Investigating Optoelectronic Characteristics and Improving the Efficiency of Mg₃AsBr₃ Perovskite Solar Cells through Machine Learning and Numerical Simulations Utilizing Diverse Charge Transport Materials

This study investigates the optoelectronic characteristics of cubic perovskite Mg₃AsBr₃ for photovoltaic (PV) applications through first-principles density functional theory (DFT), driven by the increasing interest in perovskites for renewable energy solutions. Mg₃AsBr₃ is explored as an absorber layer in conjunction with Cu₂O as the hole transport layer (HTL) and various electron transport layers (ETLs), specifically WS₂, ZnO, PC₆₀BM, and C₆₀. SCAPS-1D simulations were employed to optimize parameters including doping concentration, layer thickness, and defect density in the charge transport and absorber layers. The results show significant variations in power conversion efficiency (PCE) depending on the ETL choice. The Al/FTO/WS₂/Mg₃AsBr₃/Cu₂O/Au configuration exhibited the optimal performance, achieving a <i>V</i>_OC of 1.03 V, an FF of 88.06%, a PCE of 32.55%, and a <i>J</i>_SC of 36.01 mA/cm². Configurations utilizing ZnO, PC₆₀BM, and C₆₀ as ETLs attained PCE of 32.47, 32.21, and 31.63%, respectively. This underscores the significance of choosing the appropriate ETL for optimal perovskite solar cell (PSC) performance. The study assesses aspects including band alignment, defect density, doping concentration, and series-shunt resistances that affect device efficiency and durability. The SCAPS-1D results were validated against wxAMPS simulations, and a machine learning model was created, forecasting essential performance metrics with 84% accuracy. The proposed optimized configurations improve the efficiency and sustainability of PSCs.