Ferromagnetic van der Waals (vdW) insulators possess robust magnetic order even in a few layers of two-dimensional sheets. The heterostructures of such vdW materials prepared by molecular-beam epitaxy (MBE) are one of the ideal platforms for exploration of novel electronic/spintronic functionalities based on their ferromagnetism via an abrupt hetero-interface. Here we report successful MBE-growth of a vdW magnetic insulator Cr2Ge2Te6 thin film on a topological insulator (Bi,Sb)2Te3. Metal to insulator transition is observed in Cr–Ge–Te alloy films with increasing Ge content as tuned by the Ge and Cr flux ratio, corresponding to the structural phase change from Cr2Te3 to Cr2Ge2Te6. In the nearly stoichiometric Cr2Ge2Te6 films, a large remanent magnetization with perpendicular magnetic anisotropy appears in contrast to the bulk crystals with no discernible hysteresis. The perpendicular remanence with high Curie temperature of about 80 K remains in the thinnest 6-nm film prepared in this study. Designing of magnetic vdW heterostructures based on the Cr2Ge2Te6 thin films offers great opportunities for exploring unusual physical phenomena via proximity effect at the vdW hetero-interface.

Recently, magnetic van der Waals (vdW) materials have attracted renewed interest since the long-range ferromagnetic ordering survives down to a two-dimensional monolayer or a few monolayers in magnetic insulators, such as Cr2Ge2Te6 (Ref. 1) and CrI3 (Ref. 2). The characteristic magnetism leads to new device concepts based on the vdW structure. By applying an exfoliation technique, the electric field effect,3–6 giant magnetoresistance,7,8 and optical control9 in vdW crystal-based heterostructures have been rapidly explored, which envisions high-performance spintronic functionalities in two-dimensional materials.10,11 Significantly, concomitance of ferromagnetism and insulating features in the magnetic vdW materials provides a feasible arena for exploration of the proximity effect in the stacked heterostructures with other vdW materials such as graphene, transition metal dichalcogenides, and topological insulators.10,11 For instance, the vdW magnetic insulator can magnetize the surface of the topological insulator by the proximity effect, potentially driving them into the gap formation of the surface Dirac bands12,13 and/or the efficient magnetization control via spin-momentum locking.14 

Molecular beam epitaxy (MBE) is a powerful technique to fabricate thin films of vdW materials with finely tuned compositions and also to design heterostructures with accurate control of the respective layer thickness.15–17 In fact, topological insulators composed of a layered chalcogenide such as (Bi,Sb)2Te3 have been grown by MBE,18 realizing the quantum anomalous Hall effect in magnetically doped samples grown by tuning their chemical potential and magnetic element composition profile.19–21 Various heterostructures based on (Bi,Sb)2Te3 and ferromagnetic materials also serve for demonstrations of highly efficient charge-to-spin conversion22,23 and anomalous Hall effect24,25 through the interface. Cr2Ge2Te6, which is one of the insulating ferromagnets,1,3,26–29 is a promising vdW material compatible with such layered chalcogenides, although the MBE-growth of Cr2Ge2Te6 has remained challenging toward formation of the vdW-based heterostructure applicable to the future electronic/spintronic devices.

In this study, we report on thin film growth of the ferromagnetic insulator Cr2Ge2Te6 by MBE. In the systematic investigation with various molecular-beam flux ratios of Ge to Cr, we fabricated a series of mixed crystals from metallic Cr2Te3 (Refs. 30–32) to insulating Cr2Ge2Te6 in the Cr–Ge–Te ternary system. The fine tuning of the flux ratio is critical to obtain high resistivity and bulk-comparable Curie temperature (TC) in ferromagnetic Cr2Ge2Te6 films. Surprisingly, the nearly stoichiometric Cr2Ge2Te6 thin film possesses a large remanent magnetization with perpendicular magnetic anisotropy, which is kept even for films as thin as 6 nm.

The crystal structure of Cr2Ge2Te6 is rhombohedral with vdW stacking of hexagonal layers, as shown in Fig. 1(a) (Ref. 26). Such a layered structure is expected to be suitable to form an abrupt interface with other vdW materials, for example, Bi2Te3 and Sb2Te3. Here, we applied an insulating (Bi,Sb)2Te3 [Fig. 1(c)] as a buffer layer to form rhombohedral Te-based layer stacking. Although the (Bi,Sb)2Te3 is a well-known topological insulator possessing a conductive surface and insulating bulk states when it is thick enough (>6 nm), the ultrathin layer (1 ∼ 2 nm) becomes insulating due to the hybridization between the top and bottom surface states.33 By contrast, Cr2Te3 [Fig. 1(b)] has a trigonal structure, which is not a vdW layered compound but contains a similar Te atomic arrangement in the ab-plane. In-plane orientation of Cr2Ge2Te6 (or Cr2Te3) on (Bi,Sb)2Te3 is expected to rotate 30° with respect to the unit cell of (Bi,Sb)2Te3 to match the Te arrangement in the ab-plane, as shown in Figs. 1(d) and 1(e).27,28

FIG. 1.

