Study on the uniformity of ZnO films grown by MOCVD

Abstract This study reports the use of CFD simulation software to simulate the growth process in a ZnO-MOCVD reaction chamber. The results of the rate of ZnO deposition in the simulation were in good agreement with experimental results and the correlation coefficient in all cases indicates medium and high correlation, which verified the correctness of the theoretical model of the numerical simulation. A streamline diagram of the reaction chamber and the technicians’ observations of this growth show that the adjustment of the entrance of the source of MO is key to obtaining a substrate film of uniform thickness. In order to obtain a good film deposition rate, we created a neural network model combined with a genetic algorithm to optimize the entrance flow of the MO source and determined the optimal input process parameters for film uniformity. The optimization results and experiments were compared and analyzed to obtain satisfactory consistency. Following the optimization, the coefficient of variation decreased from 3.6% to 1.28%, which significantly improved the uniformity of film thickness of ZnO film. These results will provide a solution to process parameter adjustment and high-quality epitaxial growth in the metal-organic chemical vapor deposition process.