We investigated spin-Hall effect (SHE) and degree of MgO (100) orientation in artificially synthesized (W/Hf)-multilayer/CoFeB/MgO systems with various W thicknesses. We found that the artificially synthesized multilayer systems can enhance the spin-Hall effect and control the value of spin diffusion length. We observed a maximum magnitude in both spin-Hall angle and spin-Hall conductivity as a function of W thickness in W/Hf-multilayer systems, and found that the values of spin-Hall conductivity are larger than that for β-phase W. In addition, a more highly oriented MgO (100) texture on CoFeB is obtained for (W/Hf)-multilayer systems prepared under low-Ar-pressure condition, which would be suitable for preparation of magnetic tunnel junctions with high tunnel magnetoresistance properties on (W/Hf)-multilayer heavy metal electrode. These results suggest that the artificially synthesized multilayer system is one of the avenues for realizing spin devices using spin-orbit torque.

Current-induced spin-orbit torque (SOT) originating from the spin-Hall effect (SHE) in heavy metal/ferromagnet (HM/FM) systems has attracted attention due to their potential for application to SOT magnetoresistive random access memory (SOT-MRAM), skyrmion and domain wall devices.1–18 Study of HM materials as well as HM/FM interfaces with larger spin-orbit coupling is being actively carried out because of allowing a larger amount of spin current (Js) to be generated for manipulating the nano magneto when flowing the write charge current (JC) through the HM layers. For the application to a large-scale integration, the efficient SOT operation (absolute value of high spin Hall angle |θSH|) in low resistivity (ρxx) HM is necessary.19–21 The low |θSH| and high ρxx lead to an undesirably large energy dissipation, delay in speed, and large voltage drop during current flow in HMs. Magnitudes of the |θSH| = |JS/JC| have been determined for various HMs by measuring the spin-Hall magnetoresistance (SMR) and spin torque ferromagnetic resonance (ST-FMR)19–26 and by other means. Due to the extensive efforts, the efficiency of present SOT operations, that is, |θSH| becomes larger day by day, however, almost all HMs have large resistivity. For example, β-phase W (β-W) has a relatively large |θSH| of approximately 0.2 - 0.3, however, β-W have a very high ρxx.1,12,21,22,25–28 The magnitude of |θSH| for both intrinsic and extrinsic (side jump mechanism) terms is proportional to the magnitude of ρxx value (|θSH| ∼ σSHρxx27,28), where σSH is spin Hall conductivity. Therefore, increase in the magnitude of σSH is important from the application point of view.21 Recently, we observed that large magnitude of σSH and enhancement of perpendicular magnetic anisotropy in (W (tW)/Hf (tHf))-multilayer/CoFeB/MgO systems with tW = tHf = 0.35 nm and 0.7 nm compare to β-W/CoFeB/MgO system.21 In the previous work, we also found magnitude of σSH for (W (0.7)/Hf (0.7))-multilayer system is larger than that for (W (0.35)/Hf (0.35))-multilayer system. Therefore, when the film thickness ratio between W and Hf is optimized, further increase of σSH is expected in the (W/Hf)-multilayer system.

In this paper, we investigated W thickness dependence of σSH, θSH,ρxx and spin diffusion length (λS) in amorphous (W (tW)/Hf (0.35))-multilayer/CoFeB/MgO systems with various tW and evaluated the degree of MgO (100) orientation on the (W/Hf)-multilayer HM electrode.

