Studies were conducted on epitaxial VO2 thin films to assess to the effect of remote epitaxy on the metal–insulator transition (MIT). The epitaxial VO2 heterostructures were synthesized on both bare Al2O3 (0001) substrates and Al2O3 substrates coated with two monolayer-thick graphene. While both systems exhibit the MIT, the film grown by remote epitaxy on graphene demonstrates improved transport properties. Electrical transport measurements show that the on/off ratio is enhanced by a factor of ∼7.5 and the switching temperature window is narrower for VO2 thin films grown on graphene. By characterizing the heterostructures with a suite of structural, chemical, and spectroscopic tools, we find that the graphene interlayer inhibits oxygen vacancy diffusion from Al2O3 (0001) during the VO2 growth, resulting in improved electrical behavior at the MIT.

The transition metal oxides exhibit a wealth of properties distinct from conventional semiconductors such as the giant magnetoresistance, 2D electronic gas behavior, and superconductivity.1–4 Many of these properties are closely related to the oxygen stoichiometry which can be modulated in thin film heterostructures by altering the growth condition or by the insertion or exit of oxygen ions post deposition. VO2 is an archetypal correlated material, exhibiting a metal–insulator transition (MIT) near room temperature.5,6 The characteristics of the MIT and reconfigurability of VO2-based heterostructures via an electric field have attracted a significant amount of attention in terms of both fundamental7–14 and applied investigations—e.g., next-generation transistors,15 optical switches,16 smart windows,17 and gas sensors.18 

Vanadium dioxide thin films have been grown by a variety of physical vapor deposition techniques (magnetron sputtering,19–21 pulsed laser deposition,10,22–25 and molecular beam epitaxy26–28) and chemical solution deposition.29,30 However, the synthesis of VO2 epitaxial thin films with the desired structural quality and chemical stoichiometry is highly challenging due to the complexity of the vanadium/oxide phase diagram. An excess of oxygen vacancies in the as-grown film can lead to pronounced changes to the MIT such as a broader switching temperature window and a lower on/off ratio.10,31 Kum et al.32 have shown that oxides such as the perovskites, spinels, and garnets, can be epitaxially grown on single crystal substrates through a layer or multiple layers of graphene despite potential damage to the graphene in the oxidizing growth environment, allowing the synthesis of freestanding, single crystal films, and nanomembranes. The nature of the epitaxial growth process, termed remote epitaxy, was later studied by Kim et al. for different heterostructures.33 Guo et al.34 very recently demonstrated the creation of a reconfigurable device based on the remote epitaxial growth of VO2 through graphene.

In this work, we synthesize and compare the properties of VO2 thin films grown directly on single crystal Al2O3 (0001) and through two-monolayer-thick graphene on Al2O3 (0001) by RF magnetron sputtering. The 20 nm-thick VO2 films were sputter deposited from a V2O5 target at a substrate temperature of 550 °C. To minimize oxidation of graphene and excess oxygen in the films, O2 was not used: only flowing Ar was introduced to the chamber at 90  sccm. The pressure in the chamber was maintained at 10−2 Torr. After growth, the thin films were left in the chamber to cool to room temperature while maintaining 40  sccm Ar flow. The transfer process of polycrystalline two-monolayer graphene is described in the supplementary material.

The transport properties of the epitaxial VO2 films grown with and without the graphene interlayer were carried out in a physical property measurement system (Quantum Design PPMS). The resistance was measured in the van der Pauw geometry during heating and cooling in the 250–400 K range, with a device size of 4 mm × 8 mm. X-ray absorption spectroscopy (XAS) measurements at the vanadium K-edge were performed at room temperature at Sector 12-BM of Advanced Photon Source, with an energy resolution of ∼0.4–0.5 eV [Si (111) monochromator]. To maximize signal from the film, the XAS measurements were conducted at grazing incidence (<1°), and a four-element silicon drift detector located 90° relative to the incident beam along the direction of x-ray polarization was used to record the spectra in the total fluorescence yield (TFY) mode.

