Electrocatalytic Oxidation of Formic Acid at an Ordered ... - Wiley

0 downloads 0 Views 123KB Size Report
surface. The PtBi electrode exhibits superior properties when compared to polycrystalline platinum in terms of oxidation onset potential, current density, and a ...
ELECTROCHEMISTRY SPECIAL

Electrocatalytic Oxidation of Formic Acid at an Ordered Intermetallic PtBi Surface Emerilis Casado-Rivera, Zolta¬n Ga¬l, A. C. D. Angelo, Cora Lind, Francis J. DiSalvo,* and He¬ctor D. Abrunƒa*[a] The electrocatalytic oxidation of formic acid at a PtBi ordered intermetallic electrode surface has been investigated using cyclic voltammetry, rotating disk electrode (RDE) voltammetry and differential electrochemical mass spectrometry (DEMS). The results are compared to those at a polycrystalline platinum electrode surface. The PtBi electrode exhibits superior properties when compared to polycrystalline platinum in terms of oxidation onset potential, current density, and a much diminished poisoning effect by CO. Using the RDE technique, a value of 1.4  10 4 cm s 1 was

obtained for the heterogeneous charge transfer rate constant. The PtBi surface did not appear to be poisoned when exposed to a CO saturated solution for periods exceeding 0.5 h. The results for PtBi are discussed within the framework of the dual-path mechanism for the electrocatalytic oxidation of formic acid, which involves formation of a reactive intermediate and a poisoning pathway.

The study of the electrocatalytic oxidation of small organic molecules has seen a great deal of impetus in the recent past because such molecules can potentially be employed in fuel cell applications.[1] Much of the research, to date, has focused on the oxidation of simple, single-carbon-containing molecules, such as formaldehyde, formic acid, and methanol, since these have potentially simpler and cleaner reaction pathways. The oxidation of these molecules requires the use of catalysts in order to achieve the current densities needed for commercially viable fuel cells. This has generated a great interest in the development of new catalyst materials. As a class, precious metals such as platinum as well as their alloys represent some of the most efficient catalyst materials for the oxidation of small organic molecules. However, these materials, and especially platinum, are very susceptible to poisoning species such as CO. These poisons tend to remain strongly bound to the electrode surface, resulting in a decrease in catalytic activity and cell efficiency. The use of low-index single crystal faces has made it possible to determine the relationship between surface structure and reactivity and whether poisoning pathways are surface sensitive/ specific. For the electrooxidation of formic acid on platinum[2±8] it has been found that the Pt(111) surface has a lower propensity towards the formation of strongly bound intermediates (CO) than the Pt(110) and Pt(100) surfaces.[8] This results in the Pt(111) surface exhibiting higher current densities for the electrooxidation of formic acid as well as less poisoning effects than the traditionally more reactive Pt(110) and Pt(100) surfaces. The improvement of the catalytic activity and diminution of poisoning effects of single crystal electrode surfaces has been sought through modification with submonolayer to monolayer amounts of foreign metal adatoms.[1, 9] A system that has received a great deal of attention has been the oxidation of

formic acid on bismuth-modified Pt(111) electrodes.[5, 10±16] In this case it has been suggested that the bismuth adatoms catalyze the oxidation reaction through electronic effects with the substrate surface, since the poison formation reaction is greatly diminished even at low bismuth coverages. However, it has also been suggested that the bismuth adatoms exist as oxide and/or hydroxide species in their oxidized state.[17±19] These oxygenated species may contribute to the oxidation of formic acid and render bismuth-modified electrodes as ™bifunctional∫ catalysts. In addition, it has been proposed that the oxidation of formic acid on these surfaces follows a dual-path mechanism. As depicted in Scheme 1, this involves either the formation of a reactive intermediate, that subsequently yields CO2 as final product, or the formation of adsorbed CO via dehydration of formic acid. The adsorbed CO is then oxidized to CO2 . More recently, work has also focused on determining the identities of the reaction intermediates and poisons during the oxidation of these fuels, in order to gain an understanding of the mechanistic pathways involved in this process.[20±22] In this regard, the use of in situ Fourier transform infrared spectroscopy (FT-IR) has been especially valuable.[23±27] A very powerful, although less well-known (compared to FT-IR) technique is differential electrochemical mass spectrometry (DEMS), initially

