Dataset of SCAPS-1D simulated halide perovskite solar cells with SHAP and machine learning-based PCE optimization.

This data article describes a dataset generated by using SCAPS-1D simulation software to capture photovoltaic performance metrics for perovskite solar cells (PSC). The data collection process involved systematically modeling a range of device configurations by varying perovskite compositions, layer thicknesses, and combinations of charge transport layers. Key parameters such as open-circuit voltage, short-circuit current density, fill factor, and power conversion efficiency were recorded, with simulation outputs provided in Excel files. Additional files include detailed material databases featuring physical properties of electron and hole transport layers-such as band gap energies, electron affinities, dielectric permittivity, and carrier mobilities-organized in separate spreadsheets. Moreover, the repository offers a pre-trained CatBoost machine learning to predict the solar cell efficiency, along with a Jupyter Notebook that outlines the data preprocessing, machine learning workflow, and feature importance analysis via SHAP values. The structured dataset supports a reproducible simulation environment and is designed to facilitate further research in computational materials science and renewable energy. Its comprehensive organization makes it suitable for applications including photovoltaic device optimization, evaluation of simulation-based predictive approaches, and development of advanced data-driven models. Publicly hosted on Zenodo, this dataset provides an accessible resource for researchers and practitioners aiming to explore and enhance perovskite solar cell configurations through computational analysis and machine learning techniques.