Magnetic anisotropy of permalloy films grown on an Mo
„
001
…
stepped surface
C. C. Yu and J. C. A. Huanga)
Department of Physics, National Cheng-Kung University, Tainan, Taiwan Y. M. Hu
Department of Electric Engineering, Nan-Jeon Junior College of Technology and Commerce, Tainan, Taiwan
The Mo共001兲 stepped surface was self-assembled on Al2O3共1-102兲 substrate by molecular beam epitaxy. Ni80Fe20films were subsequently deposited on the fresh Mo共001兲 stepped surface at growth temperature (Tg) between 30 and 400 °C. Interestingly, in the low Tg regime (⬍100 °C) the Ni80Fe20films were grown as共001兲 single crystal while high Tg(⬎300 °C) results in the growth of Ni80Fe20 共110兲 quadcrystal. For low Tg samples, the easy axis of the Ni80Fe20 film is directed parallel to the underlying Mo step edge direction and the M-H loop along the hard axis reveals a two-step magnetic moment rotation process. In contrast, the high Tgsamples display isotropic-like magnetic behavior with a much enhanced coercivity due likely to the surface roughness effect. We conclude that the growth temperature of Ni80Fe20film on the Mo stepped surface plays an important role on the crystal structure and magnetic anisotropy of Ni80Fe20film. © 2001 American Institute of Physics. 关DOI: 10.1063/1.1354588兴
Bulk permalloy (Py⫽Ni80Fe20) is known as a material with high permeability, low coercivity, and small magnetic anisotropy.1 Py based multilayers2,3 and spin-valve4,5 struc-tures have attracted considerable interest due to the discovery of giant magnetoresistance phenomena with low saturation fields, which can be applied for advanced magnetic sensors. Understanding and controlling the magnetic behavior of Py films are thus important.
Step induced magnetic anisotropy plays an important role in thin film magnetism.6,7 Usually a stepped surface is made by artificially miscutting a small angle of a desired substrate. Here we study the structure and magnetism of Py grown on the self-assembled Mo stepped surface on sapphire substrate. Note that the formation of the Mo stepped surface is essentially due to the coherent tilt8,9mechanism.
The epitaxial Py films presented here were prepared in a Vacuum Product made MBE-930 chamber equipped with five electron-beam deposition sources. The base pressure of the molecular beam epitaxy system before deposition was kept about 2⫻10⫺10torr and during growth the pressure was below 5⫻10⫺9 torr. One-side polished Al2O3共1-102兲 sub-strates were outgassed at 1000 °C for 1 h prior to the initial epitaxial growth. Note that the Al2O3共1-102兲 substrates were purposely miscut 共about 1°) toward the c-axis direction. Seeding layers of 100 Å Mo were prepared on Al2O3共1-102兲 substrates at 900 °C and subsequent Py films were then grown on the Mo layer at Tg between 30 and 400 °C. The deposition rates for Py and Mo films were controlled at about 0.1 Å/s. The crystal structures and epitaxial relations of the films were examined in situ by reflection high-energy elec-tron diffraction共RHEED兲 and ex situ by synchrotron radia-tion x-ray diffracradia-tion 共XRD兲. The surface morphology and magnetic anisotropy of the Ni80Fe20 films were studied by
atomic force microscopy 共AFM兲 and longitudinal magneto-optical Kerr effect 共LMOKE兲, respectively.
High quality, single crystal bcc Mo共001兲 films were grown on Al2O3共1-102兲 substrate. Interestingly, XRD showed that the Mo共001兲 plane tilts about 3° from the Al2O3共1-102兲 plane at a growth temperature of 900 °C. In addition, the Mo共001兲 stepped surface was formed with steps along the in-plane Mo关110兴 direction. The self-assembled Mo共001兲 stepped surface structure is due to the formation of a coherent tilt boundary, which in turn results from the uniaxial strain exerted by the sapphire substrate along the step共Mo关110兴兲 direction. Note that the symmetric tilt of the Mo steps was broken by the miscut of the sapphire substrate.8,9
The stepped surface of Mo共001兲 was evidenced by RHEED observations. As shown in Fig. 1共a兲 with an electron-beam directed along the Mo关1-10兴 azimuth, there are two diffused splitting streaks resulting from by surface steps. For the reader’s convenience, the lattice geometry is provided in Fig. 1共b兲. Also an intensity contour plot of a central streak is shown in Fig. 1共c兲 for clarity. The dashed lines in Fig. 1共c兲 indicate the splitting streaks and the spacing between them is inversely proportional to the average terrace size. The average terrace size and the step height were de-termined as 30 and 1.87 Å, respectively.
