Current induced localized domain wall oscillators in NiFe/Cu/NiFe submicron wires
L. J. Chang, Pang Lin, and S. F. Lee
Citation: Applied Physics Letters 101, 242404 (2012); doi: 10.1063/1.4770306 View online: http://dx.doi.org/10.1063/1.4770306
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/101/24?ver=pdfcov Published by the AIP Publishing
Articles you may be interested in
Magnetic domain wall propagation in a submicron spin-valve stripe: Influence of the pinned layer Appl. Phys. Lett. 92, 032508 (2008); 10.1063/1.2835201
Reversible and irreversible current induced domain wall motion in CoFeB based spin valves stripes Appl. Phys. Lett. 90, 232505 (2007); 10.1063/1.2746952
Detecting domain-wall trapping and motion at a constriction in narrow ferromagnetic wires using perpendicular-current giant magnetoresistance
Appl. Phys. Lett. 85, 1562 (2004); 10.1063/1.1787154
Switching a spin valve back and forth by current-induced domain wall motion Appl. Phys. Lett. 83, 509 (2003); 10.1063/1.1594841
Switching the magnetic configuration of a spin valve by current-induced domain wall motion J. Appl. Phys. 92, 4825 (2002); 10.1063/1.1507820
Current induced localized domain wall oscillators in NiFe/Cu/NiFe
submicron wires
L. J. Chang,1Pang Lin,1and S. F. Lee2,a)
1
Department of Materials Science & Engineering, National Chiao Tung University, Hsinchu 300, Taiwan 2
Institute of Physics, Academia Sinica, Taipei 115, Taiwan
(Received 26 June 2012; accepted 26 November 2012; published online 10 December 2012) We experimentally demonstrate domain wall (DW) oscillators excited by in-plane ac current through permalloy based pseudo-spin valve wires, which contain one pair of artificial protrusions. By measuring the spin-transfer-torque induced resonance of a pinned antiparallel transverse DW, under transverse external fields, we show that the antiparallel transverse DW oscillates with a resonance frequency as high as 2.92 GHz, depending on the widths of protrusions. For DW oscillations induced by injection of dc currents, the observed peaks indV/dI associated with the reversible change of magnetoresistance are attributed to the reversible motions of the DW.VC 2012
American Institute of Physics. [http://dx.doi.org/10.1063/1.4770306]
Dynamics of magnetic domain walls (DW) in ferromag-netic nanowires has undergone intensive theoretical study in view of both fundamental research and the potential for tech-nological applications. Several DW based devices, including spintronic logic1 and magnetic memory devices,2,3 have been proposed. In such devices, the DW positions are defined by local pinning sites. These traps consist of struc-tures as artificial notches4–6 or protrusions7–9along a wire. Though notches are usually more effective traps, wires with protrusions are easier to scale down from a fabrication point of view. The DW depinning is preferably achieved by the spin transfer torque exerted by spin polarized currents instead of long-range magnetic fields. In point contact10–12 or spin valve nanopillar geometry,13,14the current is perpen-dicular to the film plane. The spin transfer torque can com-pensate the magnetic damping and causes the magnetization to precess at GHz frequencies. Recently, theoretical works on oscillators based on localized steady state DWs driven by spin polarized dc current have been investigated.15–19 These schemes are very desirable for applications in radio fre-quency (rf) sources for telecommunications and rf-assisted writing of magnetic bits in recording media.
