Particle Swarm Optimization Based Technique for Deposition Parameters Optimization of Magnetron Sputtering Process

Abstract

This study presents a soft computing based technique in the deposition parameters optimization of RF Magnetron Sputtering process. Particle Swarm Optimization (PSO) has been chosen due to its good performance in solving various optimization problems. The material used in this study was zinc oxide (ZnO) and there were four deposition parameters involved the optimization process. The deposition parameters were RF power, deposition time, oxygen flow rate and substrate temperature. The aim of the study was to obtain the optimal combination for the selected deposition parameters in order to produce the desirable ZnO thin film properties. In this study, the Desirability Function had been adapted as the fitness function for PSO. Desirability function is one of the commonly used statistical method for obtaining optimal process parameter design. The result from the PSO based optimization technique was then compared with actual laboratory result. Based on the observation made, the PSO based technique has been proven to be reliable and satisfactory in obtaining the optimal deposition parameters of ZnO thin film. It is expected that this soft computing based technique for optimizing the deposition parameters could reduce the trial and error method before the experiment is conducted in the fabrication process.