Brian E. Hayden

The adsorption of CO on Pt(111) studied by infrared reflection-adsorption spectroscopy

Abstract The adsorption of CO on Pt(111) between 85K and 300K has been studied by infrared-reflection-absorption spectroscopy together with TPD and LEED. The intensity of the absorption band due to the CO stretch of the linear species shows a maximum at the formation of the (✓3 × ✓3)R30° LEED pattern followed by a minimum at the c(4 × 2) structure during the adsorption of CO at low temperatures (≥150K). The absorption band due to the C-O stretch of the bridging species appears only after the formation of the pattern and reaches maximum intensity at the c(4 × 2) structure. Adsorption of CO to higher coverages (corresponding to the compression structures) broadens and shifts this absorption band. At higher temperatures (≥150K) a third peak is observed at ∼40cm −1 below the peak due to the bridging species and is attributed to adsorption in the three-fold sites. At 300K both peaks in this region are very broad. The intensity data differs from that measured with EELS (ref. 1) and favors a “faultline” structure of the type proposed by Avery (ref. 2). Together with the additional information from bandwidths it is possible to distinguish between the various structural models. The results obtained here may also be important in explaining data from other systems such as CO/Cu.

An infrared study of the adsorption of CO on a stepped platinum surface

Abstract The adsorption of CO at 85 K on the Pt-s[4(111)×(100)] surface has been investigated using infrared reflection-absorption spectroscopy (IRAS). TPD and LEED. Both IRAS and TPD measurements indicate that adsorption of CO takes place initially in a linearly bound configuration at sites associated with the steps (ν(CO) = 2065–2078 cm −1 ). Subsequently CO adsorbs with a reduced sticking coefficient on the (111) terraces, giving rise to a second band in the linear region (ν(CO) = 2086–2097 cm −1 ). The observed vibrational coupling is indicative of one-dimensional domains of step CO at low coverage and of two-dimensional domains, containing both step and terrace CO, at high coverages. The frequency of the step CO singleton is surprisingly close to that of the (hypothetical) terrace CO singleton. The IR extinction coefficient for CO adsorbed at the step site is a factor 2.7 greater than for CO in the two-dimensional domains at high coverage. Unlike Pt(111), adsorption on the stepped surface does not give rise to a bridging species.

Dipole coupling and chemical shifts in IRAS of CO adsorbed on Cu(110)

Results are reported of an experimental study using infra-red reflection absorption spectroscopy of mixed isotopic layers of 12C18O and 12C16O adsorbed on Cu(110). It is shown that dipole coupling leads to an upward shift of the frequency of the 12C16O layer by 50 cm−1 as the coverage is increased to saturation from an initial value of 2088 cm−1 but this is counteracted by a downward chemical shift of 44 cm−1. A general discussion of the method of separating these terms using the isotopic dilution limit is presented and some special problems involved in this separation for CO on copper are discussed. A model for the origin of the chemical shift involving through space interactions between the CO molecules due to overlap of 2 πb levels is proposed which takes into account other data on Cu, Pt and Pd surfaces.

An iras study of formic acid and surface formate adsorbed on Cu(110)

The adsorption and decomposition of formic acid on Cu(110) has been studied using infrared reflection-absorption spectroscopy and temperature-programmed desorption. A broad double peak at ∼ 1635 cm−1 was found to be characteristic of the carbonyl stretch ν(CO) of the acid monolayer after adsorption at 120 K. In agreement with earlier data, deprotonation occurred at ~ 270 K to give a surface formate species. The TPD measurements gave an estimate of formate coverage (θsat = 0.25 ± 0.05) and thus a determination of the reaction probability for deprotonation. The symmetric O-C-O vibration νs(COO) (1348–1358 cm−1) and the C-H stretch vibration ν(CH) (2891–2900 cm−1) were measured as a function of formate coverage. A vibrational band corresponding to νa(COO) was not observed under any experimental conditions; the formate species is thus thought to have C2v or Cs(1) symmetry. For the first time in IR spectroscopy of adsorbates a combination band νcomb at 2950 cm−1 was also observed. Its components (νa(COO) and δ(CH)) were not observed individually, since they have dynamic dipoles parallel to the surface. The combination band shifts at low coverages to lower frequencies, which is due to lateral interactions between parallel dynamic dipoles, and has an intensity comparable to that of the neighbouring C-H stretch, which may be due to a Fermi resonance.

Alkali metal-induced reconstruction of Ag(110)

Abstract The adsorption of Li, K and Cs on the Ag(110) surface has been studied by LEED, AES and work function measurements. It was found taht adsorption of small amounts of the three alkalis studied (0.05

The role of steps in the dynamics of hydrogen dissociation on Pt(533)

The dissociative adsorption of H2 and D2 on Pt(533) (Pt{4(111)×(100)}) has been investigated using temperature programmed desorption and supersonic molecular beams. Associative desorption of D2 from (100) step sites is observed at lowest exposures in TPD (assigned β3) at 375 K. Saturation of this peak at ΘH=0.14 corresponds to the filling of half of the available four-fold sites at the (100) step edge. At higher coverages, additional desorption takes place from the (111) terraces in a broad peak below 300 K similar to that observed (assigned β1 and β2) for the Pt(111) surface. The incident kinetic energy (Ei), surface temperature (Ts), coverage (ΘD), and incident angle (Φ) dependence of the dissociative sticking probability (S) was also measured. The initial dissociative sticking probability (S0) first decreases with increasing kinetic energy over the range 0<Ei(meV)<150 (low energy component), and subsequently increases (high energy component). Comparison with D2 dissociation on Pt(111), where (S0) incre...

