Room-temperature spin transport in InAs nanowire lateral spin valve

Abstract

With its strong spin–orbital interaction, indium arsenide (InAs) is a potential candidate material for spintronic devices. Owing to the reduced scale and quasi one-dimensional confinement, InAs nanowires are expected to show novel physics property compared with InAs two-dimensional electron gas. In this work, we report room-temperature local measurement of spin transport in an InAs nanowire synthesized from the bottom-up paradigm by molecular beam epitaxy. Due to the advantage of Co/MgO barrier contact, a large spin signal of up to 35 kΩ is observed, and long spin diffusion length (about 1.9 μm) in the unintentionally doped InAs nanowire is achieved by analysing the spin signal using the transfer matrix method.

---

1. Introduction

Electrical spin injection into semiconductors is fundamental to realize spintronic devices using the electron spin degree of freedom instead of, or in addition to, the charge.¹ Among many spin-based device concepts,^2–6 the Datta–Das spin field-effect transistor (spin-FET)⁴ is the most classical scheme, comprising a ferromagnetic spin injector and detector, while the injected spins in the semiconductor channel are manipulated by a gate electric field. Due to the high conductance mismatch between the ferromagnetic metal (FM) and semiconductor (SC), however, spin polarization of the injected current across the FM/SC interface has been predicted theoretically less than 0.1%⁷ and has been verified to be slightly higher at 1.3%.⁸

Recently, there has been considerable progress in achieving successful spin injection into semiconductors by employing Schottky^9,10 or oxide tunnel contacts,^11–15 an effective solution to the conductance mismatch problem proposed by Rashba.¹⁶ All-electrical spin injection and detection have been achieved in a variety of materials such as silicon (Si),^12,17,18 indium arsenide (InAs),^2,19 germanium (Ge),^20,21 gallium arsenide (GaAs),^14,15,22 doped-SrTiO₃,²³ graphene^11,24,25 and carbon nanotube.²⁶ Semiconductor nanowires (NWs) are ideal building blocks for future spin-based integrated circuits owing to their quasi one-dimensional confined channel and the bottom-up growth mode, avoiding complex fabrication steps. However, only a few works related to spin injection in semiconductor NWs, e.g., gallium nitride (GaN),²⁷ indium nitride (InN),²⁸ indium antimonide (InSb),²⁹ Ge,³⁰ and Si^17,31,32 have been reported. Because most of them are of weak spin–orbital interaction, research on NWs with strong spin–orbital interaction is essential for achieving the spin-FET. III–V compound semiconductors, such as InSb and InAs, have received a great deal of interest so far due to their high electron mobility and narrow band gap.^33–35 However, there has been no publication associated with spin injection into InAs NW to date. Also, it is reported that spin relaxation in III–V semiconductors can be significantly suppressed by reducing the channel width.³⁶ Owing to the reduced scale and quasi one-dimensional confinement, InAs NWs are expected to show novel physics property (e.g. longer spin diffusion length, l_sf) compared with InAs two-dimensional electron gas (2DEG) for future nanoscale spintronic devices. Nonlocal measurement is reported as a convincing method for effective detection of the spin signal ΔR (the difference in resistance measured when the magnetizations of FM electrodes are parallel and antiparallel), while local measurement is commonly considered unreliable for proving spin injection.³¹ However, the ΔR obtained by nonlocal measurement can hardly exceed a few hundred ohms; it is generally a few tens of ohms or less. Aiming at future application of spintronic devices, both a large ΔR and a large magnetoresistance (MR, the spin signal normalized by the device resistance) are desirable to discriminate them from background and spurious signals and to circumvent unpractical four-terminal device geometries.

Here, we show our measurements of efficient spin information transport and large spin signals in InAs NW-based lateral spin valve geometry with Co/MgO contact at room temperature.

