Enhanced exchange bias coupling in Fe/ Fe

(1)

Enhanced exchange bias coupling in Fe/ Fe

_x

Mn

_1−x

bilayer by reducing vertical lattice constants

Wen-Chin Lin, Bo-Yao Wang, Te-Yu Chen, Ling-Chih Lin, and Yu-Wen Liao Department of Physics, National Taiwan University, 10617 Taipei, Taiwan

and Institute of Atomic and Molecular Sciences, Academia Sinica, 10617 Taipei, Taiwan Wei Pan

Department of Physics, National Taiwan University, 10617 Taipei, Taiwan

and Department of Physics, National Chung Cheng University, Min-Hsiung Chia-Yi 62102, Taiwan Nai-Yeou Jih

Department of Physics, National Taiwan University, 10617 Taipei, Taiwan Ker-Jar Song

Institute of Atomic and Molecular Sciences, Academia Sinica, 10617 Taipei, Taiwan Minn-Tsong Lin^a兲

Department of Physics, National Taiwan University, 10617 Taipei, Taiwan

and Institute of Atomic and Molecular Sciences, Academia Sinica, 10617 Taipei, Taiwan

共Received 20 September 2006; accepted 18 December 2006; published online 29 January 2007兲 The modification of crystalline structure by epitaxial growth on different single crystals induces crucial effects on the exchange bias coupling. Due to the larger lattice constant共a0兲 of Cu3Au共100兲共a0= 3.75 Å兲, the vertical lattice constants of Fe/FexMn_1−xfilms on Cu₃Au共100兲 are much smaller than those of the Fe/ FexMn_1−x/ Cu共100兲 system 共Cu: a0= 3.61 Å兲. By reducing the vertical lattice constants, the interface exchange bias coupling energy of Fe/ FexMn_1−x/ Cu₃Au共100兲 is enhanced to be 0.12– 0.18 erg/ cm², which is approximately four times that of the Cu共100兲 system. © 2007 American Institute of Physics. 关DOI:10.1063/1.2435514兴

Due to the application of magnetoresistance in memory recording, read sensor, etc., many studies¹about the various exchange bias systems have been performed in the last de- cade. Basically the exchange bias systems are composed of ferromagnetic共FM兲 and antiferromagnetic 共AFM兲 layers. Af- ter a field-cooling process, the exchange coupling between FM and AFM layers induces a unidirectional anisotropy.

Thus a shift from the zero field position of the hysteresis loop for FM layer, the so-called exchange bias field共He兲, as well as the enhanced coercivity 共Hc兲 usually can be observed.¹

From the application’s point of view, how to advance the He in an exchange bias system is an important issue. The physical factors to promote the exchange bias coupling are also of fundamental interests. For example, in many previous studies,¹ the high temperature annealing is used to increase He. The annealing process usually induces alloy interdiffu- sion and structural transition at the FM/AFM interface. Most of the reports are largely empirical and the physics is also quite complicated.¹ Usually, the epitaxial growth of single- crystalline metallic FM film on metallic AFM Mn alloy may simplify the physical conditions and then it will be possible to characterize or manipulate the FM/AFM coupling in more detail. In this letter, single-crystalline AFM FexMn_1−x alloy films are used to detect the structural effect on the exchange bias coupling. Several previous reports are already focused on the structural and magnetic properties of single-crystalline FexMn_1−x thin films on Cu共100兲 because of the small lattice mismatch.^2–9 We tune the crystalline structure of the FM/

AFM layers through the epitaxial growth on different sub-

strates of Cu₃Au共100兲 and Cu共100兲 with the lattice constants of 3.75 and 3.61 Å, respectively. In Fig. 1 the lattice mismatch between Cu, Cu₃Au, and FexMn_1−x bulk alloy is shown as functions of the Fe concentration x.¹⁰ Since the lattice constant of the bulk Fe_xMn_1−xalloy is right between those of the Cu and Cu₃Au, FexMn_1−x films can be grown on Cu共100兲 and Cu3Au共100兲 with different strain directions, which may cause expanded and suppressed vertical lattice constants共d_⬜兲, respectively. By reducing the vertical lattice constants, the interface exchange bias coupling energy of Fe/ FexMn_1−x/ Cu₃Au共100兲 is enhanced to be 0.12– 0.18 erg/ cm², which is approximately four times that of the Cu共100兲 system.

a兲Electronic mail: [email protected]

FIG. 1. Lattice mismatch of FexMn_1−x/ Cu共100兲 and FexMn_1−x/ Cu₃Au共100兲 systems shown as functions of Fe concentration x 共Refs.2and10兲. The lattice mismatch is calculated by the definition共af− as兲/as, where afand as

are the lattice constants of the bulk Fe_xMn_1−x共Ref.10兲 and the substrate.

