5d iridium oxide as a material for spin-current detection

Abstract

Devices based on pure spin currents have been attracting increasing attention as key ingredients for low-dissipation electronics. To integrate such spintronics devices into charge-based technologies, electric detection of spin currents is essential. The inverse spin Hall effect converts a spin current into an electric voltage through spin-orbit coupling. Noble metals such as Pt and Pd, and also Cu-based alloys, have been regarded as potential materials for a spin-current injector, owing to the large direct spin Hall effect. Their spin Hall resistivity ρ_SH , representing the performance as a detector, is not large enough, however, due mainly because of their low charge resistivity. Here we report that a binary 5d transition metal oxide, iridium oxide, overcomes the limitations encountered in noble metals and Cu-based alloys and shows a very large ρ_SH ~38 μΩ cm at room temperature.

ARTICLE

Received 18 Jul 2013 | Accepted 7 Nov 2013 | Published 11 Dec 2013

Kohei Fujiwara1,w, Yasuhiro Fukuma1,w, Jobu Matsuno1,w, Hiroshi Idzuchi2, Yasuhiro Niimi2, YoshiChika Otani1,2,w & Hidenori Takagi1,3

Devices based on pure spin currents have been attracting increasing attention as key ingredients for low-dissipation electronics. To integrate such spintronics devices into charge-based technologies, electric detection of spin currents is essential. The inverse spin Hall effect converts a spin current into an electric voltage through spin-orbit coupling. Noble metals such as Pt and Pd, and also Cu-based alloys, have been regarded as potential materials for a spin-current injector, owing to the large direct spin Hall effect. Their spin Hall resistivity rSH,

representing the performance as a detector, is not large enough, however, due mainly because of their low charge resistivity. Here we report that a binary 5d transition metal oxide, iridium oxide, overcomes the limitations encountered in noble metals and Cu-based alloys and shows a very large rSHB38 mO cm at room temperature.

DOI: 10.1038/ncomms3893

5d iridium oxide as a material for spin-current detection

1 RIKEN Advanced Science Institute, Wako, Saitama 351-0198, Japan. 2 Institute for Solid State Physics, University of Tokyo, Kashiwa, Chiba 277-8581, Japan.

3 Department of Physics, University of Tokyo, Hongo, Tokyo 113-0033, Japan. w Present addresses: The Institute of Scientic and Industrial Research, Osaka University, Ibaraki, Osaka 567-0047, Japan. (K.F.); Frontier Research Academy for Young Researchers, Kyushu Institute of Technology, Iizuka,

Fukuoka 820-8502, Japan. (Y.F.); RIKEN Center for Emergent Matter Science (CEMS), Wako, Saitama 351-0198, Japan. (J.M., Y.O.). Correspondence and requests for materials should be addressed to K.F. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 4:2893 | DOI: 10.1038/ncomms3893 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms3893

Spin-orbit coupling (SOC) is a relativistic effect and relates the spin moment of an electron to its orbital momentum via a momentum-dependent effective magnetic eld. In the

presence of SOC, a charge current without any spin polarization can be converted into a pure spin current (the ow of spin angular momentum) and vice versa, known as the direct spin Hall effect and the inverse spin Hall effect (ISHE)118. The presence of these phenomena in non-magnetic semiconductors had already been recognized in 1970s (refs 1922) but recent theoretical1,2 and experimental investigations3,4 have given a boost to this eld. Spin current-based electronics with low-energy consumption has since been discussed, where spin-current injection and detection, using charge/spin conversion by direct spin Hall effect and ISHE, played a key role. A variety of materials have subsequently been explored to realize efcient charge/spin conversion518. The playground for the exploration was rst limited to semiconductors but then extended to metals, where we can take advantage of the less pronounced effect of interfacial barriers. Heavy transition metals such as Pt610,14, Au9,12,13, and Pd8,9,11 and also Cu-based alloys17,18 were found to exhibit a particularly large spin Hall angle aSH (the maximum yield of the charge/spin conversion), 0.010.1 at room temperature, owing to their pronounced SOC effects. These effects give these metals an advantage for application as a spin injector. In addition to the large aSH, their low electrical resistivity (rC), typically rC

10 710 5 O cm, allows large charge currents to be passed and hence to inject a large spin current without serious

Joule heating.

