TiSe2 is a member of the transition metal dichalcogenide family of layered van der Waals materials that exhibits some distinct electronic and optical properties. Here, we perform the Raman spectroscopy and microscopy studies on single crystal TiSe2 to investigate the thermal and photon-induced defects associated with the diffusion of selenium to the surface. Additional phonon peaks near 250 cm−1 are observed in the laser-irradiated regions that are consistent with the formation of amorphous and nanocrystalline selenium on the surface. Temperature dependent studies of the threshold temperature and laser intensity necessary to initiate selenium migration to the surface show an activation barrier for the process of 1.55 eV. The impact of these results on the properties of strongly correlated electron states in TiSe2 is discussed.

Transition metal dichalcogenides (TMDs) exhibit a series of unique electronic properties ranging from charge density wave (CDW) order to superconductivity. 1 T-TiSe2 is a quasi-2D layered material with a trigonal symmetry that has been studied for over 30 years.1–9 Recently, it was discovered that the CDW in this material has excitonic origin5,6 and has a chiral order,10,11 which may have implications on the fundamental understanding of the strongly correlated electron systems. On the other hand, the ability to separate charges in materials with reduced dimensionality is of great interest for many potential applications.7,8 Due to its specific structural and electronic properties, TiSe2 has also been considered as an alternative to graphene electronics12,13 in thermoelectric applications9,14 and as a cathode material in batteries.15 

The strongly correlated electron behavior in TMDs is usually attributed to their quasi-two-dimensional structure consisting of the X–M–X layers (M–transition metal and X–chalcogen) weakly bound together by the van der Waals forces.16 Further interest in these systems has been stimulated by the changes in their electronic properties upon reduction to a single monolayer. The modification of the electronic properties in 2D layered systems is due to the changes in the electronic and phonon band structure when transitioning to a single X-M-X layer.17,18 Up to now, the TiSe2 monolayers have been proven difficult to obtain through methods other than molecular beam epitaxy.19–21 Synthesized thin film TMDs often suffer from high concentration of defects and lack of long range crystalline order compared to their single crystal exfoliated counterparts.16 It has been established that in TiSe2 the most frequent defects are titanium self-intercalation,21,22 in which Ti ions are displaced from the lattice site and moved into the van der Waals gap23 and selenium vacancies and interstitials.1,20,22,24,25 The concentration of excess titanium and selenium vacancies depends strongly on the conditions of film growth, especially growth temperature, similar as in the case of single crystal growth.1,2 Selenium deficiencies are detrimental to the charge density wave order,16 as their presence changes the nature of the material from semiconducting to semimetallic. Therefore, the characterization of the defects and their dynamics is of utmost importance.

The Raman spectroscopy is one of the main tools utilized to study the layered TMDs, as the symmetric A1g breathing phonon mode and Eg layer shearing mode are sensitive to the inter- and intra-layer properties of the structure.26 In the present work, we study the diffusion of selenium atoms from the bulk to the surface of high quality TiSe2 single crystals. Elevated temperatures and irradiation by the Raman laser induce thermally activated diffusion of selenium atoms to the surface of the crystal, where they form selenium nanosize clusters. The temperature dependent studies allow us to estimate the activation energy for this process, an important parameter if considering TiSe2 as material for electronic and optical applications.

High quality single crystals of TiSe2 were grown using chemical vapor transport method1,2,27 and characterized using X-ray diffraction, energy dispersive X-ray spectroscopy (EDS), and variable temperature electrical transport.28 The signatures of low level of intrinsic defects in the TiSe2 single crystals are a pronounced charge density wave peak at the CDW transition in temperature dependent resistivity (an increase of 2.5–3.5 times compared to the room temperature resistivity) and low defect concentrations as observed by scanning tunneling microscopy.29 Variable temperature Raman spectroscopy measurements were conducted at the Center for Nanoscale Materials at Argonne National Laboratory using the Renishaw InVia Raman microscope with a 514 nm argon ion laser source and a ∼1.5 μm diameter spot size. The spectrometer is equipped with variable temperature cell capable of operating between 80 and 500 K. All the experiments were conducted in the presence of ultra-high pure nitrogen exchange gas at normal pressure. Fig. 1(a) shows the Raman spectra on a freshly cleaved surface of TiSe2 single crystal. At room temperature, one can see the normal phase shearing mode Eg and breathing mode A1g peaks. As the temperature is lowered below the CDW transition of ∼200 K, we see the emergence of additional peaks that reflect the change in lattice symmetry. The CDW doubles the period of the crystal lattice to 2a0 × 2a0 × 2c and induces additional Raman peaks: the 315 cm1 peak is first seen below ∼160 K, in agreement with the previous measurements.30–32 The charge density wave A1g peak at 110 cm1 becomes fully observable only at temperatures lower than ∼100 K.33 

FIG. 1.

