We evaluate the heterogeneous integration of the layered correlated electron material, 1T-TaS2, on semiconducting 2H-MoS2 for the realization of an all two-dimensional insulator-to-metal (IMT) phase transition device. First principles calculations investigate the band structure of the resulting heterostructure and confirm the existence of a charge density wave (CDW)-based bandgap. 1T-TaS2 films are synthesized via powder vapor deposition on monolayer MoS2 substrates and shown to exhibit CDW induced IMT phase transitions. Both Raman and electrical measurements display reversible commensurate to nearly commensurate CDW IMT phase transitions. Finally, a phase transition transistor device is demonstrated that harnesses the electrically triggered abrupt IMT in 1T-TaS2 and semiconducting properties of 2H-MoS2.

Materials exhibiting correlated phenomena such as electronic phase transitions are of interest for next generation energy efficient and high performance electronics applications.1–3 An electrically triggered insulator-to-metal (IMT) phase transition in correlated material systems has been coupled with conventional metal-oxide semiconductor field effect transistors (MOSFETs) to display sub-thermal (kT/q) switching slope,4–7 spiking neurons,8,9 truly random number generators,10,11 and coupled oscillator arrays.12–15 Most of the work has focused on complex oxide systems such as VO2, NbOx, and TaOx. Two-dimensional (2D) materials such as 1T-TaS2, and NbSe2 are also known to exhibit strongly correlated phenomena such as charge density waves and superconductivity.16 Unlike complex oxides, which can only marginally alter phase transition properties through strain or doping effects,17–22 the intrinsic layered nature of 2D materials allows for the intercalation of alkali ions which has been shown to drastically alter phase transition temperatures in these materials.23,24 The layered transition metal dichalcogenide 1T-TaS2 displays a charge-density wave (CDW) that couples to the lattice via electron-electron and electron-phonon interaction forming several commensurate phases through a Mott-Hubbard transition.25–27 When heated above ∼180 K, the commensurate CDW (CCDW) transforms into the nearly commensurate CDW (NCCDW) phase, identified electrically by an abrupt change in conductivity. A second transition of lesser magnitude into the incommensurate (ICCDW) phase at ∼360 K decouples the CDW from the lattice, further increasing the conductivity. It has been shown that in addition to temperature, the CDW-based phase transition can be driven using an electric field demonstrating abrupt, reversible threshold switching between the insulating Mott state and the metallic phase.28 Furthermore, the switching event can be driven electrically at nanosecond timescales, making this material a promising candidate for electronic device applications.28–30 Since the demonstration of electrically triggered switching by Hollander et al., several device concepts using 1T-TaS2 such as voltage-controlled oscillators31,32 and hybrid FET devices using graphene transistors33 have been demonstrated. However, these demonstrations all involve exfoliated materials from chemical vapor transport (CVT) grown bulk crystals, and no avenue towards a true monolithic all-2D phase transition device has been established.

Here, we demonstrate a pathway toward an all-2D phase transition device in the form of an epitaxial heterostructure of 1T-TaS2 and 2H-MoS2 capable of harnessing the CDW phase transition in 1T-TaS2 and semiconducting properties of 2H-MoS2. Figure 1(a) displays the target heterostructure phase transition device, wherein a monolayer MoS2 semiconducting layer is first deposited, followed by direct deposition of the 1T-TaS2 film. The 1T-TaS2 is separated by a van der Waals gap, allowing for the design of a hybrid device that couples the abrupt resistive switching of 1T-TaS2 with the semiconducting functionality of the MoS2. In this work, we investigate the properties of 1T-TaS2 in the 2D 1T-TaS2/MoS2 heterostructure grown via direct synthesis [Fig. 1(b)]. Both electrical and optical properties are carefully compared with those of the pure CVT and CVD grown material to verify the presence of CDW phases. Further, first principles calculations are employed to understand the effects of the 2H-MoS2 interlayer interactions on the formation of the CDW gap and the resulting magnitude of the phase transition. Finally, a device known as the Phase-FET is demonstrated to validate the integration of the phase-transition in 1T-TaS2 with the semiconducting MoS2 channel.

FIG. 1.

