Ghazi Aman Nowsherwan
This study presents a comprehensive investigation of the structural, electronic, optical, and photovoltaic properties of the lead-free double perovskite Cs 2 InSbCl 6 using density functional theory (DFT), SCAPS-1D device simulations, and machine learning (ML) modeling. DFT analysis confirms a stable cubic structure with a tunable bandgap and strong optical absorption in the visible spectrum, with corresponding mobilities of 62.9 and 12.9 cm²/V·s, respectively, indicating electron-dominated transport. The effective density of states values at 300 K were found to be Nc = 3.72 × 10¹⁹cm⁻³ and Nv = 1.26 × 10¹⁹cm⁻³. SCAPS-1D simulations revealed that the baseline Cs 2 InSbCl 6 -based device achieved a PCE of 20.71%. This performance was obtained at an active layer thickness of 500 nm and a defect density of 10¹⁵ cm⁻³, with a work function of 5.1 eV, under standard 1-sun illumination at room temperature. These results are in close agreement with the benchmarks previously reported. Through systematic optimization of thickness, defect density, doping, resistances, and contact parameters, the efficiency was significantly enhanced to 28.34%, demonstrating the strong potential of Cs 2 InSbCl 6 for high-performance, lead-free perovskite solar. Comparative validation across SCAPS-1D, COMSOL, wxAMPS, AFORS-HET, and OghmaNano confirmed consistency in predicted performance trends. ML modeling further enhanced predictive capability, where Extreme Gradient Boosting (XGBoost) outperformed other algorithms with the lowest error (MSE = 0.0036 for PCE) and highest accuracy (R² >0.9994). Feature-importance analysis identified defect density, back-contact work function, and active layer thickness as the most influential parameters. These findings demonstrate that Cs 2 InSbCl 6 is a promising candidate for eco-friendly PSCs, with simulation-driven optimization providing clear design guidelines for next-generation PSC.