APPLIED PHYSICS LETTERS 88, 112502 共2006兲
Layer-resolved study of magnetic interaction effects in heterostructure dot arrays Y. Choi,a兲 D. R. Lee,b兲 J. W. Freeland, and G. Srajer Advanced Photon Source, Argonne National Laboratory, Argonne, Illinois 60439
V. Metlushko Department of Electrical and Computer Engineering, University of Illinois at Chicago, Chicago, Illinois 60607
共Received 26 October 2005; accepted 1 February 2006; published online 14 March 2006兲 Using polarized x rays we have studied magnetic interactions in a series of patterned single-layer 共NiFe and Co兲 and multilayer 共NiFe/ Co and NiFe/ Cu/ Co兲 heterostructures. Extraction of layer-specific magnetic hysteresis loops from an array of 1-m dots allows us to separate the influence of inter- and intralayer interactions. Double layer 共NiFe/ Co兲 dots show evidence of identical vortex formation in both layers while with the spacer layer the direct coupling between the two magnetic layers is removed, and dipolar field contribution becomes significant so that the vortex formation in both layers is suppressed. © 2006 American Institute of Physics. 关DOI: 10.1063/1.2179116兴 In recent years magnetic tunnel junctions1–6 have received much attention due to their potential as a mechanism for information storage. Magnetic tunnel junctions consist of two ferromagnetic layers separated by a thin nonmagnetic layer, and these structures are being developed for magnetoresistive devices and magnetic random access memories 共MRAMs兲. For application of magnetic materials as a MRAM element, patterning on a submicron scale7–9 is necessary to achieve a memory density comparable with conventional dynamic random access memories. For patterned magnetic materials magnetic reversal takes place via a coherent rotation of the magnetization only in cases with welldefined anisotropies. More common, however, is that the reversal occurs via the domain formation at the ends of the elongated element. For arbitrarily shaped nanoscale elements, in general, it has been impossible to reliably calculate the field at which domains first form from basic principles.10 Recently, the use of circular or disk shaped elements was proposed as the storage layer in MRAM devices, where two possible vortex states with the clockwise or counterclockwise rotation of the magnetic moment 共chirality兲 realized at remanence could be used for magnetic storage.11 It was found that those states are magnetically stable, and they can be packed in high-density arrays because the flux-closed vortex state reduces interactions between neighboring elements.12–17 In this letter, we investigate the influence of the interactions between two ferromagnetic layers in patterned arrays of 1-m dots. For comparison, patterned layers without a spacer layer and patterned single layers were also studied. A Co single-layer dot array exhibits magnetic reversal characteristics associated with a vortex state while a NiFe dot array exhibits somewhat different behavior. In NiFe/ Co multilaya兲
Present address: Materials Science Division, Argonne National Laboratory, Argonne, Illinois 60439. b兲 Author to whom correspondence should be addressed; permanent address: Pohang Accelerator Laboratory, Pohang University of Science and Technology, Pohang 790-784, South Korea; electronic mail:
[email protected]
ered patterned arrays, the field dependence is modified, and, depending on the existence of a Cu spacer layer in between, the field dependence of each layer is changed. A dot array with the two magnetic layers in direct contact shows identical vortex states in both layers due to the strong ferromagnetic coupling. In another dot array with a Cu layer in between, the vortex states become unfavorable and inhibited in the both layers. With the spacer layer, the direct coupling between the two magnetic layers is removed and strong dipolar interactions between the layers become dominant. A set of four unpatterned thin film samples were evaporated using e-beam deposition: Co 共30 nm兲, NiFe 共20 nm兲, NiFe 共20 nm兲 / Co 共30 nm兲, or NiFe 共20 nm兲 / Cu 共3 nm兲 / Co 共30 nm兲. Each film was capped with 3 nm Cu to prevent oxidation. Using e-beam lithography, 1 m circular dots in a square lattice array were made on a double-layered positive type e-beam resist followed by evaporation and ultrasonic assisted lift-off process 关see Fig. 1共a兲兴. For layer-resolved magnetic characterization, x-ray resonant magnetic scattering18 共XRMS兲 measurements were performed on the continuous film and patterned samples. XRMS takes advantage of enhanced magnetic sensitivity near resonant energies of elements, and this technique allows elementspecific 共thus layer-resolved兲 magnetic hysteresis measurements on these multilayer structures. The incident photon energy was tuned to the Co L3 共778 eV兲 resonance to probe the Co layers; to probe the permalloy 共NiFe兲 layers, the incident energy was tuned to the Fe L3 共707 eV兲 resonance. The XRMS measurements were performed at beamline 4 ID-C of the Advanced Photon Source at Argonne National Laboratory.19 The scattering geometry and the applied magnetic field direction are illustrated in Fig. 1共b兲. The applied field direction was parallel to the sample surface and in the scattering plane. Using this setup and circularly polarized x rays, the laterally averaged magnetization component parallel/antiparallel to the applied field is probed. Hysteresis loops were measured on the four different continuous films, which in all cases were square in shape with coercive fields between 8 and 16 Oe. The hysteresis loops from the two single-layered dot arrays are shown in
0003-6951/2006/88共11兲/112502/3/$23.00 88, 112502-1 © 2006 American Institute of Physics Downloaded 14 Jun 2006 to 131.193.45.131. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
112502-2
Choi et al.