1T-TaS2/MoS2 Heterostructure (a) Schematic and (b) optical microscopy image of the proposed 2D heterostructure grown via direct synthesis. A monolayer 2H-MoS2 film is grown on a SiO2 substrate, followed directly by growth of 1T-TaS2. This heterostructure combines the semiconducting properties of 2H-MoS2 with the abrupt switching phenomena observed in 1T-TaS2 to realize a hybrid device. (c) Phase transition mechanism of 1T-TaS2. The insulating CCDW phase can be transformed to the NCCDW phase through the application of an external stimulus such as temperature or electric field.

FIG. 1.

1T-TaS2/MoS2 Heterostructure (a) Schematic and (b) optical microscopy image of the proposed 2D heterostructure grown via direct synthesis. A monolayer 2H-MoS2 film is grown on a SiO2 substrate, followed directly by growth of 1T-TaS2. This heterostructure combines the semiconducting properties of 2H-MoS2 with the abrupt switching phenomena observed in 1T-TaS2 to realize a hybrid device. (c) Phase transition mechanism of 1T-TaS2. The insulating CCDW phase can be transformed to the NCCDW phase through the application of an external stimulus such as temperature or electric field.

Close modal

The phase transition from CCDW to NCCDW is outlined in Fig. 1(c). In the CCDW phase, an Arrhenius type hopping transport between the thirteen Ta atom cluster forming the so-called “Star-of-David” regions gives rise to the insulating state.25 However, once a critical carrier concentration is reached through the application of an external stimulus (temperature, electric field, etc.), the CCDW phase transforms into the NCCDW phase, in which the electronic transport is dominated by conduction through metallic domain walls existing between “Star-of-David” clusters.34,35 The transformation from the CCDW to NCCDW phase is associated with the collapse of the CDW-based bandgap, resulting in the characteristic IMT phase transition in 1T-TaS2 illustrated by an abrupt change in resistivity.

To evaluate the feasibility of a CDW based heterostructure, we first theoretically investigate the band structure of the heterostructure to understand the effect of 2H-MoS2 on the formation of the CDW-based transport gap in 1T-TaS2 using first principles calculations (see supplementary material). Figures 2(a) and 2(b) display the optimized structures for both standalone multilayer 1T-TaS2 and the 1T-TaS2/2H-MoS2 heterostructure, respectively. The density of states calculated from the band structure of the standalone 1T-TaS2 structure is shown in Fig. 2(c). Formation of the Mott gap associated with the splitting of the upper and lower Hubbard bands due to the d-orbital overlap is observed, with a bandgap value of 0.18 eV that agrees well with previous reports.36 Similar calculations are performed for the heterostructure in Fig. 2(d), where a reduced bandgap of 0.11 eV is observed in the 1T-TaS2, owing to the interlayer interaction between the 1T-TaS2 and the 2H-MoS2. Although the insulating gap is reduced by 38%, the presence of a bandgap, nevertheless, is indicative of the retention of the CCDW phase in the heterostructure and paves the way for the demonstration of an electrically driven IMT phase transition in an all 2D-device.

FIG. 2.

1T-TaS2/MoS2 Heterostructure Simulation Optimized structures of (a) and (b) multilayer 1T-TaS2 and 1T-TaS2/MoS2 heterostructure used in first principles calculations, respectively. Density of states is calculated for (c) 1T-TaS2 and (d) 1T-TaS2/MoS2 heterostructures. The bandgap opening in 1T-TaS2 is associated with the presence of the commensurate CDW phase. The bandgap of 1T-TaS2 is shown to reduce by 38% from 0.18 eV to 0.11 eV when forming the heterostructure; however, the retention of the gap indicates that the CDW will remain in the heterostructure.

FIG. 2.

1T-TaS2/MoS2 Heterostructure Simulation Optimized structures of (a) and (b) multilayer 1T-TaS2 and 1T-TaS2/MoS2 heterostructure used in first principles calculations, respectively. Density of states is calculated for (c) 1T-TaS2 and (d) 1T-TaS2/MoS2 heterostructures. The bandgap opening in 1T-TaS2 is associated with the presence of the commensurate CDW phase. The bandgap of 1T-TaS2 is shown to reduce by 38% from 0.18 eV to 0.11 eV when forming the heterostructure; however, the retention of the gap indicates that the CDW will remain in the heterostructure.

