Enhancing Organic Solar Cells through Dual Electron Transport Layers: A Machine Learning Approach for Prediction of Organic and Metal Oxide Electron Transport Layer Compatibility

This research investigates the effect of incorporating dual cathode buffer layers on the performance of photovoltaic devices, focusing on ITO/PEDOT:PSS/PM6:PY‐IT/PCBM/metal oxide/Ag architectures where PCBM is combined with ZnO, TiO 2 , and SnO 2 as inorganic electron transport layers (ETLs). Key parameters such as film thickness, doping concentration, interfacial defects, Mott–Schottky behavior, and impedance characteristics are thoroughly analyzed. Among the tested configurations, the most efficient structure is PCBM/TiO 2 , which achieves an efficiency of 19.35% with an open circuit voltage of 1.14 V and a fill factor of 73.76%. To improve device optimization, this study introduces an artificial intelligence (AI)‐based approach using the SCAPS‐1D simulator. This method enables accurate efficiency prediction and optimal selection of double organic/inorganic ETL combinations. Various regression models are applied, including linear regression, decision tree regressor, random forest regressor, gradient boosting regressor, XGBoost regressor, and K‐neighbors regressor. Among them, XGBoost regressor shows the highest precision with R 2 (test) = 0.9957, root‐mean‐square error = 0.2446, and R 2 (train) = 1, highlighting the potential of AI in refining photovoltaic design.