Physics of lattice relaxation at aluminium surfaces

The authors show, on the basis of a semi-self-consistent density functional calculation of the equilibrium position of the first lattice plane at the three low index faces of aluminium, that the screening of the ionic relaxation dipole by the electrons is essentially perfect. As a result the surface energy and work function are nearly independent of small displacements of the first lattice plane. They also demonstrate that the 'frozen profile' approximation for the electron density grossly exaggerates the curvature of the potential in which the first lattice plane sits and so cannot be relied on for geometrical predictions. The results for the first interplanar spacing of the close-packed (111) surface agree with the 2% expansion (relative to bulk) found in the recent LEED analysis of Jona et al. (1980) and do not favour the 8% contraction suggested by the surface EXAFS study of Bianconi and Bachrach (1979).