KEYWORDS: electrocatalysis ¥ electrochemistry ¥ metals ¥ oxidation ¥ surfaces

[a] Prof. F. J. DiSalvo, Prof. H. D. Abrunƒa, E. Casado-Rivera, Z. Ga¬l, Prof. A. C. D. Angelo, Dr. C. Lind Department of Chemistry and Chemical Biology Baker Laboratory Cornell University Ithaca, NY, 14853-1301 (USA) Fax: (‡ 1) 607-255-9864 E-mail: [email protected], [email protected]

CHEMPHYSCHEM 2003, 4, 193 ± 199 ¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1439-4235/03/04/02 $ 20.00+.50/0

193

F. J. DiSalvo, H. D. Abrunƒa et al.

Scheme 1. Dual-path mechanism for formic acid oxidation on the surface of a Pt electrode.

developed by Bruckenstein[28] and later improved by Heitbaum.[29, 30] This technique provides virtually unambiguous identification of transient species generated during an electrochemical reaction. The very short delay times (< 1 s) for the detection of the species in the mass spectrometer allows for the potential to be slowly swept during the reaction, giving potential-dependent species-formation data in real time. We have previously carried out extensive studies of the oxidation of formic acid on bare and bismuth-modified stepped platinum surfaces with (111) terraces of varying width and monatomic (110)[16] or (100) steps.[31] The use of platinum surfaces is advantageous, since the voltammetric profile provides unambiguous information as to which sites are blocked and the extent of such blockage. Moreover the bismuth adlayer on the (111) terraces has a well-defined redox response, which can also be employed in a diagnostic way providing additional information. The data obtained suggest that for bismuthmodified platinum stepped surfaces, the electrocatalytic activity increases significantly once the steps sites are blocked/modified. This suggests that on the step sites, the bismuth adatoms act to block poison formation; typically a third-body effect. Once the step sites are blocked, increasing the surface coverage of the bismuth adatoms causes a dramatic increase (over 40  ) of the electrocatalytic current for formic acid oxidation. Since these additional bismuth adatoms are found on the terrace sites (recall that at this stage all step sites are blocked), a dramatic increase in current for a very small incremental change in bismuth surface coverage would suggest that the effect of bismuth adatoms on the terrace sites is largely electronic in nature. One of the limitations of adsorbate-modified surfaces, such as Biads/Pt, is the fact that the nature of the samples is not well defined and can change with time. Keeping this and other constraints in mind, and using our previous studies on Bimodified Pt surfaces as a point of departure, we have embarked on an investigation of the electrocatalytic activity of PtBi ordered intermetallic electrodes. From an analysis of the PtBi phase diagram we decided to initially investigate the electrocatalytic activity of PtBi. This was prompted by the fact that the 1:1 ratio of Bi and Pt was in line with our previous analysis[16] and that it is a congruently melting phase, which made it convenient from a synthesis point of view.