(a) Temperature evolution of Raman spectra of TiSe2 single crystal (1.05 mW excitation power) and (b) structure of the unit lattice of 1 T-TiSe2 and Raman active breathing mode (A1g) and shear mode (Eg) of the crystal lattice.

FIG. 1.

(a) Temperature evolution of Raman spectra of TiSe2 single crystal (1.05 mW excitation power) and (b) structure of the unit lattice of 1 T-TiSe2 and Raman active breathing mode (A1g) and shear mode (Eg) of the crystal lattice.

Close modal

Uniform heating of the TiSe2 single crystal above room temperature induces new Raman peaks that are irreversible with temperature. Fig. 2 shows the evolution of Raman spectra of the TiSe2 single crystal from room temperature and up to 400 K. At 300 K, one notices a broad Raman background signal centered at around 250 cm1. There are also enhanced fluctuations of the background intensity in the spectrum around this wave number. As the temperature is increased, the broad background slowly evolves into a set of distinct peaks. At 360 K, we observe a sudden broad increase in the background at near 250 cm1. This broad peak is consistent with the previous Raman experiments performed on non-stoichiometric TiSe2x with high concentration of selenium vacancies.14 As the temperature is raised further, the broad background evolves into distinct peaks that are made up of at least three identifiable Raman modes centered at 233, 241 and 263 cm−1 with a broad peak centered at 250 cm−1 (Fig. 2).

FIG. 2.

Raman spectra of TiSe2 single crystal at different temperatures using 1.87 mW laser excitation power (λ= 514 nm; spot diameter is 1.5 μm and averaging time is 3 min). Each spectrum is taken at a pristine location on the freshly cleaved surface of TiSe2 single crystal. Clear onset of the laser damage threshold is at around 355 K. The inset shows an expanded view of the spectrum obtained with low excitation power (∼100 μW) at a location irradiated at 400 K.

FIG. 2.

Raman spectra of TiSe2 single crystal at different temperatures using 1.87 mW laser excitation power (λ= 514 nm; spot diameter is 1.5 μm and averaging time is 3 min). Each spectrum is taken at a pristine location on the freshly cleaved surface of TiSe2 single crystal. Clear onset of the laser damage threshold is at around 355 K. The inset shows an expanded view of the spectrum obtained with low excitation power (∼100 μW) at a location irradiated at 400 K.

Close modal

Using high resolution optical imaging, we find that after exposure to elevated temperatures, new structures start nucleating on what was initially an atomically flat TiSe2 surface (Fig. 3). With the help of local EDS, we identified these as agglomerates of selenium. The selenium is observed to have migrated from the bulk of the crystal to the surface and coalesced into islands that have random dimensions from few tens to several hundreds of nanometers. In this area of interest in Fig. 3, there were two particular locations where the laser beam from the Raman microscope induced additional damage to the one created by the elevated temperature. This additional damage by the Raman laser can be seen in the optical, atomic force microscope, and scanning electron microscope images marked as “Spot 1” and “Spot 2” in Fig. 3. One can clearly observe the additional effect of the laser beam irradiation—an area of about couple of micrometers laterally where the surface of the TiSe2 crystal has become rough. The migration and aggregation of selenium on the surface is directly linked to the appearance of the Raman peaks near 250 cm1. We have assigned the Raman peaks to different solid selenium allotropes.34 The 250 cm1 Raman peak is assigned to the A1-type symmetric stretching mode in amorphous selenium,35 the 241 cm1 is assigned to the same mode in rhombohedral selenium,34 the 233 cm1 is assigned to trigonal selenium36 and the 263 cm1 peak is assigned to monoclinic selenium. Detailed description of the assigned Raman modes in selenium nanosize clusters has been calculated earlier.37 

FIG. 3.