Numerical simulation studies of the DW oscillator in magnetic nanowires show that when the magnetic field is above the Walker field3 and below the critical depinning threshold, with the properly chosen magnitude and direction of the in-plane dc current, the velocity of the DW can be tuned to zero. The spatially localized steady state DW oscil-lator has been generated at finite regions.16In soft ferromag-netic nanowires, He and Zhang15proposed that the spatially varying damping parameter, via gradient doping of rare earth impurities in the wire, can be employed to control the ampli-tude and frequency of localized DW oscillations upon the application of a magnetic field and a dc current. Ono and Nakatani16 suggested that a localized DW oscillation could be induced by injecting dc currents through permalloy wires, and the oscillation could then be converted to a microwave signal by a magnetic tunnel junction. Recent discovery in
Ref. 17 reported that DW oscillations could be maintained by the injection of a dc current through a geometrically con-strained wire with perpendicular magnetic anisotropy. More-over, the authors predicted a DW oscillator could work under a low dc current, which excites gigahertz angular pre-cession of a DW at a fixed position with a magnetic anisot-ropy step created by ion irradiation.18 Martinez et al.19 analyzed an oscillator based on pinned DWs excited by dc current in a magnetic nanowire of high perpendicular magne-tocrystalline anisotropy with the constriction of a square-shaped notch. In accordance with the results of theoretical reports, the resonance frequency was also obtained in the gigahertz level for the pinned DW as a function of ac current frequency.20–22 However, a report of an experimental result is missing. The fabrication of such devices from prediction of theory is particularly hard due to the needed narrow wire width, leading to cumbersome experimental schemes.
In this paper, we study, experimentally, the localized steady state DW oscillator driven by spin-polarized ac current. The DW was created and pinned at the artificial symmetric protrusion in a NiFe/Cu/NiFe spin valve submicron wire by application of a transverse magnetic field and was then returned to the remnant state. A transverse field was needed in the dc current case for observing the DW oscillation signature. Analysis was made and compared with micro-magnetic simu-lations to understand the underlying mechanism of the reso-nance frequency and the observeddV/dI peaks.
Submicron wires were fabricated by electron beam lithography and lift-off process on Si substrates. Our sample wires consist of a spin valve structure, NiFe(24 nm)/ Cu(12 nm)/NiFe(12 nm), and their widths are fixed at 400 nm. The length of the wires is 30 lm with a DW trap of one artificial symmetric pair of protrusions halfway along the wires. Fig. 1(a) shows a scanning electron microscopy image of a sample and the schematic measurement configu-ration. The submicron wire has variable single trap widths w¼ 200, 150, or 100 nm, and fixed 50 nm high protrusions on either side of the wire. The ground-signal-ground (GSG) coplanar waveguide with Au/Cu electrode leads having a characteristic impedance of 50 X were then fabricated by a
a)
Electronic mail: leesf@phys.sinica.edu.tw.
second lift-off process for resistance measurements. The in-plane external magnetic fieldHtwas applied transverse to
the wire to nucleate the DW in the trap. The oscillation of the DW trapped around the protrusion was excited by inject-ing an in-plane radiofrequency current Iac along the
nano-wires, and the resulting dc voltage Vdc is synchronously
obtained through a bias tee via the electrical contact probe. This DW oscillation produced a time dependent resistance DR from the giant magnetoresistance (MR) effect.23If the ac current is passed across the DW at resonance frequency fr,
the resistance reaches a maximum due to the coherent DW oscillation.24,25
We compared our measurement results with the micro-magnetic simulation, in which a unit cell of 5 5 3 nm3 and default values of material parameters for Ni80Fe20were
used, which are MS¼ 8.0 10 5
A/m, exchange stiffness con-stant A¼ 1.3 1011J/m, damping parameter a¼ 0.01, and current polarization P¼ 0.4. The spin polarized current simu-lations were carried out by numerically solving the Landau-Lifshitz-Gilbert equation with a spin momentum term.26
In Fig. 1(b), we present the experimental MR results when the transverse fieldHt was applied to the sample with
w¼ 150 nm. The curve has two stable resistance states. The low and high resistance states represent parallel and antiparal-lel magnetization configurations of the thin and thick layers between the voltage leads, respectively. The magnetization is saturated at 1000 Oe. Slow increase in the resistance occurs
from 600 Oe to 0 Oe due to the contribution of the anisotropic magnetoresistance (AMR) when the field is reduced. A large increase at 190 Oe and a plateau of resistance is seen between 200 Oe and 460 Oe. From 470 Oe, the resist-ance slowly falls to the saturation magnetization state. To understand the details of the reversal process of the magnet-ization for the MR curve, we carried out a micromagnetic sim-ulation using the OOMMF code. Fig. 1(c) shows the simulated magnetization configurations of the thin and thick NiFe layers. Starting from saturation at 1000 Oe, both the thin and thick layers are parallel to the field. When the external field is reversed to a negative value, we observed an anti-parallel transverse domain wall (TDW) between the thin and thick NiFe layers stabilized by the external field, dipole inter-action between the two layers, and the geometric constriction from 200 Oe to 370 Oe. The DWs are pinned at the left side of the trap in this simulation. Experimentally, the position of the DW depends on the details of the sample structures. The inset of Fig. 1(b) shows a minor MR hysteresis loop (þ1000 Oe to 300 Oe). It demonstrates that the anti-parallel TDW between the thin and thick NiFe layers can be stabilized in a field range from 300 Oe to þ50 Oe. At 380 Oe to 390 Oe, the thick layer shows a vortex domain state. Since the stray field from the thick layer vanishes after the nuclea-tion of the vortex DW, the magnetizanuclea-tion of the thin layer is no longer affected by the dipolar interaction from the thick layer. As a result, the width of the TDW in the thin layer becomes wider. Upon changing the field from 400 Oe to 1000 Oe, both the thin and the thick layers show a transition to negative saturation.