194

According to the literature,[32] PtBi crystallizes in the NiAs structure with Bi at (0, 0, 0) and Pt at (1³3, 2³3, 1³4). This is surprising, since it contrasts with other PtM (M ˆ metal) compounds that also form with the NiAs structure. PtSn and PtSb are reported to crystallize with Pt at the origin and Sn or Sb respectively at (1³3, 2³3, 1³4).[33] Analysis of the interatomic distances suggests that the atom positions reported for PtBi are, in fact, switched. For convenience, let us refer to the (0, 0, 0) position as the M1 site and to the (1³3, 2³3, 1³4) position as the M2 site. The erroneous assignment of the atom positions in PtBi may be due to the small scattering contrast between Pt and Bi when using laboratory X-rays, whereas the scattering contrast between Pt and Sn or Sb should be sufficient for easy differentiation of the elements. Furthermore, the structure determination was carried out in 1962, long before the advent of Rietveld refinement techniques for powder data. In order to address this problem, we have collected high quality powder data on a Scintag XDS 2000. The data was analyzed by the Rietveld method by assuming that a) the reported literature structure is correct, and b) the atom positions reported by Zhuravlev and Stepanova[32] are switched. Using the same number of refinable parameters, model b) consistently gave a better fit than model a). We will therefore base our subsequent discussion on this revised crystal structure for PtBi with Pt in the M1 and Bi in the M2 position. Herein we describe the electrocatalytic activity of PtBi towards formic acid and compare the results with those for platinum. In these studies we have employed cyclic voltammetry, rotating disk electrode (RDE) voltammetry, and differential electrochemical mass spectrometry (DEMS).

Results Voltammetry in 0.1 M H2SO4 Cyclic voltammograms obtained for Pt and PtBi electrodes in 0.1 M sulfuric acid solution are presented in Figure 1. The Pt electrode (Figure 1 a) exhibited the well-known voltammetric profile with oxide formation and reduction peaks as well as the characteristic hydrogen adsorption waves. The profile for PtBi is reminiscent of that for Pt with well-defined oxidation and reduction peaks, which we ascribe to oxide formation and reduction, respectively. In addition there was a small wave centered at about 0.2 V, which would suggest weak hydrogen adsorption. However this assignment is speculative since we have no direct evidence of this. It is also worth noting that the waves ascribed to the surface oxide formation/reduction on PtBi are shifted negative of those for Pt indicating the formation/ reduction of surface oxides at lower potentials. It is also worth noting the significantly higher current density for the PtBi electrode. At this time, we are not certain of its origin. However the PtBi electrode was polished to a mirror finish. A similarly polished Pt electrode has a roughness factor of about 30 % (determined from a comparison of the geometric area versus that obtained from the hydrogen adsorption charge). However, the large difference in current density could

CHEMPHYSCHEM 2003, 4, 193 ± 199

Oxidation at a PtBi Surface

ELECTROCHEMISTRY SPECIAL in current, presumably due to poisoning effects. On the subsequent cathodic sweep there is a sharp peak that decays as more negative potentials are reached (in the subsequent discussion these peaks will be referred to as I, II, and III, respectively). If the potential scan is limited to 0.20 to ‡ 0.20 V the forward and reverse scans are virtually superimposable, suggesting limited poisoning effects. From a qualitative analysis of these voltammetric profiles two aspects are immediately evident. First, at the PtBi surface the onset potential for formic acid oxidation is shifted negatively by over 300 mV. Second, the current density at the PtBi electrode is dramatically higher than that for the Pt electrode. As mentioned earlier, while this could be due to a difference in microscopic electrode area, it is more likely due to a difference in the activity. As mentioned above, one of the more salient features of the electrocatalytic activity of PtBi was the dramatic displacement of the onset potential for formic acid oxidation relative to a Pt electrode. In order to further study this aspect, the onset potential for formic acid oxidation was measured as a function of its bulk concentration (Figure 3). In these studies, the onset

Figure 1. Cyclic voltammograms obtained on a) Pt and b) PtBi in 0.1 M sulfuric acid solution (scan rate 10 mV s 1).

not be accounted for by roughness, although it could well be that the PtBi surface is significantly rougher. Alternatively, the Bi may contribute to a larger surface peak. These are aspects that are currently under investigation.

