Machine Learning Accelerated Discovery of Mixed Anion Materials

Mixed-anion materials are interesting counterparts to their more widely studied single-anion compounds due to the increased flexibility and tunability of properties afforded by the presence of multiple anions. Here, we demonstrate how computational approaches, based on DFT datasets can be combined with materials informatics and machine learning (ML) models to accelerate materials discovery. We utilize and compare a variety of materials representations, including a recently proposed improved crystal graph convolutional neural network (iCGCNN) model, and the Voronoi tessellation approach incorporated in the Materials-Agnostic Platform for Informatics and Exploration (MAGPIE). ML models are trained on a set of 450,000 DFT data prior calculated in the Open Quantum Materials Database (OQMD) and evaluated on a separate test set of 3,000 mixed-anion compounds where iCGCNN outperforms random forest models in predictive accuracy by ~300%. By introducing more mixed-anion compounds into the training set, the performance of the iCGCNN model is further improved, and it allows us to make predictions of a large number of stable (and hence, likely synthesizeable) new ternary oxychalcogenides, which are subsequently validated by DFT calculations.