Fig.2(a) shows the resistances as a function of the fre-quency of ac excitation current for the samples with w¼ 200, 150, and 100 nm at zero applied field. The field sequence of creating the pinned TDWs, as the minor loop in the inset of Fig. 1(b), consists of saturating samples with Ht¼ 1500 Oe,
reversing the field toHt¼ 300 Oe for the anti-parallel TDW
nucleation, after which the field is set to zero. The magnetiza-tion configuramagnetiza-tion corresponds to the state with an anti-parallel pinned TDW at the trap as shown in Fig.1(c) with Ht¼ 300 Oe. We sweep the frequency of the ac current
from 0.5 to 5 GHz with a fixed amplitude of Iac¼ 3 mA and
measureVdcforIdc¼ 20 lA as a function of frequency.
Back-ground signals, measured when saturation magnetic fields are applied as in Fig.2(b), i.e., without any domain walls, are sub-tracted in Fig.2(a). A significant resonance peak is observed in each case, which corresponds to the coherent DW oscilla-tions in the pinning potential. The resonance frequency increases as the width of the protrusion decreases. The reso-nance peak appears at 1.1 GHz with a bandwidth of 165 MHz for the sample of w¼ 200 nm. For the case of w ¼ 150 nm, a frequency peak at fr¼ 1.83 GHz is observed, and the
band-width is 90 MHz. The narrow protrusion w¼ 100 nm has a resonance peak at fr¼ 2.92 GHz and the bandwidth is
110 MHz. The resonance frequencies of the DW oscillation under the influence of spin-polarized current in the NiFe nanowire are around several GHz. Such data are in good qual-itative agreement with theoretical calculations described by the Landau-Lifshitz-Gilbert equation with spin-transfer-tor-que contribution.16,20,21,27These data show that the resonance frequencies can be controlled by modulating the width of the
FIG. 1. (a) Schematic diagram of the measurement circuit and a scanning electron microscopy image of a single trap with a 150 nm width, the height of the protrusions is 50 nm on either side of the wire. (b) Experimental mag-netoresistance curves of the sample in (a) with an applied transverse fieldHt.
Inset in (b): Minor MR hysteresis loop. The black arrows show the direction of the field sweep. (c) Simulated magnetic configurations of the thin and thick NiFe layers in the sample with w¼ 150 nm at various fields.
