Excitons in atomically thin black phosphorus

0.46 MB·

by  A. Surrente

Excitons in atomically thin black phosphorus
A. Surrente,1 A. A. Mitioglu,1,2,∗ K. Galkowski,1 W. Tabis,1,3 D. K. Maude,1 and P. Plochocka1,†
1Laboratoire National des Champs Magn´etiques Intenses,
UPR 3228, CNRS-UGA-UPS-INSA, Grenoble and Toulouse, France
2Institute of Applied Physics, Academiei Str. 5, Chisinau, MD-2028, Republic of Moldova
3AGH University of Science and Technology, Faculty of Physics and
Applied Computer Science, Al. Mickiewicza 30, 30-059 Krakow, Poland
(Dated: March 16, 2016)
Raman scattering and photoluminescence spectroscopy are used to investigate the optical prop-
6
erties of single layer black phosphorus obtained by mechanical exfoliation of bulk crystals under
1
an argon atmosphere. The Raman spectroscopy, performed in situ on the same flake as the pho-
0
toluminescence measurements, demonstrates the single layer character of the investigated samples.
2
Theemissionspectra,dominatedbyexcitoniceffects,displaytheexpectedinplaneanisotropy. The

r emission energy depends on the type of substrate on which the flake is placed due to the differ-
a
ent dielectric screening. Finally, the blue shift of the emission with increasing temperature is well
M
described using a two oscillator model for the temperature dependence of the band gap.

5
1
Blackphosphorus, themoststableofalltheallotropes simultaneous observation of charged exciton emission at

of phosphorus, has been intensively studied by different around 1.62eV.25 Using scanning tunneling microscopy
]
l experimental methods from the early fifties of the last the band gap of a single layer of black phosphorus
l
a century.1–7 Bulk black phosphorus is a semiconductor, was estimated to be 2.05eV.26 The exciton binding
h with a band gap of about 0.335eV.1,4 The orthorombic energy and consequently the emission energy strongly
-
s bulk crystal has a layered structure, with atomic lay- depend on the dielectric environment (substrate)10,27.
e ers bound by weak van der Waals interactions. A sin- This could partially explain the wide range of values
m
gle atomic layer is puckered, with the phosphorus atoms for the black phosphorus emission energy found in the
. being parallel in the (010) plane.3,8,9 Atomically thin liteTrature, possibly related also to a slightly different
t
a monolayers have been recently isolated using mechani- Fcomposition of the SiO substrates employed. Moreover,
2
m cal exfoliation,10 adding black phosphorus to the rapidly A owing to the limited life time of the samples, the various
- growing family of emerging two dimensional materialsR. characterization techniques used to identify monolayer
d
The band gap of black phosphorus is always directDand black phosphorus (e.g. atomic force microscopy, Raman
n
canbetunedfrom0.3eVtothenearlyvisiblepartofthe spectroscopy) could not always be performed on the
o
c spectrum11,12. Incontrast,grapheneisgapless,13andthe same flake where the optical response was investigated.
[ transition metals dichalcogenides (TMDs) have an indi- In this paper we present a systematic investigation of

rect gap in bulk phase and only monolayer TMDs have theopticalpropertiesofmonolayersofblackphosphorus.
2
v a direct gap.14 Moreover, black phosphorus exhibits a Weanalyzethepropertiesoftheemissionasafunctionof
0 strong in-plane anisotropy11,12,15–17, absent in graphene thedielectricconstantofthesubstrate,excitationpower,
1 and TMDs. Additionally, the relatively high mobilities polarizationandtemperature. Thesinglelayercharacter
1 measuredatroomtemperaturecombinedwiththedirect of the investigated flakes is demonstrated using in situ
1 band gap result in an on/off ratio for FET transistors of RamanmeasurementsonthesameflakeusedforthePL.
0 the order of 105.12,18,19 WeshowthatthePLemissionenergyofmonolayerblack
.