Optical (left), atomic force microscopy (middle) and scanning electron microscopy (right) images of TiSe2 single crystal surface exposed to elevated temperature and Raman microscope laser beam radiation. In the specific locations marked as “Spot 1” and “Spot 2”, Raman laser created additional damage during the Raman spectroscopy measurement. The effect of heating and laser irradiation caused the selenium to diffuse from the bulk to the surface of the crystal and create amorphous (near “Spot 2”) and crystalline (near “Spot 1”) nanosize selenium clusters. The crystal was exposed to 350 K for 10 min before the Raman measurements were taken with laser beam power of 3.32 mW (Spot 1) and 1.87 mW (Spot 2) for 3 min with the beam diameter of 1.5 μm. The scale bars indicate 1 μm.

FIG. 3.

Optical (left), atomic force microscopy (middle) and scanning electron microscopy (right) images of TiSe2 single crystal surface exposed to elevated temperature and Raman microscope laser beam radiation. In the specific locations marked as “Spot 1” and “Spot 2”, Raman laser created additional damage during the Raman spectroscopy measurement. The effect of heating and laser irradiation caused the selenium to diffuse from the bulk to the surface of the crystal and create amorphous (near “Spot 2”) and crystalline (near “Spot 1”) nanosize selenium clusters. The crystal was exposed to 350 K for 10 min before the Raman measurements were taken with laser beam power of 3.32 mW (Spot 1) and 1.87 mW (Spot 2) for 3 min with the beam diameter of 1.5 μm. The scale bars indicate 1 μm.

Close modal

To understand the relevant temperatures and activation energies for selenium diffusion, we conducted systematic experiments by varying both laser beam power and temperature. Each measurement was made on large atomically flat planes of the single crystal surface with no edges in the Raman microscope field of view. Raman spectrum at each temperature was taken for 3 min before blocking the laser beam and moving on to the next higher temperature setpoint. Each spectrum was taken on a different and pristine location on the surface of the single crystal in order to avoid effects from the previous laser irradiation. In the upper panel of Fig. 4, the Raman intensity at 250 cm1 is plotted versus temperature for multiple beam powers (Arrhenius plot), each showing different thresholds that lead to the onset of the peak increase with temperature. The horizontal dashed line indicates the noise level of the Raman signal. The threshold temperature beyond which the peak height at 250 cm1 starts increasing marks the onset of formation of selenium clusters on the TiSe2 crystal surface. The lower panel in Fig. 4 shows a linear relationship between the laser beam power and threshold temperature Tc for the detection of selenium clusters. The extrapolation of the curve to vanishing laser power leads to ordinate crossing at Tc0 = 440 K. This suggest a relatively low temperature threshold at which the process of selenium diffusion starts taking place without the assistance of any photons. Since this process is thermally activated, we fit the peak height vs. temperature with Arrhenius dependence and find that the activation energy for selenium vacancy generation is 1.55 ± 0.07 eV at the lowest laser power. This value is slightly higher from the one obtained for the generation of FeSex on the surface of Fe0.5TiSe2 (iron intercalated TiSe2) obtained by Titov et al.38 Although in that work it was not clear what was the rate determining process—the selenium or iron diffusion, both experiments indicate that the intercalation of the TMD lowers the activation barrier for the diffusion of the chalcogen.

FIG. 4.

Upper panel: Arrhenius plot of 250 cm1 Raman peak intensity at various temperatures and Raman microscope laser powers. The selenium diffusion activation energies were extracted from the fits and are shown next to each curve. The exposure time of 3 min was used for all experiments and each point was taken on unirradiated pristine area of the surface. The horizontal dashed line indicates the noise level of the Raman signal. Lower panel: dependence of the threshold temperature at which 250 cm1 Raman peak becomes observable on the Raman laser beam power. The linear fit extrapolates to 440 K at vanishing laser beam power.

FIG. 4.

Upper panel: Arrhenius plot of 250 cm1 Raman peak intensity at various temperatures and Raman microscope laser powers. The selenium diffusion activation energies were extracted from the fits and are shown next to each curve. The exposure time of 3 min was used for all experiments and each point was taken on unirradiated pristine area of the surface. The horizontal dashed line indicates the noise level of the Raman signal. Lower panel: dependence of the threshold temperature at which 250 cm1 Raman peak becomes observable on the Raman laser beam power. The linear fit extrapolates to 440 K at vanishing laser beam power.

