We present an innovative approach to the production of single-crystal iron oxide nanorings employing a solution-based route. Single-crystal hematite (alpha-Fe2O3) nanorings were synthesized using a double anion-assisted hydrothermal method (involving phosphate and sulfate ions), which can be divided into two stages: (1) formation of capsule-shaped alpha-Fe2O3 nanoparticles and (2) preferential dissolution along the long dimension of the elongated nanoparticles (the c axis of alpha-Fe2O3) to form nanorings. The shape of the nanorings is mainly regulated by the adsorption of phosphate ions on faces parallel to c axis of alpha-Fe2O3 during the nanocrystal growth, and the hollow structure is given by the preferential dissolution of the alpha-Fe2O3 along the c axis due to the strong coordination of the sulfate ions. By varying the ratios of phosphate and sulfate ions to ferric ions, we were able to control the size, morphology, and surface architecture to produce a variety of three-dimensional hollow nanostructures. These can then be converted to magnetite (Fe3O4) and maghemite (gamma-Fe2O3) by a reduction or reduction-oxidation process while preserving the same morphology. The structures and magnetic properties of these single-crystal alpha-Fe2O3, Fe3O4, and gamma-Fe2O3 nanorings were characterized by various analytical techniques. Employing off-axis electron holography, we observed the classical single-vortex magnetic state in the thin magnetite nanorings, while the thicker rings displayed an intriguing three-dimensional magnetic configuration. This work provides an easily scaled-up method for preparing tailor-made iron oxide nanorings that could meet the demands of a variety of applications ranging from medicine to magnetoelectronics.

Using transmission electron microscopy, we investigate the thermally activated motion of domain walls (DWs) between two positions in Permalloy (Ni80Fe20) nanowires at room temperature. We show that this purely thermal motion is well described by an Arrhenius law, allowing for a description of the DW as a quasiparticle in a one-dimensional potential landscape. By injecting small currents, the potential is modified, allowing for the determination of the nonadiabatic spin torque: βt=0.010±0.004 for a transverse DW and βv=0.073±0.026 for a vortex DW. The larger value is attributed to the higher magnetization gradients present.