This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
IEEE TRANSACTIONS ON MAGNETICS 1
Local Magnetoresistance at Room Temperature
in Si
100 Devices
Mizue Ishikawa
1,2, Makoto Tsukahara
1, Michihiro Yamada
1, Yoshiaki Saito
2,3, and Kohei Hamaya
1,4 1Department of Systems Innovation, Graduate School of Engineering Science, Osaka University, Suita 560-8531, Japan2Corporate Research and Development Center, Toshiba Corporation, Kawasaki 212-8582, Japan 3Center for Innovative Integrated Electronic Systems, Tohoku University, Sendai 980-0845, Japan
4Center for Spintronics Research Network, Graduate School of Engineering Science, Osaka University, Suita 560-8531, Japan
We show that there is a crystal orientation effect on the two-terminal local magnetoresistance (MR) in silicon (Si)-based lateral spin-valve (LSV) devices. When we compare the local MR effect between Si100 and Si110 LSV devices, the magnitude of the local MR signals for Si100 LSV devices is always larger than that for Si110 LSV devices. For Si100 LSV devices, the magnitude of the room-temperature MR ratio reaches approximately 0.06%. We infer that it is important to consider the tunneling anisotropic spin polarization, which is due to the magnetization direction of the ferromagnetic contacts relative to the Si crystal orientation, in the fabricated LSV devices.
Index Terms— Silicon (Si) spintronics, spin detection, spin injection.
I. INTRODUCTION
S
PINTRONIC technologies are expected to markedly improve device performances because of its nonvolatility, reconstructibility, and low power consumption [1]–[3]. For semiconductor spintronics devices [4], [5], highly efficient electrical spin injection and detection in semiconductors are the most important technologies. In general, four-terminal nonlocal (NL) spin-transport measurements in ferromagnet-semiconductor lateral spin-valve (LSV) devices have been utilized for exploring the spin injection/detection in semiconductors [6]–[17]. For spin-based logic/memory applications [4], [5], [18], [19], however, it is further important to realize the high magnetoresistance (MR) ratio obtained by two-terminal local measurements through the semiconductor channel at room temperature. Although there were several studies of observation of the local-MR effect through silicon (Si) [20]–[26] and germanium (Ge) [27], [28], the values of the local-MR ratio at room temperature were∼0.03% [24] for Si and ∼0.01% [28] for Ge. To increase
the MR ratio at room temperature in Si-LSV devices, it is necessary to explore the efficient detection of the two-terminal local spin signals.
So far, by changing the cross-sectional area of the spin-transport channel [29] and the size of the spin-injector contact [30], large spin accumulation signals detected in four-terminal NL spin-transport devices have been reported for Si. Very recently, we experimentally found the relatively efficient detection of the NL spin signals in LSV devices with the Si spin-transport channel along100 (Si100) compared to those along 110 (Si110) [31]. Since there was no Manuscript received March 8, 2018; revised May 25, 2018 and June 12, 2018; accepted June 19, 2018. Corresponding author: K. Hamaya (e-mail: [email protected]).
Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TMAG.2018.2849753
difference in the spin-diffusion length and spin lifetime between Si100 and Si110 LSV devices, we tentatively considered that the spin injection/detection efficiency depends on the crystallographic relationship between the magnetization direction of the ferromagnetic contacts and the Si crystal orientation in the used Si LSV devices [31]. However, the crystal orientation effect on the local spin signals has never been investigated in Si-based devices.
In this paper, we study the two-terminal local signals for two kinds of LSV devices consisting of the Si100 or Si110 spin-transport channel with a small size (0.305 μm2) cross section. Relatively large local spin signals can be observed for Si100 LSV devices compared to those for Si110 ones, causing the room-temperature MR ratio to become twice as large as that in the previous work [24]. We infer that it is important to consider the tunneling anisotropic spin polarization, which is due to the magnetization direction of the ferromagnetic contacts relative to the Si crystal axis, in the fabricated LSV devices.