1 Although black phosphorus has a wide range phosphorusdependsonthesubstrateused. ThePLspec-
0 of possible applications, including tunable tra, dominated by excitonic effects, exhibit the expected
6
photodetectors20, field effect transistors12,16,18,19 or inplaneanisotropy. Finally,theblueshiftoftheemission
1
: photon polarizers11,12,15–17, many of its electronic prop- energy with an increasing temperature is well described
v erties are not yet fully understood. The main difficulty with a two-oscillator model for the temperature depen-
Xi arises from the sensitivity of black phosphorus to its dence of the band gap.
environment, notably its high reactivity when exposed Singleandfewlayerblackphosphorusflakeshavebeen
r
a to air and laser light.10,21,22 Theoretical calculations obtainedbymechanicalexfoliationofabulkcrystal,pur-
predict a band gap of monolayer black phosphorus chased from Smart Elements (99.998% nominal purity).
ranging from 1eV11,19 to 2.15eV10,11, with a binding Themechanicalexfoliationwasperformedinaglovebox
energy of the neutral exciton of 0.8eV in vacuum11 filled with argon (Ar) gas (< 1ppm O2, < 1ppm H2O).
and of 0.38eV when placed on a SiO substrate10. The flakes were subsequently transferred onto a Si sub-
2
The measured photoluminescence (PL) emission energy strate, in most cases capped with a 300nm thick layer
from bilayers black phosphorus is between 1.2eV23 of SiO2 (Si/SiO2 hereafter). The samples were stored in
and 1.6eV10, while monolayer black phosphorus shows vials, in the glove box, before their transfer to the cryo-
neutral exciton emission around 1.3eV24, 1.45eV19 stat under an Ar atmosphere.
or 1.76eV25. The latter value was related to the For the optical measurements, the samples were
2
mounted on the cold finger of a He-flow cryostat,
which was then rapidly pumped to a pressure below
1×10−4mbar, minimizing the exposure of the black A 1g A 2g T = 4 K
phosphorus to air. The excitation was provided by a s) B 2 g
it
frequency-doubled diode laser, emitting at 532nm and n
u S i
focused on the sample by a 50× microscope objective .
(0.55numericalaperture),yieldingaspotsizeof∼1µm. rb m
a B u lk 2 6 0 W
The PL and Raman signals were collected through the (
y
sameobjectiveandanalyzedbyaspectrometerequipped it
s
with a liquid nitrogen cooled Si CCD camera. en M L 2 6 0 m W
t
Raman spectroscopy, which has been shown to be a In M L 1 7 m W
very precise tool for the determination of the number of
layers in TMDs14,28–31 and in graphene,32,33 can also be 3 6 0 4 0 0 4 4 0 4 8 0 5 1 0 5 4 0
usedtoidentifymonolayerblackphosphorus. Bulkblack R a m a n s h if t ( c m -1 )
phosphorus belongs to the D18 space group. Of the 12
2h
normal modes at the Γ point of the Brillouin zone, six
are Raman active.7,34 The two B modes are forbidden FIG. 1. (Color online) Typical µRaman spectra of bulk and
3g
singlelayerblackphosphorusonSi/SiO measuredatT =4K
in back scattering configuration and the B mode at 2
1g fordifferentexcitationpowersasindicated. Thedashedlines
194cm−1 is very weak. Therefore, only three Raman
indicate the position of the three Raman modes A1, A2 and
modes are expected for bulk black phosphorus: A1, A2 g g
g g B2g in bulk black phosphorus.
and B at around 365cm−1, 470cm−1 and 442cm−1,
2g
respectively.7,34 The A2 and B modes are related to
g 2g
the in plane vibrations of the atoms, while A1 mode is
g high excitation power (260µW, the same as used for the
related to the out of plane movement of the atoms7,34.
bulk crystal) does not show any additional features; the
Typical calibrated micro Raman (µRaman) spectra of peak positions remain the same as for the data obtained
the bulk and single layer black phosphorus are presented for low excitation power. Thus there is no sign of laser
in Fig. 1. For the bulk crystal three strong peaks are inducedchemicalmodificationofoursamples,10 whichis
observed at 363.7cm−1, 440.3cm−1 and 467.5cm−1 cor- importantsincetheexcitationintensityusedhereiscom-
responding well to the expected main Raman modes. A parabletothatusedinourmicroPL(µPL)spectroscopy
further peak was systematically observed at 520cm−1 measurements.