Close modal

The results on the thermal and photon induced nucleation of selenium vacancies and the diffusion of selenium to the surface of TiSe2 have broader implications for both the fundamental studies of TiSe2 as well as its potential practical applications of TMDs. The interest in TiSe2 dichalcogenide has been due to its excitonic nature of charge density wave state and superconductivity in CuxTiSe239 and PdxTiSe2,40 as well as due to the unique observation of the chiral charge density wave.10,29 The fundamental studies of these correlated states and their coexistence have been on the rise recently, as the dichalcogenide system is structurally relatively simple and can provide some answers in regard to the mechanisms of correlated electron states in other classes of materials like perovskite high-temperature superconductors and pnictides. Since TiSe2 is a compensated semimetal with very large Hall coefficient at low temperatures,1 the correlated electron states are very susceptible to the level of intrinsic doping. Therefore, any fundamental studies should take into account the selenium vacancy concentration levels and the relatively low activation energy for their nucleation and diffusion.25 The concentration and dynamics of selenium vacancies and selenium atoms could be very important in the TiSe2 thin films,13,20,41 as the Raman peaks associated with these defects are strong in synthesized films.14,41 With reduced dimensionality of the system (from quasi-3D crystals to 2D thin films), the activation energy for selenium vacancy formation and selenium diffusion could further decrease and make the 2D system metastable even at room temperatures. The intrinsic doping due to vacancies and atomic selenium on the surface might suppress the charge density wave correlations and reduce the signature of charge density wave phase normally observed resistivity vs. temperature measurements.41 A possible method to prevent selenium diffusion and at the same time enhance the CDW transition temperature in 2D TiSe2 could be the encapsulation using hexagonal boron nitride, as shown recently.42 Clearly, more work needs to be done in this area to understand the dynamics of selenium in TiSe2 and with the recent advances in the tip-enhanced Raman spectroscopy43 and X-ray photoelectron diffraction,44 one could possibly obtain sufficient information to resolve this issue.

In conclusion, the Raman spectroscopy studies on single crystal TiSe2 show a combined temperature and photon-induced defect signatures associated with the nucleation of selenium nanosize clusters towards the surface of the crystal. Additional phonon peaks at near 250 cm1 are observed in the laser irradiated regions, consistent with the Raman spectra of selenium clusters. Temperature dependent studies of the threshold laser intensity necessary to form selenium vacancies in the bulk with selenium atoms diffusing to the surface suggest thermally activated character of the process. The extrapolation of the onset temperature of the photon-induced damage shows that the process of selenium vacancy nucleation could start at temperatures as low as 440 K without any photon assistance.

We would like to acknowledge the support by the National Science Foundation under Grant No. ECCS-1408151. The use of the Center for Nanoscale Materials, an Office of Science user facility, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. We would like to thank T. Polakovic for assistance with scanning electron microscopy. G.K. would like to acknowledge support by the Ministry of Science of Montenegro, under Contract No. 01-682.

