NP JP Graphene Plasmonics

8/20/2019 NP JP Graphene Plasmonics

1/10

NATURE PHOTONICS | VOL 6 | NOVEMBER 2012 | www.nature.com/naturephotonics 749

Graphene, a two-dimensional (2D) orm o carbon in whichthe atoms are arranged in a honeycomb lattice 1,2, hasalready been shown to possess unique mechanical, elec-

tric, magnetic and thermal properties with a multitude o excit-ing applications that are being vigorously pursued by academiaand industry 3. Interestingly, it is in optics where graphene hasshown its ‘true colours’ and where the first commercial applica-

tion o graphene has been realized4. Graphene has an extremelyhigh quantum efficiency or light–matter interactions, is stronglyoptically nonlinear and contains plasmons with unusual prop-erties. Furthermore, it can be modified by gating, by doping, bychemical means and through conventional plasmonics based onnoble metals. Te optics and photonics o graphene have beenreviewed several times previously 4–8. However, the recent proli-eration o works on graphene plasmonics and the optics o 2Dsingle-layer materials calls or a resh look at graphene and itsremarkable properties.

Optical properties of grapheneAmong many o the unique properties o graphene3, probably themost peculiar is the act that quasiparticles in this material obey

a linear dispersion relation. As a result, an additional—chiral—symmetry exists or the quasiparticles, which fixes the direction opseudospin (in a given valley - a simply connected part o a Fermisurace) to be parallel or antiparallel to the directions o motiono electrons and holes, respectively 9. Tis has an immediate, anddominant, effect on the electronic and optical properties o this2D crystal. One consequence is that the optical conductivity isindependent o any material parameters: σuni = πe2 / (2h), where e is electron charge and h is Planck’s constant10,11. Tus, the opticalabsorption depends only on the fine-structure constant, πα ≈ 2.3%(Fig. 1a). Such simple behaviour is expected or undoped sam-ples at zero temperatures. Doping has a very strong effect on theoptical properties12: Pauli blocking (Fig. 1b) ensures that photonswith energy less than 2EF (where EF is the Fermi energy) are not

absorbed13

. Recently, gating with a solid electrolyte allowed carrierconcentrations as large as 1014 cm−2 to be achieved, which convertsinto EF ≈ 1 eV, such that a modulation o optical transmission inthe visible spectrum is possible12,14,15. Combining graphene withsilicon waveguides, or example, makes it possible to produce abroadband, graphene-based, waveguide-integrated optical modu-lator15 (Fig. 1c).

For photon energies greater than 3 eV, trigonal warping effectsand deviations rom linear dispersion become extremely strong.Te band structure o graphene has saddle points at the M points othe Brillouin zone, which lead to van Hove-like singularities. In the

Graphene plasmonicsA. N. Grigorenko1*, M. Polini2 and K. S. Novoselov 1

Two rich and vibrant fields of investigation—graphene physics and plasmonics—strongly overlap. Not only does graphene pos-

sess intrinsic plasmons that are tunable and adjustable, but a combination of graphene with noble-metal nanostructures prom-ises a variety of exciting applications for conventional plasmonics. The versatility of graphene means that graphene-basedplasmonics may enable the manufacture of novel optical devices working in different frequency ranges—from terahertz to thevisible—with extremely high speed, low driving voltage, low power consumption and compact sizes. Here we review the fieldemerging at the intersection of graphene physics and plasmonics.

single-electron approximation, this should result in a strong absorp-tion peak (o more than 10%) at around 5.2 eV. In practice, this sim-ple picture ails owing to many-body effects, and the peak is observedclose to 4.5 eV (res 16, 17; Fig. 1a).

Optical properties of other 2D materialsIt is ofen overlooked that the realization that graphene has a num-ber o nontrivial, interesting and possibly useul properties openeda floodgate, resulting in the discovery and study o other 2D crystals

1School of Physics and Astronomy, University of Manchester, Manchester M13 9PL, UK. 2NEST, Istituto Nanoscienze-CNR and Scuola Normale Superiore,

I-56126 Pisa, Italy. *e-mail: [email protected]

Figure 1 | Building blocks of 'flatland' optics. a, Left: the optical

conductivities of pristine graphene, with (blue) and without (purple)

electron–electron interactions taken into account; pristine bilayer graphene

(red); doped graphene (pink); doped bilayer graphene (dark green); and

fluorographene (black). The inset shows an enlarged view of the low-energy

spectral range. Right: absorption spectra of monolayer and bilayer MoS2

(left axis, normalized by the number of layers; black) and the corresponding

photoluminescence spectra (right axis, normalized by the intensity of

the peak absorption; red). The spectra are displaced in the vertical axis

for clarity. The green line shows the spectral position of the excitation

wavelength. a.u., arbitrary units. b, Schematic band structure of graphene

for various levels of doping (only one valley is shown) and Pauli blocking of

photon absorption in graphene. c, Sketch of a graphene-based, waveguide-

integrated optical modulator, where voltage applied to graphene is used to

achieve broadband and high speed modulation of the guided light (adapted

from ref. 15). Data in a are partly taken from ref. 24.

b c

a

00

0

0.1

0.2

0

0.1

0.2

1.4 1.6 1.8 2.0 2.21 2 3

Energy (eV) Photon energy (eV)

Energy (eV)

4 5 6

1

2

3

4

5

6

3

2

1

01.00.5

O p t i c a l

c o n d u c t i v i t y ( π e / 2 h )

Opticalconductivity

(πe2 /2h)

Bilayer graphene

Monolayer grapheneFluorographene

A b

s o r b a n c e p e r l a y e r

N o r m a l i z e d p h o t o l u m i n e s c e n c e ( a . u . )

FOCUS | REVIEW ARTICLESPUBLISHED ONLINE: 31 OCTOBER 2012 | DOI: 10.1038/NPHOTON.2012.262


2/10

750 NATURE PHOTONICS | VOL 6 | NOVEMBER 2012 | www.nature.com/naturephotonics

and heterostructures based on them. Over the last ew years, manyother 2D materials have been intensively investigated. Teir opti-cal properties are exciting and ofen different rom the properties otheir three-dimensional parent materials.

