Graphene as a substrate for plasmonic nanoparticles

Home

Search

Collections

Journals

About

Contact us

My IOPscience

Graphene as a substrate for plasmonic nanoparticles

This content has been downloaded from IOPscience. Please scroll down to see the full text. 2013 J. Opt. 15 114001 (http://iopscience.iop.org/2040-8986/15/11/114001) View the table of contents for this issue, or go to the journal homepage for more

Download details: This content was downloaded by: polyushkin IP Address: 62.178.75.48 This content was downloaded on 29/10/2013 at 09:08

Please note that terms and conditions apply.

IOP PUBLISHING

JOURNAL OF OPTICS

J. Opt. 15 (2013) 114001 (6pp)

doi:10.1088/2040-8978/15/11/114001

Graphene as a substrate for plasmonic nanoparticles Dmitry K Polyushkin1 , James Milton2 , Salvatore Santandrea3 , Saverio Russo2 , Monica F Craciun4 , Stephen J Green1 , Laureline Mahe1 , C Peter Winolve1 and William L Barnes1,2 1

School of Physics and Astronomy, University of Exeter, Stocker Road, Exeter, EX4 4QL, UK Centre for Graphene Science, School of Physics and Astronomy, University of Exeter, Stocker Road, Exeter, EX4 4QL, UK 3 Dipartimento di Fisica ‘E.R.Caianiello’ and Centro Interdipartimentale NANO MATES Universita’ di Salerno, and CNR-SPIN Salerno, via Ponte don Melillo, I-84084, Fisciano (SA), Italy 4 Centre for Graphene Science, School of Engineering, University of Exeter, Exeter, EX4 4QF, UK 2

E-mail: [email protected]

Received 21 March 2013, accepted for publication 22 May 2013 Published 28 October 2013 Online at stacks.iop.org/JOpt/15/114001 Abstract We report results from an investigation into the plasmonic properties of metallic nanoparticles supported by graphene fabricated using two different methods. In the first method we used electron-beam lithography to produce ordered arrays of metallic nanoparticles. In the second method we used a technique based on electrochemistry to produce random arrangements of metallic nanoparticles. Our results show that both pristine graphene and a more conducting intercalated variant are excellent substrates for plasmonic nanoparticles owing to their transparency and atomically thin nature, opening an interesting route for building plasmon-based bio- and chemical-sensors, and for developing transparent and flexible optoelectronics. Keywords: graphene, plasmonics, electrochemistry

graphene with plasmonic nanoparticles will, as for the transparent conductor indium tin oxide (ITO) [11], be a suitable platform for optical transmission-based LSPR biosensors. The use of graphene rather than ITO for transparent LSPR-based sensors opens up many possibilities. Firstly, the graphene surface could be easily modified for additional, selectivity-enhancing functionality (including bio-functionality) using the non-covalent but irreversible π –π stacking of linker molecules based on moieties such as pyrene [12]. Secondly, the ability to produce transparent, layered structures of nanoparticles interspersed with graphene sheets makes multiple-plasmon readout possible. Combinations of LSPR and electrochemical sensing are also possible on these nanoparticle-decorated graphene structures. Arrays of metallic nanoparticles have often been made on ITO by electron-beam lithography [13, 14] and electrochemistry [11, 15]; more recently they have also been

Graphene offers an interesting prospect as an optoelectronic substrate, it is transparent, can be made conductive, and is very thin. At the same time the unique properties displayed by the plasmon modes associated with metallic nanoparticles allow the manipulation and control of visible light deep into the sub-wavelength regime, with potential applications ranging from the treatment of cancer [1] and light-emitting devices [2], to data storage [3] and bio-sensing based on propagating surface plasmon resonances (PSPR) and localized surface plasmon resonances (LSPR) [4, 5]. The combination of these two classes of material, graphene and metallic nanoparticles, offers exciting prospects not available from either material on its own. With regard to sensing, graphene/plasmonicnanoparticle materials show great promise for surface enhanced Raman spectroscopy (SERS) [6–8], and have significant potential for future plasmon-based sensors and biosensors [9, 10]. This suggests that the combination of 2040-8978/13/114001+06$33.00

