Search

The surface-enhanced Raman spectroscopy (SERS) properties of thiocyanate adsorbed on platinum, palladium, rhodium, and iridium electrodes, formed by pinhole-free ultrathin film deposition onto SERS-active gold, are compared with corresponding data obtained on copper, silver, and gold, and on underpotential (upd) silver and mercury layers on gold, to assess the sensitivity of the anion coordination modes to the chemical nature of the metal surface. A crucial virtue of SERS for this purpose lies in the observability of each of the intramolecular adsorbate modes plus the metal surface−adsorbate stretching vibration, νm-a. The latter modes, inaccessible to other in-situ vibrational methods, provide particularly straightforward insight into metal-thiocyanate binding. In contrast to gold and upd mercury surfaces, which favor N-coordination only at far-negative potentials, both S- and N-binding modes are fingerprinted on Pd, Rh, and Pt surfaces from the characteristic νm-a bands. These bands, whose relative frequencies are sensitive primarily to the lead-in atomic mass, are observed throughout the range of potentials (typically −0.4 to 0.4 V vs the saturated calomel electrode) where “reversible” potential-dependent spectral behavior was obtained. Such coexisting modes of surface coordination are also indicated from the appearance of a pair of C−S stretching (νCS) vibrations indicative of S- and N-binding from the known behavior of bulk-phase metal complexes. Less diagnostic in this regard is the C−N stretching (νCN) vibration, which exhibits only a single broad band at ca. 2115−2135 cm^-1 over the above-mentioned range of potential on the Pt-group surfaces. The virtually potential-independent νCN frequencies observed on these metals, (dνCN/dE) ≤ 10 cm^-1 V^-1, contrast with the larger “Stark-tuning” slopes found on silver and copper. The potential-insensitive adsorbate frequencies and coordination modes, together with the higher νm-a frequencies compared with those on the coinage metals, are consistent with the occurrence of largely covalent transition-metal surface bonding.

The oxidation of five Pt-group metals&sbd;platinum, palladium, iridium, rhodium, and ruthenium&sbd;is examined by means of surface-enhanced Raman spectroscopy (SERS) in aqueous electrochemical and gaseous dioxygen environments as a function of electrode potential and temperature, respectively, with the objective of intercomparing systematically the conditions required for surface oxide formation and discerning the reaction mechanisms involved. The SERS strategy, utilizing ultrathin Pt-group metal films electrodeposited on a gold substrate, enables monolayer-level metal oxide vibrational spectra to readily be obtained in both the electrochemical and gaseous environments. The SER spectra obtained during positive- and then negative-going potential excursions in aqueous 0.1 M HClO4 display metal−oxygen vibrational bands signaling anodic oxide formation and subsequent removal at potentials consistent with corresponding voltammetric data. The nature of the amorphous oxides (or hydroxides) formed is deduced by comparison with bulk-phase metal oxide Raman spectra. The onset potentials for surface oxide formation are comparable to the thermodynamic potentials for the bulk-phase metal oxides. In contrast, the onset of surface oxidation even in ambient-pressure dioxygen uniformly requires elevated temperatures, ≥200 °C for each metal except for iridium, where oxide formation occurs at ca. 100 °C. While spectral differences are evident, especially on palladium and ruthenium, the nature of the oxides formed in the electrochemical and gaseous systems is largely similar. The highly activated nature of the gaseous O2 oxidations is consistent with literature reports for Pt-group surfaces in ultrahigh vacuum as well as higher-pressure conditions. Likely reasons for the markedly more efficacious metal electrooxidations are discussed. Thermodynamic factors are not responsible, since the free-energy driving forces for the gaseous O2 oxidations are larger than for the electrochemical reactions at the applied potentials where surface oxidation for the latter processes proceeds at room temperature. The electrostatic driving forces for oxygen incorporation into the metal lattice (via “high-field ion transport”) are also typically more favorable for the gaseous systems, as evidenced by a comparison of the metal−solution and metal−gas surface potentials. The intrinsically more facile electrochemical processes thereby deduced are attributed to the occurrence of direct oxide production via a metal−oxygen place-exchange mechanism, expedited by interfacial solvation and therefore being energetically unfavorable in the anhydrous gas-phase environment. Other factors, such as the formation of precursor chemisorbed oxygen, are also considered.

The utility of surface-enhanced Raman scattering (SERS) for characterizing phonon and related interfacial vibrations within ultrathin semiconductor layers on gold substrates is demonstrated, specifically for cadmium chalcogenide (CdSe, CdS) films formed by electrochemical atomic layer epitaxy. The occurrence of quantum-confinement effects for CdSe films less than ca. 10 monolayers thick is clearly evident from the observed ca. 12 cm^-1 phonon redshifts and bandshape changes. Milder phonon redshifts are seen for the ultrathin CdS films, suggesting the presence of offsetting blueshifts from lattice compression. The observed sensitivity of the phonon spectra to the order of initial Cd/Se or Cd/S monolayer electrodeposition, along with the retention of Au−Se and Au−S vibrational bands upon layer growth, signal the importance of the initial chemisorbate to the epitaxial film structure and vibronic properties.