Schematics of crystal structures for Cr2Ge2Te6 (a), Cr2Te3 (b), and (Bi,Sb)2Te3 (c). In-plane lattice configurations for Cr2Ge2Te6 (d) and (Bi,Sb)2Te3 (e). When Cr2Ge2Te6 is grown on (Bi,Sb)2Te3, the unit cells of them are rotated by 30° in-plane. (f) A cross-sectional schematic of the Cr-Ge-Te thin film on the (Bi,Sb)2Te3/InP substrate. (g) XRD patterns of five 36-nm-thick films with variation in Ge to Cr flux ratios (PGe/PCr = 0, 1, 2, 3, 4). Broad peaks around 17° and 44° possibly come from the diffraction at (006) and (0015) of the (Bi,Sb)2Te3 buffer layer. (h) d-spacing obtained from (g) as a function of PGe/PCr. The upper and lower broken lines represent bulk values for Cr2Ge2Te6 and Cr2Te3, respectively.

FIG. 1.

Schematics of crystal structures for Cr2Ge2Te6 (a), Cr2Te3 (b), and (Bi,Sb)2Te3 (c). In-plane lattice configurations for Cr2Ge2Te6 (d) and (Bi,Sb)2Te3 (e). When Cr2Ge2Te6 is grown on (Bi,Sb)2Te3, the unit cells of them are rotated by 30° in-plane. (f) A cross-sectional schematic of the Cr-Ge-Te thin film on the (Bi,Sb)2Te3/InP substrate. (g) XRD patterns of five 36-nm-thick films with variation in Ge to Cr flux ratios (PGe/PCr = 0, 1, 2, 3, 4). Broad peaks around 17° and 44° possibly come from the diffraction at (006) and (0015) of the (Bi,Sb)2Te3 buffer layer. (h) d-spacing obtained from (g) as a function of PGe/PCr. The upper and lower broken lines represent bulk values for Cr2Ge2Te6 and Cr2Te3, respectively.

Close modal

We grew Cr–Ge–Te alloy thin films on (Bi,Sb)2Te3 buffer layers on semi-insulating InP(111)A (In-termination) substrates, as schematically shown in Fig. 1(f). The films were grown at 180 °C at a base pressure of about 1 × 10−7 Pa. All fluxes were supplied from Knudsen cells, and the beam equivalent pressures, PGe, PCr, and PTe, were monitored. The growth process of the (Bi,Sb)2Te3 layer is described in Ref. 20. The Ge flux ratio against constant Cr flux (PGe/PCr) was carefully determined by their beam equivalent pressures. To suppress Te deficiency, Te was supplied with about 100 times higher pressure than Cr (PTe/PCr > 100). The thickness of the films was determined by x-ray reflectivity measurements. For comparison, bulk single crystals of Cr2Ge2Te6 were also prepared by the Bridgman method (see the supplementary material for the growth method).

Structural characterization for 36-nm-thick films (PGe/PCr = 0, 1, 2, 3, 4) was performed by x-ray diffraction (XRD), as shown in Fig. 1(g). For the film grown under PGe/PCr = 0 (bottom light blue line), the diffraction peak indices (002), (004), and (008) are observed, which indicates c-axis oriented Cr2Te3 with a trigonal structure. For the growth condition of PGe/PCr = 4 (top red line), the film shows clear diffraction peaks at (003), (006), and (0012), corresponding to the rhombohedral Cr2Ge2Te6. In between these two films, the diffraction peak around 14° systematically shifts to the smaller angle with increasing flux ratio of PGe/PCr, indicating elongation of the lattice spacing along the c-axis direction with the compositional variation [Fig. 1(h)]. If the alloyed crystals are in the same crystal structure, such a systematic peak shift is understood as Vegard’s law. In the present case, however, end compounds have different crystal structures. As the origin of such a gradual shift in the d-spacing, we speculate gradual modification of the crystal structure from Figs. 1(b) to 1(a). In other words, a gradual change from Cr2Te3 to Cr2Ge2Te6 is viewed as Ge dimer incorporation in each monolayer with the vdW gap accompanied with the reduction of Cr at the Te–Te interlayer region.

