Kinetic study of CO oxidation on step decorated Pt(1 1 1) vicinal single crystal electrodes
Qing-Song Chen,Juan Miguel Feliú,Antonio Berná,Víctor Íctor Climent,Shi-Gang Sun
DOI: https://doi.org/10.1016/j.electacta.2011.04.101
IF: 6.6
2011-01-01
Electrochimica Acta
Abstract:In this work, surface modification at atomic level was applied to study the reactivity of step sites on platinum single crystal surfaces. Stepped platinum single crystal electrodes with (1 1 1) terraces separated by monoatomic step sites with different symmetry were decorated with irreversibly adsorbed adatoms, without blocking the terrace sites, and characterized in 0.1 M HClO 4 solution. The kinetics of CO oxidation on the different platinum single crystal planes as well as on the step decorated surfaces has been studied using chronoamperometry. The apparent rate constants, which were determined by fitting the experimental data to a mean-field model, decrease after the steps of platinum single crystal electrodes have been blocked by the adatoms. This behavior indicates that steps are active sites for CO oxidation. Tafel slopes measured from the potential dependence of the apparent rate constants of CO oxidation were similar in all cases. This result demonstrates that the electrochemical oxidation of the CO adlayer on all the surfaces follows the same Langmuir–Hinshelwood model, irrespectively of step modification. Keywords Pt single crystal electrodes Stepped surfaces Adatom step decoration CO oxidation kinetics Langmuir–Hinshelwood model 1 Introduction In the past few decades, a large number of investigations related to electrocatalytic reactions have been carried out on well-defined single crystal electrodes and nanoparticles for both fundamental and practical interests. It has been demonstrated that most of these reactions are structure sensitive, and pointed out that atoms on step/defect sites are generally more reactive than those on flat surfaces due to their relative lower coordination and altered electronic structure [1–5] . Among the numerous surface sensitive electrocatalytic reactions, CO oxidation is one of the most intensively studied processes due to the simple molecular structure of CO and to the facility to form adsorbed adlayers and to characterize them by voltammetry and spectroscopy. After hydrogen, CO adsorption and stripping can be considered as the most important model species in surface electrochemistry and, because it always participates as a strong poison intermediate formed during the dissociation of small organic molecules, plays a key role in electrocatalysis. Therefore, it is of great importance to disentangle the details related to the mechanism of CO adsorption and oxidation. To achieve this, a great deal of effort both experimental and theoretical, has been devoted to study the mechanism of CO oxidation on different surface structure Pt single crystal planes. This knowledge can be used to understand the behavior of bulk polycrystalline Pt as well as dispersed nanoparticles on a suitable substrate [6–15] . Up to now, the mechanism proposed firstly by Gilman for the electrochemical oxidation of CO on Pt through a “reactant-pair” mechanism [16] , has been widely accepted in most publications. This mechanism assumes a Langmuir–Hinshelwood type reaction between the adsorbed CO and the adjacent adsorbed oxygen containing species, with the following scheme: (1) H 2 O + * ⇌ OH ads + H + + e − (2) CO ads + OH ads → COOH ads (3) COOH ads → CO 2 + H + + e − + 2* where * denotes a free surface site. The faradaic processes in Eqs. (1) and (3) are fast steps while the chemical process in Eq. (2) is considered to be the rate-determining step of the ongoing reaction. From the theoretical point of view, nucleation and growth [17,18] , Dynamic Monte Carlo simulations and mean-field approximations have been applied to simulate the experimental data [19–22] . The first model is based on the assumption that the reactants (i.e., CO) are immobile on the surface and the adsorption of OH proceeds via “nucleation” at certain places. The reaction would only take place at the interface between the two reacting phases, causing the formation and growth of islands. Conversely, the mean field model considers CO diffusion on the surface much faster than its oxidation so that CO and OH can