Exchange bias model for Fe/FeF ₂ : Role of domains in the ferromagnet

TLDR

Exchange bias is a shift of the hysteresis loop from its symmetric position, occurring when thin ferromagnetic films are deposited on antiferromagnetic materials, and is characterized by orthogonal bulk magnetizations, comparable magnitudes for compensated and uncompensated interfaces, sign reversals, and asymmetric magnetization at coercive fields. The authors propose a model describing exchange bias for a compensated interface. The model posits that near the Néel temperature a canted spin configuration in the antiferromagnetic interface freezes into a metastable state, causing the exchange bias energy to be reversibly stored in a spring‑like magnet or incomplete domain wall within the ferromagnetic slab. Analytical results and simulations from the model agree qualitatively and quantitatively with available experimental data.

Abstract

Exchange bias (EB) is a shift of the hysteresis loop from its normal position, symmetric around H = 0, to HE ≠ 0. It occurs when thin ferromagnetic (F) films are deposited on a variety of antiferromagnetic (AF) materials. EB is also associated with several additional remarkable features: i) the bulk magnetizations of the F is orthogonal to the AF easy axis; ii) HE is of similar magnitude for compensated and uncompensated AF interface layers; iii) the sign of HE can assume both positive and negative values; and, iv) the magnetization |M(H ≪ − Hc)| ≠ |M(H ≫ + Hc)| , where Hc is the coercive field. Here we propose a model that describes the EB phenomenon for a compensated interface. Based on the experimental evidence, and extensive computer simulations, we suggest that close to the Néel temperature a canted spin configuration in the AF interface freezes into a metastable state. As a consequence, the EB energy is reversibly stored in a spring-like magnet, or incomplete domain wall (IDW), in the F slab. The results we extract from our model, both analytically and through simulations, are qualitatively and quantitatively compatible with the available experimental information.