(Raman mode of the Si substrate), confirming the cor-
In Fig. 2(a) we show representative low temperature
rect calibration of our Raman setup. The slight shift µPLspectraforthemonolayerflakepreviouslycharacter-
towards lower frequencies, compared to the previously
ized using Raman spectroscopy. The measurement was
cited literature values, can be related to the low temper-
performed without removing the flake from the cryostat,
ature(4K)atwhichourmeasurementshavebeencarried
which was maintained under vacuum. The spectrum is
out.23 Underlowexcitationpower(17µW),initiallyused
dominated by a very strong emission line centered at
to avoid any risk of inducing damage by exposure to the
around 2eV identified with neutral exciton recombina-
laser light, the Raman spectrum of the monolayer flake
tion. This line is consistently observed at this energy for
is similar to that of the bulk crystal. This suggests that
all flakes transferred to Si/SiO substrates. The typical
2
the monolayer black phosphorus remains crystalline and fullwidthathalfmaximumisof(cid:46)90meV(linebroaden-
no oxidization occurred during the transfer of the sam-
ing induced by scattering with vacancies and impurities
ple from the glove box to the cryostat. Compared to
in the exfoliated flakes). The observed value of the exci-
bulk,theA1 modedoesnotshiftwithinexperimentaler-
g ton recombination energy is relatively high as compared
ror. AlthoughthisRamanmodewasfoundtosoftenina
tothetheoreticalbandgapof2.15eVcomputedforblack
monolayer sample19, it was generally found to be rather
phosphorus single layer,10,11 suggesting that the investi-
insensitive to the number of layers.35,36 In contrast, the
gated sample is indeed a single layer. This is somewhat
position of A2 and B Raman modes shift to higher fre-
g 2g larger than the values already reported for black phos-
quencies in the monolayer flake. The A2g mode shifts by phorussinglelayersof(cid:104)1.3−1.76eV19,24,25. Thiscould
1.7±0.2cm−1 while the B2g shifts by 1.0±0.2cm−1. The be partly ascribed to the different stoichiometry of the
change of the energy of the Raman modes as a function SiO layers used as a substrate, which induces a shift in
2
of the number of layers in black phosphorus has been theemissionbecauseofthedifferentdielectricconstantof
already shown for both high10,19,35 and low frequency the surrounding medium.38 To verify this hypothesis, we
modes37. The shifts we measure are in good agreement have transferred black phosphorus flakes onto a Si sub-
with the reported values for single layer black phospho- strate covered by a thin (∼ 2nm) layer of native oxide.
rus (∼ 1.9cm−1 and around ∼ 1.0cm−1 for A210,19,35,36 Suchasubstratehasalargerdielectricconstantthanthe
g
andB 10,35,respectively)demonstratingthesinglelayer standard Si/SiO substrates. Thus, the exciton binding
2g 2
character of the investigated flake. Raman measured at energy is expected to be lower and the emission energy
3
shouldbeblueshifted. AtypicalµPLspectrumofaflake (a) P=80m W, T=4.2K (b) P=300m W, T=4.2K
tWflermaahkinesislsefoieonrtnrhaeeednSmeio/rnagStiyinoOo2asfps∼uSecbi8tsr0staurmlabtesfetVe,raa.ttthuTeerehriesissirsscehosaonewmlfianrrbmgleiesnstthhFhieoifgtsd.eoe2fpo(tfeahn)ae-. L intensity mber 64 SSi/iSiO2 X X b. units) ddeett.. HV EExxcc.. HH
dmeenncte, ionfatghreeeemmeisnstiownitehnperrgevyioounstrheesudltise.l1e0c,2tr7icTheneveiffroenct- zed P Nu201.90 2.00 2.10 T nt. (ar
of the dielectric environment seems to be stronger than ali Energy (eV) T L i
foratomicallythinTMDs,inagreementwiththeoryand m P
r
experiment.39 However, we note that the effect of the No1.4 1.6 1.8 2.0 2.2 1.4 1.6 1.8 2.0 2.2
Energy (eV) Energy (eV)
dielectric environment alone is too small to account for
thelargevariationoftheemissionenergyreportedinthe
literature.19,24,25
FIG. 2. (Color online) (a) Low temperature µPL spectra of
In the spectrum measured on a Si/SiO substrate, in
2 black phosphorus flakes on different substrates. X labels the
addition to the neutral exciton recombination, we ob- neutral exciton transition, T the charged exciton. The insert
served additional features at ∼ 1.84eV, which can be showsahistogramoftheemissionenergyforthe21flakesin-
attributed to charged exciton recombination (systemat- vestigated. Themeanemissionenergyis1.991eVwithastan-
ically present in all the investigated flakes)25 together darddeviationσ=0.036eVforflakesonaSi/SiO2 substrate
with a low energy peak, typically at ∼1.72eV, and pos- and 2.102eV with σ = 0.015eV for flakes on a Si substrate.
sibly related to excitons bound to impurities. The low (b) Polarization resolved µPL spectra of a black phosphorus
energy feature is not always present, see e.g. the µPL flake on Si.
spectrum on a Si substrate in Fig. 2(a).