Bi- and trilayer graphene: materials with controllable bandgapsand controllable optical absorption. Te electronic structure obilayer graphene is drastically different rom that o the monolayer.Te band structure o graphene with A–B (Bernal) stacking is gap-less. Te valence and conduction bands have parabolic dispersionand touch at zero energy. It also contains additional subbands that,in the first approximation, are offset rom zero energy by γ1 ≈ 0.4 eV

(the nearest-neighbour hopping between layers). In optics, thisleads to a strong absorption peak at this energy (Fig. 1a).Te gapless spectra in mono- and bilayer graphene are protected

by the symmetry between the sublattices. However, in bilayer gra-phene such symmetry can be easily lifed by selective chemicaldoping o one o the layers or by applying a transverse electric field(gating). Tis leads to the opening o a significant gap, which canbe seen in optical absorption. Figure 2a shows the energy spectrumo bilayer graphene, and Fig. 2b demonstrates an effective opticalmodulator based on the tunable bandgap in bilayer graphene18.

rilayer graphene is another interesting material, which comes intwo very different configurations. Te electronic structure o Bernal-stacked (A–B–A) trilayer graphene can be viewed as a combinationo those o one monolayer and one bilayer. However, trilayers withrhombohedral (A–B–C) stacking more closely resemble bilayer

graphene. As a result, the optical properties o both are stronglydependent on the perpendicular electric field, with the A–B–Ctrilayer demonstrating a larger bandgap than the A–B–A orm.

Graphane and fluorographene. Graphene can also be considered agiant aromatic molecule that can undergo chemical reactions. wosuch chemically modified orms have already been made: graphane19 (in which a hydrogen atom is attached to each o the carbon atoms)and fluorographene20,21 (or 2D eflon, in which one fluorine atomis attached to each carbon atom). Covalently bonded hydrogen orfluorine changes the hybridization o carbon atoms rom sp2 to sp3,which removes the π orbitals rom the electronic band structure.

Tis should lead to the opening o a large gap o the order o theseparation between the σ bands. Experimentally, optical gaps o theorder o 3 eV have been observed20 or fluorographene (Fig. 1a).

2D atomic crystals and their heterostructures. Monolayer molyb-denum disulphide (MoS2) is probably the second most-studied 2Dmaterial afer graphene. It has been exoliated to the monolayer stateby both mechanical22 and liquid-phase23 exoliation. Te propertieso monolayer MoS2 are radically different rom the properties o the3D parent material. Bulk crystals have an indirect bandgap o theorder o 1.29 eV. In the 2D state, however, MoS 2 has a direct band-gap24 o the order o 1.9 eV around the K points o the Brillouinzone, which leads to a strong increase in luminescence25 (Fig. 1a).

Furthermore, owing to the absence o inversion symmetry, strongspin–orbit interactions split the valence band states at the K and Kʹ

a b

c d

Source

Drain

Top gate

20 µm

Pt (top gate)

SiO2

Si (bottom gate)

AuAl2O3

Figure 2 | Multilayered flatland optics. a, Energy spectra for pristine (left) and doped bilayer (right) graphene , with a gap opening for the latter case. b,

Optical modulator based on the tunable bandgap in bilayer graphene. c, Hypothetical multilayer structure with vertical charge separation. In this particular

structure two graphene layers are separated by an optically active barrier. An absorbed photon is converted into an electron-hole pair, which can be split

in the external electric field. d, Cavity-based graphene photodetector and graphene-based integrated interferometer. By allowing multiple reflections the

interaction between light and graphene is strongly enhanced. Images reproduced with permission from ref. 18 (b) and ref. 32 (d, adapted).

REVIEW ARTICLES | FOCUS NATURE PHOTONICS

DOI: 10.1038/NPHOTON.2012.262


3/10


points by about 0.16 eV. In addition, time reversal symmetry ensuresthat this splitting is in the opposite direction in the two valleys, lead-ing to opposite spin polarizations in the K and Kʹ valleys. Effectively,this allows optical control (by means o circularly polarized light) othe population o charge carriers separately in each o the two val-leys, leading to the realization o ‘valleytronics’26.

Te electronic and optical properties o other 2D crystals (orexample BN (re. 27), aS2, NbSe2 and WS2) are currently being

intensively investigated. It is likely that there will be more surprisesrom heterostructures created by stacking such 2D crystals one ontop o another28. Te simplest possible stacks have already revealednew physics29–31, and could revolutionize some applications. Teoptical properties o these heterostructures can be tuned very accu-rately, as materials with very different bandgaps and thicknesses canbe combined (Fig. 2c). Another direction is to combine grapheneheterostructures with standard optics, or example optical cavi-ties32–34, and optoelectronic devices, or example integrated interer-ometers5 (Fig. 2d).

Intrinsic graphene plasmonsPlasmons are ubiquitous, high-requency, collective density oscilla-tions o an electron liquid and occur in many metals and semiconduc-

tors35. Te intrinsic graphene plasmons are rereshingly different romplasmons in noble metals as they can be tuned by gating or doping.Graphene plasmonic resonances could have a pivotal role in the reali-zation o robust and cheap photodetectors o terahertz radiation 36,with important security applications. Here we consider the peculiarproperties o the intrinsic plasmon modes o the electron gas in a pris-tine graphene sheet. Our ocus is on the plasmons o doped samples,although interesting collective modes have also been predicted orundoped graphene37. Moreover, we will ocus on longitudinal modes,that is, modes whose associated electric field is parallel to the wave

vector, q. Such modes are also known as transverse magnetic modes.Te existence o a transverse collective mode in graphene (with a re-quency slightly lower than the Pauli-blocking threshold or interbandabsorption) has been also discussed38. Tis mode is also reerred to as

transverse electric modes. Te transverse electric mode in graphenehas a requency, ω, that lies in the window 1.667 < ħω/EF < 2 (whereħ denotes Planck’s constant divided by 2π) and can be tuned romthe radio-requency range to the inrared by changing the density ocharge carriers using a gate voltage38. On the other hand, the trans-

verse magnetic (longitudinal plasmon) mode is gapless in the long-wavelength limit, because its energy vanishes as q =|q| 0 (see below).

Electron plasma in two dimensions. wo-dimensional electronsystems (2DESs) have been a rich source o exciting physics ormore than our decades. As in any 2DES, electrons in graphene donot move as independent particles. Rather, their motions are highlycorrelated as a result o pairwise interactions. Tese are describedby a potential u(r ij) = u(|ri − r j|), which depends only on the absolute

value o the relative distance, rij = ri − r j, between two electrons. Teinteraction potential is sensitive to the dielectric media surroundingthe graphene sheet. For graphene with one side exposed to a mediumwith dielectric constant ε1 and the other exposed to one with dielec-tric constant ε2, we have u(r ij) = e2/εr ij, where ε = (ε1 + ε2)/2 (suchthat the 2D Fourier transorm is uq = 2πe2/εq). Te electron gas ina graphene sheet can be described at low energies by the ollowingcontinuum-model Hamiltonian:

(1)ε r

i – r

j

1 e22 ´

Ĥ = νF

σ pi+Σ

iΣi≠j

Here v F ≈ 106 m s−1 is the Fermi velocity, pi = –iħ∇ri is the canoni-cal momentum o the ith electron and σ = (σ

x

, σ y

) is a 2D vector o

the Pauli matrices. For the sake o simplicity, equation (1) has beenwritten or a single-channel, massless Dirac ermion (MDF) model;that is, it holds or electrons with given spin and valley indices. Terelative importance o electron–electron interactions is quantifiedby the ratio between the ‘magnitude’ o the second term and thato the first term. For a doped graphene sheet, the typical distancebetween electrons is o the order o the inverse o the Fermi wave-number, that is, kF−1. Te second term is thus o the order o e2kF/ε.

