Optimization of molecular beam epitaxial film thickness uniformity using Monte Carlo simulations and an artificial neural network

The thickness uniformity of the molecular beam epitaxial film is one of the most important factors affecting the quality of the film, and it is mainly influenced by the angular distribution of the molecular source, which is mainly determined by the inner wall shape of the crucible. In this paper, an optimization method based on particle swarm optimization, Monte Carlo simulations, and an artificial neural network is proposed, aiming at optimizing the epitaxial film uniformity in the molecular beam epitaxy process. The optimum angular distribution of an effusion source is obtained by using the method of particle swarm optimization for a given geometric configuration. The Monte Carlo method is used to simulate the particle evaporation process to obtain the relationship between the shape parameters of the crucible inner wall and the particle angular distribution. The optimum crucible shape parameters are subsequently obtained under a particular apparatus geometric configuration by using the artificial neural network according to the above relationship and the desired optimum angular distribution. Finally, the optimized results are compared by experiments. Published under an exclusive license by AIP Publishing.

Automated summary

In this paper, an optimization method based on particle swarm optimization, Monte Carlo simulations, and an artificial neural network is proposed to improve the thickness uniformity of molecular beam epitaxial films. The optimum angular distribution of the molecular source and the shape parameters of the crucible inner wall are determined to achieve the best film uniformity.