Because of the low symmetry of the crystal structure
In Fig. 3(b) we present the dependence of the inte-
and the screening in black phosphorus11,40, the exciton
grated PL intensity I of neutral and charged excitons
wave function is expected to be squeezed along the arm-
as a function of the excitation power. At low excita-
chair direction, resulting in a polarization dependent PL
tion power, both intensities increase linearly with the
emission24. The linear polarization dependence of the
power (indicating the absence of biexciton emission), as
emission of our flakes was investigated by mounting po-
demonstrated by the fits to the power law I ∝Pn, with
larizers/analyzers in both the excitation and the detec-
n=0.97±0.03fortheneutralexcitonandn=0.93±0.03
tion paths. For any fixed orientation of the analyzer in
forthechargedexciton. Athigherexcitationpower,both
thedetectionpath,thedetectedsignalisalwaysstronger
excitonic lines saturate at approximately the same level
for horizontally (H) polarized excitation. This stems
of excitation power [P = 1mW, see insert to Fig. 3(b)],
fromthepolarizationdependentabsorptionpropertiesof
confirming that the two observed transitions are related
blackphosphorus16. Regardlessofthepolarizationofthe
to the recombination of a single electron-hole pair.
excitation beam, whenever the analyzer in the detection
path was set to H, the intensity of the detected signal
(a) (b)
wanaissoatrmopaixcimnautmur,ewohfitchheiesxccoitnosnistinenbtlawcikthphtohsepshtorrounsg2l4y. P=2000mTW X s)200 ensity430000 X
woeFxroictTrvhitheearaxettaiiHecomvanoppllpol(ueolVt,awi)roieninzdreiFogrdeifigvclta.ethsis2oee(inrnbµfl)oPio,gfrLhwmtthesaepatsneiahocdnontwardcaleyoµtaznePesccrLeta.erndsfpuinwnegcictttthrihoaeenientxohacfetiruttherHdee sity (arb. units) PP==2200m0Wm W ensity (cnts eV/115000 Integrated int2100000 0 Po2Xw00e0r (m W40)00T
opfowtheer odbepseernvdeedntcreanofsittihoenµliPnLes.spWecetrhaafvoermseevaesruarledblathcke Inten P=2m W ed int 50 T
phosphorus monolayer flakes. In Fig. 3(a) we show rep- at
gr 0
resentative spectra measured at low, intermediate, and e
high excitation powers P at T = 4.2K. Even after ex- 1.7 1E.8ner1g.y9 (e2V.)0 2.1 Int 0 5P0ow1e0r 0(m 1W5)0 200
citing with P > 1mW, the emission efficiency did not
decrease for any of the investigated samples, suggesting
that our preparation method helps improve the stabil- FIG.3. (Coloronline)(a)µPLspectraofablackphosphorus
ity of the exfoliated black phosphorus flakes. Two peaks singlelayeronSi/SiO measuredatdifferentexcitationpow-
2
appear in the spectra of Fig. 3(a) for all excitation pow- ers. ThebrokencurvesareGaussianfitsusedtocalculatethe
ers used. The neutral exciton peak slightly blue shifts integrated intensity. (b) Integrated intensity of the charged
((cid:46) 2meV) at high excitation powers, possibly due to a (T)andneutral(X)excitonemissionversusexcitationpower.
The solid lines are fits to a power law described in the text.