Te kinetic energy is o the order o ħv FkF and, hence, the ratiobetween these two quantities defines a dimensionless parameter, αee,usually called the graphene fine-structure constant:

εħν

e2α

ee =

Tis can be expressed using the conventional fine-structure constant,α = e2/ħc, as αee = cα/εv F, which allows us quickly to estimate the mag-nitude o αee. For example, or graphene with one side exposed to air(ε1 = 1) and the other to SiO2 (ε2 ≈ 3.9), we have αee ≈ 0.9. For a sus-pended graphene sheet, ε1 = ε2 = 1 and αee ≈ 2.2. Hence, the graphenefine-structure constant can be tuned experimentally by changing thedielectric environment surrounding graphene39. From this analysis,

we conclude that, at least in principle, electrons in a doped graphenesheet interact quite strongly with each other and that interactioneffects in this material have to be analysed with great care.

Teory of 2D plasmons. Te physical origin o plasmons can beunderstood as ollows. When electrons move to screen an elec-tric field, they tend to travel slightly too ar (Fig. 3a). Tey arethen pulled back toward the charge disturbance and overshootagain, setting up a weakly damped oscillation. Te restoring orceresponsible or the oscillation is proportional to the gradient othe sel-consistent field created by all the electrons. Te plasmondispersion in a 2DES can be understood in the long-wavelength,q


4/10


Te Dirac plasmon requency thus scales like EF 1/2 ∝ n1/4 and containsPlanck’s constant. When the plasmon energy becomes larger thanthe threshold or interband transitions, plasmons become stronglyLandau damped. For a typical doping, n = 1011 cm−2, the Fermi energyin graphene is EF ≈ 37 meV and the plasmon energy or grapheneon SiO2 at q = 0.1kF ≈ 0.6 × 105 cm−1 is about 16 meV, which is inthe inrared range. One remarkable conclusion o equation (3) isthat the compression o the surace plasmon wavelength relative tothe excitation wavelength is governed by the fine-structure constantand can be strong6 because λ pl /λ0 ≈ 2αEF /(εħω) ~ α.

When q ≈ kF, the hydrodynamic approach described above isnot adequate to describe the plasmon dispersion relation quanti-

tatively. Te key ingredient o a ully quantum mechanical calcu-lation o plasmon modes is the retarded density–density responseunction35, χ nn (q,ω), which, in the ramework o linear responsetheory and within the random phase approximation (RPA), isgiven by χ nnRPA (q,ω) = χ nn(0)(q,ω) / ε(q,ω), where χ nn(0) (q,ω) is the well-known40,41,44,45 non-interacting response unction and ε(q, ω) is theRPA dynamical dielectric unction. Te unction χ nn(0) (q,ω)is usuallyreerred to as the ‘Lindhard unction’. Te RPA is not exact, but it istypically a good starting point or ordinary Fermi liquids (see, how-ever, the discussion in re. 46). Higher-order terms in the dispersionrelation in equation (3) in an expansion in powers o q/kF are givenin re. 47.

So ar, we have discussed plasmons in doped graphene sheets atzero temperature and in the absence o disorder. Te effects o finitetemperature and disorder can be incorporated in the RPA theory by

using the appropriate Lindhard unction at finite temperature48 andthe Mermin approximation49, respectively. It is worth noting thatplasmon losses in graphene are expected to be considerably smallerthan those in normal metals. Te plasmon lietime, τpl, however,should not be conused with the d.c. transport scattering time, τ tr.(Tis conusion ofen arises in quantitative analyses o plasmoniceffects in graphene: in this context, the simple Drude ormula or therequency-dependent conductivity, σ(ω), which depends only on τtr,

is typically used6,49,50.) For massive 2D electrons, it has been shown51 that τpl ≈ τtr in the high-density regime but that in the low-densityregime τpl can be much smaller than τtr. In addition, plasmons coulddecay in the absence o disorder, by emitting, or example, two elec-tron–hole pairs with opposite momenta35. Tis many-body effect isnot captured by the RPA. o the best o our knowledge, all theseissues have yet to be investigated and much more theoretical andexperimental work is needed to understand and quantiy plasmonlosses in graphene and other 2D crystals.

Te effect on Dirac plasmons of screening due to a metal gate.Here we briefly discuss the effect that screening by a metal gatehas on the Dirac plasmon dispersion (equation (3)) in a dopedgraphene sheet. Neglecting effects o hybridization between

graphene and metal, we can describe a metal as a groundedconductor screening the Coulomb interactions between elec-trons in graphene. Tis ormally leads to the replacementuq → U d (q) = 2πe2(1 − e−2qd )/q, where d is the distance between thegraphene and the gate. Because U d (q) is regular at q = 0, we expecta gapless acoustic plasmon with Ω ac(q → 0) = csq rather than an‘unscreened plasmon’ (equation (3)) with ωpl(q) ∝ √

−q. An RPA

expression or cs is derived in re. 52.

Experimental observations of intrinsic plasmons in graphene.Plasmons in 2DESs can be accessed by a variety o direct and indi-rect methods, including optical measurements, electron energy-lossspectroscopy, inelastic light scattering, angle-resolved photoemis-sion spectroscopy (ARPES) and scanning tunnelling spectroscopy.

Several experiments on electron energy-loss spectroscopy havebeen perormed on exoliated graphene sheets53 and on epitaxialgraphene samples54, which showed that Dirac plasmons in grapheneon SiC are strongly hybridized with the surace optical phononso the SiC substrate55. Dirac plasmons have also been probed bydirectly engineering their coupling to inrared light in a number ointriguing ways50,56–60.

Te study o large-area arrays o graphene microribbons grownby chemical vapour deposition has shown that inrared light polar-ized perpendicular to the ribbon axis is able to excite plasmon reso-nances o the confined MDF gas56 (Fig. 3b). In these experiments anion gel was used or gating, yielding an induced carrier concentra-tion o about 1013 cm−2, which allowed the authors to access the tera-hertz spectral range. Plasmon excitations in graphene microribbon

arrays can be varied by electrical gating58

(Fig. 3b). Furthermore,the plasmon requency scales like W −1/2, where W is the width othe ribbon, and like n1/4, in agreement with the bulk RPA predictionin equation (3) (Fig. 3c). Graphene-microribbon-based plasmonicwaveguides 6 mm long allowed the transmission o 2.5 Gbps opticalsignals with an average extinction ratio o 19 dB at a wavelength o1.31 μm (re. 60). Recently, plasmon hybridization in coupled gra-phene nanoribbons has been demonstrated61 and splitting o plas-mons in graphene nanostructures into bulk and edge modes in highmagnetic fields has been observed62.