II. LATERALSPIN-VALVEDEVICES
To prepare the same interface condition for electrical spin injection at the CoFe–MgO/Si heterostructure, we fabricated the two kinds of LSVs along the Si100 and Si110 with CoFe–MgO electrodes on a single (100)-silicon-on-insulator (SOI) (∼61 nm) wafer with a phosphorus doping (n ∼ 1.3 × 1019 cm−3), as shown in Fig. 1(a) and (b). An MgO (1.1 nm) tunnel barrier was deposited at 200 ◦C by electron beam evaporation, and then, a CoFe (10 nm) and a Ru capping layer were sputtered on top of it. Conventional processes with electron beam lithography, Ar+ ion milling, and reactive-ion etching were used to fabricate the LSV devices [29]. The sizes of the Ru–CoFe–MgO contacts were 2.0 × 5.0 μm2 and 0.5 × 5.0 μm2, and the channel width was 7.0 μm. An optical micrograph of one of the fabricated LSV 0018-9464 © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
2 IEEE TRANSACTIONS ON MAGNETICS
Fig. 1. (Color online) (a) Schematic of an LSV device with Si spin-transport channel along Si100 or Si110. (b) Relationship between the crystal orientations, 100 or 110, and the fabricated Si spin-transport channels. (c) Optical micrograph of a CoFe-MgO/Si LSV. The terminal configurations for the local voltage change (VL) and I –Vint measurements are shown. (d) and (e) I –Vintcharacteristics for all the CoFe–MgO/Si interfaces at 20 K and room temperature (303 K), respectively.
devices for two-terminal local-MR measurements is shown in Fig. 1(c). It should be noted that there is no difference in the size of the spin-injector contact between the Si100 and Si110 LSVs [30]. Finally, ohmic pads consisting of Au/Ti were formed for all contacts. We have checked that the resistivity and the Hall mobility of the Si spin-transport layer were almost the same between Si100 and Si110 Hall-bar devices.
III. RESULTS ANDDISCUSSION
Fig. 1(d) and (e) show the I –Vint characteristics for all CoFe–MgO/Si interfaces at 20 K and room temperature (303 K), respectively, where Vint is the applied voltage to the CoFe–MgO/Si interface, and the I –Vint curves can be measured by a three-terminal configuration in Fig. 1(c). Each interface shows the tunnel conduction and the I –Vintbehavior for both Si100 and Si110 LSV devices are almost the same, indicating the same interface condition at CoFe–MgO/Si
Fig. 2. (Color online) Two-terminal local-MR signals for (a) Si100 and (b) Si110 LSVs at a bias current of +0.2 mA at 20 K. (c) Schematics of a rough interpretation for the magnetization switching process of the CoFe contacts in the Si100 LSV device.
heterostructure. Thus, the quality of the spin injector and detector in both LSV devices is equivalent to explore the effect of the crystal orientation of the spin-transport channel on the local MR.
Fig. 2(a) and (b) show the two-terminal local spin signals (VL/I = RL) for Si100 and Si110 LSV devices,
respectively, at a bias current of +0.2 mA at 20 K. Here, we have already confirmed the reliable lateral spin transport detected by NL-MR and NL-Hanle measurements in the same LSV devices, similar to the data in [31]. It should be noted that the magnitude of the local signal,|RL|, for the Si100
LSV device is larger than that for the Si110 one. For Si100 LSV devices, it seems that the antiparallel magnetization state between the spin injector and detector is not stable for the application of the in-plane magnetic fields, By. Because
there is the magnetocrystalline anisotropy of the CoFe layer (easy axis: CoFe100 and hard axis: CoFe110), we should consider the correlation between the magnetization state in the CoFe contacts and By. For Si100 LSV devices, By along
Si100 ([010]) can contribute to the magnetization rotation, because the direction of By is parallel to the magnetic hard
axes in the CoFe layer (CoFe110) [31].
We roughly explain the hysteretic behavior in Fig. 2(a) in the following. When the direction of By is switched from
negative to positive, the magnetization reversal in the wider CoFe contact occurs, but the magnetization direction is pinned along a certain direction between Si110 and the direction of
By (Si 100), as shown in Fig. 2(c-1). Then, with increasing
By, the magnetization rotation in the wider CoFe contact
can be induced [see Fig. 2(c-2)], leading to the antiparallel magnetization states. Finally, as By is further increased,
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
ISHIKAWA et al.: LOCAL MR AT ROOM TEMPERATURE IN Si100 DEVICES 3
the magnetization of the narrower CoFe contact can switch toward the direction of By, as shown in Fig. 2(c-3), causing
the parallel magnetization state. Considering these magneti-zation reversal processes of the CoFe contacts in the Si100 LSV device, we can roughly interpret the hysteretic behavior in Fig. 2(a). On the other hand, the magnetization switching of the CoFe contacts for Si110 LSV devices can occur just along Si110 parallel to the easy axis of the CoFe epilayer on Si. However, the difference in the magnetization switching fields between Si100 and Si110 LSV devices cannot be understood yet. Because these measurements were conducted within±100 mT, we can expect that the magnetization of the CoFe contacts was not fully saturated. Thus, to compare the magnetization reversal process of the CoFe contacts between Si100 and Si110 LSV devices, we probably should take in account the complicated domain wall nucleation and prop-agation in the CoFe contacts fabricated epitaxially on Si. This feature was also observed in the NL spin-transport measurements in our previous work [31].
