Investigating the effectiveness of Ca3AsCl3-based Perovskite Solar Cells with optimal hole transport layer selection through numerical optimization and machine learning

Abstract

Perovskite solar cells (PSCs) are gaining attention due to their superior photovoltaic (PV) performance compared to traditional PV cells. They offer several advantages, including low cost, simple manufacturing processes, tunable bandgap energy, and excellent electrical and optical properties. In this study, SCAPS-1D software was utilized to design a novel lead-free PSC structure: FTO/TiO₂/Ca₃AsCl₃/CBTS/Ni. The absorber layer consists of lead-free Ca₃AsCl₃, and various inorganic hole transport layers (HTLs), including CBTS, CuI, CuO, CFTS, and CuSCN, were analyzed to enhance efficiency. The maximum permitted defect densities were established, and important parameters including the absorber and electron transport layers' (ETL) thickness and doping concentrations were tuned. After optimization, the PV properties of the solar cell showed significant improvements. Without the CBTS HTL layer, the device achieved a power conversion efficiency (PCE) of 17.11 %, a fill factor (FF) of 86.23 %, a short-circuit current density (J_SC) of 14.8 mA/cm² and an open-circuit voltage (V_OC) of 1.34 V. However, incorporating the CBTS HTL layer increased the PCE to 23.70 %, with a J_SC of 19.14 mA/cm², an FF of 89.28 %, and a V_OC of 1.39 V. Additionally, a Random Forest machine learning approach forecasts optimal PV parameters by examining essential material characteristics, including layer thickness, bandgap, and carrier mobility. The model predicts performance with a remarkable correlation coefficient (R²) of 0.94 for PCE. This approach improves comprehension of material optimization and provides essential data to produce cost-effective, efficient Ca₃AsCl₃-based PSCs, boosting progress in solar energy technology.