Model dipole−dipole coupling calculations of infrared spectra for various ordered CO packing arrangements on Pt(111) have been undertaken as a function of the spatial dimensions of isolated nanoscale adlayer domains. Comparisons of some of these predictions with coverage (θCO)-dependent infrared spectra obtained upon partial adlayer oxidation on low-index Pt-group surfaces in aqueous electrochemical environments enable approximate estimates of densely packed CO island sizes to be obtained. The evidence for CO island formation, described in several earlier reports, is based on the observation of C−O stretching bands, νCO, whose frequency, band shape, and relative intensity exhibit markedly smaller alterations upon progressive electrooxidative removal in comparison with corresponding νCO frequency-coverage data obtained for increasing θCO by means of dosing with dilute CO solutions. The dipole-coupling calculations indicate a significant sensitivity of the νCO frequency and related parameters to island size even for domains containing sizable numbers (100−200) of CO molecules, suggesting the value of the procedure for providing at least rough estimates of nanoscale island dimensions. While the domain size-dependent νCO spectra should be dependent inevitably upon the microscopic CO packing geometry, the (relative) changes in the νCO frequency for atop (i.e. terminal) adsorbate are predicted to be dependent on the island dimensions, yet relatively insensitive to the adlayer structure, suggesting the practical utility of this parameter for island size diagnosis. The resulting estimates of average island dimensions, along with their stability as gleaned from experimental time-dependent spectra, are seen to vary widely on different low-index Pt-group electrodes. Interestingly, the metastable CO domains formed on Pt(100), estimated to contain 20−50 atop CO's, have dimensions roughly compatible with the restructured substrate “mesas” seen to form by scanning tunneling microscopy.

The nature and magnitude of corrections to published estimates of the potentials of zero charge Epzc for Pt(111)−aqueous interfaces obtained by the “CO charge-displacement” strategy are assessed by comparing such in-situ electrochemical data with surface charge−potential information extracted from work function−surface composition data for related interfaces in ultrahigh vacuum (UHV). The corrections involve estimating the excess metal charge densities remaining upon displacement of the aqueous inner layer by chemisorbed CO. While the resulting adjustment to the so-called potential of zero total charge %@mt;sys@%%@ital@%E%@rsf@%%@sx@%pzc%@be@%t%@sxx@%%@mx@% for the Pt(111)−aqueous 0.1 M HClO4 interface is only small (ca. 25 mV), a larger correction to the corresponding “free-charge” value %@mt;sys@%%@ital@%E%@rsf@%%@sx@%pzc%@be@%f%@sxx@%%@mx@% is deduced, yielding %@mt;sys@%%@ital@%E%@rsf@%%@sx@%pzc%@be@%f%@sxx@%%@mx@% ≈ 0.2 V versus SHE. A closely concordant estimate of %@mt;sys@%%@ital@%E%@rsf@%%@sx@%pzc%@be@%f%@sxx@%%@mx@% is also extracted from UHV-based work function data. The substantial discrepancies in such Epzc values for the Pt(111)−acidic aqueous interface with those extracted by means of an ex−situ/electrode immersion procedure are also briefly considered in the light of the above analysis.

The electrode potential-dependent formation of oxygen species on copper in noncomplexing aqueous media, encompassing oxide phase films and adsorbed oxygen/hydroxide, are explored at different pH values by means of surface-enhanced Raman spectroscopy (SERS). This technique provides a monolayer-sensitive in- situ vibrational probe, which can follow potential-dependent surface speciation on voltammetric or longer time scales. In alkaline NaClO4 electrolytes (pH 13), the cyclic voltammetric peaks associated with copper oxide phase-film formation and removal are correlated quantitatively with simultaneously acquired SER spectral sequences. The latter indicate the sequential formation of Cu2O and then mixed Cu2O/Cu(OH)2 layers, diagnosed by the appearance of metal−oxygen lattice vibrations at 625/525 and 460 cm^-1, respectively. The potential-dependent speciation is in concordance with the Pourbaix diagram, certifying the “bulk-phase” nature of the films. The Raman band intensity−film thickness correlation (the latter deduced from the voltammetric Coulombic charges) indicate that the vibrational spectral responses are limited to the first 15−20 monolayers, consistent with earlier SERS observations and theoretical predictions. Weaker bands at ca. 800 and 460 cm^-1 are discernible at more negative potentials, suggestive of hydroxide adsorption. Similar, although thinner, oxide films were deduced to form in neutral 0.1 M NaClO4. In the additional presence of chloride under these conditions, a potential-sensitive competition between the formation of a CuCl and a more passivating Cu2O phase film was evident from SERS. While oxide phase films are absent on copper in 0.1 M H2SO4 and 0.1 M HClO4, an adsorbed oxygen species was nonetheless detected from a broad SERS band at ca. 625 cm^-1. This feature, which was deduced to involve oxygen rather than hydroxyl from an absence of a frequency shift upon H/D solvent isotopic substitution, is evident throughout most of the “polarizable potential” region on copper in acid, ca. −0.7 to −0.1 V vs SCE. The likely nature and reasons for its remarkable prevalence on copper in acidic media are discussed with reference to the recent literature.

