Integrating Machine Learning With <scp>SCAPS</scp> ‐ <scp>1D</scp> to Optimize Lead‐Free <scp>LCCZO</scp> Perovskites for Sustainable Solar Energy Harvesting

ABSTRACT Perovskite solar cells (PSCs) have rapidly advanced as a next‐generation photovoltaic technology due to their high efficiency and flexible fabrication routes; however, the reliance on lead‐based absorbers raises concerns regarding toxicity and long‐term stability. This study introduces La 0.5 Ce 0.5 Co 0.9 Zn 0.1 O 3 (LCCZO), a double perovskite oxide, as a lead‐free, stable absorber candidate for eco‐friendly PSCs. First‐principles calculations validate its crystallographic stability and predict a direct bandgap of ~1.8 eV, suitable for efficient visible‐light absorption. Optical analysis reveals strong absorption and favorable dielectric properties, supporting low exciton binding and effective charge separation. Device simulations using SCAPS‐1D integrating TiO 2 and NiO as transport layers demonstrate optimal performance upon tuning absorber thickness, carrier density, and defect levels. A maximum power conversion efficiency (PCE) of 23.41% is achieved, with an open‐circuit voltage ( V oc ) of 1.42 V, short‐circuit current density ( J sc ) of 18.56 mA cm −2 , and fill factor (FF) of 88.92%. The device exhibits robust tolerance to bulk and interfacial defects, attributed to favorable band offsets and intrinsic defect chemistry. A machine learning model trained on simulated datasets accelerates performance prediction across parameter regimes, achieving ~82.5% accuracy and reducing computation time significantly. These results position LCCZO as a promising lead‐free absorber and highlight the synergy of computational design and machine learning for sustainable solar energy development.