Abstract

Silicon owes its dominance as the electronic material of semiconductor technology to the high quality of its oxide layer. Understanding the nature of the SiO_x/Si interface and the SiO_x overlayer is quite important, and the problem has been investigated extensively by using a wide range of techniques, including AFM, STM, TEM, surface x-ray, and surface second-harmonic generation (SHG). An interesting, unresolved question raised by these studies is whether the SiO_x layer retains some degree of ordering or whether it is amorphous with an abrupt interface between it and the crystalline silicon surface. Earlier work seems to suggest that the latter is the case, with a transition region of about 5 Å. Recent evidence, however, indicates that the presence of crystalline SiO_x phases,^1-3 possibly tridymite^4,5 or cristobalite,⁶ at or near the interface allows the long-range order of crystalline Si to decay gradually into an amorphous SiO_x overlayer. Because most interfacial probes are not capable of observing or interpreting disorder at the interface, a definitive answer to this question remains elusive. For example, the presence of small amounts of disorder can render high-resolution transmission-electron-microscopy images uninterpretable; hence TEM measurements cannot give an unambiguous answer.

© 1995 Optical Society of America