Switchable Spin-Current Source Controlled by Magnetic Domain Walls

Letter pubs.acs.org/NanoLett

Switchable Spin-Current Source Controlled by Magnetic Domain Walls W. Savero Torres,†,‡ P. Laczkowski,†,‡ V. D. Nguyen,†,‡ J. C. Rojas Sanchez,†,‡ L. Vila,*,†,‡ A. Marty,†,‡ M. Jamet,†,‡ and J. P. Attané*,†,‡ †

INAC, CEA Grenoble, 17 avenue des Martyrs, 38054, Grenoble, France Université Grenoble Alpes, F-38000, Grenoble, France

Nano Lett. 2014.14:4016-4022. Downloaded from pubs.acs.org by IOWA STATE UNIV on 10/01/18. For personal use only.

‡

ABSTRACT: Using nonlocal spin injection, spin−orbit coupling, or spincaloritronic effects, the manipulation of pure spin currents in nanostructures underlies the development of new spintronic devices. Here, we demonstrate the possibility to create switchable pure spin current sources, controlled by magnetic domain walls. When the domain wall is located at a given point of the magnetic circuit, a pure spin current is injected into a nonmagnetic wire. Using the reciprocal measurement configuration, we demonstrate that the proposed device can also be used as a pure spin current detector. Thanks to its simple geometry, this device can be easily implemented in spintronics applications; in particular, a single current source can be used both to induce the domain wall motion and to generate the spin signal. KEYWORDS: Spintronics, magnetic domain wall, spin current, magnetic logic ure spin currents consist of the opposite flow of spin up and spin down electrons with no overall charge current. The manipulation of these pure spin currents in nanostructures generates an increasing interest, as it is notably the basis of two emerging research fields: the spincaloritronics,1 where spin currents are generated by thermal gradients, and the spin− orbitronics,2 where spin currents are manipulated using spin− orbit effects as the spin Hall or the Rashba-Edelstein effects. They can also be created in lateral devices, for example, by nonlocal injection in lateral spin-valves (LSV),3,4 which are a basic tool for studying spin transport properties in nonmagnetic materials. In addition, the absorption of pure spin currents by a ferromagnetic element is associated with a spin transfer that can excite spin-waves,5 induce magnetic oscillations,6 or even lead to magnetic switching.7,8 The manipulation of such currents is the basis of devices with enhanced characteristics, as higher working speed and lower power consumption, and could allow developing a purely spin-based information processing technology.9,10 A relatively unrelated but thriving topic is the manipulation of magnetic domain walls (DWs) in nanostructures. This research field has been the subject of intense debate over the last 15 years, particularly on fundamental questions related to current-induced DW motion.11,12 Moreover, it underlies a number of emerging spintronic technologies for memory and

P

Figure 1. (a) Schematic representation of the LSV, made of two ferromagnetic electrodes bridged by a transversal nonmagnetic channel. A nucleation pad and a constriction (50 nm long and 30 nm wide) are patterned on the right side electrode in order to introduce and to pin a DW. (b) Scanning electron microscopy (SEM) images of the device. A reservoir is inserted at the bottom of the right side electrode to inject the DW. (c) SEM image of the central part of the device. © 2014 American Chemical Society

Received: April 19, 2014 Revised: May 22, 2014 Published: May 29, 2014 4016

dx.doi.org/10.1021/nl501453p | Nano Lett. 2014, 14, 4016−4022

Nano Letters

Letter

Figure 2. (a) Nonlocal measurement in a reference sample, possessing two symmetrical 50 nm wide ferromagnetic electrodes without constrictions. The two sharp spin signal variations correspond to the magnetic switching of the electrodes. (b) Two probe measurement scheme on the right side electrode and corresponding experimental result obtained at room temperature. The sharp resistance variation corresponds to the insertion and to the depinning of a DW. (c) Nonlocal measurement in an LSV device with constriction. An intermediate state appears, corresponding to the DW pinning. (d) GMR measurement, also exhibiting the intermediate pinned state.

