Final Report Summary - CIDWM-NANOSTRIPS (Current-induced domain wall motion in magnetic nanostrips)
Project context and objectives
ately, the use and control of thermal effects in spintronic devices has attracted a lot of attention and opened up a new area of research called spin caloritronics. Since the discovery of the Spin Seebeck effect by Uchida et al. in 2008, many works have observed thermoelectric effects in magnetic metals, and even in insulating ferromagnets. This last type of material has forced consideration of other microscopic origins, such as the magnonic spin Seebeck effect and other phonon-mediated effects. This multitude of effects calls for experiments that weigh their relative magnitudes. Such identification is especially important for nanosciences, as temperature gradients that are impossible in bulk samples are easily created in nanostructures.
This project has concentrated on current-induced domain-wall motion in nanostrips of the magnetic soft alloy of Ni with Fe. According to this effect, and closely related to the celebrated giant magnetoresistance, the spin polarisation of the carriers leads to a torque on a domain-wall structure that gives rise to the motion of domain walls along the electrical current. However, the large current densities that are necessary (one ampere per square micrometre is typical) cause some heating of the sample. As a result, the nucleation of new magnetic domains under large current densities has been observed by many authors and attributed to an increase in the sample temperature above the Curie temperature. In addition, domain-wall structure transformation and/or random displacement have been associated to thermally activated Brownian motion of the wall position and magnetic moment.
Project results
The effect discovered during this project is qualitatively different. It consists of a unidirectional domain-wall displacement towards the hotter part of the nanostrip, irrespective of the current direction. By tuning the heat dissipation in the samples and modelling the heat diffusion using finite-element software, we have concluded that this unidirectional motion can only be explained by the presence of a temperature profile along the nanostrip sample. The quantitative analysis of the experiments that we have performed using micromagnetic simulations shows that, on top of the classic thermodynamic pressure on the domain wall (change of domain-wall energy with temperature), another force, probably the magnonic spin Seebeck effect, is displacing the domain walls.
In addition, we show that the nano-devices fabricated and studied in this work are well-suited to the production of very intense and localised temperature gradients, with typical values of 100 Kelvin/micrometre and durations of a nanosecond.
ately, the use and control of thermal effects in spintronic devices has attracted a lot of attention and opened up a new area of research called spin caloritronics. Since the discovery of the Spin Seebeck effect by Uchida et al. in 2008, many works have observed thermoelectric effects in magnetic metals, and even in insulating ferromagnets. This last type of material has forced consideration of other microscopic origins, such as the magnonic spin Seebeck effect and other phonon-mediated effects. This multitude of effects calls for experiments that weigh their relative magnitudes. Such identification is especially important for nanosciences, as temperature gradients that are impossible in bulk samples are easily created in nanostructures.
This project has concentrated on current-induced domain-wall motion in nanostrips of the magnetic soft alloy of Ni with Fe. According to this effect, and closely related to the celebrated giant magnetoresistance, the spin polarisation of the carriers leads to a torque on a domain-wall structure that gives rise to the motion of domain walls along the electrical current. However, the large current densities that are necessary (one ampere per square micrometre is typical) cause some heating of the sample. As a result, the nucleation of new magnetic domains under large current densities has been observed by many authors and attributed to an increase in the sample temperature above the Curie temperature. In addition, domain-wall structure transformation and/or random displacement have been associated to thermally activated Brownian motion of the wall position and magnetic moment.
Project results
The effect discovered during this project is qualitatively different. It consists of a unidirectional domain-wall displacement towards the hotter part of the nanostrip, irrespective of the current direction. By tuning the heat dissipation in the samples and modelling the heat diffusion using finite-element software, we have concluded that this unidirectional motion can only be explained by the presence of a temperature profile along the nanostrip sample. The quantitative analysis of the experiments that we have performed using micromagnetic simulations shows that, on top of the classic thermodynamic pressure on the domain wall (change of domain-wall energy with temperature), another force, probably the magnonic spin Seebeck effect, is displacing the domain walls.
In addition, we show that the nano-devices fabricated and studied in this work are well-suited to the production of very intense and localised temperature gradients, with typical values of 100 Kelvin/micrometre and durations of a nanosecond.