APPLIED PHYSICS LETTERS 90, 052502共2007兲

Downloaded 30 Jan 2007 to 140.112.101.92. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp

(2)

The studies of Cu₃Au共100兲 and Cu共100兲 systems were carried out in ultrahigh vacuum chambers with the base pres- sure better than 2⫻10⁻¹⁰torr. The substrates were prepared by cycles of sputtering and annealing.^4,11,12 Fe_xMn_1−x alloy films were grown by coevaporation, monitored by medium energy electron diffraction 共MEED兲. From MEED oscillations, the deposition rate was calibrated precisely. The lateral crystalline structure was characterized by low energy electron diffraction 共LEED兲. From the LEED-I/V curves, the average vertical lattice constant共d_⬜兲 of the film was deter- mined by kinetic approximation.¹¹Magnetic properties of the films were detected by longitudinal magneto-optical Kerr effect共MOKE兲.

Figure 2共a兲 shows the MEED oscillation curves of FexMn_1−x alloy films grown on Cu₃Au共100兲 at 300 K 共RT兲.

FexMn_1−x films reveal regular oscillations. From the clear LEED patterns of FexMn_1−x/ Cu₃Au共100兲 shown in Fig.3, as well as the previous studies on FexMn_1−x/ Cu共100兲,^2–9 FexMn_1−x exhibits well-ordered crystalline structure, coher- ent growth, and smooth surface on both Cu₃Au共100兲 and Cu共100兲 over a wide range of alloy composition. Surpris- ingly, the larger lattice mismatch in FexMn_1−x/ Cu₃Au共100兲 does not cause significant changes in the growth behavior. In Fig.2共b兲, Fe films grown on various FexMn_1−x/ Cu₃Au共100兲 reveal nearly the same features in MEED curves, due to the well-ordered structure and the flat surface of FexMn_1−x. In our previous studies, Fe films on 17 ML Fe_xMn_1−x/ Cu共100兲 also exhibit high MEED intensity until 20-25 ML.^4,13

Figure4 shows the vertical lattice constants共d_⬜兲 of the Fe and FexMn_1−xlayers grown on Cu₃Au共100兲 and Cu共100兲.

As indicated in Fig.1, the positive lattice mismatch makes the d_⬜ of FexMn_1−x/ Cu共100兲 larger than that of bulk FexMn_1−x. The large negative lattice mismatch makes the d_⬜ of Fe_xMn_1−x/ Cu₃Au共100兲 much smaller than that of bulk Fe_xMn_1−x. 15 ML Fe on 17 ML Fe_xMn_1−x/ Cu共100兲 reveal d_⬜⯝2.05 Å with x艋30%.^14,15When x⬎30%, the 15 ML Fe

is of fcc structure共c/a⯝1.0兲, which is confirmed to be non- ferromagnetic in our previous studies.^4,13For 21 ML Fe films on 15 ML FexMn_1−x/ Cu₃Au共100兲, the d_⬜is deduced to be

⯝1.47 Å.¹⁶Apparently, the epitaxial growth of Fe/ Fe_xMn_1−x bilayers on Cu共100兲 and Cu3Au共100兲 is shown to induce significant differences in d_⬜. Fe/ FexMn_1−x reveals similar cubic structures on both substrates, but with the more sup- pressed d_⬜ while grown on Cu₃Au共100兲, due to the larger in-plane lattice constant.

