Film Structures of Fe/B-doped Carbon/Fe3Si Spin Valve Junctions

Kazuki Kudo1, Kazutoshi Nakashima1, Satoshi Takeichi1, Rezwan Ahmed2, Seigi Mizuno2, Ken-ichiro Sakai3*, Masahiko Nishijima4, and Tsuyoshi Yoshitake1**

1 Department of Applied Science for Electronics and Materials, Kyushu University, Kasuga,

Fukuoka 816-8580, Japan 2 Department of Molecular and Material Sciences, Kyushu University, Kasuga,

Fukuoka 816-8580, Japan 3 Department of Control and Information Systems Engineering, National Institute of

Technology, Kurume College, Kurume, Fukuoka 830-8555, Japan 4 The Electron Microscopy Center, Technology Center for Research and Education

Activities, Tohoku University, Sendai, Miyagi 980-8577, Japan

E-mail: *kenichiro_sakai@kyudai.jp; **tsuyoshi_yoshitake@kyudai.jp

(Received January 24, 2017)

Fe/B-doped carbon/Fe3Si trilayered films were prepared on Si(111) substrates by physical vapor

deposition with a mask method, and the film-structures and magnetic properties of the films were

investigated. The Fe3Si and Fe layers were deposited by facing targets direct-current sputtering

(FTDCS), and the B-doped carbon layers were deposited by coaxial arc plasma deposition (CAPD)

with B-blended graphite targets. Here, since the B-doped carbon layers were deposited by CAPD

in the same manner as the deposition of B-doped ultrananocrystalline diamond/hydrogenated

amorphous carbon composite (UNCD/a-C:H) films, the interlayers should be B-doped UNCD/a-

C:H. The formation of a layered structure was confirmed by transmission electron microscopy

(TEM). The diffusion of Fe and Si atoms into the interlayer occurs in the range of several

nanometers. The shape of the magnetization curve has clear steps that evidently indicate the

formation of antiparallel alignment of magnetizations owing to the difference in the coercive forces

between the top Fe and bottom Fe3Si layers. It was experimentally demonstrated that B-doped

UNCD/a-C:H is applicable to Fe-Si system spin valves as interlayer materials.

1. Introduction

Spintronics devices that utilize both electric charge and spin of electrons have actively been studied.

Giant magnetoresistance (GMR) [1,2] and tunnel magnetoresistance (TMR) [3-8] effects are

representative phenomena utilizing the spin-dependent scattering of carriers.

Whereas GMR and TMR films employ nonmagnetic metal and insulator as interlayers,

respectively, there have been a few studies on spintronics devices comprising semiconductors. Spin

injection from ferromagnetic materials into semiconductors such as Si, Ge and GaAs, has been

widely investigated thus far [9-14]. Recently, spin injection into semiconducting graphene has

received much attention because of a long spin diffusion length being expected due to weak spin-

orbit interactions [15-17].

Our laboratory has structurally and electrically investigated ultrananocrystalline diamond

(UNCD)/hydrogenated amorphous carbon (a-C:H) composite (UNCD/a-C:H) films, comprising a

large number of nano-sized diamond grains and an a-C:H matrix, prepared by coaxial arc plasma

deposition (CAPD), thus far [18-22]. The UNCD/a-C:H films prepared by CAPD have the following

characteristics: (i) the production of p and n-type conduction accompanied by enhanced electrical

conductivities is possible by B and N doping, respectively; (ii) they can be grown on foreign solid

substrates; (iii) they can be grown at low substrate temperatures even on unheated substrates, while

JJAP Conf. Proc. , 011502 (2017) https://doi.org/10.7567/JJAPCP.5.0115025Proc. Asia-Pacific Conf. on Semiconducting Silicides and Related Materials 2016

© 2017 The Author(s). Content from this work may be used under the terms of the .Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

Creative Commons Attribution 4.0 license

chemical vapor deposition (CVD) requires high substrate temperatures of more than 700 °C for the

growth [23,24]. UNCD/a-C:H is a carbon-based semiconductor similarly to graphene, and it is a new

candidate as spin transport materials.

We have studied Fe-Si based artificial lattices and spin valves comprising ferromagnetic Fe3Si

and semiconducting FeSi2, prepared by sputtering, thus far [25-36]. Based on the preparation

techniques in our previous researches, spin valve junctions comprising ferromagnetic Fe3Si and Fe

layers and B-doped UNCD/a-C:H interlayers were prepared, and they were structurally investigated

by transmission electron microscopy (TEM). We report that it is possible for layered structures to be

formed owing to the low substrate temperature growth of B-doped UNCD/a-C:H interlayers.

