ELECTRON-SPECTROSCOPIC STUDIES OF THE EARLY STAGES OF THE OXIDATION OF SI

Photoelectron spectroscopy (PES) and Auger electron spectroscopy have been used to study the early stages of the oxidation of the cleaved silicon (111) surface. The Si-2p core level as well as valence emission were studied with PES, using monochromatized synchrotron radiation at a photon energy which allows maximum surface sensitivity. In the initial adsorption stage (i.e., when the surface states are removed from the band gap), oxygen can be adsorbed either molecularly or atomically. The molecularly chemisorbed oxygen was characterized by a zero shift of the Si-2p core level, while the atomic state was associated with a 2.0-eV shift. Both types of chemisorption led to a removal of the filled surface states from the band gap. A third chemisorbed state, characterized by a 2.6-eV shift of the Si-2p core level, was observed in the adsorption stage beyond monolayer coverage. A broad 3.3-eV chemically shifted peak was attributed to Si atoms bonded to three oxygen atoms with smaller contributions from other states (i.e., SiO2 and Si atoms bonded to one or two oxygen atoms). Finally, the formation of SiO2 can be unambiguously identified by a 3.8-eV shifted peak. After an SiO2 layer 12 thick had formed, asymmetric emission on the low-binding-energy side of the SiO2 shifted peak revealed that approximately two layers of silicon atoms in the SiO2-Si interface were bonded to less than four oxygen atoms. © 1979 The American Physical Society.