How a ferromagnet drives an antiferromagnet in exchange biased CoO/Fe(110) bilayers

M Ślęzak,K Matlak,M Zając,A Kozioł-Rachwał,B Matlak,J Korecki,T Ślęzak,P Dróżdż

doi:10.1038/s41598-018-37110-8

Abstract

Antiferromagnet/ferromagnet (AFM/FM) bilayers that display the exchange bias (EB) effect have been subjected to intensive material research, being the key elements of novel spintronics systems. In a commonly accepted picture, the antiferromagnet, considered as a rigid material due to its high anisotropy and magnetic hardness, controls the magnetic properties of the ferromagnet, such as a shift of the hysteresis loop or coercivity. We show that this AFM-FM master-slave hierarchy is not generally valid and that the influence of the ferromagnet on the magnetic anisotropy (MA) of the neighbouring antiferromagnet must be considered. Our computer simulation and experimental studies of EB in an epitaxial CoO/Fe(110) bilayer show that the ferromagnetic layer with strong uniaxial magnetic anisotropy determines the interfacial spin orientations of the neighbouring AFM layer and rotates its easy axis. This effect has a strong feedback on the EB effect experienced by the FM layer. Our results show new physics behind the EB effect, providing a route for grafting a desired anisotropy onto the AFM and for precise tailoring of EB in AFM/FM systems.

Highlights

The ability to modify and detect the spin structure of AFM materials is difficult but interesting from both a fundamental and an application point of view
Another promising approach is to drive the orientation of AFM spins and tune the exchange bias (EB) magnitude with epitaxial strain14; this requires the use of specific substrates or buffer layers, which in turn limit the applications in real devices
We prove that the exchange interaction-driven influence of the ferromagnetic layer on the antiferromagnetic spin structure can be more complex than that suggested by the literature data6,15 and relies on a modification of AFM magnetic anisotropy

Summary

Introduction

The ability to modify and detect the spin structure of AFM materials is difficult but interesting from both a fundamental and an application point of view. For the realization of an AFM THz oscillator, an AFM with a well-defined easy axis is strongly desired, whereas in the case of SOT devices, the main technical difficulty, namely, the need for an external magnetic field, can be overcome in AFM/FM bilayers due to the EB3–5 effect From this point of view, the possibility of simultaneously controlling the AFM easy axis and tuning the magnitude of the EB effect in AFM/FM systems is very important. The main drawback of AFM magnetic moments rotating with the external magnetic field is partial or even complete suppression of EB, which is demanded to be large and spontaneous13 Another promising approach is to drive the orientation of AFM spins and tune the EB magnitude with epitaxial strain; this requires the use of specific substrates or buffer layers, which in turn limit the applications in real devices. To trace the evolution of the interfacial AFM easy axis and the EB effect strength with the change of the effective in-plane MA of the FM layer, the magnetic hysteresis loops were measured and carefully analysed as a function of the Fe thickness

Methods

Results

Conclusion