be considered completely mixed at every stage of the reaction. According to the studies of Lebedeva et al. [21,22] , the experimental data of CO oxidation on Pt(s)[( n − 1)(1 1 1) × (1 1 0)] surfaces can be satisfactorily fitted by a mean-field approximation model, while fits to a nucleation and growth model are always inferior. The mean-field approximation has also been successfully extended to fit the current transients of CO oxidation on Pt(1 0 0), Pt(5 3 3), Pt(s)[( n − 1)(1 0 0) × (1 1 0)] and even dispersed Pt nanoparticle electrodes [23–25] . Irreversible adsorption is one of the most convenient and frequently employed techniques for the preparation of bimetallic surfaces. The irreversibly adsorbed adatoms can modify the electronic and chemical properties of the substrate and, as a consequence, improve its reactivity/selectivity for a given reaction [26–29] . For well-defined stepped surfaces, irreversibly adsorbed adatoms can be applied as atomic probes to monitor the surface structure and to obtain information on the contribution of the different terrace and step sites [30–32] . With the help of continuous recording of cyclic voltammograms, the deposition of some adatoms from a highly diluted solution can be controlled to only decorate the step sites without blocking the terrace sites [33] . This carefully controlled surface modification technique allows to discriminate the different role played by step and terrace sites in the overall electrocatalytic reaction, and to get more insight into the effect of the surface structure. In the current work, stepped surfaces of different terrace length and step symmetry were employed. Chronoamperometry was applied to study the kinetics of CO electrochemical oxidation on these surfaces, either clean or step decorated by Bi and Te. The aim is to identify the particular reactivity of the monoatomic step sites separating (1 1 1) terraces, and to provide quantitative kinetic parameters of the reaction on step decorated surfaces. 2 Experimental Hemispherical platinum single crystal electrodes, geometric area ca. 3–4 mm 2 , were prepared according to Clavilier's method [34] . Two types of crystals vicinal to Pt(1 1 1) having close-packed terrace sites separated by monoatomic step sites of different symmetry were used in this study. Pt(5 5 4) and Pt(2 2 1) with 9 and 3 atom-wide terraces, respectively, both possess (1 1 0) step sites. Pt(5 4 4) and Pt(5 3 3) have also the same terrace symmetry with 9 and 4 atom width, respectively, although the step sites have (1 0 0) symmetry. Prior to each experiment the electrodes were flame annealed and cooled down to room temperature in a H 2 + Ar atmosphere. After the flame annealing, the single-crystal electrodes were quenched with water in equilibrium with this mixture of gases and then transferred to the cell under the protection of a droplet of deoxygenated water. The experiments were carried out in two conventional three electrode glass cells: one is used for deposition of adatoms and the other one, with an additional inlet for dosing CO, was used for electrochemical characterization, CO adsorption and oxidation. The procedure for selective adatom deposition on step sites on platinum single crystal stepped surfaces has also been described elsewhere [33] . After recording the characteristic voltammogram of the platinum single crystal in the cell containing 0.1 M HClO 4 solution to confirm the cleanliness and order of the surface, the electrode was moved to the other cell containing 10 −6 to 10 −5 M of the ion being deposited in 0.5 M H 2 SO 4 solution. The deposition was carried out by means of voltammetric cycling between 0.06 and 0.80 V (for Bi) or 0.90 V (for Te) at 50 mV s −1 . The step decoration can be easily monitored by following the blockage of the characteristic hydrogen de/adsorption peaks located around 0.12 V for (1 1 0) steps or 0.28 V for (1 0 0) steps. Special attention was paid to avoid the deposition of adatoms on the terrace by stopping the process before the characteristic anion adsorption contribution around 0.5 V starts to decrease. Once the step was completely blocked, the electrode was taken out from the deposition solution, rinsed with ultra pure water and immersed into the cell with electrolyte free of the corresponding ions of the adatom, for CO