Te direct interaction o localized graphene plasmons with inra-red light has been demonstrated58 (Fig. 3d) using a stack o graphenemicrodiscs, which allowed the plasmons to be tuned through chang-ing the disk diameter, the number o discs, the filling number andthe gating (Fig. 3d). It was ound that the collective oscillation o

a b

dc

Graphene

D

S

SiO2

SiO2

Si

G

D S

Ion gel

00

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1000

10

20

30

40

50

60

50 100 150 200 250 300 350200 300 400

200 300 400

5

10

15

− Δ T / T C N P ( % )

− Δ T

I − T

( % )

(cm–1)ω

(cm–1)ω Wavenumber (cm–1)

200

15

10

5

–2.2V–1.8V–1.0V

1 µm 2 µm 4 µm 1 layer 2 layers 5 layers

–

+++

Figure 3 | Intrinsic graphene plasmons. a, The schematics of plasmon

excitation. b, Plasmon resonance in gated graphene microribbon arrays.

Top: top and side views of a typical graphene microribbon array. D, drain;

G, gate; S, source Bottom: gate-induced change of relative transmission

spectra, as a function of excitation frequency for three different gate

voltages. c, Control of plasmon resonance through microribbon width.

Transmission spectra for the samples of different widths and the same

doping concentration. d, Transparent graphene plasmonic devices.

Extinction in stacked plasmonic devices with one, two and five graphene

layers. Figures reproduced with permission from ref. 56 (b and c) and

ref. 58 (d).


DOI: 10.1038/NPHOTON.2012.262


5/10


MDFs is unambiguously quantum mechanical, with the n1/4 scalingo the plasmon requency. In addition, it was shown how stackedgraphene microdiscs could be used as an electromagnetic radiationshield with 97.5% effectiveness, a tunable, ar-inrared notch filterwith a rejection ratio o 8.2 dB and a tunable, terahertz linear polar-izer with an extinction ratio o 9.5 dB (re. 58).

Perhaps the most striking were the recent observations o intrin-sic graphene plasmons in res 50, 59, where the authors used the tip

o an atomic orce microscopy (AFM) probe in a scattering-typescanning near-field optical microscopy set-up to launch and imageDirac plasmons in real space. A schematic picture o the experimentscarried out and some o the findings are summarized in Fig. 4A.Inrared nano-imaging revealed that the plasmon wavelength com-pression ratio, λ0/λpl, can reach 40 and that the plasmon in confinedgeometries can be tuned by gating. Te strong confinement o gra-phene plasmons made it possible to use a single-molecule deect ingraphene as an atomic antenna in the petahertz requency range63.

Nanoscale morphological deects, such as atomic steps andwrinkles, in epitaxial graphene on SiC are responsible or a strongterahertz plasmonic peak in the optical response64. Plasmons in epi-taxial graphene can thus couple to terahertz light in the absence oartificial lithographic patterning. o summarize, so ar experiments

have confirmed the long-wavelength (q


6/10


continuum: interactions between quasiparticles and plasmons arestronger42 in 2D MDFs than in an ordinary non-relativistic 2DESs.

In a doped graphene sheet, it is necessary to take dynamicalscreening into account, which is typically accomplished by consid-ering the RPA effective interaction, W (q, ω) = uq/ε(q, ω). Calculating

the ‘sel-energy’ o quasiparticles within the RPA shows that at somespecific wavenumber, the bare quasiparticle velocity equals theplasmon group velocity. In the case o doped graphene, a chargecarrier scatters into a resonance consisting o a quasiparticle ‘trav-elling together’ with (that is, strongly coupled to) an undampedplasmon excitation, or plasmaron67,68. Plasmaron eatures can alsobe observed by tunnelling spectroscopy 69,70. More details on theproperties o electron–electron interactions and their influence ongraphene optics can be ound in other reviews71,72.

Graphene-based plasmonics—hybrid devicesGraphene is a versatile, broadband, adjustable and tunable opticalmaterial. However, direct applications o graphene in optics andphotonics suffer rom graphene’s relatively inefficient interactionwith light. Although it might be argued that 2.3% o absorption

o light by a single atomic layer is actually a large number, toachieve effective optical modulators and photocells it is neces-sary to enhance light–matter interactions in graphene. Te com-bination o graphene with conventional plasmonics based onnoble metals73,74 could, thereore, be beneficial or both fields o

investigation: plasmonic nanostructures can enhance the opticalproperties o graphene (stronger Raman signature, more effectivegraphene plasmonic photocells and so on), and graphene could beapplied to influence the optical response o plasmonic nanoarrays(or optical modulators and sensing) leading to graphene-basedactive plasmonics.

Raman scattering in graphene enhanced by near-fields of plas-monic nanostructures. Metallic nanostructures excited by lightofen demonstrate localized surace plasmon resonances, which arecharacterized by strongly enhanced near-fields produced by chargesstopping at the surace o the metal. Because interaction o light withgraphene is determined by the local electromagnetic fields (inducedon the graphene sheet), this interaction can effectively be increased byplacing metal nanostructures close to the graphene. Tis approach was

a b

1.0

0.5

1.0

0.5

1.0

0.5

1.0

0.5

500450 550 600 650

Wavelength (nm)Gate voltage (V)Raman shift (cm–1)

R e l a t i v e p h o t o c u r r e n t e n h a n c e m e n t ( a . u . )

R e s i s t a n c e

( k Ω )

N o r m a l

p h o t o v o l t a g e

( V W – 1 )

01,0000

5,000

10,000

15,000

20,000

25,000

30,000

1,500 2,000 2,500 3,000–1012345

–1001020304050

20 40 60 80

0.2

0.4

0.6

0.8

F’(G)

F(G)

F’(2D)

F(2D)

30

20

10

0

E n h a n c e m e n t f a c t o r

I n t e n s i t y ( a . u . )

4 5 0 5 0 0 5 5 0 6 0 0 6 5 0 7 0 0 4 5 0 5 0 0 5 5 0 6 0

0 6 5 0 7 0 0

140 nm210 nmUnpatt

Wavelength (nm)

G 2D

D

x10Dots + SLG(514 nm)

Dots + SLG(633 nm)

SLG (514 nm)

MLG

VSD

VBG

1

2

VTG

V T G

( V )

I

p–n

n-p

p–p

n–n

CNP

R (

k Ω )

3.1

1.6

0.1

2.0

1.0

0.0

–1.0

c d

Figure 5 | Hybrid graphene plasmonic devices. a, Top left: SEM images of a sample for surface-enhanced Raman spectroscopy. Top right: the total