1

c 2013 IOP Publishing Ltd Printed in the UK & the USA

J. Opt. 15 (2013) 114001

D K Polyushkin et al

evaporation and lift-off processes. The arrays consist of nano-discs of 100 nm diameter and 50 nm thickness on a square lattice of period 1.0 µm (see figure 1). To characterize the optical properties of the fabricated particles we performed dark-field scattering measurements in reflection mode (see figure 1(b)) [27]. An inverted optical microscope and spectrometer with a charge-coupled device (CCD) camera were used to acquire spectra from the scattered light. Our setup allowed spectra to be obtained from ∼20 particles at once with a spectral resolution of better than 10 nm. The scattered light spectra, Iscat (λ) were normalized with reference to the background scattering spectra, Iback (λ), obtained from the substrate as follows,

made on graphene by electrochemistry [16–18] but, to the best of the authors’ knowledge, the present work is the first instance of gold or silver nanoparticles being electrodeposited onto graphene on a transparent (glass) substrate. With regard to optoelectronics, the use of graphene as a substrate for metallic nanoparticles also opens up many possibilities. The combination of high conductivity and transparency [19] may find application in flexible optoelectronics, for example in plasmonically enhanced flexible photovoltaic cells [20]. It is also possible that graphene (and other atomically thin materials) will provide the ultimate dielectric spacer, something that is much sought-after in the quest for high optical field-enhancements in plasmonics, where high field-enhancements are known to occur when metallic nanostructures are separated by sub-nm gaps [21, 22]. To explore the properties of plasmonic nanoparticles on graphene substrates, we undertook an investigation employing two very different methods. In the first we used electron-beam lithography as a top-down nanofabrication route. Metal particles have been produced on graphene by electron-beam lithography before [14], but the effect of the graphene on the plasmon resonance of particles in direct contact with the graphene was not investigated. In the second method we used electrodeposition [23] as a bottom-up approach, attractive for low cost, large scale applications. The focus of the work reported here was to assess how the plasmonic response of metallic nanoparticles is modified by the presence of a graphene substrate. The dependence of the localized surface plasmon resonance on the refractive index of the region around the particle is central to the use of LSPRs as sensors [4]. Part of the reason for our study was to see whether the presence of graphene leads to a shift in the spectral position of the plasmon resonance: graphene has a higher refractive index than typical glass substrates so one might expect its presence to shift the spectral position of the plasmon resonance [24]. For metallic nanoparticles made by electron-beam lithography (EBL), silicon wafers coated with 300 nm of SiO2 were used as substrates. First, graphene flakes were deposited onto these substrates by mechanical exfoliation, and flakes of different thicknesses (ranging from 1 to 6 layers) were identified using optical microscopy, the number of layers being determined through examination of their optical contrast [25] and by Raman spectroscopy [26]. In addition to graphene flakes we also looked at doped graphene, produced by intercalation with ferric chloride. Intercalated graphene flakes were produced as follows. The substrate was placed in a two-zone furnace where different temperatures in the two zones created a temperature gradient. The substrate was placed in one zone and anhydrous ferric chloride powder (FeCl3 ) placed in the other, then both zones were heated inducing the FeCl3 to sublime. The temperature gradient between the two zones means that FeCl3 molecules move and fill the space between the layers of the graphene. This process allowed hole doping levels of up to 8.9 × 1014 cm−2 to be achieved [19]. Regular arrays of silver nanoparticles were patterned on the graphene flakes by standard electron-beam lithography,

Inorm (λ) =

Iscat (λ) − Iback (λ) . Iback (λ)