We prepared Ta(0.5)/artificially-synthesized (W(tW)/Hf(0.35))n multilayer (tHM)/CoFeB(tCoFeB)/MgO(1.0)/Ta(1) (n: repetition number) systems with various HM thicknesses (tHM) on high resistive Si substrates. The sputtering Ar gas pressure (PAr) for W in W/Hf multilayers employed 2.55 Pa (high-Ar-pressure condition) and 0.39 Pa (low-Ar-pressure condition), which are β- and a-phases preparation conditions in W deposition, respectively, as reported previously.19 The (W (tW)/Hf (0.35))-multilayer systems have amorphous structure as described in Ref. 21. These systems with various tHM (= 1.1∼8.4 nm) are patterned into the microscale Hall bar by photolithography and Ar ion milling. Detailed fabrication process was described elsewhere.19 The processed wafers were then annealed at 573 K in vacuum less than 1×10-4 Pa for an hour. SHE in these devices with various tHM was measured at 305 K by means of SMR. For the measurements of SMR, the current, which is less than equal to 5 μA, is passed through the devices in the x-axis direction and external magnetic field between -4 and +4 Tesla is applied to the both y- and z-axes directions in inset of Fig. 2(c). For all films, the saturation magnetization (MS) value of Co20Fe60B20 is ∼1.5 × 106 A/m. This value is nearly consistent with the nominal Co20Fe60B20 saturation magnetization.29 We also confirmed that the values of interfacial anisotropy (Ki) are nearly same value (about 1.45 [×10-3 J/m2]) for (W/Hf)- multilayer/CoFeB/MgO systems with various tW, which are much larger than that for β-W/CoFeB/MgO system. These Ki values are also consistent with previous results.21 The reason of the nearly same Ki value for samples with various tW would be originating from having the same interface structure of Hf(0.35)/CoFeB/MgO for the all (W/Hf)-multilayer/CoFeB/MgO systems prepared here.

The film structure for (W(0.7)/Hf(0.35))5/CoFeB(1.5)/MgO(1.0)/Ta(1.0) prepared by low- PAr condition was confirmed by high-resolution transmission electron microscopy (HR-TEM) image (Fig. 1(a)). Degree of the texture in MgO (1.0) layer for all systems prepared here were also investigated by reflection high energy electron diffraction (RHEED). As shown in Fig. 1(a), each film for (W(0.7)/Hf(0.35))-multilayer/CoFeB/MgO system is very flat. Figures 1(c)1(e) show the typical RHEED patterns for MgO (1.0 nm) on CoFeB (3.0) in β-W (7.0)/CoFeB (3.0)/MgO (1.0) and (W (1.0)/Hf (0.35))5 multilayer/CoFeB (3.0)/MgO (1.0) systems prepared in the conditions of high PAr and low PAr, respectively. Polycrystalline nature of MgO (1.0 nm) on CoFeB (3nm) was observed in the β-W (7nm) system (Fig. 1(c)). On the other hands, as shown in Figs. 1(d) and 1(e), the (100) oriented texture of MgO(1.0 nm) on CoFeB (3 nm) were observed in (W(1.0)/Hf(1.0))5-multilayer systems prepared by high PAr and low PAr. This result is consistent with the HR-TEM images. Rough interface between β-W and CoFeB and not clear texture of MgO (1.0 nm) on CoFeB were observed in β-W (7)/CoFeB (1.5)/MgO (1.0) system from the HR-TEM images.21 The degree of MgO(100) orientation is better for W/Hf-multilayer system prepared by low PAr condition (Fig. 1(e)) compared to the case of high PAr condition (Fig. 1(d)). The (100) oriented texture of MgO(1.0 nm) were observed in the W thickness range of tW ≤ 1.5 and tW ≤ 1.0 for low PAr and high PAr conditions, respectively (not shown). This result indicates that the (W/Hf)-multilayer HM electrode is suitable for preparation of MgO (100) oriented magnetic tunnel junctions with high tunnel magnetoresistance properties.

FIG. 1.

(a) Cross-sectional transmission electron microscopy image for (W0.7/Hf0.35)-multilayer/CoFeB(1.5)/MgO(1.0)/Ta(1.0) system. (b) Cross-section view of the sputtered film stacks for artificial (W/Hf)-multilayer/CoFeB/MgO systems. The numbers in the parenthesis show the nominal thickness in nm. (c)-(e) are reflection high energy electron diffraction (RHEED) patterns of MgO (1.0) on CoFeB (3.0) for (c) β-W (7.0) and (W (1.0)/Hf (0.35))5 multilayer systems prepared by (d) high Ar and (e) low Ar pressures.

