See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327717805

Cross-shaped nanostructures for the study of spin to charge interconversions

using spin-orbit coupling in non-magnetic materials

Preprint · September 2018

CITATIONS READS

0 21

6 authors, including:

Pham Van Tuong Paul Noël

Atomic Energy and Alternative Energies Commission ETH Zurich
52 PUBLICATIONS   800 CITATIONS    32 PUBLICATIONS   141 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

spin Hall effect in nanowire devices View project

Metamaterial perfect absorber View project

All content following this page was uploaded by Pham Van Tuong on 03 June 2019.

The user has requested enhancement of the downloaded file.

 Cross-shaped nanostructures for the study of spin to charge interconversions
using spin-orbit coupling in non-magnetic materials

V. T. Pham1, L. Vila1,*, G. Zahnd1, P. Noël1, A. Marty1, and J. P. Attané1,†

1 SPINTEC, CEA-INAC/CNRS/Univ. Grenoble Alpes, F-38000 Grenoble, France

ABSTRACT
Several spin-orbit effects allow performing spin to charge inter-conversion: the spin Hall effects, the
Rashba effect, or the spin-momentum locking in topological insulators. Here we focus on how the
detection of this inter-conversion can be made electrically, using three different cross-shaped
nanostructures. We apply these measurement configurations to the case of the spin Hall effect in Pt,
using CoFe electrodes to detect and inject spins. Both the direct and inverse spin Hall effect can be
detected, with a spin Hall signal up to two order of magnitude higher than that of nonlocal
measurements in metallic lateral spin valves, and with a much simpler fabrication protocol. We
compare the respective signal amplitude of the three proposed geometries. Finally, comparison of
the observed spin signals with finite element method calculations allows extracting the spin Hall
angle and the spin diffusion length of Pt.

TEXT

Spin current manipulation is at the core of spintronics. Recently, spin-orbit interaction (SOI)
in nonmagnetic materials has been exploited to create and detect the spin currents. For
instance, the spin Hall effect (SHE) and its inverse [1] allows performing the inter-conversion
between charge and spin currents in bulk metals [2]. Edelstein Rashba effects, at Rashba
interfaces or at the surfaces of topological insulators, can also be highly competitive for
ensuring such interconversions [3].

Direct SHE (DSHE) allows converting a charge current into a transverse pure spin current
(PSC), a flow of spin angular momentum without net charge flow. The relationship between
the produced PSC (Js), the charge current (Jc) and the spin polarization of electrons (s) can be

*
[email protected]
†
[email protected]
 ħ
written 𝑱𝑺 = 𝑒 𝛩𝑆𝐻𝐸 𝑱𝑪 × 𝒔 [1], where the conversion rate 𝛳SHE is the spin Hall angle of the

considered material. Reciprocally, a transverse charge current can be generated from a spin
current in the reverse mechanism with the same conversion rate.

A lot of techniques have been devoted to SHE detection, such as the FM-based spin
pumping/spin torque [4, 5, 6], spin Hall magnetoresistance [7], optical schemes [8, 9] and
electrical techniques on nanodevices [10, 19, 16]. The nanostructures which allow detecting
electrically the SHE are of prime interest in view of SO-based logic devices [11] and non-
volatile MRAMs [12], altogether with applications of Spin Orbit torque experiments [13, 14].

Among electrical methods, non-local techniques which exploit abilities of ferromagnets for
injecting and detecting spin current in lateral spin valves (LSV) [15, 16] is playing a significant
role in SHE researches. Notwithstanding, the complex nanostructure fabrication and the
small absolute values of the spin signals, especially for metallic systems, is hampering its
straightforward use.