logic applications13,14 in which the information storage and manipulation is linked to the existence of a DW or to its micromagnetic configuration.15 Here we show that DWs in a ferromagnetic wire can be used as injectors or detectors of pure spin currents. This allows the creation of switchable spin current sources/detectors controlled by the presence of a DW at a given position in the circuit. In usual spin valves devices, the giant magnetoresistance (GMR) tunes the resistance of the stack accordingly to the relative orientation of the two ferromagnetic elements. Our lateral spin valves (LSVs) are the lateral counterpart of these devices and consist of two NiFe electrodes bridged by a Cu spin channel (cf. Figure 1). For the purpose of this study, a nucleation pad and a constriction have been patterned in the right-side electrode so that a DW can be introduced in the wire at low field, before getting pinned underneath the nonmagnetic Cu spin channel. In all the measurements presented here, the applied field is parallel to the electrodes. Our devices were prepared by conventional e-beam lithography, e-gun deposition, and lift-off process23 on a thermally oxidized Si substrate. They consist on two NiFe stripes, a narrow one (50 nm) and a wider one (100 nm) with a constriction patterned in the middle. Both stripes are connected to a transversal 100 nm width Cu stripe (see Figure

1a). Prior to the Cu evaporation, an Ar ion etching was carriedout, resulting in high-quality transparent interfaces with low resistances (200 mΩ at room temperature). The transport measurements were conducted using standard lock-in techniques with current densities around 1011 A/m2 or less to avoid heating effects. Devices without constrictions, used as reference, were fabricated on the same sample with the width of both the electrodes and channel being 50 nm. Figure 2a shows the nonlocal spin signal on such a device at room temperature. In this standard nonlocal measurement,4 an electrical current is injected from one of the electrodes, the injector, into the channel. The current is then driven away to the left part of the device. As the current is spin polarized in the ferromagnetic lead, a spin accumulation is created at the interface between the injector and the channel, inducing the appearance of a pure spin current that diffuses along the channel. The voltage measured between the detector and the channel is then proportional to the spin accumulation at the vicinity of the detector/channel interface. The two sharp nonlocal resistance variations correspond to the magnetic switching of the electrodes from parallel to antiparallel states and vice versa. We will now focus on the devices possessing a constriction. In such devices, the DW pinning can be observed in the 4017


Nano Letters

Letter

Figure 3. (a) Nonlocal measurement configuration used to inject a pure spin current from a DW and corresponding spin signal (300 K). For magnetic field values corresponding to the presence of the DW on the constriction, a spin current is injected in the channel and detected by the voltmeter. (b) Reciprocal measurement. A spin accumulation is created in the channel by the flow of an electrical current through the NiFe/Cu interface. The voltage change occurs when a DW is pinned. (c) Use of the proposed injection method in the case of the device of Figure 2a, which possess no constriction on which a DW could get pinned. Although the usual nonlocal configuration exhibits a clear spin signal, here the signal is flat; in this measurement configuration, a pinned DW on the left side electrode is required to inject a pure spin current.

presented in Figure 3a. In this experiment at room temperature, the current is applied along the right-side electrode. The leftside electrode is only used to detect the spin accumulation far from the injection zone. If one assumes that the charge current flows only in the NiFe wire, there should not be any spin current created in the structure. However, a clear spin signal variation of 1 mΩ appears for magnetic fields corresponding to the pinning of the DW on the constriction. This shows that when a DW is pinned on the constriction, a pure spin current is injected in the channel. The spin signal due to the injection by a DW is found to be comparable with those obtained using the classical nonlocal injection in which the spin signal amplitude was of 1.9 mΩ. This implies that this injection process possesses a very high efficiency, comparable to those of the direct injection of a current through a ferromagnetic− nonmagnetic interface. The pinned DW can also be used to detect a spin accumulation using the reciprocal measurement configuration. Figure 3b shows the experimental spin signal obtained using the DW as a spin accumulation detector. In this case, the spin current is generated by a current crossing the left side NiFe/Cu interface. As expected in nonlocal measurements, the permutation of voltage and current contacts between Figure 3 panels a and b lead to similar spin signals. To demonstrate that a pinned DW is required to inject a spin current, we show in Figure 3c results obtained using an injector without pinning site. In this case, no spin signal is observed, showing that there is no injected spin current. Reciprocally, in the detector configuration a pinned DW is required to detect the spin current.