Figure5共a兲exhibits two examples of the MOKE hysteresis loops for Fe/ Fe_xMn_1−x bilayers on Cu₃Au共100兲 and Cu共100兲, after field cooling from 300 to 100–110 K along 关010兴. In the Cu3Au共100兲 system, the bias field 共He兲 ranges between 200 and 300 Oe with x = 0 % – 54% at 100 K. In the Cu共100兲 system, He ranges only 60– 80 Oe with x

= 20% – 30% at 110 K. Due to the nonferromagnetism of fcc Fe films, no hysteresis loop can be observed with x

⬎30%.^4,13

For the comparison between different studies, as shown in Fig. 5共b兲, the interface exchange bias coupling energy E_ex= H_et_FMM_FM is calculated,¹ where t_FM and M_FM are the thickness and saturated magnetization of the ferromagnetic Fe layer, respectively. By substituting M_FM= 2.2␮B,¹⁷ E_exis evaluated as 0.03– 0.04 erg/ cm² for Fe/ FexMn_1−x/ Cu共100兲

FIG. 2.共a兲 MEED oscillation curves of various FexMn_1−xfilms grown on Cu₃Au共100兲 at RT. 共b兲 MEED curves of Fe films grown on 15 ML FexMn_1−x/ Cu₃Au共100兲 at RT.

FIG. 3. 共Color online兲 LEED patterns of the various 15 ML FexMn_1−x/ Cu₃Au共100兲 taken at 100 K with the beam energy of 120 eV.

FIG. 4. 共Color online兲共a兲 Vertical lattice constants of 17 ML FexMn_1−x/ Cu共100兲共square兲共Ref. 13兲, bulk FexMn_1−x共solid circle兲共Ref.

10兲, and 15 ML FexMn_1−x/ Cu₃Au共100兲共open circle兲 as functions of Fe concentration x. All the FexMn_1−xfilms are grown at RT.共b兲 Vertical lattice constants of 15 ML Fe films grown on 17 ML FexMn_1−x/ Cu共100兲共circle兲共Ref. 13兲 and 21 ML Fe films grown on 15 ML FexMn_1−x/ Cu₃Au共100兲共square兲. The arrows indicate the lattice constants of bulk Cu, Cu3Au, and Fe.

FIG. 5.共Color online兲共a兲 Normalized hysteresis loops of 21 ML Fe/15 ML Fe_0.34Mn_0.66/ Cu₃Au共100兲 and 15 ML Fe/17 ML Fe0.3Mn_0.7/ Cu共100兲 mea- sured at 100 and 110 K, respectively.共b兲 Interface exchange bias coupling energy共Eex兲 of the Fe/FexMn_1−xbilayers on Cu₃Au共100兲 and Cu共100兲 calculated from the MOKE data.

052502-2 Lin et al. Appl. Phys. Lett. 90, 052502共2007兲

(3)

with x = 20% – 30%. For Fe/ FexMn_1−x/ Cu₃Au共100兲, Eex

= 0.12– 0.18 erg/ cm² with x = 0 % – 54%, which is approxi- mately four times that of the Cu共100兲 system. Besides, the Fe/ FexMn_1−x/ Cu₃Au共100兲 also reveals superiority over other similar systems, e.g., Co/ FexMn_1−x and Ni-Fe/ FexMn_1−x bilayers on Cu共100兲. In the study by Offi et al.’s, a change in easy axis from the具110典 to the 具001典 azimuth directions is observed at room temperature in a Co film coupled to a Fe₅₀Mn₅₀ film thicker than 10 ML.² It originates from the different orientations of the easy axis for Co film,具110典, and the pinning direction of AFM Fe₅₀Mn₅₀,具001典.³This discrep- ancy adds more complexity in Co/ FexMn_1−x exchange bias system and the bias field is shown to be small.²In the studies of Ni– Fe/ fcc-Fe₅₀Mn₅₀bilayers,^1,7,18,19 where the Fe₅₀Mn₅₀ is prepared in either single crystalline or 共100兲 texture, the exchange bias coupling energy E_ex is evaluated as 0.04– 0.07 erg/ cm², close to Fe/ FexMn_1−x/ Cu共100兲 and much smaller than Fe/ FexMn_1−x on Cu₃Au共100兲.

From the results of Figs.4 and5, the suppressed d_⬜in Fe/ FexMn_1−xlayers on Cu₃Au共100兲 could be strongly corre- lated to the enhanced exchange bias energy. Intuitively when the FM and AFM layers are arranged more compactly, the overlap of the electron wave functions might be also in- creased. Thus a stronger exchange bias coupling could be obtained. However, the detailed correlations between the d_⬜ in crystalline structure and the exchange bias coupling still need further theoretical studies.

