Vortex domain wall depinning by polarized current in submicron half-ring wires

(1)

Vortex domain wall depinning by polarized current in submicron half-ring wires

Y. S. Chen, K. W. Cheng, C. Yu, S. F. Lee, D. C. Chen, S. H. Wu, M. T. Lin, Y. Liou, K. T. Wu, and Y. D. Yao

Citation: Journal of Applied Physics 99, 08G516 (2006); doi: 10.1063/1.2173624

View online: http://dx.doi.org/10.1063/1.2173624

View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/99/8?ver=pdfcov Published by the AIP Publishing

Articles you may be interested in

Asymmetric domain wall depinning under current in spin valves with perpendicular anisotropy Appl. Phys. Lett. 98, 232512 (2011); 10.1063/1.3592844

Current-driven domain wall in giant magnetoresistance half-ring series wires with varied linewidth J. Appl. Phys. 105, 07D115 (2009); 10.1063/1.3068134

Controlled depinning of domain walls in a ferromagnetic ring circuit Appl. Phys. Lett. 90, 242504 (2007); 10.1063/1.2748339

Magnetization reversal study in submicron half-ring patterned wires with different corner structures J. Appl. Phys. 99, 08G309 (2006); 10.1063/1.2173223

Injection of a magnetic domain wall into a submicron magnetic wire Appl. Phys. Lett. 75, 2815 (1999); 10.1063/1.125159

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP: 140.113.38.11 On: Thu, 01 May 2014 02:02:07

(2)

Vortex domain wall depinning by polarized current in submicron

half-ring wires

Y. S. Chen,a兲 K. W. Cheng, C. Yu, and S. F. Lee

Institute of Physics, Academia Sinica, Taipei 115, Taiwan

D. C. Chen

Department of Materials Science and Engineering, National Chiao Tung University, Hsinchu 300, Taiwan

S. H. Wu, M. T. Lin, and Y. Liou

K. T. Wu

Department of Physics, Fu Jen Catholic University, Taipei Hsien 242, Taiwan

Y. D. Yao

共Presented on 2 November 2005; published online 25 April 2006兲

Domain wall pinning force in the junctions共corners兲, with different shapes of square, semicircle, or triangle, of half-ring in-series wires is considered to study the current injection induced wall movements. This geometry has less thermal activation at the region of domain wall nucleation in contrast to notch structures. The wires with square corners have the largest domain pinning force to resist polarized current-induced magnetization reversal, judging from the largest slope in the current-field dependence共⌬I/⌬H=0.274兲. © 2006 American Institute of Physics.

关DOI:10.1063/1.2173624兴

I. INTRODUCTION

The influence of a high current density injected into a single magnetic film was predicted1 and proved2 in many studies on current-induced magnetization reversal. It is ex-pected to be applied to the local magnetization switching method, which is different from switching by external field in the present magnetic random access memories共MRAMs兲 and other magnetoelectronic devices. These current-induced phenomena have attracted much attention not only in inter-esting scientific research but also in applications. From the original predictions of Berger,3 the current flow exerting a torque on the magnetic moment of the ferromagnet excites spin waves or directly flips the magnetic moment because of the transfer of spin angular momentum. This means that the polarized current can also be used for current-induced do-main wall motion. The critical current density to drive the domain wall is dependent on two factors4 the domain wall pinning force and the material-dependent Gilbert damping coefficient. Permalloy共Ni80Fe20兲 is studied in our experiment

for its shape anisotropy. The effect of current injection to different domain wall pinning forces is the main purpose of our study. Recently, structures such as notches5 were em-ployed to trap domain walls. However, current density on the cross section of notches increases and raises temperature lo-cally when polarized current is injected. Although most re-searches confirmed that the current-induced magnetization reversal was by the mechanism of spin momentum transfer, the local influence of thermal activation in these kinds of structures may reduce its practical use. Therefore, we make

use of discontinuous corners of half-ring in-series wires6 to nucleate and trap domain walls in this experiment. Bound-ary conditions of different shapes 关the left side of Figs. 1共a兲–1共c兲兴 in the ends of corners were designed to

cre-a兲_{Electronic mail: [email protected]}

FIG. 1. SEM images and micromagnetic simulation configurations at zero field of共a兲 SQ, 共b兲 SC, and 共c兲 TA shape corners at the half-ring in-series wires. There is one corner producing domain wall between electrodes. The SQ sample has different domain structure from the other two shapes.共d兲 Spin configuration simulation image of the SC sample on the field near the switching field showing the domain wall nucleation at the corner region and the beginning of domain wall motion共the larger arrows indicate the direc-tions of motion兲.