To study the effect of the graphene layers on VO2, the temperature dependent resistance was measured for films grown on bare Al2O3 and graphene-coated Al2O3. The metal–insulator–transitions for both types of thin films are evident from the heating and cooling profiles shown in Fig. 1(a). Interestingly, a higher resistance modulation (ΔRs/Rs ∼ 2.6 × 103, where Rs is the sheet resistance) was observed across the MIT for the VO2 film grown on graphene as compared to a film grown directly on Al2O3 (0001) (ΔRs/Rs ∼ 3.5 × 102), an increase in nearly 7.5×. In addition to the resistance ratio, other metal–insulator transition characteristics were measured: d(log10(Rs))/dT, Tc, Th, ΔTc, ΔTh, and ΔH, as shown in Figs. 1(b) and 1(c); their definitions are described below. The profiles for d(log10(Rs))/dT were fit with Gaussian functions, providing the transition temperatures Tc and Th, which correspond to the peak positions measured during cooling and heating, respectively. The difference between these gives the transition width, ΔH. The sharpness of the transitions during cooling and heating, ΔTc and ΔTh, was determined from the full-width-at-half-maximum at the peak. Values for the width and sharpness for both types of VO2 films are summarized in Fig. 1(d). We find that the sharpness of the transition as well as the resistance modulation are significantly improved for the film grown on graphene (11 and 13.2 K for ΔTc and ΔTh, respectively) vs the film grown directly on Al2O3 (18.8 and 25.6 K for ΔTc and ΔTh, respectively). The transition width of the film grown on graphene is 2.1 K larger than that without graphene. This represents a change in hysteretic behavior for films grown on graphene vs without graphene.

FIG. 1.

(a) Sheet resistance as a function of temperature for 20-nm-thick, epitaxial VO2 films grown on Al2O3 (0001) (blue) and on two-monolayer-graphene/Al2O3 (0001) (red). The dependence of the rate of change, dlog(Rs)/dT on temperature elucidates different properties of the metal–insulator transition, namely, the transition point (Th and Tc), the transition sharpness (ΔTh and ΔTc), and the transition width (ΔH); the results are shown for VO2 films grown epitaxially without graphene (b) and with graphene (c). The subscripts h and c indicate measurements taken during heating or cooling, respectively. (d) Comparison of the metal–insulator transition properties in different VO2 thin films.

FIG. 1.

(a) Sheet resistance as a function of temperature for 20-nm-thick, epitaxial VO2 films grown on Al2O3 (0001) (blue) and on two-monolayer-graphene/Al2O3 (0001) (red). The dependence of the rate of change, dlog(Rs)/dT on temperature elucidates different properties of the metal–insulator transition, namely, the transition point (Th and Tc), the transition sharpness (ΔTh and ΔTc), and the transition width (ΔH); the results are shown for VO2 films grown epitaxially without graphene (b) and with graphene (c). The subscripts h and c indicate measurements taken during heating or cooling, respectively. (d) Comparison of the metal–insulator transition properties in different VO2 thin films.

Close modal

To probe the enhancement of the MIT in VO2 films by remote epitaxy, we measured x-ray absorption spectra at the V K-edge to quantify the variation of V oxidation state and the local structure with the presence of graphene interlayers for the growth of VO2 films; the spectra are shown in Fig. 2(a). The various absorption features are identified and labeled as a–e. The pre-edge feature a represents the dipole-forbidden transition 1s → 3d.35,36 This is due to hybridization of the V 3d orbitals and the O 2p states, which leads to the p component of the dipole transition to the hybridized states. The intensity and position of feature a, therefore, provides information on both the local coordination environment and oxidation state.36 The main absorption edge b represents the excitation of a core photoelectron into the unoccupied continuum. The near-edge features c–e above the absorption edge result from not only multiple scattering but also from the dipole-allowed excitation of a core 1s electron to a localized 4p state.36,37 The difference between the two spectra is shown in Fig. 2(b). The red shaded region and the green shaded region correspond to the pre-edge and the post-edge of feature a, respectively. The blue shaded region includes the feature b and the pre-edge of feature c. The results indicate that the energy of V K-edge valence for epitaxial films grown directly on Al2O3 shifts toward lower energy, demonstrating that the oxidation state of vanadium here is lower than in films grown via remote epitaxy. According to the previous work,38 the decrease in oxidation state can be caused by oxygen vacancies in the film. It is well known that the oxygen vacancies can suppress the MIT of VO2, and our transport properties measurement results are consistent with this, as in Ref. 10. As for the extended x-ray absorption fine structure (EXAFS), since the first and second neighbor atoms of a central vanadium correspond to six O and two V atoms, respectively, the distances between V–O and V–V can be quantitatively determined. Figure 2(c) shows the Fourier transform modulus of the EXAFS profiles in real space to elucidate the local atomic structure of the two types of films. The profile for the VO2 film grown by traditional epitaxy exhibits different amplitudes and distances from the film grown by remote epitaxy. The atomic distances of V–O and V–V in VO2/Al2O3 (0001) are 0.167 and 0.133 Å longer than in VO2/graphene/Al2O3 (0001), indicating an expanded lattice for VO2 film grown directly on the Al2O3 substrate without the presence of graphene interlayers. Moreover, we performed synchrotron x-ray diffraction (XRD) to confirm the lattice structure variation in both thin films. The 00L specular reflections of both films are shown in Fig. 2(d). The (020) reflection in the VO2 diffraction is coupled to the (0006) Al2O3 substrate Bragg peak. Obviously, the out-of-plane lattice spacing is expanded for VO2 directly grown on Al2O3 (without graphene) as compared to that grown with the presence of graphene interlayers.

