石墨烯, 掺杂, graphene, doping
Observation of Raman G-band splitting in top-doped few-layer graphene
Matteo Bruna1and Stefano Borini1,*
1Electromagnetic Division,INRIM,Strada delle Cacce91,I-10135Torino,Italy
?Received31August2009;revised manuscript received16February2010;published19March2010?An experimental study of Raman scattering in N-layer graphene as a function of the top layer doping is reported.At high doping level,achieved by a CHF3plasma treatment,we observe a splitting of the G band in the spectra of bilayer and4-layer graphene?N even?,whereas the splitting is not visible in case of monolayer and trilayer graphene?N odd?.The different behaviors are related to distinct electron-phonon interactions, which are affected by symmetry breaking and Fermi-level position in different ways in the various N-layer graphenes.In trilayer graphene,a weakening of the electron-phonon coupling as a function of the Fermi energy induces a hardening of all zone-center in-plane optical-phonon modes,such as in monolayer graphene.On the other hand,in4-layer graphene two distinct trends are observed in the G band as a function of doping, suggesting the presence of two different groups of electron-phonon interactions,such as in bilayer graphene. DOI:10.1103/PhysRevB.81.125421PACS number?s?:78.30.Na,63.20.K?
I.INTRODUCTION
Since the demonstration of the isolation of a single atomic plane of graphite?graphene?on a standard SiO2/Si substrate,1it has suddenly become possible to experimentally verify many theoretical predictions about the peculiar physi-cal behavior of graphene-based systems.Indeed,the linear E?k dispersion curves,and the consequent relativisticlike behavior of charge carriers in monolayer graphene,have been broadly con?rmed by several experimental observations.2,3Interestingly,stacking a number N of graphene layers on top of each other can lead to new physi-cal systems exhibiting completely different properties.For instance,some peculiar gate-tunable electronic and optical properties have been recently reported in the bilayer case.4–7 In fact,the interlayer coupling induces a gradual depar-ture from the electronic bands of monolayer graphene,8until the bulk limit?graphite?is reached for N large enough. Therefore,there is a range of N where the physical properties of graphene stacks are sensitive even to a variation?N=1. For instance,a qualitative difference between the transport properties of bilayer and trilayer graphene was evidenced in recent experiments.9Moreover,the optical-absorption spec-tra are predicted to systematically vary with the layers num-ber within the effective-mass approximation for1?N?6,8 re?ecting the gradual modi?cation of the band structure. Here we show that a clear splitting of the Raman G band is observed for N=2and N=4,when the multilayer graphene symmetry is broken by heavy doping of the top layer, whereas the splitting is not observed for N=1and N=3.The presence or absence of splitting highlights different electron-phonon interactions,which are in?uenced by doping and symmetry breaking in distinct ways in the various N-layer graphenes.The reported results con?rm two experimental reports in literature about the G band splitting in bilayer graphene?obtained by gate?eld effect?,10,11adding new im-portant information such as a systematic study of the split-ting as a function of the layers number and of the doping.
Raman spectroscopy is a very powerful tool for studying graphene,yielding information on the electronic structure and on the electron-phonon coupling?EPC?in the material.12,13In fact,this technique allows to clearly distin-
guish a monolayer and a bilayer from a few-layer graphene,
through the analysis of the2D band,14and to estimate the
charge-carrier density and type in monolayer graphene from
the spectral positions and relative intensities of the G and2D
bands.15The G band?at?1580cm?1?is due to a?rst-order Raman-scattering process involving zone-center in-plane op-
tical phonons.In stacked graphene layers,the vibrations in
different atomic planes can combine with each other in vari-
ous ways,depending on the number of layers and on the
stacking order.Moreover,the EPC is affected by the number
and symmetry of the stacked layers so that the G band may
be used to study the effects of symmetry breaking and dop-
ing on the electronic and phononic properties of multilayer
graphenes.The knowledge of these effects is of fundamental
importance for the development of graphene-based?eld-
effect electronic devices.
