A Numerical and Experimental Study on the Fabrication GaN Films by Chemical Vapor Deposition

Abstract Computational modeling and simulation are employed to study a rotating susceptor vertical impinging chemical vapor deposition (CVD) reactor to predict GaN film deposition. Many metal-organic chemical vapor deposition reactor manufacturers use prior experience to design and fabricate CVD reactors without a fundamental basis for the process and information on the optimal conditions for the deposition. Through trial and error, they fine tune the gas flow parameters, heater temperatures, chamber pressure, and concentration of species gases for optimal growth. However, expensive raw precursor gas and time are wasted through this method. A computational model is an important step in the CVD reactor design and GaN growth prediction. It can be used to model and optimize the reactor to yield favorable operating conditions. In this paper, a simple geometry consisting of a rotating susceptor and flow guide is considered. The focus is on gallium nitride (GaN) thin films. The study shows how the computational model can benefit reactor design. It also presents comparisons between model prediction results and experimental data from a physical, practical, system. Commercially available software is used, with appropriate modifications, and the results obtained are discussed in detail.