2. Experimental methods

Fe (100 nm)/B-doped ultrananocrystalline diamond (10 nm)/Fe3Si (100 nm) trilayered spin valve

junction were deposited on p-type Si(111) substrates by a mask method, as shown in Fig. 1(a). First,

the Si(111) substrate with specific resistance range of 30008000 cm, produced by a floating zone

(FZ) method, was cleaned with 1 hydrofluoric acid and rinsed in deionized water before it was set

into the vacuum chamber of a facing target direct-current sputtering (FTDCS) apparatus. The bottom

Fe3Si layer (100 nm) was deposited on the Si(111) substrate at an Ar pressure of 1.33 101 Pa and

substrate temperature of 300 °C, by FTDCS using the 1st mask. Subsequently, after the exposure to

air for the mask exchange, the UNCD/a-C:H interlayer (10 nm) was deposited with the 2nd mask at

a H2 pressure of 53.3 Pa and substrate temperature of 300 °C by CAPD with a 10 at.% B-blended

graphite target. After the exposure to air for the mask exchange again, the top Fe layer (100 nm) was

deposited with the 3rd mask at an Ar pressure of 1.33 101 Pa and room temperature by FTDCS

apparatus. The optical top-view image of the resultant junction is shown in Fig. 1(b). The B content

of the UNCD/a-C:H interlayer was estimated to approximately 5 at. % from the X-ray photoemission

spectra. The microstructure of the trilayered films were structurally observed by TEM combined with

energy dispersive X-ray (EDX) spectroscopy. The magnetization curves of the junctions were

measured at room temperature using a vibrating sample magnetometer (VSM).

3. Results and discussion

Figure 2 shows the XRD pattern of the junction, measured in 2θ-θ. The junction exhibit diffraction

Fig. 1. (a) Schematic diagram of deposition procedure of spin valve junction comprising Fe3Si, B-doped

UNCD/a-C:H, and Fe trilayers and (b) optical top-view image of the spin valve junction.

(a) (b)

Fe3Si Fe

UNCD

011502-2JJAP Conf. Proc. , 011502 (2017) 5

peaks of Fe3Si-222 and Fe-110 and 200. In our previous study, we confirmed that Fe3Si oriented

grains are also in-plane ordered on Si(111) substrates, by X-ray diffraction (XRD) [35,36]. From a

pole figure pattern concerning the Fe3Si-422 plane with a rotation axis of Fe3Si [222], the in-plane

orientation of the Fe3Si layer was confirmed. Totally considering this results and our previous

research results [26], the bottom Fe3Si layer is epitaxially grown on the Si(111) substrate. Since the

top Fe layer exhibits Fe-110 and 220 peaks and it is deposited on the non-oriented UNCD/a-C:H

layer, the Fe layer should be non-oriented polycrystalline. Here, the Fe-110 peak might be buried in

the background pattern of the Si substrate.

Figure 3(a) shows the cross-sectional high-angle annular dark field scanning transmission

electron microscopic (HAADF-STEM) image of the junction. A trilayered structure is certainly

formed, in other words, the UNCD/a-C:H layer structurally acts as an interlayer that clearly separates

the top Fe layer from the bottom Fe3Si layer. Figure 3(b) shows a bright-field scanning transmission

electron microscopic (BF-STEM) image of an Fe3Si/Si(111) interface in the junction. The interface

is not sharp in atomic scale, which might be because of atomic interdiffusion induced by the

bombardment of sputtering species and the HF pretreatment of the Si substrate was imperfectly made.

Fig. 2. XRD pattern of junction comprising Fe, B-doped UNCD/a-C:H, and Fe3Si

layers, measured in 2θ-θ.

(a) (b)

Fig. 3. (a) Cross-sectional HAADF-STEM image of trilayered junction and (b) BF-

STEM image of interface between Fe3Si bottom-layer and B-doped UNCD/a-C:H

interlayer in the junction.

011502-3JJAP Conf. Proc. , 011502 (2017) 5

From the electron diffraction pattern, the epitaxial growth of the Fe3Si layer on the Si(111) substrate

was confirmed, as expected from the XRD measurement results. This is consistent with results in our

previous research, wherein Fe3Si thin films are epitaxially grown on Si(111) substrate even at room

temperature [26]. The cross-sectional HAADF-STEM image of a magnified area around the interlayer of the

junction and the EDX profile of C, O, Si, and Fe in the depth direction are shown in Fig. 4(a) and

4(b), respectively. The STEM image of the interlayer can be divided into three areas. At the side

areas of A and B, O atoms preferentially exist, which means that the interfaces are oxidized due to temporal exposures to air for the replacements of the masks. Additionally, C, Si, and Fe atoms coexist in

the areas of A and B. This is because of the diffusions of Fe and Si atoms into the interlayer. Surprisingly,

the center area between the A and B areas hardly contains all atoms. The TEM observation sample was

prepared by a focused ion beam (FIB) apparatus. As a probable reason for it, we consider that the center

C-rich areas were preferentially etched away by Ga ions bombardment during the FIB process. Since the

interlayer exists in layer structure, the UNCD/a-C:H layer should have been present before the FIB etching.