FIG. 2.

(a) Normalized V K-edge XANES spectra for epitaxial VO2 thin films grown on two-monolayer-graphene/Al2O3 (0001) (red) and directly on Al2O3 (0001) (blue). (b) The difference between the XANES spectra in the two types of 20-nm-thick, epitaxial VO2 thin films. The color shading highlights different regions of interest. (c) The magnitude of the Fourier transformed EXAFS profiles for VO2 thin films grown with graphene (red) and VO2 thin films without graphene (blue). (d) 00L specular XRD data (H, K = 0) for both films. The VO2 thin film Bragg peak (020) is plotted in the reciprocal lattice unit (r.l.u.) of Al2O3. L = 6 corresponds to Al2O3 (0006) substrate Bragg peak.

FIG. 2.

(a) Normalized V K-edge XANES spectra for epitaxial VO2 thin films grown on two-monolayer-graphene/Al2O3 (0001) (red) and directly on Al2O3 (0001) (blue). (b) The difference between the XANES spectra in the two types of 20-nm-thick, epitaxial VO2 thin films. The color shading highlights different regions of interest. (c) The magnitude of the Fourier transformed EXAFS profiles for VO2 thin films grown with graphene (red) and VO2 thin films without graphene (blue). (d) 00L specular XRD data (H, K = 0) for both films. The VO2 thin film Bragg peak (020) is plotted in the reciprocal lattice unit (r.l.u.) of Al2O3. L = 6 corresponds to Al2O3 (0006) substrate Bragg peak.

Close modal

We present a schematic of the structures in Figs. 3(a) and 3(b), demonstrating how oxygen vacancies can induce the lattice expansion process: Fig. 3(a) shows the original lattice structure and Fig. 3(b) illustrates how the neighboring atoms of oxygen vacancies move outwards, affecting the EXAFS and XRD measurements. We then find that the VO2 films grown by remote epitaxy can exhibit improved MIT characteristics compared to those grown by traditional epitaxy. In previous reports,33,39 researchers indicated that oxygen migration through graphene is severely impeded. In Figs. 3(c) and 3(d), we present a simple schematic summarizing our conclusions. The graphene layers allow epitaxial growth of VO2/Al2O3 (0001) via penetrating ionic interactions but hinder migration of oxygen ions between the film and substrate, helping VO2 thin films to maintain good oxygen stoichiometry. Since oxygen vacancies are known to be deleterious to the properties of the MIT, growth by remote epitaxy appears to be a viable strategy for achieving improved thin film properties.

FIG. 3.

Schematic of the lattice expansion process due to oxygen vacancies (a) and (b) and the two epitaxial VO2 thin films grown on Al2O3 (0001) (c) and (d). Graphene layers can impede the interdiffusion of oxygen defects between VO2 and the Al2O3 substrates. This is illustrated by the absence of vacancies (c) and the appearance of oxygen vacancies (Ov) in VO2 (d).

FIG. 3.

Schematic of the lattice expansion process due to oxygen vacancies (a) and (b) and the two epitaxial VO2 thin films grown on Al2O3 (0001) (c) and (d). Graphene layers can impede the interdiffusion of oxygen defects between VO2 and the Al2O3 substrates. This is illustrated by the absence of vacancies (c) and the appearance of oxygen vacancies (Ov) in VO2 (d).