Formic Acid Oxidation Figure 2 presents the voltammetric profile for formic acid oxidation on Pt and PtBi electrode surfaces. The profile at the Pt electrode (Figure 2 a) is well established with the onset potential on first anodic sweep being about ‡ 0.2 V. Poisoning effects by CO are evident and upon oxidation of the adsorbed CO, the current increases significantly in the cathodic sweep with a well-defined hysteresis. On the PtBi surface the initial anodic sweep exhibits an onset potential of about 120 mV (see below) and reaches an apparent steady-state current (of about 5 mA cm 2) at about ‡ 0.10 V. Scanning to more positive potentials results in an additional peak followed by a decrease

Figure 3. Dependence of the onset potential (measured at 10 mA) on the formic acid concentration in 0.1 M sulfuric acid solution.

potential was taken as that at which the current reached 10 mA. Although the choice of the current magnitude is clearly arbitrary, it does provide for an estimate of the limiting value. In these studies the electrode was rotated at 3000 rpm in order to minimize the blockage of the electrode surface from gas (ostensibly CO2) bubbles formed. In addition, the potential was swept slowly, typically at a rate of 10 mV s 1. As can be ascertained, there is initially an approximately linear displacement of the onset potential to more negative values as the formic acid concentration is increased. At higher concentrations there is a diminution of the dependence and at the highest concentration employed, the onset potential for formic acid oxidation asymptotically reaches a value of about 0.18 V. As Figure 2. Cyclic voltammograms obtained for formic acid oxidation on a) Pt and b) PtBi in 0.1 M mentioned previously, this represents a dramatic 1 sulfuric acid plus 0.125 M formic acid solution (scan rate 10 mV s ). The steady-state currents are improvement over Pt, with a difference of over a) 0.063 and b) 2.4 mA cm 2. Note that in b) the current scale is significantly magnified compared to 300 mV. a).

CHEMPHYSCHEM 2003, 4, 193 ± 199

195

F. J. DiSalvo, H. D. Abrunƒa et al. CO Poisoning Effects As mentioned earlier, one of the major drawbacks of the use of platinum as an electrocatalyst in fuel cell applications is its high propensity to poisoning by CO. In order to assess the susceptibility of PtBi to CO poisoning we carried out studies in sulfuric acid solution saturated with CO and compared the results to those on Pt. The propensity of CO to adsorb to platinum surfaces is well known and documented. Exposure of a Pt surface to a CO containing solution results in a rapid and irreversible adsorption by CO. The modification of the voltammetric profile for a platinum surface in sulfuric acid solution before and after (Figures 4 a and 4 b) exposure to CO is dramatic. As shown in Figure 4 b, the surface is strongly passivated to the point there is essentially no current up to about ‡ 0.60 V, at which there is a sharp peak ascribed to the oxidation of adsorbed CO to CO2 . On the PtBi surface, the effects are dramatically different as shown in Figure 4 c, where we present the cyclic voltammetric profiles prior to and after exposure to CO. As can be ascertained, the voltammetric profiles are qualitatively very similar with

somewhat of an enhancement of peak IV in the anodic sweep. Peak V, on the other hand, remained essentially unchanged. These results indicate a dramatic diminution in the affinity of CO towards the PtBi surface relative to Pt. From DEMS studies (see below) we believe that the small current enhancement of peak IV could be due to oxidation of CO to CO2 with the former likely to arise from the dehydration pathway of formic acid as presented in Scheme 1. We believe that the dramatic drop in the affinity of CO for PtBi is a direct consequence of its structure, specifically, the Pt Pt distance in Pt versus the Pt Pt distance in PtBi. In Pt the interatomic distance is 2.77 ä whereas in PtBi the Pt Pt bond length is dramatically increased to 4.32 ä (see Figure 5). This expansion makes it very difficult for CO to bind in bridge site or three-fold hollow site configurations, thus decreasing its binding affinity.

Figure 5. Structures of Pt and PtBi surfaces.