Close modal

The synthesis of a 1T-TaS2/2H-MoS2 heterostructure involves a two-step growth process. Monolayer MoS2 is first synthesized via powder vapor (PV) deposition, followed by PV growth of 1T-TaS2 outlined by Zhao et al. in Ref. 37 (see supplementary material). A cross-sectional transmission electron microscopy (TEM) image of the heterostructure film is shown in Fig. 3(a), where multilayer 1T-TaS2 and monolayer MoS2 are separated only by a van der Waals gap. An energy dispersive x-ray (EDX) spectroscopy scan across the interface is presented in Fig. 3(b), highlighting the near ideal stoichiometry of the 1T-TaS2 and an atomically precise interface between the 1T-TaS2 and 2H-MoS2. We verify the CDW phases of the TaS2 film in the heterostructure by measuring the Raman spectrum as shown in Figs. 3(c) and 3(d). Albertini et al. calculated and verified the acoustic phonon modes associated with the formation of the thirteen Ta atom “Star of David” structure in the commensurate phase.38 These modes have a distinct signature between 50 and 100 cm−1 in the Raman spectrum and have since been used to optically verify the temperature dependent CDW phase transitions in 1T-TaS2.37,39,40 We measure the Raman spectrum at both 80 K and 300 K for the heterostructure films as well as the exfoliated (CVT) and CVD films (grown via powder vapor deposition). In all variations of the 1T-TaS2 investigated, the broadened peak around 70 cm−1 at 300 K splits into multiple peaks as temperature is reduced to 80 K, indicating the formation of the commensurate phase.

FIG. 3.

Metrology of the 1T-TaS2 Heterostructure. (a) TEM cross-section of the as-grown 1T-TaS2/MoS2 heterostructure on SiO2 substrate.41 Reproduced with permission from B. Grisafe et. al, IEEE Symposium on VLSI Technology, Systems, and Applications (2018). Copyright 2018 IEEE and (b) corresponding EDX scan across the interface. Raman characteristics for 1T-TaS2 at (c) 80 K and (d) 300 K. The single peak at 300 K splitting to multiple peaks at 80 K between 70 and 100 cm−1 is indicative of the formation of the commensurate CDW phase. All forms of 1T-TaS2 show CCDW formation, although the synthetic materials show slight peak broadening commonly associated with material defects.

FIG. 3.

Metrology of the 1T-TaS2 Heterostructure. (a) TEM cross-section of the as-grown 1T-TaS2/MoS2 heterostructure on SiO2 substrate.41 Reproduced with permission from B. Grisafe et. al, IEEE Symposium on VLSI Technology, Systems, and Applications (2018). Copyright 2018 IEEE and (b) corresponding EDX scan across the interface. Raman characteristics for 1T-TaS2 at (c) 80 K and (d) 300 K. The single peak at 300 K splitting to multiple peaks at 80 K between 70 and 100 cm−1 is indicative of the formation of the commensurate CDW phase. All forms of 1T-TaS2 show CCDW formation, although the synthetic materials show slight peak broadening commonly associated with material defects.

Close modal

To characterize the electrical properties of the CDW phase transition, two-terminal devices were fabricated on exfoliated 1T-TaS2, CVD (grown using the PV method) 1T-TaS2 and the 1T-TaS2/MoS2 heterostructure [Fig. 4(a)]. Individual exfoliated and CVD regions are first identified, with contact pads subsequently patterned and deposited using Ti/Au metal evaporation and a lift-off process. Heterostructure devices are fabricated in a similar fashion; however, in this case, the individual devices are first isolated using a XeF2 etch to remove the underlying MoS2 and then capped with 10 nm thick Al2O3. Contact vias are then patterned and etched through the Al2O3, followed by contact metallization. This process ensures that the top metal contact is made only to the 1T-TaS2, preventing any unwanted subsurface leakage pathways through the MoS2 layer in the heterostructure devices. We first investigate the CCDW to NCCDW phase transition using resistivity vs. temperature measurements over the temperature range of 80 K to 400 K [Fig. 4(b)]. As we increase the temperature above 200 K, both the exfoliated and CVD 1T-TaS2 transform from the CCDW to NCCDW (insulating to metallic state), indicated by an abrupt reduction in resistivity. Upon cooling, we recover the CCDW phase once the temperature drops below 120 K. The resistivity change for the CVD film is comparable to that for the exfoliated films, indicating a high quality synthetic material. We measure the two-terminal current-voltage (I-V) characteristics for the 1T-TaS2 films at 80 K in Fig. 4(c). The exfoliated, CVD, and heterostructure films all display abrupt, reversible, electrically driven IMT transitions, with compliance levels set to 5 mA in the case of exfoliated and CVD to prevent premature electrical breakdown of the devices. Figure 4(d) benchmarks the performance of the three films. We observe that while the CVD film displays similar characteristics to the exfoliated film, the heterostructure film grown in the same fashion suffers from a degraded ON/OFF ratio and increased metallic state resistivity (ρmetallic). In addition, we extract the critical resistivity (ρcritical) and critical carrier concentrations (ncritical) for each of the films, in order to evaluate whether the Mott criterion for resistive switching holds. The ρcritical in each case displays resistivities approximately in the range of ∼8 mΩ cm, which agrees well with the measured values presented in Ref. 28. Both insulating state and critical resistivities in our work are similar to those reported previously in Refs. 25 and 28. Assuming a carrier mobility of 15 cm2/V s, we estimate the critical carrier density, ncritical, to be approximately 5 × 1019 cm−3 for each TaS2 film. The ncritical values indicate that the Mott criterion for the abrupt IMT phase transition holds true in the case of the heterostructure. The differences in the ON/OFF resistivity ratio and ρmetallic between the heterostructure and other films can be attributed to two possible co-existing effects: (i) impurities introduced between layers when the sample experiences exposure to the atmosphere between the MoS2 and TaS2 growth leads to reduced electron mobility, and to higher metallic state resistivity, and (ii) the interlayer interaction between 1T-TaS2 and 2H-MoS2 alters the band structure in the heterostructure configuration resulting in the reduced bandgap of 1T-TaS2 (as predicted in Fig. 2). Both factors lead to a reduction in the overall ON/OFF resistivity ratio.

