[13a-A403-1~9] 16.3 Bulk, thin-film and other silicon-based solar cells

△ [13a-A403-6] Improvement of Si Solar Cells by Bayesian Optimization for Diffusion Process

〇Shogo Kunieda¹, Kentaro Imamura¹, Hikaru Kobayashi¹ (1.ISIR, Osaka Univ.)

Keywords:crystalline Si solar cell, Bayesian optimization, machine learning

Optimization of experimental parameters such as etching duration and diffusion temperature is of great importance to improve solar cell characteristics. In recent years, machine-learning-assisted process optimization draws much attention in various fields. In this study, optimization of diffusion process for black Si solar cells having an ultra-low reflectance nanocrystalline Si layer has been accelerated using the Bayesian optimization method, and the conversion efficiency higher than 20% has been achieved for simply structured Si solar cells without an anti-reflection coating layer.