Understanding Interfacial and Spin-Orbit Torque Effects in Thin Film Magnetic Multilayers.

Abstract

The understanding of both the fundamentals of spin-orbit torques and the potential avenues for spin-orbit exploitation is crucial for the development of future spintronic devices. Accurate quantification of the relevant parameters, such as the spin-orbit effective field strengths, is vital for both performance comparisons and the understanding of the physical origins. In this work, a common effective field measurement technique is probed for its potential inaccuracies, a new method for achieving field-free magnetic reversal is demonstrated and both the structural and magnetic property variations in structurally inverted thin films are investigated. The design, construction and commissioning of a multi-functional MOKE instrument has been presented. The multi-functional nature of the instrument allowed for sample environments which included the simultaneous application of external fields and sample charge currents for spin-orbit torque effect measurements. Raster scanned 2D MOKE data were collected for multiple samples based upon a 20 µm-wide perpendicularly magnetised Pt/Co/Pt rectangular bar. Each rectangular bar had a differing lateral displacement with respect to the 60 µm wide electrical contacts, with positions varying from the centre of the electrical contacts to being parallel with the edge. It was shown that the series of samples exhibit deterministic bi-directional field-free magnetic reversal. After discussing the previously known methodologies for field-free magnetic reversal, a new methodology was required in order to explain the symmetry of the results, where the charge current Oersted field is the cause of both the initial domain nucleation and the overall bi-directional magnetic reversal pattern. The accuracy of a common methodology for quantifying the spin-orbit effective fields was studied by probing the effect of magnetic reversal. It was shown that if magnetic reversal occurs when otherwise unwanted, the resultant effective field values can be exaggerated by over an order of magnitude compared to their typical values. This inaccuracy can be undetectable, particularly in thin films with non unity magnetic remanence. Structural and magnetic property variations in Pt/CoFeTaB/Ir, together with its structural inverse, were probed with a variety of techniques including polarised neutron reflectometry. Both structural and magnetic variations were observed, with the most dominant change being a Curie temperature variation of over 100 K, where Ir/CoFeTaB/Pt is paramagnetic at room temperature whereas Pt/CoFeTaB/Ir is ferromagnetic.