FIG. 4.

Electrical characterization of 1T-TaS2 and (a) Schematic of the fabricated 1T-TaS2 devices. (b) Normalized Resistivity vs. Temperature characteristics for exfoliated and synthetic 1T-TaS2. We observe comparable insulating and metallic states for the synthetic material to that of exfoliated. (c) Current-Voltage (I-V) measurements normalized to the device cross sectional area for the various 1T-TaS2 films show the insulator-to-metal phase transition from the commensurate to nearly commensurate phase at 80 K. The CDW phase transition is retained for the heterostructure when electrically driven, although the ON/OFF ratio is degraded. (d) Benchmarking of several characteristic properties of the CDW phase transition in 1T-TaS2.

FIG. 4.

Electrical characterization of 1T-TaS2 and (a) Schematic of the fabricated 1T-TaS2 devices. (b) Normalized Resistivity vs. Temperature characteristics for exfoliated and synthetic 1T-TaS2. We observe comparable insulating and metallic states for the synthetic material to that of exfoliated. (c) Current-Voltage (I-V) measurements normalized to the device cross sectional area for the various 1T-TaS2 films show the insulator-to-metal phase transition from the commensurate to nearly commensurate phase at 80 K. The CDW phase transition is retained for the heterostructure when electrically driven, although the ON/OFF ratio is degraded. (d) Benchmarking of several characteristic properties of the CDW phase transition in 1T-TaS2.

Close modal

To take advantage of the abrupt, reversible phase transition in 1T-TaS2, we experimentally explore a 2D Phase-FET as shown in Fig. 5(a) by electrically coupling the phase transition material (1T-TaS2) on the source side of a back-gated 2H-MoS2 transistor. The operation principle of the Phase-FET is shown in Fig. 5(b). By placing the IMT material on the source, one can control the voltage across the IMT through the application of gate to source voltage (VGS).4,7 Initially, the IMT material in the insulating (OFF) state lowers the effective VGS to the Phase-FET, reducing the OFF-state current in comparison to the standalone MOSFET. However, as the VGS increases the IMT material abruptly transforms to the metallic (ON) state, recovering the ON-state of the MOSFET leading to an overall enhancement in the ON/OFF ratio of the Phase-FET. Figures 5(c) and 5(d) display the output characteristics for the standalone MoS2 and MoS2/TaS2 heterostructure, respectively, with a physical gate length, LG, of 750 nm and MoS2 thickness, tMoS2, of 5 nm. A change in the slope of the output drain family curves can be observed at VDS = 4 V and VGS = 5 V in the Phase-FET case where the 1T-TaS2 undergoes the IMT phase transition, which is not observed in the standalone MoS2 FET case. An abrupt phase transition is not observed in the Phase-FET output characteristics due to a lack of an abrupt voltage snapback in the current-mode characteristics of the 1T-TaS2 [inset Fig. 5(d)]. In addition, the large trigger currents of ∼2 mA and consequently low insulating resistance of 1T-TaS2 prevent enhancement of the ON/OFF ratio. The added series resistance is not comparable to that of the MoS2 OFF-state resistance and, hence, does not reduce the Phase-FET OFF-state resistance significantly.4,6

FIG. 5.

