Inorganic Cs2TeI6/Cs2AgInBr6 Dual‐Absorber Perovskite Solar Cells: SCAPS‐1D Optimization and Multiple Machine Learning Comparison Analysis

Abstract

Perovskite solar cells (PSCs) have garnered sustained focus of the academic researchers owing to their exceptional light‐harvesting capabilities and solution manufacturing fabrications. Here, nontoxic inorganic materials Cs2TeI6 and Cs2AgInBr6 (CAIB) are employed as a dual‐absorber layer, which enhances the spectral range, while accelerating the extraction efficiency of carriers through the distribution of space charges, with an initial power conversion efficiency (PCE) of 10.99%. On a Solar Cell Capacitance Simulator in One Dimension (SCAPS‐1D) platform, after an in‐depth examination the bulk defect density, the interface defect density, the bandgap, and the thickness of the absorber layer, an exciting PCE at 31.60% is achieved, with the back electrode substituted by high‐efficiency and durable carbon. Afterward, incorporating realistic radiative recombination and resistance effects, PCE is refined to 24.70% with enhanced experimental feasibility. Notably, the proposed PSC demonstrates exceptional property in temperature and illumination evaluations, followed by discussions of capacitance. Furthermore, five algorithms are employed to construct models of PSC, striving to accurately represent the relationship between various property parameters and certain indicators of the absorber layer. Among them, eXtreme gradient boosting is identified as the most appropriate algorithm, with CAIB layer defect density dominating PCE, providing the optimal optimization strategy for future PSC designers.