PhotonicsandNanostructures–FundamentalsandApplications15(2015)109–115
Availableonlineatwww.sciencedirect.com
ScienceDirect
E-Beam
lithography
designed
substrates
for
surface
enhanced
Raman
spectroscopy
夽
Neval
A.
Cinel
a,∗,
Semih
Cakmakyapan
a,b,
Serkan
Butun
c,
Gulay
Ertas
d,
Ekmel
Ozbay
a,b,eaBilkentUniversity,NanotechnologyResearchCenter,06800,Ankara,Turkey bBilkentUniversity,PhysicsDepartment,06800,Ankara,Turkey
cTechnologicalInstitute,ElectricalEngineeringandComputerScienceDepartment,NorthwesternUniversity,Evanston,IL60208,USA dMETU,ChemistryDepartment,06800,Ankara,Turkey
eBilkentUniversity,ElectricalandElectronicsEngineeringDepartment,06800,Ankara,Turkey
Received29September2014;receivedinrevisedform24October2014;accepted24November2014 Availableonline25January2015
Abstract
SurfaceEnhancedRamanSpectroscopy(SERS)isapopularmethodthatamplifiesweakRamansignalsfromRaman-activeanalyte
moleculesmakinguseofcertainspecially-preparedmetallicsurfaces.ThemainchallengeinSERSistodesignandfabricatehighly
repeatable,predictable,andsensitivesubstrates.Therearemanyfabricationmethodsthatstrivetoachievethisgoal,whichare
brieflysummarizedinthispaper.TheE-beamlithographymethodisproposedtobesuperiortothementionedtechniques.Inthis
paper,wereviewhowEBLcanbeutilizedinthepreparationofSERSsubstratesandwediscussthecontributionstothefieldby
theÖzbaygroup.
©2015ElsevierB.V.Allrightsreserved.
Keywords: Electronbeamlithography;SurfaceenhancedRamanspectroscopy;Nano-cone;Nano-ring;Bowtie;Fractal
1. Introduction
SERShasbeenextensivelyusedasapowerfulmethod forthesensitiveandspecificdetectionofawiderange ofbio-moleculessinceitwasdiscoveredin1977[1,2]. SERSdependsonaphenomenoncalledRaman scatter-ing.Whenaphotoninteractswithanatomormolecule, averysmallpercentageofthephotonsarescatteredwith afrequencylessthantheincomingphotons.Theenergy
夽 ThearticlewasacceptedfromPECS–XIconference. ∗Correspondingauthor.
E-mailaddress:nyilmaz@ee.bilkent.edu.tr(N.A.Cinel).
differenceisdissipatedatthevibrationalmodesofthe moleculeininteraction.Ifthe scatteredlight intensity isplottedagainstfrequency,theRamanspectrumofthe moleculecanbeobtained.Everychemicalbondinthe molecule stands for a different peak in the spectrum. TheRamanspectra arejustlike thefingerprintofthe moleculeinthismanner.Thisphenomenonisenhanced atmetallic,nano-structuredsurfacesby105–106andis namedSERS.
Electromagnetic(EM)andchemicalmechanismsare consideredasthefactorsthatenablesurfaceenhanced Raman scattering [3,4]. The dominant mechanisms amongthetwoaretheEMmechanismwhichisdueto theexcitationofthelocalizedsurfaceplasmons (LSP) onnano-roughenedsurfaces.TheLSPisthecollective
http://dx.doi.org/10.1016/j.photonics.2014.11.003
tothis, toenhancethe signalintensity ofSER scatter-ing,allthefactorsthataffectLSPresonanceshouldbe controlled.
After the discovery of SERS, several fabrication techniqueshavebeendevelopedtoattainstabilized sub-stratesthathavelongshelflivesandhighquality.The first SERS-active surfaces were obtained by chemi-cal methods such as oxidation-reduction cycling [5], electro-chemicalroughening[6]andislandlithography [7].Inoxidation-reductioncycling,repeated oxidation-reductioncyclesareappliedtothemetallic(Ag)surfaces endingupwithroughenedsurfacesatthenano-scale.In electro-chemicalroughening the metallic surfaces are immersed in highly concentrated acidic solutions for briefperiods tochemically roughenthe surfaces.The metalisland lithographymethod,takes itsname from theislandlike distributionof the Agfilms evaporated ontoapropersubstrate.Inallthementionedtechniques, unreproducibleandheterogeneous substratesthat have a large surface area but low SERS enhancement are obtained.
