Computational Analysis of the Interplay between Deep Level Traps and Perovskite Solar Cell Efficiency

New deposition methods of halide perovskites are being developed with the aim of improving solar cell power conversion efficiency by controlling the physiochemical properties of the perovskite film. In the case of methylammonium lead iodide (MAPbI₃), deep level traps limit efficiency by participating in charge carrier recombination. Prior work has shown that the solar cell efficiency of MAPbI₃ solar cells varied significantly with deposition method; specifically, efficiencies of 13.5 and 17.7% were observed for MAPbI₃ processed with a one- and two-step method, respectively. However, the origin of the difference in efficiency remains unclear. In this study, we analyze the interplay between deep level traps and efficiency by simulating the photoexcited charge carrier pathway across solar cells processed via the one- and two-step method using finite-element drift-diffusion modeling. We determined that in the case of one-step processing, the traps propagate throughout the bulk, while for two-step, the traps congregate at the interface where the MAPbI₃ was grown (mesoporous TiO₂). Composition and structural analysis are used to propose a plausible explanation as to why the difference in processing changes the spatial distribution of the traps.