Spin currents can be detected by using the ISHE. In contrast to the case for spin injectors, the low rC of the heavy metals with a large aSH is a disadvantages when it comes to spin detection. In the ISHE, the electric voltage DVISHE generated by a spin current IS is in proportion to rC (refs 23,24):

DVISHE / aSHrCIS rSHIS: 1 Spin Hall resistivity given by rSH aSHrC determines the

efciency of spin-current detection. This means that a sensitive detection of spin current could be achieved in materials with both a large aSH and a high rC. In typical heavy metals, their low rC imposes strong constraints in tuning the value of rSH as it remains

as low as o1 mO cm. One of the obvious approaches to improve the performance of spin detectors could be to increase the rC by

alloying. Alloying can in fact increase not only the rC but also the aSH through an extrinsic spin Hall effect2527. At room temperature where the inelastic scattering of charge carriers is dominant, however, it should be hard to anticipate a drastic increase of rC by

alloying, for example, more than an order of magnitude increase as compared with the pristine material. Therefore, the improvement of rSH would not be achieved only by alloying.

In this work, we propose 5d transition metal oxides (TMOs) as alternative metals to realize both large aSH and large rC, which should enhance rSH. The uniqueness of 5d TMOs is characterized by the extremely strong SOCB0.51 eV originating from the predominant 5d character of the conduction band. The SOC in 5d TMOs is in fact as strong as to reconstruct the electronic structures drastically, as rst discussed in the case of Jeff 1/2

Mott state Sr2IrO4 (ref. 28). The analogous predominant SOC effect has been commonly observed in a broad range of Ir oxides including Ba2IrO4 (ref. 29), CaIrO3 (ref. 30) and Ir2O4 (ref. 31). In those Ir oxides, in addition to the strong SOC, the localized character of d orbitals gives rise to a moderately high charge resistivity rC even in the metallic state. Typical rC values of conductive 5d Ir TMOs such as IrO2 and SrIrO3 at room temperature are of the order of 10 410 3 O cm, at least one order of magnitude higher than those of normal s metals. As the variety of materials continues to grow rapidly, 5d TMOs offer a

unique opportunity to explore the giant rSH for efcient spin detection. We selected a simple binary oxide, rutile IrO2 (refs 3234), as the basis for our exploration. IrO2 with Ir4 shares the same 5d5 conguration with many other Ir oxides, where the strong SOC dominates the electronic states and the resultant electronic properties34. This binary oxide has been long used as an electrode in various device applications, ranging from non-volatile ferroelectric memories to electrochemical devices. The excellent electrode properties are due to the formation of a clean, well-dened interface with other materials owing to its high chemical and thermal stability, and superior barrier properties for oxygen diffusion. In such a clean interface, it may be possible to inject a spin current through a diffusion process; IrO2 is therefore an ideal platform for testing the ISHE of 5d TMOs. Here, we describe successful spin-current injection into IrO2 using a lateral spin-valve (LSV) device geometry and demonstrate the ISHE with a remarkably high rSH as a metal not only in the polycrystalline state but also in the practically important amorphous state.

ResultsDevice structure. We fabricated cross-junction-type devices consisting of an IrO2 wire and a permalloy(Ni80Fe20, Py)/Ag/Py

LSV35, shown in Fig. 1a,b (see also Methods). Spin currents are passed in the Ag strip by a non-local charge current injection (IC)

from the ferromagnetic Py electrode and are perpendicularly injected in part into the IrO2 wire through the diffusion (Fig. 1a). This effect, called spin absorption, enables us to conduct a quantitative analysis of the ISHE as have been demonstrated for various materials7,8,17,18. We examined the potential of polycrystalline and amorphous IrO2 wires as the spin detector element. Their resistivities were 2.0 10 4 O cm and

5.7 10 4 O cm at 300 K, respectively, which are 12 orders of

magnitude higher than those of metals that have been studied to date as a spin Hall material.

