High-throughput search of magnetic topological insulators and semi-metals using spin-orbit spillage and machine-learning methods

Magnetic topological insulators and semi-metals have a variety of properties that make them attractive for applications including spintronics and quantum computation. However, very few such materials are known to date. In this work, we use a systematic high-throughput density functional theory based computational workflow to identify possible intrinsic Chern insulators and semi-metals among 40000 three-dimensional materials in the JARVIS-DFT database (https://jarvis.nist.gov/jarvisbdft). First, we screen materials with net magnetic moment>0.5 μB, and spin-orbit spillage > 0.5. Then we carry out Wannier charge centers, Chern number, anomalous Hall conductivity, surface bandstructure, Fermi-surface and magnetic ordering analysis to determine the exotic characteristics of the screened compounds. We also train machine learning model for predicting spillage, bandgap and magnetic moment using JARVIS-ML models to further accelerate the screening process. The computational prediction if realized experimentally will open a new paradigm in the physics of topological quantum materials.