FIG. 1.

(a) Cross-sectional transmission electron microscopy image for (W0.7/Hf0.35)-multilayer/CoFeB(1.5)/MgO(1.0)/Ta(1.0) system. (b) Cross-section view of the sputtered film stacks for artificial (W/Hf)-multilayer/CoFeB/MgO systems. The numbers in the parenthesis show the nominal thickness in nm. (c)-(e) are reflection high energy electron diffraction (RHEED) patterns of MgO (1.0) on CoFeB (3.0) for (c) β-W (7.0) and (W (1.0)/Hf (0.35))5 multilayer systems prepared by (d) high Ar and (e) low Ar pressures.

Close modal

Figures 2(a) and 2(b) show the sheet conductance (Gxx = L/(wRxx)) as a function of the HM layer thickness (tHM) in (W(h)/Hf)-multilayer and (W(l)/Hf)-multilayer systems, respectively, with reference result of β-W system,21 which also show the result of β-W for comparison. The values of L and w in devices are L = 205 μm and w = 5.1 μm as shown in inset of Fig. 2(c). The W (h) and W (l) mean tungsten (W) films prepared at the high (h)-Ar-pressure (PAr=2.55 Pa) and low (l)-Ar-pressure (PAr=0.39 Pa) conditions, respectively. Since the slope in Figs. 2(a) and 2(b) is the inverse of the resistivity of HM (1/ρxx), we can see from Figs. 2(a) and 2(b) that the resistivity ρxx values for the all (W/Hf)-multilayer systems are smaller than that for β-W. As shown in Fig. 2(b), there are phase transition from amorphous phase to that containing α-phase W for the device with (W(l) (tW)/Hf (0.35))-multilayer systems with tW = 1.0 and 1.2 nm. The phase transition thicknesses (tT) are tT ∼ 3.9, 3.8 nm for (W(l) (tW)/Hf (0.35))-multilayer systems with tW = 1.0 and 1.2 nm, respectively, whereas there is no anomaly for other devices. From here, we use the data of less than tT in (W (l) (tW)/Hf (0.35))-multilayer systems with tW=1.0 and 1.2 nm for analyzing the value of ρxx and data of SMR for estimating the |θSH|, σSH and spin diffusion length (λS). The values of the resistivity (ρxx) for (W(h) (tW)/Hf (0.35))-multilayer (tW = 0.35, 0.7, 1.0 nm) and (W(l) (tW)/Hf (0.35))-multilayer (tW = 0.35, 0.7, 1.0, 1.2 nm) systems as a function of W thickness (tW) are shown in Fig. 2(c). The estimated value of ρxx for β-W obtained by the least-square-fit in Fig. 2(a) is also plotted in the Fig. 2(c) (red line). As shown in Fig. 2(c), the resistivity values for (W/Hf)-multilayer systems are smaller than that for β-W system and slightly decreases monotonically with increasing tW (see the black solid line in Fig. 2(c)). We did not prepare (W(h) (1.2)/Hf (0.35))-multilayer system, because we could not observe MgO(100) texture in (W(h) (1.2)/Hf (0.35))5 multilayer/CoFeB/MgO system (MgO has an amorphous structure). Because the metastable polycrystalline β-W has a higher resistance, amorphous W has a lower resistance compared to that of β-W. The ρxx values are nearly the same between artificially synthesized (W(h) (tW)/Hf (0.35)) and (W(l) (tW)/Hf (0.35))-multilayers prepared at PAr=2.55 Pa and PAr=0.39 Pa as shown in Fig. 2(c). This would because the multilayers have a same amorphous structure.

FIG. 2.