localized heating effect induced by the high power of the
incoming laser beam (see also below for a discussion of
temperature dependence of the black phosphorus PL). In view of the potential use of black phosphorus in
4
a wide variety of electronic and optoelectronic appli- tion method preserves the optical properties of the black
cations, we have investigated the temperature depen- phosphorus flakes. The shift of the Raman modes mea-
dence of its optical properties. In Fig. 4(a) we show sured on the exfoliated flakes with respect to the bulk
normalized µPL spectra measured at different temper- (a) P=160m W (b) 2.06
atsohtunifrtepsse.(amkWebairstouharedindecnrdse.EasgTi/nhgdeTtnem∼eupt3erar.1alt×euxr1ce0it−toh4neeVecmh/aKisrsgibeodentwebxelcueine- s) 226000KK ergy (eV) 22..0042
4p0orKtsaonfdth1e60teKm),pewrahtiuchreidsepcoennsdiestnecnetowfitthheebaarnlidergraep- b. unit 160K ak en 2.00
oorf b2u.3lk3×bla1c0k−4pehVos/pKh4o1r)u.s T(dhEeg/bedhTav=io2r.8of×t1h0e−e4meViss/iKon2 sity (ar 80K (Pec) 1.0 Tem1p0e0rature2 0(K0)
energy of the excitonic peak is shown more in detail in n
Fig. 4(b). In bulk semiconductors, the variation of the Inte 50K sity 0.8
n 0.6
band gap as a function of the temperature is direct con- e
nt 0.4
sequence of the renormalization of the band gap via the 4K m. i 0.2
electron-phononinteractionandofthethermalexpansion or
of the lattice.42 In the framework of the two-oscillator 1.6 1.7 1.8 1.9 2.0 2.1 N
0.0 0.1 0.2
model43–46, the band gap E is approximated by Energy (eV) 1/T (K-1)
g
(cid:32) (cid:33) (cid:32) (cid:33)
2 2
E (T)=E +E +1 +E +1 , FIG.4. (Coloronline)(a)µPLspectraofablackphosphorus
g 0 1 (cid:126)ω1 2 (cid:126)ω2
ekBT −1 ekBT −1 single layer on Si/SiO2 measured at different temperatures.
The broken lines are Gaussian fits used to extract the emis-
where E0 is the bare band gap (i.e. the low tempera- sion energy. (b) Emission energy of the neutral exciton as a
turebandgapexhibitedintheabsenceofzeropointmo- function of the temperature. (c) Integrated intensity of the
tion), E +E is the renormalization energy and (cid:126)ω = neutralexcitonversustemperature. Thesolidlinesarefitsto
1 2 1
17.23meV and (cid:126)ω = 52.82meV denote the two oscil- the models described in the text.
2
lator energies, as extracted from the computed phonon
density of states of monolayer black phosphorus.47 The
neutralexcitonemissionenergyisthengivenbyE (T)− black phosphorus confirms the single layer character of
g
E , where E is the exciton binding energy. The solid the exfoliated black phosphorus flakes. The measured
X X
curve shown in Fig. 4(b) is obtained by fitting the ex- µPL spectra exhibit strong emission lines, attributed
perimental data, yielding (E − E ) = 2.22 ± 0.02eV to the recombination of neutral and charged excitonic
0 X
(larger than the observed low T transition energy, sug- complexes. It is not clear if the enhanced emission en-
gesting the occurrence of band gap renormalization due ergy, compared to the previously reported values,19,24,25
to electron-phonon interaction), E = 72.6 ± 0.3meV, is linked with a larger intrinsic band gap or a reduced
1
and E =−297±3meV. exciton binding energy. The excitonic nature of the ob-
2
The integrated intensity of the excitonic transition as served PL was confirmed by the observed polarization
a function of the inverse substrate temperature T−1 is dependenceandbythenearlylinearincreaseoftheemis-
shown in Fig. 4(c). With increasing temperature, the sion intensity with the excitation power. The increase in
emissionintensitydecreases,owingtothethermalactiva- the emission energy with temperature was modeled with
tion of non-radiative recombination centers. To quantify a two-oscillator model to account for the temperature
theactivationenergyE , theexperimentaldataisfitted dependence of semiconductor band gap.
A
using I(T)/I(T = 0) = 1/(1+ae−EA/kBT),48 where a
is related to the ratio of the radiative and non-radiative
lifetimes,49 to give E = 10±1meV, and a = 6.8±2.
A
ACKNOWLEDGMENTS
The significantly lower value of E as compared to the
A
computedexcitonbindingenergy10 confirmsthatthede-
crease of the PL intensity at high T is brought about by The authors gratefully acknowledge Baptiste Vignolle
the thermal occupation of non radiative recombination for his assistance with the glove box and for his careful
centers rather than the dissociation of the excitons. proof reading and Geert Rikken for providing the bulk
In summary, we have performed a detailed investiga- black phosphorus. AAM acknowledges financial support
tion of the optical properties of monolayer black phos- from the French foreign ministry. This work was par-
phorus mechanically exfoliated in an Ar atmosphere. No tially supported by ANR JCJC project milliPICS, the
significant degradation of the PL emission was induced R´egion Midi-Pyr´en´ees under contract MESR 13053031
by the laser illumination, suggesting that this prepara- and STCU project 5809.