Close modal

In summary, through a series of electrical transport, structural, and spectroscopic studies, we reveal improvements to the behavior of metal-insulator transition in VO2 films with the use of remote epitaxy. Electrical transport measurements show that both the ratio of the change in resistance and the sharpness of the transition improve significantly for epitaxial VO2 films grown on graphene, as compared to films grown directly on Al2O3 (0001). With the aid of synchrotron x-ray techniques, it is shown that the graphene interlayers can help to prevent oxygen loss during VO2 deposition.

See the supplementary material for information on the graphene transfer procedure, electrical measurements from additional VO2 samples, and additional XANES/EXAFS results.

The authors acknowledge Brandon L. Fisher and David Gosztola from the Center for Nanoscale Materials and 12-BM beamline scientist Benjamin Reinhart from the Advanced Photon Source (APS). This research used resources of the Advanced Photon Source and Center for Nanoscale Materials, both U.S. Department of Energy (DOE) Office of Science User Facilities and is based on the work supported by the Laboratory Directed Research and Development (LDRD) funding from Argonne National Laboratory, provided by the Director, Office of Science, of the U.S. DOE under Contract No. DE-AC02-06CH11357.

The authors have no conflicts to disclose.

Hui Cao: Data curation (lead); Writing – original draft (lead). Xi Yan: Data curation (equal). Yan Li: Data curation (supporting). Liliana Stan: Data curation (equal); Writing – review and editing (supporting). Wei Chen: Writing – review and editing (supporting). Nathan Guisinger: Writing – review and editing (supporting). Hua Zhou: Conceptualization (equal); Writing – review and editing (equal). Dillon D. Fong: Conceptualization (equal); Writing – review and editing (equal).

The data that support the findings of this study are available within the article and its supplementary material.