(patterned) enhancement factors F' for the G and 2D Raman peaks for the structures with particle diameter of 140 nm and 210 nm. The black dotted line

is the corresponding interference (unpatterned) enhancement factor F. Bottom left: schematics of plasmonic metamaterial (coupled golden nanodots

on a glass substrate in this particular case) and graphene. Bottom right: Raman spectra of single-layer graphene (SLG) placed on top of gold nanodots,

measured at 514 and 633 nm. b, Top left: SEM images of graphene photodetectors with plasmonic nanostructures. L, θ and TR represent the state of

light polarization. Top right: the corresponding resistance and photovoltaic characteristics for devices with (red squares) and without (blue) plasmonicnanostructures. Solid bars indicate the enhancement ratio (right-hand scale). Bottom: multicolour photodetection using graphene devices coupled to

plasmonic nanostructures (left, SEM images; right, corresponding photoresponses). c, Device geometry and optoelectronic characteristics of the graphene

p–n junction. Left: experimental schematics. Here V BG, V TG and V SD are the voltages of the back, top and side gates, respectively. Right: resistance versus

V BG and V TG at V SD = 1.4 mV and T = 175 K. (CNP stands for charge neutrality point). d, Schematics of charge separation for graphene doped by metallic

contacts Built-in electric field (created at the metal-graphene interface due to the difference in the work functions) separates electron-hole pairs, thus

creating photocurrent. Figures reproduced with permission from refs 75, 76 (a), refs 81, 82 (b) and ref. 86 (c).


DOI: 10.1038/NPHOTON.2012.262


7/10


taken in res 75–78, where the localized surace plasmon resonanceso metallic nanodots were used to increase significantly the Ramanintensity (Fig. 5a). Graphene provides the ideal prototype test mate-rial or investigation o surace-enhanced Raman spectroscopy 79. ItsRaman spectrum is well known80, and graphene samples are entirelyreproducible and can be made virtually deect ree. Te 2D nature ographene allows a closed-orm description o the Raman enhance-ment based on the electromagnetic mechanism74, in agreement withexperiments75. Te development o a generic procedure or graphenetranser on top o a preabricated plasmonic nanostructure76 paves theway or the use o graphene as the platorm o choice or studying andquantiying field amplification in conventional plasmonics.

Plasmonic enhancement of photovoltage in graphene. Near-field enhancement by plasmonic nanostructures was used toimprove the efficiency o photovoltage conversion significantly ingraphene81 and to achieve a spectral selectivity that enables mul-ticolour photodetection82 (Fig. 5b). Graphene-based photodetec-tors have excellent characteristics in terms o quantum efficiencyand reaction time83, because o the very large room-temperaturemobility and high Fermi velocity o graphene’s charge carriers. Acombination o graphene with plasmonic nanostructures makes itpossible to enhance the photovoltage by a actor o 15–20 withoutcompromising the operational speed o a graphene photodetector.It is worth noting that the exact mechanism or light-to-current

b

d

Single layer graphene

Metamaterialstructure

Silicon nitridemembrane

x

y

D

c

a 0.5

0.4

0.3

0.2

0.1

01.4 1.6 1.8 2.0

(α)

(β)

Metamaterial

Wavelength (µm)

T r a n s m i t t a n c e

R a

m a n i n t e n s i t y ( a . u . )

Metamaterialwith graphene

–2 20 –5 50

(β)(α)

d

m

2,000

1,500

1,000

500

0500 1,000 1,500 2,000 2,500 3,000 3,500

D G

2D

|E|

Wavelength (nm)

Wavelength (nm)

Wavelength (nm)

ω (cm–1) |E|

35

30

25

20

15

10

5

590 600 610 620600 610 6200 0

2

–60

–30

30

60

0

4

6

8

10

Ψ ( d e g )

Ψ ( d e g )

( d e g )

Ψ (

d e g )

4000

10

20

30

40

600 800

Figure 6 | Hybrid graphene plasmonic devices. a, Top: schematic of the fabricated complementary split-ring metamaterial. D, is the unit-cell size. Bottom:

Raman spectrum corresponding to the marked region. b, Top: experimental transmission spectrum of an array before (dashed black line) and after (red

solid line) deposition of graphene, for D = 711 nm. Bottom: electric field maps at the plasmon resonances. Vectors d→

and m→

represent dipole and magnetic

moments respectively. c, Evaluation of sensitivity for singular-phase plasmonic detectors. Left: ellipsometric spectra in the region of the collective plasmon

resonance for the pristine double-dot array (black curve) and for the array with graphene transferred on top (red curve). Inset, entire spectrum for the

pristine case. Right: Ellipsometric parameters Ψ and Δ for the case of 1% hydrogenated, green curves, and pristine graphene, red curves, as functions of

wavelength (incidence angle of 70°). One-sided graphene hydrogenation by 1% corresponds to a change in areal mass density of ~1 pg mm−2.

d, Hypothetical graphene-based active plasmonic interferometer, with one of the branches covered with graphene on boron nitride, thus allowing for

modulation of optical signals. Figures reproduced with permission from ref. 91 (a, b) and ref. 94 (c).

FOCUS | REVIEW ARTICLESNATURE PHOTONICS

DOI: 10.1038/NPHOTON.2012.262


8/10


conversion is still debated84,85. Recently, the importance o ther-moelectric effects was highlighted in direct measurements86 o thesix-old symmetry o photovoltage patterns as unctions o bot-tom- and top-gate voltages (Fig. 5c). Tese patterns, together withthe measured spatial and density dependences o the photore-sponse, provide evidence that at low temperatures nonlocal hotcarrier transport dominates the intrinsic photoresponse in plaingraphene with non-structured electrodes86.

Important building blocks o graphene-based photodetec-tors are p–n junctions, which are usually required to separate thephoto-generated electron–hole pairs. Such p–n junctions are ofencreated close to the contacts, because o the difference in the workunctions o metal and graphene87,88. Te geometrical position (and‘strength’) o the p–n junction can be varied by gating, which in turncan be used to modulate the photoresponse o graphene (Fig. 5d).Te presence o electric contact between the graphene and theplasmonic nanostructures is important because it could result inresistive coupling o local surace plasmon resonances89 and leadto additional doping o the graphene sheet. In this respect, verticalgeometries such as those shown in Fig. 2c are more beneficial orlight harvesting90.

Graphene-based plasmonics for modulation and sensing. Tetunability o graphene afforded by gating is vital or the field o activeplasmonics. Active optical elements are o great importance in di-erent areas o science and technology, with applications rangingrom displays to high-tech requency modulators. Despite a greatdeal o progress in optical disciplines, active optics still relies heavilyon liquid crystals, which guarantee deep modulation in inexpen-sive, small cells but are quite slow, and nonlinear optical crystals,which are ast but quite bulky and expensive. For this reason, thedevelopment o inexpensive, ast and small active optical elementswould be o considerable interest.

