Xenon and Krypton Adsorption on the SILICON(111) 7X7 Surface.

The thermodynamically stable (111) surface of Si has been known for many years to reconstruct with a unit mesh seven times larger in each direction than the unit mesh of an unreconstructed (111) plane. Despite extensive experimental and theoretical work done on the Si(111)7x7 surface, its atomic arrangement remains unknown. A study of adsorption of Xe and Kr on the Si(111)7x7 surface was made with the motivation of using the weakly bound physisorbed atoms as a probe of some of the structural features of the surface. It was hoped that, unlike in chemisorption systems, the weak adatom-substrate interaction would not appreciably alter the 7x7 reconstruction. The experimental techniques included low energy electron diffraction, Xe Auger spectroscopy, and measurements of isobars, i.e., the amount of gas adsorbed as a function of temperature at fixed pressure. The low energy electron diffraction indicates that, as anticipated, the substrate structure is not appreciably modified by the adsorption. The Xe Auger spectra shift and broaden with coverage indicating that Xe atoms are being adsorbed in sites farther from the substrate as the adsorption proceeds. The isobars show the stepwise adsorption at each of the first several adatoms per surface unit mesh. From the temperature positions of the steps in coverage, the binding energies were measured. These binding energies show appreciable differences. Several general conclusions about the 7x7 structure can be drawn without detailed reference to specific models. First, the 7x7 unit mesh presents a singly unique site of highest binding energy followed by a resolved sequence of sites of successively smaller binding energy. Secondly, the large differences between the binding energies of the first adsorbed atoms and those at higher coverage indicate the existence of some sites in which the adatom has significantly larger coordination. This result alone seems to preclude "smooth" models. Pairwise sums of 6-12 potentials have been used to estimate binding energies at sites present in proposed structural models. None of the currently proposed models give detailed agreement with the data. However, modification of the recent triangular double-layer models could possibly offer the necessary variety and sequences of sites and binding energies.