Magnetization and electrical transport were measured in a superconducting quantum interference device magnetometer (MPMS, Quantum Design) and a physical property measurement system (PPMS, Quantum Design), respectively, where magnetic fields (B) are applied perpendicular to the films (B||c) unless otherwise specified. For magnetization measurements, the diamagnetic contributions from the InP substrates were subtracted by their B-linear components above the saturation fields. The temperature dependent magnetization curves (B = 50 mT) of the films are shown in Fig. 2(a). While all the films represent ferromagnetic behaviors, TC widely varies from 170 to 80 K with increasing PGe/PCr. The higher TC’s (∼170 K) for the films with PGe/PCr ∼ 0 and 1 are close to the bulk and thin-film values of Cr2Te3 (Refs. 30–32), while the lower TC’s (∼80 K) for the films with PGe/PCr = 3 and 4 are comparable to the bulk value (60–70 K) of Cr2Ge2Te6 (Refs. 1, 3, 26, 27, and 29). The variation of the magnetization at the lowest temperature (2 K) among these films comes from the differences in remanent magnetization at zero field as well as perpendicular magnetic anisotropy. We note that the slightly higher TC of 80 K for the films may be related to the enhancement of perpendicular magnetic anisotropy as compared to the bulk; the enhanced magnetic anisotropy would result in the suppression of thermal fluctuations for low dimensional magnetism in the MBE-grown films. Magnetization curves at T = 10 K are plotted in Fig. 2(b). Well-developed hysteresis loops are observed in all the films indicating the long-range ferromagnetic order with perpendicular magnetic anisotropy. Moreover, the rectangular shape of the hysteresis loop becomes clearer with increasing PGe/PCr, i.e., toward the Cr2Ge2Te6 end. The saturation magnetizations (Ms) for the four films except for PGe/PCr = 0 (Cr2Te3 film) are around 2.0–2.4 µB/Cr atom, being consistent with the bulk value of 2.2–3 µB/Cr in the Cr2Ge2Te6 crystals.26,27,29 In addition, the reduction of Ms to about 1-2 µB/Cr for the Cr2Te3 film is possibly explained by a small antiferromagnetic contribution from Cr at the Te–Te interlayer region.30,31

FIG. 2.

(a) Temperature (T) dependence of magnetization (M) for 36-nm-thick films grown with various PGe/PCr ratios from 0 to 4 measured under a magnetic field of B = 50 mT applied along the c-axis. (b) Magnetization curves taken at T = 10 K for the respective films shown in (a). (c) T dependence of resistivity (ρxx) for 36-nm-thick films grown with PGe/PCr = 0, 1, 2, 3, 3.2, 4. (d) Curie temperature (TC) as a function of the d-spacing obtained from XRD patterns shown in Fig. 1(d) in comparison with bulk values of Cr2Te3 (Ref. 30) (blue circle) and Cr2Ge2Te6 (Ref. 26) (blue square). The inset shows TC as a function of ρxx at T = 50 K obtained from (c).

FIG. 2.

(a) Temperature (T) dependence of magnetization (M) for 36-nm-thick films grown with various PGe/PCr ratios from 0 to 4 measured under a magnetic field of B = 50 mT applied along the c-axis. (b) Magnetization curves taken at T = 10 K for the respective films shown in (a). (c) T dependence of resistivity (ρxx) for 36-nm-thick films grown with PGe/PCr = 0, 1, 2, 3, 3.2, 4. (d) Curie temperature (TC) as a function of the d-spacing obtained from XRD patterns shown in Fig. 1(d) in comparison with bulk values of Cr2Te3 (Ref. 30) (blue circle) and Cr2Ge2Te6 (Ref. 26) (blue square). The inset shows TC as a function of ρxx at T = 50 K obtained from (c).