(a) (b) show sheet conductance (Gxx) as a function of HM thickness (tHM). The solid lines in (a) and (b) are linear fits to the data. (c) shows the estimated resistivity (ρXX) as a function of W thickness. The black solid line in (c) is the result of linear fit to the data and the red solid line in (c) is the plot of ρxx for β-W obtained by the least-square-fit in (a). Inset in (c) is schematic diagram of a prepared device.

FIG. 2.

(a) (b) show sheet conductance (Gxx) as a function of HM thickness (tHM). The solid lines in (a) and (b) are linear fits to the data. (c) shows the estimated resistivity (ρXX) as a function of W thickness. The black solid line in (c) is the result of linear fit to the data and the red solid line in (c) is the plot of ρxx for β-W obtained by the least-square-fit in (a). Inset in (c) is schematic diagram of a prepared device.

Close modal

Figures 3(a) and 3(b) show the typical longitudinal resistance (Rxx) versus external magnetic field (H) measured at 305 K for the devices with amorphous (W(h) (1.0)/Hf (0.35)) and amorphous (W(l) (1.0)/Hf (0.35))-multilayer systems, respectively. As shown in Figs. 3(a) and 3(b), the values of Rxx in the magnetic field directions along z-axis: Hz > 0 T and Hz < 0 T are nearly the same (for example, Rxx (Hz = 4 T) ∼ Rxx (Hz = -4 T)), however, the values of Rxx in the magnetic field directions along y-axis: Hy > 0 T and Hy < 0 T are different from each other for the both devices with amorphous (W(h) (1.0)/Hf (0.35)) and amorphous (W(l) (1.0)/Hf (0.35))-multilayer systems. For both devices with amorphous (W (1.0)/Hf (0.35))-multilayers, the value of Rxx at Hy = 4 T is smaller than that at Hy = -4 T. These are related to the anomalous Nernst voltage (VNernst) due to the thermal hot electron current flow from the film to high resistive Si substrate as discussed in Ref. 19. The degree of the difference of Rxx values between for Hy > 0 T and Hy < 0 is smaller in the case of (W(l) (1.0)/Hf (0.35))-multilayer system compared with the case of (W(h) (1.0)/Hf (0.35))-multilayer systems. This would be related to the slight difference of the absolute value of Rxx for between (W(l) (1.0)/Hf (0.35)) and (W(h) (1.0)/Hf (0.35))-multilayers as shown in Figs. 3(a) and 3(b), because we flowed the same current value of I = 3 μA during measurements. The VNernst sign for amorphous W/Hf multilayers is the same with that for crystalline W systems.19 Therefore, the current would mainly flow in the amorphous W in (W/Hf)-multilayer systems.

FIG. 3.

Typical longitudinal resistance Rxx versus external magnetic field H oriented along the y axis (open and closed circles) and z axis (open and closed rectangulars) measured at 305 K for the device with (W(1.0)/Hf(0.35))-multilayers prepared (a) high and (b) low-Ar-pressures. SMR ΔRXX/RXXH=0 plotted against the HM layer thickness (tHM) for (W/Hf)-multilayer systems prepared by (c) high Ar gas pressure and (d) low Ar gas pressure. The solid lines show the fitting results using drift diffusion model.

FIG. 3.

Typical longitudinal resistance Rxx versus external magnetic field H oriented along the y axis (open and closed circles) and z axis (open and closed rectangulars) measured at 305 K for the device with (W(1.0)/Hf(0.35))-multilayers prepared (a) high and (b) low-Ar-pressures. SMR ΔRXX/RXXH=0 plotted against the HM layer thickness (tHM) for (W/Hf)-multilayer systems prepared by (c) high Ar gas pressure and (d) low Ar gas pressure. The solid lines show the fitting results using drift diffusion model.

Close modal

In order to neglect the anomalous Nernst effect to analyze the SMR, we define the SMR by19,21

SMR=ΔRXX/RXXH=0=ΔRXX1+ΔRXX2/2RXXH=0,
(1)
ΔRXX1=RXXHy=1.6TRXXHZ=1.6T,
(2)
ΔRXX2=RXXHy=+1.6TRXXHZ=+1.6T,
(3)

where RXXH=0 is the longitudinal resistance at H = 0 T.