242404-2 Chang, Lin, and Lee Appl. Phys. Lett. 101, 242404 (2012)
protrusion in a fixed current density. We suggest that the strong restoring force acting on the DW for the narrow protru-sion, leading to higher frequency of resonance, is due to the curvature of pinning potential.28
Fig.3(a)shows the dependence on width of the protru-sion of the resonance peak by simulation in hollow triangles and experimental data in solid circles. In this simulation, the ac current in the form of Iac(t)¼ Iacsin2pft was applied
to excite DW oscillation in the nanowire, where f is the
frequency of the ac current,Iac¼ 3 mA, and t is the time. The
resonance frequency is determined from the maximum am-plitude of time evolutions of the DW motion29 as shown in the inset of Fig.3(a). The solid circles in Fig.3(a) obtained by measurement are in qualitative agreement with simulation data. The discrepancy could be due to defects or edge rough-ness in the sample.30,31To quantitatively identify the pinning potential profile dependence on the width of the DW trap, micromagnetic simulations (OOMMF) have been used to plot the curve of the potential landscape of our samples. As in the previous study,7the potential landscape can be sepa-rated into three regimes for the nanowire with a symmetric protrusion, including the side barriersH1, side wellsH2, and
center barrier Hc. The absolute values of H1, H2, and Hc
increase when the protrusion becomes narrower, as shown in Figs.3(b)–3(d). We obtained the critical fields with different parameters of the trap from the minor loop technique.7,32–34 The curvature of the pinning potential can be controlled by changing the width of the protrusion, thereby changing the DW resonance frequency. The higher resonance frequency for the narrow trap is due to the steeper potential landscape, which enhances the restoring force on the pinned DW.24,35
We have also performed experiments for the DW reso-nance frequency as functions of ac current density and the transverse external field. We found the resonance peak is quite stable against small variation of these two parameters. The anti-parallel TDW between the two ferromagnetic layers forms a stable flux closure state to reduce energy because of magnetostatic coupling, while keeping the structure of the pinned DW unchanged. The potential landscape is hardly modulated from the applied current up to our measuring cur-rent and magnetic field up to 350 Oe. As a result, the eigen-frequencies for each of the samples were observed.
Since resonant states can be achieved by ac current, we also test whether dc currents can excite trapped DW oscilla-tors. The differential resistancedV/dI with a fixed transverse magnetic fieldHtand in-plane dc current of the sample with
w¼ 150 nm is shown for selected values of Htin Fig.4(a).
All curves display reversible peak structure at different dc current. Similar behavior is also obtained in samples with w¼ 200 and 100 nm. As shown in the inset of Fig.4(a), our data show a step of resistance, which is about one tenth of the total MR ratio in Fig.1(b), and the corresponding peak in dV/dI at Ht¼ 210 Oe. Because this step is reversible with
applied current, it is not due to the spin transfer torque induced DW motion propagating through the whole wire. It has been taken as evidence of current induced coherent DW oscillations between the pinned regimes.36We plot the criti-cal currentIccorresponding to the peak indV/dI at different
absolute values of the external field Ht from 200 Oe to
460 Oe as shown in Fig.4(b). The error inIc, mainly due to
uncertainties in geometry, is about 10%. From these results, we defined the boundary for excitation of the DW oscillation by dc current. The critical current varied symmetrically, which is expected from the DW trap with a symmetric poten-tial landscape.7
The map of dV/dI versus transverse field and dc current as well as the phase diagram are shown in Figs. 4(b) and 4(c), respectively. These data indicate there are three regions for different behaviors ofIcversusHt. For region A FIG. 2. (a) Experimental measurement of the ac current induces resonance
excitation of pinned DW trapped at the protrusion. Resistance change as a function of ac excitation current frequency for the submicron wires contain-ing artificial symmetric protrusions with three different widths of protrusion w¼ 200, 150, and 100 nm. (b) The response curve measured at the saturation field with a uniform state of submicron wires (without DW). The DR is observed unchanged with frequency for each of the samples.
FIG. 3. (a). Resonance frequency of pinned DW dependence on the width of trap w, the solid circles and the open triangles indicate the experiment and simulation results, respectively. The inset shows the simulated time evolu-tions of the DW motion with w¼ 150 nm. (b)-(d) Potential landscape of pinned DW from micromagnetic simulation with three different width of protrusion w¼ 200, 150, and 100 nm.
