Volume
Volume 4, Issue 2 (2024) – 5 articles
Cover Picture: Light-induced segregation limits the practical application of mixed halide perovskites in solar cells. Herein, we adopted a data-driven approach to predict perovskite experimental bandgaps for further studying halide segregation in perovskites. Firstly, we utilized graph neural networks to establish a unique database of mixed perovskite bandgaps with continuous compositions. Data analysis revealed that the bandgap increases as the ionic radii at the A and X sites decrease. Subsequently, by leveraging this database, we discovered a strategy to suppress halide segregation. Specifically, using a higher Cs content at the A-site, rather than MA, can reduce the bandgap difference, which can enhance the performance of perovskite solar cells. The proposed data-driven approach facilitates the targeted design of novel mixed-composition perovskites and aids in the investigation of halide perovskite segregation.
view this paper