We used the values of RXX at |H| = 1.6 T, which is the saturation magnetic field value for CoFeB in the magnetic hard-axis direction, for the estimation of SMR, because we think that the slight increase in ΔRXX with increasing |H| above 1.6 T may originate from contribution of the Hanle magnetoresistance.30,31

Figures 3(c) and 3(d) show the ΔRXX/RXXH=0 as a function of tHM for (W(h) (tW)/Hf (0.35))-multilayer systems (tW = 0.35, 0.7, 1.0) and (W(l) (tW)/Hf (0.35))-multilayer systems (tW = 0.35, 0.7, 1.0, 1.2), respectively. The solid lines in Figs. 3(c) and 3(d) are the results fitted the measured data by using the equations:22,26

SMR=ΔRXX/RXXH=0θSH2λStHMtanh(tHM/2λS)1+ξ×11cosh(tHM/λS),
(4)
ξρHMtCoFeBρCoFeBtHM,
(5)

where λS is spin diffusion length and ρCoFeB = 139.9 μΩcm and ρHM are the resistivity estimated by the least-square-fitting shown in Figs. 2(a) and 2(b). As shown in Figs. 3(c) and 3(d), the thicknesses values of tHM at which minimum magnitude of ΔRXX/RXXH=0 for the fitted solid lines in amorphous (W (tW)/Hf (0.35))-multilayer systems with small tW are thinner than those for thick tW. This indicates the λS values in (W/Hf)-multilayers systems increase with increasing tW in W (tW)/Hf (0.35) multilayer systems. The applied SMR model is based on the drift diffusion model,32 therefore the estimated θSH and the λs are all effective values. The magnitudes of |θSH| and λS of the amorphous (W/Hf)-multilayer HM electrodes are successfully obtained as shown next.

Figures 4(a)-4(c) show the results of the magnitudes of σSH, |θSH| and λS as functions of tW and artificial-cycle-film thickness (tAFC = tW + 0.35 nm) for W/Hf multilayer systems. The estimated values of σSH, |θSH| and λS for β-W21 are also plotted in the Figs. 4(a)-4(c) (red lines). A maximum magnitude in both σSH and |θSH| as a function of tW are observed for (W (tW)/Hf (0.35))-multilayer systems. Thus, we observed σSH values is 15% higher than the previous value,21 when investing the tW dependence detailly. As shown in Fig. 4(a), the all estimated values of σSH for (W/Hf)-multilayer systems are larger than that for β-W. This is due to the lower resistivity values for (W/Hf)-multilayer systems compared to that for β-W as shown in Fig. 2(c). The maximum magnitude of θSH value for amorphous (W (tW)/Hf (0.35))-multilayer systems is -0.21 for tW = 0.7, 1.0 nm. We found that the magnitude of |θSH| for amorphous (W/Hf)-multilayer systems with tW = 0.7, 1.0 nm is nearly the same with that for β-W (θSH = -0.207) and found the decrease in |θSH| for tW = 1.2 nm (θSH = -0.17).

FIG. 4.

(a) Estimated magnitude of the spin Hall conductivity (σSH) (open and closed black circles), (b) spin Hall angle |θSH| (open and closed blue rectangulars) and (c) spin diffusion length (λs) (open and closed light blue triangles) as functions of W thickness (tW) and artificial-cycle-film thickness (tAFC) for the (W/Hf)-multilayer systems. The solid black and dark and light blue lines in (a)-(c) are guides for the eyes. The red solid lines in (a)-(c) are the plots of σSH, |θSH| and λs for β-W system.

FIG. 4.

