AVS 46th International Symposium
    Electronic Materials and Processing Division Thursday Sessions
       Session EM2-ThM

Paper EM2-ThM6
Study of GaAs Oxidation in H@sub 2@O@sub 2@ and H@sub 2@O using Spectroscopic Ellipsometry

Thursday, October 28, 1999, 10:00 am, Room 612

Session: Dielectric Passivation/Oxides on Compound Semiconductors
Presenter: S.-J. Cho, University of Nebraska, Lincoln
Authors: S.-J. Cho, University of Nebraska, Lincoln
P.G. Snyder, University of Nebraska, Lincoln
Correspondent: Click to Email
Oxidation of GaAs in hydrogen peroxide (H@sub 2@O@sub 2@) and deoinized H@sub 2@O (DH@sub 2@O) at room temperature was studied using in-situ real time spectroscopic ellipsometry (RTSE) and ex-situ spectroscopic ellipsometry (SE). GaAs samples were immersed in H@sub 2@O@sub 2@ for periods of up to 2 hours while RTSE data were recorded, then rinsed in flowing DH@sub 2@O and blown dry. SE data (1.5-5.5 eV) were taken before and after immersion. Analysis of the SE data indicated the development of a 2-3 nm interface (modeled as porous GaAs) between the oxide and GaAs, which was not present before immersion. The oxide itself had also become more porous as well as thicker. Accurate modeling of the interface and oxide layers required data in both the E@sub 2@ (~4-5 eV) and E@sub1@ (~2.5-3.5 eV) spectral regions. The RTSE spectral range was limited to below 3.5 eV by UV absorption in the H@sub 2@O@sub 2@ ambient, so the interface could not be included in the real time analysis. Another problem was the formation of bubbles on the surface, which became visible after about 30 minutes. Within these limitations on the RTSE data, their analysis indicated that the oxide growth was approximately logarithmic over at least the first 10-15 minutes. The oxidation rate varied from 0.21 to 0.54 nm/decade (minutes). In DH@sub 2@O no bubbles formed, but an interface again appeared to be present. Oxidation was nearly linear in time, at a higher rate than in the H@sub 2@O@sub 2@, and with a higher void volume (more porous oxide). Growth rates again varied considerably, with a typical rate 0.16 nm/min. Research supported by AFOSR Grant #49620-96-1-0480.