Numerical simulation and performance enhancement of CsBi3I10-based heterojunction solar cell with various semiconductor layers (CZTS, CZTGS, Al0.8Ga0.2Sb, GaAs) along with machine learning-based analysis

Abstract

Strategies to boost the efficiency of bismuth halide-based photovoltaic devices are being investigated, along with the positive ecological impacts of these solar cells. This study thoroughly examines the efficiency of a CsBi₃I₁₀-based heterojunction solar cell by employing diverse bottom absorber layers, with an emphasis on the impact of several aspects such as thickness and doping density of various layers, operating temperature and work function of the back contact on device performance. Efficiency has been elevated by determining an extremely effective GaAs semiconductor layer via an accepter concentration of 5 × 10¹⁶ cm⁻³ and enhancing its thickness. In the presented work, a novel CsBi₃I₁₀-based heterojunction PSC is designed as Au/NiO/GaAs/CsBi₃I₁₀/ZnSe/ITO. Optimizing a precise semiconductor layer for the excellent performance zone of our device is prior to progressing to the HTL and ETL layers. It has been identified that ZnSe and NiO exhibit the most efficient electron transport layer and hole transport layer properties. In addition, a Machine Learning model was employed to ascertain the optimized device performance by observing the progression of the output on input matrices. The heterojunction solar cell demonstrates superior performance, achieving an impressive efficiency of 27.40 %, an open circuit voltage (V_oc) of 1.03 V, a short circuit current density (J_sc) of 30.2 mA/cm^2, and a fill factor of 88.1 %. This represents a substantial improvement in efficiency, far exceeding that of the conventional CsBi₃I₁₀-based heterojunction solar cell. In that vein, this PSC architecture has emerged as a promising future device that is crucial to the fabrication of lead-free heterojunction PV devices.