Kinetic Studies In order to gain some insight on the kinetics of the formic acid oxidation at the PtBi electrode surface, RDE experiments were carried out. In these studies the potential was swept (at 10 mV s 1) from 0.20 to ‡ 0.25 V; this region is of interest in fuel cell applications. It should also be mentioned that due to a significant gas evolution (ostensibly CO2) the surface might be partially covered by gas (especially at the slower rates of rotation). The inset to Figure 6 shows the Levich plot (IL versus w1/2) for the oxidation of formic acid at the PtBi electrode. In these experiments the limiting current was measured at ‡ 0.20 V. As can be seen, the plot exhibits significant curvature, suggesting a reaction that is kinetically, and not transport, limited. For a

Figure 4. Cyclic voltammograms obtained for CO oxidation studies in 0.1 M H2SO4 : a) Pt before exposure to CO, b) Pt after exposure to CO (scan rate 50 mV s 1) and c) PtBi before and after exposure to CO (scan rate 10 mV s 1).

196

Figure 6. Koutecky ± Levich plot for formic acid oxidation on PtBi in 0.1 M sulfuric acid plus 0.125 M formic acid solution. Inset: Levich plot for formic acid oxidation on PtBi in 0.1 M sulfuric acid plus 0.125 M formic acid solution (current measured at ‡ 0.25 V).

CHEMPHYSCHEM 2003, 4, 193 ± 199

Oxidation at a PtBi Surface

ELECTROCHEMISTRY SPECIAL

system that is kinetically controlled, theory predicts that a Koutecky ± Levich plot of 1/IL versus w 1/2 should be linear and from its intercept the value of the rate constant can be obtained. The main panel of Figure 6 shows such a plot and the linearity is immediately apparent. From the intercept we obtained a value of 1.4  10 4 cm s 1 for the rate constant. This value is comparable to values obtained for various platinum surfaces under conditions that were intended to minimize poisoning by CO as well as surface blocking from CO2 evolution.[34, 35] As mentioned earlier, it is quite likely that under our experimental conditions the surface of the PtBi electrode is partially blocked by the evolution of CO2 . Thus, the value presented above likely represents a lower limit. In addition, the fact that it is comparable to platinum (even when the surface is partially blocked) attests to its electrocatalytic activity. Moreover, as has been indicated before, the onset potential is significantly shifted to less negative values and the surface appears to be relatively immune and/or tolerant to CO. All of these attributes clearly make PtBi a superior electrocatalyst for formic acid oxidation. DEMS Studies In order to characterize the products from the electrocatalytic oxidation of formic acid, we have carried out DEMS experiments. As mentioned earlier DEMS allows for the detection of the (neutral) products of electrochemical reactions in real time. In these experiments the potential applied to the PtBi electrode was varied at a slow scan rate (2 mV s 1) while the ion intensities at mass values of 28 and 44 corresponding to CO (as well as N2) and CO2 , respectively, were monitored. Figure 7 a shows the cyclic voltammogram (at 2 mV s 1) for the oxidation of formic acid at a PtBi electrode over the potential range of 0.2 to 1.0 V. Figures 7 b and 7 c present the ion current intensities corresponding to CO2 and CO, respectively. It is clear that the oxidation of formic acid gives rise to the generation of both CO2 and CO and that there is a clear correspondence between the voltammetry in Figure 7 a and the profiles for the ion current intensities. The fact that both CO2 and CO are generated would suggest that the dual path mechanism for formic acid oxidation (see Scheme 1) is operating at the PtBi surface. In addition, the detection of CO also indicates that it is not strongly bound to the PtBi surface, again reinforcing our earlier assumption that CO does not appear to significantly poison the PtBi surface.

Discussion The results obtained in these studies indicate that the PtBi ordered intermetallic phase has properties and reactivities that are dramatically different from those of bare platinum surfaces, with regards to formic acid oxidation. Of particular note, the onset potential for the electrocatalytic oxidation of formic acid is significantly shifted (by over 300 mV) to more negative values and the current density (at a given potential) is significantly enhanced when compared to bare platinum. Moreover, the PtBi surface appears to have a dramatically lower sensitivity to poisoning by CO.