2D phase-FET demonstration (a) Schematic of the fully 2D Phase-FET. The 1T-TaS2 device is electrically integrated with the source of the MoS2 transistor. (b) Operating principle of the Phase-FET in the ON and OFF state. In the OFF state, the insulating state of the IMT material reduces the OFF-state current of the Phase-FET due to added series resistance that lowers the effective VGS' of the Phase-FET. Once the effective VGS reaches a voltage high enough for the phase transition to take place, the IMT material transforms into the low resistance state, and the current of the standalone MOSFET is recovered. (c) Output characteristics for the 2D MoS2 transistor to be used in the Phase-FET configuration. (d) Output characteristics for the 1T-TaS2/MoS2 Phase-FET. The IMT takes place when VGS is greater than 5 V and VDS is greater than 4 V.

FIG. 5.

2D phase-FET demonstration (a) Schematic of the fully 2D Phase-FET. The 1T-TaS2 device is electrically integrated with the source of the MoS2 transistor. (b) Operating principle of the Phase-FET in the ON and OFF state. In the OFF state, the insulating state of the IMT material reduces the OFF-state current of the Phase-FET due to added series resistance that lowers the effective VGS' of the Phase-FET. Once the effective VGS reaches a voltage high enough for the phase transition to take place, the IMT material transforms into the low resistance state, and the current of the standalone MOSFET is recovered. (c) Output characteristics for the 2D MoS2 transistor to be used in the Phase-FET configuration. (d) Output characteristics for the 1T-TaS2/MoS2 Phase-FET. The IMT takes place when VGS is greater than 5 V and VDS is greater than 4 V.

Close modal

We have fabricated and characterized a 1T-TaS2/MoS2 heterostructure grown using powder vapor deposition. The 1T-TaS2 in the heterostructure configuration displays an abrupt, reversible CDW based IMT phase transition that can be triggered both thermally and electrically. A reduced ON/OFF resistivity ratio is observed for the heterostructure compared to the standalone films, due to the reduction of the band-gap in 1T-TaS2 caused by the 2H-MoS2 underlayer. Finally, a hybrid device that harnesses abrupt phase transitions is demonstrated in an all-2D structure driven by electrically coupling the IMT property of 1T-TaS2 with the semiconducting property of 2H-MoS2. This work provides a pathway and understanding to the development of heterogeneous integration of phase transition materials with traditional 2D semiconducting material systems.

See supplementary material for additional information on first principle calculations and materials synthesis methods.

This work was supported by the National Science Foundation Emerging Frontiers in Research and Innovation program under Award No. 1433307.

