Stochastic theory of spin-transfer oscillator linewidths

TL;DR

This paper develops a stochastic theoretical framework to analyze the linewidths of magnetization oscillations in spin-valve structures, linking thermal noise effects to spectral linewidths and validating with experimental data.

Contribution

It introduces a nonlinear oscillator model with Langevin equations to describe amplitude and phase fluctuations, providing a new approach to understanding linewidths in spin-transfer oscillators.

Findings

Linewidths are inversely proportional to spin-wave intensities.

A lower bound on linewidths is set by frequency modulations.

Quantitative agreement with experimental results is achieved.

Abstract

We present a stochastic theory of linewidths for magnetization oscillations in spin-valve structures driven by spin-polarized currents. Starting from a nonlinear oscillator model derived from spin-wave theory, we derive Langevin equations for amplitude and phase fluctuations due to the presence of thermal noise. We find that the spectral linewidths are inversely proportional to the spin-wave intensities with a lower bound that is determined purely by modulations in the oscillation frequencies. Reasonable quantitative agreement with recent experimental results from spin-valve nanopillars is demonstrated.