adsorption and electrochemical oxidation. Adsorption of CO was carried out following the same methodology used for CO displacement experiments that has been described elsewhere [35] . After recording the initial voltammetric profile, a flow of CO was introduced at 0.10 V over the meniscus formed between the electrode and the solution. The displacement current–time transient was recorded simultaneously until the displacement current dropped to zero, signaling the attainment of a fully blocked CO adlayer. After CO dosing, Ar was bubbled through the solution for at least 8 min to remove the dissolved CO, while keeping the electrode at 0.1 V in the bulk of the solution. Therefore, in all data reported here, CO electrooxidation took place in a solution that did not contain CO. Removing CO from the solution may result in a small decrease of the saturation coverage [36] . However, the blockage of the surface remained complete. Finally, the working electrode was returned to the meniscus configuration for the following potential step experiments. The CO oxidation was initiated by stepping the potential to the desired potential value 0.55 ≤ E ≤0.79 V with an interval 0.03 V. After the transient, once CO oxidation current had dropped to zero, the potential was stepped to 0.1 V, and a blank voltammetric profile was recorded to check for a possible readsorption of CO or any other solution contamination as well as for the stability of the adatoms. The cell and all glassware were immersed in a potassium permanganate solution overnight, followed by rinsing with water and a solution of hydrogen peroxide and sulfuric acid. Finally, everything was boiled and rinsed with ultra-pure water several times. A platinum wire was used as a counter electrode and a reversible hydrogen electrode was used as a reference. Working solutions were prepared from concentrated HClO 4 , H 2 SO 4 (Merck Suprapur), Bi 2 O 3 , TeO 2 (Merck) and ultrapure Elga–Vivendi water (18.2 MΩ cm). Argon (N50, Air Liquide in all gas used) was used to deoxygenate all solutions and CO (N47) to dose CO. All electrochemical experiments were carried out using a waveform generator (EG&G PARC 175) together with a potentiostat (Amel 551 or eDAQ EA161) and a digital recorder (eDAQ, ED401). All experiments were performed at room temperature. 3 Results and discussion In our previous study [37] , CO voltammetric stripping and in situ infrared reflection–absorption spectra (IRAS) have been applied to reveal the local properties of step sites on stepped surface decorated or not by irreversibly adsorbed adatoms. It has been found that CO coverage on (1 0 0) steps is much lower than that on (1 1 0) steps. In situ IRAS studies confirmed that the (1 1 0) step sites are dominated by atop CO while bridged bonded CO are prevalent on (1 0 0) step sites. In order to gain further insight into the CO oxidation mechanism on step decorated surfaces, in particular on the reactivity of step sites, chronoamperometric measurements for the oxidation of saturated CO adlayers were carried out for all studied surfaces at a series of constant oxidation potentials. The corresponding current–time transients of CO oxidation in 0.1 M HClO 4 are shown in Figs. 1 and 2 . It is evident that all these transients, corresponding to different stepped surfaces both with and without adatom decoration, have a similar shape: fast charging of the double layer and a constant plateau at short times, followed by a main oxidation peak, which is also similar to the features observed in a number of previous studies on the kinetics of the CO adlayer oxidation in 0.5 M H 2 SO 4 [22] . In the current plateau region, the constant current implies a constant coverage of OH which is very small due to the limited availability of free sites for OH adsorption and the relaxation of the CO adlayer [38] . The apparent reaction order with respect to adsorbed CO is either zero or “quasi zero”, as described by Eq. (4) or (5) , respectively [21] . (4) j ∝ d θ C O d t = − k (5) j ∝ d θ CO d t = − k θ CO , with θ CO ≈ constant The reason for the existence of a constant plateau is not clear in the present understanding of CO oxidation. The most accepted interpretation is that the CO adlayer relaxes during the initial stages of the oxidation with the consequence of charge flowing without