2. Experiment

The InAs NWs under investigation were grown in a catalyst-free manner on n-type Si (111) substrates by molecular-beam epitaxy. The growth was performed at a temperature of 505 °C, with a V/III beam equivalent pressure ratio of 120 and growth time of 120 min. The NW morphology and crystal structure were characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). As shown in Fig. 1(a), the InAs NWs were vertically grown on the Si (111) substrate, and the diameter varied from ∼40 to 120 nm. As shown in Fig. 1(b), high-resolution TEM image of the InAs NW indicates that the NW has a mixture of wurtzite and zinc-blende crystal structure. After the growth, the NWs were mechanically transferred on a p-type Si (100) substrate covered with a 300 nm-thick SiO₂ layer. For each NW, two contact regions were defined using electron beam lithography after locating individual wires. Prior to the deposition of electrodes, the sample was etched in buffered hydrofluoric acid to remove the native oxide layer on the NW, followed by a (NH₄)₂S_x solution passivation to avoid any re-oxidation of the contact area. For typical back-gate measurement, Cr/Pt (20 nm/60 nm) electrodes were deposited by magnetron sputtering and a subsequent lift-off process, as shown in Fig. 1(c). For spin transport measurement, however, 1.5 nm MgO oxide barrier, 100 nm Co and 20 nm Au capping layer were fabricated by electron beam evaporation (Fig. 1(d)). Finally, this sample was annealed at 350 °C for 30 minutes. It should be noted that the width of the two FM electrodes is 200 nm and 500 nm, respectively, and the channel length L is 650 nm.

Fig. 1 (a) Side view SEM image of catalyst-free InAs NWs grown on an n-type Si(111) substrate by molecular-beam epitaxy. (b) High-resolution TEM image of a single InAs NW. (c) Back-gate InAs NW device with two Cr/Pt electrodes. (d) Schematic diagram of an InAs NW two-terminal device.

3. Results and discussion

Fig. 2(a) and (b) show the typically measured room-temperature transfer and output curves of an InAs NW transistor, indicating the stable FET characteristics. We calculated the field effect mobility as³⁷

μ_FE = g_mL_G²/CV_ds (1)

where g_m is the peak transconductance obtained from the slope of I_ds vs. V_g in the linear region, L_G is the gate length, V_ds is the source-drain voltage, ε is the insulator dielectric constant, t_ox is the gate insulator thickness (here the thickness of SiO₂ layer), a is the nanowire radius, and C is gate capacitance, which can be reduced to when t_ox ≫ a. We then calculated the μ_FE to be 380 cm² V⁻¹ s⁻¹; this small value may be due to the large contact resistance induced by the residual oxide layer on the surface of InAs NW. Using the expression , we obtain the two-dimensional carrier concentration of 9.4 × 10¹¹ cm⁻², comparable to the previously reported value of InAs NW devices with similar crystal structure.³⁸ The inset of Fig. 2(a) shows that the I_on/I_off ratio is over 10³. It should be noted that the nanowire device exhibits hysteresis in the transfer curve, which can be attributed to the large number of traps in the InAs NW surface amorphous layer.³⁹

Fig. 2 (a) Room-temperature transfer curve of an InAs NW back-gate device at V_ds = 10 mV. The inset is the logarithmic I_ds vs. V_g. (b) Output curves of the same device at different gate voltages.

We then performed the local measurement on the two-terminal devices with two ferromagnetic electrodes. The resistance (R) was measured as a function of external magnetic field (H) while applying a fixed injection current (150 nA) between the two electrodes. Direction of the applied magnetic field was along the easy axes of the electrodes (parallel to the long axes of FM electrodes). Fig. 3(a) shows the measured R–H curve. While sweeping the magnetic field between ±1000 Oe, we can observe pronounced resistance change. At H = 1000 Oe, the magnetization of two electrodes are both aligned along the positive magnetic field direction. When the field decreases to ∼−200 Oe, the magnetization of the wider electrode changes its direction to the negative magnetic field direction and becomes antiparallel to the narrower one, causing an increase of resistance. With further decreasing H to ∼−600 Oe, the magnetization of the narrower electrode also changes its direction to the negative magnetic field direction and again becomes parallel to the wider electrode, causing a decrease in resistance. Sweeping the field reversely results in a similar procedure. Anisotropy magnetoresistance (AMR) curves of the two electrodes were carried out in order to estimate their coercive fields. From Fig. 4(a) and (b), we can clearly see that the switching fields are about 230 Oe and 540 Oe for the wider and narrower FM electrodes, respectively, which agree with the R–H curves in Fig. 3(a).

