STATE OF UPD-TIN ON A PLATINUM STUDIED WITH SURFACE RAMAN SPECTROSCOPY HOLZE
Institute of Physical Chemistry. University of Bonn, Wegelerstr. 12, 5300 Bonn 1. West Germany (Received
18 June 1987; in revised
Abstract-The oxidation state and chemical composition of the coordination shell of upd-tin atoms or ions present on the platinum electrode in acidic electrolyte have been studied using platinum and tin electrodes with Surface Ran-tan Spectroscopy, for reference purposes tin sulfate containing solutions were studied for the first time with Normal Raman Spectroscopy. Evidence was found suggesting the tin ad-ion present on the electrode to becoordinated with sulfate ions as in the solution phase. This implies an oxidation state of 2 + or 4+ for the tin ad-ion as the most likely one under these conditions.
INTRODUCTION In recent years the electrocatalysis of several electrochemical processes like eg oxygen reduction or methanol oxidation on metal electrodes with ad-atoms or ad-ions (upd-metal) of a different metal present on the electrode has been studied intensively, the results have by Kolb[ 11, Adzic, been reviewed Kokkinidis and most recently by Andricacos et aI.. Although the most important feature of an electrode/upd-metal system is the enhanced catalytic activity for a given process any explanation of this enhancement will include a consideration of the oxidation state of the upd-metal and conceivable changes of the oxidation state within the electrocatalytic process under investigation. This is of particular importance when the enhanced activity is either based upon changed adsorption properties for eg OH--adsorption or upon the assumption of a redox mediation of the electrocatalytic process involving various redox states of the ad-atom. In the case of upd-tin present on a platinum electrode in acidic electrolyte solution the effect upon oxidation of methanol is particularly pronounced and well known[S, 63. Several methods have been applied to identify the oxidation state of tin. With the exception of Bowles and Cranshaw, who used in-situ MoDbauer Spectroscopy, all papers published so far are essentially based on electrochemical measurements and therefore the results varying from oxidation state 0 to 4+ are based upon more or less indirect conclusions from these measurements. Bowles and CranshawC7) found metallic tin at all electrode coverages investigated, the degree of platinum-tin interaction derived from the isomer shift depending upon the coverage. Based upon electrochemical measurements Janssen and Moolhuysen arrived at the same conclusion. Vassiliev er a/.[93 suggested a formal oxidation state 1 of the tin ad-ion, since this is an unknown species in the solution phase the actual oxidation state should be partially 2 and 0. Sobkowski et al.[lO] concluded, that 353
Sn4+ is slowly adsorbed on the electrode and reduced to Sn’+, which is still coordinated with solution phase species. Szabo[ll] found, that St-?+ disproportionates during adsorption on a platinum electrode and forms a monolayer very similar to a layer formed during adsorption from a SnCl, solution. Upon anodic polarisation tin is dissolved as Sn4+. From these results the author concludes that tin is present in the oxidation state 0. In order to obtain further more direct information on the oxidation state and chemical coordination of the tin on the platinum electrode Surface Raman Spectroscopy is used in the work reported here. Since any platinum-tin vibrations are difficult to detect (only recently in situ IR spectroscopy has been applied to the study of lithium ad-atom deposition on gold by Li et al.) the vibrational spectra of sulfate ions present on the electrode surface have been studied. Vibrational spectra of the sulfate ion in various states (solvated various sulfate containing SO:-, HSO,, H,SO,, complexes) have been published and could be used as reference[13-151. The adsorbed sulfate ion on a metal electrode has been studied using Electrode Modulated Infrared Spectroscopy (EMIRS) by Bewick et a/.[16181 on a gold electrode. Surface Raman spectra of ionic species adsorbed on a platinum electrode have been studied by Cooney et al. with adsorbed J-, CO and by Heitbaum for an intermediate of the oxidation of phenylhydrazine. Whether any surface enhancement is operative under the experimental conditions employed by these investigators is still under discussion. In this paper Surface Raman Spectroscopy is used to study the adsorption of sulfate and hydrogensulfate ions on the tin and platinum electrode in sulfuric acid electrolyte and on a platinum electrode in an electrolyte containing Sn4+ ions. Because of the capability of sulfate ions to coordinate with metal ions like eg Sn4+ and to adsorb specifically on the electrode and to show corresponding shifts of the molecular vibration modes indicating the type of coordination this ion is used as a probe to test the electrode surface.
