Structure and Adsorption Characteristics of Clean Surfaces of Germanium and Silicon

Publisher Website
Google Scholar
Abstract
Diffraction patterns obtained from atomically clean germanium surfaces contained half‐integral order beams in (110) azimuths for both (100) and (110) surfaces and in all azimuths for the (111) surface. These results are considered to be due to displacements of surface atoms from their normal bulk lattice positions in the surface plane. Adsorption of oxygen on all of these surfaces extinguished all of the diffraction beams which were not integral order. In addition to the normal surface lattice spacings of clean (111) and (100) surfaces of silicon, there were surface structures with larger spacings, most of which depended on the conditions of ion bombardment and/or subsequent heat treatment. Two such structures have been observed for the (100) surface of silicon; one is a double‐spaced lattice in the (110) azimuth, similar to that for germanium, and the other has a spacing about 8% greater than that of normal silicon and was obtained only after radiation quenching of the crystal from 1000°C. Two large‐spaced structures were observed for the (111) surface. All of these structures were extinguished by exposure to oxygen. Evidence is presented which indicates that these structures were not due to contamination but to the silicon itself. The method of determining the kinetics of gas adsorption from the low‐energy electron diffraction beams is outlined and the calculations of the fractional coverage and sticking probability are presented for oxygen. For silicon crystals, the calculations of surface coverages and sticking coefficients for oxygen were found to depend on the preceding treatments of the crystals. The rate of adsorption of oxygen was proportional to the pressure, at least for pressures below 10‐6 mm Hg, and depended on the preceding treatments of the crystals. After oxygen adsorption, the clean germanium surface could be regenerated by heating at 500°C for 30 min and for silicon it was regenerated by heating at 900°C for a few minutes. A comparison with results of other observers is given.