II.EXPERIMENTAL
Graphene layers studied in this work were deposited on a
285nm thick SiO2on Si substrates by adhesive tape exfolia-
tion of natural graphite.Then,the samples were analyzed by
optical microscopy,in order to estimate the number of
graphene layers composing the deposited thin?akes.It can
be seen in Fig.1that the contrast?de?ned as1?R G/R S, where R G and R S are the re?ected light intensities from the SiO2/Si substrate with and without graphene,respectively?measured on many semitransparent?akes increases in a step-wise manner.The analysis of the2D Raman band con?rmed that the lowest two steps correspond to monolayer and bi-layer graphenes,indicating that few graphene layers can be counted by contrast analysis.Such a behavior is related to the optical absorption of graphene,which was found to be directly proportional to the number of layers for N small enough.16Using appropriate?lters in order to select the wavelengths at which the contrast variation is high,we were able to distinguish up to6layers.Moreover,the experimental contrast values were checked by theoretical calculations within the Fresnel coef?cients approach.17After a prelimi-nary annealing in vacuum?1?10?5mbar for2days?in or-
PHYSICAL REVIEW B81,125421?2010?
der to remove possible adsorbed impurities from the graphene surface,the ?rst run of Raman measurements was carried out.Raman spectra were acquired by means of a Jobin-Yvon U1000Raman spectrometer equipped with a mi-croscope ?100?objective ?and with an Ar-Kr laser,using the excitation wavelength ?=514.5nm.The incident laser power focused on the sample was adjusted to be less than 5mW to avoid any local heating effect.Various N -layer graphene ?akes were analyzed,displaying the standard G and 2D bands reported in literature.The D band at ?1350cm ?1,related to lattice defects,was never observed in the experiments,con?rming the good quality of our graphene samples.Moreover,the analysis of the monolayer spectra ?G peak at ?1582cm ?1with full width at half maxi-mum ?FWHM ??13cm ?1?indicates that the unintentional doping level in the pristine samples was relatively low ?about 1?1012cm ?2?.18Then,the samples were subjected to a radio-frequency ?rf ?plasma treatment in CHF 3gas and im-mediately ?within a few minutes ?recharacterized by Raman spectroscopy.
III.RESULTS AND DISCUSSION
A.CHF 3plasma
Previous studies reported in literature 19,20have shown that the following radical species can be found in a CHF 3plasma:F atoms and CF x ?x =1,2,3?radicals.In dry etching pro-
cesses,F atoms normally act as the reactive species ?respon-sible for the etching ?,whereas CF x radicals are passivation precursors giving rise to polymer deposition.It has been ob-served that,in the case of CHF 3plasma processes,the etch rate decreases with increasing ?ow ?in contrast with what happens in the case of CF 4plasma ?,due to the low-concentration ratio ?F ?/?CF x ?.21This means that the action of F atoms becomes more effective with decreasing the ?ow rate,because the passivating action of CF x radicals is re-duced.In our experiments,we have observed an increasing modi?cation of the graphene Raman spectra,with decreasing the ?ow rate in the plasma treatment.
In particular,we performed a preliminary study on mono-layer graphene processed at various CHF 3?ow rates.Figure 2shows that the plasma treatment induces a blueshift of both G and 2D peaks,which increases with decreasing the gas ?ow.Moreover,below a ?ow-rate threshold ?about 6SCCM ?,two new peaks arise at about 1350cm ?1?D peak ?and 1620cm ?1?D ?peak ?,which indicate the presence of defects in the sp 2C lattice.Also the effect on the Raman spectra of N -layer graphenes,which is going to be discussed in detail in the following section,was remarkably reduced with increasing the gas ?ow rate.
In analogy with the mechanism involved in the etching processes,the interaction of F atoms with the graphene sur-face is likely reduced at high ?ow rate,because of the pas-sivating action of the CF x radicals.Therefore,the effect of CHF 3plasma on graphene Raman spectra can be ascribed to the adsorption of F atoms on the surface.With increasing the F coverage,the modi?cation of the graphene properties be-comes more important,passing from a p -type doping effect ?blueshifted G and 2D peaks ??Ref.15?to a structural modi-?cation ?D and D ?peaks ?at a very low ?ow rate ?less than 6SCCM ?.The doping effect can be ascribed to an electron transfer from graphene to adsorbed F atoms,i.e.,a mecha-nism analogous to the observed charge transfer between graphene and adsorbed K atoms.22At very low gas ?ow,when the passivating action of the CF x radicals is minimized,chemical modi?cation ??uorination ?of graphene may even-tually occur,with a transition from sp 2to sp 3C hybridization similar to that observed in graphane formation by plasma hydrogenation.23This evolution may be analogous to the transition from semi-ionic to covalent C-F bonding observed in carbon nanotubes treated in CF 4plasma.