H

V

IC

Ag

Py1

Py2

IrO2

IrO2

Py1 Py2

Ag

Figure 1 | Inverse spin Hall effect measurements using a spin absorption effect in a lateral spin-valve geometry. (a) The structure of device used for ISHE measurements is shown schematically. A LSV with two Py electrodes (Py1: spin source, Py2: spin detector) bridged by an Ag spin transport layer was formed on an IrO2 wire. Spin-polarized charge current,

IC, was injected along the arrow to accumulate pure spin currents inthe Ag. The diffusion of spin current into the IrO2 wire, namely, the spin absorption, gives rise to the ISHE in IrO2. (b) Scanning electron microscopy image of a typical device. Scale bar, 500 nm. The dotted lines indicatethe side edges of the IrO2 wire for clarity.

2 NATURE COMMUNICATIONS | 4:2893 | DOI: 10.1038/ncomms3893 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms3893 ARTICLE

To check that the Ag layer is not oxidized by directly contacting with IrO2, we rst characterized the Ag/IrO2 interface by means of interface resistance measurements36. The current

voltage characteristics of the Ag/IrO2 interface showed an ohmic behaviour, typical of metal/metal interface. The slope of the data yielded a resistancearea product (RA) of 2.3 fO m2 at 300 K, which is as low as that of the metallic, transparent Ag/Py interface (RA 1.9 fO m2 at 300 K). An estimate of the thickness of the

possibly oxidized Ag layer, using the resistivities of Ag2O and

AgO in the literature37, never exceeded 0.1 nm. These observations imply that the Ag/IrO2 interface in the present devices is very sharp as in metal/metal interfaces.

Spin-current injection into IrO2. The spin absorption by IrO2 across the interface was then conrmed by non-local spin-valve (NLSV) experiments7,8,17,18,35. The schematic diagram of the measurement is shown in Fig. 2a. The pure spin currents accumulated in the Ag strip by injecting spin-polarized charge currents IC from Py1 to Ag is detected as a voltage VS ( V V ) between Py2 (V ) and Ag (V ) (upper panel).

The spin accumulation signal DRS is given by the difference in VS/ IC (RS) between parallel (high RS) and antiparallel (low RS) magnetizations of the two Py wires. If an IrO2 middle wire is in contact with the Ag, the accumulated spin diffuses in part into it (lower panel), resulting in the reduction in DRS (ref. 7).

Figure 2b and c show eld dependences of RS at 300 and 10 K, respectively. The decrease of DRS by the insertion of the IrO2 middle wire indicates the spin absorption effect. This is more clearly demonstrated in the temperature (T) dependence of DRS with (DRSwith) and without (DRSwithout) the middle wire, as displayed in Fig. 2d. At all the temperatures, DRSwith is systematically smaller than DRSwithout, while both are enhanced

at low temperatures owing to the increase of the spin-diffusion length of Ag (lAg)35. By analysing the data based on the three-dimensional spin-diffusion model18 (see Supplementary Methods and Supplementary Fig. S1), we estimated a spin-diffusion length in polycrystalline IrO2, lIrO2 3.8 nm at 300 K and 8.4 nm at

10 K, which are comparable to those of Pt (310 nm at 300 K (refs 610,14)) and CuIr (520 nm at 10 K (ref. 17)) where relatively large spin Hall angles were observed. This indicates that spin scattering of IrO2 is indeed very strong as expected from the 5d conduction band34.

ISHE and qSH of IrO2. By injecting spin currents through the spin absorption effect, we successfully observed ISHE in IrO2. In Fig. 3a,b, we plot the room-temperature ISHE signal RISHE

( DVISHE/IC) for polycrystalline and amorphous samples as a

function of the magnetic eld (The temperature dependence of the ISHE is shown in Supplementary Fig. S2 and is explained in Supplementary Note 1). Here the magnetic eld was applied along the hard axis of the Py spin source (see Fig. 1a), and DVISHE

was induced along the long-axis direction of the IrO2 wire. Reecting the magnetization process in hard axis of Py, RISHE

linearly increased up to B1,500 Oe and then saturated. The antisymmetric response to the applied magnetic eld excludes any spurious effects such as anisotropic magnetoresistance of Py; the ISHE of IrO2 is responsible for the resistance change (2DRISHE) between positive and negative magnetic elds. The in-plane angular dependence of magnetic eld effect on RISHE was

conrmed fully consistent with the ISHE origin. Figure 3c shows the variation in RISHE under a magnetic eld direction rotated by f with respect to the ordinary set-up (Fig. 3d, inset). The sign change of RISHE occurs between f 0 and p by magnetization

switching of Py and concomitant polarization reversal of the spin

2.0

0

R S(m)