In summary, this study indicates the possibility of opti- mizing the exchange bias by tuning the vertical lattice constants of the FM/AFM bilayers. As compared with the Cu共100兲 system, the larger lattice constant of Cu3Au共100兲 results in significant suppression of the vertical lattice constants in Fe and FexMn_1−xlayers. The reduced vertical lattice constants in Fe/ FexMn_1−x/ Cu₃Au共100兲 enhance the interface exchange bias coupling energy E_exto be 0.12– 0.18 erg/ cm², which is approximately four times that of the Cu共100兲 system. The large exchange bias and well-ordered crystalline structure also suggest Fe/ FexMn_1−x/ Cu₃Au共100兲 as one of

the model systems for the further studies of exchange bias coupling.

This work was supported by the National Science Coun- cil of Taiwan under Grant Nos. NSC 94-2112-M-002-005, 95-2120-M-002-015, 95-2112-M-002-MY3, and 94-2112-M- 001-045.

1J. Nogués and I. K. Schuller, J. Magn. Magn. Mater. 192, 203共1999兲.

2F. Offi, W. Kuch, and J. Kirschner, Phys. Rev. B 66, 064419共2002兲.

3W. Kuch, F. Offi, L. I. Chelaru, M. Kotsugi, K. Fukumoto, and J.

Kirschner, Appl. Phys. Lett. 86, 122504共2005兲.

4C. C. Kuo, W. Pan, Y. C. Chen, and Minn-Tsong Lin, J. Appl. Phys. 93, 8743共2003兲.

5C. Won, Y. Z. Wu, H. W. Zhao, A. Scholl, A. Doran, W. Kim, T. L.

Owens, X. F. Jin, and Z. Q. Qiu, Phys. Rev. B 71, 024406共2005兲.

6F. Matthes, A. Rzhevskii, L.-N. Tong, L. Malkinski, Z. Celinski, and C. M.

Schneider, J. Appl. Phys. 93, 6504共2003兲.

7R. Jungblut, R. Coehoorn, M. T. Johnson, J. aan de Stegge, and A.

Reinders, J. Appl. Phys. 75, 6659共1994兲.

8R. Thamankar, S. Bhagwat, and F. O. Schumann, Phys. Rev. B 69, 054411共2004兲.

9R. Thamankar, S. Bhagwat, and F. O. Schumann, Phys. Rev. B 69, 054419共2004兲.

10Yasuo Endoh and Yoshikazu Ishikawa, J. Phys. Soc. Jpn. 30, 1614共1971兲.

11W. C. Lin, L. C. Lin, T. Y. Chen, B. Y. Wang, Ker-Jar Song, and Minn-Tsong Lin, J. Appl. Phys. 97, 10K112共2005兲.

12B. Y. Wang, W. C. Lin, Y. W. Liao, Ker-Jar Song, and Minn-Tsong Lin, Surf. Sci. 600, 4517共2006兲.

13Wei Pan, W. C. Lin, N. Y. Jih, Y. C. Chen, C. H. Chuang, C. C. Kuo, and Minn-Tsong Lin, Phys. Rev. B 74, 224430共2006兲.

14S. Muller, P. Bayer, C. Reischl, K. Heinz, B. Feldmann, H. Zillgen, and M.

Wuttig, Phys. Rev. Lett. 74, 765共1995兲.

15K. Kalki, D. D. Chambliss, K. E. Johnson, R. J. Wilson, and S. Chiang, Phys. Rev. B 48, 765共1993兲.

16M.-T. Lin, J. Shen, W. Kuch, H. Jenniches, M. Klaua, C. M. Schneider, and J. Kirschner, Phys. Rev. B 55, 5886共1997兲.

17K. Schwarz and D. R. Salahub, Phys. Rev. B 25, 3427共1982兲.

18R. Jungblut, R. Coehoorn, M. T. Johnson, Ch. Sauer, P. J. van der Zaag, A.

R. Ball, Th. G. S. M. Rijks, J. aan de Stegge, and A. Reinders, J. Magn.

Magn. Mater. 148, 300共1995兲.

19T. Lin, D. Mauri, N. Staud, C. Hwang, J. K. Howard, and G. L. Gorman, Appl. Phys. Lett. 65, 1183共1994兲.

052502-3 Lin et al. Appl. Phys. Lett. 90, 052502共2007兲