Enhancement of Sulfide Based Absorber and Charge Transport Layer Solar Cell Performances Using Machine Learning and the SCAPS-1D Simulator

Abstract

Lead (II) sulfide (PbS)-integrated solar devices are now attracting attention from scientists because of their extraordinary semiconducting attributes. The anticipated outcome is not obtained due to a number of problems, including incompatibility of the band topologies at the absorber/ETL and HTL/absorber interfaces, recombination of carriers at the front and rear metal contacts, and others. In addition to examining the impacts of the SnS HTL and SnS2 ETL layers on the performance parameters, the primary focus of this work is to optimize the layer’s properties of the lately suggested Al/FTO/SnS2/PbS/SnS/Ni photocell. The SCAPS simulation program was used to conduct this study. A higher performance has been achieved by analysing the performance characteristics, which include change in defect concentration of every stratum, thickness, doping concentration, capacitance (C)-voltage (V), interfacial defects, operating temperature, resistance, and front and back metals. In a thin (900 nm) PbS layer’s thickness, this device works very well at lower acceptor density (1×1017 cm-3). The Al/FTO/SnS2/PbS/Ni reference cell's PCE, VOC, JSC, and FF values were calculated to be 22.96%, 0.99 V, 26.99 mA/cm2, and 84.08%. Besides, the Al/FTO/SnS2/PbS/SnS/Ni recommended structure, which introduces SnS between the PbS and Ni, has PCE, VOC, JSC, and FF values of 31.43%, 1.12 V, 31.46 mA/cm2, and 89.10%. After that, we developed a machine learning (ML) model to predict the output parameters of the photo devices. By using ML, the performance matrix of the photocells under study was predicted with an accuracy rate of almost 83.75%. The suggested study could provide light on the matter and a workable method for yielding a PbS-based photovoltaic cell at a reasonable price.