R S(m)

R S(m)

0.2

Without IrO2

With IrO2 VS

VS

H

V

0.4

IC

0.6

w/o IrO2w/IrO2 300 K

10 K

Ag

0

400 200 0 200 400H (Oe)

Py1

0.5

Py2

1.0

1.5

H

V

IC

1.5

1.0

0.5

0 300

200

100

0T (K)

IrO2

w/IrO2

w/o IrO2

Figure 2 | Non-local spin-valve measurements. (a) The set-up of measurement is schematically depicted. The magnetic eld H is applied along the easy axis of the Py electrodes. (b,c) NLSV signals measured at 300 K and 10 K for Py/Ag/Py LSVs without and with the polycrystalline IrO2 middle wire.

(d) Temperature dependences of the NLSV signals.

NATURE COMMUNICATIONS | 4:2893 | DOI: 10.1038/ncomms3893 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms3893

0.2

IrO2 Polycrystalline w = 170 nm

Amorphousw = 100 nm w = 170 nm

2RISHE

0.1

R ISHE(m)

R ISHE (m)

0

0

=

0.2

0.1

4,000 0 4,000 H (Oe)

0.2

0.1

R ISHE (m)

0.1

0

2R ISHE (m)

0

H

0.1

0.1

0.2

4,000 2,000 0 2,000 4,000 H (Oe)

0

/2

(rad)

= 0

Figure 3 | Inverse spin Hall effect of IrO2 at 300 K. (a) The ISHE signal RISHE ( DVISHE/IC) for a polycrystalline IrO2 wire with a width w of 170 nm

was measured as a function of the applied magnetic eld H. (b) The results for amorphous IrO2 samples. Two devices with different w values (100 and 170 nm) were measured. The 100 nm-width device showed larger RISHE because of the increase of spin-current density. RISHE of the two amorphous devices are smaller than that of the polycrystalline sample shown in a. (c,d) Angular dependence of RISHE and the resistance change 2DRISHE by the ISHE.

The in-plane azimuthal angle f with respect to the long axis of Ag is rotated to change the polarization direction of the spin current (inset in d). The solid line in d is a cos f curve tted to the results.

current. As shown in Fig. 3d, the resistance change 2DRISHE

vanishes at f p/2 and exhibits a cos f relation. This char

acteristic angular dependence is the hallmark of ISHE7,12, evidencing spin-dependent electron scattering in IrO2.

Because of the large difference in rC between Ag and IrO2, the charge current induced in the IrO2 wire by the ISHE is partially shunted by the adjacent Ag layer. To take into account such geometric effects and to precisely determine rSH, we adopted the three-dimensional spin-diffusion model (see also Supplementary Methods). Using the ISHE signal 2DRISHE and data obtained from

the NLSV measurements, aSH and rSH are, respectively, calculated to be 0.040 and 8.0 mO cm for polycrystalline IrO2, and 0.065 and37.5 mO cm for amorphous IrO2 at 300 K.

Figure 4 summarizes spin Hall resistivity rSH and electrical

resistivity rC for a variety of metals with a large spin Hall angle aSH reported so far, together with that for IrO2. It mimics the point of this work in that rSH of IrO2 is distinctly large compared with typical heavy metals and their alloys of which rSH is

B0.161.2 mO cm (refs 614,17,18), and is now comparable to the large rSH that was recently discovered in ultrathin lms of b-phase (A15 crystal structure) W (ref. 15) and Ta (ref. 16).

DiscussionTo generate a large electric voltage using the ISHE of IrO2 from existing metal-based spin-current devices, it is important to reduce the shunting effect at the Ag/IrO2 interface due to their large difference in rC. One of the promising approaches is to use Al (ref. 38) and Mg (ref. 39) as the non-magnetic spin transport layer. Their rC values are one order of magnitude higher than that of Ag we used, of the order of 10 mO cm at room temperature, while a comparably long spin-diffusion length is obtained. Another interesting route is a spin transport layer based on a conductive oxide composed of light elements with small

SOC. For example, in ZnO (ref. 40), doping with Al yields rC of

the order of 100 mO cm (ref. 41), which can match with that of IrO2. The high chemical stability of IrO2 could be advantageous in fabricating such oxide-based junctions.

