Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Copper wires each one containing a pseudo-spin valve of Co/Cu/Co were produced by electrodeposition inside nanoporous polycarbonate membranes. These wires have a diameter of 40 nm and a length of 6000 nm of which only 50 nm relates to the spin valve. In order to be able to measure the magnetic behavior of a single wire, the entire deposition was carried out in a Co/Cu multibath and then electrical contact of a single wire was done with a special equipment using a pure Copper bath. Knowing that such wire has a resistance of the order of 500 Ω, with a Giant Magneto-Resistance ratio of about 0.2 %, one needs a sufficiently sensitive measuring apparatus be able to detect such variations. After having carried out a magnetoresistive characterization, we used a set-up built with a Wheatstone bridge in which the sample is integrated. The resistance variation associated with the magnetization reversal being weak, we were very concerned with the sample quality during these three years of research. As high current densities are injected, a behavior of "Two Level Fluctuation" appeared between parallel and antiparallel states. Such phenomena can be treated only by statistics. Therefore, several thousands of measurements were taken in order to be able to understand the statistical behaviors of the current induced magnetization switching. To better understand the mechanisms of the reversal, we studied its dynamics with a model of double potential well by measuring the average time of residence in parallel (respectively in antiparallel) state as a function of current and applied field for different amplitudes and signs. From our analyses, it arises clearly that the dynamics of the magnetization reversal induced by a high current density, in pseudo-spin valves, could be formulated in terms of Spin-Transfer-Torque. We show that the agreement with this model is remarkable.
Dirk Grundler, Andrea Mucchietto, Korbinian Baumgärtl