1.
A.
Zylbersztejn
and
N. F.
Mott
,
Phys. Rev. B
11
,
4383
(
1975
).
2.
J.
Del Valle
,
Y.
Kalcheim
,
J.
Trastoy
,
A.
Charnukha
,
D. N.
Basov
, and
I. K.
Schuller
,
Phys. Rev. Appl.
8
,
054041
(
2017
).
3.
C. H.
Ahn
,
J.-M.
Triscone
, and
J.
Mannhart
,
Nature
424
,
1015
(
2003
).
4.
N.
Shukla
,
A. V.
Thathachary
,
A.
Agrawal
,
H.
Paik
,
A.
Aziz
,
D. G.
Schlom
,
S. K.
Gupta
,
R.
Engel-Herbert
, and
S.
Datta
,
Nat. Commun.
6
,
7812
(
2015
).
5.
J.
Song
,
J.
Woo
,
S.
Lee
, and
A.
Prakash
,
IEEE Electron Device Lett.
37
,
932
(
2016
).
6.
N.
Shukla
,
B.
Grisafe
,
R. K.
Ghosh
,
N.
Jao
,
A.
Aziz
,
J.
Frougier
,
M.
Jerry
,
S.
Sonde
,
S.
Rouvimov
,
T.
Orlova
,
S.
Gupta
, and
S.
Datta
, in
2016 IEEE International Electron Devices Meeting
(IEEE,
2016
), p.
34.6.1
34.6.4
.
7.
J.
Frougier
,
N.
Shukla
,
D.
Deng
,
M.
Jerry
,
A.
Aziz
,
L.
Liu
,
G.
Lavallee
,
T. S.
Mayer
,
S.
Gupta
, and
S.
Datta
, in
Digest of Technical Papers-Symposium VLSI Technology
, September,
2016
.
8.
M.
Jerry
,
W. Y.
Tsai
,
B.
Xie
,
X.
Li
,
V.
Narayanan
,
A.
Raychowdhury
, and
S.
Datta
, in
Device Research on Conference-Conference Digest
, DRC, August,
2016
.
9.
M.
Jerry
,
A.
Parihar
,
B.
Grisafe
,
A.
Raychowdhury
, and
S.
Datta
, in
IEEE Symposium VLSI Circuits Digest of Technical Papers
(
2017
), Vol.
8
, p.
T186
.
10.
M.
Jerry
,
A.
Parihar
,
A.
Raychowdhury
, and
S.
Datta
, in
2017 75th Annual Device Research Conference
(
2017
), Vol.
47
, p.
1
.
11.
M.
Jerry
,
IEEE Electron Device Lett.
39
,
139
(
2018
).
12.
N.
Shukla
,
A.
Parihar
,
E.
Freeman
,
H.
Paik
,
G.
Stone
,
V.
Narayanan
,
H.
Wen
,
Z.
Cai
,
V.
Gopalan
,
R.
Engel-Herbert
,
D. G.
Schlom
,
A.
Raychowdhury
, and
S.
Datta
,
Sci. Rep.
4
,
4964
(
2014
).
13.
N.
Shukla
,
W.-Y.
Tsai
,
M.
Jerry
,
M.
Barth
,
V.
Narayanan
, and
S.
Datta
, in
2016 IEEE Symposium on VLSI Technology
(IEEE,
2016
), pp.
1
2
.
14.
D.
Lee
,
E.
Cha
,
J.
Park
,
C.
Sung
,
K.
Moon
,
S. A.
Chekol
, and
H.
Hwang
,
IEEE J. Electron Devices Soc.
6
,
250
(
2018
).
15.
A. A.
Sharma
,
Y.
Li
,
M.
Skowronski
,
J. A.
Bain
,
S.
Member
, and
J. A.
Weldon
,
IEEE Trans. Electron Devices
62
,
3857
(
2015
).
16.
J. A.
Wilson
,
F. J.
Di Salvo
, and
S.
Mahajan
,
Adv. Phys.
24
,
117
(
1975
).
17.
G. H.
Jaffari
and
W.
Mahmood
,
AIP Adv.
7
,
115312
(
2017
).
18.
H.
Kim
,
N. S.
Bingham
,
N. A.
Charipar
, and
A.
Piqué
,
AIP Adv.
7
,
105116
(
2017
).
19.
H.
Futaki
and
M.
Aoki
,
Jpn. J. Appl. Phys., Part 1
8
,
1008
(
1969
).
20.
M.
Kang
,
S.
Yu
, and
J.
Son
,
J. Phys. D. Appl. Phys.
48
,
095301
(
2015
).
21.
T.
Yajima
,
T.
Nishimura
, and
A.
Toriumi
,
Nat. Commun.
6
,
10104
(
2015
).
22.
W. A.
Vitale
,
E. A.
Casu
,
A.
Biswas
,
T.
Rosca
,
C.
Alper
,
A.