The discovery of the ISHE in amorphous IrO2, which can be grown at room temperature, is a signicant step towards spintronics devices based on such oxide materials. The comparably large aSH in the amorphous form may point to that spin scattering in this compound occurs at the length scale of as short as a unit cell, potentially enabling the spin detection at the atomic layer thickness. Although the physics of spin scattering in amorphous states has not been understood yet, the availability of the amorphous form should make it easier to study SOC-induced transport phenomena in 5d TMOs not used before in spintronics.

In conclusion, 5d TMOs are the most promising materials for spin detection in spintronics. We demonstrated that the combination of strong SOC and a moderately high rC is a key to improve the performance of spin detection. These achievements in the simple binary oxide manifest the potential of 5d TMOs as a new class of spintronic materials. Strong SOC in this class of materials produces a rich variety of intriguing physical properties, including correlated topological insulator42, Weyl semi-metal43 and Kitaev spin liquid44. Our discovery of the large spin Hall resistivity may make 5d TMOs even more fascinating.

Methods

Device fabrication. The Py/Ag/Py LSV with an IrO2 middle wire was fabricated on a SiO2/Si substrate. Au/Ti leads for electrical measurements were prepared on the substrate by a photolithography and an e-beam deposition. IrO2 lm was grown by a reactive sputtering with a 99.9% pure Ir target. A wire structure of IrO2 was formed using a posi-type resist patterned by an e-beam lithography. The typical wire width and thickness were 170 and 15 nm, respectively. During the deposition, the substrate was not heated, and the pressure was xed at 0.7 Paof Ar/O2 (90:10) mixture gas. X-ray diffraction analysis indicated that the as-deposited IrO2 lm was amorphous. The amorphous lm turned into

4 NATURE COMMUNICATIONS | 4:2893 | DOI: 10.1038/ncomms3893 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms3893 ARTICLE

V

104

-W15

a-IrO2

-Ta16

| SH| (cm)

105

106

poly-IrO2

CuBi18 Pt14

SH=0.01

CuIr17

Ta8

Au12

Pt10 Nb8

Pd8

Pt8 Mo8

Pt9

Pd11

SH=0.1

107106

105 104 103

C ( cm)

Figure 4 | Spin Hall resistivity of metals. Experimentally measured values of spin Hall resistivity rSH for various metals are plotted as a function of electrical resistivity rC. Data include Pt (refs 9,10) and Pd (ref. 11) evaluated at room temperature by spin pumping, Pt (ref. 14) and atomic layerthin lms of b-W (ref. 15) and b-Ta (ref. 16) at room temperature by spin-transfer-torque ferromagnetic resonance, Au at 295 K using a planar

Hall device12, Ta, Nb, Pd, Mo, Pt (ref. 8), 12% Ir-doped Cu (ref. 17), 0.5% Bi-doped Cu (ref. 18) at 10 K, and polycrystalline and amorphous IrO2 at

300 K (this work) by spin absorption.

polycrystalline after a post-annealing at 400 C in air. The LSV with a PyPy separation of 700 nm (centre-to-centre) was then formed onto the IrO2 wire by the shadow evaporation method35,36. To prevent the surface oxidation of Ag, a 2-nm-thick MgO protection layer was e-beam deposited on the surface without breaking vacuum. The width and thickness of wire were 170 and 100 nm for Ag, and 170 and 20 nm for Py, respectively. Their rC and spin-diffusion length were 3.0 mO cm and 330 nm for Ag and 47 mO cm and 5 nm for Py, respectively, at 300 K. One of the two Py electrodes was made longer than the other (not shown in Fig. 1b) to make the contrast in the elds for magnetization reversal.

Electrical measurements. All electrical measurements were performed by a standard lock-in technique in a He ow cryostat. An alternating current IC of

200 mA with a frequency of 79 Hz was used for NLSV measurements. In ISHE measurements, IC of 400 mA was applied. We conrmed the linearity between the output voltage DVISHE and IC in the range of 100400 mA.

References

1. Hirsch, J. Spin Hall effect. Phys. Rev. Lett. 83, 18341837 (1999).2. Zhang, S. Spin Hall effect in the presence of spin diffusion. Phys. Rev. Lett. 85,

393396 (2000).

3. Kato, Y. K., Myers, R. C., Gossard, A. C. & Awschalom, D. D. Observation of the spin Hall effect in semiconductors. Science 306, 19101913 (2004).