Here Iscat is the signal recorded by the CCD camera with the sample in place, and Iback is the measured spectrum of light scattered by the bare substrate, which we found to be indistinguishable (to within experimental error) to the scattering arising from pristine graphene and multilayer graphene on an otherwise bare substrate. Representative normalized scattering spectra from arrays of silver nanoparticles on graphene flakes and intercalated graphene flakes are shown in figures 2 and 3 respectively. Both figures show spectra obtained from particle arrays on top of few-layer graphene flakes consisting of different numbers of layers. For single layers of both types of graphene—pristine and intercalated—a broad (150 nm) peak is seen in the scattering spectra, centred at ∼530 nm. This peak is a result of the localized surface plasmon resonance associated with the silver-nanoparticle. As the number of graphene layers is increased the major difference in the scattering spectra is that the plasmon resonance becomes weaker. This attenuation of the scattering spectra with increasing number of graphene layers is highlighted in the insets to figures 2 and 3, from which we can see that the degree of attenuation for the pristine graphene is somewhat greater than for the intercalated graphene. Here attenuation refers the numerical factor by which the intensity is decreased, an attenuation of 2 means the intensity is halved. Previous experimental studies have shown that an increase of the effective refractive index (permittivity) of the region surrounding metallic nanoparticles leads to a red-shift in the spectral position of the particle plasmon resonance [24]. For few-layer graphene one might expect that increasing the thickness of the graphene substrate would increase the effective refractive index of the region surrounding the particle, thus leading to a red-shift in the spectral position of the plasmon resonance. However, this expectation is in contrast with our observation that the spectral position of the plasmon resonance does not shift, rather it is only attenuated as the number of graphene layers is increased, irrespective of the graphene doping level (from 0 up to 1014 holes cm−2 , see figures 2 and 3). This difference between expectation and observation can be explained by considering the thickness of 6 layers of graphene, it is just ∼2 nm, at least an order of magnitude less than the optical field penetration depth 2

J. Opt. 15 (2013) 114001

D K Polyushkin et al

Figure 1. Silver nanoparticles made on graphene by electron-beam lithography. (a) Scanning electron micrograph of an array of silver nanoparticles deposited onto a graphene flake by electron-beam lithography and sample structure and geometry. The regions of the flake visible through their different grey scales are: thick, ∼10 layers of graphene, darkest; 2 layers, middle; and 1 layer, lightest. The scale bar is 5 µm. (b) Arrangement used to acquire dark-field scattering spectra. Light is sent through the outer cladding of the objective so as to illuminate samples at grazing angles of incidence. Scattered light is collected and directed to the spectrometer. The design details of the circular disc shaped silver nanoparticles are shown in (d) and the arrays into which they were formed in (c).

Figure 3. Experimental scattering spectra from the EBL particles on intercalated graphene of different thicknesses (1–6 layers), and on the bare substrate. Inset: extent of attenuation (factor by which intensity is reduced) of the scattered intensity as a function of layer thickness.

Figure 2. Experimental scattering spectra from the EBL particles on pristine graphene of different thicknesses (1–6 layers), and on the bare substrate. Inset: extent of attenuation (factor by which intensity is reduced) of the scattered intensity as a function of layer thickness.

associated with localized surface plasmon resonances [28]. The key feature that we need to account for in these data is the decrease in scattering strength with increasing thickness of the graphene layer. One way in which the graphene can modify the scattering is to alter the strength of the optical field at the surface of the substrate—the presence of the graphene

layer, of relatively high refractive index, is to alter the strength of the reflected field. This reflected field interferes with the incident field, and it is this interference that can lead to a modified strength of the field at the surface—the field that drives the scattering process. To investigate this possibility 3

J. Opt. 15 (2013) 114001

D K Polyushkin et al

Figure 5. A typical SEM image of a graphene/gold (G/Au) electrode, the scale bar is 500 nm. (The growth conditions for this layer are described in the main text.)