In Fig. 3, bias-voltage (Vbias) dependence of |RL| is
examined at 20 K, where the inset shows the two-terminal
I –Vbias characteristics through the CoFe–MgO/Si/MgO–CoFe double barrier structures. Despite nearly the same I –Vbias behavior between Si100 and Si110 LSV devices and the almost symmetric characteristics with respect to Vbias = 0, we can clearly observe the difference in the Vbiasdependence of |RL| at 20 K, and the value of |RL| in the Si100
LSV device is markedly larger than that in the Si110 one in Vbias > 0. These are the first observations of the crystal orientation effect on the local spin signals in Si-based LSV devices.
Even at room temperature (303 K), the enhancement in the local spin signals can be observed in Si100 LSV devices, as shown in Fig. 4(a)–(d), although the noise level is relatively large compared to the low-temperature measurements. Also, the difference in the magnetization switching between Si100 and Si110 LSV devices becomes obscure. As can be seen, a large|RL| of ∼2 at room temperature can be obtained
for Si100 LSV devices in both positive and negative Vbias conditions. Note that the value of |RL| is larger than that
by the Kyoto group [24]. Namely, this is the largest |RL|
value reported so far. In addition, the local-MR ratio of approximately 0.06% estimated here is twice as large as that observed so far [24]. Although the value of the local-MR ratio is insufficient to realize the spintronics devices, it is important to consider the crystal orientation effect on the local-MR signals in Si-based LSV devices.
In [31], the spin-diffusion length and spin lifetime between Si100 and Si110 LSV devices were comparable. Thus, we now infer that the observed features are attributed to the difference in the spin injection/detection efficiency depending on the magnetization direction of the ferromagnetic contacts relative to the crystal orientation of Si [31]. As shown in Fig. 1(b), the magnetization switching of the CoFe layers for Si100 LSV devices occurs nearly along the direction parallel to Si100, whereas that for Si110 LSV devices occurs along the direction parallel to the Si110. Considering these differences, we speculate that the observed crystal
Fig. 3. (Color online) Vbiasdependence of|RL| at 20 K for Si100 and Si110. The inset shows the I–Vbiascharacteristics for Si100 and Si110 at 20 K.
Fig. 4. (Color online) Two-terminal local-MR signals for (a) and (c) Si100 and (b) and (d) Si110 LSVs at +0.5 mA (Vbias = 2.3 V) and −0.5 mA (Vbias= −2.2 V), respectively, at room temperature.
orientation effect is able to be understood by the presence of the tunneling anisotropic spin polarization, which was discov-ered in (Ga, Mn)As/GaAs LSV devices [32], in the CoFe/MgO tunnel contacts. Even in a single device, the anisotropy of the tunneling spin polarization can be observed [32]–[34] and is related to the crystallographic relationship between the magnetization direction of the ferromagnetic contact and the
This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
4 IEEE TRANSACTIONS ON MAGNETICS
crystal orientation of semiconductors. Further investigations should be performed for understanding this crystal orientation effect in Si-based LSV devices.
IV. CONCLUSION
We have studied the local spin signals for two kinds of LSV devices consisting of the Si100 or Si110 spin-transport channel with the epitaxial CoFe/MgO electrodes. We found that the local spin signals in Si100 LSV devices were always larger than those in Si110 ones. The largest MR of 0.06% was observed in a Si100 LSV device at room temperature. For local MR effect in Si-based LSV devices, it may be important to consider the tunneling anisotropic spin polarization, which is due to the magnetization direction of the ferromagnetic contacts relative to the crystal orientation of Si.
