Machine Learning-Assisted Optimization of Cu-Based HTLs for Lead-Free Sr 3 PBr 3 Perovskite Solar Cells Achieving Over 30% Efficiency via SCAPS-1D Simulation

Abstract

This study investigates the optimization of Sr 3 PBr 3 -based lead-free perovskite solar cells (PSCs) by evaluating five copper-based hole transport layers (HTLs): Cu 2 O, CuI, CuSbS 2 , CuSCN, and CBTS (Cu 2 BaSnS 4 ). The device structure, FTO/SnS 2 /Sr 3 PBr 3 /HTL/Au, was simulated using SCAPS-1D to assess the influence of each HTL on photovoltaic performance. Results reveal that HTL selection plays a critical role in determining output parameters, strongly impacting efficiency and stability. Among the candidates, CBTS delivered superior performance, achieving power conversion efficiency (PCE) of 30.78%, open-circuit voltage (V OC ) of 1.32 V, shortcircuit current density (J SC ) of 26.82 mA/cm², and fill factor (FF) of 87.05%. Additional optimization considered absorber thickness, doping level, defect density, series resistance, and operating temperature. A machine learning model was employed to predict device performance with exceptional accuracy, averaging 99.6%. To interpret parameter significance, SHAP (SHapley Additive exPlanations) analysis identified the most influential factors governing efficiency. These findings highlight CBTS as a promising, non-toxic HTL alternative and demonstrate the combined power of simulation and AI-driven analysis for device optimization.This work provides valuable guidance for designing stable, high-efficiency, and environmentally sustainable perovskite solar cells.