In-situ infrared reflection−absorption spectra (IRAS) measured for the C−O stretch (νCO) of carbon monoxide on ordered Ir(111) in aqueous 0.1 M HClO4 as a function of the dosed coverage, θCO, are compared with corresponding extant IRAS data on Ir(111) in ultrahigh vacuum (UHV) in order to elucidate the physical influences exerted by the electrochemical double layer on chemisorbate vibrational properties. Unlike most metal surfaces, CO chemisorption on Ir(111) exhibits a lone νCO band consistent with a single (atop) binding site in both electrochemical and UHV environments over the entire coverage range up to saturation and features sizable θCO-dependent changes in the νCO frequency and other spectral parameters, reflecting adsorbate−adsorbate interactions. This surface−chemisorbate system therefore provides an unusual opportunity to explore in detail solvation and related environmental factors at the same stable (unreconstructed) metal substrate in the absence of additional “chemical” effects associated with coverage- and/or solvent-induced alterations in CO binding geometry. The nature of the vibrational interactions is assessed in part by comparing the θCO-dependent spectral parameters with predictions from a numerical dipole-coupling analysis. The electrochemical νCO peak frequencies, %@mt;sys@%ν%@sx@%CO%@be@%p%@sxx@%%@mx@% , exhibit a marked dependence upon θCO that is comparable to that observed for the Ir(111)−UHV system. Similar saturation coverages, θCO ≈ 0.7, are also attained. Furthermore, the θCO-dependent %@mt;sys@%ν%@sx@%CO%@be@%p%@sxx@%%@mx@% values are approximately coincident once the differences in surface work function in these two environments are taken into account by means of a Stark-tuning analysis. However, the integrated νCO band absorbance Ai displays a notably different dependence on θCO at the electrochemical and UHV-based interfaces, the former plot having a markedly less nonlinear shape than the latter. These differences can be understood in terms of solvent dielectric-screening effects, which attenuate the former Ai values at low θCO. The bandwidths, Δν1/2, at the electrochemical interface are markedly (2−3-fold) larger at low and moderate θCO values compared to the UHV case. The former Δν1/2−θCO behavior is indicative of stochastic broadening associated with microscopic variations in oscillator density incurred by random chemisorption. The Lorentzian νCO band shapes (for θCO > 0.4) suggest that solvation-induced inhomogeneous line broadening is less important. The smaller Δν1/2 values and curvilinear Δν1/2−θCO behavior observed for the UHV system are consistent with local chemisorbate island formation. Nevertheless, for coverages approaching saturation (θCO ≈ 0.7) the electrochemical Δν1/2 values diminish to a value, 7 cm^-1, only slightly broader than that (5 cm^-1) observed for the solvent-free UHV system. The Ai value attained at saturation for the electrochemical interface approaches that observed for the UHV system, the remaining dissimilarity being roughly consistent with the likely differences in the IRAS optical geometry.

