Origin of the reduced exchange bias in epitaxial FeNi(111)/CoO(111) bilayer

TL;DR

This study uses advanced scattering techniques to reveal that the reduced exchange bias in epitaxial FeNi/CoO bilayers is due to anisotropic antiferromagnetic domains and their irreversible reconfiguration, affecting the magnetic properties.

Contribution

It provides detailed insights into the magnetic interface and domain structure, clarifying the origin of reduced exchange bias in epitaxial bilayers with unprecedented precision.

Findings

Antiferromagnetic domain size is approximately 30 nm.

Irreversible domain reconfiguration causes training effects.

Correlation found between domain size, exchange bias, and frozen-in spins.

Abstract

We have employed Soft and Hard X-ray Resonant Magnetic Scattering and Polarised Neutron Diffraction to study the magnetic interface and the bulk antiferromagnetic domain state of the archetypal epitaxial Ni$_{81}$Fe$_{19}$(111)/CoO(111) exchange biased bilayer. The combination of these scattering tools provides unprecedented detailed insights into the still incomplete understanding of some key manifestations of the exchange bias effect. We show that the several orders of magnitude difference between the expected and measured value of exchange bias field is caused by an almost anisotropic in-plane orientation of antiferromagnetic domains. Irreversible changes of their configuration lead to a training effect. This is directly seen as a change in the magnetic half order Bragg peaks after magnetization reversal. A 30 nm size of antiferromagnetic domains is extracted from the width the (1/2 1/2 1/2) antiferromagnetic magnetic peak measured both by neutron and x-ray scattering. A reduced blocking temperature as compared to the measured antiferromagnetic ordering temperature clearly corresponds to the blocking of antiferromagnetic domains. Moreover, an excellent correlation between the size of the antiferromagnetic domains, exchange bias field and frozen-in spin ratio is found, providing a comprehensive understanding of the origin of exchange bias in epitaxial systems.