Recently, plasmonic metamaterials have established themselvesas a versatile tool or creating new optical devices. Teir opticalproperties can be easily controlled by changing the electric coupling

between the plasmonic nanoresonators that constitute a nanomate-rial. Te extraordinary optical, electrical and mechanical propertieso graphene can be used to achieve such control. Te combinationo graphene with plasmonics could result in ast, relatively cheapand small active optical elements and nanodevices. Tere are twomain challenges to this: combining graphene with plasmonic ele-ments and achieving effective control over graphene’s propertiesand the optical response o hybrid optical devices.

In re. 91, a low-pressure chemical vapour deposition processwas used to grow graphene92 on polycrystalline Cu oils and apoly(methyl methacrylate) wet transer procedure was then usedto transer graphene on top o preabricated plasmonic meta-materials (Fig. 6a). Te graphene changed the spectral positiono the plasmonic resonances o the metamaterial as well as the

absolute value o spectral transmission (Fig. 6b). Simple finite-difference time domain calculations93 provided general supportor the experimental data, although the details o graphene resis-tive coupling to plasmons and graphene doping by the metal arestill lacking.

Mechanically exoliated graphene and a poly(methyl meth-acrylate) wet transer procedure have been used to place graphene ontop o regular plasmonic nanoarrays76,94. Tese plasmonic nanoar-rays possess ultranarrow diffractive-coupled plasmon resonances95 (or geometric resonances), which show extremely high phase sen-sitivity to the external environment96. Te presence o graphenestrongly influenced the position o coupled resonances and otheroptical properties o the samples (Fig. 6). o modiy the grapheneproperties, reversible graphene hydrogenation was perormed inwhich an areal mass sensitivity limit o 0.1 g mm−2 was recorded94.

Tis is our orders o magnitude better than the areal sensitivity othe surace plasmon resonance technique97 and thus opens up thepossibility o realizing single-molecule label-ree detection. Owingto the well-defined 2D geometry and the possibility o monitor-ing the hydrogenation level using Raman spectroscopy 98, graphenecould become the platorm o choice or measuring areal mass sen-sitivity in plasmonic nanosensors.

Te ultimate goal or hybrid graphene–plasmonic elements is to

achieve light modulation using graphene gating; see Fig. 6d, wherea hypothetical plasmonic intererometer99 governed by graphene isshown. Te presence o metallic nanostructures makes the problemo gating quite complicated owing to electrostatics and electric dis-charge. Te problem o graphene gating in hybrid plasmonic nan-odevices is currently under intense investigation.

PerspectiveGraphene, its derivatives and other atomic-monolayer materialscould become the building blocks o new generations o multilayeroptical devices. Te range o building blocks now includes all kindso optical materials: dielectrics, semiconductors, semimetals, met-als, those that are gapless, those that have small and large opticalgaps, and so on. We hope that their combination will result in new

optical applications (efficient photocells, ultraast optical modula-tors and graphene-based 2D lasers) in the near uture. Te potentialo graphene plasmons is also being realized, and they show veryattractive eatures including extremely high localization, strongconfinement, efficient and strong light–matter interactions, rela-tively long lietimes, tunability and electrical controllability. Lightmanipulation with intrinsic graphene plasmons and the accessibil-ity o quantum optical regimes promise a revolution in optoelec-tronics and optical computing. A combination o graphene withconventional plasmonic nanostructures and metamaterials willprovide ultrasensitive chemical sensors and biosensors, new non-linear optical elements and effective photodetectors. We oreseegraphene becoming the platorm o choice or quantiying fieldamplification in, and measuring areal sensitivity o, plasmonic

devices. Furthermore, the diverse properties o graphene quasipar-ticles and their interaction with plasmonic vibrations promise newdiscoveries in undamental physics.

References1. Novoselov, K. S. et al. Electric field effect in atomically thin carbon films.

Science 306, 666–669 (2004).2. Novoselov, K. S. et al. wo-dimensional gas o massless Dirac ermions in

graphene. Nature 438, 197–200 (2005).3. Geim, A. K. & Novoselov, K. S. Te rise o graphene. Nature Mater. 6,

183–191 (2007).4. Bonaccorso, F., Sun, Z., Hasan, . & Ferrari, A. C. Graphene photonics and

optoelectronics. Nature Photon. 4, 611–622 (2010).5. Vakil, A. & Engheta, N. ransormation optics using graphene. Science 332,

1291–1294 (2011).6. Koppens, F. H. L., Chang, D. E. & Javier Garcia de Abajo, F. Graphene

plasmonics: a platorm or strong light-matter interactions. Nano Lett. 11, 3370–3377 (2011).7. assin, P., Koschny, ., Kaesaki, M. & Soukoulis, C. M. A comparison o

graphene, superconductors and metals as conductors or metamaterials andplasmonics. Nature Photon. 6, 259–264 (2012).

8. Bao, Q. & Loh, K. P. Graphene photonics, plasmonics, and broadbandoptoelectronic devices. ACS Nano 6, 3677–3694 (2012).

9. Castro Neto, A. H., Guinea, F., Peres, N. M. R., Novoselov, K. S. & Geim, A. K.Te electronic properties o graphene. Rev. Mod. Phys. 81, 109–162 (2009).

10. Kuzmenko, A. B., van Heumen, E., Carbone, F. & van der Marel, D. Universaloptical conductance o graphite. Phys. Rev. Lett. 100, 117401 (2008).

11. Nair, R. R. et al. Fine structure constant defines visual transparency ographene. Science 320, 1308 (2008).

12. Li, Z. Q. et al. Dirac charge dynamics in graphene by inrared spectroscopy.Nature Phys. 4, 532–535 (2008).

13. Bao, Q. L. et al. Atomic-layer graphene as a saturable absorber or ultraastpulsed lasers. Adv. Funct. Mater. 19, 3077–3083 (2009).


DOI: 10.1038/NPHOTON.2012.262


9/10


14. Wang, F. et al. Gate-variable optical transitions in graphene. Science 320, 206–209 (2008).

15. Liu, M. et al. A graphene-based broadband optical modulator. Nature 474, 64–67 (2011).

16. Kravets, V. G. et al. Spectroscopic ellipsometry o graphene and an exciton-shifed van Hove peak in absorption. Phys. Rev. B 81, 155413 (2010).

17. Yang, L., Deslippe, J., Park, C. H., Cohen, M. L. & Louie, S. G. Excitonic effectson the optical response o graphene and bilayer graphene. Phys. Rev. Lett. 103, 186802 (2009).

18. Zhang, Y. B. et al. Direct observation o a widely tunable bandgap in bilayergraphene. Nature 459, 820–823 (2009).19. Elias, D. C. et al. Control o graphene’s properties by reversible hydrogenation:

evidence or graphane. Science 323, 610–613 (2009).20. Nair, R. R. et al. Fluorographene: a two-dimensional counterpart o eflon.

Small 6, 2877–2884 (2010).21. Robinson, J. . et al. Properties o fluorinated graphene films. Nano Lett. 10,

3001–3005 (2010).22. Novoselov, K. S. et al. wo-dimensional atomic crystals. Proc. Natl Acad. Sci.