magnetoresistance (MR) curve of the right-side electrode (cf. Figure 2b); as expected for a 100 nm wide wire, it exhibits a loop mostly due to the magnon magnetoresistance16 with an additional anisotropic magnetoresistance (AMR) contribution due to the presence of a DW. The drop of MR corresponds to the pinning of a DW in the constriction and the sudden increase to the depinning. Both these events occur at fields lower than the switching field of the left side electrode. Figure 2c shows a nonlocal measurement similar to that in Figure 2a but made in a device possessing a constriction. Three magnetic states can be identified: in addition to the antiparallel and parallel configurations of the electrodes magnetizations, an intermediate state appears during the switching of the right-side electrode, corresponding to the pinning of the DW on the constriction. This intermediate state can also be observed in the GMR measurement of Figure 2d. Its effect on the spin signal can be qualitatively understood by considering that when a DW is pinned, the magnetic state of the detector is only partially antiparallel to those of the injector, so with a reduced change of resistance. These results show that nonlocal measurements constitute a DW detection method, free from Oersted fields and Joules effect. Indeed, they do not require the flow of a charge current through the DW, contrarily to usual transport techniques (anisotropic, giant, tunnel, DW and magnon magnetoresistances, extraordinary Hall effect, and so forth). As demonstrated by Ilgaz et al.,17 this measurement configuration can also be used to induce a spin-transfer torque on the DW by spin current absorption. We now focus on the key point of this study, which is the use of a DW to inject or detect spin currents. The main results are 4018


Nano Letters

Letter

Figure 4. (a) Principle of the spin injection by a DW. Blue spheres represent electrons, and the associated black arrows represent their spins. The green arrows indicate the directions of diffusion of the electrons. The gray arrows represent the magnetization within the ferromagnet. The spinpolarized current flowing in the right-side ferromagnetic electrode (red dotted line) is deflected in the channel. When a DW is pinned, majority spins are injected in the channel, and a spin accumulation is created at the vicinity of the constriction. Consequently, a pure spin current is injected in the channel; in the nonferromagnetic material, electrons of opposite spins diffuse in opposite direction. In our experiment, this pure spin current is detected nonlocally by the left-side electrode. (b) A 3D view of the geometry used to compute our simulations. (c) Simulation of the current lines deviation at the vicinity of the constriction. The current (I = 1 A) is applied along the ferromagnetic wire possessing the constriction. The view is a vertical cross section along the wire, and the arrow lengths represent the current density on a logarithmic scale. (d) Simulation of the spin accumulation (electrochemical potential splitting for I = 1 A) generated in the nanostructure when both ferromagnetic wires are in the parallel state and (e) in the case in which the DW is pinned on the constriction.

The pure spin current injection by a DW is based on the following: although the current is supposed to flow along the ferromagnetic wire, at the vicinity of the constriction the current lines are locally deflected into the Cu (cf. Figure 4a), generating a pure spin current in the channel when a DW is pinned. Three-dimensional (3D) finite elements method (FEM) simulations have been performed that reproduce our experimental observations. The FEM simulations were carried out within the framework of a 2 spin-current drift diffusion model18 with collinear magnetization of the electrodes along the easy axis of the ferromagnetic wires. The spin up (J+) and spin down (J−) currents are proportional to the gradients of the electrochemical potentials (ECP) μ+ and μ−, respectively. In ferromagnetic materials, the conductivities σ asymmetry is given by the polarization p J+⃗ = −σ

J−⃗ = −σ

1−p ∇⃗μ− 2

Because we are mainly interested in comparing simulations with measurements, the ECP are expressed in voltage unit rather than in energy units. Charge conservation and spin flip processes with the spin diffusion length lsf lead to the following divergence equations div J+⃗ = −div J−⃗ =

σ(1 − p2 ) (μ+ − μ−) 4 lsf2

The ferromagnet-normal metal interfaces are assumed to be transparent with the ECP being continuous along the entire device. The boundary conditions are of Neumann type and account for the current imposed to the device at the contact interfaces. The simulations were done using the free softwares GMSH19 for geometry construction, mesh generation and postprocessing, and with the associated solver GETDP.20 As seen in Figure 4c, the deflected current can be important as the resistivity of Cu is much lower than those of NiFe (at

1+p ∇⃗μ+ 2 4019


Nano Letters

Letter

Figure 5. (a) Top: SEM image of a simple nanostructure where both the spin current injection and its detection are done using DWs. Inset: measurement configuration and orientation of the magnetization when two DWs are pinned. Bottom: corresponding magnetoresistance curve, showing a clear spin signal when (and only when) the two DWs are pinned simultaneously. (b) Similar device and measurement as in (a) but with nucleation pads on opposite sides. In this case, the domain walls are of opposite types (head-to-head and tail-to-tail) and thus the spin signal variation is negative. Note that AMR measurements show that the small plateau seen at positive fields in the measurement of panel a is probably a residual spin accumulation, present when only one DW is pinned.