The transient reduction kinetics of surface rhodium oxide (Rh2O3) by gaseous H2 and CO have been probed in situ by surface-enhanced Raman spectroscopy (SERS). The Rh surfaces are ultrathin films electrodeposited onto SERS-active gold, enabling surface vibrational spectroscopic information to be obtained with high temporal resolution (≈1 s) at elevated temperatures (up to 500 °C) and under ambient-pressure flow-reactor conditions. Surface Rh2O3 is formed by heating Rh in flowing O2 at 1 atm and fingerprinted by an intense 530 cm^-1 νRh-O feature. The reduction of such oxidized surfaces upon sudden exposure to either H2 or CO over a range of partial pressures (1−760 Torr) was monitored from the decay kinetics of the νRh-O band intensity. Surface oxide is readily reduced by both reductants, although striking differences in the observed kinetic behavior indicate the occurrence of distinct reaction pathways. In the case of H2, at low partial pressures (≤7.6 Torr) below 200 °C a temperature-dependent induction period is observed prior to rapid first-order oxide reduction. Such “autocatalytic” behavior is indicative of a “nucleation/growth” mechanism, where H2 dissociatively adsorbs to form reaction centers, followed by facile reaction between adsorbed (or lattice) oxygen and (possibly subsurface) hydrogen atoms. Immeasurably fast H2-induced oxide reduction, however, occurs at higher temperatures (≥200 °C) and pressures (≥76 Torr). In contrast, CO-induced oxide reduction was found to be substantially (at least 10-fold) slower than with H2 at similar pressures and temperatures, yet no induction period was detected. At lower temperatures (≤250 °C), the oxide reduction kinetics by CO exhibit fast initial removal followed by a much more sluggish zero-order response. Such “autoinhibited” kinetic behavior, along with the lack of an appreciable CO pressure dependence, suggests that oxide removal is hindered by the extensive formation of adsorbed CO, which is observed to develop rapidly under these conditions from the characteristic Rh−CO and C−O stretching vibrations. This mechanism is further supported by the first-order kinetic response observed throughout oxide removal at temperatures (≥300 °C) where buildup of adsorbed CO is not detected. The transient kinetics for CO-induced oxide reduction are linked to the well-known poisoning effect that Rh surface oxidation exerts on catalytic CO oxidation by O2. Comparisons are also made between the present results and those recently obtained for methanol-induced oxide removal in order to elucidate which chemisorbed alcohol fragment(s) constitute the reactive oxygen “scavenger”.

The adsorption of carbon monoxide and nitric oxide on Ir(100) in acidic aqueous solutions has been probed by voltammetry together with in-situ infrared reflection−absorption spectroscopy, with the objective of assessing the substrate as well as adlayer structure and bonding. The ordered Ir(100) surface was prepared by flame annealing followed by cooling in a H2/Ar stream. Similar to, and consistent with, the reported behavior in ultrahigh vacuum (UHV), at potentials E ≥ 0 V vs SCE (i.e., in the presence of coadsorbed water) the C−O stretch (νCO) for adsorbed CO occurs at frequencies, 1970−2040 cm^-1, that are indicative of atop (or near-atop) surface coordination. At lower potentials where coadsorbed hydrogen is present, however, the νCO spectra at intermediate CO coverages (0.2 < θCO < 0.5) include a lower-frequency band suggestive of bridge-bonded CO. The voltammetric features for hydrogen adsorption−desorption, dominated by a sharp current−potential peak at −0.04 V, are consistent with the formation of localized H islands in the presence as well as absence of coadsorbed CO. The νCO spectral form at intermediate CO coverages, however, indicates that some CO/H intermixing is also present. The occurrence of partial NO dissociative chemisorption is evident from the voltammetry for irreversibly adsorbed layers. This is also consistent with the N−O stretching (νNO) infrared spectra which exhibit relatively weak bands at frequencies ca. 1620−1650 and 1790−1810 cm^-1, suggestive of the presence of bridging as well as atop NO. The infrared spectral as well as the voltammetric findings are diagnostic of the presence of an unreconstructed Ir(100) surface by comparison with the reported behavior for the (1 × 1) and hexagonal reconstructed forms of Ir(100) in UHV.

The adsorption of carbon monoxide and nitric oxide on an ordered Ir(110) electrode surface in aqueous 0.1 M HClO4 has been probed by voltammetry together with in-situ infrared reflection−absorption spectroscopy (IRAS). Exclusively atop coordination of both CO and NO is suggested from the relatively high C−O and N−O stretching (νCO, νNO) frequencies observed, 1980−2060 and 1820−1840 cm^-1, respectively, that upshift with increasing coverage. Adsorption of NO as well as CO is essentially molecular, with near-unity saturation coverages, as deduced from voltammetry as well as infrared spectrophotometry. The potential-dependent νCO frequencies for the saturated CO adlayer are closely compatible with that for the corresponding Ir(110)/CO interface in ultrahigh vacuum (UHV) once the differences in surface potential are taken into account. In contrast to the case of the latter system, however, the electrochemical Ir(110)/CO interface exhibits a pair of νCO bands at intermediate CO coverages (θCO), suggestive of a difference in substrate-induced adlayer domains in the two environments. Closely similar θCO-dependent νCO spectra and voltammetric oxidation profiles were obtained for adlayers formed by either partial electroxidative stripping from a saturated adlayer or by direct dosing from a dilute CO solution. This unusual behavior indicates that extensive CO “islands” are not formed by partial adlayer electrooxidation, in contrast to the behavior of most ordered low-index Pt-group electrodes, suggesting that the substrate morphology features nanoscale domains rather than large terraces. The νCO and νNO frequencies for saturated adlayers on Ir(110) and (111) are similarly red-shifted from the gas-phase νCO and νNO values. However, the νCO−E and especially the νNO−E dependences (“Stark-tuning” slopes) are markedly larger than the predicted gas-phase values. The larger dνNO/dE values are ascribed to more extensive potential-dependent dπ−2π* back-donation for adsorbed atop NO compared with CO.