Figure 4. Calculated ratio of the square of the relative electric field strength on the substrate/air interface, and the graphene/air interface for different numbers of layers. The data points correspond to an incident angle of 64◦ . The dotted lines indicate the range appropriate for incident angles of 58◦ and 72◦ (bottom and upper lines, respectively). (The attenuation is the factor by which intensity is reduced.)

contact was prepared on the edge of the graphene film by vapour deposition under vacuum through an appropriate mask. A tinned copper wire contact was made to this gold using conductive epoxy. Finally, the exposed gold and conducting epoxy were covered with electrically insulating epoxy resin. This basic electrode design was used to fabricate, by electrodeposition, silver-nanoparticle decorated graphene electrodes (G/Ag) and gold-nanoparticle decorated graphene electrodes (G/Au). In a further system, a second layer of graphene was added (G/Au/G) onto which silver nanoparticles were electrodeposited to give the multilayered structure (G/Au/G/Ag). The addition of the second graphene layer followed the same procedure (see below) as the initial transfer of graphene to glass. The edges of this two-layer structure were sealed with epoxy to prevent any ingress of solution through capillarity during the second electrodeposition step of silver nanoparticles. A typical SEM image of a graphene/gold (G/Au) electrode is shown in figure 5. To facilitate the transfer of the CVD graphene from its copper catalyst backing onto glass microscope slides, it was firstly spin coated at 4000 rpm with a 1:1 by volume solution of 950 K molecular weight poly(methylmethacrylate) (PMMA) in anisole, giving about a 70 nm layer of the polymer. The PMMA layer was then cured on a hotplate at 120 ◦ C for 20 min, before being allowed to cool to room temperature. The copper was etched from this PMMA-coated graphene by immersion in an aqueous solution of FeCl3 (2 g in 80 cm3 ) for 16 h. Then, following four, 15 min periods of rinsing with de-ionized water, the now free-floating PMMA/graphene film was transferred (graphene face down) onto a glass slide dipped into the water beneath it and dried on a hotplate at 60 ◦ C for 45 min. The addition of a few drops of the PMMA in anisole solution to the existing polymer layer, followed by heating on a hotplate at 180 ◦ C for 5 min, was used to ease the subsequent removal of the PMMA by immersion in boiling acetone for 45 min. Finally, the glass-backed graphene film was rinsed with 2-propanol and blown dry in a stream of nitrogen gas.

we conducted a simple calculation based on Fresnel reflection coefficients. In the Fresnel model we incorporated the Si wafer substrate (permittivity = 16 + 0.5i) covered with a 300 nm thick layer of SiO2 (permittivity = 2.13 + 0.0i). We then added graphene of different thicknesses (1–6 layers) and finally included air above the structure. The optical constants of the graphene were taken to be n = 2.6, k = 1.3 [29] and each layer thickness to be 0.35 nm. The dark-field objective used in the experiments had a numerical aperture (NA) 0.90, corresponding to maximum angle of 20 = arcsin(0.90) ≈ 64◦ . To collect only scattered light the structure should be illuminated at angles higher than 20 . These data (figure 4) show that one should expect an attenuation of the scattered signal as the number of graphene layers is increased, here an attenuation of 1.3 times for 6 layers. (As for the insets to figures 2 and 3, attenuation in figure 4 refers the numerical factor by which the intensity is decreased, an attenuation of 2 means the intensity is halved.) For comparison results for different angles of incidence are also included, specifically data for illumination angles 21 = arcsin(0.85) and 21 = arcsin(0.95), see figure 4. This prediction of attenuation by a factor of 1.3 compares with the observed attenuation, for which the factor is 2.3 (figure 2). The lack of quantitative agreement is to be expected, the angle dependence of the scattering process makes the quantitative comparison of dark-field spectra with numerical models an involved task [30], one that was beyond the scope of the present study. Whilst flakes of graphene were sufficient for the studies reported above on metallic nanoparticles produced by electron-beam lithography, for electrochemical deposition the need to use the graphene as an electrode in an electrochemical cell meant that larger area graphene samples were required. We used CVD graphene grown on copper foil and transferred it onto a glass substrate. A gold 4

