R. G. Tobin

CO adsorption site occupation on Pt(335): a quantitative investigation using TPD and EELS

Abstract In a study of CO on Pt(335) at various coverages using TPD and EELS we observe bridge-bonded CO. Bridge-bonded CO on Pt(335) is inconsistent with existing models of CO adsorption on highly stepped Pt surfaces. Previous studies of CO on Pt(335) and on the similar Pt(112) surface identified two atop CO species: edge atop and terrace atop. We observe two additional species: edge bridge and terrace bridge. Edge atop and edge bridge CO both adsorb before CO adsorbs at terrace sites. Activation energies are estimated for CO desorption from edge and terrace sites. A model that explains our data, based on a quantitative analysis of the TPD and EELS spectra, is presented. Possible overlayer structures are proposed at selected coverages. At saturation, all edge sites are occupied with atop CO, half of the terrace sites are occupied, and on the terrace the atop-to-bridge ratio is 2:1. At low coverage the overlayer reaches equilibrium within a few minutes at 100 K. At high coverage equilibrium is not reached in tens of minutes at 100 K, but annealing 10 s at 280 K does allow the overlayer to reach equilibrium.

Vibrational linewidths of adsorbed molecules: Experimental considerations and results

The current understanding of the causes of vibrational line-broadening in adsorbates is discussed, with particular emphasis on experimental considerations. Recent experimental determinations of line-broadening mechanisms are reviewed. While there has been great progress in the theoretical understanding of line-broadening by energy and phase relaxation, it remains difficult to determine experimentally the relative importance of the various proposed mechanisms. Inhomogeneous broadening, which is certainly important in many cases, has received little theoretical attention.

Adsorption and dissociation of oxygen on Pt(335)

Abstract Oxygen adsorption and dissociation on Pt(335) have been studied with high-resolution electron energy loss spectroscopy (HREELS) and temperature-programmed desorption. Five vibrational peaks are found: from atomic oxygen on the terraces at 470 cm −1 and on the steps at 560 cm −1 , from molecular oxygen on the terraces at 860 cm −1 and on the steps at 830 cm −1 , and, under limited conditions, from a distinct molecular step species at 700 cm −1 . Four desorption features correspond to the first four vibrational peaks: terrace O 2 at 125 K, step O 2 at 210 K, terrace O at 750 K and step O at 850 K. Our analysis of the HREELS intensities suggests that the step species are bound on the exposed step edges, not in the trough as suggested by previous studies. Dissociation occurs almost exclusively at the steps: At saturation coverage about 95% of the step O 2 dissociates before desorbing, while 95% of the terrace O 2 desorbs directly. Dissociation of the edge O 2 occurs in part directly and in part through the metastable 700 cm −1 state. Direct dissociation produces step O and terrace O with equal probabilities, while the 700 cm −1 state produces predominantly step O. Equilibration between step and terrace O occurs only above 400 K.

An infrared emission study of the molecule-substrate mode of CO on Pt(111)

Abstract We have used infrared emission spectroscopy to study the molecule-substrate vibrational mode of on-top CO on Pt(111), as a function of coverage and of temperature, in the range from 210 to 400 K. Little change in the line is observed as a function of coverage, except for an abrupt shift, of ~ 6 cm −1 , to lower frequency, at coverages > 0.5. The line shape is asymmetric, and the linewidth varies much more slowly with temperature than is predicted by theories of lifetime broadening by phonon emission. Based on these observations, our conclusion is that the line is inhomogeneously broadened. Our measurements place an upper limit on the homogeneous linewidth for this mode. The conclusion that the line is inhomogeneously broadened is in contrast to suggestions that the corresponding mode of CO on Ni(100) is lifetime broadened by a two-phonon decay process. We also report the observation of the atom-substrate mode of O on Pt(111).

Surface morphology and electric conductivity of epitaxial Cu(100) films grown on H-terminated Si(100)

We have investigated the crystal structure, surface morphology, and electrical conductance of Cu films grown on H‐terminated Si(100). The films were grown by thermal evaporation at 10−8 Torr, at deposition rates from 0.1 to 3.5 nm/s and at substrate temperatures from room temperature up to 200 °C. Typical film thicknesses were ∼100 nm. Epitaxial growth was verified by x‐ray diffraction for films grown near room temperature at rates up to 2 nm/s. The root‐mean‐square surface roughness was measured to be 1–2 nm using atomic force microscopy (AFM). Higher growth rates or deposition temperatures above 100 °C produced poor quality, nonepitaxial films. Postdeposition annealing at temperatures up to 175 °C did not improve the surface roughness, and at higher annealing temperatures rapid silicide formation destroyed the Cu film. In situ electrical resistance measurements and AFM images suggest that for about the first 6 nm of growth the film consists of disconnected islands. X‐ray‐diffraction data show that the i...