(a) Estimated magnitude of the spin Hall conductivity (σSH) (open and closed black circles), (b) spin Hall angle |θSH| (open and closed blue rectangulars) and (c) spin diffusion length (λs) (open and closed light blue triangles) as functions of W thickness (tW) and artificial-cycle-film thickness (tAFC) for the (W/Hf)-multilayer systems. The solid black and dark and light blue lines in (a)-(c) are guides for the eyes. The red solid lines in (a)-(c) are the plots of σSH, |θSH| and λs for β-W system.

Close modal

As show in Fig. 4(c), we also found that the value of λs increases with increasing tW. This would correlate with the tAFC in amorphous (W (tW)/Hf (0.35))-multilayer systems. The estimated value of λs in (W (1.2)/Hf (0.35))-multilayer system is nearly same with that in β-W (λs =1.05 nm) (red line in Fig. 4(c)). The low λs for (W/Hf)-multilayer systems with tW = 0.35, 0.7, 1.0 would be related to the increase in the interfacial scattering of multilayer systems, and the nearly same in the λs for (W/Hf)-multilayer system with tW = 1.2 would indicate the decrease in the interfacial scattering of multilayer systems due to the thick in W thickness. As shown in Figs. 4(a) and 4(b), it can be also seen that the spin-Hall effects (both σSH and |θSH|) decrease at the film thickness where the scattering mechanism at the interface disappears (the tAFC is thicker than equal to 1.2+0.35 nm (tAFC ≥ 1.55 nm)). These results clearly suggest that artificially synthesized multilayer system can enhance the spin-Hall effect and control the value of λs.

We prepared artificially synthesized (W/Hf)-multilayer/CoFeB/MgO systems and observed a maximum magnitude in spin Hall effect as a function of tW in amorphous (W (tW)/Hf (0.35))-multilayer systems. We found 15% enhancement of the magnitude of σSH by investigating the tW dependence of spin Hall effect. In addition, we found that the value of λs correlates with the artificial-cycle-film thickness in (W (tW)/Hf (0.35))-multilayer systems. These results clearly suggest that the artificially synthesized multilayer system can enhance the spin-Hall effect and control the value of λs. We also found that the degree of (100) oriented texture of MgO is higher for the artificially (W/Hf)-multilayer systems prepared by low Ar pressure condition, which would be better for preparation of MTJs with (W/Hf)-multilayer HM electrode in order to realize a high tunnel magnetoresistance resulting from coherent tunneling of Δ1 electrons through the MgO (100) barrier. These results suggest that artificially synthesized multilayer system is one of the avenues for realizing spin devices using spin-orbit torque.