Close modal

Figure 2(c) shows the temperature (T) dependence of longitudinal resistivity (ρxx), which strongly depends on the flux ratio of Ge to Cr. In the film grown under PGe/PCr = 0, less temperature dependence of ρxx is observed in accord with the reported features for Cr2Te3 bulk crystals and thin films.30–32 With increasing Ge flux, on the other hand, ρxx increases at a whole temperature range. Indeed, the insulating behavior is pronounced when PGe/PCr is optimized to be 3.2. Judging from the saturation magnetization of 2.4 µB/Cr, TC of about 80 K, and the insulating behavior, the quality or phase-purity of Cr–Ge–Te films grown under PGe/PCr = 3–4 appears comparable to that of the bulk single crystal of Cr2Ge2Te6 (Refs. 1, 3, 26, 27, and 29). In Fig. 2(d), TC for the Cr–Ge–Te films is plotted as a function of the d-spacing along the c-axis direction in comparison with bulk values for Cr2Te3 (blue circle)30 and Cr2Ge2Te6 (blue square).26TC of films with PGe/PCr = 0 and PGe/PCr = 3–4 correspond to that of Cr2Te3 and Cr2Ge2Te6, respectively. However, the TC values for the films grown under various PGe/PCr could not be interpolated with a linear relationship against the d-spacing between the two bulk values for Cr2Te3 and Cr2Ge2Te6. The observed sudden change may originate not only from the structural change where Cr atoms are replaced by Ge dimers to form the vdW gap with weakened Cr–Cr magnetic interaction but also from the carrier-mediated ferromagnetic interaction occurring in metallic Cr2Te3. To see the possible latter effect, we plot TC versus resistivity (ρxx at 50 K) for the respective films in the inset of Fig. 2(d), which confirms the strong correlation between the two quantities.

Hereafter, we focus on the structure and magnetic properties of the highly resistive Cr–Ge–Te film grown under PGe/PCr = 3.2 in detail. Figures 3(a)–3(c) show high-angle annular dark-field (HAADF) images taken by a scanning transmission electron microscopy (STEM). Te, In, Bi and Sb are mainly observed due to their large atomic numbers compared with Cr, Ge, and P. In the fairly large area shown in Figs. 3(a) and 3(b), there is discerned a well-ordered layer-stacking structure with neither dislocation nor segregation. In the magnified image [Fig. 3(c)], a contrast between the interlayer and the intralayer of Cr–Ge–Te layers can be observed, suggesting that the interlayers correspond to the vdW gap. Furthermore, a sharp interface is observed between Cr–Ge–Te and (Bi,Sb)2Te3 layers with the expected crystal-orientation relationship. By performing the fast Fourier transform (FFT) along the lateral direction of the STEM image [Fig. 3(c)], the lateral atomic distance of each layer was estimated, as shown in Fig. 3(d). The observed in-plane Te–Te distances of Cr–Ge–Te [=6.89(5) Å] and (Bi,Sb)2Te3 [=7.38(5) Å] agree well with the reported bulk values of Cr2Ge2Te6 and (Bi,Sb)2Te3 with identical Sb composition.26,27,34 At the interface, the FFT color plot shows a sharp change in the lateral atomic distance, reflecting the character of the vdW interface. Figures 3(e)–3(i) show uniform distribution of the Cr, Ge, Te, In, and P atoms as probed by STEM energy dispersive x-ray spectroscopy (EDX). In Fig. 3(j), the averaged composition fraction profiles of Ge and Te against Cr along the growth direction are plotted. The averaged fractions of Ge/Cr and Te/Cr are approximately unity and three, respectively. Judging from the structural and compositional characterizations, the Cr–Ge–Te film grown under PGe/PCr = 3.2 can be identified to be as the Cr2Ge2Te6 phase. Additionally, the slight deviation of stoichiometry [Fig. 3(j)] and/or lattice expansion [Fig. 2(d)] is one of the possible reasons for the higher TC of about 80 K and the remanent perpendicular magnetization.

FIG. 3.

[(a) and (b)] Cross-sectional HAADF-STEM images of a 36-nm-thick Cr–Ge–Te film on the (Bi,Sb)2Te3/InP substrate protected by an AlOx capping layer. The scale bars correspond to 50 nm (a) and 10 nm (b) in length. (c) Expanded image showing the Cr–Ge–Te/(Bi,Sb)2Te3 interface. The scale bar corresponds to 5 nm. (d) Lateral atomic distance of each layer obtained by FFT of (c) plotted along the growth direction. Elemental distributions of Cr (e), Ge (f), Te (g), In (h), P (i) obtained by EDX. The scale bar corresponds to 5 nm. (j) Averaged compositional fraction profiles for Ge and Te normalized by Cr along the growth direction deduced from [(e)–(g)]. Gray-shaded regions correspond to the InP substrate (lower) and the AlOx capping layer (upper).

FIG. 3.

