Antiferromagnetic Spintronics: Understanding the strain

点击翻译按钮获取中文摘要

Ferromagnetic materials have underpinned data storage technology for decades. This is not just for the data storage but also for 'spintronic' components such as magneto-resistive field sensors and magnetic tunnel junctions. With the desire to increase the speed of devices and pack components more densely, researchers are now turning to anti-ferromagnetic materials. Externally, these materials look non-magnetic, but on the atomic level they consist of exactly cancelling opposing magnetic moments. For a long-time they have been overlooked due to the difficulties in controlling the internal order (they do not couple strongly to magnetic fields) but recent discoveries in spin-orbit torques show that electrical manipulation and readout is possible paving the way to antiferromagnetic spintronics. Even though this is a recently created field, it is rapidly evolving, and proof of concept desktop memory devices have already been demonstrated.In ferromagnetic devices, the demagnetising field causes the shape of a device to influence the equilibrium magnetisation state and determine the domain state. The non-locality of the field means analytic solutions are only available for trivial cases. Calculating and understanding the demagnetising field of ferromagnets is one of the most important modelling problems in magnetism, spawning many competing 'micromagnetism' software packages.In antiferromagnets, there is no demagnetising field because there is no net magnetisation, yet antiferromagnets also contain complex magnetic domains. It is believed that magnetostriction and magnetoelastic effects may be the driving force. These too are non-local and shape dependent. Recent experiments in antiferromagnetic spintronics have shown the domain state and the motion of domains are important in the interpretation of electrical signals, manipulation of domains and the propagation of spin currents.We propose to solve the coupled magnetostriction, magnetoelastic, antiferromagnetic order problem to understand domain formation, shape and motion. This will allow us to provide clear interpretations of experimental results, for example why antiferromagnet domain formation and behaviour is radically different depending on the crystallographic cut of a thin film. We also anticipate modelling the dynamics when spin-orbit torques are applied, to understand domain nucleation, destruction, expansion and migration. In the long term it would also be interesting to understand how dynamical stresses (such as applied by piezo actuators) could be used to manipulate the properties of antiferromagnets