CHEMPHYSCHEM 2003, 4, 193 ± 199

Figure 7. Mass spectrometry/cyclic voltammograms obtained for formic acid oxidation in 0.1 M sulfuric acid plus 0.125 M formic acid solution: a) cyclic voltammogram, b) mass spectrometry/cyclic voltammogram for CO2 , and c) mass spectrometry/cyclic voltammogram for CO.

We ascribe these differences to both electronic and geometric effects. We believe that the shift in onset potential and increase in current density are dominated by electronic effects. In essence, we believe that formation of the PtBi ordered intermetallic results in a charge redistribution (as a first approximation arising from the work function differences), which enhances the affinity of PtBi towards formic acid and further gives rise to the formation of surface oxides at much lower potentials. These two effects combine to give rise to the enhanced performance observed. In addition, these effects are consistent with the generally accepted concept that in these types of systems the catalytic performance comes as the result of a combination of enhanced activation and the presence of surface oxides at low potentials. In terms of geometric effects, we believe that the greatly reduced propensity of PtBi towards poisoning by CO (relative to Pt) arises, to a significant extent, as a result of the increased Pt Pt distance in PtBi (Pt Pt ˆ 4.32 ä) in relation to Pt (2.77 ä; Figure 5). However there could also be some electronic effects involved. Clearly both electronic and geometric effects play key roles in the enhanced activity of PtBi towards the electrocatalytic oxidation of formic acid. We believe that the use of ordered

197

F. J. DiSalvo, H. D. Abrunƒa et al. intermetallics represents a new paradigm in the judicious and deliberate development and design of new electrocatalytic materials for fuel cell applications. This is in contraposition to adatom modified surfaces and bulk alloys. In the former, only the surface layer is modified and upon loss of the ad-layer the bulk material will not exhibit any activity. In bulk alloys, the distribution of the constituent elements will be compositiondependent and random. In most random Pt alloys, the metal ± metal distance remains close to that found in Pt. Thus poisoning is not greatly reduced. There can also be segregation effects. Thus, bulk alloys can have a propensity to aging in unpredictable ways. In fact, it has been recently shown that whereas platinum surfaces modified with adsorbed ruthenium are active in the electrocatalytic oxidation of methanol, bulk Ru/Pt alloys are not.[36] We are currently exploring the activity of other Bi/Pt ordered intermetallic phases as well as other related materials; results will be presented elsewhere.

Conclusions The activity of the PtBi ordered intermetallic phase towards the electrocatalytic oxidation of formic acid has been investigated and the results compared to those on platinum. Relative to platinum, PtBi exhibits a greatly enhanced performance in terms of the current density and the onset potential. Moreover, PtBi was virtually immune to poisoning by CO. We ascribe the enhanced performance on the ordered intermetallic phase performance to electronic and geometric factors. We believe that the current enhancement and onset potential are most influenced by electronic effects whereas the immunity to CO poisoning is largely controlled by the geometric and structural factors.

Experimental Section PtBi Intermetallic Phase: Platinum powder (Johnson Matthey 99.999 %) and bismuth powder (Alfa Aesar, 99.9999 %) from pellets were thoroughly mixed in the appropriate molar ratios (total 1.5 g) in an agate mortar. Pellets (6 mm OD, 4 mm length) were pressed in a hydraulic press. The pellets were sealed in evacuated fused silica tubes (10 mm ID, 100 mm length) and heated, oriented vertically, in a box furnace to 800 8C for 24 h and annealed at 650 8C for 48 h. The pellets were ground and the above procedure was repeated twice. Powder X-ray diffraction (Scintag XDS2000 equipped with an energy sensitive detector, CuKa radiation) indicated the presence of single phase PtBi[32] (P63/mmc, a ˆ 4.315 ä, c ˆ 5.490 ä). Electrode Preparation: The final pellets were cut into cylinders of  3 ± 5 mm length and were press-fitted into Teflon rods. Electrical contact was made through a graphite felt plug in turn connected to a stainless steel holder. The former could be connected to the rotating arbor of a Pine electrode rotator. Contact resistances between the stainless steel holder and the surface of the electrode were typically 3 ± 4 W. Once mounted, the electrodes were polished with 400 and 600 grid emery paper (Buehler). They were subsequently polished with 1 mm diamond paste (Buehler) to a mirror finish. Prior to each experiment, the electrodes were polished with