There are also lithographic techniques such as nanospherelithography[8],obliqueangledeposition[9]. Innanosphere lithography,theSERS-activesubstrates are obtained by coating a self-assembled monolayer of polystyrene nanoparticles with metal evaporation. In oblique angle deposition, the sample is heldat an angle during metal evaporation. Cylindrical, irregular shaped and randomly distributed nano-rod arrays are obtained due to the surface diffusion and shadowing effect[10].Othercommonlyusedmethodsaretodirectly usecolloidalmetallicsolutionsorusingthemfixedon asubstrate[11].Thecolloidalsolutionsareobtainedby thereduction of dilutedmetallicsaltsolutions.In this method,thedifferences inmetallicparticlessizes and theirshapesduetothedifferencesinpreparationrecipes canleadtoseveralordersofmagnitudedifferenceinthe SERSenhancementfactor.
Electron Beam Lithography overcomes all the aforementioned difficulties. The substrates have high repeatabilityandstability.Theycanbedesignedtowork atdesiredwavelengthsandtheSERSenhancement fac-torcanbecalculatedwiththehelpofsimulationstosave manpowerandmoney.The verificationof the designs
substratepreparation,makeEBLanidealcandidatefor fabrication[12].
Thebestmetricusedinthecomparisonofthe SERS-active surfaces is the SERS enhancement factor. The most commonly used enhancement factor calculation canbewrittenasEF=(ISER/Nsurf)/(IR/Nvol)[13–15].
HereNVol isthe numberof moleculesinthedetection
volume that contributes tothe Raman signal, Nsurf is
the number of moleculesattached on the surface that contributetotheSERsignal.IRandISERarethe
intensi-tiesoftheRamanandSERsignals.Themoleculeswith knownpackingdensitiesarepreferentiallyusedinthese calculations.Thebestknownmoleculeusedforthisaim isBenzenethiol.Twomeasurements,onefromaknown detection volumeof liquid benzenethiol andtheother from benzenethiolapplied onthe nano-structured sur-face are taken andthe mentioned formula isused for the calculationof the SERSenhancementfactor.This enables an objective comparison tobe made between twolithographicsurfaces.Suchacomparisonisnotclear forsurfacespreparedwithothermethodssincethetotal metallicareaand,therefore,Nsurfcannotbedetermined
accurately.
Intheliterature,therearemanySERS-activesurface designspreparedwithEBL[12,16–18].Thesedesigns aremainlyperiodicarraysofsimplenano-structuresand generallytherelationbetweentheLSPresonance wave-lengthandSERsignalenhancementisstudied.Togive severalexamples,LeRu,hastakenSERSmeasurements, fromperiodicgolddot,squareandtrianglearraysfrom “Rhodamine 6G” [19]. In a very similar study, Gun-narsson,studiedsimilarAgstructuresonsiliconwafer for the samemoleculeandreported that betterresults areobtained whencomparedwithnano-roughenedAg film[20].Kahl,hasshownthattheSERSmeasurements of “Rhodamine6G” ongoldperiodic nano-dot arrays andgratingstructuresresultedinanorderofmagnitude betterSERSenhancementwhencomparedwith metal-islandfilmsubstrates[21].Finally,Hatap,showedthat anSERSenhancementof1011hasbeenobtainedfrom free-standinggoldbowtienano-antennas[22].The diver-sityofdesignsisendlesswhenfabricationwithEBLis considered. Below,we summarizethe contributionsto substratedesignwithEBLforSERSbyÖzbaygroup,
Fig.1.(a)physicaldimensionsofasinglenano-cone(b)SEMimageofanarrayofnano-coneswithperiodof200nmandabasediameterof 100nm.
namelytandemstructures,concentricrings,andfractal bowtieantennas.
2. Tandemstructures
Inthisstudy,periodicAu-SiO2-AuNano-conearrays
are fabricatedwithEBLonsapphiresubstrateandare showntoprovidea10timeshigher SERsignal inten-sitywhencomparedwithperiodicAunano-conearrays fabricated with the same method [23].Au nano-cone arraysprovidesingleresonancetunedtothelaser wave-length.Ontheotherhand,Au-SiO2-Aunano-conearrays
asshowninFig.1canprovidedoubleresonancetuned to the laser and the Stokes wavelength, respectively. ThehightunabilityoftheAu-SiO2-Aunano-conearrays
comes from the used fabrication method. The nano-dots written by EBL were then coatedwith goldand SiO2byelectronbeamevaporationmethod.Thedetails
of the fabrication are described in previous reports and, therefore, not repeated here [23]. Several dif-ferent basediameter nano-cones were fabricatedwith 5nm increments in a basediameter successfully that enabledtuningtheresonancebehaviorofthesubstrate inadditiontothemetalheightswhichcanbevaried eas-ilywiththeE-Beamevaporationmethod.Suchhighly tunable and predictable substrates are impossible to obtain with other techniques described earlier in this report.