1.
A. F.
Santander-Syro
,
O.
Copie
,
T.
Kondo
,
F.
Fortuna
,
S.
Pailhès
,
R.
Weht
,
X. G.
Qiu
,
F.
Bertran
,
A.
Nicolaou
,
A.
Taleb-Ibrahimi
,
P.
Le Fèvre
,
G.
Herranz
,
M.
Bibes
,
N.
Reyren
,
Y.
Apertet
,
P.
Lecoeur
,
A.
Barthélémy
, and
M. J.
Rozenberg
,
Nature
469
,
189
(
2011
).
2.
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
).
3.
H. Y.
Hwang
,
Y.
Iwasa
,
M.
Kawasaki
,
B.
Keimer
,
N.
Nagaosa
, and
Y.
Tokura
,
Nat. Mater.
11
,
103
(
2012
).
4.
C.
Liu
,
X.
Yan
,
D.
Jin
,
Y.
Ma
,
H.-W.
Hsiao
,
Y.
Lin
,
T. M.
Bretz-Sullivan
,
X.
Zhou
,
J.
Pearson
,
B.
Fisher
,
J. S.
Jiang
,
W.
Han
,
J.-M.
Zuo
,
J.
Wen
,
D. D.
Fong
,
J.
Sun
,
H.
Zhou
, and
A.
Bhattacharya
,
Science
371
,
716
(
2021
).
5.
F. J.
Morin
,
Phys. Rev. Lett.
3
,
34
(
1959
).
6.
K.
Kosuge
,
J. Phys. Soc. Jpn.
22
,
551
(
1967
).
7.
M.
Liu
,
H. Y.
Hwang
,
H.
Tao
,
A. C.
Strikwerda
,
K.
Fan
,
G. R.
Keiser
,
A. J.
Sternbach
,
K. G.
West
,
S.
Kittiwatanakul
,
J.
Lu
,
S. A.
Wolf
,
F. G.
Omenetto
,
X.
Zhang
,
K. A.
Nelson
, and
R. D.
Averitt
,
Nature
487
,
345
(
2012
).
8.
M. F.
Jager
,
C.
Ott
,
P. M.
Kraus
,
C. J.
Kaplan
,
W.
Pouse
,
R. E.
Marvel
,
R. F.
Haglund
,
D. M.
Neumark
, and
S. R.
Leone
,
Proc. Natl. Acad. Sci.
114
,
9558
(
2017
).
9.
M. M.
Qazilbash
,
M.
Brehm
,
B.-G.
Chae
,
P.-C.
Ho
,
G. O.
Andreev
,
B.-J.
Kim
,
S. J.
Yun
,
A. V.
Balatsky
,
M. B.
Maple
,
F.
Keilmann
,
H.-T.
Kim
, and
D. N.
Basov
,
Science
318
,
1750
(
2007
).
10.
J.
Jeong
,
N.
Aetukuri
,
T.
Graf
,
T. D.
Schladt
,
M. G.
Samant
, and
S. S. P.
Parkin
,
Science
339
,
1402
(
2013
).
11.
G.
Stefanovich
,
A.
Pergament
, and
D.
Stefanovich
,
J. Phys.: Condens. Matter
12
,
8837
(
2000
).
12.
S.
Biermann
,
A.
Poteryaev
,
A. I.
Lichtenstein
, and
A.
Georges
,
Phys. Rev. Lett.
94
,
026404
(
2005
).
13.
T.
Yao
,
X.
Zhang
,
Z.
Sun
,
S.
Liu
,
Y.
Huang
,
Y.
Xie
,
C.
Wu
,
X.
Yuan
,
W.
Zhang
,
Z.
Wu
,
G.
Pan
,
F.
Hu
,
L.
Wu
,
Q.
Liu
, and
S.
Wei
,
Phys. Rev. Lett.
105
,
226405
(
2010
).
14.
M.
Imada
,
A.
Fujimori
, and
Y.
Tokura
,
Rev. Mod. Phys.
70
,
1039
(
1998
).
15.
M.
Nakano
,
K.
Shibuya
,
D.
Okuyama
,
T.
Hatano
,
S.
Ono
,
M.
Kawasaki
,
Y.
Iwasa
, and
Y.
Tokura
,
Nature
487
,
459
(
2012
).
16.
X.
Chen
,
J.
Li
, and
Q.
Lv
,
Optik
124
,
2041
(
2013
).
17.
J.
Zhou
,
Y.
Gao
,
Z.
Zhang
,
H.
Luo
,
C.
Cao
,
Z.
Chen
,
L.
Dai
, and
X.
Liu
,
Sci. Rep.
3
,
3029
(
2013
).
18.
E.
Strelcov
,
Y.
Lilach
, and
A.
Kolmakov
,
Nano Lett.
9
,
2322
(
2009
).
19.
F. J.
Wong
,
Y.
Zhou
, and
S.
Ramanathan
,
J. Cryst. Growth
364
,
74
(
2013
).
20.
C.
Ko
,
Z.
Yang
, and
S.
Ramanathan
,
ACS Appl. Mater. Interfaces
3
,
3396
(
2011
).
21.
D.
Ruzmetov
,
K. T.
Zawilski
,
V.
Narayanamurti
, and
S.
Ramanathan
,
J. Appl. Phys.
102
,
113715
(
2007
).
22.
K.
Martens
,
N.
Aetukuri
,
J.
Jeong
,
M. G.
Samant
, and
S. S. P.
Parkin
,
Appl. Phys. Lett.
104
,
081918
(
2014
).
23.
M. R.
Bayati
,
R.
Molaei
,
F.
Wu
,
J. D.
Budai