Electrical devices which can produce high spin signals are thus of prime importance not only
for basics research but also for applications. Recently, a simple F/N bilayer device with two
ferromagnetic electrodes probing directly the spin accumulation induced by the SHE of an
underneath metallic nanowire with strong SOC has been proposed to achieve large spin-
charge inter-conversion signals [17]. In this paper, we compare DSHE and ISHE measurement
on different cross-shaped nanostructures. Thanks to the local spin detection technique, our
experiments show mΩ-DSHE/ISHE signals, up to two orders of magnitude larger compared to
classical non local devices, with a much simpler structure that could be implemented into
SHE-based computational devices [11, 12]. The extracted spin Hall angle of Pt in these
CoFe/Pt systems by the FEM simulations shows a good agreement with the spin Hall angle of
Pt extracted in these Co/Fe Systems by FEM simulations [17].

The samples have been fabricated by conventional e-beam lithography, e-gun deposition,
and lift-off processes on the thermal oxidized SiO2/Si substrate. The chosen materials are
Co60Fe40 and Pt, as they are archetypal ferromagnetic and SHE materials. Their magnetic and
spin transport properties have been previously characterized [17, 18]. The depositions were
performed in high vacuum, about 4e-8 mTorr. The top surface of the Pt nanowires is cleaned
by Ar ion etching prior to the deposition of the CoFe electrode. They are connected by Au/Ti
electrical pads. In all the samples the thickness of the CoFe layer is 15 nm, whereas that of Pt
is 7 nm for the device of Fig. 1(a), and 8 nm for the samples of Figs. 1(c) and 1(e). The widths
of the CoFe and Pt nanowires are 50 nm and 400 nm, respectively. The transport
measurements were performed using a lock-in amplifier working at 332 Hz, with an applied
current of 100 or 200 µA. In the presented measurement configurations, the spin signal in
Ohm is the ratio between the detected voltage and the applied current. In all experiments,
the field is applied along the easy axis of the ferromagnetic electrodes. Figures 1(a), 1(c) and
1(e) show the SEM images and the measurement principle of the studied devices in view of
DSHE configuration, whereas ISHE measurement setups are sketched on Figs. 1(b, d, f).

Figures 1(a) and 1(b) represent the device proposed in ref. 17. In Fig. 1(a), a vertical PSC is
created by DSHE when flowing an electrical current along the Pt wire. The PSC is probed by
the two CoFe electrodes, providing that the magnetizations of the two ferromagnetic-
electrodes are opposite. Otherwise they probe the same electrochemical potential and no
voltage difference develops between them. Reciprocally in Fig. 1(b), a non-zero net spin
polarized current is injected vertically at the two interfaces of CoFe/Pt only when the two
magnetizations of the CoFe electrodes are opposite. The ISHE of Pt wire converts this spin
current into a transverse charge current which can be detected as a voltage in open circuit
condition.

A simplified design of a cross-shaped device is shown in Figs. 1(c) and 1(d), with only a single
CoFe electrode on top of a T-shaped pattern made of Pt. In Fig. 1(c), the charge current
flowing in the Pt creates a spin accumulation at its top surface. The transverse contacts
probe the electrochemical potential associated to the spin accumulation at the CoFe/Pt top
interface. The polarization of the ferromagnetic electrode controlled by an external
magnetic field, defines if majority or minority spins are probed. The inverse SHE is measured
by permuting the current and voltage leads [Fig. 1(d)]. The charge current flowing in the
CoFe allows injecting a spin current into the Pt stripe. This spin current is converted into a
transverse charge current by ISHE, which leads to the development of a transverse voltage in
open circuit conditions. The direction of the produced charge current is reversed when the
polarization of the spin current changes with the reversal of the magnetization of the
ferromagnetic electrode. Figures 1(e) and 1(f) illustrate a simple cross made of two straight
CoFe and Pt wires, respectively designed for ISHE and DSHE measurements using the same
principles.

