Fig. 1: Left) MFM images (20 μm×20μm) of a device patterned in a FePt/Pt/FePt spinvalve. The wires are 200 nm wide. The dark areas correspond to the reversed domain, which has nucleated in the reservoir on the top left of the MFM image. Here, the DW is pinned between two Hall bars. Right) Resistance of the wire during a field loop. The plateaus correspond to the anti parallel state.
Current induced domain walls motion is one important issue emerging in the fields of spintonics with potential applications. Larger effects are expected for narrow walls which appear in materials with high anisotropy. From our know-how on thin layers with high perpendicular magnetic anisotropy we have developed devices to investigate magnetic domain walls dynamic. Our results provide useful key parameters which are debated among theoretical models.
The magnetic configuration of ferromagnets consists in domains with uniform magnetization separated by domain walls (DW). When an external field is applied, the magnetization reversal occurs through nucleation and DW motion. The DWs can be pinned on structural defects, leading to remanence and coercivity. It has been recently shown that electric currents can “push” the DW, acting on the magnetization as a torque called spin-torque. This phenomenon opens large opportunities of applications in the field of memories and magnetologic devices, however the spin-torque efficiency has to be enhanced, and fundamental question still need to be addressed. Particularly, applications require the precise control of DW depinning. In this context, if most experiments have been conducted in using permalloy layers with planar magnetization and large domain walls, it has been predicted that the force exerted on the DW should increase for narrower DWs, as a very strong magnetization gradient is present. To get enhanced effect, we use magnetic materials with high perpendicular magnetic anisotropy, where DW widths (3.6 nm) are smaller than the typical length of electronic transport such as mean free path, spin diffusion and Larmor precession lengths. Such an ultra narrow domain wall is thus an interesting playground for testing the theoretical models under construction.
The NM group acquired an in-depth mastering of the growth of high anisotropy – high magnetization thin films based of chemically ordered alloys (FePt…) by Molecular Beam Epitaxy. An effort has been devoted to the understanding of the depinning dynamic of a single DW from identified defects in the thermally activated regime (field just below coercitive field). In this regime, the depinning process under applied field and current is stochastic, the time spent on the defect being a random variable, similarly to the crossing of an energy barrier by a particle due to thermal activation.
The depinning events can be measured at an high rate using electric measurements. However to investigate the DW depinning statistics from an individual defect the determination of its position is required. In order to measure the position of the DW in the wire, we have grown spin-valves structures FePt/NM/FePt. The proportion of the wire where magnetization is reversed by the passage of the DW in the soft layer is directly related to its position within the wire. Due to the GMR effect in the spinvalve, this proportion may be deduced from the electrical resistance (Fig1). Also, we have recently shown that magnons scattering in thin layers with high perpendicular anisotropy gives rise to a magnetoresistance enabling the measurement of the magnetization in a single FePt layer. This provides another and simpler way to determine DW position from electrical resistance.
Maj : 17/10/2013 (483)