HOWE AND BARBARA
EXPERIMENTAL Experiments were done using a tin electrode (tin 99.999 “/,, Fluka) and a platinum electrode (99.999 %, Heraeus) of 5 mm dia. embedded in epoxy (Araldite, Ciba-Geigy) polished with abrasive paper and AI,O, down to 1 pm. Electrolyte solutions were prepared from sulfuric acid (suprapure, Merck), Sn(SO,), x 2 H,O (Ventron, Alfa Products) and 18 MOhm water purified in a Millipore unit. They were carefully deaerated before each measurement. Additional solutions for Normal Raman spectroscopy of SnSO, (2.8 mM, neutral, Riedel deHaen, Merck, G.R.), Sn(SO,), (7.7 mM, acidic), K,SO, and concentrated sulfuric acid were prepared from p.A. chemicals or practical grade reagents. All spectra and cyclic voltammograms were recorded at room temperature. The standard electrochemical equipment incl.uded a BANK 61 TR potentiostat, a X-Y recorder HOUSTON-Instruments 2000 and a function generator built in our laboratory. A platinum electrode separated by a glass frit from the cell main compartment was used as a counter electrode, a reversible hydrogen electrode prepared according to Will er a1. was used as reference electrode, all potentials are quoted us this electrode. For spectroelectrochemical as well electrochemical measurements a cell previously described was used. The power of the exciting laser beam (SPECTRA-PHYSICS 164-00,488 nm) was measured at the position of the cell with a laser power meter (Laser Instrumentation Ltd.) and adjusted to 40 mW, in case of the Normal Raman spectra (of solutions) a cell and collection optics supplied by CODERG were used, the incident power was slightly higher. All spectra were recorded on a CODERG LRT 800 triple monochromator spectrometer equipped with a photomultiplier tube (EM1 6256 S) and a dc-amplifier at a scan rate of 50cm-’ min-‘. The spectral resolution was adjusted to 4cm-’ if not indicated otherwise, the given wavenumbers are believed to be correct within + 3 cm-‘.
Fig. 1. Cyclic voltammogramof a polished polycrystallinetin electrode in 0.1 N H,SO,, scan rate 20mVs-‘, nitrogen purged,arrowsindicatepotentialswhereSRS weremeasured.
Tin electrode in H,SO,-soft&ion The cyclic voltammogram of a tin electrode in 0.1 N H,SOI solution is presented in Fig. 1, the arrows indicate the potentials, where in situ Surface Raman spectra were measured. A typical Raman spectrum measured at E= -265 mV,,, obtained during the anodic scan after a few oxidation reduction cycles is given in Fig. 2. The broad feature centered around 420 czr - ’ was found to be due to luminescence of the cell window, which is very close to the working electrode (approx. 0.2 mm), this feature was found in ex situ spectra of a freshly polished tin electrode, too (see below) and has been reported in the literature. The peaks observed in the in situ spectrum at 485 and 930 cm- I correspond’ to modes of adsorbed HSO; ions, they are shifted due to ion-surface interaction (for details see Table 1, normal Raman spectra of reference solutions recorded for the purpose of comparison and the band assignment based on literature data[13-151
1000 Raman shift /cm’
Fig. 2. SurfaceRaman spectrum of a cycled tin electrodeat E = -265 mV,,, in 0.1 N H,SO, solution, excitation wavelength 488 nm, resolution 10 cm- ’
are collected therein). The peak at 810cm-’ is presumably caused by some not yet identified surface species on the tin electrode surface also observed in exsitu experiments (see below). When the electrode potential is shifted more cathodically only weak bands at 810cm-’ (see below) and 1064 cm-’ are found, the latter one may be due to a mode of weakly adsorbed HSO, This is in partial agreement with electrosorption studies by Barteniev et al. , who found no specific adsorption of sulfate ions from dilute neutral solution of Na,SO, on polycrystalline tin; other results
Oxidation state of upd-tin on a platnium electrode
reviewed by Galus do not entirely support this result, but indicate weak adsorption, especially from acidic solutions similar to those used in this work. In theexsituspectrumaveryweakpeakat 810 cm-’ was found, its intensity increased with the time of exposure to air. This indicates that the peak may be due to surface oxides or hydroxides, since SnSO, or Sn(OH), do not show any band in this range[13-151 this peak may correspond to a vibration of a lower valency hydroxide or oxide, which has not been studied with Normal Raman Spectroscopy so far. Infrared studies of tin oxide SnO showed an absorption band at 81 l-816 cm-’ (the position depends upon experimental conditions), since the corresponding stretching mode is both IR and Raman active (SnO belongs to the point group C,,) the band observed in this work at 810cm-’ may be caused by this mode. Platinum-electrode
The Surface Raman spectrum of the platinum electrode (polished with 1 pm Al,O,, electrochemically roughened with fast cyclic voltammetry until a brownish colour appeared on the etectrode (roughness factor approx. 2.7)) at E = +900 mV,,, in 1 N H,SO, solution is given in Fig. 3. Bands at 895, 980 and 1060 cm -’ are observed, based on a comparison with the NR spectrum of 1 N H,SO,, which contains mainly HSO;-ions, they can be attributed to modes of weakly adsorbed HSO;-ions. A very rough estimate of the Raman signal intensities obtained in this experiment with those observed in the corresponding NR spectra does not yield evidence for any substantial Surface Enhancement effect beyond the contribution of surface roughening (see also). When the electrode potential is shifted to more cathodical values
600 Ramanshlf t /cm1
Fig. 3. Surface Raman spectrum of a polished platinum electrode, electrochemically roughened. in 1 N H*SO, electrolyte solution at E = 900 mV,,, (1) and E = 200 mVRHE (2). excitation wavelength 488 nm.
