Reducing Search Space for Halide Perovskites: Comparing New Simple Material Model (NSMM), Coin-Flip, Goldschmidt’s Tolerance Factor Formalism (GTFF), and SPuDS Algorithm

Abstract

Halide perovskites are of recent interest for increasing the efficiency of photovoltaic energy conversion devices. The search space for halide perovskites is daunting considering there are over 1700 potential “simple” halide perovskites, not to mention their solid solutions and the additional enormous number of more complex halide perovskites and their solid solutions. Therefore, some questions become: which combination of atoms will form halide perovskites, which will not, and for those that do form halide perovskites, what will their properties be? Goldschmidt’s tolerance factor formalism (GTFF), including the more recent Structure Prediction Diagnostic Software (SPuDS) algorithm based on GTFF, have been used over roughly the past century to guide discovery and development of perovskites. Here, four tractable methods: GTFF; SPuDS; coin-flip; and, “new simple” material model (NSMM) are used to investigated which of 864 potential simple halide perovskites should and which should not form. Output of GTFF, SPuDS, coin-flip and NSMM are compared to the literature, including the Joint Committee on Powder Diffraction Standards (JCPDS) files, and scored in a straightforward manner. From analysis of scoring, NSMM is shown to significantly outperform GTFF, including SPuDS, and the coin-flip methods for identifying which simple halide perovskites will and will not form.