The magnetoresistance loops of the DSHE and the ISHE are plotted in Figs. 2. (a, c, and e)
and Figs. 2. (b, d, and f) respectively, using the configurations shown in Figure 1. All the
figures have the same range of signal amplitude to ease their direct comparison, an arrow on
Fig. 2(c) depicts a 5 mΩ amplitude. The drop and rise of the transverse resistance
corresponds to the magnetization reversal of the CoFe contact, leading to the change of the
probed electrochemical potential (majority or minority) in DSHE, or to the change of sign of
injected spin current polarization in ISHE. The amplitude of the spin signals is the difference
between the different plateaus.

These simple nanostructures show that the absolute value of the spin-to-charge signal is in
the 10 mΩ range at room temperature, around two orders of magnitude larger than
standard nonlocal measurement technique in metallic LSVs [15, 19, 20]. For instance, the
signal is even comparable to the tunnelling spin Hall signals obtained in LSVs having a 2D-
graphene channel that owns superior spin transport properties [21, 22]. The enhancement of
the spin signals in our device takes also advantage of the stronger spin current injection of
CoFe compare to NiFe [18].

As expected for ISHE and DSHE, the obtained signals have similar amplitude when changing
the measurement configuration, as expected by their reciprocal relation. We believe that the
small offset resistance can be explained by the slight misalignment between the different
electrodes, as seen in the SEM image in Fig. 1(d).

The results illustrated in Figs. 2(a) and 2(b) correspond for the measurement setups carried
out on device geometry similar to that of ref. 17. The total spin signal, the difference of
resistance level for the two anti-parallel states, is of the order of 16 mΩ. The amplitude of
the signal is larger by nearly a factor of 2 than the 7 mΩ for the device shown in Figs. 2 (c)
and 2(d). Indeed, in the measurements of Fig. 2(a-b) the spin current is injected twice at the
F/NM interface in ISHE or the spin accumulation measured twice in DSHE, the signal is thus
roughly a factor of 2 larger in the sample geometry presented in Figs. 1(a-b) with respect to
that of Figs. 1(c-d). Nonetheless, the geometry of Figs.1 (c-d) has larger tolerance in the
alignment error between the ferromagnetic wires and the Pt slab, and might thus be easier
to scale down.
Figures 2(e) and 2(f) illustrate the magnetoresistance loops corresponding to the geometry
device and probe configuration of Figs. 1(e) and 1(f), respectively. A zero-spin signal is
obtained in both cases. In the DSHE configuration, the same electrochemical potential is
probed along the ferromagnetic wire when it is homogenously magnetized and thus no
voltage develops [Fig. 2(e)]. In the ISHE configuration, the injected spin current into the Pt
wire is zero in average, thus producing no transverse voltage. We will show later how to
access the spin signal in this device.

As shown in Figures 3, these cross nanostructures allow studying the spin signal using other
probe configurations, as depicted in the insets, with the voltage and current leads circulating
by pairs around the sample. This type of measurement has been used in experiments
involving Topological Insulators [23, 24, 25]. The first observation is that Figs. 3(a) and 3(b)
combines DSHE and ISHE signals. They can be distinguished by the two switching events and
the plateau on each branch of the hysteresis, corresponding to the achievement of the
antiparallel magnetic states of the CoFe electrodes. Since a nucleation pad is patterned at
the end of the bottom ferromagnetic electrode (F1), it has a lower reversal field. Depending
on this contact being either injecting a current or detecting a voltage, the corresponding
reversal event (at low absolute field value) corresponds to either ISHE or DSHE signal. This is
the opposite for the reversal event at larger absolute field. The second observation is that
the spin signal, around 7 mΩ, is roughly two times lower than that of Figs. 2(a-b). Indeed,
now only one detection (DSHE) and injection (ISHE) mechanism occurs here.