SinceEBLisahighlyprecisemethodinwhichthe intended andfabricatedstructuresshowagoodmatch inshapeandsize,theirresonancebehaviorcanbe pre-dictedbysimulations.Inthisstudy,thedoubleresonance behavior of the nano-coneswas first shownby simu-lationsandthenverified byopticalmeasurements. By thisway,onecanhaveagoodestimateoftheresonance behavior of the intended nanostructures aheadof fab-rication, andthereby reduce the fabrication costsand time.The simulationsare donebyLumerical software that usesthe Finite DifferenceTimeDomain (FDTD)
method. The simulation and measurement results are showninFig.2fornanoconesthathaveabottomradius of55nm.Thebroadeningandintensityofthepeaks dif-fer dueto the discrepanciesin shape and size of the nanoparticles fabricated and simulated. However, the positions of the resonance dips are in good correla-tion.
TheSERSenhancementfactorcanalsobepredicted by simulations for EBL fabricated structures via the use of 3D electric field monitors. The electric field distribution obtained at excitation and Stokes shifted wavelengths are used to calculate Σ|Eexc|2.|Estokes|2
overtheexposedvolume.Theresultsarethen normal-izedwithrespecttothetotalsurfacearea.Inthisstudy, the results obtained for “tandem” structures are com-paredwith“onlygold”structurestomakeaguessahead of SERS experiments. The resultant enhancement is approx.12fora55nmradiusarrayfortheRamanline at1575cm−1.Theexperimentalvaluesobtainedforthe sameconditionsareapprox.9fora55nmradiusarray whichisingoodaccordancewiththesimulations.
ThecomparisonofSERsignalintensityforanarray of tandem nano-coneswith aperiod of 200nm anda
Fig.2.Comparisonofthetransmissionandsimulationresultsforan arrayofnano-coneswithaperiodof200nmandabasediameterof 110nm.
Fig.3.ComparisonofSERsignalintensityforanarrayoftandem nano-coneswithaperiodof200nmandabasediameterof110nm andits singleresonancecounterpart. Theexcitationwavelengthis 632.8nm.
basediameterof110nmanditssingleresonance coun-terpart are shown in Fig. 3. The SERS enhancement predictedthroughsimulationsandmeasuredviaSERS experiments have shown that the tandem nano-cone designprovided 10 times higherSER signalintensity whencomparedwithperiodicAunano-conearrays fab-ricatedwiththesamemethod.Theenhancementwould be moreprominent if the comparison was made with respecttosimplynano-roughenedmetallicsurfaces,as showninanotherstudyconductedbytheOzbaygroup [24].
3. Concentricringstructures
Concentric arcs and rings are used for various purposes in the literature mainly for their focusing properties [25] or as plasmonic lenses [26]. In our
obtainstructuresbothindentedandprotruding.Toetch theringsfromplaingoldanegativetoneresistwereused inthee-beamlithographyandthenanO2plasmaetch
wasconducted.Thedetailsofthefabricationstepsare describedelsewhere[24].
Fig. 4 depicts the SER spectrum of Benzenethiol fromthe“coupled”ringsandthe“etched”rings.Here, the etched ring design is optimized to maximize the focusing property by choosing its period to be equal to the surface plasmon wavelength.The coupled ring designisfabricatedwiththesamephysicaldimensions for bettercomparison. TheSERSmeasurements show thatthecoupledringdesigncanprovide630times bet-terenhancementthanplaingoldfilmand8timesbetter enhancement than the etched film. This is attributed totheincreased numberof“hotspots” duetothe cou-plingbetweenupperandlowerrings.Theenhancement factor calculated as described in the introduction is 1.67×107 for the coupled-concentric ring structures. TheSERSenhancementfactorsexperimentallyobtained are comparedwiththosecalculatedbyFDTD simula-tionswheretheSERSintensityistakentobeproportional tothe fourthpower of E-fielddistributionatthe laser excitation frequency. The ratio |E|4coupled/|E|4etched is
12 for the simulations which is in accordance with the experimentalratio,eight timeslargersignal inten-sity.