5
∗ Present address: High Field Magnet Laboratory (HFML- Nanotechnology 10, 517 (2015).
EMFL), Institute for Molecules and Materials, Rad- 25 J. Yang, R. Xu, J. Pei, Y. W. Myint, F. Wang, Z. Wang,
boud University, Toernooiveld 7, 6525 ED Nijmegen, The S.Zhang,Z.Yu, andY.Lu,Light-Science&Applications
Netherlands 4, e312 (2015).
† [email protected] 26 L. Liang, J. Wang, W. Lin, B. G. Sumpter, V. Meunier,
1 R. W. Keyes, Physical Review 92, 580 (1953). and M. Pan, Nano Letters 14, 6400 (2014).
2 D. Warschauer, Journal of Applied Physics 34, 1853 27 A.H.Woomer,T.W.Farnsworth,J.Hu,R.A.Wells,C.L.
(1963). Donley, and S. C. Warren, ACS Nano 9, 8869 (2015).
3 Y. Maruyama, S. Suzuki, K. Kobayashi, and S. Tanuma, 28 C. Lee, H. Yan, L. E. Brus, T. F. Heinz, J. Hone, and
Physica B+C 105, 99 (1981). S. Ryu, ACS Nano 4, 2695 (2010).
4 H. Asahina, Y. Maruyama, and A. Morita, Physica B+C 29 H. R. Gutirrez, N. Perea-Lpez, A. L. Elas, A. Berkdemir,
117118, Part 1, 419 (1983). B.Wang, R.Lv, F.Lpez-Uras, V.H.Crespi, H.Terrones,
5 J. C. Jamieson, Science 139, 1291 (1963). and M. Terrones, Nano Letters 13, 3447 (2013).
6 J. Wittig and B. T. Matthias, Science 160, 994 (1968). 30 H. Li, Q. Zhang, C. C. R. Yap, B. K. Tay, T. H. T. Ed-
7 S.SugaiandI.Shirotani,SolidStateCommunications53, win, A.Olivier, andD.Baillargeat,AdvancedFunctional
753 (1985). Materials 22, 1385 (2012).
8 S.Narita,S.Terada,S.Mori,K.Muro,Y.Akahama, and 31 S.-L. Li, H. Miyazaki, H. Song, H. Kuramochi, S. Naka-
S.Endo,JournalofthePhysicalSocietyofJapan52,3544 harai, and K. Tsukagoshi, ACS Nano 6, 7381 (2012).
(1983). 32 A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi,
9 Y.Takao,H.Asahina, andA.Morita,JournalofthePhys- M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S.
ical Society of Japan 50, 3362 (1981). Novoselov, S. Roth, and A. K. Geim, Physical Review
10 A.Castellanos-Gomez,L.Vicarelli,E.Prada,J.O.Island, Letters 97, 187401 (2006).
K.L.Narasimha-Acharya,S.I.Blanter,D.J.Groenendijk, 33 A.Gupta,G.Chen,P.Joshi,S.Tadigadapa, andEklund,
M. Buscema, G. A. Steele, J. V. Alvarez, H. W. Zandber- Nano Letters 6, 2667 (2006).
gen,J.J.Palacios, andH.S.J.vanderZant,2DMaterials 34 S.Sugai,T.Ueda, andK.Murase,JournalofthePhysical
1, 025001 (2014). Society of Japan 50, 3356 (1981).
11 V. Tran, R. Soklaski, Y. Liang, and L. Yang, Physical 35 A. Favron, E. Gaufres, F. Fossard, A.-L. Phaneuf-
Review B 89, 235319 (2014). L’Heureux, N. Y.-W. Tang, P. L. Levesque, A. Loiseau,
12 X. Ling, H. Wang, S. Huang, F. Xia, and M. S. Dres- R. Leonelli, S. Francoeur, and R. Martel, Nature Materi-
selhaus, Proceedings of the National Academy of Sciences als 14, 826 (2015).