24
FIG.1.?Color online ?Optical microscope image of one of the ?akes analyzed in this work ?top ?.The number of layers estimated by contrast analysis is indicated.In the bottom graph,the discrete behavior of the contrast as a function of the number of layers is shown.Experimental data,obtained at ?=550nm,are compared to theoretical
values.
FIG.2.Raman spectra of monolayer graphene after plasma treatments at various CHF 3?ow rates
MATTEO BRUNA AND STEFANO BORINI PHYSICAL REVIEW B 81,125421?2010?
Here,we discuss the results of processes carried out at a gas ?ow of 6SCCM ?pressure of 100mTorr ?for 5min,at rf power=15W.Such experimental conditions lead to a very high doping without structural modi?cation of the graphenes.Moreover,the symmetry of stacked graphenes is broken by the dipole moment generated by the charge transfer from graphene to the adsorbed F atoms.Therefore,the situation under study is very similar to that found in ?eld-effect ex-periments,and the results here reported may be useful for the study of gated graphene-based devices.
B.G band splitting and electron-phonon coupling
in N -layer graphene
We focus now on the effect of top doping on the Raman G band,which displays very distinct features depending on the number N of stacked graphene layers.The change in the G band induced by plasma treatment in the various cases is visible in Fig.3,where the spectra at t 0and t 1were acquired on the same substrate before and immediately after the treat-ment,respectively.The G peak of monolayer graphene is largely blueshifted ?to ?1590cm ?1?and narrowed ?FWHM ?6cm ?1?.Both the observations are consistent with an increase in the doping level,which induces a hard-ening of the mode,due to the nonadiabatic removal of a Kohn anomaly for zone-center optical phonons,25and a re-duction in the linewidth,due to Pauli exclusion principle which inhibits phonon decay into electron-hole pairs when
the Fermi level surpasses half the phonon energy.26On the other hand,the bilayer and 4-layer spectra display a very evident splitting of the G mode,whereas the behavior of trilayer spectrum is similar to that of monolayer.These re-sults have been con?rmed on different ?akes on the same sample and on different samples.
Furthermore,we observed that the modi?cation induced by the plasma treatment was not stable under ambient con-ditions,as the Raman spectra changed with the passing of time,slowly tending to their pristine form ?Fig.3?.Indeed,the initial condition can be restored by a vacuum annealing,so that the plasma treatment can be repeated for several times in a reproducible way.Such a reversible behavior is consis-tent with the absence of the D peak,which indicates the lack of structural modi?cation of the material ?in contrast with the case of the chemical modi?cation obtained by plasma treat-ment at very low gas ?ow ?.As previously discussed,the monolayer spectra,showing the hardening of both the G mode and the 2D mode,tell us that a strong p -type doping is achieved 15upon plasma processing.Therefore,we were able to gradually vary the doping level on top of each N -layer graphene,studying the effect for different values of N .
The effect of charged adsorbates has already been studied on epitaxial bilayer graphene on SiC by angle resolved photoemission spectroscopy measurements.22It was shown that the electronic bands of bilayer graphene are strongly affected by a potassium atoms coverage on the top surface due to the n -type doping induced by the adsorbates.
Indeed,
FIG.3.?Color online ?Evolution of the Raman G band of N -layer graphene after CHF 3plasma treatment.The spectra at t 0were acquired before the treatment while those at t i ?1?i ?6?were taken at various time intervals after the treatment ?t 1?15min;t 2?6h;t 3?24h;t 4?48h;t 5?72h;and t 6?144h ?.The lines connecting peaks are guides for the eye.
OBSERV ATION OF RAMAN G -BAND SPLITTING IN …PHYSICAL REVIEW B 81,125421?2010?
because of the short screening length along the c axis of stacked graphenes,27the majority of the doping charge is localized in the top layer and a dipole moment is formed, giving rise to a symmetry breaking and to the consequent band-gap opening.28In our case,an analogous effect is likely to occur:the doping charge density rapidly decreases going from the top layer to the bottom,thus altering the symmetry of the multilayers.The symmetry breaking can affect very much the physical properties of the system,as it eliminates the symmetry constraints which rule both phonon Raman activity and EPC.In these conditions,all phonon modes in-cluded in the G band can become Raman active and strongly mixed with each other,29and the EPC is affected by the modi?cation of the electronic bands.