USA 102, 10451–10453 (2005).23. Coleman, J. N. et al. wo-dimensional nanosheets produced by liquid

exoliation o layered materials. Science 331, 568–571 (2011).24. Mak, K. F., Lee, C., Hone, J., Shan, J. & Heinz, . F. Atomically thin MoS2: a new

direct-gap semiconductor. Phys. Rev. Lett. 105, 136805 (2010).25. Splendiani, A. et al. Emerging photoluminescence in monolayer MoS2. Nano

Lett. 10, 1271–1275 (2010).26. Mak, K. F., He, K., Shan, J. & Heinz, . F. Control o valley polarization in

monolayer MoS2 by optical helicity. Nature Nanotech. 7, 494–498 (2012).27. Gorbachev, R. V. et al. Hunting or monolayer boron nitride: optical and

Raman signatures. Small 7, 465–468 (2011).28. Novoselov, K. S. Graphene: materials in the flatland. Rev. Mod. Phys. 83,

837–849 (2011).29. Dean, C. R. et al. Boron nitride substrates or high-quality graphene

electronics. Nature Nanotech. 5, 722–726 (2010).30. Mayorov, A. S. et al. Micrometer-scale ballistic transport in encapsulated

graphene at room temperature. Nano Lett. 11, 2396–2399 (2011).31. Ponomarenko, L. A. et al. unable metal–insulator transition in double-layer

graphene heterostructures. Nature Phys. 7, 958–961 (2011).32. Furchi, M. et al. Microcavity-integrated graphene photodetector. Nano Lett. 12,

2773–2777 (2012).33. Engel, M. et al. Light–matter interaction in a microcavity-controlled graphene

transistor. Nature Commun. 3, 906 (2012).34. Ferreira, A., Peres, N. M. R., Ribeiro, R. M. & Stauber, . Graphene-based

photodetector with two cavities. Phys. Rev. B 85, 115438 (2012).35. Giuliani, G. F. & Vignale, G. Quantum Teory of the Electron Liquid

(Cambridge Univ. Press, 2005).36. Vicarelli, L. et al. Graphene field-effect transistors as room-temperature

terahertz detectors. Nature Mater. 11, 865–871 (2012).37. Gangadharaiah, S., Farid, A. M. & Mishchenko, E. G. Charge response unction

and a novel plasmon mode in graphene. Phys. Rev. Lett. 100, 166802 (2008).38. Mikhailov, S. A. & Ziegler, K. New electromagnetic mode in graphene. Phys.

Rev. Lett. 99, 016803 (2007).39. Ponomarenko, L. A. et al. Effect o a high-kappa environment on charge carrier

mobility in graphene. Phys. Rev. Lett. 102, 206603 (2009).40. Wunsch, B., Stauber, ., Sols, F. & Guinea, F. Dynamical polarization o

graphene at finite doping. N. J. Phys. 8, 318 (2006).41. Hwang, E. H. & Das Sarma, S. Dielectric unction, screening, and plasmons in

two-dimensional graphene. Phys. Rev. B 75, 205418 (2007).42. Polini, M. et al. Plasmons and the spectral unction o graphene. Phys. Rev. B

77, 081411 (2008).

43. Abedinpour, S. H. et al. Drude weight, plasmon dispersion, and ac conductivityin doped graphene sheets. Phys. Rev. B 84, 045429 (2011).44. Shung, K. W. K. Dielectric unction and plasmon structure o stage-1

intercalated graphite. Phys. Rev. B 34, 979–993 (1986).45. Barlas, Y., Pereg-Barnea, ., Polini, M., Asgari, R. & MacDonald, A. H. Chirality

and correlations in graphene. Phys. Rev. Lett. 98, 236601 (2007).46. Reed, J. P. et al. Te Effective Fine-Structure Constant o Freestanding

Graphene Measured in Graphite. Science 330, 805–808 (2010).47. Principi, A., Polini, M. & Vignale, G. Linear response o doped graphene sheets

to vector potentials. Phys. Rev. B 80, 075418 (2009).48. Ramezanali, M. R., Vazieh, M. M., Asgari, R., Polini, M. & MacDonald, A. H.

Finite-temperature screening and the specific heat o doped graphene sheets. J. Phys. A 42, 214015 (2009).

49. Jablan, M., Buljan, H. & Soljacic, M. Plasmonics in graphene at inraredrequencies. Phys. Rev. B 80, 245435 (2009).

50. Fei, Z. et al. Gate-tuning o graphene plasmons revealed by inrared nano-imaging. Nature 487, 82–85 (2012).

51. Belitz, D. & Das Sarma, S. Plasmon linewidth in metals and semiconductors:a memory-unction approach. Phys. Rev. B 34, 8264–8269 (1986).

52. Principi, A., Asgari, R. & Polini, M. Acoustic plasmons and composite hole-acoustic plasmon satellite bands in graphene on a metal gate. Solid StateCommun. 151, 1627–1630 (2011).

53. Eberlein, . et al. Plasmon spectroscopy o ree-standing graphene films. Phys.Rev. B 77, 233406 (2008).

54. Liu, Y., Willis, R. F., Emtsev, K. V. & Seyller, . Plasmon dispersion anddamping in electrically isolated two-dimensional charge sheets. Phys. Rev. B 78,

201403 (2008).55. Koch, R. J., Seyller, . & Schaeer, J. A. Strong phonon-plasmon coupledmodes in the graphene/silicon carbide heterosystem. Phys. Rev. B 82, 201413 (2010).

56. Ju, L. et al. Graphene plasmonics or tunable terahertz metamaterials. NatureNanotech. 6, 630–634 (2011).

57. Fei, Z. et al. Inrared nanoscopy o Dirac plasmons at the graphene-SiO2 interace. Nano Lett. 11, 4701–4705 (2011).

58. Yan, H. et al. unable inrared plasmonic devices using graphene/insulatorstacks. Nature Nanotech. 7, 330–334 (2012).

59. Chen, J. et al. Optical nano-imaging o gate-tunable graphene plasmons. Nature 487, 77–81 (2012).

60. Kim, J. . & Choi, S-Y. Graphene-based plasmonic waveguides or photonicintegrated circuits. Opt. Express 19, 24557–24562 (2011).

61. Christensen, J., Manjavacas, A., Tongrattanasiri, S., Koppens, F.H. L. & García de Abajo, F. J. Graphene plasmon waveguiding andhybridization in individual and paired nanoribbons. ACS Nano 6,

431–440 (2012).62. Yan, H. et al. Inrared spectroscopy o tunable Dirac terahertz magneto-

plasmons in graphene. Nano Lett. 12, 3766–3771 (2012).63. Zhou, W. et al. Atomically localized plasmon enhancement in monolayer

graphene. Nature Nanotech. 7, 161–165 (2012).64. Crassee, I. et al. Intrinsic terahertz plasmons and magnetoplasmons in large

scale monolayer graphene. Nano Lett. 12, 2470–2474 (2012).65. Hwang, E. H. & Das Sarma, S. Quasiparticle spectral unction in doped

graphene: electron-electron interaction effects in ARPES. Phys. Rev. B 77, 081412 (2008).