Fig. 3 (a) Resistance vs. in-plane magnetic field, measured in the two-terminal configuration. The blue (red) curve corresponds to the positive (negative) sweep direction. (b) Two-terminal I–V characteristics of the device at room temperature. The solid line serves as a visual guide.

Fig. 4 AMR measurement of single Co wires with the width of (a) 500 nm and (b) 200 nm.

Fig. 3(b) shows the I–V curve of this device. The nonlinear I–V behavior indicates the tunneling nature of the contacts. The overall resistance of the two-terminal device is much larger than that of the InAs channel (a few kΩ), proving that the transport behavior is dominated by the tunnel barrier.

When the magnetizations of the two FM electrodes are in parallel alignment, the device resistance (R_p) is ∼115 kΩ. The resistance difference between parallel and antiparallel states (ΔR) is ∼35 kΩ, a rather large spin signal, almost two orders of magnitude greater than previously reported spin injection devices. This large spin signal can hardly be explained by spurious effects such as AMR, local Hall effect (LHE)⁴⁰ or magneto-Coulomb effect (MCE).⁴¹ As in Fig. 3(a) and (b), the signals of AMR for the two FM electrodes are all quite small and less than 1 Ω. The LHE signal is generally induced by the strong fringe field near the edge of an FM electrode in a nanoscaled semiconducting system. In our case, the LHE signal should be significantly reduced because the InAs NW is far away from the edge of two FM electrodes. Besides, MCE should play a role in two-terminal devices only at low temperature rather than room temperature. Also considering the fact that all the signals originated from these spurious effects generally cannot exceed 1 kΩ, which are far below the observed signal ΔR, then we attribute the large signal to spin transport instead of from spurious effects.

Our experiment results can be analyzed by employing eqn (3)⁴²

with R_b = R_p/2(1 − γ²) and R^s_ch = ρl_sf/S, in the limit where R_F ≪ R_b. Here, R^s_ch and R_b are the spin resistances of the channel and contact resistance, respectively, γ the spin asymmetry for a magnetic material, L the distance between the two FM electrodes, and l_sf the spin diffusion length (SDL). Firstly, the SDL of the InAs NW can be derived by eqn (3). There are only two free parameters, l_sf and γ. Here we assume γ to be 55%, slightly larger compared with the experimental value of Co in literature (0.47 ± 0.02).⁴³ This is reasonable if we take into account the calculated value of spin polarization of Co (∼78%). We could get a γ closer to the calculated value due to the tunneling nature of the Co/MgO barrier. Note that larger values of γ would lead to shorter l_sf, and then the calculated l_sf of 1.9 μm should be the lower limit of SDL in the InAs NW, slightly larger than the reported value in an InAs quantum well (1.6 μm).⁴⁴ This can be understood by the fact that the SDL in the quasi one-dimensional confinement channel is larger due to the suppressed spin relaxation compared with 2DEG.³⁶

In Fig. 5, we show the simulated R_b dependence of ΔR using eqn (4) with typical parameters of our device (L = 650 nm, ρ = 0.01 Ω cm), γ = 0.55 for Co/MgO tunnel barrier and different l_sf. It is found that the observation of spin signal ΔR above 10 kΩ requires not only very large l_sf (typically above 1 μm) but also a high contact resistance (>30 kΩ). If both conditions are not satisfied, ΔR is limited to only a few hundred ohms.

Fig. 5 Calculated spin signal ΔR vs. tunnel barrier resistance R_b by using eqn (4) with different l_sf.