1.
F. J.
Di Salvo
,
D. E.
Moncton
, and
J. V.
Waszczak
,
Phys. Rev. B
14
,
4321
(
1976
).
2.
J. A.
Wilson
,
Phys. Status Solidi B
86
,
11
(
1978
).
3.
H.
Guo
,
T.
Yang
,
M.
Yamamoto
,
L.
Zhou
,
R.
Ishikawa
,
K.
Ueno
,
K.
Tsukagoshi
,
Z.
Zhang
,
M. S.
Dresselhaus
, and
R.
Saito
,
Phys. Rev. B
91
,
205415
(
2015
).
4.
E.
Lorchat
,
G.
Froehlicher
, and
S.
Berciaud
,
ACS Nano
10
,
2752
(
2016
).
5.
H.
Cercellier
,
C.
Monney
,
F.
Clerc
,
C.
Battaglia
,
L.
Despont
,
M. G.
Garnier
,
H.
Beck
,
P.
Aebi
,
L.
Patthey
,
H.
Berger
, and
L.
Forro
,
Phys. Rev. Lett.
99
,
146403
(
2007
).
6.
C.
Monney
,
E. F.
Schwier
,
M. G.
Garnier
,
N.
Mariotti
,
C.
Didiot
,
H.
Cercellier
,
J.
Marcus
,
H.
Berger
,
A. N.
Titov
,
H.
Beck
, and
P.
Aebi
,
New J. Phys.
12
,
125019
(
2010
).
7.
M. M.
Fogler
,
L. V.
Butov
, and
K. S.
Novoselov
,
Nat. Commun.
5
,
4555
(
2014
).
8.
T.
Korn
,
S.
Heydrich
,
M.
Hirmer
,
J.
Schmutzler
, and
C.
Schueller
,
Appl. Phys. Lett.
99
,
102109
(
2011
).
9.
R.
Bhatt
,
R.
Basu
,
S.
Bhattacharya
,
A.
Singh
,
D. K.
Aswal
,
S. K.
Gupta
,
G. S.
Okram
,
V.
Ganesan
,
D.
Venkateshwarlu
,
C.
Surgers
,
M.
Navaneethan
, and
Y.
Hayakawa
,
Appl. Phys. A - Mater. Sci. Process.
111
,
465
(
2013
).
10.
J.
Ishioka
,
Y. H.
Liu
,
K.
Shimatake
,
T.
Kurosawa
,
K.
Ichimura
,
Y.
Toda
,
M.
Oda
, and
S.
Tanda
,
Phys. Rev. Lett.
105
,
176401
(
2010
).
11.
J.-P.
Castellan
,
S.
Rosenkranz
,
R.
Osborn
,
Q.
Li
,
K. E.
Gray
,
X.
Luo
,
U.
Welp
,
G.
Karapetrov
,
J. P. C.
Ruff
, and
J.
van Wezel
,
Phys. Rev. Lett.
110
,
196404
(
2013
).
12.
P.
Chen
,
Y. H.
Chan
,
X. Y.
Fang
,
Y.
Zhang
,
M. Y.
Chou
,
S. K.
Mo
,
Z.
Hussain
,
A. V.
Fedorov
, and
T. C.
Chiang
,
Nat. Commun.
6
,
8943
(
2015
).
13.
L. J.
Li
,
E. C. T.
O'Farrell
,
K. P.
Loh
,
G.
Eda
,
B.
Ozyilmaz
, and
A. H. C.
Neto
,
Nature
529
,
185
(
2016
).
14.
R.
Bhatt
,
S.
Bhattacharya
,
R.
Basu
,
S.
Ahmad
,
A. K.
Chauhan
,
G. S.
Okram
,
P.
Bhatt
,
M.
Roy
,
M.
Navaneethan
,
Y.
Hayakawa
,
A. K.
Debnath
,
A.
Singh
,
D. K.
Aswal
, and
S. K.
Gupta
,
ACS Appl. Mater. Interfaces
6
,
18619
(
2014
).
15.
Y.
Gu
,
Y.
Katsura
,
T.
Yoshino
,
H.
Takagi
, and
K.
Taniguchi
,
Sci. Rep.
5
,
12486
(
2015
).
16.
R.
Lv
,
J. A.
Robinson
,
R. E.
Schaak
,
D.
Sun
,
Y.
Sun
,
T. E.
Mallouk
, and
M.
Terrones
,
Acc. Chem. Res.
48
,
56
(
2015
).
17.
Z. Y.
Zhu
,
Y. C.
Cheng
, and
U.
Schwingenschlogl
,
Phys. Rev. B
84
,
153402
(
2011
).
18.
S.
Lebegue
and
O.
Eriksson
,
Phys. Rev. B
79
,
115409
(
2009
).
19.
K.
Sugawara
,
Y.
Nakata
,
R.
Shimizu
,
P.
Han
,
T.
Hitosugi
,
T.
Sato
, and
T.
Takahashi
,
ACS Nano
10
,
1341
(
2016
).
20.
J.-P.
Peng
,
J.-Q.
Guan
,
H.-M.
Zhang
,
C.-L.
Song
,
L.
Wang
,
K.
He
,
Q.-K.
Xue
, and
X.-C.
Ma
,
Phys. Rev. B
91
,
121113(R)
(
2015
).
21.
H.
Nishikawa
,
T.
Shimada
, and
A.
Koma
,
J. Vac. Sci. Technol. A
14
,
2893
(
1996
).
22.