66. Bostwick, A. et al. Observation o plasmarons in quasi-reestanding dopedgraphene. Science 328, 999–1002 (2010).

67. Walter, A. L. et al. Effective screening and the plasmaron bands in graphene.Phys. Rev. B 84, 085410 (2011).

68. Lundqvist, B. I. Single-particle spectrum o the degenerate electron gas. Phys.Kondens. Mater. 6, 193–205 (1967).

69. Brar, V. W. et al. Observation o carrier-density-dependent many-bodyeffects in graphene via tunneling spectroscopy. Phys. Rev. Lett. 104, 036805 (2010).

70. LeBlanc, J. P. F., Hwang, J. & Carbotte, J. P. Distinguishing Coulomb andelectron-phonon interactions or massless Dirac ermions. Phys. Rev. B 85, 115126 (2012).

71. Peres, N. M. R. Te transport properties o graphene: an introduction. Rev. Mod. Phys. 82, 2673–2700 (2010).

72. Kotov, V. N., Uchoa, B., Pereira, V. M., Guinea, F. & Castro Neto, A. H.Electron-electron interactions in graphene: current status and perspectives.Rev. Mod. Phys. 84, 1067–1125 (2012).

73. Barnes, W. L., Dereux, A. & Ebbesen, . W. Surace plasmon subwavelengthoptics. Nature 424, 824–830 (2003).

74. Maier, S. A. Plasmonics: Fundamentals and Applications (Springer, 2007).75. Schedin, F. et al. Surace-enhanced Raman spectroscopy o graphene. ACS

Nano 4, 5617–5626 (2010).76. Kravets, V. G. et al. Surace hydrogenation and optics o a graphene sheet

transerred onto a plasmonic nanoarray. J. Phys. Chem. C 116, 3882–3887 (2012).77. Lee, J., Shim, S., Kim, B. & Shin, H. S. Surace-enhanced Raman scattering

o single- and ew-layer graphene by the deposition o gold nanoparticles.Chemistry 17, 2381–2387 (2011).

78. Lee, J., Novoselov, K. S. & Shin, H. S. Interaction between metal and graphene:dependence on the layer number o graphene. ACS Nano 5, 608–612 (2011).

79. Ling, X. et al. Can graphene be used as a substrate or Raman enhancement?Nano Lett. 10, 553–561 (2010).

80. Ferrari, A. C. et al. Raman spectrum o graphene and graphene layers. Phys.Rev. Lett. 97, 187401 (2006).

81. Echtermeyer, . J. et al. Strong plasmonic enhancement o photovoltage ingraphene. Nature Commun. 2, 458 (2011).

82. Liu, Y. et al. Plasmon resonance enhanced multicolour photodetection bygraphene. Nature Commun. 2, 579 (2011).

83. Mueller, ., Xia, F. N. A. & Avouris, P. Graphene photodetectors or high-speedoptical communications. Nature Photon. 4, 297–301 (2010).

FOCUS | REVIEW ARTICLESNATURE PHOTONICS

DOI: 10.1038/NPHOTON.2012.262


10/10


84. Park, J., Ahn, Y. H. & Ruiz-Vargas, C. Imaging o photocurrent generation andcollection in single-layer graphene. Nano Lett. 9, 1742–1746 (2009).

85. Xu, X. D., Gabor, N. M., Alden, J. S., van der Zande, A. M. & McEuen, P. L.Photo-thermoelectric effect at a graphene interace junction. Nano Lett. 10, 562–566 (2010).

86. Gabor, N. M. et al. Hot carrier–assisted intrinsic photoresponse in graphene.Science 334, 648–652 (2011).

87. Giovannetti, G. et al. Doping graphene with metal contacts. Phys. Rev. Lett. 101, 026803 (2008).

88. Blake, P. et al. Influence o metal contacts and charge inhomogeneity ontransport properties o graphene near the neutrality point. Solid State Commun. 149, 1068–1071 (2009).

89. Kravets, V. G., Schedin, F. & Grigorenko, A. N. Fine structure constant andquantized optical transparency o plasmonic nanoarrays. Nature Commun. 3, 640 (2012).

90. Atwater, H. A. & Polman, A. Plasmonics or improved photovoltaic devices.Nature Mater. 9, 205–213 (2010).

91. Papasimakis, N. et al. Graphene in a photonic metamaterial. Opt. Express 18, 8353–8359 (2010).

92. Li, X. S. et al. Large-area synthesis o high-quality and uniorm graphene filmson copper oils. Science 324, 1312–1314 (2009).

93. Zou, Y., assin, P., Koschny, . & Soukoulis, C. M. Interaction betweengraphene and metamaterials: split rings vs. wire pairs. Opt. Express 20, 12198–12204 (2012).

94. Kravets, V. G. et al. Singular-phase nanooptics: towards label-ree single moleculedetection. Nature Mater. (submitted).

95. Kravets, V. G., Schedin, F. & Grigorenko, A. N. Extremely narrow plasmonresonances based on diffraction coupling o localized plasmons in arrays ometallic nanoparticles. Phys. Rev. Lett. 101, 087403 (2008).

96. Kravets, V. G., Schedin, F., Kabashin, A. V. & Grigorenko, A. N. Sensitivity ocollective plasmon modes o gold nanoresonators to local environment. Opt.Lett. 35, 956–958 (2010).

97. Kabashin, A. V., Patskovsky, S. & Grigorenko, A. N. Phase and amplitude

sensitivities in surace plasmon resonance bio and chemical sensing. Opt.Express 17, 21191–21204 (2009).98. Elias, D. C. et al. Control o graphenes properties by reversible hydrogenation:

evidence or graphane. Science 323, 610–613 (2009).99. Bozhevolnyi, S. I., Volkov, V. S., Devaux, E., Laluet, J-Y. & Ebbesen, .

W. Channel plasmon subwavelength waveguide components includingintererometers and ring resonators. Nature 440, 508–511 (2006).

Acknowledgements We thank A. K. Geim, V. I. Fal’ko, M. I. Katsnelson, R. Asgari, R. Fazio, F. Guinea, A.H. MacDonald, V. Pellegrini, E. Rotenberg, F. addei, A. redicucci and G. Vignaleor conversations. M.P. was supported by MIUR through the FIRB programme, grantno. RBFR10M5B. A.N.G. was partly supported by an FP7 Metachem grant and theSamsung GRO programme.


DOI: 10.1038/NPHOTON.2012.262

NP JP Graphene Plasmonics

Documents

Transcript of NP JP Graphene Plasmonics