On the Mechanism of Ultra Thin Silicon Oxide Film Growth During Thermal Oxidation
- PDF / 408,539 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 51 Downloads / 229 Views
E.P. GUSEV 1,3, H.C. LU 2, T. GUSTAFSSON 2 and E. GARFUNKEL 1 'Department of Chemistry, and Laboratory for Surface Modification, Rutgers University, Piscataway, NJ 08855. 2 Department of Physics, and Laboratory for Surface Modification, Rutgers University, Piscataway, NJ 08855. 3 Department of Molecular Physics, Moscow Engineering Physics Institute, Kashirskoe shosse 31, Moscow 115409, Russia. ABSTRACT
The growth of ultra-thin oxide films by the thermal oxidation of silicon has been studied by low and medium energy ion scattering spectroscopies (LEIS and MEIS) and X-ray photoelectron spectroscopy (XPS). To help elucidate the diffusional and mechanistic aspects of oxide growth we have used sequential isotope oxidation (1802 followed by 1602). LEIS demonstrates that both 180 and 160 atoms are on the silicon surface under our growth conditions. MEIS also distinguishes 180 from 160 and gives a depth distribution for both with high accuracy. Our results show that several key aspects of the Deal-Grove model (oxygen diffusion to the Si-SiO 2 interface and oxide formation at the interface) are consistent with our results for 50A films. For very thin oxide films (15A or less), we found a mixed isotopic distribution in the film, demonstrating more complex oxidation behavior. INTRODUCTION
Investigations of thermal oxide growth and properties of the Si/SiO2 interface region are of great interest to the applied microelectronics, basic surface science and thin films communities1 5" . Problems pertaining to ultra-thin film oxidation become more acute as MOS gate oxides drop well below 100A. However, despite many publications on this subject, the growth mechanism for very thin (100A) films is known to be described by the Deal-Grove Model (DGM) 6. According to this model, the oxide grows via molecular oxygen diffusion through the oxide film and reaction at the Si/Si0 2 interface. It has been shown",3, 5, however, that the oxidation kinetics 3 23 7 for ultrathin films differ from a simple DGM extrapolation. Several models ,8" have been proposed to account for this deviation, although most are based exclusively on kinetic data. The Reactive Layer Model (RLM) 5,24 was advanced recently as an attempt to integrate the observation of microcrystallinity near the interface25' 26 , a thin transition region of non-stoichiometric oxide 27 and results of experiments on silicon oxidation using oxygen isotopes 28. In contrast to the DGM, the RLM posits that oxidation takes place at the top (internal) surface of a 5-20A "reactive layer" between crystalline Si and a-Si0 2 . Although the RLM and other alternative models fit the experimental data on thin film oxidation kinetics better than the simple DGM, the large number of fitting parameters makes the physical interpretation questionable. Also, most of the models do not have decisive 69 Mat. Res. Soc. Symp. Proc. Vol. 318. ©1994 Materials Research Society
experimental support for their initial assumptions. This paper is intended to help elucidate the mechanism of very thin oxide growth during
Data Loading...