The early stages of oxidation of magnesium single crystal surfaces

The early stages of oxidation of Mg(001) and Mg(100) single crystal surfaces at 300 K have been investigated by LEED, ELS, work function and ellipsometric measurements. A sharp decrease in work function on both surfaces during the first 12 L exposure indicates the incorporation of oxygen m the earliest stages of the interaction. The incorporated oxygen on Mg(001) gives rise to a broadening of the integral order LEED spots for an exposure < 6 L indicating a (1 × 1) oxygen underlayer. Initial interaction of oxygen with the Mg(100) surface leads to the disappearance of the substrate spots. After 6 L exposure and a coverage of between 0.5 and 1 monolayer, ellipsometric and ELS results indicate that on both surfaces incorporated oxygen together with oxygen from the gas phase form islands of MgO. Ellipsometric results further show that this transformation of incorporated oxygen into the three-dimensional oxide is complete at ~12 L exposure. LEED indicates that the oxide grows epitaxially on both surfaces, the square lattice of MgO(100) being observed in three equivalent domains on Mg(001) and in two on Mg(100). The (1 × 1) oxygen underlayer seems to co-exist with the epitaxial oxide islands on Mg(001) up to exposures of ~1 × 103 L.

Stepped single-crystal surfaces as models for small catalyst particles

Abstract Infrared spectroscopy is used to study the vibration of CO adsorbed on small, silica-supported platinum particles. Different samples have average Pt-particle sizes varying from 11 to 105 A in diameter. The infrared spectrum shows that all CO molecules are linearly bound to single Pt atoms; however, three bands can be detected on each of the different samples. The bands appear at 2081, 2070 and 2063 cm−1 and shift to higher wavenumbers (by about 6 cm −1) as the coverage increases. The intensities of the three bands vary with particle size in the same manner as the relative number of face, corner and edge atoms in our samples vary with particle size. This correlation leads to the assignment of the three bands to CO, linearly bound to face, corner and edge atoms in order of decreasing wavenumber. Some confirmation of these assignments are obtained from infrared reflection-absorption measurements of CO adsorbed on three extended-crystal surfaces of Pt: a (111) crystal, a crystal with (111) terraces and (100) steps, and a crystal with (111) terraces and kinked steps. These surfaces are described as (111), (533) and (432). The stepped crystals give bands for CO molecules associated with terrace sites appearing at about 2085 cm−1 and with step and kink sites appearing at about 2065 cm −1. Within ± 2 cm −1 there is no difference between the frequencies of CO adsorbed on step and on kink sites.

Particle size and support effects in electrocatalysis

Researchers increasingly recognize that, as with standard supported heterogeneous catalysts, the activity and selectivity of supported metal electrocatalysts are influenced by particle size, particle structure, and catalyst support. Studies using model supported heterogeneous catalysts have provided information about these effects. Similarly, model electrochemical studies on supported metal electrocatalysts can provide insight into the factors determining catalytic activity. High-throughput methods for catalyst synthesis and screening can determine systematic trends in activity as a function of support and particle size with excellent statistical certainty. In this Account, we describe several such studies investigating methods for dispersing precious metals on both carbon and oxide supports, with particular emphasis on the prospects for the development of low-temperature fuel-cell electrocatalysts. One key finding is a decrease in catalytic activity with decreasing particle size independent of the support for both oxygen reduction and CO oxidation on supported gold and platinum. For these reactions, there appears to be an intrinsic particle size effect that results in a loss of activity at particle sizes below 2-3 nm. A titania support, however, also increases activity of gold particles in the electrooxidation of CO and in the reduction of oxygen, with an optimum at 3 nm particle size. This optimum may represent the superposition of competing effects: a titania-induced enhanced activity versus deactivation at small particle sizes. The titania support shows catalytic activity at potentials where carbon-supported and bulk-gold surfaces are normally oxidized and CO electrooxidation is poisoned. On the other hand, platinum on amorphous titania shows a different effect: the oxidation reduction reaction is strongly poisoned in the same particle size range. We correlated the influence of the titania support with titania-induced changes in the surface redox behavior of the platinum particles. For both supported gold and platinum particles in electrocatalysis, we observe parallels to the effects of particle size and support in the equivalent heterogeneous catalysts. Studies of model supported-metal electrocatalysts, performs efficiently using high throughput synthetic and screening methodologies, will lead to a better understanding of the mechanisms responsible for support and particle size effects in electrocatalysis, and will drive the development of more effective and robust catalysts in the future.

The electrooxidation of carbon monoxide on ruthenium modified Pt(110)

Abstract The effect of adsorbed and incorporated (alloyed) sub-monolayers of Ru on Pt(110) on the electrooxidation of CO has been investigated using UHV-electrochemical transfer techniques. Overlayers of metal vapour deposited (MVD) Ru were characterised using LEED, XPS and LEISS. Both adsorbed and incorporated Ru overlayers were prepared, the alloy layers formed by annealing in the temperature range 400–1100 K. The reversible hydrogen adsorption characteristics of Pt(110) were suppressed in the cyclic voltammetry when Ru was adsorbed, incorporated in the top layer, or when absent from the top layer where there was a significant concentration of Ru in the second/third layers. Surfaces exhibited significant redox behaviour in the double layer region (0.3–0.6 V) of Pt(110) also when ruthenium was present in the adsorbed or top surface phases, but to a much lesser extent when incorporated in the second/third layer. CO electrooxidation, however, was promoted primarily by Ru incorporated in the top layer. Cycling to potentials of 1.2 V resulted in facile oxidative dissolution of adsorbed Ru, but only slow dissolution of Ru when incorporated in the top layer. The effect of cycling to 1.2 V on surfaces with negligible Ru in the top layer, but Ru in the second and third layers, was to modify the CO oxidation behaviour substantially, producing a significant coverage of a Pt(110)–Ru phase which promoted CO oxidation in a peak at 0.5 V.