JOURNAL OF APPLIED PHYSICS 99, 08G516共2006兲

0021-8979/2006/99共8兲/08G516/3/$23.00 99, 08G516-1 © 2006 American Institute of Physics [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

(3)

ate different domain wall pinning force, which resists against current-induced domain wall motion.

II. EXPERIMENT

We fabricated several sets of 0.4␮m wide Ni80Fe20

wires of two inner radius 0.5␮m half-rings in series with the corners shapes7 of square共SQ兲, semicircle 共SC兲, or triangle 共TA兲 by using electron beam lithography and lift-off tech-nique, as shown in the scanning electron microscope共SEM兲 images on the left hand side of Figs. 1共a兲–1共c兲. By using magnetron sputtering with the working pressure of 1 ⫻10−3_{Torr, the ferromagnetic layer with about 25 nm}

thick-ness was deposited onto 50 nm SiO₂ coated silicon 共100兲 substrates. Ti/ Au leads for the four-point measurement were made by similar techniques on two sides of each wire. Be-fore dc current from 10␮A to 3 mA with steps of 10␮A was applied to induce magnetization reversal by domain wall motion, the switching field 共coercive field兲 of each sample was found first from the magnetoresistance 共MR兲 loops. Then the current-driven wall motion at different external fields lower than its switching field was carried out. The experiments were performed at room temperature 共295 K兲. Different sets of samples showed the same trend of behav-iorsdiscussed in the following.

III. RESULTS AND DISCUSSIONS

Typical longitudinal MR loops of Ni80Fe20 samples

nor-malized by saturation resistance are shown in Fig. 2. These MR loops can be understood by the anisotropic MR共AMR兲 effect during magnetization reversal process. When the ex-ternal magnetic field was swept through zero from positive

共or negative兲 saturation field, 6000 Oe, the gradual drop in the MR curve represents that the transverse components of the magnetization increased slowly. The right hand side of Figs. 1共a兲–1共c兲 shows the simulation results at the remanent state. It is clearly seen that the SQ sample has the end do-main structure typically seen in a rectangular shape. How-ever, the SC and TA samples have different domain struc-tures due to the local shape anisotropy. Figure 1共d兲 shows the simulation results in the SC sample at a field of −224 Oe, lower than the simulated switching field. The magnetization reversal caused by domain walls being pushed away from the wire at the switching field results in the MR curve jumps back to the high resistance state. This switching process 共re-versible followed by irre共re-versible兲 has been observed in each sample. The SQ sample has the largest longitudinal switch-ing field H_sw= 268± 2 Oe, while H_sw= 230± 2 Oe for the SC sample and Hsw= 228± 2 Oe for the TA sample.

Figure 3 shows that there was a resistance jump at the critical current Ic= 2.32 mA for the resistance versus current

curve of sample SQ at external field Hex= −221 Oe. The Ic

data at different Hex of sample SQ is listed in Table I. The

variation of resistance共⌬R兲 caused by current injection was about 200– 230 m⍀, which was close to the value of ⌬R of the MR loop at the switching field. The agreement of ⌬R’s suggests that the changes of magnetization states caused by external field and by dc current are similar. The background resistance by Joule heating can also be clearly seen to in-crease with the applied dc current in Fig. 3. So, the differen-tial resistance共⌬R=Rn− Rn−1兲 as a function of injected

cur-rent at diffecur-rent fields is a good approach to eliminate this effect, as shown in Fig. 4. The resistance Rn represents the nth point of resistance measurement. The larger field is

ap-plied on the samples; the smaller critical current is needed to induce the magnetization reversal 共domain wall depinning兲. Through the linear fit of critical current versus external field, the samples SQ, SC, and TA have theslopes 共mA/Oe兲 of

FIG. 3. The SQ sample resistance as a function of the injecting current at external fields Hex= −221 Oe.