The effects of dosing various solvents along with potassium upon the infrared reflection−absorption spectra (IRAS) for saturated chemisorbed CO adlayers on Pt(111) at 90 K in ultrahigh vacuum (UHV) have been examined with the objective of assessing the roles of interfacial solvation upon the vibrational Stark effect (i.e., the adsorbate frequency−electrode potential dependence) for this archetypical Pt/CO electrochemical system. The solvents overlayers chosen−water, methanol, acetonitrile, acetone, and ammonia−used also in our earlier “UHV electrochemical modeling” studies, span a range of polarity and other solvating properties. Potassium dosage, with coverages θK ≤ 0.08, yields cations with the electron transferred to the metal surface and is therefore analogous to electrochemical double-layer charging. Kelvin-probe measurements of the work-function changes, ΔΦ, attending the addition of cations and solvent were undertaken so to evaluate the desired relationship between shifts in C−O vibrational frequencies, ΔνCO, and the overall surface potential. Complete solvation of the CO adlayer, usually requiring 1−2 solvent monolayers on top of the chemisorbate, induced substantial and solvent-sensitive downshifts of the atop νCO frequency and the work function (up to 50 cm^-1 and 1.5 eV, respectively). Addition of solvent to saturated CO adlayers with predosed K⁺ yielded profound changes in the νCO spectra. First, low solvent dosages, corresponding to solvent/cation stoichiometries ≥3, essentially eliminated the markedly red-shifted νCO band component due to short-range K⁺−CO interactions, apparently as a result of primary cation solvation. Additional solvent dosing red-shifted further the atop and bridging νCO bands that are characteristic of longer-range cation−CO interactions, finally yielding the simple linear ΔνCO−ΔΦ dependence familiar for in-situ electrochemical interfaces. In contrast to the addition of solvent to cation-free interfaces, the νCO frequency shifts induced by progressive cation addition in the presence of large solvent dosages are relatively insensitive to the nature of the solvent. Furthermore, the ΔνCO−ΔΦ dependences are essentially independent of the solvent for dosages, θs* ~ 2 monolayers, sufficient to complete the νCO spectral changes. This last finding indicates that much of the potential drop falls across the CO adlayer under these conditions, the solvent acting as a dielectric in screening the cation charge. Such simple solvent-independent ΔνCO−ΔΦ behavior, which contrasts the complex spectral patterns seen upon alkali-metal addition in the absence of solvent, is in good agreement with the ΔνCO−electrode potential dependences observed by in-situ IRAS at the corresponding electrochemical interfaces. The likely multifaceted roles of the solvating medium in influencing the nature as well as magnitude of the electrochemical Stark effect are discussed in light of these findings.

The utility of surface-enhanced Raman spectroscopy (SERS) as an in situ adsorbate probe of heterogeneous catalytic systems at high gas pressures (here up to 1 atm) in conjunction with mass spectrometry (MS) is demonstrated for reactions on transition-metal catalysts, at temperatures up to 500 °C. This simultaneous SERS−MS approach enables the relationships between the formation of specific adsorbed species (as sensed by SERS) and reaction products (as detected by MS) to be explored on a common (≈1 s) time scale. The low-frequency (<1000 cm^-1) spectral region, where metal−adsorbate vibrations are located, is especially informative. The environmentally important reactions of NO reduction by CO or H2 and the oxidation of methanol on transition metals (Rh, Pd, Pt) are considered here. The possible mechanistic significance of surface oxide formation observed during such ambient-pressure catalytic reactions is also discussed.

The outstanding performance of current military and civil aero gas turbine engines is linked closely to the way in which modern design and manufacturing techniques have become totally integrated with materials designed specifically for operation within the hostile environment of a gas turbine. Advanced titanium alloys are used extensively throughout the compressor and nickel-base superalloys dominate materials application in the turbine. In spite of current achievements, the engine designer is still under severe competitive pressure to improve engine performance still further and this will inevitably lead to even more demanding material requirements. The present paper outlines the continuing trends in engine development and describes the impact these are having on materials technology in general and the mechanical property requirements of nickel-base superalloys in particular.MST/512