Ω.nm, and p = 0.3. lsf(Cu) and p have been obtained by studying the gap dependence of the spin signal amplitude in lateral spin-valves, and the resistivities were obtained using four probe resistance measurements on the nanostructures. Still, using experimentally measured spin transport parameters in the FEM simulations, we obtain spin signal amplitudes close to the experimental value (0.6 mΩ instead of 1 mΩ). Larger spin injections should be obtained by reducing the DW width (for example, by using materials with large anisotropies, that is, with out-of-plane magnetization). Indeed, the spin signal dependence with the DW width relies fundamentally on the same physical processes as the enhancement of the nonadiabatic spin-torque contribution at low widths.12,21,22 For thin DWs, the magnetization rotates abruptly and hence accumulated spins are either up or down along a single quantization axis. For larger DWs, noncollinearities should restrain the spin accumulation and thus the spin signal amplitude. Furthermore, the efficiency of the injection process might be improved by tuning the resistivities of the materials, or by perfecting the device geometry. Also, we expect the spin signal amplitude to increase at lower sizes, as in LSVs it typically scales with 1/w3 with w being the wire width.23 Figure 5a shows a device using both proposed injection and detection principles. A short Cu wire is used as a spin channel, and there are two NiFe wires with patterned constrictions. In the measurement configuration of Figure 5a, the left side electrode is used as an injector, and the right side electrode as a detector. As seen previously, if there is no pinned DW in the injector, the spin accumulation created into the channel is negligible. As a consequence, the voltage measured on the detector is equal to zero. Accordingly, if there is no DW at the

300 K ρcu = 35 Ω.nm and ρPy = 220 Ω.nm) and as the Cu wire is thicker than the NiFe wire and because of the geometry of the constriction. When the right side electrode is magnetically saturated, the magnetizations are parallel in both sides of the constriction. As seen in Figure 4d, a spin accumulation is created in the zone where the spin-polarized current enters the Cu, and a spin accumulation of the opposite sign is created where the current leaves the Cu. Seen from the detector, these spin accumulations cancel each other, so that there is no spin current injected in the Cu channel. In other words, the spin polarized current that flows in the Cu wire can easily reenter the ferromagnetic wire. However, when a DW is pinned on the constriction, the magnetizations are antiparallel on the two sides of the constriction. As seen in Figure 4d, the spin accumulations created on both sides of the constriction have the same sign and reinforce each other, so that a spin current is then injected in the Cu channel. The reciprocal effect, that is, the detection of a spin accumulation by a DW (Figure 3b) can be understood by considering the split of the electrochemical potential of spins up and down at the vicinity of the constriction. The pinning of the domain wall leads to a voltage difference between both sides of the constriction, since the two sides of the ferromagnetic wire align their Fermi levels with different electrochemical potentials. The simulation uses the simplifying assumption of an infinitely narrow DW; in a real DW with finite width, the magnetization is not always collinear to the wires axis and can act as a spin sink. The parameters used for the simulations are lsf(Cu) = 290 nm, lsf(NiFe) = 4 nm, ρCu = 35 Ω.nm, ρNiFe = 220 4020


Nano Letters

Letter

A.M., who proposed the original experiment. Finally, W.S.T., L.V., A.M., and J.P.A. took care of the data analysis and wrote the manuscript.