The same behaviour is observed in the T-shape device. The spin signals in Fig. 3(c) and (d)
are roughly a factor of 2 smaller than that of Figs. 2c and 2d, 3 mΩ compared to 7 mΩ. The
signal is decreased because only half of the transverse voltage is measured in ISHE, or half of
the spin accumulation in DSHE. In contrast the probe configuration of Figs. 3(e) and 3(f)
allows now the observation a signal for the device with a straight CoFe wire. The results can
be explained by using the same mechanism in Figs. 2(a) and 2(b), i.e. being a combination of
DSHE and ISHE. However, there is no distinction between DSHE and ISHE contributions in the
signal here because both the injector and detector have the same magnetic state.
Additionally, because the ferromagnetic-wire is continuous the transverse resistance is
decreased and the spin signal amplitude is much reduced compare to the results in Figs. 3(a)
and 3(b), where instead the CoFe wire doesn’t fully cover the Pt strip.
Note that the anomalous Hall effect (AHE) [26] can also appear with the same principle as
the SHE in all of the measurements. However, the AHE plays an insignificant role (less than
10 %) in those thin samples [see more details in the supplementary material of ref. 17].

The device presented in Fig. 1(c) has thus the advantage of showing a much larger signal
than that of Fig. 1(e), and a simpler fabrication process than that of Fig. 1(a). In order to
estimate the spin-charge conversion rate of this system, we carried out finite element
method simulations, similarly as in ref. 17 for the sample of Figs. 1(a-b), within the
framework of a 2 spin-current drift diffusion model [27].

Figure 4 describes the simulations in ISHE configuration for the scheme shown in Fig. 1(c).
The distribution of the applied charge current in Fig. 4(a) allows injecting the spin polarized
current in Z direction at the CoFe/Pt interface. The contour of the spin accumulation in Figs.
4(b) and 4(c) is for a spin current flowing in the – Z direction when the magnetization is
polarized along + X. The amplitude of the spin signal is the difference of the two calculated
values of the transverse voltage. We retrieved the same spin Hall angle and the spin
diffusion length (λs) of Pt reported in ref. 17 for a system made of CoFe/Pt, ΘSHE = 0.19 and λs
= 3.0 nm, with the measured resistivity ρPt = 28 μΩcm for Pt at a thickness of 8 nm. The
resistivity of CoFe is ρCoFe = 20 μΩcm, for a thickness of 15 nm.

To conclude, we studied the spin-charge inter-conversion on simple cross-shaped

nanostructures. The milliohm-spin signals are obtained in both the DSHE and ISHE
configurations thanks to this local measurement technique. Additionally, we can measure
DSHE and ISHE in a simple cross junction between CoFe and Pt wires. The FEM simulations
were performed to analyse the spin Hall conversion rate. Apart from giving a
straightforward technique for the metrology of SHE/ISHE, these results are also promising
for developing computational magnetic devices, such as a non-volatile memory with a single
nano-magnet [12] or spin-orbit based logic circuits [11]. These device geometries can also be
used to study the spin-charge interconversion using Rashba interfaces or Topological
Insulators.
Figure Captions:
Figure 1: (a), (c), and (e) show the SEM image of the studied devices and the schemes for ISHE
measurements. In (a) and (c), the ferromagnetic electrodes, CoFe (in dark-grey color) are patterned
on top of the transverse nanowire and of a T-shaped stripe made of Pt (in wine color). An applied
charge current Jc injects a spin current Js from the CoFe electrode into the Pt layer along – Z-direction
at the interfaces. Thanks to the ISHE in the Pt layer, the spin current is deflected orthogonally to the
plane defined by the spin S (X) of electron and Js (Z), i.e., a transverse charge current is generated
along the Y direction, which is a Y–Z cross-section of the device. A voltmeter is used to detect the ISHE
voltage (VISHE) in the nanowire along the Y-direction. (e) The opposite spin current in each side of the
interface expects a negligible spin signal. The external magnetic field (yellow arrow) controls the
magnetization in the FM electrode, represented by the black arrows. (b), (d), and (f) are sketches of
the devices showing the direct SHE mechanism and measurement setup, reciprocal to (a), (c), and (e).

Figure 2: (a) – (f) are the loops recorded in direct or inverse spin Hall effect experiments
corresponding to the measurement schemes shown in Figs. 1(a) – 1(f), respectively. All the curves are
have the same size and an arrow on (c) shows a 5 m signal amplitude. The black arrows represent
the magnetization states. The dashed arrows denote the direction of the applied magnetic field.