ACKNOWLEDGMENT
This work was supported in part by the Japan Society for the Promotion of Science through a Grant-in-Aid for Scientific Research (A) (16H02333), and in part by the Ministry of Education, Culture, Sports, Science, and Technology through a Grant-in-Aid for Scientific Research on Innovative Areas, Nano Spin Conversion Science (26103003).
REFERENCES
[1] I. Žuti´c, J. Fabian, and S. D. Sarma, “Spintronics: Fundamentals and applications,” Rev. Mod. Phys., vol. 76, no. 2, p. 323, 2004.
[2] S. Yuasa and D. D. Djayaprawira, “Giant tunnel magnetoresistance in magnetic tunnel junctions with a crystalline MgO(001) barrier,”
J. Phys. D, Appl. Phys., vol. 40, p. R337, Nov. 2007.
[3] A. Hirohata and K. Takanashi, “Future perspectives for spintronic devices,” J. Phys. D, Appl. Phys., vol. 47, p. 193001, Apr. 2014. [4] H. Dery, P. Dalal, L. Cywi´nski, and L. J. Sham, “Spin-based logic in
semiconductors for reconfigurable large-scale circuits,” Nature, vol. 447, pp. 573–576, May 2007.
[5] M. Tanaka and S. Sugahara, “MOS-based spin devices for reconfigurable logic,” IEEE Trans. Electron Devices, vol. 54, no. 5, pp. 961–976, May 2007.
[6] X. Lou et al., “Electrical detection of spin transport in lateral ferromagnet–semiconductor devices,” Nature Phys., vol. 3, pp. 197–202, Feb. 2007.
[7] T. Uemura, T. Akiho, M. Harada, K.-I. Matsuda, and M. Yamamoto, “Non-local detection of spin-polarized electrons at room temperature in Co50Fe50/GaAs Schottky tunnel junctions,” Appl. Phys. Lett., vol. 99, p. 082108, Aug. 2011.
[8] T. Saito, N. Tezuka, M. Matsuura, and S. Sugimoto, “Spin injection, transport, and detection at room temperature in a lateral spin transport device with Co2FeAl0.5Si0.5/n-GaAs Schottky tunnel junctions,” Appl.
Phys. Express, vol. 6, p. 103006, Oct. 2013.
[9] O. M. J. van’t Erve et al., “Electrical injection and detection of spin-polarized carriers in silicon in a lateral transport geometry,” Appl. Phys.
Lett., vol. 91, p. 212109, Nov. 2007.
[10] S. Toshio, S. Tomoyuki, O. Tohru, S. Masashi, S. Yoshishige, and N. Kiyoshi, “Room-temperature electron spin transport in a highly doped Si channel,” Appl. Phys. Express, vol. 4, no. 2, p. 023003, 2011. [11] Y. Saito et al., “Spin-based MOSFETs for logic and memory applications
and spin accumulation signals in CoFe/tunnel barrier/SOI devices,” IEEE
Trans. Magn., vol. 48, no. 11, pp. 2739–2745, Nov. 2012.
[12] Y. Zhou et al., “Electrical spin injection and transport in germa-nium,” Phys. Rev. B, Condens. Matter, vol. 84, no. 12, p. 125323, Sep. 2011.
[13] Y. Fujita, M. Yamada, S. Yamada, T. Kanashima, K. Sawano, and K. Hamaya, “Temperature-independent spin relaxation in heavily doped
n-type germanium,” Phys. Rev. B, Condens. Matter, vol. 94, p. 245302,
Dec. 2016.
[14] A. Bhattacharya, M. Z. Baten, and P. Bhattacharya, “Electrical spin injection and detection of spin precession in room temperature bulk GaN lateral spin valves,” Appl. Phys. Lett., vol. 108, p. 042406, Jan. 2016. [15] Y. Fujita et al., “Spin transport and relaxation up to 250 k in heavily
doped n-type Ge detected using Co2FeAl0.5Si0.5electrodes,” Phys. Rev.
Appl., vol. 8, p. 014007, Jul. 2017.
[16] M. Yamada, Y. Fujita, M. Tsukahara, S. Yamada, K. Sawano, and K. Hamaya, “Large impact of impurity concentration on spin transport in degenerate n-Ge,” Phys. Rev. B, Condens. Matter, vol. 95, p. 161304(R), Apr. 2017.