[(a) and (b)] Cross-sectional HAADF-STEM images of a 36-nm-thick Cr–Ge–Te film on the (Bi,Sb)2Te3/InP substrate protected by an AlOx capping layer. The scale bars correspond to 50 nm (a) and 10 nm (b) in length. (c) Expanded image showing the Cr–Ge–Te/(Bi,Sb)2Te3 interface. The scale bar corresponds to 5 nm. (d) Lateral atomic distance of each layer obtained by FFT of (c) plotted along the growth direction. Elemental distributions of Cr (e), Ge (f), Te (g), In (h), P (i) obtained by EDX. The scale bar corresponds to 5 nm. (j) Averaged compositional fraction profiles for Ge and Te normalized by Cr along the growth direction deduced from [(e)–(g)]. Gray-shaded regions correspond to the InP substrate (lower) and the AlOx capping layer (upper).

Close modal

Finally, we investigate the thickness (t) dependence of magnetic properties for the Cr2Ge2Te6 films grown under the nearly stoichiometric condition with PGe/PCr = 3.2 [Fig. 4(a)]; the film thickness (t = 6, 18, 36 nm) was determined from the x-ray reflectivity fringes shown in Fig. 4(b). Figure 4(c) displays the temperature dependence of magnetization for the films by applying B = 50 mT in comparison with the bulk Cr2Ge2Te6 single crystal. The TC values of the films are almost constant about 80 K, as shown in the inset of Fig. 4(c). Only for the t = 6 nm film, the TC slightly decreases possibly due to a dimensionality effect as also observed for mechanically exfoliated thin flakes.1 One significant finding in the present study is the large remanent magnetization with rectangular hysteresis loops in thin films [Figs. 4(e)–4(g)], in sharp contrast to the magnetization curve of a Cr2Ge2Te6 bulk crystal with no discernible hysteresis [Fig. 4(d)] as is the case for previous studies even in flakes as thin as two monolayers.1,3,26,27,29 The saturation magnetizations Ms of the present thin film samples are about 2.4 µB/Cr atom, almost irrespective of thickness [see the inset of Fig. 4(c)]. By contrast, the coercive fields increase with decreasing thickness; this is contrary to the behavior for conventional ferromagnets with perpendicular magnetic anisotropy because the demagnetization field increases with decreasing thickness.

FIG. 4.

(a) Cross-sectional schematic of t-nm-thick Cr2Ge2Te6 grown with a flux ratio of PGe/PCr = 3.2 on the (Bi,Sb)2Te3/InP substrate. (b) X-ray reflectivity scans and deduced sample thicknesses. (c) Temperature (T) dependence of magnetization (M). Inset: thickness dependence of the Curie temperature (TC) (shown by red circles) and saturation magnetization (Ms) (shown by blue circles). Magnetic field (B) dependence of M normalized by Ms of bulk crystal (d), 36 nm (e), 18 nm (f), and 6 nm (g) films measured at T = 2 K. Magnetic field is applied along the c-axis (shown in corresponding color) and perpendicular to the c-axis (shown in black). Black arrows for Bc indicate the saturation field. Irregular behavior in the B||c magnetization curve in (g) around ±0.1 T is merely the artifact coming from the loss of the magnetization due to the cancellation of the ferromagnetic signal of the film by the diamagnetism of the substrate.

FIG. 4.

(a) Cross-sectional schematic of t-nm-thick Cr2Ge2Te6 grown with a flux ratio of PGe/PCr = 3.2 on the (Bi,Sb)2Te3/InP substrate. (b) X-ray reflectivity scans and deduced sample thicknesses. (c) Temperature (T) dependence of magnetization (M). Inset: thickness dependence of the Curie temperature (TC) (shown by red circles) and saturation magnetization (Ms) (shown by blue circles). Magnetic field (B) dependence of M normalized by Ms of bulk crystal (d), 36 nm (e), 18 nm (f), and 6 nm (g) films measured at T = 2 K. Magnetic field is applied along the c-axis (shown in corresponding color) and perpendicular to the c-axis (shown in black). Black arrows for Bc indicate the saturation field. Irregular behavior in the B||c magnetization curve in (g) around ±0.1 T is merely the artifact coming from the loss of the magnetization due to the cancellation of the ferromagnetic signal of the film by the diamagnetism of the substrate.