4. Wunderlich, J., Kaestner, B., Sinova, J. & Jungwirth, T. Experimental observation of the spin-Hall effect in a two-dimensional spin-orbit coupled semiconductor system. Phys. Rev. Lett. 94, 047204 (2005).

5. Valenzuela, S. O. & Tinkham, M. Direct electronic measurement of the spin Hall effect. Nature 442, 176179 (2006).

6. Saitoh, E., Ueda, M., Miyajima, H. & Tatara, G. Conversion of spin current into charge current at room temperature: inverse spin-Hall effect. Appl. Phys. Lett. 88, 182509 (2006).

7. Kimura, T., Otani, Y., Sato, T., Takahashi, S. & Maekawa, S. Room-temperature reversible spin Hall effect. Phys. Rev. Lett. 98, 156601 (2007).

8. Morota, M. et al. Indication of intrinsic spin Hall effect in 4d and 5d transition metals. Phys. Rev. B 83, 174405 (2011).

9. Mosendz, O. et al. Detection and quantication of inverse spin Hall effect from spin pumping in permalloy/normal metal bilayers. Phys. Rev. B 82, 214403 (2010).

10. Ando, K. et al. Electric manipulation of spin relaxation using the spin Hall effect. Phys. Rev. Lett. 101, 036601 (2008).

11. Ando, K. & Saitoh, E. Inverse spin-Hall effect in palladium at room temperature. J. Appl. Phys. 108, 113925 (2010).

12. Seki, T. et al. Giant spin Hall effect in perpendicularly spin-polarized FePt/Au devices. Nat. Mater. 7, 125129 (2008).

13. Mihajlovi, G., Pearson, J., Garcia, M., Bader, S. & Hoffmann, A. Negative nonlocal resistance in mesoscopic gold Hall bars: absence of the giant spin Hall effect. Phys. Rev. Lett. 103, 166601 (2009).

14. Liu, L., Moriyama, T., Ralph, D. & Buhrman, R. Spin-torque ferromagnetic resonance induced by the spin Hall effect. Phys. Rev. Lett. 106, 036601 (2011).

15. Pai, C.-F. et al. Spin transfer torque devices utilizing the giant spin Hall effect of tungsten. Appl. Phys. Lett. 101, 122404 (2012).

16. Liu, L. et al. Spin-torque switching with the giant spin Hall effect of tantalum. Science 336, 555558 (2012).

17. Niimi, Y. et al. Extrinsic spin Hall effect induced by iridium impurities in copper. Phys. Rev. Lett. 106, 126601 (2011).

18. Niimi, Y. et al. Giant spin Hall effect induced by skew scattering from bismuth impurities inside thin lm CuBi alloys. Phys. Rev. Lett. 109, 156602 (2012).

19. Chazalviel, J.N. & Solomon, I. Experimental evidence of the anomalous Hall effect in a nonmagnetic semiconductor. Phys. Rev. Lett. 29, 16761679 (1972).

20. Chazalviel, J.N. Spin-dependent Hall effect in semiconductors. Phys. Rev. B 11,

39183934 (1975).

21. Dyakonov, M. I. & Perel, V. I. Possibilty of orienting electron spins with current. Sov. Phys. JETP Lett. 13, 467469 (1971).

22. Bakun, A. A., Zakharchenya, B. P., Rogachev, A. A., Tkachuk, M. N. & Flesher, V. G. Observation of a surface photocurrent caused by optical orientation of electons in a semiconductor. Sov. Phys. JETP Lett. 40, 12931295 (1984).

23. Ando, K. & Saitoh, E. Observation of the inverse spin Hall effect in silicon. Nat. Comm. 3, 629 (2012).

24. Vila, L., Kimura, T. & Otani, Y. Evolution of the spin Hall effect in Pt nanowires: size and temperature effects. Phys. Rev. Lett. 99, 226604 (2007).

25. Gu, B. et al. Surface-assisted spin Hall effect in Au lms with Pt impurities. Phys. Rev. Lett. 105, 216401 (2010).

26. Fert, A. & Levy, P. Spin Hall effect induced by resonant scattering on impurities in metals. Phys. Rev. Lett. 106, 157208 (2011).

