Microstructure and Magnetic Properties of Multilayer [FePt/Os]n Films

Delivered by Ingenta to:

INST. FOR APPLIED SCIENCE & ENGINEERING RESEARCH/ACADEMIA SINICA

IP : 140.109.229.234

Wed, 04 May 2011 16:03:14

Copyright © 2011 American Scientific Publishers All rights reserved

Printed in the United States of America

Nanoscience and Nanotechnology Letters Vol. 3, 285–288, 2011

Microstructure and Magnetic Properties of

Multilayer [FePt/Os]

_n

Films

D. P. Chiang

_{, S. Y. Chen}

_{, Y. D. Yao}

_{, C. C. Yu}

_{, Y. Y. Chen}

_{, and H. M. Lin}

1_{Center of General Education, Ming Hsin University of Science and Technology, Hsinchu 304, Taiwan} 2_{Department of Materials Engineering, Tatung University, Taipei 104, Taiwan}

3_{Institute of Physics, Academia Sinica, Taipei 115, Taiwan}

4_{Institute of Applied Science and Engineering, Fu Jen Catholic University, Taipei 242, Taiwan} 5_{Department of Applied Physics, University of Kaohsiung, Kaohsiung 811, Taiwan}

The microstructure and magnetic properties of multilayer [FePt(x)/Os]nfilms on glass and Si

sub-strates by dc-magnetron sputtering technique have been studied as a function of the annealing temperatures between 300 and 800 C. Here, x varied from 10, 20, 25, 50, to 100 nm with its associated n value of 10, 5, 4, 2, and 1, respectively. On glass, no diffusion evidence was found in all the samples. However, on Si, the insertion of a 10 nm Os layer into the FePt and Si interface results in better thermal stability. The Os underlayer can effectively prevent the diffusion of the inter-mixing between FePt layer and the Si(100) substrate for temperatures up to 800C. The grain size of the multilayer films can be well controlled by both the annealing temperature and the thickness of the FePt layers between the Os layers. The Os layers can effectively prevent the diffusion of the intermixing among the FePt layers and the Si(100) substrate. This means that the diffusion effect can be efficiently prevent in the multilayer [FePt(x)/Os]nfilms by the Os layers.

Keywords:

1. INTRODUCTION

SeveralL1₀structure alloys, such as FePt, FePd, and CoPd films, are potential candidates for ultrahigh density mag-netic recording media because of the high magnetocrys-talline anisotropy;1–7 _{among them, the ordered L1}

0 phase

of FePt exhibits one of the highest magnetic anisotropy energy and is a promising candidate for high density mag-netic recording media and high energy product materials in microelectromechanical systems. Generally speaking, as-deposited FePt films have a disordered fcc structure with a soft magnetic phase. The formation of the ordered fctL1₀ FePt hard magnetic phase requires preheating substrate or post-annealing the as-deposited film at a high temperature of above 500 C.8 _{However, this high temperature}

pro-cess results in the grain growth, poor surface roughness, inter diffusion between layers, which has the disadvan-tage of decreasing the recording density of the films and raises the production costs. So, it is a key challenge how to efficiently prevent the diffusion in a magnetic multilayer. From our recent works of the noble metal Osmium (Os) systems,9–12 _{Os has high melting and boiling point, which}

∗_{Author to whom correspondence should be addressed.}

is predicted to have good effect on preventing inter dif-fusion between layers induced by sputtering process. We have reported that the Os layer is not a good layer for magnetic exchange coupling,9 _{but it may be an excellent}

candidate for buffer and space layers for a hard magnetic system.10
11 _{These motivated us to study the thermal}

sta-bility of magnetic FePt/Os films on Si and glass substrates with/without insertion of Os spacer layers, especially on the magnetic properties and microstructure. In this inves-tigation, we report the Os buffer layer and inserted layer effects for FePt/Os multilayers on Si(100) and glass sub-strates from the variation of magnetic properties of the films.