B.
Hildebrand
,
C.
Didiot
,
A. M.
Novello
,
G.
Monney
,
A.
Scarfato
,
A.
Ubaldini
,
H.
Berger
,
D. R.
Bowler
,
C.
Renner
, and
P.
Aebi
,
Phys. Rev. Lett.
112
,
197001
(
2014
).
23.
E.
Pehlke
and
W.
Schattke
,
Z. Phys. B: Condens. Matter
66
,
31
(
1987
).
24.
A. S.
Razinkin
,
A. N.
Enyashin
,
T. V.
Kuznetsova
,
A. N.
Titov
,
M. V.
Kuznetsov
, and
A. L.
Ivanovskii
,
J. Struct. Chem.
51
,
737
(
2010
).
25.
A. M.
Novello
,
B.
Hildebrand
,
A.
Scarfato
,
C.
Didiot
,
G.
Monney
,
A.
Ubaldini
,
H.
Berger
,
D. R.
Bowler
,
P.
Aebi
, and
C.
Renner
,
Phys. Rev. B
92
,
081101(R)
(
2015
).
26.
W.
Zhao
,
P. H.
Tan
,
J.
Liu
, and
A. C.
Ferrari
,
J. Am. Chem. Soc.
133
,
5941
(
2011
).
27.
C. S.
Oglesby
,
E.
Bucher
,
C.
Kloc
, and
H.
Hohl
,
J. Cryst. Growth
137
,
289
(
1994
).
28.
P.
Husanikova
, “
Competing order parameters in CuxTiSe2
,” Ph.D. thesis,
Comenius University
, Bratislava,
2014
.
29.
M.
Iavarone
,
R.
Di Capua
,
X.
Zhang
,
M.
Golalikhani
,
S. A.
Moore
, and
G.
Karapetrov
,
Phys. Rev. B
85
,
155103
(
2012
).
30.
C. S.
Snow
,
J. F.
Karpus
,
S. L.
Cooper
,
T. E.
Kidd
, and
T. C.
Chiang
,
Phys. Rev. Lett.
91
,
136402
(
2003
).
31.
S.
Sugai
,
K.
Murase
,
S.
Uchida
, and
S.
Tanaka
,
Solid State Commun.
35
,
433
(
1980
).
32.
J. A.
Holy
,
K. C.
Woo
,
M. V.
Klein
, and
F. C.
Brown
,
Phys. Rev. B
16
,
3628
(
1977
).
33.
H.
Barath
,
M.
Kim
,
J. F.
Karpus
,
S. L.
Cooper
,
P.
Abbamonte
,
E.
Fradkin
, and
R. J.
Cava
,
Phys. Rev. Lett.
100
,
106402
(
2008
).
34.
K.
Nagata
,
T.
Ishikawa
, and
Y.
Miyamoto
,
Jpn. J. Appl. Phys., Part 1
22
,
1129
(
1983
).
35.
G.
Lucovsky
,
A.
Mooradian
,
W.
Taylor
,
G.
Wright
, and
R.
Keezer
,
Solid State Commun.
5
,
113
(
1967
).
36.
M.
Quintana
,
E.
Haro-Poniatowski
,
J.
Morales
, and
N.
Batina
,
Appl. Surf. Sci.
195
,
175
(
2002
).
37.
S.
Kohara
,
A.
Goldbach
,
N.
Koura
,
M.-L.
Saboungi
, and
L.
Curtiss
,
Chem. Phys. Lett.
287
,
282
(
1998
).
38.
A. N.
Titov
,
E. G.
Shkvarina
,
L. N.
Zelenina
,
T. P.
Chusova
, and
M. R.
Sharafutdinov
,
Phys. Solid State
58
,
735
(
2016
).
39.
E.
Morosan
,
H. W.
Zandbergen
,
B. S.
Dennis
,
J. W. G.
Bos
,
Y.
Onose
,
T.
Klimczuk
,
A. P.
Ramirez
,
N. P.
Ong
, and
R. J.
Cava
,
Nat. Phys.
2
,
544
(
2006
).
40.
E.
Morosan
,
K. E.
Wagner
,
L. L.
Zhao
,
Y.
Hor
,
A. J.
Williams
,
J.
Tao
,
Y.
Zhu
, and
R. J.
Cava
,
Phys. Rev. B
81
,
094524
(
2010
).
41.
P.
Goli
,
J.
Khan
,
D.
Wickramaratne
,
R. K.
Lake
, and
A. A.
Balandin
,
Nano Lett.
12
,
5941
(
2012
).
42.
L. J.
Li
,
W. J.
Zhao
,
B.
Liu
,
T. H.
Ren
,
G.
Eda
, and
K. P.
Loh
,
Appl. Phys. Lett.
109
,
141902
(
2016
).
43.
N.
Jiang
,
D.
Kurouski
,
E. A.
Pozzi
,
N.
Chiang
,
M. C.
Hersam
, and
R. P. V.
Duyne
,
Chem. Phys. Lett.
659
,
16
(
2016
).
44.
M.
Kuznetsov
,
I.
Ogorodnikov
,
A.
Vorokh
,
A.
Rasinkin
, and
A.
Titov
,
Surf. Sci.
606
,
1760
(
2012
).