Chemical Vapor Deposition: Simulation and Optimization

The deposition of thin films on to a solid substrate has become an important materials processing technique and is of interest in many applications such as those involved with the fabrication of micro-electronic circuits, optical and magnetic devices, etc. Thin films are generally deposited by the process of Chemical Vapor Deposition (CVD). This work is directed at the simulation and optimization of the CVD process in a vertical impinging CVD reactor for material fabrication, focusing on the rate of deposition and on the uniformity of the thin film obtained. This work considers the deposition of Silicon from Silane. Numerical simulations are used to determine the effect of important design variables on the deposition rate and film characteristics. Then Response surfaces are created using the Compromise Response Surface Method (CRSM) to approximate the responses for the range of design variables considered. Then stochastic optimization is performed using the Mean Value Method (MVM) on the resulting response surfaces to find the optimal values of the design variables for various levels of uncertainty.