2. EXPERIMENTAL DETAILS

The multilayer [FePt(x)/Os]_n films were deposited on a Si(100) and glass substrates with/without a 10-nm-thick Os underlayer at room temperature by dc-magnetron sput-tering system. Here, x varied from 10, 20, 25, 50, to 100 nm with its associated n value of 10, 5, 4, 2, and 1, respectively. The Os layers were used to control the tex-ture of hard magnetic layers and the grain size by proper annealing temperature and film thickness. The chamber

Delivered by Ingenta to:

INST. FOR APPLIED SCIENCE & ENGINEERING RESEARCH/ACADEMIA SINICA

IP : 140.109.229.234

Wed, 04 May 2011 16:03:14

Microstructure and Magnetic Properties of Multilayer [FePt/Os]_nFilms Chiang et al.

base pressure was less than 10 × 10−8 torr and the work-ing pressure was 50×10−5torr with the high purity argon source (99.999%). The FePt films were deposited by using Fe₅₀Pt₅₀ alloying target and the composition of Pt and Fe checking by inductively coupled plasma spectra is 48.6 and 51.4%, respectively. All the films were deposited at room temperature and then annealed subsequently. A post annealing procedure with temperature (T_A) ranged between 300 and 800 C for one hour was taken such that the FePt layers could be converted into its magnetically hard phases. The magnetic properties were measured at room temperature with a vibrating sample magnetometer (VSM) with a maximum applied field equal to 2 Tesla. Speci-mens for cross-section transmission electron microscopy (TEM) observation were prepared by mechanical grinding to a thickness of about 15m, followed by appropriate ion milling. The microstructures of the films were observed by the TEM with a nano-beam energy dispersive spectrome-ter. The crystal structure of the samples is characterized by a Philips PW3040/60 X-ray diffraction (XRD) with Cu-K radiation. The following analyses will be focused on the thermal stability of films structures on glasses or on Si substrates with/without an Os underlayer, especially on the magnetic properties and diffusion between the layers.

3. RESULTS AND DISCUSSION

Heat treatments of the as-deposited FePt films with/without an Os underlayer on both Si and glass substrates were studied between 300 and 800 C. Figure 1 shows the XRD patterns of the multilayer [FePt(x)/Os(5)]_n films on the Si substrate (a) for as-deposited Os(10)/FePt(100)/Os(5) samples, (b) for

Fig. 1. The XRD patterns of the FePt single-layer films on the Si(100) substrate. (a) as-deposited samples with/without the Os underlayer, (b) sample with a 10 nm Os underlayer after annealing at 700C for 1 hour, (c) FePt single-layer films without the Os underlayer after anneal-ing at 700C for 1 hour, and (d) [FePt(25)/Os(5)]4films without a Os underlayer after annealing at 700C for 1 hour.

Os(10)/FePt(100)/Os(5) samples after annealing at TA= 700 C for 1 hour, (c) for FePt(100)/Os(5)

sam-ples annealed at T_A= 700 C for 1 hour, and (d) for [FePt(25)/Os(5)]₄ samples annealed at T_A = 700 C for 1 hour. For the as-deposited films with a 10 nm Os underlayer, the XRD pattern displays a disordered face-centered-cubic (fcc) phase with (111) orientation only at 40.4 as shown in Figure 1 curve (a). For sample with Os underlayer (curve b), it is clear that the (111) peak of the post-annealing treated FePt flim shifted to 41 and the superlattice (001), (200), and (002) peaks showed up after the annealing at 700C; this indicates the formation of the L1₀ ordered phase. However, for samples without an 10 nm Os underlayer after annealing at 700 C, the FePt(111) peak is vanished as shown in curves (c) and (d), and some extra Pt₂Si₃(004), Pt₂Si₃(101), and Fe₅Si₃(100) peaks are identified according to the JCPDS file 89-2047 and 41-0874 as shown in curves (c) and (d). This means that no hard magnetic phases exist in these samples. Figure 2 shows the XRD patterns of the [FePt(20)/Os(5)]₅ films with a 10 nm Os underlayer, and (a) for the as deposited sample, (b) to (d) for samples annealed at TA= 400, 600, and 800 C, respectively. It is clear that

there is a shoulder at the right side of the peak near 41 for all the curves. By using a Lorentzian fit, a second peak at 42 could be obtained; this means that the Os(002) diffraction peak is embedded in this shoulder. For the as deposited sample, curve (a) displays a disordered fcc phase with (111) orientation only. After annealing at 400, 600, and 800C, a peak for (200) was observed as shown in curves (b) to (d). Some additional peaks of (001), (200), and (002) were observed. Kim et al.13_{reported that}

the origin of fct-FePt (001) texture evolution in FePt films was due to the occurrence of anisotropic strain during ordering transformation.