J. Opt. 15 (2013) 114001

D K Polyushkin et al

For the electrodeposition, silver nitrate (AgNO3 ) (Fluka, 99.0%) and sulfuric acid (H2 SO4 ) (Sigma-Aldrich, 95–98%, reagent grade) were used as received; hydrogen tetrachloroaurate (HAuCl4 ) was prepared by a standard literature procedure [31]. Single-layer, chemical-vapour deposition (CVD) graphene on copper foil was purchased from the Graphene Supermarket (Graphene Laboratories Inc.). Electrodeposition experiments were performed using graphene as the working electrode (approximate area 0.5 cm2 ) in a three-electrode cell, with Pt counter-electrode and either an Ag-wire pseudoreference or an Ag/AgCl/3M NaCl reference (BASi). The cell was controlled using a BASi 100 W electrochemical workstation. Gold nanoparticles were electrodeposited onto graphene from a nitrogen-purged solution of 1.3 mM HAuCl4 in 0.5 M H2 SO4 . Deposition occurred over four successive cyclic voltammograms recorded at 0.1 V s−1 , between 0 V (start and finish) and −2 V relative to the Ag/AgCl/3M NaCl reference. Silver nanoparticles were electrodeposited onto graphene from a nitrogen-purged solution of 2.0 mM AgNO3 in 0.5 M H2 SO4 . Deposition occurred at a fixed potential, held for 60 s, of −2 V relative to the silver wire pseudoreference. To characterize the optical properties of the electrodeposited samples, absorbance spectra were obtained using a UV–visible spectrophotometer. Scanning electron microscopy (SEM) was used to determine the size of the Au particles deposited on the graphene as 30–60 nm. Figure 6 shows UV–visible absorbance spectra recorded for a G/Au/G/Ag (Ag on top) electrode at the various stages of its construction. The first spectrum recorded was for G/Au in air and shows a broad peak at a wavelength of 650 nm, attributed to the gold nanoparticles. The addition of a second graphene layer, to give G/Au/G, produced only minor changes in the peak shape and position in air, indicating that the presence of the graphene over-layer did not significantly alter the plasmonic properties of the gold nanoparticles. Further, replacing air with water did not change the peak shape or position for the G/Au/G, indicating that the graphene-covered gold particles were isolated from this change in dielectric, both physically (the system was sealed to prevent any ingress of liquid between the graphene sheets) and in terms of their plasmonic response. The lack of sensitivity of the plasmonic response to the change from air to water indicates that the second graphene sheet was not in sufficiently intimate contact with the underlying gold particles, particularly in any so-called ‘hot spot’ regions between particles, to provide sensitivity to the air/water dielectric change. The graphene sheet is likely to have made only point contacts with the gold particles and then only with the larger particles within the considerable size distribution on the surface, thus bridging the smaller particles to leave them with no contact with the graphene at all. Such an explanation has been offered to account for a lack of surface enhanced Raman (SERS) activity in a similarly graphene-veiled gold-nanoparticle structure [32]. Here [32], SERS activity was imparted by thermal annealing to create a ‘wrinkled’ graphene sheet that was therefore able to make contact not only with more of the underlying gold particles, but also, by moulding to

Figure 6. UV–visible absorption spectra recorded for the graphene/gold/graphene/silver (G/Au/G/Ag) layered electrode at the various stages of its construction. For ease of comparison, 0.3 and 0.5 absorbance units have been subtracted from the spectra for G/Au/G and the spectrum for G/Au/G/Ag, respectively. The spectrum shown for a separate graphene/silver (G/Ag) electrode has also been adjusted downward in absorbance to facilitate easy comparison of peak shapes and positions with those given by the layered structure.