This work was supported by the JST OPERA (JPMJOP1611) and JSPS KAKENHI (19H00844).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
L.
Liu
,
C.-F.
Pai
,
Y.
Li
,
H. W.
Tseng
,
D. C.
Ralph
, and
R. A.
Buhrman
,
Science
336
,
555
(
2012
).
2.
G.
Yu
,
P.
Upadhyaya
,
Y.
Fan
,
J. G.
Alzate
,
W.
Jiang
,
K. L.
Wong
,
S.
Takei
,
S. A.
Bender
,
L.-T.
Chang
,
Y.
Jiang
,
M.
Lang
,
J.
Tang
,
Y.
Wang
,
Y.
Tserkovnyak
,
P. K.
Amiri
, and
K. L.
Wang
,
Nat. Nanotechnol.
9
,
548
(
2014
).
3.
W.
Jiang
,
P.
Upadhyaya
,
W.
Zhang
,
G.
Yu
,
M. B.
Jungfleisch
,
F. Y.
Fradin
,
J. E.
Pearson
,
Y.
Tserkovnyak
,
K. L.
Wang
,
O.
Heinonen
,
S. G. E.
te Velthuis
, and
A.
Hoffmann
,
Science
349
,
283
(
2015
).
4.
P. P. J.
Haazen
,
E.
Murè
,
J. H.
Franken
,
R.
Lavrijsen
,
H. J. M.
Swagten
, and
B.
Koopmans
,
Nat. Mater.
12
,
299
(
2013
).
5.
A.
Chernyshov
,
M.
Overby
,
X.
Liu
,
J. K.
Furdyna
,
Y.
Lyanda-Geller
, and
L. P.
Rokhinson
,
Nat. Phys.
5
,
656
(
2009
).
6.
I. M.
Miron
,
K.
Garello
,
G.
Gaudin
,
P.-J.
Zermatten
,
M. V.
Costache
,
S.
Auffret
,
S.
Bandiera
,
B.
Rodmacq
,
A.
Schuhl
, and
P.
Gambardella
,
Nature
476
,
189
(
2011
).
7.
J.
Kim
,
J.
Sinha
,
M.
Hayashi
,
M.
Yamanouchi
,
S.
Fukami
,
T.
Suzuki
,
S.
Mitani
, and
H.
Ohno
,
Nat. Mater.
12
,
240
(
2013
).
8.
S.
Fukami
,
T.
Anekawa
,
C.
Zhang
, and
H.
Ohno
,
Nat. Nanotechnol.
11
,
621
(
2016
).
9.
K.-S.
Lee
,
S.-W.
Lee
,
B.-C.
Min
, and
K.-J.
Lee
,
Appl. Phys. Lett.
104
,
072413
(
2014
).
10.
K.
Garello
,
C. O.
Avci
,
I. M.
Miron
,
M.
Baumgartner
,
A.
Ghosh
,
S.
Auffret
,
O.
Boulle
,
G.
Gaudin
, and
P.
Gambardella
,
Appl. Phys. Lett.
105
,
212402
(
2014
).
11.
C.
Zhang
,
S.
Fukami
,
H.
Sato
,
F.
Matsukura
, and
H.
Ohno
,
Appl. Phys. Lett.
107
,
012401
(
2015
).
12.
M.-H.
Nguyen
,
C. F.
Pai
,
K. X.
Nguyen
,
D. A.
Muller
,
D. C.
Ralph
, and
R. A.
Buhrman
,
Appl. Phys. Lett.
106
,
222402
(
2015
).
13.
S.
Fukami
,
T.
Anekawa
,
C.
Zhang
, and
H.
Ohno
,
Nat. Nanotechnol.
11
,
621
(
2016
).
14.
S. V.
Aradhya
,
G. E.
Rowlands
,
J.
Oh
,
D. C.
Ralph
, and
R. A.
Buhrman
,
Nano Lett.
16
,
5987
(
2016
).
15.
M.
Baumgartner
,
K.
Garello
,
J.
Mendil
,
C. O.
Avci
,
E.
Grimaldi
,
C.
Murer
,
J.
Feng
,
M.
Gabureac
,
C.
Stamm
,
Y.
Acremann
,
S.
Finizio
,
S.
Wintz
,
J.
Raabe
, and
P.
Gambardella
,
Nanotechnol
12
,
980
(
2017
).
16.
Y.
Kato
,
Y.
Saito
,
H.
Yoda
,
T.
Inokuchi
,
S.
Shirotori
,
N.
Shimomura
,