ERHElmV Fig. 4. Normalised Raman signal intensity of the S-O-stretch mode (around 1060 cm- ‘) as a function of electrode potential (the signal was normalised by setting the highest intensity equal to one) for various electrode-electrolyte combinations: -o platinum in 1 N H,SO,-solution. +-+ platinum electrode in an electrolyte solution of 1 N H,SO, + 0.244 mM Sn(SO&, x - x platinum electrode with adsorbed tin in a 1 N H,SO,-solution.
these bands decrease in intensity sharply (see Fig. 4); this supports the assumption, that these bands are due to species on the electrode surface and not to solution phase species. The potential dependent signal intensity and consequently the degree of coverage with hydrogensulphate ion is in agreement with simple electrostatic considerations and earlier radiotracer studies by Horanyi et al.. Platinum-electrode
in Sn4+ containing
;’ I 0
I I 1200
ERHE ‘m” Fig. 5. Cyclic voltammogram ofa polished, electrochemically roughened platinum electrode in a 1 N H,SO, + 0.244 mM Sn(SO& solution. scan rate 50 mVs-‘, nitrogen purged, (---) supporting electrolyte. (-) Sn(SO& added.
Electrode potential E > Ep”., The cyclic voltammogram of the roughened p atmum electrode m the supporting 1 N H,SO, electrolyte solution after addition of 0.244mM Sn(SO,), is given in Fig. 5, it resembles closely previously published CVs see ref and [IO] and shows only minor distortions caused by IR-drop inside the spectroelectrochemical cell. At the potentials marked in the CV Raman spectra were recorded. At anodic potentials (E r Epzc) the spectrum is very similar to the spectrum of 1 NH,SO,, only minor band shifts (see Table 1) are observed (eg see Fig. 6, spectrum recorded at E = 1191 mV,,,), this indicates weak (unspecific) adsorption. Electrode Potentials E < E, (Evidence of the positive oxidation state of the adsorbed tin ion) At less positive potentials, where a sharp decrease of signal intensity was observed in the tin-free solution, the spectrum remains almost constant in signal intensity (see Fig. 4), but the band observed previously at becomes weaker, whereas the peak at 890cn-’ 908 cm-t increases. This peak is close to the one observed at 915cm-’ in Sn(SO& solution. Sn4+ is present in this solution containing large excess of sulphate and hydrogen-sulfate ions mainly as SnSO: + ion. This ion will adsorb at cathodic potentials , which has been determined (with respect to the E tensammetry[%] to be EpzF = 584+/ with ), it obviously keeps its coordmated sul-20 mV,,, fate ion, since otherwise the strong Raman bands found at these potentials attributed to vibrations of the sulfate ion cannot be explained properly. A similar coadsorption of sulphate ions on a rough platinum
Fig. 6. Surface Raman spectrum of a polished, electrochemitally roughened platinum electrode. in a 1 N H2SOI and 0.244 mM Sn(SO,), solution at E = 1191 mV,,,
electrodeat cathodic potentials in the presence of Cd’+ has been observed by Horanyi. By comparison of these Surface Raman spectra recorded at potentials negative to the EpIx with the Surface Raman spectrum obtained with a tm electrode the presence of a tin adsorbate with tin in the oxidation state 0 and surface properties similar to those of a bulk tin electrode can be ruled out, since the respective spectra are distinctly different (compare Fig. 2 and Fig. 6). The similarity of the SR spectra obtained with the platinum electrode in a tin free solution at anodic potentials (see Fig. 3) and of the spectrum of the platinum electrode measured at cathodic potentials in a tin containing solution suggests, that the same ion is adsorbed on the electrode in both cases, this is conceivable only, if adsorption of the SnSO:+ ion or a