,
Y.
Liu
,
R. J.
Narayan
, and
J.
Narayan
,
Acta Mater.
61
,
7805
(
2013
).
24.
D. K.
Lee
,
Y.
Park
,
H.
Sim
,
J.
Park
,
Y.
Kim
,
G.-Y.
Kim
,
C.-B.
Eom
,
S.-Y.
Choi
, and
J.
Son
,
Nat. Commun.
12
,
5019
(
2021
).
25.
D.
Lee
,
B.
Chung
,
Y.
Shi
,
G.-Y.
Kim
,
N.
Campbell
,
F.
Xue
,
K.
Song
,
S.-Y.
Choi
,
J. P.
Podkaminer
,
T. H.
Kim
,
P. J.
Ryan
,
J.-W.
Kim
,
T. R.
Paudel
,
J.-H.
Kang
,
J. W.
Spinuzzi
,
D. A.
Tenne
,
E. Y.
Tsymbal
,
M. S.
Rzchowski
,
L. Q.
Chen
,
J.
Lee
, and
C. B.
Eom
,
Science
362
,
1037
(
2018
).
26.
H.-T.
Zhang
,
L.
Zhang
,
D.
Mukherjee
,
Y.-X.
Zheng
,
R. C.
Haislmaier
,
N.
Alem
, and
R.
Engel-Herbert
,
Nat. Commun.
6
,
8475
(
2015
).
27.
L. L.
Fan
,
S.
Chen
,
Z. L.
Luo
,
Q. H.
Liu
,
Y. F.
Wu
,
L.
Song
,
D. X.
Ji
,
P.
Wang
,
W. S.
Chu
,
C.
Gao
,
C. W.
Zou
, and
Z. Y.
Wu
,
Nano Lett.
14
,
4036
(
2014
).
28.
S.
Chen
,
Z.
Wang
,
H.
Ren
,
Y.
Chen
,
W.
Yan
,
C.
Wang
,
B.
Li
,
J.
Jiang
, and
C.
Zou
,
Sci. Adv.
5
,
eaav6815
(
2019
).
29.
Y. F.
Wu
,
L. L.
Fan
,
S. M.
Chen
,
S.
Chen
,
C. W.
Zou
, and
Z. Y.
Wu
,
AIP Adv.
3
,
042132
(
2013
).
30.
S.
Lu
,
L.
Hou
, and
F.
Gan
,
Adv. Mater.
9
,
244
(
1997
).
31.
Q.
Lu
,
C.
Sohn
,
G.
Hu
,
X.
Gao
,
M. F.
Chisholm
,
I.
Kylänpää
,
J. T.
Krogel
,
P. R. C.
Kent
,
O.
Heinonen
,
P.
Ganesh
, and
H. N.
Lee
,
Sci. Rep.
10
,
18554
(
2020
).
32.
H. S.
Kum
,
H.
Lee
,
S.
Kim
,
S.
Lindemann
,
W.
Kong
,
K.
Qiao
,
P.
Chen
,
J.
Irwin
,
J. H.
Lee
,
S.
Xie
,
S.
Subramanian
,
J.
Shim
,
S.-H.
Bae
,
C.
Choi
,
L.
Ranno
,
S.
Seo
,
S.
Lee
,
J.
Bauer
,
H.
Li
,
K.
Lee
,
J. A.
Robinson
,
C. A.
Ross
,
D. G.
Schlom
,
M. S.
Rzchowski
,
C.-B.
Eom
, and
J.
Kim
,
Nature
578
,
75
(
2020
).
33.
H.
Kim
,
K.
Lu
,
Y.
Liu
,
H. S.
Kum
,
K. S.
Kim
,
K.
Qiao
,
S.-H.
Bae
,
S.
Lee
,
Y. J.
Ji
,
K. H.
Kim
,
H.
Paik
,
S.
Xie
,
H.
Shin
,
C.
Choi
,
J. H.
Lee
,
C.
Dong
,
J. A.
Robinson
,
J.-H.
Lee
,
J.-H.
Ahn
,
G. Y.
Yeom
,
D. G.
Schlom
, and
J.
Kim
,
ACS Nano
15
,
10587
(
2021
).
34.
Y.
Guo
,
X.
Sun
,
J.
Jiang
,
B.
Wang
,
X.
Chen
,
X.
Yin
,
W.
Qi
,
L.
Gao
,
L.
Zhang
,
Z.
Lu
,
R.
Jia
,
S.
Pendse
,
Y.
Hu
,
Z.
Chen
,
E.
Wertz
,
D.
Gall
,
J.
Feng
,
T.-M.
Lu
, and
J.
Shi
,
Nano Lett.
20
,
33
(
2020
).
35.
G.
Giuli
,
E.
Paris
,
J.
Mungall
,
C.
Romano
, and
D.
Dingwell
,
Am. Miner.
89
,
1640
(
2004
).
36.
J.
Wong
,
F. W.
Lytle
,
R. P.
Messmer
, and
D. H.
Maylotte
,
Phys. Rev. B
30
,
5596
(
1984
).
37.
C.
Marini
,
M.
Bendele
,
B.
Joseph
,
I.
Kantor
,
M.
Mitrano
,
O.
Mathon
,
M.
Baldini
,
L.
Malavasi
,
S.
Pascarelli
, and
P.
Postorino
,
Europhys. Lett.
108
,
36003
(
2014
).
38.
H.
Cao
,
H.
Guo
,
Y.-C.
Shao
,
Q.
Liu
,
X.
Feng
,
Q.
Lu
,
Z.
Wang
,
A.
Zhao
,
A.
Fujimori
,
Y.-D.
Chuang
,
H.
Zhou
, and
X.
Zhai
,
Nano Lett.
21
,
3981
(
2021
).
39.
J. S.
Bunch
,
S. S.
Verbridge
,
J. S.
Alden
,
A. M.
van der Zande
,
J. M.
Parpia
,
H. G.
Craighead
, and
P. L.
McEuen
,
Nano Lett.
8
,
2458
(
2008
).

Supplementary Material