Fig.4.(a)SERSmeasurementstakenfromcoupledringandplaingold(inset:SEMfiguretakenforcoupledringstructure,periodoftherings: 500nm,innerringdiameter:975nm).(b)SERSmeasurementstakenfromcoupledringandplaingold.Periodoftheringsisapproximately615nm andthediameterofthecenterdiscisaround1.45m.(Inset:conceptualfiguresofcoupledringandetchedringstructures)Theexcitationwavelength usedinSERSmeasurementsis632.8nm.
Fig.5.(a)Transmissionmeasurementsversussimulations(inset:SEMfiguresofthefabricatedstructures.)(b)SERSmeasurementsforbowtie nano-antennasandtheirfirstdegreecounterparts(gap=65nm,sidesoftriangles∼400nmforbothfigures).TheexcitationwavelengthusedinSERS measurementsis785nm.
4. Bowtieandfractalbowtiestructures
Bowtie structures offer a wide diversity of tuning mechanismsforenhancedE-fieldgenerationatdesired wavelengths duetotheexcitation of localizedsurface plasmons.Inbowtiestructures,theLSPresonancecan betunedbychangingthespacingbetweenthetwo tri-angle prisms [22,27], the triangle angles [28,29] and sizes[30]toobtain betterSERSenhancementfactors. EBL,offersadvantagessuchasdecreasingthespacing downto5nm [31]andsensitivelytuning thephysical parameters.Obtainingfractalstructuresisevenpossible.
To ourknowledge,the Özbaygrouprecently reported thefabricationoffirstandseconddegreefractalbowtie nano-antennasforthefirsttime[32].
Thedesignofthebowtieandfractalstructureswere donebyFDTDsimulationsandverifiedbytransmission measurements.Thecalculatedandmeasured transmis-sion spectra of the structures are ingood accordance as shown in Fig. 5(a). The resonance wavelength of the structures are in NIR, which enable them to be used in NIR-SERS applications, which are preferred especially to overcome the disadvantages of visible laserssuchasphotochemicalreactions,backgroundfrom
Theelectricfielddistributionisalsoobtainedthrough simulations at both the excitation and Stokes wave-lengthsasshowninFig.6.TheintensifiedE-fieldatboth wavelengthsatthetipofthetrianglesandthegapsleadto appreciableSERSenhancement.TheincreaseinSERS signalintensityforfractalnano-antennasisattributedto thegeneratedhotspotsinthecavitiesofthefractal struc-tureasaresultoftheplasmoniccouplingbetweenthe subwavelengthfeatures.TheSERSenhancementfactor isestimatedbycalculatingΣ|Eexc|2.|Estokes|2overthe
goldareas.Thesimulationsshowedthattheintegrated E-fieldis24timeslargerforthefirstdegreefractalstructure whencomparedwiththebowtiedesign.
5. Summary
ThehighsensitivityandtunabilityobtainedwithEBL canbeused toreducethe time,effort,andcostinthe designphaseoftheSERSsubstrateswiththehelpof sim-ulations.Throughsimulations,thephysicaldimensions required to obtain the desired resonance wavelengths canbestudiedbeforethefabricationstarts.TheSERS enhancement can even be predicted through calcula-tionsusing electric fielddistributions observedby3D monitors. This approach can be applied to a variety ofgeometriessuchasnano-cones,concentricringsand bowtienano-antennas.
Acknowledgements
ThisworkissupportedbytheprojectsDPT-HAMIT, DPT-FOTON, and NATO-SET-193 and TUBITAK undertheprojectnos.113E331,109A015and109E301. Oneoftheauthors(E.O.)alsoacknowledgespartial sup-portfromtheTurkishAcademyofSciences.
References
[1]D.L.Jeanmaire,R.P.VanDuyne,SurfaceRaman spectroelectro-chemistry:partI.Heterocyclic,aromatic,andaliphaticamines adsorbedontheanodizedsilverelectrode,J.Electroanal.Chem. InterfacialElectrochem.84(1977)1–20.
[2]M.G. Albrecht, J.A. Creighton,Anomalously intense Raman spectraofpyridineatasilverelectrode,J.Am.Chem.Soc.99 (1977)5215–5217.
violet sub-monolayers on electrochemically roughened silver substrates: dodye molecules adsorbpreferentially onhighly SERS-activesites?Chem.Phys.Lett.414(2005)271–275(Oct 14).
