| Summary: | The aim of this work is to assess the potential usefulness of photoelectron diffraction in the determination of surface structure. A review is first made of the theory of photoelectron diffraction from crystalline surfaces. Then, calculations of the variation of intensity with azimuthal angle are performed for various model systems, in order to investigate the sensitivity of the diffraction patterns to structural and non-structural parameters. It is found that for detector angles lying approximately half-way between the surface and surface normal, and for some photon energies, the diffraction patterns are sensitive to structure. For some energies, the patterns sure found to be sensitive to non-structural parameters, making it necessary to know these parameters to a high precision. Then, am analysis is made dr data taken in the azimuthal mode for 4d emission from c(C2 x 2) and p(2 x 2)Te adsorbed on Ni(OOl) and for 4d emission from (√3 x √3) R30°I adsorbed on AgClll). By transfering parameters from the relevant LEED calculation quite good agreement between theory and experiment is obtained for I on Agllll) but the agreement is poor for both coverages of Te on Ni (001). Subsequently the case of normal emission from the 4d levels of c(2 x 2)Te on Ni(001) is looked at, the aim being to resolve the discrepancy that has been found to exist between theory and experiment. This discrepancy is resolved by using the energy dependent Sara potential rather than Slater exchange in the construction of the Te potential. The use of this new potential for Te also leads to improved agreement with the experimental data taken with the azimuthal mode. Then, am investigation is made as to whether the Lee and Beni potential, which takes into account screened exchange and correlation effects is an improvement over the Hara potential in the description of electron scattering. To this end calculations are performed of the total and differential cross-sections for atomic Xe, Kr and Ar using the Slater, Hara and Lee and Beni potentials, and a comparison is made with experiment gas phase data. Hara exchange is found to be superior to the Lee and Beni potential especially at low energies. It is suggested that the inadequacy of the Lee and Beni potential is due to a breakdown of the Local Density Approximation and the success of the Hara potential is due to a cancellation of errors. Angle-resolved photoemission from adsorbed molecules is then looked at in order to assess the importance of multiple scattering. It is found that near to a resonance it is essential to incorporate multiple scattering. It is found that the Mattheis prescription of superposing atomic charge densities is inadequate in describing the emission and scattering properties of molecules, and it is preferable to use the Scattered-Wave Xa method which more realistically describes the distribution of charge within the molecule. It is concluded that if core-state photoelectron diffraction is going to be used as a surface structural technique then it is essential that energy dependent potentials be used in the description of the emission and scattering properties of the surface atoms. For valence emission from adsorbed molecules it is suggested that the Mattheis prescription should be dispensed with in constructing the molecular potential.
|
|---|
