Topological magnetic materials of the (MnSb$_2$Te$_4$)$\cdot$(Sb$_2$Te$_3$)$_n$ van der Waals compounds family

TL;DR

This study uses density functional theory to predict a family of van der Waals compounds with tunable topological magnetic phases, including antiferromagnetic topological insulators and Weyl semimetals, with potential spintronics applications.

Contribution

It introduces the (MnSb$_2$Te$_4$)$\cdot$(Sb$_2$Te$_3$)$_n$ family as topologically nontrivial magnetic materials with novel phases and field-driven topological transitions.

Findings

MnSb$_2$Te$_4$ is a 3D antiferromagnetic topological insulator.

For $n \geq 3$, partial magnetic disorder coexists with nontrivial topology.

Materials can be driven into ferromagnetic Weyl semimetal or axion insulator states by magnetic field.

Abstract

Combining robust magnetism, strong spin-orbit coupling and unique thickness-dependent properties of van der Waals crystals could enable new spintronics applications. Here, using density functional theory, we propose the (MnSb$_2$Te$_4$)$\cdot$(Sb$_2$Te$_3$)$_n$ family of stoichiometric van der Waals compounds that harbour multiple topologically-nontrivial magnetic phases. In the groundstate, the first three members of the family, i.e. MnSb$_2$Te$_4$, ($n=0$), MnSb$_4$Te$_7$, ($n=1$), and MnSb$_6$Te$_{10}$, ($n=2$), are 3D antiferromagnetic topological insulators (AFMTIs), while for $n \geq 3$ a special phase is formed, in which a nontrivial topological order coexists with a partial magnetic disorder in the system of the decoupled 2D ferromagnets, whose magnetizations point randomly along the third direction. Furthermore, due to a weak interlayer exchange coupling, these materials can be field-driven into the FM Weyl semimetal ($n=0$) or FM axion insulator states ($n \geq 1$). Finally, in two dimensions we reveal these systems to show intrinsic quantum anomalous Hall and AFM axion insulator states, as well as quantum Hall state, achieved under external magnetic field, but without Landau levels. Our results provide a solid computational proof that MnSb$_2$Te$_4$, is not topologically trivial as was previously believed that opens possibilities of realization of a wealth of topologically-nontrivial states in the (MnSb$_2$Te$_4$)$\cdot$(Sb$_2$Te$_3$)$_n$ family.