Oxidation state of upd-tin on a platnium electrode similar species containing sulfate or hydrogensulfate (the presence of the SOH-stretching mode indicates the presence of at least some of the latter ions on the electrode) takes place. In order to check the possibility of the tin ion being present in the oxidation state 2 + the NR spectrum of a saturated solution of S&O, was recorded. The spectrum (neutral solution, see Fig. 7) is markedly different, it shows in particular a splitting of the degenerate bending mode of the 0 = S = 0 unit at 613 cm-’ in neutral sulfate solution (see NR spectrum of Na,SO,solution, Table 1) into two bands located at 559 and 632 cm-‘. This splitting suggests, that Snzt is coordinated with sulfate ions in neutral solution, the rather low solubility in neutral solution agrees with this assumption. A similar splitting of degenerate sulfate modes has been observed with crystalline Sn(S0,) by Steger and Schmidt. These bands are not observed in Surface Raman spectra under the experimental conditions described above. On the contrary a NR spectrum of a solution of 4.6 mM SnSO, in 1 N H,SO, resembles exactly a NR spectrum of 1 N H,SO, alone, the influence of the Sn2 + ion cannot be detected in the presence of the large excess of HSO,-ions. Thus no distinction between the oxidation state 4+ and 2+ based on NR and Surface Raman spectra alone is possible at the time being, an electrochemical study of the electrosorption of tin ions on a platinum electrode is presently carried out, the preliminary results indicate an oxidation state of the adsorbed tin depending on the potential of the platinum electrode, a full discussion will be published soon. Platinum electrode with tin adsorbate in a tin free H,SO,-electrolyte solution. In order to check the catalytic activity of Sn4+ adsorbed from an acidic solution of Sn(S04)2 on a platinum electrode the
solution of electrode was exposed to a 0.244mM Sn(SO,), at E = 400 mVa,, for 15 min, removed under potential control from the cell and carefully washed with water. The electrode was inserted into a cell containing an electrolyte solution of 1 M methanol and 1 N H,SO,, the stationary cyclic voltammogram recorded subsequently is in close agreement with CVs recorded in this electrolyte solution with Sn4+ present in the solution phase. A spectroscopic examination of the platinum electrode with adsorbed tin in a tin-free solution gave essentially the same results as obtained previously with a tin-contaning solution; in particular the high Raman signal intensity of the band around 1060cm-’ at cathodic potentials (see Fig. 4). Spectra were recorded within less than ten minutes after transfer of the electrode to the spectroelectrochemicaal cell, within this time no time dependence of the scattered Raman signal at the various working electrode potentials was observed. This observation is reasonable if the slow (desorption times of several minutes up to more than two hours have been reported[ 10,351) and incomplete desorption of the adsorbed tin, which has been reported previously, are considered. Thus it can be assumed that the tin ion adsorbed on the platinum electrode is the catalytically active species. The electrosorption behaviour of Sn’+, which has been claimed to show similar catalytic activity as Sn4+ (see above), remains to be studied. The results obtained so far with Sn4+ present in the solution phase and/or adsorbed on the electrode surface indicate strongly, that a tin ion in the oxidation state 4+ or 2f is adsorbed with its coordinate sulphate ion. Acknowledgements-This work was in part supported by a Community under contract grant of the European EN3EOO72D. REFERENCES M. Kolb, in Advances in Electrochemistry and Elecrrochemical Engineering, (Edited by H. Gerischer and C. W. Tobias), Vol, 1. Wiley, New York (1978). in Electrochemistry and 2. R. R. Adzic, in Advances Electrochemical Engineering. (Edited by H. Gerischer and C. W. Tobias) Vol. 13, Wiley, New York (1984). 