198

diamond paste (METADI-Buehler, f 1 mm), rinsed with ultrapure water (Millipore Milli-Q, 18 MW cm 1), and placed in an ultrasound bath for 10 min. All the solutions were prepared with ultrapure water. Sulfuric acid solutions (0.1 M ; J. T. Baker Ultrapure Reagent) solutions were used as supporting electrolyte; formic acid was obtained from Fisher Chemical (88 %). All solutions were deaerated with nitrogen for at least 15 min and measurements were conducted at room temperature. Carbon monoxide (CP grade) was obtained from Matheson Air Products. Electrochemistry: The setup used to characterize the electrodes was as described earlier.[19] All potentials are referenced to a saturated Ag/ AgCl electrode without regard for the liquid junction. Differential electrochemical mass spectroscopy (DEMS) was used to determine the production of CO and CO2 from the oxidation of formic acid on the PtBi and bare Pt electrode surfaces. The ionization chamber used in the DEMS experiments was pumped at 250 L s 1 with a turbomolecular pump (Varian) supported with a dry pump (Triscroll, Varian) in order to avoid contamination from oil vapors. Due to the high pressure in the ionization chamber during the electrochemical experiments, a turbomolecular pump (65 L s 1, Pfeiffer) supported with a dry diaphragm pump (Pfeiffer) differentially pumped the mass spectrometer analysis chamber. This chamber was isolated from the ionization chamber through a pressure converter (Leybold Inficon IPC2A). The quadrupole mass spectrometer (Leybold Inficon Transpector H-100 M) was connected to the analysis chamber and contained an electron multiplier/Faraday cup detector (Channeltron). The very short delay times (< 1 s) allowed the measurement of the mass intensity of the products as a function of the potential by slowly sweeping the potential (2 mV s 1).

This work was supported by the Natural Science Foundation (DMR9805719). The use of facilities in the Cornell Center for Materials Research (MRSEC grant DMR-0079992), the Office of Naval Research, and discussion with Sean P. E. Smith, are all gratefully acknowledged. A. C. D. A. thanks FundaÁaƒo CoordenaÁaƒo de AperfeiÁoamento de Pessoal de NÌvel Superior (CAPES) of Brazil for financial support during his scientific visit to Cornell University.

[1] R. Parsons, T. VanderNoot, J. Electroanal. Chem. 1988, 257, 9. [2] J. Clavilier, R. Parsons, R. Durand, C. Lamy, J. M. Leger, J. Electroanal. Chem. 1981, 124, 321. [3] A. Fernandezvega, J. M. Feliu, A. Aldaz, J. Clavilier, J. Electroanal. Chem. 1989, 258, 101. [4] T. Iwasita, X. H. Xia, E. Herrero, H. D. Liess, Langmuir 1996, 12, 4260. [5] S. A. Campbell, R. Parsons, J. Chem. Soc. Faraday Trans. 1992, 88, 833. [6] C. Lamy, J.-M. Le¬ger, J. Clavilier, R. Parsons, J. Electroanal. Chem. 1983, 150, 71. [7] J. Clavilier, S. G. Sun, J. Electroanal. Chem. 1986, 199, 471. [8] R. R. Adzic, A. V. Tripkovic, V. B. Vesovic, J. Electroanal. Chem. 1986, 204, 329. [9] R. R. Adzic, Reaction Kinetics and Mechanisms on Metal Single Crystal Electrodes, Vol. 21 (Eds.: R. E. White, B. E. Conway), Plenum Press, New York, NY, 1990, Chapt. 5. [10] J. Clavilier, A. Fernandez-Vega, J. M. Feliu, A. Aldaz, J. Electroanal. Chem. 1989, 258, 89. [11] E. Herrero, A. Fernandez-Vega, J. M. Feliu, A. Aldaz, J. Electroanal. Chem. 1993, 350, 73. [12] S. C. Chang, Y. H. Ho, M. J. Weaver, Surf. Sci. 1992, 265, 81. [13] E. Leiva, T. Iwasita, E. Herrero, J. M. Feliu, Langmuir 1997, 13, 6287. [14] T. J. Schmidt, R. J. Behm, B. N. Grgur, N. M. Markovic, P. N. Ross, Langmuir 2000, 16, 8159. [15] T. J. Schmidt, V. R. Stamenkovic, C. A. Lucas, N. M. Markovic, P. N. Ross, Phys. Chem. Chem. Phys. 2001, 3, 3879. [16] S. P. E. Smith, H. D. Abrunƒa, J. Electroanal. Chem. 1999, 467, 43.