the particle shape, with the inter-particle hot spots. This approach could have been adopted in the present case, but here the plasmonic insensitivity of the G/Au/G is considered of potential use in providing a fixed reference peak in multiple-layer (e.g. G/Au/G/Ag), multi-plasmon sensors. The electrodeposition of a top layer of silver nanoparticles, to give G/Au/G/Ag, resulted in the appearance of an absorption peak (in water) at 405 nm, attributed to the silver nanoparticles (and similar to that recorded for a separate G/Ag electrode); the blue shift of the gold peak to around 585 nm is an apparent shift due to the superposition of the gold and silver peaks. In this configuration, the top layer of silver particles is exposed to the surrounding medium and so available to act as an LSPR sensor surface. It is also straightforward to extend the methods employed here to produce the alternative configurations of G/Au/G/Au, G/Ag/G/Ag and G/Ag/G/Au as befits the particular sensing application. In conclusion, we have shown that graphene is an excellent substrate for plasmonic (metallic) nanoparticles, produced by electron-beam lithography or by electrodeposition. We have demonstrated that the electrodeposition of plasmonic nanoparticles onto graphene can be used to fabricate multiplelayer structures, such as graphene/gold/graphene/silver, on a glass (and thus transparent) substrate. This offers the prospect of a simple, cost-effective and versatile route to the fabrication of transmission-based sensors relying on localized surface plasmon resonance (LSPR), with multiple-plasmon readout, in possible combination with electrochemistry. The graphene/metal top surface of such structures allows for future modification of either material (or both) with chemical and biological sensing elements. These results show that the combination of graphene with plasmonic metallic nanoparticles, particularly in multiple-layer structures, offers great promise as a platform for bio-/chemical-sensors and, more generally, for transparent, physically flexible (for 5

J. Opt. 15 (2013) 114001

D K Polyushkin et al

example using plastic substrates) and chemically-tunable materials in electronics applications.

[14] Yurtsever A, Van der Veen R M and Zewail A H 2012 Subparticle ultrafast spectrum imaging in 4D electron microscopy Science 335 59–64 [15] Wang Y, Deng J, Di J and Tu Y 2009 Electrodeposition of large size gold nanoparticles on indium tin oxide glass and application as refractive index sensor Electrochem. Commun. 11 1034–7 [16] Wang L, Zhu H, Hou H and Zhang Z 2012 A novel hydrogen peroxide sensor based on Ag nanoparticles electrodeposited on chitosan–graphene oxide/cysteamine-modified gold electrode J. Solid State Electrochem. 16 1693–700 [17] Liu C, Wang K, Luo S, Tang Y and Chen L 2011 Direct electrodeposition of graphene enabling the one-step synthesis of graphene–metal nanocomposite films Small 7 1203–6 [18] Jiang B, Wang M, Chen Y, Xie J and Xiang Y 2012 Highly sensitive electrochemical detection of cocaine on graphene/AuNP modified electrode via catalytic redox-recycling amplification Biosens. Bioelectron. 32 305–8 [19] Khrapach I, Withers F, Bointon T H, Polyushkin D K, Barnes W L, Russo S and Craciun M F 2012 Novel highly conductive and transparent graphene based conductors Adv. Mater. 24 2844–9 [20] Echtermeyer T J, Britnell L, Jasnos P K, Lombardo A, Gorbachev R V, Grigorenko A N, Geim A K, Ferrari A C and Novoselov K S 2011 Strong plasmonic enhancement of photovoltage in graphene Nature Commun. 2 458 [21] Brunazzo D, Descrovi E and Martin O J F 2009 Narrowband optical interactions in a plasmonic nanoparticle chain coupled to a metallic film Opt. Lett. 34 1405–7 [22] Ciraci C, Hill R T, Mock J J, Urzhumov Y, Fernandez-Dominguez A I, Maier S A, Pendry J B, Chilkoti A and Smith D R 2012 Probing the ultimate limits of plasmonic enhancement Science 337 1072–4 [23] Dahlin A B, Dielacher B, Rajendran P, Sugihara K, Sannomiya T, Zenobi-Wong M and Voros J 2012 Anal. Bioanal. Chem. 402 1773–84 [24] Murray W A, Augui´e B and Barnes W L 2009 Sensitivity of localized surface plasmon resonances to bulk and local changes in the optical environment J. Phys. Chem. C 113 5120–5 [25] Craciun M F, Russo S, Yamamoto M, Oostinga J B, Morpurgo A F and Tarucha S 2009 Trilayer graphene is a semimetal with a gate-tunable band overlap Nature Nanotechnol. 4 4–6 [26] Jhang S H et al 2011 Stacking-order dependent transport properties of trilayer graphene Phys. Rev. B 84 161408 [27] S¨onnichsen C, Franzl T, Wilk T, Von Plessen G and Feldmann J 2002 Drastic reduction of plasmon damping in gold nanorods Phys. Rev. Lett. 88 077402 [28] Murray W A, Suckling J R and Barnes W L 2006 Overlayers on silver nanotriangles: field confinement and spectral position of localized surface plasmon resonances Nano Lett. 6 1772–7 [29] Bruna M and Borini S 2009 Optical constants of graphene layers in the visible range Appl. Phys. Lett. 94 3 [30] Knight M W, Fan J, Capasso F and Halas N J 2010 Influence of excitation and collection geometry on the dark field spectra of individual plasmonic nanostructures Opt. Express 18 2579–87 [31] Brauer G 1965 Handbook of Preparative Inorganic Chemistry (New York: Academic) [32] Xu W, Xiao J, Chen Y, Chen Y, Ling X and Zhang J 2013 Graphene-veiled gold substrate for surface-enhanced Raman spectroscopy Adv. Mater. 25 928–33

