Anisotropy of spin relaxation in metals

TL;DR

This paper introduces the concept of anisotropy in spin relaxation in non-magnetic metals, showing it can reach extremely high values due to Fermi surface features, with implications for spintronic applications.

Contribution

It reveals the anisotropy of spin relaxation linked to the Elliott-Yafet parameter and Fermi surface topology, providing a new understanding of spin dynamics in metals.

Findings

Anisotropy of spin relaxation can reach 830% in uniaxial transition metals.

Extended spin-flip hot areas on the Fermi surface cause large anisotropy.

No theoretical upper limit for the anisotropy effect.

Abstract

The concept of anisotropy of spin relaxation in non-magnetic metals with respect to the spin direction of the injected electrons relative to the crystal orientation is introduced. The effect is related to an anisotropy of the Elliott-Yafet parameter, arising from a modulation of the decomposition of the spin-orbit Hamiltonian into spin-conserving and spin-flip terms as the spin quantization axis is varied. This anisotropy, reaching gigantic values for uniaxial transition-metals (e.g. 830% for hcp Hf) as density-functional calculations show, is related to extended "spin-flip hot areas" on the Fermi surface created by the proximity of extended sheets of the surface, or "spin-flip hot loops" at the Brillouin zone boundary, and has no theoretical upper limit. Possible ways of measuring the effect as well as consequences in application are briefly outlined.