In Fig.3,the G band dependence on the doping is visible in all cases but it looks different for different N values.An estimation of the doping level may be obtained from mono-layer graphene spectra,basing on the G peak position and linewidth,and on the intensity ratio of the G and2D peaks,15 but we have veri?ed that different monolayers can display slightly different G peak positions on the same sample,due to the dif?culty of controlling the native doping of graphene in ambient atmosphere.30Furthermore,monolayer graphene is likely to have a different reactivity with respect to that of few-layer graphene,as experimentally observed in the case of hydrogenation.23,31Also in our experiments,we have seen that,after plasma treatment at very low?ow rate,the D peak is hardly observed in the few-layers spectra,whereas the monolayer spectrum displays a very evident D peak.There-fore,it seems not correct to extrapolate the doping values from the monolayer analysis to interpret the few-layer spec-tra.However,although a precise quantitative estimate of car-rier concentration in each case is not possible in our experi-ments,we can monitor the different behavior of N-layer graphene Raman spectra with decreasing the top doping, starting from high doping levels?more than1?1013cm?2?as suggested by the analysis of monolayers spectra.
The effect in the bilayer can be interpreted on the basis of some recent literature.Indeed,the splitting of the Raman G band was recently observed in gated bilayer graphene,10and ascribed to the inversion symmetry breaking induced by the gate?eld effect and to two distinct EPC involved in the G band.This is due to the fact that the G band of bilayer graphene includes two doubly degenerate modes,E2g and E u, which are symmetric and antisymmetric with respect to the inversion symmetry,respectively.Consequently,only the E2g
mode is normally Raman active,unless inversion symmetry breaking switches the E u mode on,too.It has been shown by Ando and Koshino29that,in the presence of asymmetry in the potential of the two stacked layers,symmetric and anti-symmetric modes are strongly mixed with each other and two peaks appear in the Raman spectrum.Moreover, phonons can be considerably modi?ed by resonant electronic interband transitions,when the asymmetry opens up a gap comparable to the phonon energy.Ab initio calculations have been performed to compute the G band of bilayer graphene in asymmetric conditions as a function of the carrier concen-tration in the top and bottom layer?n top and n bot?,32predict-ing the behavior of the two modes which become Raman visible at certain values of n top and n bot.
In Fig.4,we report the analysis of the two Lorentzian peaks which can?t the bilayer G band at various stages of the plasma modi?cation,for two different samples.The ex-perimental evolution of the peaks position has been?tted by the theoretical curves obtained from Ref.32,assuming that about the85%of the total charge carriers is con?ned in the top layer,according to Ref.27,and using the total carrier concentration n as the free parameter in the?tting.Then,the other spectral features?intensity and linewidth?have been compared to the behavior predicted in Ref.32,considering the carrier-concentration values obtained from the best?t of the peaks position curve.It can be seen that a good qualita-tive agreement is obtained for all the analyzed parameters.In particular,the threshold at about n=1?1013cm?2,
after FIG.4.Features of the two peaks?tting the bilayer G band as a function of doping.Circles and triangles are experimental data from two different samples,and lines are the theoretical?ndings of Ref.
32.In?a?,the peaks positions have been?tted by the theoretical curves,assuming that about the85%of the total charge carriers is con?ned in the top layer.In?b?and?c?,the width and intensity ratios of the low?L?frequency and high?H?frequency peak are reported.
MATTEO BRUNA AND STEFANO BORINI PHYSICAL REVIEW B81,125421?2010?
which a steep variation in both the intensity and linewidth
ratios is predicted by theory,is well reproduced by the ex-perimental data,thus con?rming the consistence of the carrier-concentration results obtained by ?tting the peaks po-sition curves.Therefore,the behavior of the bilayer G band is well interpreted in the framework of an asymmetric carrier distribution model.
In order to interpret the spectra for N ?2,it is worth to consider,?rst of all,the evolution of the G band with N as predicted by group theory.Indeed,applying the group theory to Bernal stacked graphene layers,the irreducible represen-tations of the infrared ?IR ?and Raman-active modes at the ?point for N -layer graphene can be obtained,33,34as listed in Table I .The modes related to the G band are the E g ,E u ,E ?,and E ?in-plane modes,while the A modes are related to out-of-plane phonons.
