrapolated to 1/ 01/2-,0) is inversely proportional to the film thickness (Fig. 12; the film thickness is expressed in terms of the charge used for polymerization). The results indicate that in the case of the coated elec- trodes i, defines the rate of diffusion of H+ ions across the film. This process may be described1g2 by the formula: FADi, = -cP$?d where A is the electrode area and D is the diffusion coefficient of H in the polymer. Using eqn. (1l), the diffusion coefficient + of H+ ions in the polymer film may be determined. To do so, the polymerization charge, 26.2 mC cm-2, is again taken to correspond to a film thickness d = 0.1 pm,” and the ion- 0.15 I -10 :0 -1.2 -1.0 -0.8 -0.6 50 100 150 200 electrode potential, €/V vs.SCE polymerisation charge/mC Fig. 10 Current-potential curves obtained on (a) an uncoated and Fig. 12 Dependence of i, (obtained from Levich-Koutecky plots) (b) a PMPy-PSS-coated rotating-disk (gold) electrode, in 0.02 mol on PMPy-PSS-film thickness (in mC used for polymerization; 73 dm-3 HCI. Polymerization charge, 90 mC for an electrode of area mC correspond to ca. 1 pm2’).Electrolyte: aqueous solution of 20 0.28 an2;rotation speed, 900 rpm; scan rate, 1 mV s-’. mmol dm-3 HCl-1 mol dm-3 KCI. 750 exchange equilibrium constant between H+ and K+ is assumed to be l.28*29The value of the diffusion coefficient determined from the slope of the line in Fig.12 is D = 1 x cm2 s-’,i.e. lower than in aqueous solution by a factor of about 102.30 Conclusions The cathodic hydrogen evolution reaction was studied on gold electrodes coated with poly(N-methylpyrrole) (PMPy), and with a polymer mixture of PMPy and poly(Cstyrenesu1- fonate) (PMPy-PSS). In all cases, the H+ ions were found to permeate the film and react at the metal-electrode surface. In particular, for low concentrations of HCl, the PMPy film hinders significantly the movement of the depolarizer to the electrode. As the HCl concentration in the electrolyte is increased, the permeability of the PMPy film rises consider- ably, probably because of the commencement of protonation and swelling of the PMPy matrix.In the case of PMPy-PSS, H+ transport across the film proceeds readily, because the H+ ions constitute counter-ions to the immobilized sulfonate groups, present at a concentration of ca. 1 mol dm-3. The diffusion coefficient of H+ ions in the PMPy-PSS matrix was found to be D = 1 x cm2 s-’. The effect of the partitioning equilibria and the Donnan potential on the observed potential dependence of the charge- transfer rate is discussed in detail. It is shown that the pres- ence of the cation-exchanger coating (PMPy-PSS) enhances the rate of electrochemical H+ reduction (relative to the uncoated electrode) when the concentration of H+ in the electrolyte is small compared with the fixed-charge concen- tration.The basic reason for this is the fact that the H+ con- centration in this polymer, which is effective in the electron-transfer reaction, is almost unaffected by changes of the H+ concentration in the electrolyte (see Fig. 6). Thus, a change of the electrolyte concentration (at constant E) will affect the charge-transfer rate only uia the Donnan potential. For example, when at constant applied E the electrolyte con- centration is reduced from lOOc, to lOc, (see Fig. 8), the CT rate will be reduced by a factor of about three, compared with a factor of 10 in the absence of the fixed-charge polymer film. K.M. is grateful to Alexander von Humboldt-Stiftung for financial support of his research stay in Germany. References (a) W.J. Albery and A. R. Hillman, Annu. Rep. Prog. Chem., Sect. C, 1981, 78, 377; (b) A. R. Hillman, in Electrochemical Science and Technology of Polymers, Elsevier, London, 1987,vol. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 1, pp. 103-292;(c) W. J. Albery and A. R. Hillman, J. Electro-anal. Chem., 1984,170,27. 2 C. P.Andrieux, J-M. Dumas-Bouchiat and J-M. Savant, J. Elec-troanal. Chem., 1982,131,l; 1982,142,l; 1984,169,9. 3 D.L. Miller and J. O’M. Bockris, J. Electrochem. SOC., 1992, 139, 967. 4 A. Deronzier and J-C. Moutet, Acc. Chem. Res., 1989,22,249. 5 G. Bidan, E. M. Genies and M. Lapkowski, J. Chem. SOC.,Chem. Commun., 1988,533. 6 M. Fabrizio, G.Mengoli, M. M. Musiani and F. Paolucci, J. Electroanal. Chem., 1991,300, 23.7 J. Desilvestro and 0.Haas, Electrochim. Acta, 1991,36, 361. 8 F. Mitzutani, S. Iijima, Y. Tanabe and K. Tsuda, J. Chem. SOC., Chem. Commun., 1983,148,299. 9 (a) T. Ikeda, C. R. Leidner and R. W. Murray, J. Electroanal. Chem., 1982, 138, 343; (b) C.R. Leidner and R. W. Murray, J. Am. Chem. SOC., 1984, 106, 1606; (c) B. J. Feldman, A. G. Ewing and R. W. Murray, J. Electroanal. Chem., 1985, 194,63. 10 Y.Xie and F. C. Anson, J. Electroanal. Chem., 1993,344,405. 11 N. Oyama, Y. Ohnuki, K. Chiba and T. Ohsaka, Chem. Lett., 1983,1759. 12 V. E. Kazarinov, M. D. Levi, A. M. Skundin and M. A. Vorotyntsev, J. Electroanal. Chem., 1989,271, 193. 13 M. M. Lohrengel, J. W. Schultze and A. Thyssen, in Elektrisch leitende Kunststoffe, ed.H. J. Mair and S. Roth, Hanser Verlag, Munchen, 1989,p. 377. 14 M.D.Levi and E. Yu. Pisarevskaya, Synth. Met., 1991,45,309. 15 K. Maksymiuk and K. Doblhofer, Electrochim. Acta, 1994, 39, 217. 16 K. Maksymiuk and K. Doblhofer, Synth. Met., 1993,55, 1382. 17 R. A. Bull, F-R. F. Fan and A. J. Bard, J. Electrochem. SOC., 1982,129,1009. 18 R. C. M. Jakobs, L. J. J. Janssen and E. Barendrecht, Electro-chim. Acta, 1985,30, 1085. 19 K. Doblhofer, J. Electroanal. Chem., 1992,331, 1015. 20 B. Pfeiffer, A. Thyssen and J. W. Schultze, J. Electroanal. Chem., 1989,260,393. 21 D.Stockert, M. M. Lohrengel and J. W. Schultze, Synth. Met., 1993,55,1323. 22 Q-X. Zhou, L. L. Miller and J. R. Valentine, J. Electroanal. Chem., 1989,261,147. 23 C. Zhong, K. Doblhofer and G. Weinberg, Faraday Discuss. Chem. SOC., 1989,88,307. 24 C. Zhong and K. Doblhofer, Electrochim. Acta, 1990,35, 1971. 25 K. Doblhofer, in Electrochemistry of Novel Materials, ed. J. Lip- kowski and P. N. Ross, VCH, New York, in the press. 26 Y. Chiang and E.B. Whipple, J. Am. Chem. Soc., 1963,852763; H.Beyer and W. Walter, Lehrbuch der Organischen Chemie, S. Hirzel Verlag, Stuttgart, 21st edn., 1988,p. 708. 27 K.Gossner and C. Lomer, 2.Phys. Chem., Neue Folge, 1963,37, 115. 28 T. R. E. Kressman and J. A. Kitchener, J. Chem. SOC., 1949, 1190; 1201; 1208. 29 S.Peterson, Ann. N. Y. Acad. Sci., 1953,57, 144. 30 G.J. Brug, H. Sluyters-Rehbach, J. H. Sluyters and A. Hamelin, J. Electroanal. Chem., 1984,181,245. Paper 3/05961H; Received 5th October, 1993