Bachelor project: Bias-dependence

Determining the spin-torque bias dependence of the spin-transfer torques in MgO-based magnetic tunnel junctions

A spin-polarized current flowing through a ferromagnet exerts a torque on the magnetization at the nanoscale, thereby providing means of manipulating it. In a nano-size magnet, spin-transfer torques can induce either magnetization reversal or steady-state precession. These phenomena have been proposed as write method for non-volatile magnetic memory devices and operating mechanism for tuneable radio-frequency nano-oscillators, respectively. Given their good scaling perspectives, spin-torque devices have recently been identified as one of the prime candidates for beyond Moore technologies. Spin-torque devices based on MgO-based magnetic tunnel junctions show great promise for applications, although fundamental questions remain to be answered.

In particular, recent theoretical models predicted that the bias dependence of the spin-transfer torque, as well as its magnitude, depends on a variety of factors, such as the symmetry of the structure, the respective layer thicknesses involved, the magnitude of the exchange splitting, the type and distribution of interfacial defects and/or the density of states’ energy dependence. Partially conflicting experimental results have been reported, obtained by several groups based on different approaches.

The focus of this study will be to carry out, for the first time, a direct comparison between two techniques which can be utilize to determine the spin-torque bias dependence, namely spin-torque ferromagnetic resonance, and thermally excited ferromagnetic resonance. Both methods rely on high-frequency magneto-transport measurements. The results will be considered in the context of previously reported experimental results, and will provide means to validate theoretical predictions.