DID YOU KNOW? DTIC has over 3.5 million final reports on DoD funded research, development, test, and evaluation activities available to our registered users. Click
HERE to register or log in.
Accession Number:
ADA326093
Title:
Comparison of Electron Cyclotron Resonance Dry Etch Chemistries for Silicon Carbide.
Descriptive Note:
Master's thesis,
Corporate Author:
AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH
Report Date:
1997-06-03
Pagination or Media Count:
91.0
Abstract:
Plasma etching using an electron cyclotron resonance ECR source has been used to study dry etching of SiC. Gas chemistries studied include C12Ar, Cl2 H2, SF6Ar, IBrAr, and Ar. The variables studied include ECR power 400 to 1000 W, RF power 50 to 350 W, chamber pressure 1.5 to 10 mTorr, and active to total gas flow rate ratios 0 to 1. Etch rates in excess of 2000 Amin. have been achieved with SF6Ar ECR discharges at low power 400 Watts and moderate RF power 100-150 Watts. The rates for Cl2Ar are comparable to SF6Ar rates with those for Cl2 H2 being somewhat lower. IBrAr is found to be a very poor chemistry for SiC with etch rates reaching only 1000 Amin. Pure Ar sputtering achieved rates of less than 800 Amin. Smooth surfaces 0.6-0.8 nm RMS roughness have been achieved in all plasma chemistries with no dependence of roughness on plasma conditions. Auger electron analysis confirmed carbon enrichment of the surface in most cases. High RF biases resulted in stoichiometric termination of the SiC surface. Silicon depletion was noted for low RF biases and high chlorine content plasmas. Contamination by F and S was noted in SF6Ar discharges. All gas chemistries studied favored the chemically dominant side of the theoretical etch rate curve with the exception of pure argon, which favored the sputter dominant side. Chlorine showed the most synergistic effects and covered the largest range of the theoretical etch rate curve for the experimental conditions studied.
Distribution Statement:
APPROVED FOR PUBLIC RELEASE