The relationship between surface speciation and catalytic activity/selectivity during methanol oxidation on polycrystalline rhodium under ambient-pressure flow-reactor conditions was studied from 25 to 500 °C by means of surface-enhanced Raman spectroscopy (SERS) along with parallel mass spectrometric (MS) measurements. By utilizing SERS-active Rh films formed by electrodeposition onto gold, the former technique provides in situ surface vibrational spectra with unique sensitivity under these demanding conditions, enabling adsorbed species to be probed in real time (≈1 s) for comparison with the overall kinetics as evaluated by MS. Exposure of Rh to O2-free methanol yielded no detectable vibrational bands between 25 and 500 °C, although methanol decomposition to form CO and H2 was evident from MS. The presence of even subunity molar ratios of oxygen, however, yielded rich SER spectra, highlighted by bands indicative of CO(ads) (νRh-CO = 465 cm^-1, νRh-CO ≈ 2000 cm^-1). The catalytic selectivity toward CO2 (versus CO) gaseous product formation decreased markedly around the desorption temperature of CO(ads), ≈ 350 °C under these conditions. This is consistent with the facilitation of CO2 production by the presence of CO(ads). Complete selectivity toward exhaustive methanol oxidation (i.e., CO2, H2O formation) was observed in oxygen-rich methanol mixtures, adsorbed CO now being absent at all temperatures. The CO2 production occurs partly via methanolic C−O cleavage as deduced by ¹⁸O2 substitution. The presence of rhodium oxide (Rh2O3) was diagnosed for such reactant mixtures above ca. 300 °C from the characteristic 500−580 cm^-1 νRh-O bands. The kinetics of formation and removal of the oxide were probed by gas flow-switching coupled with transient SERS measurements. The oxide formation rates following O2 exposure are slowed markedly (>100-fold) by the presence of even a small (5%) methanol mole fraction. Switching to pure methanol results in very rapid oxide reduction, so that, for example, removal is complete within ca. 1 s at 350 °C with 100 Torr of CH3OH. Examination of the transient oxide removal kinetics as a function of temperature and methanol pressure revealed a transition from strongly activated to essentially T-independent behavior at lower pressures and/or higher temperatures. This is indicative of a change from rate-determining removal of oxygen from the oxide lattice to a subsequent step involving formation of and/or reaction with an adsorbed methanol scavenger. While such reactivity earmarks the oxide as a potential reaction intermediate, the overall catalytic turnover rates for methanol oxidation are nonetheless faster than can readily be accommodated on this basis.

Some virtues of modeling electrochemical systems by dosing interfacial components onto clean metal surfaces in ultrahigh vacuum (UHV) are discussed, with an emphasis on elucidating the nature of double-layer solvation and how solvent molecules influence the intermolecular interactions. This “non situ” strategy (as distinct from ex situ approaches involving electrode transfer to/and from UHV) allows each interfacial component (solutes, ions, solvent) to be added sequentially and in controlled amounts, enabling the various molecular (and hence intermolecular) ingredients that constitute the double layer to be accessed in incremental fashion. The approach also provides an invaluable means of understanding the differences in structure and bonding between analogous electrochemical interfaces and the constituent metal−UHV systems. Such issues are particularly germane with the recent advent of microscopic-level structural information for in situ electrochemical systems. Described specifically here is the UHV-based vibrational characterization of solvent and chemisorbate modes by employing infrared reflection−absorption spectroscopy (IRAS), together with work-function measurements, as a function of interfacial composition. The former provides a sensitive monitor of intermolecular interactions as well as being applicable (albeit with more restrictions) to in situ systems, whereas the latter yields insight into surface-potential profiles and also links the potential scales of metal−UHV and electrochemical interfaces. Several distinct examples aimed at elucidating double-layer solvation effects on Pt(111), recently scrutinized in our laboratory, are discussed. These include examining the progressive solvation of cations, adsorbed anions, and combinations thereof, by water and methanol. Comparisons with vibrational spectra for the solvation of gas-phase (i.e., isolated) ions enables the substantial influence of the metal surface upon double-layer solvation to be explored in detail. The converse role of double-layer charge upon the inner-layer solvent orientation (as exemplified for acetone and acetonitrile) is also found to be considerable and involves long-range forces. The combined influences of solvent and double-layer charge upon chemisorbate structure and bonding are also considered for the archetypical example of carbon monoxide. Marked electrostatic effects of solvation upon CO structure and bonding are seen even in the absence of net charge. The complex short-range influences of added cationic (K⁺) charge upon the CO adlayer are quenched upon partial K⁺ solvation. The longer-range electrostatic effects are progressively modified as the chemisorbate layer as well as the ionic charges become fully solvated, so to reveal a simple “Stark-tuning” frequency−potential behavior identical with that familiar in electrochemistry. Some more general implications and applications of such “UHV double-layer modeling” tactics are also briefly considered.