3. G. Kokkinidis, J. electrounal. Chem. 201, 217 (I 986). 4. P. C. Andricacos, M. Krishnan and D. Rath, Ext. Abstracts, 170th Electrochem. Sot. Meeting, San Diego, USA, 19-24. 10 (1986). 5. K. J. Cathro, .I. electrochem. Sot. 116, 1608 (1969). 6. M. Shibata and S. Motoo, J. electroannl Chem. 209, 151 (1986). 7. B. J. Bowles and T. E. Cranshaw, Phys. L&r. 17, 258 (1965). 8. M. M. Janssen and J. Moolhuysen, J. Cotal. 46, 289 (1977). 9. Yu. B. Vassiliev, V. S. Bagotzky, N. V. Osetrova and A. A. Mikhailova, J. ebxtroanal. Chem. 97, 63 (1979). 10. J. Sobkowski, K. Franaszczuk and A. Piasecki, J. Electroonol. Chem. I%, 145 (1985). 11. S. Szabo, J. Electroanal Chem. 172, 359 (1984). 12. J. Li, J. Daschbach, J. J. Smith, M. D. Morse and S. Pons, J. electroanal. Chem. 209, 387 (1986). in 13. H. Siebert. Amvendungen der Schwingwgsspektroskopie der anorganischsn Chemie, Springer-Verlag, Berlin 1966. 13a. H. Siebert, Z. Anorg. A&. Chem. 289, 15 (1957). 14. K. Nakamoto, infrared Spectra of Inorganic and 1. D.
Ramon shtftlcm-’ Fig. 7. Normal
Raman spectrum of a saturated (c = 2.8 mM).
solution of S&O,,
RUDOLF HOUE AND BARBARA BITTINS-CATTANEO
Coordination (1986). 15. S. D.
19. 20. 21. 22.
Ross, Inorganic Infrared and Raman Spectra, Hill, London (1972). A. Bewick and S. Pans, in Advances in Infrared and Raman Spectroscopy, (Edited by R. J. H. Clark and R. E. Hester), Vol. 12, Wiley, New York (1985). and Physics of A. Bewick, in The Chemistry Electrocafalysis, (Edited by J. D. E. McIntyre, M. J. Weaver and E. B. Yeager). The Electrochemical Society, Princeton (1984). A. Bewick, in Trends in Interfacial Electrochemistry. (Edited by A. F. Silva), D. Reidel Publishing Comp. p. 331 (1986). R. P. Cooney, P. J. Hendraand M. Fleischmann, J. Raman Spec. 6, 264 (1977). R. P. Cooney, M. Fleischmann and P. J. Hendra, J. Chem. Sot. Chem. Comm. 7,235 (1977). J. Heitbaum, 2. Phys. Chem. NF, 105, 307 (1977). S. Efrima, in Modern Aspects of Electrochemistry, 16, (Edited by B. Conway, R. White and J. GM Bockris), Plenum Press, New York (1985). F. G. Will and H. J. Hess, J. electrochem. Sot. 120, 1 (1973).
Chem. (in print). 24. R. Holze, J. Electroanal. 25. T. M. Cotton, S. G. Schultz and R. P. Van Duyne, J. Am. Chem. Sot. 102, 7960 (1980). 26. V. J. Barteniev, E. S. Sevastijanov and D. I. Leikis Elektrokhimiya 6, 1868 (1970). 27. 2. Galus, in Encyclopedia of Electrochemistry of the Elements, (Edited by A. J. Bard), M. Dekker, New York (1975). 28. G. Horanyi, Electrochim. Acta 25, 43 (1980). 29. E. W. Abel, in Comprehensive Inorganic Chemistry Vol. 2 (Edited by J. C. Bailar, H. J. Emeleus, R. Nyholm and A. F. Trotman-Dickinson), Pergamon Press, Oxford (1973). 30. R. H01m and M. Beltowska-Brzezinska, J. electroana~. Chem. 201. 387 (19861. 31. G. Horanyi, J. Solt andG. VBrtes,J. electroamd. Chem. 32, 271 (1971). 32. E. Steger and W. Schmidt, Ber. Buns. Phys. Chem. 68, 102 (1964). 33. J. S. Gdgen and M. J. Ricks, J. Chem. Phys. 53,896 (1970). 34. B. Bittins-Cattaneo and T. Iwasita, ElectrocataIysis OJ Methanol Oxidation by Tin Adsorbed on Platinum, (to be oublished). 35. &I. M. Stefenel, T. Chierchieand C. Mayer, Z. Phys. Chem. NF, 135, 251 (1983).