The magnetization curve of the junction measured at room temperature is shown in Fig. 5. The

magnetization curve has clear two steps that indicate the antiparallel alignment formation of

(a)

(b)

Fig. 4. (a) Cross-sectional HAADF-STEM image and (b) EDX profiles of C, O, Si, and Fe corresponding

to the depth direction in the STEM image.

57.5

C

O

Si

Fe

0Distance (nm)

Fe3Si layer Fe layerAB

011502-4JJAP Conf. Proc. , 011502 (2017) 5

magnetizations owing to the difference in the coercive forces between the top Fe and the bottom

Fe3Si layers. The polycrystalline Fe layer has a larger coercive force than that of the epitaxial Fe3Si

layer [37]. The spin valve behavior is realized, which proves that the UNCD/a-C:H layer act as an

interlayer for the spin valve action.

4. Conclusion

Trilayered junctions comprising Fe, B-doped UNCD/a-C:H, and Fe3Si layers were prepared on

Si(111) substrates by physical vapor deposition combined with a mask method, and the film-

structures and magnetic properties were investigated. The TEM observation revealed that the

interlayer exists in layered structure and it contains oxidized layers at interfaces with ferromagnetic

layers due to the exposure to air for the replacements of the masks. From the shape of the

magnetization curve, the formation of antiparallel alignment of magnetizations owing to the

difference in the coercive forces between the top Fe and bottom Fe3Si layers were confirmed. It was

experimentally demonstrated that UNCD/a-C:H is applicable to Fe-Si system spin valves as

interlayer materials although the diffusion of Si and Fe atoms from the ferromagnetic layers into the

interlayer occurs in the range of several nanometers.

Acknowledgment

This work was partially supported by JSPS KAKENHI Grant Number JP16K14391, JP15K21594,

Yoshida Grant from Yoshida Academic Education Promotion Association, and Tohoku University

Advanced Characterization Nanotechnology Platform in Nanotechnology Platform Project

sponsored by the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.

References

[1] G. Binasch, P. Grunberg, F. Saurenbach, and W. Zinn, Phys. Rev. B 39, 4828 (1989).

[2] M. N. Baibich, J. M. Broto, A. Fert, F. Nguyen Van Dau, F. Petroff, P. Etienne, G. Creuzet, A. Friederich,

and J. Chazelas, Phys. Rev. Lett. 61, 2472 (1988).

[3] M. Julliere, Phys. Lett. 54A, 225 (1975).

[4] S. Maekawa, and U. Gafvert, IEEE Trans. Magn. 18, 707 (1982).

[5] T. Miyazaki and N. Tezuka, J. Magn. Magn. Mater. 139, L231 (1995).

[6] J. S. Moodera, L. R. Kinder, T. M. Wong, and R. Meservey, Phys. Rev. Lett. 74, 3273 (1995).

Fig. 5. Magnetization curve of junction, measured at room temperature.

–20 0 20

–1

0

1

Magnetic field (Oe)

M/M

s

UNCD (10 nm)

Fe3Si (100 nm)

Fe (100 nm)

011502-5JJAP Conf. Proc. , 011502 (2017) 5

[7] S. Yuasa, T. Nagahama, A. Fukushima, Y. Suzuki, and K, Ando, Nat. Mater. 3, 868 (2004).

[8] S. S. P. Parkin, C. Kaiser, A. Panchula, P. M. Rice, B. Hughes, M. Samant, and S. H. Yang, Nat. Mater. 3,

862 (2004).

[9] I. Appelbaum, B. Huang, and D. J. Monsma, Nature 447, 295 (2007).

[10] B. Huang, D. J. Monsma, and I. Appelbaum, Phys. Rev. Lett. 99, 177209 (2007).

[11] B. T. Jonker, G. Kioseoglou, A. T. Hanbicki, and C. H. Li, Nat. Phys. 3, 542 (2007).

[12] Y. Ando, K. Kasahara, K. Yamane, K. Hamaya, K. Sawano, T. Kimura, and M. Miyao, Appl. Phys.

Express 3, 093001 (2010).

[13] K. Kasahara, Y. Fujita, S. Yamada, K. Sawano, M. Miyao, and K. Hamaya, Appl. Phys. Express 7, 033002

(2014).

[14] X. Lou, C. Adelmann, S. A. Crooker, E. S. Garlid, J. Zhang, K. S. M. Reddry, S. D. Flexner, C. J.