Close modal

To reveal the peculiar behavior in the magnetic hysteresis loops, we measured the magnetization by applying the magnetic field along the in-plane direction (Bc) [black lines in Figs. 4(d)–4(g)]. The saturation field (indicated by a black arrow) increases with decreasing thickness. It is consistent with the enhancement of the coercive fields if we take into account a reversal model of simple magnetization rotation. Considering the analogy of the perpendicular magnetic anisotropy as observed for Co/Pt films,35 the enhancement of the magnetic anisotropy may be ascribed to the large spin-orbit coupling in the adjacent (Bi,Sb)2Te3 buffer layer. However, this scenario cannot account for the present case because a Cr2Ge2Te6 film grown directly on an InP substrate also shows a similar rectangular hysteresis loop (see the supplementary material). Other possible scenarios to be considered are as follows: a certain amount of stacking fault/twin formation that may reduce the domain size reserving in-plane magnetic correlations, a possible tensile strain (∼1%) as derived from the TEM image [Figs. 3(c) and 3(d)], charge transfer from the substrate or buffer layer, and a small composition deviation by Ge deficiency perhaps indicated in Fig. 3(j). The tendency of Ge deficiency is a major difference in MBE-grown films from bulk single crystals possibly due to their different growth processes: non-equilibrium and equilibrium growth processes. Incidentally, similar thickness dependent enlargement of hysteresis loops is also observed in the exfoliation samples of ferromagnetic vdW Fe3GeTe2 (Refs. 36 and 37), which may indicate a unique feature in two-dimensional magnetism.

In summary, vdW Cr2Ge2Te6 thin films are prepared by MBE with careful tuning of the Ge/Cr flux ratio under the Te-rich condition. Ge incorporation into Cr2Te3 elongates the d-spacing along the c-axis direction and makes it electrically insulating. Optimized stoichiometric Cr2Ge2Te6 films show the well-defined ferromagnetism with TC of about 80 K and perpendicular magnetic anisotropy even in the thinnest 6-nm film in this study. The perpendicular anisotropic ferromagnetic Cr2Ge2Te6 films with large remanent magnetization are compatible to other chalcogenide vdW materials, which will enable us to fabricate innovative functional devices based on the magnetic proximity effect by the MBE growth method.

See supplementary material for the growth method of Cr2Ge2Te6 bulk single crystals and the influence of the (Bi,Sb)2Te3 buffer layer on magnetism of Cr2Ge2Te6 thin films.

We are grateful to Yoshihiro Iwasa for fruitful discussions and to Kiyou Shibata for helpful advice on STEM analysis. This research was partly supported by JSPS/MEXT Grant-in-Aid for Scientific Research (Nos. 15H05853, 17J03179, and 18H04229) and JST CREST (No. JPMJCR16F1).

