Computational fluid dynamics modeling of SIC chemical vapor deposition.

摘要

Computational fluid dynamics (CFD) modeling was employed to simulate chemical vapor deposition (CVD) of silicon carbide. Influences of precursor concentration, feed flow rate, temperature, pressure, heat transfer and reactor geometry on SiC CVD were investigated systematically. The increase in SiC deposition rate with increasing precursor concentration, increasing feed flow rate and decreasing deposition temperature can all be attributed to reactant depletion along the gas flow direction. Silicon vapor homogeneous nucleation Silicon leads to a lower SIC deposition rate. The Soret effect reduces the SiC deposition rate by ∼30%. SiC deposition uniformity along the flow direction can be improved by using the hot wall condition. A low pressure reduces the silicon vapor nucleation. Narrowing of reactor tube gives higher deposition rate. Some of the simulation results were examined experimentally and a good agreement was obtained.; A solid vanadium source was employed for in situ vanadium doping of SiC epilayer. Vanadium doping at the 1015 cm−3 level was achieved. This is the first step for semi-insulating SIC epilayer growth.