Figure 3: Alternative measurement configurations for all three proposed devices. (a) and (b) present
the recorded loop for the sample of Figure 1(a,b), (c) and (d) corresponds to the device of Figure 1(c,d)
and that of (e) and (f) to that of Figure 1 (e,f). The corresponding measurement schemes are shown
in the inset of the figures.

Figure 4: (a) Charge current lines calculated by finite element method simulations for the sample of
Fig 1 (d) and showing the spread of the current line in the Pt strip. A charge current is applied along
the ferromagnetic electrode through the Pt wire in X direction of the T-shape pattern. (b) and (c) are
lateral cut-planes (X-Y cut-plane) of the electrochemical potential contour in the case of + X-
magnetization, at the interface (z=0) and at a distance z = – 7.5 nm. The magnetization directions are
represented by the bright arrows.
Acknowledgements
The devices were fabricated in the Platforme Technologie Amont in Grenoble. We
acknowledge the support from the labex laboratory LANEF of Univ. Grenoble Alpes, and
funding from the ANR TOPRISE.

References

[1] J. E. Hirsch. Phys. Rev. Lett. 83, 1834 (1999).

[2] A. Hoffmann. IEEE Trans. Mag. 49, 5172-5193 (2013).
[3] A. Soumyanarayanan, N. Reyren, A. Fert, and C. Panagopoulos. Nature 539, 509 (2016)
[4] A. Manchon, H. C. Koo, J. Nitta, S. M Frolov, and R. A. Duine. Nature Mater.14, 871 (2015).
[5] Y. Tserkovnyak, A. Brataas, and G.E. Bauer. Phys. Rev. Lett. 88, 117601 (2002).
[6] E. Saitoh, M. Ueda, H. Miyajima, and G. Tatara. Appl. Phys. Lett. 88, 182509 (2006).
[7] C.O. Avci, K. Garello, A. Ghosh, M. Gabureac, S.F. Alvarado, and P. Gambardella. Nat. Phys. 11,
570 (2015).
[8] V. Sih, R.C. Myers, Y.K. Kato, W.H. Lau, A.C. Gossard, and D.D. Awschalom. Nat. Phys. 1, 31 (2005).
[9] J. Wunderlich, B.G. Park, A.C. Irvine, L.P. Zârbo, E. Rozkotová, P. Nemec, V. Novák, J. Sinova, and
T. Jungwirth. Science 330, 1801 (2010).
[10] E. S. Garlid, Q. O. Hu, M. K. Chan, C. J. Palmstrøm, and P. A. Crowell. Phys. Rev. Lett., 2010, 105,
156602.
[11] S. Manipatruni, D. E. Nikonov, and I. A. Young. Nat. Phys. 14, 338–343 (2018)
[12] S. Sayed, S. Hong, E.E. Marinero, and S. Datta. IEEE Electron Device Letters (2017).
[13] I.M. Miron, G. Gaudin, S. Auffret, B. Rodmacq, A. Schuhl, S. Pizzini, J. Vogel, and P. Gambardella.
Nat. Mat. 9, 230-234 (2010)
[14] L. Liu, C.F. Pai, Y. Li, H.W.Tseng, D.C. Ralph, and R.A. Buhrman. Science 336, 555 (2012).
[15] T. Kimura, Y. Otani, T. Sato, S.Takahashi, and S. Maekawa, Phys. Rev. Lett. 98, 156601 (2007).
[16] S. O. Valenzuela, and M. Tinkham. Nature 442, 176 (2006).
[17] V.T. Pham, L. Vila, G. Zahnd, A. Marty, W. Savero-Torres, M. Jamet, and J. P. Attané. Nano Lett.
16, 6755-6760 (2016).
[18] G. Zahnd, L. Vila, V. T. Pham, A. Marty, P. Laczkowski, W.S. Torres, C. Beigné, C. Vergnaud, M.
Jamet, and J.P. Attané. Nanotech. 27, 035201 (2015).
[19] L. Vila, T. Kimura, and Y. Otani. Phys. Rev. Lett. 99, 226604 (2007).
[20] M. Morota, Y. Niimi, K. Ohnishi, D.H. Wei, T. Tanaka, H. Kontani, T. Kimura, and Y. Otani. Phys.
Rev. B 83, 174405 (2011).
[21] W. Savero-Torres, J. F. Sierra, L. A. Benítez, F. Bonell, M. V. Costache, and S. O. Valenzuela. 2D
Mater. 4, 041008 (2017).
[22] W. Yan, E. Sagasta, M. Ribeiro, Y. Niimi, L. E. Hueso, and F. Casanova. Nat. Comm. 8, 661 (2017).
[23 ] C. H. Li, O. M. J. van ‘t Erve, J. T. Robinson, Y. Liu, L. Li & B. T. Jonker. Nat. Nanotech. 9, 218
(2014).
[24] J. Tang, L. Chang, X. Kou, K. I Murata, E. S. Choi, M. Lang, Y. Fan, Y. Jiang, M. Montazeri, W. Jiang,
Y. Wang, L. He, and Kang L. Wang. Nano Lett. 14, 5423 2014
[25] Y. Ando, T. Hamasaki, T. I Kurokawa, K. Ichiba, F. Yang, M. Novak, S. Sasaki, K. Segawa, Y.Ando,
and M. Shiraishi. Nano Lett., 2014, 14 (11), pp 6226–6230
[26] Nagaosa, N., Sinova, J., Onoda, S., MacDonald, A. H., & Ong, N. P. (2010). Anomalous hall effect.
Reviews of modern physics, 82(2), 1539.
[27] T.S. Kim, B.C. Lee, and H.W. Lee.Phys. Rev. B 78, 214427 (2008).
(a) (c) (e)
z y
z y z y
x x x
Jc Jc Jc
VSHE VSHE VSHE