detector, the detected voltage is always zero whether or not a spin current is injected from the left electrode. Thus, a nonzero voltage appears at the detector only when two DWs are pinned simultaneously, one in the left-side electrode to inject a pure spin current, and another DW in the right-side detector to detect it. Now, if we consider the presences of the DWs on the constrictions as inputs, and the spin signal as output, this simple low impedance device thus acts as an AND gate. Note that the left and right parts of the channel have been removed, as there is no need to contact the nonmagnetic channel. Depending on the material used, it can even enhance the spin signal by avoiding part of the spin relaxation.24 Additionally, in Figure 5b we show that contrary to classical transport measurement techniques the proposed detection method is sensible to the DW type (head-to-head and tail-totail). Using a similar device as in Figure 5a but with nucleation pads on opposite sides, a tail-to-tail is pinned in the detector leading this time to a negative spin signal variation. These results show that DWs can be used both as switchable injector and detector of spin currents for nonlocal measurements or eventually to create logical or memory applications combining lateral geometries and DW motion. This could benefit from the fact that a spin current created by a pinned DW could also be used to induce the depinning of another DW17 or to induce the switching of a magnetic dot,7 which allows envisioning reprogrammable spin-logic circuits. Also, for racetrack memory-like devices, this provides a way to probe the presence of a DW with a simple Cu wire, using the same current source to induce the DW motion and to generate the spin signal. The development of DW-based devices requires artificial pinning defects to stabilize the DW between two motion events. Numerous experiments have shown that the micromagnetic structure of the pinned DW (type of DW, chirality, position in the constriction, and so forth) determines the depinning field and the threshold current,25−28 and that this parameter could thus be controlled along the magnetic circuit using current pulses29 or nearby magnetic elements.15,30 It should be thus interesting to apply the proposed detection technique for the investigation of the pinned DW structure and the associated spin signal and spin torque. In conclusion, we have shown that a DW in a nanowire is an efficient injector/detector of pure spin currents in nonmagnetic wires with a high sensibility to the DW configuration. For spintronics applications like DW shift registers or threeterminals DW-MRAM, this technique allows using a single current source to induce both the domain wall motion (writing operation) and the spin signal generation (reading operation). Beyond possible applications for DW-based logic and memory devices, the proposed method can be used to obtain an efficient spin injection in various materials and systems. In particular, DWs could be combined with spin Hall effect wires or Rashba interfaces in order to convert the injected spin currents into charge currents and vice versa.

■

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge L. Notin, C. Beigné, C. Vergnaud, A. Brenac, and M. Cubuckcu for assistance during the experiments. Also, financial support from the French Agence National de la Recherche ANR-10-BLANC-SOspin and from the Institut Universitaire de France are acknowledged.

■

REFERENCES

(1) Bauer, G. E. W.; Saitoh, E. B.; Van Wees, J. Spin Caloritronics. Nat. Mater. 2012, 11, 391−399. (2) Jungwirth, T.; Wunderlich, J.; Olejník, K. Spin hall effect devices. Nat. Mater. 2012, 11, 382−390. (3) Johnson, M.; Silsbee, R. H. Interfacial charge-spin coupling: Injection and detection. Phys. Rev. Lett. 1985, 55, 1790−1793. (4) Jedema, F. J.; Filip, A. T.; Van Wees, B. J. Electrical spin injection and accumulation at room temperature in an all-metal mesoscopic spin valve. Nature 2001, 410, 345−348. (5) Kajiwara, Y.; Harii, K.; Takahashi, S.; Ohe, J.; Uchida, K.; Mizuguchi, M.; Umezawa, H.; Kawai, H.; Ando, K.; Takanashi, K.; Maekawa, S.; Saitoh, E. Transmission of electrical signals by spin-wave interconversion in a magnetic insulator. Nature 2010, 464, 262. (6) Liu, L.; Moriyama, T.; Ralph, D. C.; Buhrman, R. A. Spin-Torque Ferromagnetic Resonance Induced by the Spin Hall Effect. Phys. Rev. Lett. 2011, 106, 036601. (7) Yang, T.; Kimura, T.; Otani, Y. Giant spin-accumulation signal and pure spin-current-induced reversible magnetization switching. Nat. Phys. 2008, 4, 851−854. (8) Liu, L.; Pai, C.; Li, Y.; Tseng, H. W.; Ralph, D. C.; Buhrman, R. A. Spin-Torque Switching with the Giant Spin Hall Effect of Tantalum. Science 2012, 336, 555. (9) Behin-Aein, B.; Datta, D.; Salahuddin, S.; Datta, S. Proposal for an all-spin logic device with built-in memory. Nat. Nano. 2010, 5, 266. (10) Sinova, J.; Ž utić, I. New moves of the spintronics tango. Nat. Mater. 2012, 11, 368−371. (11) Thiaville, A.; Nakatani, Y.; Miltat, J.; Susuki, Y. Micromagnetic understanding of current-driven domain wall motion in patterned nanowires. Europhys. Lett. 2005, 69, 990. (12) Boulle, O.; Malinowski, G.; Kläui, M. Current-induced domain wall motion in nanoscale elements. Mater. Sci. Eng., R 2011, 72, 159. (13) Parkin, S. S. P.; Hayashi, M.; Thomas, L. Magnetic Domain-Wall Racetrack Memory. Science 2008, 320, 190. (14) Allwood, D. A.; Xiong, G.; Faulkner, C. C.; Atkinson, D.; Petit, D.; Cowburn, R. P. Magnetic Domain-Wall Logic. Science 2005, 309, 1688. (15) Lewis, E. R.; Petit, D.; Jausovec, A.-V.; O’Brien, L.; Read, D. E.; Zeng, H. T.; Cowburn, R. P. Measuring Domain Wall Fidelity Lengths Using a Chirality Filter. Phys. Rev. Lett. 2009, 102, 057209. (16) Nguyen, V. D.; Vila, L.; Laczkowski, P.; Marty, A.; Faivre, T.; Attané, J. P. Detection of Domain-Wall Position and Magnetization Reversal in Nanostructures Using the Magnon Contribution to the Resistivity. Phys. Rev. Lett. 2011, 107, 136605. (17) Ilgaz, D.; Nievendick, J.; Heyne, L.; Backes, D.; Rhensius, J.; Moore, T. A.; Niño, M. Á .; Locatelli, A.; Menteş, T. O.; Schmidsfeld, A. v.; Bieren, A. v.; Krzyk, S.; Heyderman, L. J.; Kläui, M. Domain-wall depinning assisted by pure spin currents. Phys. Rev. Lett. 2010, 105, 076601. (18) Kim, T. S.; Lee, B. C.; Lee, H. W. Effect of ferromagnetic contacts on spin accumulation in an all-metallic lateral spin-valve system: Semiclassical spin drift-diffusion equations. Phys. Rev. B 2008, 78 (21), 214427.

