Low-temperature epitaxial silicon using hot wire chemical vapor deposition

This thesis answers the question that precipitated from the recent discovery of a low-temperature epitaxial silicon growth process at the National Renewable Energy Laboratory (NREL). The question is: Can hot wire chemical vapor deposition (HWCVD) be "used" to grow electronic-quality, single-crystal layers on crystalline silicon? To answer this question I demonstrate the first reproducible growth of epitaxial silicon using HWCVD. Then to answer questions pertaining to process conditions which determine crystallographic properties of the epitaxial layers, I develop first-principles solutions to the problems of diffusion through a moving boundary, a theory of the limited epitaxial thickness at low-temperatures, a preliminary theory for the effects of filament condition/lifetime as a function of growth parameters as well as an empirical formulation of HWCVD growth kinetics in terms of growth parameters. These results are combined into a composite model which defines the reactor conditions requisite for successful epitaxy. This model allows one to both understand the HWCVD epitaxial process and formulate an answer to the thesis question. The answer being that under current process conditions, HWCVD epitaxial silicon growth can not produce epitaxial layers with enough crystallographic perfection to form electronic-grade silicon. However, the analysis performed in this work points to process modifications which may allow successful epitaxial growth using this technique.