Appl. Phys. Lett. 88, 112502 共2006兲
FIG. 1. 共Color online兲 共a兲 SEM image of the dot array. 共b兲 XRMS experimental setup. Circularly polarized 共CP兲 x rays were used to probe the sample magnetization M along the applied field H. At specular condition 共i = f 兲, reflected intensities 共I+,−兲 were measured as a function of angle 共i兲, incident x-ray energy 共E兲, and field 共H兲.
Fig. 2. The field dependence for these two arrays shows increased coercive and saturation field values as compared with the unpatterned films. While the Co array shows a hysteresis loop behavior typical to a vortex state, the NiFe dot array shows a square loop with enhanced coercivity. This difference can be understood by considering the energies competing to form the magnetic ground state. The primary energy driving the formation of the vortex state is due to the shape anisotropy, which favors the removal of magnetic surface charge.20 The lower saturation magnetization of NiFe reduces this energy and continues to favor the single-domain state while the higher saturation magnetization of Co results
FIG. 3. 共Color online兲 Normalized XRMS signal as a function of applied field from NiFe/ Co dot array. The film structure is shown as the inset. 共a兲 Measured hysteresis loop of the NiFe layer. 共b兲 Measured hysteresis loop of the Co layer.
in the formation of a vortex state 共or at least closer to a vortex state兲. In NiFe/ Co multilayered dot arrays, the field dependence of each layer is changed due to the coupling effect between the two layers. First, the case where NiFe and Co layers are in direct contact is studied. As shown in Fig. 3, the NiFe and Co layers were probed separately using XRMS measurements. The NiFe and Co layers are expected to be strongly coupled ferromagnetically, and the measured hysteresis loops show that the field dependence of the NiFe layer mimics that of the Co layer. In comparison with the single NiFe layer dot array 关Fig. 2共a兲兴, the NiFe layer in direct contact with a Co layer in Fig. 3共b兲 follows the vortex behavior of the Co layer. The field dependence of the Co layer is also modified from the single layer to multilayer dots. For the single Co layer dot, the nucleation field for the vortex state is lower than 200 Oe while, for the Co layer in the multilayered dots, the nucleation field is higher than 200 Oe. Additionally, the remanent magnetization is lower in the NiFe/ Co system as compared to the single-layer dots 关 Fig. 3共b兲 versus Fig. 2共b兲兴. This indicates that the balance of magnetic energies is favoring the flux closure state of the vortex. When a Cu spacer layer was inserted between the two ferromagnetic layers, the field dependence of the two layers becomes modified as shown in Fig. 4. With the removal of the strong direct ferromagnetic interaction, the field dependence of the two layers deviates from that of the typical vortex states as shown in the previous case 共Fig. 3兲. What is noticeable here is that as the applied field was switched from −H to +H, the magnetization of the Co layer remains negative near zero while that of the NiFe layer is already switched to positive. This indicates that at remanence the NiFe and Co layers are magnetized antiparallel. This behavior was also observed by Bonfim et al.21 on the squarepatterned arrays with a similar film structure. Such inverted loops result from a strong antiferromagnetic dipolar interac-
FIG. 2. 共Color online兲 Normalized XRMS signal as a function of applied field from single-layer arrays. The data points as the field was changed from −H to +H are represented as solid circles, and the points as the field was changed from +H to −H are represented as hollow circles. The film structures are shown as the insets. 共a兲 Measured hysteresis loop from the NiFe dot array. 共b兲 Measured hysteresis loop from the Co dot array. Downloaded 14 Jun 2006 to 131.193.45.131. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
112502-3
Appl. Phys. Lett. 88, 112502 共2006兲
Choi et al.
compensated by the stability introduced by the vortex formation in both of the NiFe and Co layers 共Fig. 3兲. In conclusion, we have studied interlayer dipolar field effects in multilayered dot arrays. Typically vortex states are expected in these dot shaped arrays, but due to altered magnetic interactions in multilayered elements, the formation of vortex states become less favorable. Stray interlayer dipolar fields can alter the balance of exchange and shape anisotropy making the single domain state more favorable. Work at Argonne is supported by the U.S. Department of Energy, Office of Science, under Contract No. W-31-109ENG-38. V.M. is supported by the U.S. National Science Foundation, Grant No. DMR-0210519. 1
FIG. 4. 共Color online兲 Normalized XRMS signal as a function of applied field from a NiFe/ Cu/ Co dot array. The film structure is shown as the inset. 共a兲 Measured hysteresis loop of the NiFe layer. 共b兲 Measured hysteresis loop of the Co layer.