(anti-parallel TDW), the external fieldHtis from 200 Oe to
370 Oe. The Ic slightly increases with increasing Ht. We
obtain a linear variation with a small slope. In this situation, there is only one pinned DW oscillator at the trap in the soft NiFe layer responsible for the behavior. The transverse component of DW magnetization and Ht are in the same
direction. The DW is slightly wider whenHt is parallel to
the wall magnetization. From the theoretical calculation,37 the value of critical currentIcwhen the DW starts to
oscil-late is proportional to the DW width. Our data are consist-ent with theoretical predictions. Region B is very narrow, 380 Oe ⬉Ht ⬉ 390 Oe, where Ic sharply increases in the
original current direction. As shown in Fig. 4(a) for Ht¼ 380 Oe, we find another peak appears at the opposite
current direction. Simulation results show the nucleation of a vortex DW (VDW) in the hard layer, which is pinned at the right side of the trap for this current and field.
We suggest this new peak is due to a VDW oscillator in the hard NiFe layer induced by dc current at the large trans-verse field, and the sign of Ic is determined by the VDW
pinned location. In region C (parallel TDW) for 400 Oe⬉Ht
⬉ 460 Oe, both peaks sharply decrease with increasing Ht.
The peaks of the soft layer are at the opposite current direc-tion. From the results of the simulation, we find the TDW of the soft layer is twisted to the right side of the trap. In this situation, the DW oscillation was necessarily excited by dc current in the opposite direction. Unfortunately, the signal to noise ratio is too small to be resolved on frequency, and time domain measurements for the dc current induces localized DW oscillations. More efforts are needed to obtain direct observations of the dc current induced dynamic of DW oscillators.
We have demonstrated that the DW oscillators can be resonantly excited at the GHz frequency level by in-plane ac current at remnance after an applied transverse field. A strong pinning potential via artificial symmetric protrusions is used to localize the DW, and the oscillation of the wall is simply induced by the injection of in-plane ac current. The well-defined eigenfrequency is dependent on the width of the protrusion. The DW oscillator excited by the injection of in-plane dc current, proposed theoretically in these samples has been studied. We found the boundary for exciting DW oscillation by dc current with a transverse field.
The financial support of the Academia Sinica and the National Science Council of Taiwan, Republic of China are gratefully acknowledged.
1
D. A. Allwood, G. Xiong, C. C. Faulkner, D. Atkinson, D. Petit, and R. P. Cowburn,Science309, 1688 (2005).
2M. Hayashi, L. Thomas, R. Moriya, C. Rettner, and S. S. P. Parkin,
Science320, 209 (2008).
3
S. S. P. Parkin, M. Hayashi, and L. Thomas,Science320, 190 (2008).
4M. Hayashi, L. Thomas, C. Rettner, R. Moriya, X. Jiang, and S. S. P.
Parkin,Phys. Rev. Lett.97, 207205 (2006).
5
S. Lepadatu, A. Vanhaverbeke, D. Atkinson, R. Allenspach, and C. H. Marrows,Phys. Rev. Lett.102, 127203 (2009).
6J. Akerman, M. Mu~noz, M. Maicas, and J. L. Prieto,Phys. Rev. B82,
064426 (2010).
7
H. T. Zeng, D. Read, D. Petit, A. V. Jausovec, L. O’Brien, E. R. Lewis, and R. P. Cowburn,Appl. Phys. Lett.94, 103113 (2009).
8E. R. Lewis, D. Petit, L. O’Brien, A.-V. Jausovec, H. T. Zeng, D. E. Read,
and R. P. Cowburn,Appl. Phys. Lett.98, 042502 (2011).
9
L. J. Chang, Y. D. Yao, P. Lin, and S. F. Lee,IEEE Trans. Magn.47, 2519 (2011).
10M. Tsoi, A. G. M. Jansen, J. Bass, W. C. Chiang, M. Seck, V. Tsoi, and
P. Wyder,Phys. Rev. Lett.80, 4281 (1998).
11
M. Tsoi, A. G. M. Jansen, J. Bass, W. C. Chiang, V. Tsoi, and P. Wyder,
Nature (London)406, 46 (2000).
12Y. Ji, C. L. Chien, and M. D. Stiles, Phys. Rev. Lett. 90, 106601
(2003).