112, 4523 (2015). 36 W.Lu,H.Nan,J.Hong,Y.Chen,C.Zhu,Z.Liang,X.Ma,
13 K. S. Novoselov, A. K. Geim, S. V. Morozow, D. Jiang, Z.Ni,C.Jin, andZ.Zhang,NanoResearch7,853(2014).
M. I. Katsnelson, I. V. Grigorieva, S. V. Dubonos, and 37 X.Luo,X.Lu,G.K.W.Koon,A.H.C.Neto,B.O¨zyilmaz,
A. A. Firsov, Nature 438, 197 (2005). Q.Xiong, andS.Y.Quek,NanoLetters15,3931(2015).
14 Q. H. Wang, K. Kalantar-Zadeh, A. Kis, J. N. Coleman, 38 J. H. Choi and M. S. Strano, Applied Physics Letters 90,
and M. S. Strano, Nature Nanotechnology 7, 699 (2012). 3114 (2007).
15 T. Low, A. S. Rodin, A. Carvalho, Y. Jiang, H. Wang, 39 Y. Lin, X. Ling, L. Yu, S. Huang, A. L. Hsu, Y.-H. Lee,
F. Xia, and A. H. Castro Neto, Physical Review B 90, J.Kong,M.S.Dresselhaus, andT.Palacios,NanoLetters
075434 (2014). 14, 5569 (2014).
16 F. Xia, H. Wang, and Y. Jia, Nature Communications 5, 40 P. Li and I. Appelbaum, Physical Review B 90, 115439
4458 (2014). (2014).
17 J. Qiao, X. Kong, Z.-X. Hu, F. Yang, and W. Ji, Nature 41 M. Baba, Y. Nakamura, K. Shibata, and A. Morita,
Communications 5, 4475 (2014). Japanese Journal of Applied Physics 30, L1178 (1991).
18 L. Li, Y. Yu, G. J. Ye, Q. Ge, X. Ou, H. Wu, D. Feng, 42 M. Cardona and M. L. W. Thewalt, Reviews of Modern
X. H. Chen, and Y. Zhang, Nature Nanotechnology 9, Physics 77, 1173 (2005).
372 (2014). 43 L. Vin˜a, S. Logothetidis, and M. Cardona, Physical Re-
19 H. Liu, A. T. Neal, Z. Zhu, Z. Luo, X. Xu, D. Toma´nek, view B 30, 1979 (1984).
and P. D. Ye, ACS Nano 8, 4033 (2014). 44 M.CardonaandR.K.Kremer,ThinSolidFilms571,680
20 M.Buscema,D.J.Groenendijk,S.I.Blanter,G.A.Steele, (2014).
H. S. J. van der Zant, and A. Castellanos-Gomez, Nano 45 P. Dey, J. Paul, J. Bylsma, D. Karaiskaj, J. M. Luther,
Letters 14, 3347 (2014). M. C. Beard, and A. H. Romero, Solid State Communi-
21 S. P. Koenig, R. A. Doganov, H. Schmidt, A. H. Cas- cations 165, 49 (2013).
tro Neto, and B. O¨zyilmaz, Applied Physics Letters 104, 46 H. J. Lian, A. Yang, M. L. W. Thewalt, R. Lauck, and
103106 (2014). M. Cardona, Physical Review B 73, 233202 (2006).
22 Z.X.Gan,L.L.Sun,X.L.Wu,M.Meng,J.C.Shen, and 47 Y. Aierken, D. C¸akır, C. Sevik, and F. M. Peeters, Phys-
P. K. Chu, Applied Physics Letters 107, 021901 (2015). ical Review B 92, 081408 (2015).
23 S. Zhang, J. Yang, R. Xu, F. Wang, W. Li, M. Ghufran, 48 M. Leroux, N. Grandjean, B. Beaumont, G. Nataf, F. Se-
Y.-W. Zhang, Z. Yu, G. Zhang, Q. Qin, and Y. Lu, ACS mond, J. Massies, and P. Gibart, Journal of Applied
Nano 8, 9590 (2014). Physics 86, 3721 (1999).
24 X. Wang, A. M. Jones, K. L. Seyler, V. Tran, Y. Jia, 49 Y.Fang,L.Wang,Q.Sun,T.Lu,Z.Deng,Z.Ma,Y.Jiang,
H. Zhao, H. Wang, L. Yang, X. Xu, and F. Xia, Nature H.Jia,W.Wang,J.Zhou, andH.Chen,ScientificReports
6
5, 12718 (2015).