Acknowledgments The authors would like to acknowledge D Anderson and D Horsell for technical support and D Hudson for help with device fabrication. This work was supported in part by the Centre for Graphene Science award (EPSRC/HEFCE, EP/G036101/1), and through an EPSRC grant under the EPSRC’s UK–Japan scheme, EP/J000396/1. SR, MFC and WLB would also like to acknowledge support from The Royal Society. WLB would also like to acknowledge support from The Leverhulme Trust.

References [1] Loo C, Lowery A, Halas N J, West J and Drezek R 2005 Immunotargeted nanoshells for integrated cancer imaging and therapy Nano Lett. 5 709–11 [2] Barnes W L 2004 Turning the tables on surface plasmons Nature Mater. 3 588–9 [3] Zijlstra P, Chon J W M and Gu M 2009 Five-dimensional optical recording mediated by surface plasmons in gold nanorods Nature 459 410–3 [4] Willets K A and Van Duyne R P 2007 Localized surface plasmon resonance spectroscopy and sensing Annu. Rev. Phys. Chem. 58 267–97 [5] Svedendahl M, Chen S, Dmitriev A and Kall M 2009 Refractometric sensing using propagating versus localized surface plasmons: a direct comparison Nano Lett. 9 4428–33 [6] Sun S and Wu P 2011 Competitive surface-enhanced Raman scattering effects in noble metal nanoparticle-decorated graphene sheets Phys. Chem. Chem. Phys. 13 21116–20 [7] Xu W, Ling X, Xiao J, Dresselhaus M S, Kong J, Xu H, Liu Z and Zhang J 2012 Surface enhanced Raman spectroscopy on a flat graphene surface Proc. Natl Acad. Sci. USA 109 9281–6 [8] Schedin F, Lidorikis E, Lombardo A, Kravets V G, Geim A K, Grigorenko A N, Novoselov K S and Ferrari A C 2010 Surface-enhanced Raman spectroscopy of graphene ACS Nano 4 5617–26 [9] Wu L, Chu H S, Koh W S and Li E P 2010 Highly sensitive graphene biosensors based on surface plasmon resonance Opt. Express 18 14395–400 [10] Kravets V G, Schedin F, Jalil R, Britnell L, Novoselov K S and Grigorenko A N 2012 Surface hydrogenation and optics of a graphene sheet transferred onto a plasmonic nanoarray J. Phys. Chem. C 116 3882–7 [11] Deng J, Song Y, Wang Y and Di J 2010 Label-free optical biosensor based on localized surface plasmon resonance of twin-linked gold nanoparticles electrodeposited on ITO glass Biosens. Bioelectron. 26 615–9 [12] Kodali V K, Scrimgeour J, Kim S, Hankinson J H, Carroll K M, De Heer W A, Berger C and Curtis J E 2011 Nonperturbative chemical modification of graphene for protein micropatterning Langmuir 27 863–5 [13] F´elidj N, Laurent G, Aubard J, L´evi G, Hohenau A, Krenn J R and Aussenegg F R 2005 Grating-induced plasmon mode in gold nanoparticle arrays J. Chem. Phys. 123 221103

6