A universal scaling law of exchange bias training effect

TL;DR

This paper introduces a universal scaling law for the exchange bias training effect in ferromagnetic/antiferromagnetic bilayers, derived from a thermodynamic model, applicable across different magnetization reversal mechanisms.

Contribution

A generalized model based on a discretized Landau-Khalatnikov equation is proposed, unifying the description of the training effect beyond empirical fits.

Findings

The model accurately describes the training effect in various systems.

It extends the understanding of exchange bias training beyond the empirical 1/√n law.

The approach is independent of the magnetization reversal mechanism.

Abstract

Exchange bias training effect in ferromagnetic/antiferromagnetic bilayers is investigated. In some systems the evolution of the exchange bias field $H_E$ with the number of cycle $n$ cannot be fitted by the empirical $1/\sqrt{n}$ function. A unified expression is derived from a discretized Landau-Khalatnikov equation in the framework of the thermodynamics model which is proposed by Ch.\ Binek. This generalized model describes well training effect independent of the magnetization reversal mechanism in the ferromagnetic layers.