liberation of active sites for OH adsorption [21,22,38] . Strictly speaking, constant OH coverage would lead to a first order decay of the current. However, for the small variation of the CO coverage in the plateau region, smaller than a 6%, a pseudo zero order behavior can be accepted (Eq. (5) ). Besides, it is also likely that the small decrease of CO coverage is compensated with a small increase of OH coverage resulting in the experimentally observed zero order behavior (Eq. (4) ). This situation is clearly different from the reaction order of the main peak, where a more complex dependence on the CO and OH coverage leads to a symmetric relationship of the current vs. time, as described by Eq. (6) [40] . (6) j ∝ d θ CO d t = − k θ CO θ OH The current in the plateau region does decrease only slightly after the Pt( h k l ) steps were decorated with Bi or Te (see Table 1 ). According to previous studies, a small amount of defect sites significantly enhance the CO oxidation in the current plateau region. Nevertheless, further increase of the step density does not lead to a significant rise of the plateau current [22] . This points out the possibility that a very small amount of defects could still remain in the step decorated surfaces and play an active role for CO oxidation [37] . For all Pt electrodes, including the corresponding step decorated ones, the plateau currents increase with increasing the step potential. The fittings of the plots of the logarithm of the plateau current (log( j pl )) vs. the electrode potential give a set of Tafel slopes which have been summarized in Table 1 . In the case of non-decorated Pt single crystal planes, the Tafel slopes, ranging from 67 ± 5 mV of Pt(5 4 4) to 89 ± 2 mV of Pt(5 5 4), do not show a clear trend with the step type and step density. However, there is a clear decreasing trend in Tafel slopes for both surfaces with (1 1 0) steps and (1 0 0) steps after they were decorated, especially for Pt(5 3 3) where the value decreases to near half, 42 ± 1 mV. The only exception would be Pt(5 4 4)/Te for which the Tafel slope increases to 83 ± 1 mV. Taking into account the low current values measured in the plateau, as compared to those in the mean peak, quantitative results should be taken cautiously because the experimental error may be important. In fact, the plateau contribution to the overall process of CO stripping can be considered small although it might be a key step in the formation of OH adsorption sites. It is worthwhile pointing out that, up to now, although the current plateau region has been observed in a lot of previous studies, the reaction mechanism about the initial CO oxidation is still unclear. It has been proposed in recent and closely related studies [22] , that the mechanism of CO oxidation in the plateau region should be also assigned to a Langmuir–Hinshelwood mechanism (CO ad + OH ad → CO 2 + H + + e − ) rather than an Eley–Rideal type (CO ad + H 2 O → CO 2 + 2H + + 2 e − ) [22,39] . That is, the initial CO oxidation may undertake a Langmuir–Hinshelwood mechanism with no effective liberation of sites for OH adsorption after the first few CO molecules are oxidized, because the relaxation of the CO adlayer would compensate the first removal of CO molecules. In good agreement with results reported by Lebedeva et al. [22] , the decrease of j pl and the Tafel slope after the surface steps were blocked by decoration with extra adatoms may well indicate the critical role of steps in the initial stage of CO oxidation and demonstrate that the reaction is started at steps, since the anion (OH) preferentially adsorbs on the step sites. In the main oxidation peak, a clear dependence of the peak position with the step potential can be observed. After step decoration, normally longer time is necessary to reach the maximum current. The CO oxidation in the main peak region observed in this work is in qualitative agreement with those reported earlier. This indicates that the proposed mechanism is valid regardless the supporting electrolyte or the surface structure [21–25] . The Langmuir–Hinshelwood reaction type has been widely employed to interpret this reaction kinetics and good coincidence between different reported results has been obtained. According to previous studies, the mean-field approximation for Langmuir–Hinshelwood