[17] M. Yamada et al., “Room-temperature spin transport in n-Ge probed by four-terminal nonlocal measurements,” Appl. Phys. Express, vol. 10, p. 093001, Aug. 2017.
[18] Y. Saito, T. Inokuchi, M. Ishikawa, H. Sugiyama, T. Marukame, and T. Tanamoto, “Spin-based MOSFET and its applications,” J.
Elec-trochem. Soc., vol. 158, no. 10, pp. H1068–H1076, 2011.
[19] T. Tanamoto et al., “Scalability of spin field programmable gate array: A reconfigurable architecture based on spin metal-oxide-semiconductor field effect transistor,” J. Appl. Phys., vol. 109, p. 07C312, Mar. 2011. [20] I. Appelbaum, B. Huang, and D. J. Monsma, “Electronic measurement
and control of spin transport in silicon,” Nature, vol. 447, pp. 295–298, May 2007.
[21] T. Sasaki, T. Oikawa, T. Suzuki, M. Shiraishi, Y. Suzuki, and K. Noguchi, “Local and non-local magnetoresistance with spin preces-sion in highly doped Si,” Appl. Phys. Lett., vol. 98, p. 262503, Jun. 2011. [22] Y. Saito et al., “Local magnetoresistance through Si and its bias voltage dependence in ferromagnet/MgO/silicon-on-insulator lateral spin valves,” J. Appl. Phys., vol. 115, p. 17C514, Feb. 2014.
[23] T. Sasaki et al., “Spin transport in nondegenerate SI with a spin MOSFET structure at room temperature,” Phys. Rev. Appl., vol. 2, no. 3, p. 034005, Sep. 2014.
[24] T. Sasaki et al., “Local magnetoresistance in Fe/MgO/Si lateral spin valve at room temperature,” Appl. Phys. Lett., vol. 104, no. 5, p. 052404, Feb. 2014.
[25] T. Tahara et al., “Observation of large spin accumulation voltages in nondegenerate Si spin devices due to spin drift effect: Experiments and theory,” Phys. Rev. B, Condens. Matter, vol. 93, no. 21, p. 214406, Jun. 2016.
[26] Y. Saito, T. Inokuchi, M. Ishikawa, T. Ajay, and H. Sugiyama, “Spin accumulation and transport signals in CoFe/MgO/Si devices with con-fined structure of n+-Si layer,” AIP Adv., vol. 7, p. 055937, Mar. 2017. [27] P. Li, J. Li, L. Qing, H. Dery, and I. Appelbaum, “Anisotropy-driven spin relaxation in germanium,” Phys. Rev. Lett., vol. 111, p. 257204, Dec. 2013.
[28] M. Kawano et al., “Electrical detection of spin accumulation and relaxation in p-type germanium,” Phys. Rev. Mater., vol. 1, p. 034604, Aug. 2017.
[29] M. Ishikawa, T. Oka, Y. Fujita, H. Sugiyama, Y. Saito, and K. Hamaya, “Spin relaxation through lateral spin transport in heavily doped n-type silicon,” Phys. Rev. B, Condens. Matter, vol. 95, p. 115302, Mar. 2017.
[30] A. Spiesser et al., “Giant spin accumulation in silicon nonlocal spin-transport devices,” Phys. Rev. Appl., vol. 8, p. 064023, Dec. 2017. [31] M. Ishikawa et al. (2017). “Efficient spin transport along Si100 at
room temperature.” [Online]. Available: https://arxiv.org/abs/1712.07789 [32] A. Einwanger, M. Ciorga, U. Wurstbauer, D. Schuh, W. Wegscheider, and D. Weiss, “Tunneling anisotropic spin polarization in lateral (Ga,Mn)As/GaAs spin Esaki diode devices,” Appl. Phys. Lett., vol. 95, p. 152101, Oct. 2009.
[33] S. Sharma, S. P. Dash, H. Saito, S. Yuasa, B. J. van Wees, and R. Jansen, “Anisotropy of spin polarization and spin accumulation in Si/Al2O3/ferromagnet tunnel devices,” Phys. Rev. B, Condens. Matter, vol. 86, p. 165308, Oct. 2012.
[34] S. Sharma, A. Spiesser, H. Saito, S. Yuasa, B. J. van Wees, and R. Jansen, “Crystal-induced anisotropy of spin accumulation in Si/MgO/Fe and Si/Al2O3/ferromagnet tunnel devices,” Phys. Rev. B,