FIG. 2. Typical longitudinal MR loops of the wires measured at 295 K.

TABLE I. The critical current density data at different Hexof sample SQ.

External field

共Oe兲 −219 −221 −223 −225 −227

Critical current density共A/cm2_兲

2.88⫻107 _2.32_⫻107 _2.08_⫻107 _1.38_⫻107 _5.5_⫻106

FIG. 4. The differential resistance of the SQ sample as a function of the injected current at external fields Hex= −219, −221, −223, −225, and

−227 Oe.

08G516-2 Chen et al. J. Appl. Phys. 99, 08G516共2006兲

(4)

⬃0.274±0.005, ⬃0.206±0.005, and ⬃0.24±0.005, respec-tively, as shown in Fig. 5. The magnitude of ⌬I/⌬H indi-cates the ease of domain wall depinning. The larger the ⌬I/⌬H, the more polarized current is needed to drive the domain wall away from the corners, thus suggesting the larger domain wall pinning force. Obviously, the square shape corners have the largest domain wall pinning force, which resists current-induced domain wall motion. This re-sult is in agreement with the switching field of MR loops. However, for samples SC and TA, there is a deviation of trend which is different from the result of switching field in the MR loops. The shunt effect of current distribution around the corner could be responsible for this result.

IV. CONCLUSION

The structure of the designed wire including two half-ring parts with strong shape anisotropy can confine domain wall nucleation at corner 共connected part兲 easily. Through the MR loops and current-induced domain wall motion, we have proved that the domain wall pinning force can change with the designed corner shapes of the samples. Using half-ring in-series wires with different designed corners to create different domain wall pinning force offers a way to study the pinning force. Our results offer good alternatives to notch structures for trapping domain wall with different pinning force.

1_{L. Berger, J. Appl. Phys. 55, 1954}_{共1984兲; C.-Y. Hung and L. Berger, J.}

Appl. Phys. 63, 4276共1988兲; J. C. Slonczewski, J. Magn. Magn. Mater. 159, L1共1996兲; L. Berger, Phys. Rev. B 54, 9353 共1996兲.

2_{M. Tsoi, A. G. M. Jansen, J. Bass, W.-C. Chiang, M. Seck, V. Tsoi, and P.}

Wyder, Phys. Rev. Lett. 80, 4281 共1998兲; J. Z. Sun, J. Magn. Magn. Mater. 202, 157共1999兲; E. B. Myers, D. C. Ralph, J. A. Katine, R. N. Louie, and R. A. Buhrman, Science 285, 867共1999兲; F. J. Albert, J. A. Katine, R. A. Buhrman, and D. C. Ralph, Appl. Phys. Lett. 77, 3809 共2000兲.

3_{L. Berger, J. Appl. Phys. 49, 2156}_共1978兲.

4_{S. Zhang, P. M. Levy, and A. Fert, Phys. Rev. Lett. 88, 236601}_共2002兲. 5_{K. Miyake, K. Shigeto, K. Mibu, and T. Shinjo, J. Appl. Phys. 91, 3468}

共2002兲; M. Tsoi, R. E. Fontana, and S. S. P. Parkin, Appl. Phys. Lett. 83, 2617共2003兲.

6_{C. Yu, S. F. Lee, Y. D. Yao, Y. R. Ma, E. W. Huang, J. L. Tsai, T. Y. Chen,}

and C. R. Chang, J. Appl. Phys. 93, 7619共2003兲.

7_{P. Vavassori, O. Donzelli, V. Metlushko, M. Grimsditch, B. Llic, P. Neuzil,}

and R. Kumar, J. Appl. Phys. 88, 999共2000兲. FIG. 5. The critical current of three samples as a function of the external

magnetic field. The least square fits to the data and their slopes are also shown.

08G516-3 Chen et al. J. Appl. Phys. 99, 08G516共2006兲