Fig. 2. The XRD patterns of the [FePt(20)/Os(5)]5films on Si substrate with a 10 nm Os underlayer. (a) is for the as-deposited film. (b), (c), and (d) are for the films annealed atTA= 400C, 600C, and 800C, respectively.

Delivered by Ingenta to:

INST. FOR APPLIED SCIENCE & ENGINEERING RESEARCH/ACADEMIA SINICA

IP : 140.109.229.234

Wed, 04 May 2011 16:03:14

Chiang et al. Microstructure and Magnetic Properties of Multilayer [FePt/Os]_nFilms

The inter diffusion behavior between the FePt layers and the Si(100) substrate can be studied by a HRTEM cross-sectional observation. As an example, Figure 3 shows the HRTEM cross-sectional image of a multilayer [FePt(25)/Os(5)]₄ on the Si(100) substrate (a) without and (b) with a 10 nm Os underlayer after annealing at 700C for 1 h, respectively. As shown in Figure 3(a) for the sam-ple without an Os underlayer, the interfaces between the layers were destroyed significantly due to the heavy inter-diffusion between the layers. Blurred interfaces between the FePt layers and the Si(100) substrate after high tem-perature annealing were displayed at the initial stage, but the FePt diffusion could be effectively blocked by the inserted Os layers. Roughly speaking, only the two top-most FePt layers can be recognized in Figure 3(a). How-ever, for the multilayer [FePt(25)/Os(5)]₄ sample with a 10 nm Os underlayer, well defined interfaces between all the FePt, Os, and Si(100) layers were observed, as depicted in Figure 3(b). Again, we have experimentally demon-strated that the diffusion effect can efficiently prevent in the hard magnetic FePt films by the Os buffer and space layers. For identifying the inhibition effect of the grain growth in FePt layers by the Os space layer, again as an example, Figure 4 shows the cross-section HRTEM images of a multilayer [FePt(25)/Os(5)]₄sample on glass substrate after annealing at 700 C for one hour. In general, the vari-ous gray regions are originated from the different crystallo-graphic orientations of the isotropically distributed grains. From Figures 3 and 4, It is clear that the Os space layers can limit the grain growth at the thickness direction.

To study the effect of Os layers on the magnetic prop-erty of the multilayer FePt/Os samples, as an example, Figure 5 shows the hysteresis loops of the multilayer [FePt(10)/Os(5)]₁₀ films on Si(100) substrate without an Os buffer layer annealed up to 700 C for 1 hour. For all the samples without an Os underlayer, the interfaces between the layers were destroyed significantly due to the heavy interdiffusion between the layers. Therefore, the coercivity increases quite small even after annealing

Fig. 3. Cross-section HRTEM images of the [Os(5)/FePt(25)]4films on the Si(100) substrate (a) without, and (b) with a 10 nm Os underlayer after annealing at 700C for one hour, respectively.

Fig. 4. Cross-section HRTEM images of the [Os(5)/FePt(25)]4films on the glass substrate after annealing at 700C for one hour.

above 700 C, its value is roughly near 2 kOe. For com-parison, Figure 6 plots the hysteresis loops of the mul-tilayer [FePt(20 nm)/Os(5 nm)]₅ films on glass substrate annealed up to 800 C for 1 hour. The coercivity increases with increasing annealing temperature very slowly below roughly 400 C. However, between 400 C and 600 C, it increases very fast from roughly 2 kOe at 400 C and up to roughly 9 kOe at 600 C. After 800 C annealing, the coercivity can up to roughly 11 kOe. This is explained by the different growth rate and grain size of the hard mag-netic FePt controlled by both annealing temperature and thickness of the FePt layers. This enhancement of coer-civity is related to the increase of the FePt L1₀ phase after annealing. The enhancement of coercivity can be understood from the fact that for a FePt/Os films with fixed thickness of Os spacer layers, the increasing num-ber of Os spacer layer will inhibit the grain growth of FePt grains and enriches the grain boundary. The layer by layer structure of the films can control the hard magnetic

Fig. 5. The coercivity versus the annealing temperature of the [Os(5)/FePt(10)]10 films on the Si(100) substrate without an Os under-layer. The external magnetic field is parallel to the film plane.

Delivered by Ingenta to:

INST. FOR APPLIED SCIENCE & ENGINEERING RESEARCH/ACADEMIA SINICA

IP : 140.109.229.234

Wed, 04 May 2011 16:03:14

Microstructure and Magnetic Properties of Multilayer [FePt/Os]_nFilms Chiang et al.

