UC San Diego

One of the most interesting topics in material science is the effect of geometric confinement on nanoscale magnetic structures. Confinement produces new phenomena not observed in bulk materials and has led to a large number of technologies. An important structure featuring a geometric confinement effect is a thin film multilayer. In particular, the exchange bias effect, which is present in some ferromagnet / antiferromagnet bilayers, is used in applications such as magnetic recording, sensing, and spintronic devices. Although exchange bias has been the focus of many experimental and theoretical investigations, the details of its mechanism are still elusive. Above all, the contributions of magnetic structures in regions away from the interface remain ambiguous. This dissertation studies the individual contributions of each region of the ferromagnetic/antiferromagnetic bilayer on the exchange bias. We consider the exchange bias effect in magnetic thin films as defined by the three components of the spin structure : i) at the erromagnetic/antiferromagnetic interface, ii) in the antiferromagnetic layer, and iii) in the ferromagnetic layer. We performed experiments to address individually these three constituents and found that the exchange bias is significantly affected by the bulk of both the antiferromagnetic and the ferromagnetic layers. We used Ni/FeF₂ bilayers as a prototypical exchange bias system. First, we explored magnetization reversal in cross shaped nano-junctions using the planar Hall Effect. We demonstrated how to find the direction of magnetization rotation and obtained the angular distribution of the unidirectional anisotropy axes. Second, we investigated the contribution from the antiferromagnetic bulk using ion bombardment to induce controllable depth dependent defects. We unambiguously showed that the exchange bias can be tuned without altering the interface or the ferromagnetic layer. Third, we investigated the ferromagnetic layer thickness dependence on the exchange bias. We found a strong deviation from the inverse proportionality law and demonstrated the importance of the spin structure in the ferromagnetic layer. An incomplete domain wall model was used to quantitatively account for the experimental results. Our studies show that the bulk of the antiferromagnetic and the ferromagnetic layers, together with the interface, are crucial ingredients in the establishment and the control of the exchange bias