Oxide perovskites and their derivatives are attractive candidates for the diverse applications in renewable energy conversions due to their unique structural and compositional flexibility and high material stability.

Stability is of central importance in current perovskite solar cell research and applications. Goldschmidt tolerance factor (t) recently provided qualitative guidance for experimentalists to engineer stable ABX perovskite by tuning effective ionic size with mixing cations or anions and for theorists to search emerging perovskites. Through first-principles calculations, we have calculated decomposition energies of 138 perovskite compounds of potential solar cell applications. Instead of t, we have found that (μ + t), where μ and η are the octahedral factor and the atomic packing fraction, respectively, demonstrates a remarkably linear correlation with thermodynamic stability. As a stability descriptor, (μ + t) is able to predict the relative stability among any two perovskites with an accuracy of ∼90%. This trend is then used to predict decomposition energies of another 69 perovskites, and the results are in excellent agreement with first-principles calculations, indicating the generalization of the trend. This thermodynamic stability trend may help the efficient high-throughput search for emerging stable perovskites and precise control of chemical compositions for stabilizing current perovskites.

Development of lead-free inorganic perovskite material, such as Cs2AgBiBr6, is of great importance to solve the toxicity and stability issues of traditional lead halide perovskite solar cells. However, due to a wide bandgap of Cs2AgBiBr6 film, its light absorption ability is largely limited and the photoelectronic conversion efficiency is normally lower than 4.23%. In this text, by using a hydrogenation method, the bandgap of Cs2AgBiBr6 films could be tunable from 2.18 eV to 1.64 eV. At the same time, the highest photoelectric conversion efficiency of hydrogenated Cs2AgBiBr6 perovskite solar cell has been improved up to 6.37% with good environmental stability. Further investigations confirmed that the interstitial doping of atomic hydrogen in Cs2AgBiBr6 lattice could not only adjust its valence and conduction band energy levels, but also optimize the carrier mobility and carrier lifetime. All these works provide an insightful strategy to fabricate high performance lead-free inorganic perovskite solar cells.

Symbolic regression (SR) is an approach of interpretable machine learning for building mathematical formulas that best fit certain datasets. In this work, SR is used to guide the design of new oxide perovskite catalysts with improved oxygen evolution reaction (OER) activities. A simple descriptor, μ/t, where μ and t are the octahedral and tolerance factors, respectively, is identified, which accelerates the discovery of a series of new oxide perovskite catalysts with improved OER activity. We successfully synthesise five new oxide perovskites and characterise their OER activities. Remarkably, four of them, Cs 0.4 La 0.6 Mn 0.25 Co 0.75 O 3 , Cs 0.3 La 0.7 NiO 3 , SrNi 0.75 Co 0.25 O 3 , and Sr 0.25 Ba 0.75 NiO 3 , are among the oxide perovskite catalysts with the highest intrinsic activities. Our results demonstrate the potential of SR for accelerating the data-driven design and discovery of new materials with improved properties.