Machine learning the lattice constant of cubic pyrochlore compounds

Abstract Pyrochlores, with a general formula A 2 B 2 X 7 , are promising candidates in many potential applications due to their wide varieties of physical properties. Different combinations of cations and anions in the crystal structure enable the tailoring of their properties and functionalities. For cubic pyrochlores, the lattice constant, a, is an integrated result of stoichiometry, ionic radii, and electronegativities of alloying elements. It also has significant impacts on stabilities, electronic structures, ionic conductivities, and thus performance of materials. Here, we develop the Gaussian process regression model to shed light on the relationship among ionic radii, electronegativities, and lattice constants for cubic pyrochlores. The dataset includes ternary pyrochlores and mixed pyrochlores with co‐doping and multi‐doping situations. One hundred and thirty‐nine samples with lattice constants between 9.287 and 11.150 Å are examined. The model is highly stable and accurate that contributes to efficient and low‐cost lattice constant estimations.