mechanism gave better results than fitting with instantaneous/progressive nucleation and growth model, indicating a fast diffusion of CO on all studied surfaces, independent of step decoration. Thus, the mean-field approximation was applied to simulate all the current transients in this study using the following equation [40] : (7) j ( t ) = q k exp ( − k ( t − t max ) ) [ 1 + exp ( − k ( t − t max ) ) ] 2 In Eq. (7) , j ( t ) is the current density, q is the charge density associated with CO adlayer oxidation in the main peak, including contributions from anion re-adsorption, which is less important in this case, k is the rate constant for the Langmuir–Hinshelwood reaction step, and t max is the time at which the maximum is observed in the transient. The fittings for different surfaces are shown in Fig. 3 . It can be seen that the experimental data can be satisfactorily fitted by using this approach. The observation of a tailing on surfaces with higher step density (for example Pt(2 2 1)) is similar to that reported previously [22] , and it has been ascribed to the slow oxidation of the CO molecules that remain on the steps after the oxidative removal of CO from the terraces. A series of rate constant, k , at different step potential for each surface can be determined by the mean-field approximation. The logarithm of k (log( k )) was plotted vs. the step potential and shown in Fig. 4 for the different studied surfaces. Table 1 lists the log( k ) values of all studied surfaces at 0.73 V. It can be inferred from this table that the rate constant k increases with the step density for a given step symmetry at a given potential, and also that it is higher for the surfaces having (1 1 0) steps than for those having (1 0 0) steps with similar step density, e.g. the rate constant (log( k s −1 )) of Pt(5 5 4) at 0.73 V is 0.69 while it is lower for Pt(5 4 4) (0.52). After the surface was decorated with Bi or Te, all the rate constants decrease, and the variation is more apparent for the surfaces with higher step density. The magnitude of the decrease of the rate constant k after step decoration depends on the symmetry of the step. For a similar step density, the average decrease of k is more important for (1 1 0) steps than for (1 0 0) steps. This can be clearly seen in Fig. 5 , where the ratio between k for the surfaces with and without step decoration is plotted as a function of potential. It is clearly observed that the curve corresponding to Pt(5 5 4) lies below the curve for Pt(5 4 4) in almost all cases studied. Similar conclusion is obtained both with Te and Bi decoration and that proves that differences in the type of adatom do not play a major role in the kinetics. These results demonstrate the different catalytic activity of step sites for CO oxidation: the catalytic effect of the (1 1 0) step is always higher than that of the (1 0 0) step. This may be related to a higher adsorption energy of CO on (1 0 0) steps, because the main CO species on (1 0 0) steps is bridge bonded CO while on (1 1 0) steps is atop CO [37] . In addition, OH adsorption energy is also higher on (1 1 0) steps (the adsorption takes place at lower potentials) than on (1 0 0) steps [41] . This is likely a consequence of a more negative local pzfc and a higher step dipole associated with the former step symmetry. Since the only effect of adatoms is to neutralize the charge at step sites while showing negligible effect towards (1 1 1) terraces [37] , if the steps were perfectly decorated by the adatoms, the activity of the decorated surfaces should have been similar to that of an ideal basal (1 1 1) plane, which is very slow [22] . Hence, the kinetics of CO oxidation on decorated surfaces should have decreased dramatically and become lower than any of the unmodified vicinal step surfaces. However, as shown in Table 1 , the log( k / s −1 ) for surfaces with shorter terraces after step decoration (Pt(2 2 1)–Bi is 0.89, Pt(5 3 3)–Bi is 0.73) is still higher than that of the corresponding non-modified surfaces (Pt(5 5 4) is 0.69, Pt(5 4 4) is 0.52). This maybe due to the fact that adatoms cannot block completely the step sites and they retain significant activity for CO oxidation, similarly to what has been also observed in the current plateau region, or, alternatively, it may indicate