Injection of orbital angular momentum into transition metals from first-principles

TL;DR

This paper uses first-principles quantum calculations to study how orbital angular momentum injected into transition metals decays, revealing that orbital currents decay much faster than spin currents and are partially converted into spin currents due to spin-orbit coupling.

Contribution

The study demonstrates that orbitally-polarized currents decay within a few atomic layers, challenging previous interpretations of experimental decay lengths and providing new insights into the orbital Hall effect.

Findings

Orbital currents decay within a few atomic layers.

Injected orbital current is partially converted into spin current with spin-orbit coupling.

Contradicts previous experimental interpretations of exponential decay on longer length scales.

Abstract

We use quantum mechanical scattering calculations implemented in a basis of tight-binding muffin-tin orbitals to calculate nonequilibrium spin and orbital currents in transition metals with a view to understanding the length scale on which they decay. In the case of spin currents, the relaxation length, called the spin-flip diffusion length, is reasonably well understood. We apply our experience with spin currents to study orbitally-polarized currents and find that they behave qualitatively differently. Upon injection from a lead, orbital currents decay within a few atomic layers contradicting the current interpretation of experimental results which appear to show exponential decay on the length scale of the spin-flip diffusion length and longer. When spin-orbit coupling is included, the injected orbital current is partially converted into a spin current within a few atomic layers. This insight provides a new perspective on the physics of the orbital Hall effect.