4. Conclusions

In summary, we have fabricated the two-terminal InAs NW-based lateral spin valve with Co/MgO tunnel contacts. The R–H measurement shows a large spin signal of up to 35 kΩ and MR of 30.4% at room temperature. The SDL in the unintentionally doped InAs NW is calculated to be 1.9 μm. We believe that these results open a practical way to design InAs NW-based spintronic devices to realize more reliable, faster and less energy-consuming information processing.

Acknowledgements

This work was supported by the National Basic Research Program of China (Grant No. 2012CB932702, 2015CB921502), the National Science Foundation of China (Grant No. 51371024, 51325101, 51271020, 51471029, 61504133, 11304381).

Notes and references

1. I. Žutić, J. Fabian and S. D. Sarma, Rev. Mod. Phys., 2004, 76, 323.
2. H. C. Koo, J. H. Kwon, J. Eom, J. Chang, S. H. Han and M. Johnson, Science, 2009, 325, 1515–1518.
3. M. Tanaka and S. Sugahara, IEEE Trans. Electron Devices, 2007, 54, 961–976.
4. S. Datta and B. Das, Appl. Phys. Lett., 1990, 56, 665–667.
5. T. Schäpers, J. Nitta, H. Heersche and H. Takayanagi, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 64, 125314.
6. A. Fert, J. M. George, H. Jaffres and R. Mattana, IEEE Trans. Electron Devices, 2007, 54, 921–932.
7. G. Schmidt, D. Ferrand, L. Molenkamp, A. Filip and B. Van Wees, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 62, R4790.
8. W. Han, K. Pi, W. Bao, K. McCreary, Y. Li, W. Wang, C. Lau and R. Kawakami, Appl. Phys. Lett., 2009, 94, 222109.
9. Y. Ando, K. Kasahara, S. Yamada, Y. Maeda, K. Masaki, Y. Hoshi, K. Sawano, M. Miyao and K. Hamaya, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 85, 035320.
10. A. Spiesser, H. Saito, R. Jansen, S. Yuasa and K. Ando, Phys. Rev. B: Condens. Matter Mater. Phys., 2014, 90, 205213.
11. W. Han, K. Pi, K. McCreary, Y. Li, J. J. Wong, A. Swartz and R. Kawakami, Phys. Rev. Lett., 2010, 105, 167202.
12. B. T. Jonker, G. Kioseoglou, A. T. Hanbicki, C. H. Li and P. E. Thompson, Nat. Phys., 2007, 3, 542–546.
13. K. R. Jeon, B. C. Min, Y. H. Jo, H. S. Lee, I. J. Shin, C. Y. Park, S. Y. Park and S. C. Shin, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 84, 165315.
14. S. Liang, T. Zhang, P. Barate, J. Frougier, M. Vidal, P. Renucci, B. Xu, H. Jaffres, J.-M. George and X. Devaux, Phys. Rev. B: Condens. Matter Mater. Phys., 2014, 90, 085310.
15. S. H. Shim, H. J. Kim, H. C. Koo, Y. H. Lee and J. Chang, Appl. Phys. Lett., 2015, 107, 102407.
16. E. Rashba, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 62, R16267.
17. O. Van'tErve, A. Friedman, E. Cobas, C. Li, J. Robinson and B. Jonker, Nat. Nanotechnol., 2012, 7, 737–742.
18. S. P. Dash, S. Sharma, R. S. Patel, M. P. de Jong and R. Jansen, Nature, 2009, 462, 491–494.
19. W. Y. Choi, H. J. Kim, J. Chang, S. H. Han, H. C. Koo and M. Johnson, Nat. Nanotechnol., 2015, 10, 666–670.