Fig. 6. The coercivity versus the annealing temperature of the [Os(5)/FePt(20)]5films on the glass substrate. The external magnetic field is parallel to the film plane.

Fig. 7. The coercivity of the [Os(t)/FePt(x)]nfilms withn = 4 and 10;

andt = 1 and 5 nm as a function of the annealing temperature between 300 and 800C for one hour.

behaviors and the increase of coercivity of the films with Os spacer layers could be attributed to the pinning of mag-netic domains at the grain boundary.

The coercivity as a function of annealing temperature for all the [FePt(x)/Os(t)]_n films withn = 4 and 10; t = 1 and 5 nm is depicted in Figure 7. From this figure, we demonstrate experimentally that the magnetic behavior of the multilayer films with the thickness of the inserted Os layers varied between 1 and 5 nm is almost the same; and it also indicated that the multilayer [FePt(x)/Os(t)]_n films even with a very thin Os spacer layer of 1 nm can efficiently prevent the diffusion of the FePt layers

and exhibit good hard magnetic properties. Because Os is the densest known element (22.59 g/cm3_{), that the}

Os layers could prevent the diffusion in the multilayer [FePt(x)/Os(t)]_n films is related to its dense atomic pack-ing in the lattice.

In summary, we have reported experimentally the ther-mal stability, the interlayer diffusion behavior and the mag-netic properties of the multilayer [FePt(x)/Os(t)]_nfilms on both Si(100) and glass substrates with/without a 10-nm-thick Os underlayer as functions of the annealing temper-atures between 300 and 800 C. The insertion of the Os layers into the FePt films on a Si(100) substrate results in better thermal stability as seen from the X-ray, TEM, and magnetic studies. The Os layers can effectively prevent the diffusion of the intermixing among the FePt layers and the Si(100) substrate for temperatures below 800C. This means that the diffusion effect can be efficiently prevent in the hard magnetic FePt films by the Os buffer and inter layers.

Acknowledgment: We are grateful for the financial support by Fu Jen Catholic University, the Sapintia Edu-cation Foundation and the National Science Council of Taiwan under Grant No. NSC99-2119-M-030-001.

References and Notes

1. D. Weller, A. Moser, L. Folks, M. E. Best, W. Lee, M. F. Toney, M. Schwickert, J. U. Thiele, and M. F. Doerner, IEEE Trans. Magn. 36, 10 (2000).

2. J. P. Liu, C. P. Luo, Y. Liu, and D. J. Sellmyer, Appl. Phys. Lett. 72, 483 (1998).

3. V. Nandwana, G. S. Chaubey, Y. Zhang, and J. P. Liu, J. Nanosci.

Nanotechnol. 10, 2979 (2010).

4. D. Goll and S. Macke, Appl. Phys. Lett. 93, 152512 (2008). 5. A. Perumal, Y. K. Takahashi, T. O. Seki, and K. Hono, Appl. Phys.

Lett. 92, 132508 (2008).

6. T. Seki, Y. Hasegawa, S. Mitani, S. Takahashi, H. Imamura, S. Maekawa, J. Nitta, and K. Takanashi, Nature Mater. 7, 125 (2008).

7. X. D. Liu, Z. Xu, R. X. Gao, Z. F. Chen, T. S. Lai, J. Du, and S. M. Zhou, J. Appl. Phys. 106, 053907 (2009).

8. C. M. Kuo, P. C. Kuo, H. C. Wu, Y. D. Yao, and C. H. Lin, J. Appl.

Phys. 85, 4886 (1999).

9. S. Y. Chen, Y. D. Yao, and J. M. Wu, J. Magn. Magn. Mater. 310, 1914 (2007).

10. T. Y. Peng, C. K. Lo, S. Y. Chen, and Y. D. Yao, J. Appl. Phys. 99, 08C907-1 (2006).

11. T. Y. Peng, C. K. Lo, S. Y. Chen, and Y. D. Yao, IEEE Trans. Magn. 43, 894 (2007).

12. T. Y. Peng, C. K. Lo, Y. D. Yao, and S. Y. Chen, Appl. Phys. Lett. 90, 121904 (2007).

13. J. S. Kim, Y. M. Koo, and B. J. Lee, J. Appl. Phys. 99, 053906 (2006).

Received: 20 July 2010. Accepted: 20 November 2010.