CHEMPHYSCHEM 2003, 4, 193 ± 199

Oxidation at a PtBi Surface

ELECTROCHEMISTRY SPECIAL

[17] G. Wittstock, A. Strubing, R. Szargan, G. Werner, J. Electroanal. Chem. 1998, 444, 61. [18] U. Hamm, D. Kramer, R. S. Zhai, D. M. Kolb, Electrochim. Acta 1998, 43, 2969. [19] S. P. E. Smith, H. D. Abrunƒa, J. Phys. Chem. B 1998, 102, 3506. [20] S. G. Sun, J. Clavilier, J. Electroanal. Chem. 1987, 235, 95. [21] T. Iwasita, U. Vogel, Electrochim. Acta 1988, 33, 557. [22] S. C. Chang, M. J. Weaver, J. Phys. Chem. 1990, 94, 5095. [23] D. S. Corrigan, L. W. Leung, M. J. Weaver, Anal. Chem. 1987, 59, 2252. [24] S. C. Chang, L. W. H. Leung, M. J. Weaver, J. Phys. Chem. 1990, 94, 6013. [25] S. C. Chang, A. Hamelin, M. J. Weaver, J. Chim. Phys. Phys.-Chim. Biol. 1991, 88, 1615. [26] S. G. Sun, Studying Electrocatalytic Oxidation of Small Organic Molecules with In-Situ Infrared Spectroscopy in Electrocatalysis (Eds.: J. Lipkowski, P. N. Ross), Wiley, NY, 1998, pp. 243. [27] S. M. Stole, D. D. Popenoe, M. D. Porter, Infrared Spectroelectrochemistry: A Probe of the Molecular Architecture of the Electrochemical Interface in

CHEMPHYSCHEM 2003, 4, 193 ± 199

[28] [29] [30] [31] [32] [33] [34] [35] [36]

Electrochemical Interfaces: Modern Techniques for In-Situ Interface Characterization (Ed.: H. D. Abrunƒa), VCH, New York, NY, 1991, pp. 339. S. Bruckenstein, J. Corneau, Discuss. Faraday Soc. 1973, 56, 285. O. Wolter, J. Heitbaum, Ber. Bunsen-Ges. 1984, 88, 2. J. Willsau, O. Wolter, J. Heitbaum, J. Electroanal. Chem. 1985, 185, 163. S. P. E. Smith, K. F. Ben-Dor, H. D. Abrunƒa, Langmuir 1999, 15, 7325. N. N. Zhuravlev, A. A. Stepanova, Sov. Phys.-Crystallogr. 1962, 7, 241. P. Durussel, R. Massara, P. Feschotte, J. Alloys Compd. 1994, 215, 175. S. G. Sun, Y. Y. Yang, J. Electroanal. Chem. 1999, 467, 121. Y. Y. Yang, Z. Y. Zhou, S. G. Sun, J. Electroanal. Chem. 2001, 500, 233. J. W. Long, R. M. Stroud, K. E. Swider-Lyons, D. R. Rolison, J. Phys. Chem. B 2000, 104, 9772.

Received: August 27, 2002 [C 496]

199