E g and E u modes are found in case of inversion symmetry of the system ?N even ?,whereas E ?and E ?modes appear for mirror symmetry ?N odd ?.Importantly,only the inversion symmetry inhibits the Raman activity of antisymmetric modes.It can be obtained by ab initio calculations 35that in trilayer graphene two E ?modes and one E ?mode can be found at the G -band frequencies,whereas in 4-layer graphene two E g modes and two E u modes vibrate at the G -band frequencies.The inversion symmetry breaking,pro-duced by the top doping,switches on the previously Raman silent antisymmetric modes E u in N even-layer graphene,so that all the phonon modes become Raman active and mixed in N -layer graphene for every N value.Moreover,like in the bilayer case previously discussed,the EPC can be strongly affected by the lack of symmetry constraints and by the change in the electronic band structures.
The absence of splitting for trilayer graphene can be qualitatively interpreted considering the allowed electronic interband transitions which give rise to the phonon energy renormalization ?Kohn anomaly ?when E F ?0?low doping ?.In Fig.5it is shown that both E ?and E ?phonons can couple with electronic transitions when the Fermi level is at the Dirac point,whereas in bilayer graphene only the symmetric E g mode can ef?ciently create electron-hole pairs,due to energy conservation and symmetry selection rules.Indeed,?rst-principles calculations 36have shown that in trilayer graphene the phonon linewidths of symmetric and antisym-metric modes are all of the same order of magnitude,whereas in the bilayer the antisymmetric linewidth is two orders of magnitude smaller than that of the symmetric mode.Therefore,almost the same EPC strength is expected for the three phonon modes in trilayer graphene,giving rise to the phonon-energy renormalization at E F =0and to the consequent hardening of all phonons with moving the Fermi level,such as in monolayer graphene.This is a qualitative
different case with respect to that of bilayer,where only the symmetric phonon energy is renormalized at E F ?0.The presence or absence of the G -band splitting in the two cases re?ects the presence or absence of distinct EPC for phonons of distinct symmetry.
The observation of splitting in the 4-layer spectrum sug-gests the presence of distinct EPC for the different phonon modes in analogy with the bilayer behavior.In Fig.6we show the experimental position of the two Lorentzian peaks ?tting the 4-layer G band,as a function of doping as ob-tained from the bilayer analysis.It is worth noting that a minimum can be clearly identi?ed in the curve of the low-frequency peak,indicating that a maximum is likely to occur in the EPC for some phonon modes at a given doping value ?n m ?9?1012cm ?2?.Such a feature can be interpreted by taking into account the band structure of the 4layer,which can be approximated by two bilayer-type band structures.8In particular,intraband transitions occurring between two sub-bands separated by ?E ?0.24eV,which corresponds to an experimentally observed absorption by IR spectroscopy,37may give rise to an almost resonant coupling with phonons ?E ph ?0.2eV ?,and to a consequently strong renormalization of phonon energy.This EPC is expected to have a maximum
TABLE I.Irreducible representations of the IR and Raman-active modes at the ?point for N -layer graphene.
N ?IR
?Raman
Even ?N ?1?A 2u ?N ?1?E u
NA 1g NE g
Odd
NA 2? NE ?
?N ?1?A 1? NE ? ?N ?1?E ?
FIG.5.?Color online ?Electronic transitions allowed by symme-try rules in bilayer ?left ?and trilayer ?right ?graphene when ?F =0.The
electronic bands are taken from Ref.8.The transitions indi-cated by dotted lines are not involved in the phonon-energy renor-malization,because their energy is much higher than the G -band phonon energy ??0.196eV ?,which is reported as a scale bar on the left.
FIG. 6.?Left ?Electronic intraband transitions in 4-layer graphene for ?F =0.24eV.The electronic bands are taken from Ref.8and the G -band phonon energy ??0.196eV ?is reported as a scale bar.?Right ?Positions of the two peaks ?tting the splitted G band of 4-layer graphene as a function of doping.The observed dip may be related to a strong EPC due to the intraband transitions shown on the left.
OBSERV ATION OF RAMAN G -BAND SPLITTING IN …PHYSICAL REVIEW B 81,125421?2010?
when the Fermi level reaches the high-energy subband?see Fig.6?,when the number of possible intraband transitions is maximized.Therefore,the doping value n m,at which a mini-mum for the Raman peak position is observed,is likely to correspond to the Fermi energy touching the high-energy subband at E F?0.24eV.A theoretical analysis of the 4-layer band structure and density of states in the presence of asymmetric doping may con?rm this hypothesis.
