Temperature dependence of the mean magnon collision time in a spin Seebeck device

TL;DR

This paper models how magnon collision times vary with temperature in a spin Seebeck device, revealing two dominant scattering processes and predicting their effects on thin film performance.

Contribution

It introduces a temperature-dependent model for magnon collision times based on experimental data, distinguishing defect and magnon-magnon scattering mechanisms.

Findings

Low temperature scattering follows a T^{-1/2} dependence.

High temperature scattering follows a T^{-4} dependence.

Predicted thickness dependence of the spin Seebeck coefficient.

Abstract

Based on the relaxation time approximation, the mean collision time for magnon scattering $\tau_c(T)$ is computed from the experimental spin Seebeck coefficient of a bulk YIG / Pt bilayer. The scattering results to be composed by two processes: the low temperature one, with a $T^{-1/2}$ dependence, is attributed to the scattering by defects and provides a mean free path around 10 $\mu$m; the high temperature one, depending on $T^{-4}$, is associated to the scattering by other magnons. The results are employed to predict the thickness dependence of the spin Seebeck coefficient for thin films.