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Author Contributions

J.P.A. and L.V. planned and managed the project. W.S.T. performed the experiments with the help of P.L., V.D., J.C.R.S., L.V., M.J., and J.P.A. Simulations were made by W.S.T. and by 4021


Nano Letters

Letter

(19) Geuzaine, C.; Remacle, J.-F. Gmsh: A 3-D finite element mesh generator with built-in pre- and post-processing facilities, J. Numer. Methods Eng. 2009, 79, 1309. (20) Geuzaine, C. GetDP: a general finite-element solver for the de Rham complex. Proc. Appl. Math. Mech. 2007, 7, 1010603. (21) Tatara, G.; Kohno, H.; Shibata, J.; Lemaho, Y.; Lee, K.-J. Spin Torque and Force due to Current for General Spin Textures. J. Phys. Soc. Jpn. 2007, 76, 054707. (22) Lee, K. J.; Stiles, M. D.; Lee, H. W.; Moon, J. H.; Kim, K. W.; Lee, S. W. Self-consistent calculation of spin transport and magnetization dynamics. Phys. Rep. 2013, 531, 89. (23) Laczkowski, P.; et al. Spin Signal in Metallic Lateral Spin Valves Made by a Multiple Angle Evaporation Technique. Appl. Phys. Express 2011, 4, 063007. (24) Laczkowski, P.; Vila, L.; Nguyen, V. D.; Marty, A.; Attané, J.-P.; Jaffrès, H.; George, J.-M.; Fert, A. Enhancement of the spin signal in permalloy/gold multiterminal nanodevices by lateral confinement. Phys. Rev. B 2012, 85, 220404(R). (25) Petit, D.; Jausovec, A.; Read, D.; Cowburn, R. P. Domain wall pinning and potential landscapes created by constrictions and protrusions in ferromagnetic nanowires. J. Appl. Phys. 2008, 103, 114307. (26) He, K.; Smith, D. J.; McCartney, M. R. Observation of asymmetrical pinning of domain walls in notched Permalloy nanowires using electron holography. Appl. Phys. Lett. 2009, 95, 182507. (27) Hayashi, M.; Thomas, L.; Rettner, C.; Moriya, R.; Jiang, X.; Parkin, S. S. P. Dependence of Current and Field Driven Depinning of Domain Walls on Their Structure and Chirality in Permalloy Nanowires. Phys. Rev. Lett. 2006, 97, 207205. (28) Akerman, J.; Muñoz, M.; Maicas, M.; Prieto, J. L. Stochastic nature of the domain wall depinning in permalloy magnetic nanowires. Phys. Rev. B 2010, 82, 064426. (29) Kläui, M.; et al. Direct Observation of Domain-Wall Configurations Transformed by Spin Currents. Phys. Rev. Lett. 2005, 95, 026601. (30) Bogart, L. K.; Atkinson, D.; O’Shea, K.; McGrouther, D.; McVitie, S. Dependence of domain wall pinning potential landscapes on domain wall chirality and pinning site geometry in planar nanowires. Phys. Rev. B 2009, 79, 054414.

4022


Switchable Spin-Current Source Controlled by Magnetic Domain Walls

Recommend Documents