Oxide electronics

Abstract:

We present the discovery and refinement by neutron powder diffraction of a new magnetic phase in the Na1-xCaxMn7O12 quadruple perovskite phase diagram, which is the incommensurate analogue of the well-known pseudo-CE phase of the simple perovskite manganites. We demonstrate that incommensurate magnetic order arises in quadruple perovskites due to the exchange interactions between A and B sites. Furthermore, by constructing a simple mean field Heisenberg exchange model that generically describes both simple and quadruple perovskite systems, we show that this new magnetic phase unifies a picture of the interplay between charge, magnetic and orbital ordering across a wide range of compounds.

Abstract:

Whirling magnetic textures can have topological properties, enhancing their stability over and above that derived from energetic considerations. Such structures have been proposed as data carriers in next-generation post-Moore computing. Whilst abundantly observed in ferromagnets, their antiferromagnetic counterparts are more elusive. Interest in antiferromagnetic topological textures for device applications is growing, due to their predicted ultra-fast, deflection-free dynamics whilst being robust against external fields. In this thesis, I develop processes for imaging, nucleating and controlling topological textures in antiferromagnets, targeted towards their integration in next-generation racetrack-based oxide electronics. The prototypical canted antiferromagnet α-Fe2O3 is used throughout as an interesting test case, due to the family of topological textures present at room temperature that can be repeatedly nucleated via a Kibble-Zurek-like quench.

I developed analytical and micromagnetic models for topological textures in A-type antiferromagnets, focusing on the scaling of textures with relevant material parameters, allowing us to push towards the ultra-small sizes relevant for device applications. This was also used to predict the existence of the long sought-after topological antiferromagnetic skyrmions. I investigated freestanding crystalline α-Fe2O3 nanomembranes, a novel form of matter developed by my collaborators. One key conclusion of these experiments was that defects strongly affect the first-order Morin transition, whilst maintaining the Kibble-Zurek phenomenology observed in thin films attached to substrates. Magnetic fields cause domain repopulation in this canted AFM, but topological textures were observed to be stable in the presence of moderate field perturbations. Finally, freestanding crystal membranes can host relatively large strains compared to attached thin films or bulk crystals, which have similar lateral dimensions but the latter are drastically thicker. This was used to produce an athermal route to nucleate topological textures and tune domain populations, opening novel pathways for exploring Kibble-Zurek phenomenology in crystal membranes, as well as providing an interesting route towards device applications.

Abstract:

Vortices are among the simplest topological structures, and occur whenever a flow field `whirls' around a one-dimensional core. They are ubiquitous to many branches of physics, from fluid dynamics to superconductivity and superfluidity, and are even predicted by some unified theories of particle interactions, where they might explain some of the largest-scale structures seen in today's Universe. In the crystalline state, vortex formation is rare, since it is generally hampered by long-range interactions: in ferroic materials (ferromagnetic and ferroelectric), vortices are only observed when the effects of the dipole-dipole interaction is modified by confinement at the nanoscale, or when the parameter associated with the vorticity does not couple directly with strain. Here, we present the discovery of a novel form of vortices in antiferromagnetic (AFM) hematite ($\alpha$-Fe$_2$O$_3$) epitaxial films, in which the primary whirling parameter is the staggered magnetisation. Remarkably, ferromagnetic (FM) topological objects with the same vorticity and winding number of the $\alpha$-Fe$_2$O$_3$ vortices are imprinted onto an ultra-thin Co ferromagnetic over-layer by interfacial exchange. Our data suggest that the ferromagnetic vortices may be merons (half-skyrmions, carrying an out-of-plane core magnetisation), and indicate that the vortex/meron pairs can be manipulated by the application of an in-plane magnetic field, H$_{\parallel}$, giving rise to large-scale vortex-antivortex annihilation.