1.
C.
Gong
,
L.
Li
,
Z.
Li
,
H.
Ji
,
A.
Stern
,
Y.
Xia
,
T.
Cao
,
W.
Bao
,
C.
Wang
,
Y.
Wang
,
Z. Q.
Qiu
,
R. J.
Cava
,
S. G.
Louie
,
J.
Xia
, and
X.
Zhang
,
Nature
546
,
265
(
2017
).
2.
B.
Huang
,
G.
Clark
,
E.
Navarro-Moratalla
,
D. R.
Klein
,
R.
Cheng
,
K. L.
Seyler
,
D.
Zhong
,
E.
Schmidgall
,
M. A.
McGuire
,
D. H.
Cobden
,
W.
Yao
,
D.
Xiao
,
P.
Jarillo-Herrero
, and
X.
Xu
,
Nature
546
,
270
(
2017
).
3.
W.
Xing
,
Y.
Chen
,
P. M.
Odenthal
,
X.
Zhang
,
W.
Yuan
,
T.
Su
,
Q.
Song
,
T.
Wang
,
J.
Zhong
,
S.
Jia
,
X. C.
Xie
,
Y.
Li
, and
W.
Han
,
2D Mater.
4
,
024009
(
2017
).
4.
S.
Jiang
,
J.
Shan
, and
K. F.
Mak
,
Nat. Mater.
17
,
406
(
2018
).
5.
B.
Huang
,
G.
Clark
,
D. R.
Klein
,
D.
MacNeill
,
E.
Navarro-Moratalla
,
K. L.
Seyler
,
N.
Wilson
,
M. A.
McGuire
,
D. H.
Cobden
,
D.
Xiao
,
W.
Yao
,
P.
Jarillo-Herrero
, and
X.
Xu
,
Nat. Nanotechnol.
13
,
544
(
2018
).
6.
S.
Jiang
,
L.
Li
,
Z.
Wang
,
K. F.
Mak
, and
J.
Shan
,
Nat. Nanotechnol.
13
,
549
(
2018
).
7.
T.
Song
,
X.
Cai
,
M. W.-Y.
Tu
,
X.
Zhang
,
B.
Huang
,
N. P.
Wilson
,
K. L.
Seyler
,
L.
Zhu
,
T.
Taniguchi
,
K.
Watanabe
,
M. A.
McGuire
,
D. H.
Cobden
,
D.
Xiao
,
W.
Yao
, and
X.
Xu
,
Science
360
,
1214
(
2018
).
8.
D. R.
Klein
,
D.
MacNeill
,
J. L.
Lado
,
D.
Soriano
,
E.
Navarro-Moratalla
,
K.
Watanabe
,
T.
Taniguchi
,
S.
Manni
,
P.
Canfield
,
J.
Fernandez-Rossier
, and
P.
Jarillo-Herrero
,
Science
360
,
1218
(
2018
).
9.
D.
Zhong
,
K. L.
Seyler
,
X.
Linpeng
,
R.
Cheng
,
N.
Sivadas
,
B.
Huang
,
E.
Schmidgall
,
T.
Taniguchi
,
K.
Watanabe
,
M. A.
McGuire
,
W.
Yao
,
D.
Xiao
,
K.-M. C.
Fu
, and
X.
Xu
,
Sci. Adv.
5
,
e1603113
(
2017
).
10.
A. K.
Geim
and
I. V.
Grigorieva
,
Nature
499
,
419
(
2013
).
11.
K. S.
Novoselov
,
A.
Mishchenko
,
A.
Carvalho
, and
A. H. C.
Neto
,
Science
353
,
aac9439
(
2016
).
12.
M. Z.
Hasan
and
C. L.
Kane
,
Rev. Mod. Phys.
82
,
3045
(
2010
).
13.
X.-L.
Qi
and
S.-C.
Zhang
,
Rev. Mod. Phys.
83
,
1057
(
2011
).
14.
D.
Pesin
and
A. H.
MacDonald
,
Nat. Mater.
11
,
409
(
2012
).
15.
A.
Koma
,
K.
Sunouchi
, and
T.
Miyajima
,
Microelectron. Eng.
2
,
129
(
1984
).
16.
A.
Koma
,
K.
Saiki
, and
Y.
Sato
,
Appl. Surf. Sci.
41-42
,
451
(
1990
).
17.
S.
Liu
,
X.
Yuan
,
Y.
Zhou
,
Y.
Sheng
,
C.
Huang
,
E.
Zhang
,
J.
Ling
,
Y.
Liu
,
W.
Wang
,
C.
Zhang
,
J.
Zou
,
K.
Wang
, and
F.
Xiu
,
npj 2D Mater. Appl.
1
,
30
(
2017
).
18.
J.
Zhang
,
C.-Z.
Chang
,
Z.
Zhang
,
J.
Wen
,
X.
Feng
,
K.
Li
,
M.
Liu
,
K.
He
,
L.
Wang
,
X.
Chen
,
Q.-K.
Xue
,
X.
Ma
, and
Y.
Wang
,
Nat. Commun.
2
,
574
(
2011
).
19.
C.-Z.
Chang
,
J.
Zhang
,
X.
Feng
,
Z.
Zhang
,
M.
Guo
,
K.
Li
,
Y.
Ou
,
P.
Wei
,
L.-L.
Wang
,
Z.-Q.
Ji
,
Y.
Feng
,
S.
Ji
,
X.
Chen
,
J.
Jia
,
X.
Dai
,
Z.
Fang
,
S.-C.
Zhang
,
K.
He
,
Y.
Wang
,
L.
Lu
,
X.-C.
Ma
, and
Q.-K.
Xue
,
Science
340
,
167
