Mathieu Pascal

The structure of oxygen on Cu(100) at low and high coverages

The local adsorption structure of oxygen on Cu(1 0 0) has been studied using O 1s scanned-energy mode photoelectron diffraction. A detailed quantitative determination of the structure of the 0.5 ML (√2×2√2)R45°-O ordered phase confirms the missing-row character of this reconstruction and agrees well with earlier structural determinations of this phase by other methods, the adsorbed O atoms lying only approximately 0.1 A above the outermost Cu layer. At much lower coverages, the results indicate that the O atoms adopt unreconstructed hollow sites at a significantly larger O–Cu layer spacing, but with some form of local disorder. The best fit to these data is achieved with a two-site model involving O atoms at Cu–O layer spacings of 0.41 and 0.70 A in hollow sites; these two sites (also implied by an earlier electron-energy-loss study) are proposed to be associated with edge and centre positions in very small c(2×2) domains as seen in a recent scanning tunnelling microscopy investigation.

The local adsorption geometry of CO and NH3 on NiO (100) determined by scanned-energy mode photoelectron diffraction

The local adsorption structures of CO and NH3 on NiO(1 0 0) have been determined by C 1s and N 1s scanned-energy mode photoelectron diffraction. CO adsorbs atop Ni surface atoms through the C atom in an essentially perpendicular geometry (tilt angle 12±12°) with a C–Ni nearest-neighbour distance of 2.07±0.02 A. NH3 also adsorbs atop Ni surface atoms with a N–Ni distance of 2.06±0.02 A. These bondlengths are only very slightly longer than the comparable values for adsorption on metallic Ni surfaces. By contrast theoretical values obtained from total energy calculations, which exist for CO adsorption on NiO(1 0 0) (2.46 A and 2.86 A) are very much longer than the experimental value. Similar discrepancies exist for the N–Ni nearest-neighbour bondlength for NO adsorbed on NiO(1 0 0). Combined with the published measurements of the desorption energies, which also exceed the calculated bonding energies, these results indicate a significant failure of current theoretical treatments to provide an effective description of molecular adsorbate bonding on NiO(1 0 0).

Photoelectron diffraction study of the Ag(110)-(2×1)-O reconstruction

The structure of the (2×1)-O adsorption phase on Ag(110) has been determined using scanned-energy mode photoelectron diffraction. The oxygen atoms have been found to occupy the long-bridge site and are almost coplanar with the top layer of silver atoms. The best agreement between multiple-scattering theory and experiment has been obtained for a missing-row (or equivalently an ‘added-row’) reconstruction in which the first-to-second and second-to-third interlayer spacings are 1.55±0.06 A and 1.33±0.06 A, respectively. Alternative buckled-row and unreconstructed surface models can be excluded.

Quantitative determination of the local adsorption structure of carbonate on Ag(110)

The local geometry of carbonate (CO3) on Ag(1 1 0), formed by the reaction of CO2 with preadsorbed oxygen, has been investigated using C Is scanned-energy mode photoelectron diffraction. The carbonate species is essentially planar and adsorbs almost parallel to the surface in an off-atop site relative to an outermost layer Ag atom, at a C-Ag layer spacing of 2.64 +/- 0.09 Angstrom, with a well-defined azimuthal orientation. This geometry is best understood in terms of the added-row model proposed by Guo and Madix in which additional Ag atoms lie adjacent to the carbonate, such that bonding can occur through at least one of the oxygen atoms. The distance between this oxygen and its nearest neighbour Ag adatom is 1.90 +/- 0.42 Angstrom. The C-O distances are in the range 1.26-1.30 Angstrom. While the symmetry of the carbonate in the optimum structure is reduced, the D-3h symmetry of the isolated species lies within the limits of Precision

Structure determination of methanethiolate on unreconstructed Cu(111) by scanned-energy mode photoelectron diffraction

The local structure of methanethiolate, CH3S–, on an unreconstructed Cu(1 1 1) surface at low temperature, has been investigated by S 2p and C 1s scanned-energy mode photoelectron diffraction, with chemical state sensitivity. 71(+14/−16)% of the methanethiolate was found to occupy bridge sites, 29±14% to occupy fcc hollow sites and 0+19% to occupy hcp hollow sites. In the bridge site the layer spacing of the sulphur atom to the outermost substrate layer is 1.87±0.03 A giving a Cu–S bondlength of 2.27±0.03 A. The methanethiolate adsorbed in the fcc hollow site has a Cu–S layer spacing of 1.73±0.04 A, corresponding to the same bondlength of 2.27±0.04 A. The S–C bondlength was found to be 1.92±0.10 A. These conclusions are consistent with the results of previous X-ray standing wave and scanning tunnelling microscopy studies for a common model involving co-occupation of bridge and hollow sites, although differing relative occupations and long-range ordering are thought to arise from different preparation conditions. The new data favour a model in which the S–C bond axis of the bridge-bound thiolate is tilted by 45±12° away from the surface normal in the azimuth directed towards the fcc hollow site.

The structure of the Ni(100)c(2×2)–N2 surface: a chemical-state-specific scanned-energy mode photoelectron diffraction determination

Using the chemical shift in the N 1s photoemission peak from the two inequivalent N atoms of N2 adsorbed on Ni(1 0 0) we have performed a scanned-energy mode photoelectron diffraction (PhD) structure determination of the Ni(1 0 0)c(2 × 2)–N2 weak chemisorption system. The N2 is found to adsorb atop surface Ni atoms with the N–N axis perpendicular to the surface at a Ni–N nearest-neighbour distance of 1.81 ± 0.02 A. This is very significantly shorter than the value (2.25 A) found in an earlier published study. An independent density-functional theory slab calculation yields a value of 1.79 A, in excellent agreement with the results of the current experiment. Analysis of the PhD modulations of the N 1s photoemission satellite peak show that these are consistent with this comprising separable components localised at the two N atoms as has previously been assumed in an earlier investigation based on (angle-scan) X-ray photoelectron diffraction. Both experiment and theory indicate a small extension of the N–N distance due to the adsorption (0.03 ± 0.03 A and 0.02 A respecti