Investigating magnetic van der Waals materials using data-driven approaches

Abstract

In this work, we investigate magnetic monolayers of the form AⁱAⁱⁱB₄X₈ based on the well-known intrinsic topological magnetic van der Waals (vdW) material MnBi₂Te₄ (MBT) using first-principles calculations and machine learning techniques. We select an initial subset of structures to calculate the thermodynamic properties, electronic properties, such as the band gap, and magnetic properties, such as the magnetic moment and magnetic order using density functional theory (DFT). Data analytics approaches are used to gain insight into the microscopic origin of materials’ properties. The dependence of materials’ properties on chemical composition is also explored. For example, we find that the formation energy and magnetic moment depend largely on A and B sites whereas the band gap depends on all three sites. Finally, we employ machine learning tools to accelerate the search for novel vdW magnets in the MBT family with optimized properties. This study creates avenues for rapidly predicting novel materials with desirable properties that could enable applications in spintronics, optoelectronics, and quantum computing.