Orbital magnetization in two-dimensional materials from high-throughput computational screening

TL;DR

This study computationally screens 822 2D magnetic materials to analyze their orbital magnetization, revealing significant orbital moments in 5d compounds, the influence of crystal fields, and the impact of Hubbard corrections on magnetic anisotropy.

Contribution

It provides a comprehensive high-throughput analysis of orbital magnetization in 2D materials, highlighting the importance of including Hubbard corrections and spin-orbit coupling for accurate magnetic property predictions.

Findings

Orbital moments of 0.3-0.5 μ_B in 5d compounds.

Alignment of orbital moments follows Hund's rule with deviations explained by crystal field effects.

Hubbard corrections can induce large unquenched orbital moments and magnetic anisotropies.

Abstract

We calculate the orbital magnetization of 822 two-dimensional magnetic materials from the Computational 2D Materials Database (C2DB). For compounds containing 5$d$ elements we find orbital moments of the order of 0.3-0.5 $\mu_\mathrm{B}$, which points to the necessity of including these in any type of magnetic modeling and comparison with experiments. It is also shown that the alignment of orbital moments with respect to the spin largely follows the predictions from Hund's rule and that deviations may be explained by the $d$-band splitting originating from the crystal field - for example in the important case of CrI$_3$. Finally, we show that for certain insulators, Hubbard corrections may lead to large and fully unquenched orbital moments that are pinned to the lattice rather than the spin and that these moments can lead to enormous magnetic anisotropies. Such unquenched ground states are only found from density functional theory calculations that include both Hubbard corrections and self-consistent spin-orbit coupling and largely invalidates the use of the magnetic force theorem for calculating magnetic anisotropies.