Palmstorm, and P. A. Crowell, Nat. Phys. 3, 197 (2007).

[15] M. Ohishi, M. Shiraishi, R. Nouchi, T. Nozaki, T. Shinjo, and Y. Suzuki, Jpn. J. Appl. Phys. 46, L605

(2007).

[16] D. Huertas-Hernando, F. Guinea, and A. Brataas, Phys. Rev. Lett. 103, 146801 (2009).

[17] V. K. Dugaev, E. Y. Sherman, and J. Barna´s, Phys. Rev. B 83, 085306 (2011).

[18] T. Yoshitake, A. Nagano, M. Itakura, N. Kuwano, T. Hara, and K. Nagayama, Jpn. J. Appl. Phys. 46, L936

(2007).

[19] T. Yoshitake, Y. Nakagawa, A. Nagano, R. Ohtani, H. Setoyama, E. Kobayashi, K. Sumitani, Y. Agawa,

and K. Nagayama, Jpn. J. Appl. Phys. 49, 015503 (2010).

[20] S. Ohmagari, T. Yoshitake, A. Nagano, R. Ohtani, H. Setoyama, E. Kobayashi, T. Hara, and K. Nagayama,

Jpn. J. Appl. Phys. 49, 031302 (2010).

[21] S. Ohmagari and T. Yoshitake, Jpn. J. Appl. Phys. 51, 090123 (2012).

[22] Y. Katamune, S. Ohmagari, S. Al-Riyami, S. Takagi, M. Shaban, and T. Yoshitake, Jpn. J. Appl. Phys. 52,

065801 (2013).

[23] S. Bhattacharyya, O. Auciello, J. Birrell, J. A. Carlisle, L. A. Curtiss, A. N. Goyette, D. M. Gruen, A. R.

Krauss, J. Schlueter, A. Sumant, and P. Zapol, Appl. Phys. Lett. 79, 1441 (2001).

[24] C. Popov, W. Kulisch, S. Boycheva, K. Yamamoto, G. Ceccone, and Y. Koga, Diamond Relat. Mater. 13,

2017 (2004).

[25] T. Yoshitake, T. Ogawa, D. Nakagauchi, D. Hara, M. Itakura, N. Kuwano, and Y. Tomokiyo, Appl. Phys.

Lett. 89, 253110 (2006).

[26] K. Takeda, T. Yoshitake, D. Nakagauchi, T. Ogawa, D. Hara, M. Itakura, N. Kuwano, Y. Tomokiyo, T.

Kajiwara, and K. Nagayama, Jpn. J. Appl. Phys. 46, 7846 (2007).

[27] K. Takeda, T. Yoshitake, Y. Sakamoto, T. Ogawa, D. Hara, M. Itakura, N. Kuwano, T. Kajiwara, and K.

Nagayama, Appl. Phys. Express 1, 021302 (2008).

[28] S. Hirakawa, T. Sonoda, K. Sakai, K. Takeda, and T. Yoshitake, Jpn. J. Appl. Phys. 50, 08JD06 (2011).

[29] K. Sakai, T. Sonoda, S. Hirakawa, K. Takeda, and T. Yoshitake, Jpn. J. Appl. Phys. 51, 028004 (2012).

[30] K. Sakai, Y. Noda, K. Takeda, M. Takeda, and T. Yoshitake, Phys. Status Solidi C 10, 1862 (2013).

[31] K. Sakai, Y. Noda, D. Tsumagari, H. Deguchi, K. Takeda, and T. Yoshitake, Phys. Status Solidi A 211,

323 (2014).

[32] K. Sakai, Y. Noda, T. Daio, D. Tsumagari, A. Tominaga, K. Takeda, and T. Yoshitake, Jpn. J. Appl. Phys.

53, 02BC15 (2014).

[33] K. Sakai, Y. Asai, M. Takeda, K. Ishibashi, Y. Noda, K. Takeda, and T. Yoshitake, JJAP Conf. Proc. 3,

011503 (2015).

[34] K. Sakai, Y. Asai, Y. Noda, T. Daio, A. Tominaga, K. Takeda, and T. Yoshitake, JJAP Conf. Proc. 3,

011502 (2015).

[35] Y. Asai, K. Sakai, K. Ishibashi, K. Takeda, and T. Yoshitake, JJAP Conf. Proc. 3, 011501 (2015).

[36] Y. Asai, K. Sakai, K. Ishibashi, K. Takeda, and T. Yoshitake, JJAP Conf. Proc. 3, 011504 (2015).

[37] K. Ishibashi, K. Nakashima, K. Sakai, and T. Yoshitake, APS Conf. Proc., 2015, GT1.00151.

011502-6JJAP Conf. Proc. , 011502 (2017) 5