| Summary: | Although Low Energy Electron Diffraction (LEED) has become relatively well established aa a technique for the study of' some aspects of' the surfaces of crystalline systems, it is only comparatively recently that close aattention has been paid to the physics of the diffraction process. The work presented in this thesis was therefore carried out in an attempt to elucidate some of these aspects of the process. Firstly, we show that it is not necessary tor the surface under examination to be an atomically perfect crystal surface, or even a macroscopically perfect surface over the entire area sampled by the electron beam, by a consideration of the diffraction patterns produced from faceted metal toils. This shows that provided the ssurface contains small area of good crystal order larger than the coherence area of the electron beam used to produce the diffraction pattern only as little as 10% of the sampled area of surface need be occupied by these areas of good crystalline order to produce no visible degradation of the diffraction pattern. A short discussion of the properties of' an ideal set of experimental data against which theoretical calculations or LEED intensity-energy spectra should be checked is followed by a set of 'experimental results which goes some way towards meeting these criteria. This comprises a complete set of intensity-energy spectra tor the specular LEED beam from a single crystal copper (100) surface in an azimuth close to (110) , covering a large planar area of k-space bounded by the approximate energies 20 and 350eV,and the angles of incidence 6 and 22º , the spectra being taken at 1º intervals, together with a similar set of data tor the (111) surface in a (112) azimuth, the spectra being taken in this case at 2° intervals. The (110) surface of copper proved to be impossible to clean satisfactorily in a UHV environment, and a sshort discussion of its adsorption properties in that environment is given. These sets of spectra demonstrate that although the main features or the intensity-energy spectrum lie close to the kinematical Bragg scattering conditions, the details of the spectra can only be explained by the use of proper dynamical calculations, in accord with the latest theories of LEED. Some correlations between the features of the spectra and the emergence of new non-specular diffraction beams are discussed. Finally, a detailed discussion is presented on the effects of specimen temperature on the intensity-energy spectra obtained from the Cu(100) surface and the Cu(111) surface, the results for the (100) surface being rather more detailed than those for the (111) surface. These results demonstrate that the principal effect of temperature is the diminution in intensity of the diffraction peaks in the intensity-energy spectra, and that this diminution in intensity can be characterised by an effective Debye temperature. As the specimen temperature is changed, these peaks also shift in energy in a non-simple way, which, while not in itself important when comparing experimental results with rigid lattice theoretical spectra, must be taken into account in evaluating the Debye temperature for each peak. However, the width and overall shape of diffraction peaks which are well isolated from other peaks in the spectra appear to be independent of temperature. The detailed study of the measured Debye temperatures and peak shifts tor all the diffraction peaks as a function or the experimental diffraction parameters - energy and crystal orientation - proves them to be highly sensitive to the diffraction conditions. Such results cannot therefore be interpreted in the traditional kinematical manner, and hence cannot be used with confidence in extracting information about surface atom vibrations, surface lattice expansions, etc. The importance of including thermal effects when comparing theory with experiment is emphasised, and a particularly interesting experimental result is discussed in the light of a very recent calculation.
|
|---|