Surface-enhanced Raman spectroscopy (SERS) in conjunction with mass spectroscopy (MS) has been utilized to investigate the adsorption and hydrogenation of carbon monoxide on polycrystalline rhodium surfaces. The SERS-active Rh substrates were prepared by electrodeposition of ultrathin films on electrochemically roughened gold and display remarkably robust SERS activity over a wide range of temperatures (up to 400 °C) and pressures (here up to 1 atm). The SER spectra reveal that CO adsorbed primarily on atop sites (νRh-C = 470 cm^-1) and desorbed by about 250−300 °C under all gas-phase conditions examined. Partial dissociation of the CO adlayer, however, was obtained at temperatures as low as 100 °C, most likely facilitated by the large number of steps and kinks present on these roughened surfaces. This was evidenced by a partial removal of CO at temperatures (ca 100 °C) well below those expected for thermal desorption (200−250 °C) and supported by the observed formation of surface carbonate (665 cm^-1) under these conditions. The CO dissociation, however, is hampered at lower temperatures (<200 °C) when gas-phase H2 and/or CO are present, most likely due to blocking of site ensembles necessary for decomposition to proceed. Interestingly, heating a CO adlayer in pure H2 resulted in the formation of an adsorbed oxygen species (νRh-O = 295 cm^-1) at temperatures above 250 °C. The CO hydrogenation reaction was examined over a wide range of gas-phase conditions (H2:CO = 99:1 to 4:1 at 1 atm), with methane being the only hydrocarbon product detectable with MS. In addition to the presence of adsorbed CO observed up to 250 °C under all H2/CO reaction ratios, the adsorbed oxygen species noted above was detected at higher temperatures (>250 °C) when a low percentage of CO (≤1%) in the feed reactant stream was used. The influence of the adsorbed species on the overall methanation rates is discussed in light of these findings. Utilizing the seconds time-scale resolution of SERS, the exchange between gas-phase and adsorbed CO was also studied. The results of such transient ¹³CO/¹²CO exchange experiments reveal that this desorption pathway is weakly activated (≈1 kcal mol^-1), first order with respect to CO coverage, yet independent of CO partial pressure in the regime studied (8−760 Torr). This contrasts the first-order pressure dependence for much lower CO partial pressures (≤10^-5 Torr) reported earlier in the literature. A rate law and mechanism are proposed which account for these differences and rationalize the observed behavior.

The adsorption of nitric oxide and carbon monoxide on polycrystalline iridium in aqueous 0.1 M HClO4 was probed by in-situ surface-enhanced Raman spectroscopy (SERS) and infrared reflection−absorption spectroscopy (IRAS) with the primary objective of probing the molecular and dissociative chemisorption of the former adsorbate in this electrochemical environment. The latter adsorbate was examined partly as a means of characterizing the microscopic nature of the iridium electrode, given the well-documented sensitivity of the CO intramolecular vibration to the surface structure. The SERS vibrational technique was harnessed by utilizing ultrathin iridium films electrodeposited on a gold substrate. The availability of this method enables vibrational features associated with surface-adsorbate as well as intramolecular bonds to be scrutinized, facilitating detection of adsorbed NO fragments. Saturation adsorption of CO yields a single C−O stretching (νCO) band at ca. 2040−2070 cm^-1 in both the Raman and infrared spectra, suggestive of exclusive atop (or near-atop) binding as on monocrystalline Ir surfaces. Significant (15−20 cm^-1) discrepancies in the potential-dependent νCO frequencies measured by SERS and IRAS are evident, however, indicating that the ensemble distribution of iridium surface microenvironments sensed by these techniques is dissimilar. However, the νCO frequencies are approximately consistent with those evaluated for corresponding iridium−vacuum interfaces once the differences in surface potentials are taken into account. In contrast to the νCO band, the frequency of the metal−CO (νM-CO) vibration decreases with increasing electrode potential, in harmony with SERS findings on other transition-metal surfaces and theoretical bonding expectations. Adsorption of NO yields a weaker N−O stretching (νNO) band at 1780−1810 cm^-1 in both the SERS and IRAS spectra, indicative of the presence of (probably atop) molecular NO chemisorption. The SERS lower-frequency region is dominated by a well-defined band at 570 cm^-1, attributed to a metal−oxygen stretch from the chemisorbed oxygen fragment formed by NO dissociation. This assignment arises in part from the markedly different potential-dependent stability as well as intensity of the 570 cm^-1 band in comparison with the νNO feature. Supporting evidence includes the observation of a near-identical vibrational band upon oxygen dissociation (and NO adsorption) on Ir surfaces in vacuum and for the present Ir films upon gas-phase O2 dosing. While the extent of NO dissociation on Ir cannot be estimated quantitatively, consideration of band intensities suggests that the coverages of molecular NO and the atomic-oxygen fragment are not greatly different. Comparisons are briefly made with NO adsorption on other transition-metal surfaces in electrochemical and vacuum environments.

