Discovery of Lead-Free Halide Double Perovskites for Solar Cells via Interpretable Machine Learning

Developing novel lead-free perovskite materials with suitable band gaps and high stability is essential for advancing the use of perovskite materials in solar cells application. This research paper aims to identify lead-free halide double perovskites for solar cells using machine learning with the SISSO (sure independence screening and sparsifying operator) method. The study uses a dataset of 540 double halide perovskite compounds to apply the SISSO method for feature selection and identify key descriptors that can accurately predict the performance of these materials as solar cell materials. The SISSO method screens and selects the most important features from the dataset, and machine learning algorithms are then employed to identify patterns and relationships among the selected features. The use of machine learning and SISSO in conjunction with a large dataset of halide double perovskites allows for a more efficient and effective identification of promising solar cell materials compared to traditional methods. The results of this study may provide valuable insights into the design and optimization of lead-free halide double perovskites as solar cell materials.