On the fresh Mo共001兲 stepped surfaces, the structure of the subsequent Ni80Fe20 共Py兲 films depends strongly on the growth temperature (Tg). As shown by XRD in Fig. 2共a兲, in the low Tg regime (⬍100 °C) the Py film was grown as a 共001兲 oriented single crystal with the in-plane azimuth rela-tion Py关110兴//Mo关100兴. In the high Tg regime (⬎300 °C),
however, a Py共110兲 phase was formed as shown in Fig. 2共b兲. At intermediate Tg, mixed 共001兲 and 共110兲 structures were
observed. The in-plane XRD scans further reveal that the Py共110兲 film was grown as a 共110兲 quadcrystal structure in a兲Electronic mail: [email protected]
JOURNAL OF APPLIED PHYSICS VOLUME 89, NUMBER 11 1 JUNE 2001
7380
0021-8979/2001/89(11)/7380/3/$18.00 © 2001 American Institute of Physics
the high Tg regime. The four equivalent domains are rotated by ⫹10°/⫺10° from the underlying Mo关100兴 and Mo关001兴 azimuth, as schematically illustrated in Fig. 3. It is believed that the Mo step edges assist the growth of a Py共001兲 single crystal or a Py共110兲 quadcrystal in different temperature regimes. This is in marked contrast to the growth of the bicrystal Py共110兲 structure on the nonstepped Mo共001兲 sur-face关on MgO共001兲 for example兴.10
For the Py films grown on Mo surface stepped surfaces, uniaxial and isotropic like magnetic behaviors were observed for low Tg(⬍100 °C) and high Tg(⬎300 °C), respectively.
As shown in Fig. 4共a兲, at low Tgthe easy axis of the Py film
is directed parallel to the underlying Mo step edge direction 共//Mo关1-10兴 or ⌽⫽0°). Along the hard axis 共//Mo关110兴 or ⌽⫽90°), in contrast, the M-H loops reveal a two-step mag-netic moment rotation process, a characteristic of a magmag-netic film on a stepped surface.
On the other hand, the high Tg Py films display
isotropic-like magnetic behavior with a much enhanced co-ercive field, as shown, for example, in Fig. 4共b兲. The AFM scans reveal that the surface roughness of the Py films in-creases with increasing Tg. The growth temperature depen-dence of the surface roughness together with the coercive field of the Py films are shown in Fig. 5. We believe that at higher Tg the surface roughness 共due to three-dimensional nucleation兲 can suppress the step-induced magnetic anisot-ropy and provide a stronger pinning force for the motion of Py domains.
FIG. 1. 共a兲 RHEED image showing the Mo共001兲 stepped surface with the electron-beam directed parallel to the Mo关1-10兴 azimuth as illustrated in 共b兲. 共c兲 An intensity contour plot of the central streak.
FIG. 2. X-ray diffraction spectra of 100 Å Py/100 Å Mo/Al2O3共1-102兲 for
Tg⫽共a兲100 °C, and 共b兲300 °C.
FIG. 3. Schematic diagram showing the Py共110兲 quadcrystal grains on the Mo共001兲 stepped surface.
FIG. 4. LMOKE hysteresis loops of 100 Å Py grown on the Mo共001兲 stepped surface with Tg⫽共a兲100 °C, and 共b兲300 °C.
FIG. 5. The surface roughness and coercivity as a function of growth tem-perature for Py films grown on the Mo共001兲 stepped surface.
7381
J. Appl. Phys., Vol. 89, No. 11, 1 June 2001 Yu, Hunag, and Hu
Finally, the annealing effect of the low Tg samples was also measured. The loop splitting along the hard axis de-creases with increasing annealing temperature and disappears at about 250 °C. This suggests that the uniaxial anisotropy behavior of the Py共001兲 film is due to the step effect, which can be smeared out by either the high temperature growth or annealing treatment. We conclude that the Mo stepped sur-face plays an important role in the crystal structure and mag-netic anisotropy of Py overlayers.
We are grateful for the financial support from the ROC NSC under Grant No. 89-2112-M-006-037.
1E. E. Huber, Jr., D. O. Smith, and J. B. Goodenough, J. Appl. Phys. 29, 294共1958兲.
2K. Rook, A. M. Zeltser, J. O. Artman, D. E. Laughlin, and R. M. Chrenko, J. Appl. Phys. 69, 5670共1991兲.
3T. Valet, J. C. Jacquet, P. Galtier, J. M. Coutelier, L. G. Pereira, R. Morel, D. Lottis, and A. Fert, Appl. Phys. Lett. 61, 3178共1992兲.
4
B. Dieny, V. S. Sperious, S. S. P. Parkin, B. A. Gurney, D. R. Wilhoit, and D. Mauri, Phys. Rev. B 43, 1297共1991兲.
5Th. G. S. M. Rijks, W. J. M. de Jonge, W. Folkerts, J. C. S. Kools, and R. Coehoorn, Appl. Phys. Lett. 65, 916共1994兲.
6
R. K. Kawakami, E. J. Escorcia-Aparicio, and Z. Q. Qiu, Phys. Rev. Lett.
77, 2570共1996兲.
7J. S. Suen and J. L. Erskine, Phys. Rev. Lett. 78, 3567共1997兲. 8R. R. Du and C. P. Flynn, J. Phys.: Condens. Matter 2, 1338共1990兲. 9J. C. A. Huang, R. R. Du, and C. P. Flynn, Phys. Rev. B 44, 4060共1991兲. 10
J. C. A. Huang, C. C. Yu, and C. H. Lee, J. Appl. Phys. 87, 4921共2000兲.
7382 J. Appl. Phys., Vol. 89, No. 11, 1 June 2001 Yu, Hunag, and Hu