27. Gradhand, M. et al. Perfect alloys for spin Hall current-induced magnetization switching. SPIN 2, 1250010 (2012).

28. Kim, B. J. et al. Phase-sensitive observation of a spin-orbital Mott state in Sr2IrO4. Science 323, 13291332 (2009).

29. Okabe, H. et al. Ba2IrO4: a spin-orbit Mott insulating quasi-two-dimensional antiferromagnet. Phys. Rev. B 83, 155118 (2011).

30. Ohgushi, K. et al. Resonant X-ray diffraction study of strongly spin-orbit-coupled Mott insulator CaIrO3. Phys. Rev. Lett. 110, 217212 (2013).

31. Kuriyama, H. et al. Epitaxially stabilized iridium spinel oxide without cations in the tetrahedral site. Appl. Phys. Lett. 96, 182103 (2010).

32. Ryden, W., Lawson, A. & Sartain, C. Electrical transport properties of IrO2 and RuO2. Phys. Rev. B 1, 14941500 (1970).

33. Mattheiss, L. Electronic structure of RuO2, OsO2, and IrO2. Phys. Rev. B 13, 24332450 (1976).

34. Hirata, Y. et al. Complex orbital state stabilized by strong spin-orbit coupling in a metallic iridium oxide IrO2. Phys. Rev. B 87, 161111 (2013).

35. Idzuchi, H., Fukuma, Y., Wang, L. & Otani, Y. Spin relaxation mechanism in silver nanowires covered with MgO protection layer. Appl. Phys. Lett. 101, 022415 (2012).

36. Fukuma, Y. et al. Giant enhancement of spin accumulation and long-distance spin precession in metallic lateral spin valves. Nat. Mater. 10, 527531 (2011).

37. Abe, Y., Hasegawa, T., Kawamura, M. & Sasaki, K. Characterization of Ag oxide thin lms prepared by reactive RF sputtering. Vacuum 76, 16 (2004).

38. Valenzuela, S. O. & Tinkham, M. Spin-polarized tunneling in room-temperature mesoscopic spin valves. Appl. Phys. Lett. 85, 5914 (2004).

39. Idzuchi, H., Fukuma, Y., Wang, L. & Otani, Y. Spin diffusion characteristics in magnesium nanowires. Appl. Phys. Express 3, 063002 (2010).

40. Althammer, M., Karrer-Mller, E.-M., Goennenwein, S. T. B., Opel, M. & Gross, R. Spin transport and spin dephasing in zinc oxide. Appl. Phys. Lett. 101, 082404 (2012).

41. Suzuki, A., Matsushita, T., Wada, N., Sakamoto, Y. & Okuda, M. Transparent conducting Al-doped ZnO thin lms prepared by pulsed laser deposition. Jpn J. Appl. Phys. 35, L56L59 (1996).

42. Shitade, A. et al. Quantum spin Hall effect in a transition metal oxide Na2IrO3. Phys. Rev. Lett. 102, 256403 (2009).

43. Wan, X., Turner, A. M., Vishwanath, A. & Savrasov, S. Y. Topological semimetal and Fermi-arc surface states in the electronic structure of pyrochlore iridates. Phys. Rev. B 83, 205101 (2011).

44. Jackeli, G. & Khaliullin, G. Mott insulators in the strong spin-orbit coupling limit: from Heisenberg to a quantum compass and Kitaev models. Phys. Rev. Lett. 102, 017205 (2009).

NATURE COMMUNICATIONS | 4:2893 | DOI: 10.1038/ncomms3893 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms3893

Acknowledgements

We thank K. Ohgushi for helpful discussions. This work was supported by Grants-in-Aid (number 22340108, 23103518 and 24224010) from MEXT, Japan.

Author contributions

K.F., Y.F. and J.M. designed the experiments. K.F., Y.F. and H.I. fabricated devices and collected the data. K.F. and Y.N. analysed the data. K.F. wrote the manuscript with input from J.M. Y.O., and H.T. H.T. and Y.O. planned and supervised the project. All authors discussed the results.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Fujiwara, K. et al. 5d iridium oxide as a material for spin-current detection. Nat. Commun. 4:2893 doi: 10.1038/ncomms3893 (2013).

6 NATURE COMMUNICATIONS | 4:2893 | DOI: 10.1038/ncomms3893 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.