Identifying Pb-free perovskites for solar cells by machine learning

Hybrid halide perovskites are one of the new-age solar cells that are expected to solve the world's energy problems. In particular, lead-based halogen compounds with Pb2+ at the B-site, which have been most widely studied in photovoltaic applications. These materials can be easily produced at low cost, but they have drawbacks such as chemical instability and toxicity. Nakajima et al. search for novel materials for lead-free perovskite solar cells using the computational screening technique. In this study, we attempt to explore the candidate compounds of perovskite materials suitable for solar cells using statistics and multiple regression analysis, building predictive models and machine learning among these new material candidates.