Optically-Induced, Room-Temperature Oxidation of Gallium Arsenide
- PDF / 251,447 Bytes
- 5 Pages / 420.48 x 639 pts Page_size
- 3 Downloads / 216 Views
OPTICALLY-INDUCED,
ROOM-TEMPERATURE OXIDATION OF GALLIUM ARSENIDE
Chien-Fan Yu, Michael T. Schmidt, Dragan V. Podlesnik, and Richard M. Osgood New York, Jr., Microelectronics Sciences Laboratories, Columbia University, NY 10027
ABSTRACT Room-temperature, optically-induced oxidation of the gallium arsenide surface has been studied with laser radiation of different wavelengths. It was found that deep-ultraviolet light is much more effective in enhancing oxidation than near-ultraviolet or visible light. The growth rate of the oxide was also found to be drastically increased by the presence of chemisorbed water molecules on the surface.
INTRODUCTION The study of optically-induced oxidation of the gallium arsenide surface has been a subject of intensive investigation[1-5]. These studies have shown that simultaneous exposure of a semiconductor surface to oxygen and to laser light can stimulate the uptake of oxygen. It is also known that a thin oxide layer formed on the gallium arsenide surface can be used to passivate the surface or to alter the interface properties[6]. A better understanding and control of this room-temperature process can lend itself to many applications in the fabrication of semiconductor devices. The enhancement of oxidation on the gallium arsenide surface through aboveband-gap photon illumination has been generally attributed to the effect of photogenerated carriers[1-5]. We report here our recent study on the wavelength dependence and the effect of water molecules adsorbed on the gallium arsenide surface in this photon-induced process. The use of deep ultraviolet light, below the oxygen dissociation limit, and/or the presence of chemisorbed water molecules on the surface were found to increase significantly the growth rate of the oxide.
EXPERIMENTAL Experiments were carried out with either an argon-ion laser tuned to one of its visible or near-uv lines, or with a frequency-doubled argon-ion laser to generate 257-nm deep uv light. An excimer laser was also used as an alternative to generate 248-nm (KrF) light. In all cases the dissociation limit of oxygen molecules (242 nm) was not exceeded. The laser power was kept sufficiently low so that the temperature rise on the sample durin§ exposure was negligible. The typical cw power density used was 50 mW/cm . The gallium arsenide saT9les used ýn this experiment were n-type of (100) orientation with - 10 atoms/cm doping density. Surfaces were cleaned by the following steps before ex-situ experiments: degreasing in warm trichloroethylene, immersion in acetone, immersion in methanol, immersion in DI water, and 50% NH4 OH rinse. Samples were then blown dry and attached to the mount with silver paste. For in-situ experiments, performed inside an UHV chamber, additional surface cleaning was done by heating to 540'C, followed by 500 eV argon-ion bombardment and then 540'C annealing for 5 minutes to reduce sputtering damage. Surface analysis was performed using a multiprobe surface spectroscopy system equipped with a concentric hemispherical energy analyzer
Data Loading...