Measurements of the work-function changes, ΔΦ, on Pt(111) for continuously increasing solvent exposures θs* and in the presence of various coverages of potassium, θK, in ultrahigh vacuum (UHV) at 90 K are reported with the objective of ascertaining how the surface charge−potential properties of such “UHV electrochemical model” interfaces are altered by progressive solvation. The solvents&sbd;water, methanol, acetonitrile, acetone, and ammonia&sbd;span a range of dipolar and other solvating properties and have been utilized in related vibrational spectroscopic studies from this laboratory. Since potassium dosage yields interfacial electron transfer to form K⁺ together with surface electronic charge, the corresponding ΔΦ−θK plots for various solvent dosages extracted from the above data provide surface charge−potential (σ−φ) curves for systematically varying extents of interfacial solvation. In contrast to the large (1−3 eV) monotonic solvent-induced Φ decreases observed in the absence of ionic charge, the presence of predosed K⁺ yields initial Φ increases, associated with cation solvation, followed by Φ decreases due primarily to the ensuing metal surface solvation. Examination of the corresponding ΔΦ−θK traces obtained for these different solvent dosage regions shows that the basic charge−potential features characteristic of the solvated double layer require only ionic solvation, even though complete metal surface solvation modifies significantly the electrostatic behavior. While surface solvation by the different species examined in the absence of charge yield substantially dissimilar Φ values (i.e., differing “potentials of zero charge”), the charge−potential characteristics are relatively insensitive to the solvent. This finding, comparable to that obtained for in-situ electrochemical interfaces, indicates that the effective “interfacial solvent dielectric constant” varies by only 2-fold or less. ΔΦ−θK data obtained by K dosing after solvent addition yielded larger −ΔΦ values (i.e., smaller capacitances), consistent with more complete K⁺ solvation and/or larger K⁺− surface separations. Corresponding ΔΦ−θK data for CO-saturated Pt(111) indicates that the CO adlayer plays a role in dielectric screening. Effectively θK-independent ΔΦ responses were obtained with ammonia-solvated Pt(111), however, suggestive of the formation of solvated electrons. Specific comparisons are made between the UHV-based charge−potential behavior with that for in-situ electrochemical interfaces and for ionizable high-nuclearity Pt carbonyl clusters in nonaqueous media. The latter systems, in particular, exhibit closely similar surface charge−potential characteristics to the corresponding UHV-based Pt(111) interfaces.

In-situ infrared reflection−absorption spectra are reported for mixed nitric-oxide−carbon-monoxide adlayers along with the constituent chemisorbates separately as a function of coverage on ordered Ir(111) at 0.4−0.45 V vs standard hydrogen electrode in aqueous 0.1 M HClO4, with the objective of assessing the composition-dependent nature of the coadsorbate vibrational interactions. This substrate−coadsorbate combination provides an informative model system since both chemisorbates appear to bind exclusively in atop (or near-atop) surface sites on the basis of their simple N−O (νNO) and C−O (νCO) vibrational fingerprints, and composition-dependent mixed adlayers can readily be formed via partial replacement of saturated irreversible adsorbed NO layers by exposure to dilute CO solutions. Increasing the CO coverage, θCO, both in the absence and presence of coadsorbed NO, yields marked progressive blueshifts in the νCO band frequency (ca. 2020−2075 cm^-1), attributable chiefly to enhanced dipole−dipole coupling. While adsorption of NO alone exhibited virtually coverage (θNO)-independent νNO frequencies, ca. 1835 cm^-1, indicative of chemisorbate island formation, dilution within mixed CO/NO adlayers yields progressive νNO redshifts (down to ca. 1790 cm^-1). The composition-dependent νCO and νNO frequencies within the CO/NO adlayers are consistent with molecular intermixing, as supported by comparison with numerical simulations extracted from conventional dipole-coupling theory, although the observed nonlinear νCO−θCO dependence suggests the formation of locally enriched CO regions at intermediate compositions. Evidence supporting coadsorbate intermixing is obtained by comparing the composition-dependent CO and NO band absorbances with the dipole-coupling predictions. In particular, the presence of coadsorbed CO yields marked (up to 3-fold) decreases in the NO band absorbance, especially toward lower νNO values, which arise from band-intensity transfer to neighboring higher-frequency (CO) oscillators. Despite the large (ca. 250 cm^-1) difference in νNO and νCO singleton frequencies, this striking effect is in approximate agreement with dipole-coupling theory, again presuming molecular CO/NO intermixing. The observed marked increases in the νCO bandwidth toward lower θCO values, along with pronounced asymmetric band shapes, are in good agreement with theoretical predictions that include stochastic fluctuations of the local adsorbate population density. Moreover, the larger intermediate-θCO νCO bandwidths observed in the presence of coadsorbed NO are also quantitatively accounted for on this basis in terms of coadsorbate intensity transfer. The more broad-based utility of such dipole-coupling analyses for elucidating local interactions within mixed adlayers is considered in light of these findings.