Surface Structure Determination by Low-Energy Electron Diffraction

Four aspects of low-energy electron diffraction (LEED) have been investigated: interpretation of spot patterns to determine the surface unit cell and possible ambiguities, development of a photographic method for measuring angles of incidence and determining alignment of the LEED instrument, study of reliability factors used for intensity analysis, and justification for equivalent beam averaging. The interpretation of LEED spot patterns to determine the geometry of the surface unit cell can be involved when there are several symmetrically equivalent structural domains contributing to the pattern. Complex patterns can be deciphered by the algorithm described in Chapter II. The algorithm determines a surface unit cell that is often unique but not always, as where a p(2 x 2) pattern from a fcc(111) surface can be produced by a true (2 x 2) overlayer or by three domains of a (2 x 1) structure. This ambiguity arises on surfaces with 6m symmetry, such as fcc(111) and hcp(0001), for spot patterns with threefold rotational symmetry. In Chapter III, a broadly applicable photographic method for measuring angles of incidence and determining the alignment of the LEED instruments is described. Two published methods for determining the angle of incidence are special cases of this general procedure. The procedure extends the photographic methods and facilitates the verification of the alignment of the components of the LEED instrument. Reliability factors are used to evaluate correspondence between computed and observed LEED intensity spectra. Zanazzi and Jona, Pendry, and Sobrero and Weinberg have proposed reliability factors that are examined in Chapter IV. Chapter V provides a theoretical analysis and shows that averaging over momentum space gives the best resolution of the surface structure while energy averaging smears out information in the intensity spectra. Chapter VI provides a theoretical basis for the procedure of equivalent beam averaging, which provides a first-order correction to LEED intensities for systematic error due to angular misalignment of the incident beam and corrects for misorientation (where the actual surface plane is at a slight angle to the desired crystal plane). The potential of higher-order corrections is discussed.