(
2013
).
20.
M.
Mogi
,
R.
Yoshimi
,
A.
Tsukazaki
,
K.
Yasuda
,
Y.
Kozuka
,
K. S.
Takahashi
,
M.
Kawasaki
, and
Y.
Tokura
,
Appl. Phys. Lett.
107
,
182401
(
2015
).
21.
M.
Mogi
,
M.
Kawamura
,
A.
Tsukazaki
,
R.
Yoshimi
,
K. S.
Takahashi
,
M.
Kawasaki
, and
Y.
Tokura
,
Sci. Adv.
3
,
eaao1669
(
2017
).
22.
A. R.
Mellnik
,
J. S.
Lee
,
A.
Richardella
,
J. L.
Grab
,
P. J.
Mintun
,
M. H.
Fischer
,
A.
Vaezi
,
A.
Manchon
,
E.-A.
Kim
,
N.
Samarth
, and
D. C.
Ralph
,
Nature
511
,
449
(
2014
).
23.
K.
Kondou
,
R.
Yoshimi
,
A.
Tsukazaki
,
Y.
Fukuma
,
J.
Matsuno
,
K. S.
Takahashi
,
M.
Kawasaki
,
Y.
Tokura
, and
Y.
Otani
,
Nat. Phys.
12
,
1027
(
2016
).
24.
C.
Tang
,
C.-Z.
Chang
,
G.
Zhao
,
Y.
Liu
,
Z.
Jiang
,
C.-X.
Liu
,
M. R.
McCartney
,
D. J.
Smith
,
T.
Chen
,
J. S.
Moodera
, and
J.
Shi
,
Sci. Adv.
3
,
e1700307
(
2017
).
25.
F.
Katmis
,
V.
Lauter
,
F. S.
Nogueira
,
B. A.
Assaf
,
M. E.
Jamer
,
P.
Wei
,
B.
Satpati
,
J. W.
Freeland
,
I.
Eremin
,
D.
Heiman
,
P.
Jarillo-Herrero
, and
J. S.
Moodera
,
Nature
533
,
513
(
2016
).
26.
V.
Carteaux
,
D.
Brunet
,
G.
Ouvrard
, and
G.
André
,
J. Phys.: Condens. Matter
7
,
69
(
1995
).
27.
H.
Ji
,
R. A.
Stokes
,
L. D.
Alegria
,
E. C.
Blomberg
,
M. A.
Tanatar
,
A.
Reijnders
,
L. M.
Schoop
,
T.
Liang
,
R.
Prozorov
,
K. S.
Burch
,
N. P.
Ong
,
J. R.
Petta
, and
R. J.
Cava
,
J. Appl. Phys.
114
,
114907
(
2013
).
28.
L. D.
Alegria
,
H.
Ji
,
N.
Yao
,
J. J.
Clarke
,
R. J.
Cava
, and
J. R.
Petta
,
Appl. Phys. Lett.
105
,
053512
(
2014
).
29.
X.
Zhang
,
Y.
Zhao
,
Q.
Song
,
S.
Jia
,
J.
Shi
, and
W.
Han
,
Jpn. J. Appl. Phys., Part 1
55
,
033001
(
2016
).
30.
T.
Hashimoto
,
K.
Hoya
,
M.
Yamaguchi
, and
I.
Ichitsubo
,
J. Phys. Soc. Jpn.
31
,
679
(
1971
).
31.
T.
Hamasaki
,
T.
Hashimoto
,
Y.
Yamaguchi
, and
H.
Watanabe
,
Solid State Commun.
16
,
895
(
1975
).
32.
A.
Roy
,
S.
Guchhait
,
R.
Dey
,
T.
Pramanik
,
C.-C.
Hsieh
,
A.
Rai
, and
S. K.
Banerjee
,
ACS Nano
9
,
3772
(
2015
).
33.
Y.
Zhang
,
K.
He
,
C.-Z.
Chang
,
C.-L.
Song
,
L.-L.
Wang
,
X.
Chen
,
J.-F.
Jia
,
Z.
Fang
,
X.
Dai
,
W.-Y.
Shan
,
S.-Q.
Shen
,
Q.
Niu
,
X.-L.
Qi
,
S.-C.
Zhang
,
X.-C.
Ma
, and
Q.-K.
Xue
,
Nat. Phys.
6
,
712
(
2010
).
34.
R. J.
Cava
,
H.
Ji
,
M. K.
Fuccillo
,
Q. D.
Gibson
, and
Y. S.
Hor
,
J. Mater. Chem. C
1
,
3176
(
2013
).
35.
M. T.
Johnson
,
P. J. H.
Bloemen
,
F. J. A.
den Broeder
, and
J. J.
de Vries
,
Rep. Prog. Phys.
59
,
1409
(
1996
).
36.
C.
Tan
,
J.
Lee
,
S.-G.
Jung
,
T.
Park
,
S.
Albarakati
,
J.
Partridge
,
M. R.
Field
,
D. G.
MacCulloch
,
L.
Wang
, and
C.
Lee
,
Nat. Commun.
9
,
1554
(
2018
).
37.
Z.
Fei
,
B.
Huang
,
P.
Malinowski
,
W.
Wang
,
T.
Song
,
J.
Sanchez
,
W.
Yao
,
D.
Xiao
,
X.
Zhu
,
A.
May
,
W.
Wu
,
D.
Cobden
,
J.-H.
Chu
, and
X.
Xu
,
Nat. Mater.
17
,
778
(
2018
).

Supplementary Material