CO on Pt(335): Vibrational Stark effect, mode coupling, and local field effects on a stepped surface

Abstract We have studied CO on Pt(335) using infrared diode laser spectroscopy and electric field modulation. The spectra cover the 2000–2160 cm−1 range. The Pt(335) surface is stepped, with (111) terraces four atoms wide. At low coverage only one species is observed: CO adsorbed on step edges. In terms of the applied field, the Stark tuning rate of a low coverage of this species is (6.0 ± 0.8) × 10−7cm−1/(V/cm). At the same coverage, the screened dynamic dipole moment ~e = (0.94 ± 0.11)e. The rate calculated from ~e is (7.0 ± 1.1) × 10−7cm−1/(V/cm), agreement with the measured value. At higher coverages te also observed. Terrace CO is similar to CO on Pt(111) in vibrational frequency, dynamic dipole moment and desorption temperature. However, the Stark tuning rate of terrace CO is at least an order of magnitude smaller than expected. We discuss why the tuning rate might be so small. When both species are present, the vibrational spectra are strongly affected by dipole-dipole coupling. As a result, the electroreflectance spectrum is not the derivative of the polarization modulation spectrum. The spectra are analyzed using a phenomenological coupled-oscillator model. Fitting the spectra to the model, we find that for CO adsorbed on step edges, the Stark tuning rate in terms of local field is(7.5 ± 2.0) × 10−7cm−1/(V/cm), independent of coverage. The saturation coverage clean surface at 300 K is (8.3 ± 0.6) × 1014 cm−2. With saturation coverage of CO the linewidth is very small, only 4.5–5.0 cm−1 FWHM, even though the sample is at 300 K. Also, while cleaning the surface, we observed a correlation between Sn coverage and saturation coverage of CO. One possible explanation for the correlation, although not proven, is that Sn blocks CO adsorption sites. If so, the data suggests that each Sn atom blocks 11 ± 6 CO adsorption sites, which may explain how Sn helps prevent CO poisoning of Pt electrodes in fuel cells.

The C=O stretching vibration of CO on Ni(100) by infrared emission spectroscopy

Abstract Infrared emission spectroscopy has been used to study the C =0 stretching vibration of CO on Ni(100) at 310 K. In contrast to previous electron energy loss and infrared measurements, no bridge-bonded CO is observed at any coverage below θ =0.5. Variations in the lineshape with exposure suggest the presence of c(2×2) islands even at low coverage, though low energy electron diffraction shows no sign of such ordering.

Electric field screening in an adsorbed layer: CO on Pt(111)

Abstract The IR cross section and Stark tuning rate of CO on Pt(111) versus coverage are measured. (The Stark tuning rate is the variation of vibrational frequency with applied electric field.) We interpret the observations as coverage-dependent screening of applied static and IR fields at COs adsorption site. In contrast to the standard dipole—dipole coupling model, we find that static and IR screening depend differently on coverage. We also find that the local field Stark tuning rate for CO on Pt(111) in vacuum is half that at the electrochemical double layer. Possible explanations of our results are discussed.

REACTIONS OF N AND NO ON PT(335)

As part of a study of species important in automotive exhaust chemistry, the reactivity of atomic N and NO on Pt(335) at low temperature has been studied. The atomic N was produced by dissociating adsorbed NO with a 76 eV electron beam. Cross sections for electron-stimulated desorption and dissociation are estimated for NO on terrace and step sites. Terrace NO is at least five times more likely to desorb than to dissociate. Step NO has a lower desorption cross section than terrace NO, but probably a higher dissociation cross section. Temperature-programmed desorption was used to monitor desorption, dissociation, and the formation of N2 and N2O from adsorbed N and NO. Five distinct desorption states of N2 formed by NO dissociation are identified. The dominant N2 peak (435 K) comes from electron-dissociated step NO; its desorption temperature is higher than the N2 peaks from electron-dissociated terrace NO. Coadsorbed N and NO react to form N2O even below 100 K, with an activation barrier of ∼6 kcal/mol. On...

An infrared emission study of the C=O stretch vibration of bridge-bonded CO on Pt(111)

We have studied the C=O stretch vibration of bridge-bonded CO on Pt(111) in the temperature range from 225 to 300 K. Our spectra are qualitatively different from those previously reported by Hayden and Bradshaw [Surface Sci. 125 (1983) 787] in that we do not observe the line near 1810 cm−1 that they attributed to three-fold bonded CO. We observe only a single line at 1849 cm−1. As the coverage is increased beyond 0.5 monolayer, the linewidth decreases, reaching its minimum value near ϱ=0.54. Little change in the frequency or integrated intensity of the line is seen. These observations tend to support “fault line” models for the high coverage structure of CO on Pt(111). The minimum linewidth of the line is 11.5±0.6 cm−1 for temperatures below 275 K, rising rapidly to ≈ 34 cm−1 at 300 K. This temperature dependence cannot be explained by current models of homogeneous broadening. We attribute it to inhomogeneous broadening associated with an order-disorder transition in the overlayer.