that shorter terraces exhibit higher activity towards CO oxidation. The plots of log( k ) vs. the final step potential give a set of Tafel slopes, included also in Table 1 . The magnitude of this series of Tafel slopes is mainly located between 60 and 80 mV, relatively close to 60 mV. This indicates that the mechanism of CO oxidation is the same for all different surface structures, with and without step decoration, including a slow limiting chemical step (Eq. (2) ) as indicated in the reaction scheme mentioned above with some changes that could be attributed to the CO adlayer compactness, especially for short terraces. The value of the Tafel slope for Pt(5 5 4) is 75 ± 2 mV, while it is 94 ± 5 mV for Pt(2 2 1). These values are coincident with those reported by Lebedeva et al. using 0.1 M H 2 SO 4 as working solution (78 ± 4 mV for Pt(5 5 4) and 97 ± 4 mV for Pt(5 5 3) which has a step density close to that of Pt(2 2 1) electrode surface) [22] . The Tafel slope for Pt(5 3 3) is 88 ± 4 mV, somewhat higher than that reported by Inkaew et al. (79 ± 4 mV) [24] . According to our data and that reported by Lebedeva et al., we can reach the conclusion that the same trend was observed for the Tafel slope that increases with increasing the step density for both Pt(s)[( n − 1)(1 1 1) × (1 1 0)] and Pt(s)[ n (1 1 1) × (1 0 0)] single crystals, at least for shorter terrace surfaces. Similar results have been observed on Pt(s)[( n − 1)(1 0 0) × (1 1 0)] surfaces reported by Vidal-Iglesias et al., e.g. the Tafel slope is about 75 mV for surfaces having wide terraces (Pt( n 1 0) when n ≥ 7), while it can be over 100 mV for shorter terrace surfaces (Pt(3 1 0) and Pt(2 1 0)) [25] . Indeed, the Tafel slope seems also to depend on the symmetry of the surface steps, i.e. under similar step density, the surface with (1 1 0) steps leads to a higher Tafel slope than that with (1 0 0) steps. After the step was decorated with irreversibly adsorbed Bi or Te, the Tafel slope for Pt(s)[( n − 1)(1 1 1) × (1 1 0)] electrodes decreased to ca. 70 mV and it also decreases to ca. 60 mV for Pt(s)[ n (1 1 1) × (1 0 0)] electrodes. Since the adatoms decorating steps show negligible electronic effect towards (1 1 1) terrace sites, and they only neutralize the charge associated with step sites [37] , step decoration may just lead to a decrease of step density. In this regard, the diminution of the Tafel slope after decoration of the steps follows the same trend as the step density dependence of the Tafel slope observed on the unmodified surfaces, as mentioned above. The change of Tafel slope related to the variation of step sites (type, density and modification) can be attributed to changes in the adsorption isotherm of OH, since the experimental Tafel slope is greatly affected by the potential dependence of OH coverage. As has been discussed previously in CO voltammetric stripping and in situ IRAS studies [37] , the adsorbed CO species and their coverage at (1 1 0) and (1 0 0) steps are quite different. This points out that the OH adsorption isotherm and coverage may be somewhat different at steps with different symmetry, because the adsorption of OH is greatly affected by the adsorbed CO. It has been pointed out that OH adsorption isotherm and coverage is different for surfaces having wide and narrow terraces, i.e., with different step density [25] . It is worth to remark that the variation trend of the Tafel slope of log( j pl ) vs. E is quite coincided with that of log( k ) vs. E , which confirms that CO oxidation in the current plateau region may take place with the same mechanism as that in the main peak region, i.e., both with a Langmuir–Hinshelwood mechanism, but just at a different stage of the reaction. 