20. Y. Zhou, W. Han, L. T. Chang, F. Xiu, M. Wang, M. Oehme, I. A. Fischer, J. Schulze, R. K. Kawakami and K. L. Wang, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 84, 125323.
21. K. Kasahara, Y. Fujita, S. Yamada, K. Sawano, M. Miyao and K. Hamaya, Appl. Phys. Express, 2014, 7, 033002.
22. T. Saito, N. Tezuka, M. Matsuura and S. Sugimoto, Appl. Phys. Express, 2013, 6, 103006.
23. W. Han, X. Jiang, A. Kajdos, S. H. Yang, S. Stemmer and S. S. Parkin, Nat. Commun., 2013, 4, 2134.
24. N. Tombros, C. Jozsa, M. Popinciuc, H. T. Jonkman and B. J. Van Wees, Nature, 2007, 448, 571–574.
25. B. Dlubak, M.-B. Martin, C. Deranlot, B. Servet, S. Xavier, R. Mattana, M. Sprinkle, C. Berger, W. A. De Heer and F. Petroff, Nat. Phys., 2012, 8, 557–561.
26. L. E. Hueso, J. M. Pruneda, V. Ferrari, G. Burnell, J. P. Valdes-Herrera, B. D. Simons, P. B. Littlewood, E. Artacho, A. Fert and N. D. Mathur, Nature, 2007, 445, 410–413.
27. H. Kum, J. Heo, S. Jahangir, A. Banerjee, W. Guo and P. Bhattacharya, Appl. Phys. Lett., 2012, 100, 182407.
28. S. Heedt, C. Morgan, K. Weis, D. Bürgler, R. Calarco, H. Hardtdegen, D. Grützmacher and T. Schapers, Nano Lett., 2012, 12, 4437–4443.
29. S. Bandyopadhyay, M. Hossain, H. Ahmad, J. Atulasimha and S. Bandyopadhyay, Small, 2014, 10, 4379–4385.
30. E.-S. Liu, J. Nah, K. M. Varahramyan and E. Tutuc, Nano Lett., 2010, 10, 3297–3301.
31. S. Zhang, S. A. Dayeh, Y. Li, S. A. Crooker, D. L. Smith and S. Picraux, Nano Lett., 2013, 13, 430–435.
32. J. Tarun, S. Huang, Y. Fukuma, H. Idzuchi, Y. Otani, N. Fukata, K. Ishibashi and S. Oda, Appl. Phys. Express, 2012, 5, 045001.
33. D. Pan, D. Fan, N. Kang, J. Zhi, X. Yu, H. Xu and J. Zhao, Nano Lett., 2016, 16, 834–841.
34. N. Guo, W. Hu, L. Liao, S. Yip, J. C. Ho, J. Miao, Z. Zhang, J. Zou, T. Jiang and S. Wu, Adv. Mater., 2014, 26, 8203–8209.
35. D. Zheng, J. Wang, W. Hu, L. Liao, H. Fang, N. Guo, P. Wang, F. Gong, X. Wang and Z. Fan, Nano Lett., 2016, 16, 2548–2555.
36. A. Holleitner, V. Sih, R. Myers, A. Gossard and D. Awschalom, Phys. Rev. Lett., 2006, 97, 036805.
37. S. A. Dayeh, D. P. Aplin, X. Zhou, P. K. Yu, E. T. Yu and D. Wang, Small, 2007, 3, 326–332.
38. C. Blömers, T. Grap, M. Lepsa, J. Moers, D. Grützmacher, H. Lüth and T. Schäpers, Appl. Phys. Lett., 2012, 101, 152106.
39. X. Zhang, M. Fu, X. Li, T. Shi, Z. Ning, X. Wang, T. Yang and Q. Chen, Sens. Actuators, B, 2015, 209, 456–461.
40. F. Monzon, D. Patterson and M. Roukes, J. Magn. Magn. Mater., 1999, 195, 19–25.
41. F. Zwanenburg, D. Van der Mast, H. Heersche, L. Kouwenhoven and E. Bakkers, Nano Lett., 2009, 9, 2704–2709.
42. H. Jaffrès, J. M. George and A. Fert, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 82, 140408.
43. C. Kant, O. Kurnosikov, A. Filip, P. LeClair, H. Swagten and W. de Jonge, Phys. Rev. B: Condens. Matter Mater. Phys., 2002, 66, 212403.
44. T. Ishikura, L. K. Liefeith, Z. Cui, K. Konishi, K. Yoh and T. Uemura, Appl. Phys. Express, 2014, 7, 073001.

---