IV.CONCLUSIONS
In summary,we have experimentally investigated the Ra-man G band for N-layer graphene?1?N?4?in the presence of high asymmetric doping,?nding two different types of behavior.For N odd,the G band is always?tted by a single Lorentzian peak,which is blueshifted with increasing the doping level.This is due to a strong EPC for all phonon modes when E F?0,which decreases with increasing the Fermi energy.On the other hand,for N even,an evident splitting of the G band is observed,related to the presence of distinct EPC for phonons of distinct symmetry.In particular, in the4-layer case a signature of the van Hove singularity at E?0.24eV is likely to be observed as a minimum of the low-energy peak position.
Insights into the electron-phonon interactions in N-layer graphenes in the presence of top doping can be useful for the study of?eld-effect graphene-based devices.Moreover,the CHF3plasma treatment may be a powerful technique for the study of graphene in the presence of a coverage of highly electronegative atoms such as?uorine.
Finally,the variety of the results reported in the literature about the Raman G band in heavily doped bilayer graphene10,11,38,39suggests that the repartition of the addi-tional charge carriers is not well understood in most experi-ments.Therefore,suspended graphene samples may be a good test bed to further investigate the distribution of the doping,especially in the bilayer case.40,41
ACKNOWLEDGMENTS
This work was carried out within the EURAMET Joint Research Project“ULQHE.”The research within this EURAMET JRP receives funding from the EC FP7,ERA-NET Plus,under Grant Agreement No.217257.
*s.borini@inrim.it
1K.S.Novoselov,D.Jiang,F.Schedin,T.J.Booth,V.V.Khot-kevich,S.V.Morozov,and A.K.Geim,Proc.Natl.Acad.Sci.
U.S.A.102,10451?2005?.
2A.H.Castro Neto,F.Guinea,N.M.R.Peres,K.S.Novoselov, and A.K.Geim,Rev.Mod.Phys.81,109?2009?.
3A.K.Geim and K.S.Novoselov,Nature Mater.6,183?2007?. 4Y.Zhang,T.T.Tang,C.Girit,Z.Hao,M.C.Martin,A.Zettl, M.F.Crommie,Y.R.Shen,and F.Wang,Nature?London?459, 820?2009?.
5K.F.Mak,C.H.Lui,J.Shan,and T.F.Heinz,Phys.Rev.Lett.
102,256405?2009?.
6A.B.Kuzmenko,L.Benfatto,E.Cappelluti,I.Crassee,D.van der Marel,P.Blake,K.S.Novoselov,and A.K.Geim,Phys.
Rev.Lett.103,116804?2009?.
7T.T.Tang,Y.Zhang,C.H.Park,B.Geng,C.Girit,Z.Hao, M.C.Martin,A.Zettl,M.F.Crommie,S.G.Louie,Y.R.Shen, and F.Wang,Nat.Nanotechnol.5,32?2010?.
8M.Koshino and T.Ando,Solid State Commun.149,1123?2009?.
9M.F.Craciun,S.Russo,M.Yamamoto,J.B.Oostinga,A.F.
Morpurgo,and S.Tarucha,Nat.Nanotechnol.4,383?2009?. 10L.M.Malard,D.C.Elias,E.S.Alves,and M.A.Pimenta,Phys.
Rev.Lett.101,257401?2008?.
11J.Yan,T.Villarson,E.A.Henriksen,P.Kim,and A.Pinczuk, Phys.Rev.B80,241417?R??2009?.
12A.C.Ferrari,Solid State Commun.143,47?2007?.
13L.M.Malard,M. A.Pimenta,G.Dresselhaus,and M.S.
Dresselhaus,Phys.Rep.473,51?2009?.
14A.C.Ferrari,J.C.Meyer,V.Scardaci,C.Casiraghi,https://www.wendangku.net/doc/b511757743.html,zzeri,
F.Mauri,S.Piscanec,D.Jiang,K.S.Novoselov,S.Roth,and
A.K.Geim,Phys.Rev.Lett.97,187401?2006?.
15A.Das,S.Pisana,B.Chakraborty,S.Piscanec,S.K.Saha,U.V.
