Performance assessment and machine learning-driven optimization of Ca3NBr3-based bifacial perovskite solar cells: improving VOC via HTL and charge transport layer analysis

This study examines the optoelectronic characteristics of Calcium Nitride Bromide (Ca3NBr3) as a prospective absorber material for heterojunction solar cells. This work assesses the efficacy of two-hole transport layers (HTLs), MASnBr3 and P3HT, in conjunction with two electron transport layers (ETLs), C60 and ZnO, through SCAPS-1D modeling. By optimizing layer thickness, doping concentrations, defects, and recombination parameters, a peak power conversion efficiency (PCE) of 28.76 % was obtained using MASnBr3 as the HTL and ZnO as the ETL. The enhancement in performance is ascribed to the reduced recombination losses at the absorber/HTL interface, resulting in increased open-circuit voltage (VOC) and overall efficiency. The device exhibited notable bifacial performance, with an efficiency of 32.83 % and a bifacial gain of 17.69 %. A machine learning model was developed to predict solar cell performance, attaining an accuracy of 82.75 %. The findings indicate that Ca3NBr3 may improve the performance of perovskite-based solar cells.