High-throughput computations and machine learning for halide perovskite discovery

Halide perovskites are materials of considerable interest for solar cells, photodiodes, LEDs, photocatalysis, and photorechargeable batteries. One the most attractive features this class is sheer tunability their stability, electronic bandgaps, optical absorption behavior, defect properties, via composition engineering, phase transformation, change in dimensionality, surface interface octahedral rotation distortion. Due to ease simulating well-defined crystal structures systematically investigating compositional structural factors that affect first-principles-based density functional theory (DFT) computations frequently used studying halide perovskites, leading high-throughput data sets, screening promising materials, training machine learning (ML) models accelerated prediction optimization. In article, we present an overview computational data-driven discovery novel using some examples from literature believe best represent success field. Specific methods optimization across large chemical spaces, automated design compositions, highlighted. DFT-ML-based frameworks have been instrumental expanding pool stable with desired optoelectronic properties will continue inform new close synergy targeted experiments.