Computational Modeling for the Development of CVD SiC Epitaxial Growth Processes

Computational modeling is a powerful tool for development of new SiC CVD epitaxial growth processes. Growth dependencies that can be reliably predicted using simulation were investigated by correlating simulation results with results of epitaxial growth. It was shown that for the mass transfer-controlled growth typical for SiC epitaxy, the growth rate can be predicted with a sufficient precision even if elaborate models of the surface reactions are not available. Depletion of the precursors in the gas phase along the growth direction was shown to be one of the most important sources for the growth rate non-homogeneity. The effective Si/C ratio determined by simulation can help in predicting doping non-homogeneity. Simulation of the degree of precursor supersaturation above the growth surface can be used as a good estimate of the morphology degradation, which can help in specifying the window for the optimal growth parameters during new process development. Introduction The major requirements for commercial grade SiC epitaxial layers are low concentrations of both defects and undesirable impurities, good surface morphology, and precision thickness and doping control. The critical characteristics of epitaxial layers are influenced by the temperature in the reactor and temperature distribution, the rate of precursor cracking (e.g., through the degree of Si and C supersaturation above the substrate), the degree of concurrent surface etching by the carrier gas (hydrogen), the effective C/Si ratio above the substrate, the gas flow velocity, etc. Even moderate changes in the reactor size and geometry can influence at least some of the growth conditions mentioned above. Any change in the reactor requires a tedious search for the new optimal growth settings. Computational Fluid Dynamics (CFD) simulation is an effective approach for increasing the efficiency of new CVD process development [1,2,3]. In our work, CFD simulations were used to determine the factors that are the most critical for arriving at optimal process conditions. Extensive practical understanding of the CVD epitaxial growth process in the cold wall CVD reactor was applied to identify and refine the most relevant simulation models to be used in further computational experiments for new process development. Experimental verification of key simulation results was accomplished. Experimental and Simulation A traditional silane/propane precursor-based CVD systems with a hydrogen carrier gas was used to grow SiC epilayers on commercial SiC substrates. The growth was conducted at 1550C in an atmospheric pressure reactor with a 100-mm diameter quartz reaction tube equipped with an outer water jacket to permit “cold wall” processing. A 120-mm silicon-carbide coated graphite susceptor was supported on a quartz boat inserted into the reaction tube. A solenoidal induction coil surrounds the quartz reaction tube and radio-frequency electrical current is used to heat the susceptor to the growth temperature. The thickness of the grown epilayers was characterized using FTIR reflectance spectroscopy. Surface morphology was examined by optical microscopy. Mercury-probe Capacitance-Voltage (C-V) measurements were used to determine the net free carrier concentration. Low temperature Materials Science Forum Online: 2003-09-15 ISSN: 1662-9752, Vols. 433-436, pp 177-180 doi:10.4028/www.scientific.net/MSF.433-436.177 © 2003 Trans Tech Publications Ltd, Switzerland All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans Tech Publications Ltd, www.scientific.net. (Semanticscholar.org-12/03/20,10:56:42) photoluminescence (PL) measurements were performed in order to evaluate quality of the epilayers and incorporation of impurities during epitaxial growth. A commercial Computational Fluid Dynamics (CFD) software package was used in this work. The computational models simulated the gas flow pattern, thermal and mass diffusion, as well as gas phase reactions and simplified surface reactions. A reduced set of chemical reactions was developed on the basis of literature data [4,5]. Post-processing of the simulation results included analysis of the mass fractions of the gas phase reaction products at different points in the simulated reactor as well as the mass flux of the species adsorbed onto the surface. This information was used to estimate the degree of supersaturation of the growth species above the substrate, to calculate the growth rate on the basis of known density of the deposited material, and also to determine the effective C/Si ratio at different positions along the susceptor length. Results Comparison of the simulated and experimental growth rates is shown in Fig. 1. Five identical samples of 4H-SiC substrate were placed at different locations along the susceptor, with Sample A located 3 mm away from the leading edge of the susceptor and Sample E located 3 mm away from the downstream edge of the susceptor. A few times higher growth rate at the upstream region of the