13
S. Urazhdin, N. O. Birge, W. P. Pratt, Jr., and J. Bass,Phys. Rev. Lett.91, 146803 (2003).
14S. I. Kiselev, J. C. Sankey, I. N. Krivorotov, N. C. Emley, R. J. Schoelkopf,
R. A. Buhrman, and D. C. Ralph,Nature(London)425, 380 (2003).
15
J. He and S. Zhang,Appl. Phys. Lett.90, 142508 (2007).
16
T. Ono and Y. Nakatani,Appl. Phys. Express1, 061301 (2008).
17A. Bisig, L. Heyne, O. Boulle, and M. Kl€aui,Appl. Phys. Lett.95, 162504
(2009).
18
J. H. Franken, R. Lavrijsen, J. T. Kohlhepp, H. J. M. Swagten, and B. Koopmans,Appl. Phys. Lett.98, 102512 (2011).
19E. Martinez, L. Torres, and L. Lopez-Diaz, Phys. Rev. B 83, 174444
(2011).
20
E. Martinez, L. Lopez-Dias, O. Alejos, and L. Torres,Phys. Rev. B77, 144417 (2008).
21J. Yoon, C. Y. You, Y. Jo, S. Y. Park, and M. H. Jung,Appl. Phys. Express
4, 063006 (2011).
22
J. Yoon, C. Y. You, Y. Jo, S. Y. Park, and M. H. Jung,J. Korean Phys. Soc.57, 1594 (2010).
23J. C. Sankey, P. M. Braganca, A. G. F. Garcia, I. N. Krivorotov, R. A.
Buhrman, and D. C. Ralph,Phys. Rev. Lett.96, 227601 (2006).
24
S. Lepadatu, O. Wessely, A. Vanhaverbeke, R. Allenspach, A. Potenza, H. Marchetto, T. R. Charlton, S. Langridge, S. S. Dhesi, and C. H. Marrows,
Phys. Rev. B81, 060402 (2010).
25C. T. Boone, J. A. Katine, M. Carey, J. R. Childress, X. Cheng, and I. N.
Krivorotov,Phys. Rev. Lett.104, 097203 (2010).
FIG. 4. (a) Differential resistance vs. current density at different external transverse fields Ht,
enlarged in the inset for V/I vs. j at Ht¼ 210 Oe.
(b) Map of dV/dI versus transverse field and dc cur-rent. (c) Critical currentIcvs.Ht.
242404-4 Chang, Lin, and Lee Appl. Phys. Lett. 101, 242404 (2012)
26J. Xiao, A. Zangwill, and M. D. Stiles, Phys. Rev. B
70, 172405 (2004).
27
D. V. Berkov, C. T. Boone, and I. N. Krivorotov,Phys. Rev. B83, 054420 (2011).
28E. Saitoh, H. Miyajima, T. Yamaoka, and G. Tatara, Nature
432, 203 (2004).
29
S. Kasai, Y. Nakatani, K. Kobayashi, H. Kohno, and T. Ono,Phys. Rev. Lett.97, 107204 (2006).
30
Y. Nakatani, A. Thiaville, and J. Miltat,Nature Mater.2, 521 (2003).
31
T. W. Chiang, L. J. Chang, C. Yu, S. Y. Huang, D. C. Chen, Y. D. Yao, and S. F. Lee,Appl. Phys. Lett.97, 022109 (2010).
32D. Petit, A. V. Jausovec, H. T. Zeng, E. Lewis, L. O’Brien, D. Read, and
R. P. Cowburn,Phys. Rev. B79, 214405 (2009).
33
L. K. Bogart, D. Atkinson, K. O’Shea, D. McGrouther, and S. McVitie,
Phys. Rev. B79, 054414 (2009).
34D. Petit, A. V. Jausovec, D. Read, and R. P. Cowburn,J. Appl. Phys.
103, 114307 (2008).
35
S. Lepadatu, J. S. Claydon, D. Ciudad, C. J. Kinane, S. Langridge, S. S. Dhesi, and C. H. Marrows,Appl. Phys. Lett.97, 072507 (2010).
36
S. Laribi, V. Cros, M. Munoz, J. Grollier, A. Hamzic´, and C. D. A. Fert,
Appl. Phys. Lett.90, 232505 (2007).
37