S.
Oikawa
,
A.
Tiwari
,
M.
Ishikawa
,
M.
Shimizu
,
B.
Altansargai
,
H.
Sugiyama
,
K.
Koi
,
Y.
Ohsawa
, and
A.
Kurobe
,
Phys. Rev. Appl.
10
,
044011
(
2018
).
17.
H.
Honjo
,
T. V. A.
Nguen
,
T.
Watanabe
,
T.
Nasuno
,
C.
Zhang
,
T.
Tanigawa
,
S.
Miura
,
H.
Inoue
,
M.
Niwa
,
T.
Yoshiduka
,
Y.
Noguchi
,
M.
Yasuhira
,
A.
Tamakoshi
,
M.
Natsui
,
Y.
Ma
,
H.
Koike
,
Y.
Takahashi
,
K.
Furuya
,
H.
Shen
,
S.
Fukami
,
H.
Sato
,
S.
Ikeda
,
T.
Hanyu
,
H.
Ohno
, and
T.
Endoh
,
IEDM Tech. Dig.
28
,
5
(
2019
).
18.
M.
Natsui
,
A.
Tamakoshi
,
H.
Honjo
,
T.
Watanabe
,
T.
Nasuno
,
C.
Zhang
,
T.
Tanigawa
,
H.
Inoue
,
M.
Niwa
,
T.
Yoshiduka
,
Y.
Noguchi
,
M.
Yasuhira
,
Y.
Ma
,
H.
Shen
,
S.
Fukami
,
H.
Sato
,
S.
Ikeda
,
H.
Ohno
,
T.
Endoh
, and
T.
Hanyu
, in
2020 Symposia on VLSI Technology and Circuits, CM2.2
,
2020
.
19.
Y.
Saito
,
N.
Tezuka
,
S.
Ikeda
,
H.
Sato
, and
T.
Endoh
,
Appl. Phys. Exp.
12
,
053008
(
2019
).
20.
Y.
Saito
,
N.
Tezuka
,
S.
Ikeda
,
H.
Sato
, and
T.
Endoh
,
AIP. Advances
9
,
125312
(
2019
).
21.
Y.
Saito
,
N.
Tezuka
,
S.
Ikeda
,
H.
Sato
, and
T.
Endoh
,
Appl. Phys. Lett.
116
,
132401
(
2020
).
22.
J.
Liu
,
T.
Ohkubo
,
S.
Mitani
,
K.
Hono
, and
M.
Hayashi
,
Appl. Phys. Lett.
107
,
232408
(
2015
).
23.
C.-F.
Pai
,
L.
Liu
,
Y.
Li
,
H. W.
Tseng
,
D. C.
Ralph
, and
R. A.
Buhrman
,
Appl. Phys. Lett.
101
,
122404
(
2012
).
24.
A.
Ganguly
,
K.
Kondou
,
H.
Sukegawa
,
S.
Mitani
,
S.
Kasai
,
Y.
Niimi
,
Y.
Otani
, and
A.
Barman
,
Appl. Phys. Lett.
104
,
072405
(
2014
).
25.
Y.
Wang
,
P.
Deorani
,
X.
Qiu
,
J. H.
Kwon
, and
H.
Yang
,
Appl. Phys. Lett.
105
,
152412
(
2014
).
26.
J.
Kim
,
P.
Sheng
,
S.
Takahashi
,
S.
Mitani
, and
M.
Hayashi
,
Phys. Rev. Lett.
116
,
097201
(
2016
).
27.
C.
Zhang
,
S.
Fukami
,
K.
Watanabe
,
A.
Ohkawara
,
S.
DuttaGupta
,
H.
Sato
,
F.
Matsukura
, and
H.
Ohno
,
Appl. Phys. Lett.
109
,
192405
(
2016
).
28.
L.
Wang
,
R. J. H.
Wesselink
,
Y.
Liu
,
Z.
Yuan
,
K.
Xia
, and
P. J.
Kelly
,
Phys. Rev. Lett.
116
,
196602
(
2016
).
29.
R. M.
Bozorth
,
Ferromagnetism
(
Wiley-IEEE Press
,
New York
,
1993
).
30.
M. I.
Dyakonov
,
Phys. Rev. Lett.
99
,
126601
(
2007
).
31.
S.
Velez
,
V. N.
Golovach
,
A.
Bedoya-Pinto
,
M.
Isasa
,
E.
Sagasta
,
M.
Abadia
,
C.
Rogero
,
L. E.
Hueso
,
F. S.
Bergeret
, and
F.
Casanova
,
Phys. Rev. Lett.
116
,
016603
(
2016
).
32.
V. P.
Amin
and
M. D.
Stiles
,
Phys. Rev. B
94
,
104420
(
2016
).