4 Conclusions In this work the effects of step sites on platinum single crystal towards CO electrochemical oxidation has been studied by step decoration with adatoms. Chronoamperometric experiments demonstrate that step sites are the reactive sites of Pt single crystal planes vicinal to Pt(1 1 1), and (1 1 0) step sites possess higher catalytic effect towards CO oxidation than (1 0 0) step sites. This can be evidenced by the different effects of decrease of the rate constant after the steps were blocked by foreign adatoms decoration. Two processes (the current plateau and the main peak regions) of electrooxidation of a saturated CO adlayer can be observed in the current–time transients for all surfaces studied. These two processes may follow the same mechanism, based on the Langmuir–Hinshelwood model, which can be simulated satisfactorily for the main peak region using a mean-field approximation. The Tafel slopes were found located between 60 and 90 mV which confirms that the mechanism of CO oxidation, irrespectively of step decoration, is the same, i.e., Langmuir–Hinshelwood type with a slow chemical step in the reaction scheme. The change of the Tafel slopes is in good agreement with the variation of the properties of step sites (symmetry, density and decoration), indicating the sensitivity of step sites for the adsorption of oxygen-containing species and pointing out the important role of step sites towards CO oxidation. Acknowledgements This work has been financially supported by the Ministerio de Educación y Ciencia of Spain through the project CTQ2010-16271 (Feder). QSC acknowledges the fellowship support of the China Scholarship Council and the support of NSFC (grant no. 20833005 ). References [1] G.A. Somorjai Introduction to Surface Chemistry and Catalysis 1994 John Wiley & Sons New York [2] R.A. van Santen Acc. Chem. Res. 42 2009 57 [3] J. Tersoff L.M. Falicov Phys. Rev. B 24 1981 754 [4] J.T. Yates Jr. J. Vac. Sci. Technol. A 13 1995 1359 [5] B. Hammer J.K. Nørskov Adv. Catal. 45 2000 71 [6] B. Beden C. Lamy N.R. de Tacconi A.J. Arvia Electrochim. Acta 35 1990 691 [7] S.C. Chang M.J. Weaver J. Chem. Phys. 92 1990 4582 [8] J. Xu P. Henriksen J.T. Yates Jr. J. Chem. Phys. 97 1992 5250 [9] E. Herrero J.M. Feliu S. Blais Z. Radovic-Hrapovic G. Jerkiewicz Langmuir 11 2000 4779 [10] M.T.M. Koper N.P. Lebedeva C.G.M. Hermse Faraday Discuss. 121 2002 301 [11] J. Meier K.A. Friedrich U. Stimming Faraday Discuss. 121 2002 365 [12] B. Andreaus F. Maillard J. Kocylo E.R. Savinova M. Eikerling J. Phys. Chem. B 110 2006 21028 [13] J. Solla-Gullon F.J. Vidal-Iglesias E. Herrero J.M. Feliu A. Aldaz Electrochem. Commun. 8 2006 189 [14] F. Maillard E.R. Savinova U. Stimming J. Electroanal. Chem. 599 2007 221 [15] Q.S. Chen S.G. Sun Z.Y. Zhou Y.X. Chen S.B. Deng Phys. Chem. Chem. Phys. 10 2008 3645 [16] S. Gilman J. Phys. Chem. 68 1964 70 [17] C. McCallum D. Pletcher J. Electroanal. Chem. 70 1976 277 [18] B. Love J. Lipkowski ACS Symp. Ser. 378 1988 484 [19] M.T.M. Koper A.P.J. Jansen R.A. van Santen J.J. Lukkien P.A.J. Hilbers J. Chem. Phys. 109 1998 6051 [20] M.T.M. Koper A.P.J. Jansen J.J. Lukkien Electrochim. Acta 45 1999 645 [21] N.P. Lebedeva M.T.M. Koper J.M. Feliu R.A. van Santen J. Electroanal. Chem. 524–525 2002 242 [22] N.P. Lebedeva M.T.M. Koper J.M. Feliu R.A. van Santen J. Phys. Chem. B 106 2002 12938 [23] P. Inkaew W. Zhou C. Korzeniewski J. Electroanal. Chem. 614 2008 93 [24] P. Inkaew C. Korzeniewski Phys. Chem. Chem. Phys. 10 2008 3655 [25] F.J. Vidal-Iglesias J. Solla-Gullon J.M. Campina E. Herrero A. Aldaz J.M. Feliu Electrochim. Acta 54 2009 4459 [26] R. Parsons T. VanderNoot J. Electroanal. Chem. 257 1988 9 [27] R.R. Adzic J.O. Bockris R.E. White B.E. Conway Modern Aspects of Electrochemistry vol. 21 1990 Plenum Press New York (Ch. 5) [28] E. Herrero A. Fernandez-Vega J.M. Feliu A. Aldaz J. Electroanal. Chem. 350 1993 73 [29] S.P.E. Smith H.D. Abruna J. Electroanal. Chem. 467 1999 43 [30] J.M. Feliu R. Gomez M.J. Llorca A. Aldaz Surf. Sci. 289 1993 152 [31] J.M. Feliu R. Gomez M.J. Llorca A. Aldaz Surf. Sci. 297 1993 209 [32] J. Solla-Gullon F.J. Vidal-Iglesias P. Rodriguez E. Herrero J.M. Feliu J. Clavilier A. Aldaz J. Phys. Chem. B 108 2004 13573 [33] E. Herrero V. Climent J.M. Feliu Electrochem. Commun. 2 2000 636 [34] J. Clavilier D. Armand S.G. Sun M. Petit J. Electroanal. Chem. 205 1986 267 [35] J. Clavilier R. Alabalat R. Gomez J.M. Orts J.M. Feliu A. Aldaz J. Electroanal. Chem. 330 1992 489 [36] I. Villegas M. Weaver J. Chem. Phys. 101 1994 1648 [37] Q.S. Chen A. Berna V. Climent S.G. Sun J.M. Feliu Phys. Chem. Chem. Phys. 12 2010 11407 [38] N.M. Markovic B.N. Grgur C.A. Lucas P.N. Ross J. Phys. Chem. B 103 1999 487 [39] D.D. Eley E.K. Rideal Nature 146 1940 401 [40] M. Bergelin E. Herrero J.M. Feliu M. Wasberg J. Electroanal. Chem. 467 1999 74 [41] N. García-Aráez V. Climent J.M. Feliu Electrochim. Acta 54 2009 966