Machine Learning Lattice Constants for Cubic Perovskite Compounds
Yun Zhang, Xiaojie Xu
Abstract
Abstract Cubic perovskites have attracted great attention in the past decade due to unique and tunable optical, mechanical, and electrical properties, which are promising candidates for various applications such as solar cells, light emitting diodes, actuators, and laser cooling devices. The lattice constant, a, as the only variable parameter among the six parameters in the crystal structure, has a significant impact on the structural stability, bandgap structure, and thus materials performance. In this study, we develop the Gaussian process regression (GPR) model to shed light on the statistical relationship between ionic radii and lattice constants for cubic perovskite compounds. A total of 135 samples with lattice constants ranging from 3.680 Å to 6.330 Å are explored. The model has a high degree of accuracy and stability that contributes to fast and robust lattice constant estimations.