Advancing HTL‐Free Cs ₂ PtI ₆ Carbon Perovskite Solar Cells: Insights from Hybrid Simulation and Machine Learning

ABSTRACT The primary causes of the high cost of perovskite solar cells are metal electrodes and hole transport layers. In this theoretical work, we examine the outputs of a hole transport layer‐free carbon‐based solar cell with an FTO/ETL/Cs 2 PtI 6 /Carbon electrode structure using the Solar Cell Capacitance Simulator (SCAPS‐1D). The paper studied various carbon electrode types‐Graphene/Carbon Black (G/CB) (5 eV), Graphene (4.9 eV), Graphene Oxide (GO) (4.8 eV), and Bio‐carbon (4.5 eV)‐ and electron transport layers‐SnO 2 , TiO 2 , LBSO, and WO 3 . The studied parameters included perovskite and ETL layer thickness, doping density, and defect density. The outputs showed that the best PCE of 15.50% resulted from using G/CB electrode and TiO 2 as the ETL, with a thickness of 0.09 µm, and a doping density of 10 × 10 19 cm −3 . Additionally, for the Cs 2 PtI 6 absorber layer, a Cs 2 PtI 6 composition with a thickness of 1.2 µm, a defect density of 1× 10 15 cm −3 , and a doping density of 10 × 10 12 cm −3 demonstrated superior performance, resulting in a PCE of 15.50%. These findings suggest that the FTO/TiO 2 /Cs 2 PtI 6 /G/CB structure, particularly with optimized TiO 2 and Cs 2 PtI 6 layers, holds great potential for hole transport layer‐free‐carbon‐based solar cell fabrication. Furthermore, machine learning models with a random forest algorithm evaluated the relative importance of the features on cell efficiency, and predicted the efficiency of the suggested configuration with R 2 of 0.93 underscoring the potential of machine learning in enhancing solar cell design and performance.