The semiclassical theory for spin dynamics in a disordered system

TL;DR

This paper develops a semiclassical theory for spin dynamics in disordered two-dimensional systems with spin-orbit coupling, analytically calculating spin-Hall conductivities and comparing them with experimental data.

Contribution

It introduces a semiclassical framework capturing both Larmor and non-Larmor spin precession effects and provides analytical expressions for spin-Hall conductivity in disordered Rashba systems.

Findings

Non-universal intrinsic spin-Hall conductivity that vanishes with zero Rashba coupling.

Analytical calculation of spin-Hall conductivities matching experimental values.

Short-time response of the system aligns with Kubo formula results.

Abstract

We investigate the Drude model of spin dynamics in two-dimensional spin-orbit coupled systems. In the absence of an applied electric field, the spin aligns with the k-dependent effective magnetic field. The influence of disorder (the momentum relaxation time $\tau$) on the system is considered. In the presence of an electric field, the change in momentum causes a change in the effective magnetic field. The in-plane spin can precess around the successive change in the orientation of the effective magnetic field. We find that up to the linear order of the electric field, the spin orientation undergoes Larmor-like and non-Larmor like precession. Furthermore, we find that the non-Larmor motion over a very short time ($t\ll\tau$) exactly equals the result obtained from the Kubo formula. This means that the Kubo formula only captures the system's response over a very short evolution. The spin-Hall conductivity for Larmor and non-Larmor precession in the Rashba system is analytically calculated. We find that the intrinsic spin-Hall current is not a universal constant and correctly drops to zero when the Rashba spin-orbit coupling drops to zero. We also calculate the time-averaged Larmor and non-Larmor spin-Hall conductivities (SHCs) for the k-cubic Rashba system and compare them to experimental values. The time-averaged Larmor SHC vanishes and the non-Larmor SHC is given by $2.1(q/8\pi)$ which is very close to the experimental value $2.2(q/8\pi)$ by Wunderlich et al. [Phys. Rev. Lett. {\bf 94}, 047204 (2005)].