B B B
400 nm 400 nm
400 nm SiO2/Si SiO2/Si
SiO2/Si

(b) (d) (f)

Figs. 1(a), (b), (c), (d), (e) and (f)

 (a) (c) (e)
4015 365
315

310
4010 5 m 360
RISHE (m

RISHE(m
RISHE (m
4005 355
305

4000 350
300

3995 345
-300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300

(b) B (mT) (d)-3370

B (mT) (f) 285 B (mT)
304

-3375 280
300

RSHE(m
RSHE (m

RSHE (m

296 -3380 275

292
-3385 270
288
-3390 265
-300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300
B (mT) B (mT) B (mT)
Figs. 2 (a), (b), (c), (d), (e) and (f)
 (a) (c) -9310 (e)
-5710
-9295 Jc
5 m Jc

[RISHE + RSHE ](m

[RISHE + RSHE ](m

DSHE ISHE -9315 -5715

RSHE(m
-9300 V
Jc F2 V
ISHE DSHE -9320 -5720
-9305

F1 V
-9325 -5725
-300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300
B (mT) B (mT) B (mT)
(b) (d) V (f)
-9580 -8460 -5890 V
V F2

[RISHE + RSHE ](m

[RISHE + RSHE ](m

Jc
-9585 -8465 Jc -5895
RISHE(m

DSHE ISHE
Jc
F1
-9590 ISHE DSHE -8470 -5900

-9595 -8475 -5905

-300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300

B (mT) B (mT) B (mT)

Figs. 3 (a), (b), (c), (d), (e) and (f)

 (a)
V+

z x
y JC

V−
2.5 x 10-5 0.014 7.5
JC (mA/m2)

(b) 400 nm (c)

y y
x z=0 x z = - 7.5 nm
μs (V) 0.0 0.3 0.6
Figs. 4 (a), (b) and (c)

View publication stats