Substitutional Alloying Using Crystal Graph Neural Networks

The use of machine learning (ML) Michalski et al (2013); LeCun et al (2015) methods in material science to accelerate materials discoveryCurtarolo et al (2013) is at the base of the so-called material informatics (MI) Ramakrishna et al (2019); Ramprasad et al (2017); Takahashi and Tanaka (2016); L. Ward (2017); Rajan (2005). By training ML models on large databases, such as OQMD or the Materials Project high-throughput electronic structure calculation databases Saal et al (2013); Jain et al (2013); Curtarolo (2012); Hachmann et al (2011); NOMAD (https://nomad-coe.eu), the goal is to achieve predictions of material properties with quantum accuracy. As in statistical mechanics with the need for identifying appropriate order parameters of novel phases and structures, the key challenge in ML algorithms is to identify effective system descriptors that can function as structure identifiers. A large variety of descriptors have been proposed, including fixed-length feature vectors of material elemental or electronic properties Seko et al (2015); Xue et al (2016); Isayev et al (2017), as well as structural descriptors, based on rotational and traslational invariant transformations of atomic coordinates, like the Coulomb matrix Rupp et al (2012), atom-centered symmetry functions (ACSFs) Behler (2011), social permutation invariant coordintes (SPRINT) Pietrucci and Andreoni (2011), smooth overlap of atomic positions (SOAP) De et al (2016) and global minimum of root mean-square distance Sadeghi et al (2013). However, these solutions are often system-specific, and are not suitable for vast compositional and structural space exploration.