Waghmare,K.S.Novoselov,H.N.Krishnamurthy,A.K.Geim, A.C.Ferrari,and A.K.Soo,Nat.Nanotechnol.3,210?2008?. 16R.R.Nair,P.Blake,A.N.Grigorenko,K.S.Novoselov,T.J. Booth,T.Stauber,N.M.R.Peres,and A.K.Geim,Science 320,1308?2008?.
17M.Bruna and S.Borini,Appl.Phys.Lett.94,031901?2009?. 18S.Pisana,https://www.wendangku.net/doc/b511757743.html,zzeri,C.Casiraghi,K.S.Novoselov,A.K. Geim,A.C.Ferrari,and F.Mauri,Nature Mater.6,198?2007?. 19K.Tachibana,H.Kamisugi,and T.Kawasaki,Jpn.J.Appl. Phys.,Part138,4367?1999?.
20G.Hancock and J.P.Sucksmith,J.Vac.Sci.Technol.A20,270?2002?.
21D.-Y.Choi,S.Madden,A.Rode,R.Wang,and B.Luther-Davies,J.Appl.Phys.104,113305?2008?.
22T.Ohta,A.Bostwick,T.Seyller,K.Horn,and E.Rotenberg, Science313,951?2006?.
23D.C.Elias,R.R.Nair,T.M.G.Mohiuddin,S.V.Morozov, P.Blake,M.P.Halsall,A.C.Ferrari,D.W.Boukhvalov,M.I. Katsnelson,A.K.Geim,and K.S.Novoselov,Science323,610?2009?.
24N.O.V.Plank,L.Jiang,and R.Cheung,Appl.Phys.Lett.83, 2426?2003?.
https://www.wendangku.net/doc/b511757743.html,zzeri and F.Mauri,Phys.Rev.Lett.97,266407?2006?. 26J.Yan,Y.Zhang,P.Kim,and A.Pinczuk,Phys.Rev.Lett.98, 166802?2007?.
27T.Ohta,A.Bostwick,J.L.McChesney,T.Seyller,K.Horn,and E.Rotenberg,Phys.Rev.Lett.98,206802?2007?.
28E.McCann,Phys.Rev.B74,161403?R??2006?.
29T.Ando and M.Koshino,J.Phys.Soc.Jpn.78,034709?2009?. 30C.Casiraghi,S.Pisana,K.S.Novoselov,A.K.Geim,and A.C. Ferrari,Appl.Phys.Lett.91,233108?2007?.
31S.Ryu,M.Y.Han,J.Maultzsch,T.F.Heinz,P.Kim,M.L. Steigerwald,and L.E.Brus,Nano Lett.8,4597?2008?.
MATTEO BRUNA AND STEFANO BORINI PHYSICAL REVIEW B81,125421?2010?
32P.Gava,https://www.wendangku.net/doc/b511757743.html,zzeri,A.M.Saitta,and F.Mauri,Phys.Rev.B 80,155422?2009?.
33J.-W.Jiang,H.Tang,B.-S.Wang,and Z.-B.Su,Phys.Rev.B 77,235421?2008?.
34L.M.Malard,M.H.D.Guimaraes,D.L.Mafra,M.S.C. Mazzoni,and A.Jorio,Phys.Rev.B79,125426?2009?.
35S.K.Saha,U.V.Waghmare,H.R.Krishnamurthy,and A.K. Sood,Phys.Rev.B78,165421?2008?.
36J.-A.Yan,W.Y.Ruan,and M.Y.Chou,Phys.Rev.B79,115443?2009?.37K.Mak,M.Sfeir,J.Misewich,and T.Heinz,arXiv:0908.0154?unpublished?.
38J.Yan,E.A.Henriksen,P.Kim,and A.Pinczuk,Phys.Rev.Lett.
101,136804?2008?.
39A.Das,B.Chakraborty,S.Piscanec,S.Pisana,A.K.Sood,and A.C.Ferrari,Phys.Rev.B79,155417?2009?.
40S.Berciaud,S.Ryu,L.E.Brus,and T.F.Heinz,Nano Lett.9, 346?2009?.
41Z.H.Ni,T.Yu,Z.Q.Luo,Y.Y.Wang,L.Liu,C.P.Wong, J.Miao,W.Huang,and Z.X.Shen,ACS Nano3,569?2009?.
OBSERV ATION OF RAMAN G-BAND SPLITTING IN…PHYSICAL REVIEW B81,125421?2010?