tion between the two layers. Since the behavior is not observed in the unpatterned films, we can rule out interlayer exchange via spin-polarization in the Cu spacer. This means the antiparallel alignment is induced by the dipolar interaction resulting from shape effects. In the low fields, the dipolar interaction between the NiFe and Co layers dominates so that the antiparallel alignment between the two layers becomes energetically favorable. In comparison with the direct contact case shown in Fig. 3, the changes caused by the dipolar field interactions are more drastic in the NiFe hysteresis loop than in the Co loop since the NiFe moment is much smaller than the Co moment. In Fig. 4共a兲, the hysteresis curve of the NiFe layer crosses the M / M s = 0 at 190 Oe. For the NiFe layer the total field experienced by the layer consists of both the applied field and the dipolar field due to the Co layer. If the coercivity is negligible, the point where the magnetization is zero in the NiFe layer, HApplied, is balanced by HDemag. These points correspond to the points where the dipolar field from the bottom Co layer is canceling out the contribution from the applied field. At the same time, the dipolar field from the top NiFe layer influences the bottom Co layer so that the Co layer hysteresis loop in Fig. 4共b兲 is different from the previous ones in Figs. 2 and 3. In comparison, for the NiFe/ Co dot array without the spacer layer, the dipolar field effect is
S. S. P. Parkin, K. P. Roche, M. G. Samant, P. M. Rice, R. B. Beyers, R. E. Scheuerlein, E. J. O’ Sullivan, S. L. Brown, J. Bucchigano, D. W. Abraham, Y. Lu, M. Rooks, P. L. Trouilloud, R. A. Wanner, and W. J. Gallagher, J. Appl. Phys. 85, 5828 共1999兲. 2 W. J. Gallagher, S. S. P. Parkin, Y. Lu, X. P. Bian, A. Marley, R. A. Altman, S. A. Rishton, K. P. Roche, C. Jahnes, T. M. Shaw, and G. Xiao, J. Appl. Phys. 81, 3741 共1997兲. 3 S. Tehrani, J. M. Slaughter, E. Chen, M. Durlam, J. Shi, and M. DeHerrera, IEEE Trans. Magn. 35, 2814 共1999兲. 4 T. Miyazaki and N. Tezuka, J. Magn. Magn. Mater. 139, L231 共1995兲. 5 J. S. Moodera, L. R. Kinder, T. M. Wong, and R. Meservey, Phys. Rev. Lett. 74, 3273 共1995兲. 6 J. Barnas, A. Fuss, R. Camley, P. Grunberg, and W. Zinn, Phys. Rev. B 42, 8110 共1990兲. 7 Yu Lu, R. A. Altman, A. Marley, S. A. Rishton, P. L. Trouilloud, Gang Xiao, W. J. Gallagher, and S. S. P. Parkin, Appl. Phys. Lett. 70, 2610 共1997兲. 8 J. M. Daughton, J. Appl. Phys. 81, 3758 共1997兲. 9 B. A. Everitt, A. V. Pohm, and J. M. Daughton, J. Appl. Phys. 81, 4020 共1997兲. 10 M. Grimsditch, A. Berger, J. Johnson, V. Metlushko, B. Ilic, P. Neuzil, and R. Kumar, Europhys. Lett. 54, 813 共2001兲. 11 J. G. Zhu, Y. F. Zheng, and G. A. Prinz, J. Appl. Phys. 87, 6668 共2000兲. 12 T. Shinjo, T. Okuno, R. Hassdorf, K. Shigeto, and T. Ono, Science 289, 930 共2000兲. 13 J. Raabe, R. Pulwey, R. Sattler, T. Schweinböck, J. Zweck, and D. Weiss, J. Appl. Phys. 88, 4437 共2000兲. 14 M. Schneider, H. Hoffmann, and J. Zweck, Appl. Phys. Lett. 77, 2909 共2000兲. 15 R. P. Cowburn, D. K. Koltsov, A. O. Adeyeye, M. E. Welland, and D. M. Tricker, Phys. Rev. Lett. 83, 1042 共1999兲. 16 V. Novosad, M. Grimsditch, K. Y. Guslienko, P. Vavassori, Y. Otani, and S. D. Bader, Phys. Rev. B 66, 052407 共2002兲. 17 K. Y. Guslienko, V. Novosad, Y. Otani, H. Shima, and K. Fukamichi, Phys. Rev. B 65, 024414 共2002兲. 18 C. Kao, J. B. Hastings, E. D. Johnson, D. P. Siddons, G. C. Smith, and G. A. Prinz, Phys. Rev. Lett. 65, 373 共1990兲. 19 J. W. Freeland, J. C. Lang, G. Srajer, R. W. Winarski, D. Shu, and D. M. Mills, Rev. Sci. Instrum. 73, 1408 共2002兲. 20 K. Y. Guslienko, V. Novosad, Y. Otani, H. Shima, and K. Fukamichi, Appl. Phys. Lett. 78, 3848 共2001兲. 21 M. Bonfim, G. Ghiringhelli, F. Montaigne, S. Pizzini, N. B. Brookes, F. Petroff, J. Vogel, J. Camarero, and A. Fontaine, Phys. Rev. Lett. 86, 3646 共2001兲.
Downloaded 14 Jun 2006 to 131.193.45.131. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