Krammer
,
G. V.
Luong
,
Q. T.
Zhao
,
S.
Mantl
,
A.
Schüler
, and
A. M.
Ionescu
,
Sci. Rep.
7
,
355
(
2017
).
23.
A. C.
Finnefrock
,
M.
Sutton
, and
J. D.
Brock
,
Phys. Rev. B
50
,
8157
(
1994
).
24.
H. J.
Crawack
and
C.
Pettenkofer
,
Solid State Commun.
118
,
325
(
2001
).
25.
R.
Inada
,
Y.
Onuki
, and
S.
Tanuma
,
Physica B+C
99
,
188
(
1980
).
26.
A. W.
Tsen
,
R.
Hovden
,
D.
Wang
,
Y. D.
Kim
,
J.
Okamoto
,
K. A.
Spoth
,
Y.
Liu
,
W.
Lu
,
Y.
Sun
,
J. C.
Hone
,
L. F.
Kourkoutis
,
P.
Kim
, and
A. N.
Pasupathy
,
Proc. Natl. Acad. Sci.
112
,
15054
(
2015
).
27.
R.
He
,
J.
Okamoto
,
Z.
Ye
,
G.
Ye
,
H.
Anderson
,
X.
Dai
,
X.
Wu
,
J.
Hu
,
Y.
Liu
,
W.
Lu
,
Y.
Sun
,
A. N.
Pasupathy
, and
A. W.
Tsen
,
Phys. Rev. B
94
,
201108
(
2016
).
28.
M. J.
Hollander
,
Y.
Liu
,
W. J.
Lu
,
L. J.
Li
,
Y. P.
Sun
,
J. A.
Robinson
, and
S.
Datta
,
Nano Lett.
15
,
1861
(
2015
).
29.
I.
Vaskivskyi
,
I. A.
Mihailovic
,
S.
Brazovskii
,
J.
Gospodaric
,
T.
Mertelj
,
D.
Svetin
,
P.
Sutar
, and
D.
Mihailovic
,
Nat. Commun.
7
,
11442
(
2016
).
30.
S.
Hellmann
,
T.
Rohwer
,
M.
Kalläne
,
K.
Hanff
,
C.
Sohrt
,
A.
Stange
,
A.
Carr
,
M. M.
Murnane
,
H. C.
Kapteyn
,
L.
Kipp
,
M.
Bauer
, and
K.
Rossnagel
,
Nat. Commun.
3
,
1069
(
2012
).
31.
A.
Khitun
,
G.
Liu
, and
A.
Balandin
,
IEEE Trans. Nanotechnol.
16
,
860
(
2017
).
32.
G.
Liu
,
B.
Debnath
,
T. R.
Pope
,
T. T.
Salguero
,
R. K.
Lake
, and
A. A.
Balandin
,
Nat. Nanotechnol.
11
,
845
(
2016
).
33.
S.
Zheng
,
F.
Liu
,
C.
Zhu
,
Z.
Liu
, and
H. J.
Fan
,
Nanoscale
9
,
2436
(
2017
).
34.
D.
Cho
,
G.
Gye
,
J.
Lee
,
S. H.
Lee
,
L.
Wang
,
S. W.
Cheong
, and
H. W.
Yeom
,
Nat. Commun.
8
,
392
(
2017
).
35.
L.
Ma
,
C.
Ye
,
Y.
Yu
,
X. F.
Lu
,
X.
Niu
,
S.
Kim
,
D.
Feng
,
D.
Tománek
,
Y. W.
Son
,
X. H.
Chen
, and
Y.
Zhang
,
Nat. Commun.
7
,
10956
(
2016
).
36.
P.
Fazekas
and
E.
Tosatti
,
Physica B+C
99
,
183
(
1980
).
37.
R.
Zhao
,
B.
Grisafe
,
R. K.
Ghosh
,
S.
Holoviak
,
B.
Wang
,
K.
Wang
,
N.
Briggs
,
A.
Haque
,
S.
Datta
, and
J.
Robinson
,
2D Mater.
5
,
025001
(
2018
).
38.
O. R.
Albertini
,
R.
Zhao
,
R. L.
McCann
,
S.
Feng
,
M.
Terrones
,
J. K.
Freericks
,
J. A.
Robinson
, and
A. Y.
Liu
,
Phys. Rev. B
93
,
214109
(
2016
).
39.
R.
Zhao
,
Y.
Wang
,
D.
Deng
,
X.
Luo
,
W. J.
Lu
,
Y. P.
Sun
,
Z. K.
Liu
,
L. Q.
Chen
, and
J.
Robinson
,
Nano Lett.
17
,
3471
(
2017
).
40.
W.
Fu
,
Y.
Chen
,
J.
Lin
,
X.
Wang
,
Q.
Zeng
,
J.
Zhou
,
L.
Zheng
,
H.
Wang
,
Y.
He
,
H.
He
,
Q.
Fu
,
K.
Suenaga
,
T.
Yu
, and
Z.
Liu
,
Chem. Mater.
28
,
7613
(
2016
).
41.
B.
Grisafe
,
R.
Zhao
,
M.
Jerry
,
J. A.
Robinson
, and
S.
Datta
, in
IEEE Symposium on VLSI Technology, Systems, and Applications (VSLI-TSA)
(IEEE,
2018
).

Supplementary Material