[7]M.Green,F.M.Liu,SERSsubstratesfabricatedbyisland lithog-raphy:thesilver/pyridinesystem,J.Phys.Chem.B107(2003) 13015–13021(Nov27).
[8]C.L.Haynes,R.P.VanDuyne,Nanospherelithography:aversatile nanofabricationtoolforstudiesofsize-dependentnanoparticle optics,J.Phys.Chem.B105(2001)5599–5611(Jun21).
[9]S.B.Chaney,S.Shanmukh,R.A.Dluhy,Y.-P.Zhao,Alignedsilver nanorodarraysproducehighsensitivitysurface-enhancedRaman spectroscopysubstrates,Appl.Phys.Lett.87(2005)031908.
[10]T. Karabacak, G.-C.Wang,T.-M. Lu,Physical self-assembly and the nucleation of three-dimensional nanostructures by obliqueangledeposition,J.VacuumSci.Technol.A22(2004) 1778–1784.
[11]R.G. Freeman, K.C.Grabar, K.J. Allison, R.M.Bright, J.A. Davis,A.P.Guthrie,etal.,Self-assembledmetalcolloid mono-layers: an approachto SERSsubstrates, Science267 (1995) 1629–1632.
[12]M.Kahl,E.Voges,S.Kostrewa,C.Viets,W.Hill,Periodically structuredmetallicsubstratesforSERS,Sens.ActuatorsB-Chem. 51(1998)285–291(Aug31).
[13]E.C. Le Ru, E. Blackie,M.Meyer, P.G. Etchegoin, Surface enhanced Raman scattering enhancement factors: a compre-hensive study, J. Phys. Chem. C 111 (2007) 13794–13803 (Sep20).
[14]C.L. Haynes, R.P. Van Duyne, Plasmon-sampled surface-enhancedRamanexcitationspectroscopy,J.Phys.Chem.B107 (2003)7426–7433(Jul31).
[15]A.D.McFarland,M.A.Young,J.A.Dieringer,R.P.VanDuyne, Wavelength-scannedsurface-enhancedRamanexcitation spec-troscopy,J.Phys.Chem.B109(2005)11279–11285(Jun9).
[16]J.Beermann,S.M.Novikov,K.Leosson,S.I.Bozhevolnyi, Sur-faceenhancedRamanimaging:periodicarraysandindividual metalnanoparticles,Opt.Express17(2009)12698–12705(Jul 20).
[17]N.Guillot,H.Shen,B.Fremaux,O.Peron,E.Rinnert,T.Toury, etal.,SurfaceenhancedRamanscatteringoptimizationofgold nanocylinderarrays:influenceofthelocalizedsurfaceplasmon resonanceandexcitationwavelength,Appl.Phys.Lett.97(2010) (Jul12).
[18]J.M.Oran,R.J.Hinde,N.AbuHatab,S.T.Retterer,M.J.Sepaniak, NanofabricatedperiodicarraysofsilverellipticaldiscsasSERS substrates,J.RamanSpectrosc.39(2008)1811–1820(Dec).
[19]E.LeRu,P.Etchegoin,J.Grand,N.Félidj,J.Aubard,G.Lévi, etal.,SurfaceenhancedRamanspectroscopyon nanolithography-preparedsubstrates,Curr.Appl.Phys.8(2008)467–470.
[20]L.Gunnarsson,S.Petronis,B.Kasemo,H.Xu,J.Bjerneld,M. Käll,Optimizingnanofabricatedsubstratesforsurfaceenhanced Ramanscattering,Nanostr.Mat.12(1999)783–788.
[21]M.Kahl,E.Voges,S.Kostrewa,C.Viets,W.Hill,Periodically structuredmetallicsubstratesforSERS,Sens.ActuatorsB:Chem. 51(1998)285–291.
[22]N.A.Hatab,C.-H.Hsueh,A.L.Gaddis,S.T.Retterer,J.-H.Li,G. Eres,etal.,Free-standingopticalgoldbowtienanoantennawith variablegapsizeforenhancedRamanspectroscopy,NanoLett. 10(2010)4952–4955.
[23]N.A.Cinel,S.Butun,G.Ertas,E.Ozbay,Fairychimney’-shaped tandem metamaterials as doubleresonance SERS substrates, Small9(Feb252013)531–537.
[24]N.A.Cinel,S.Cakmakyapan,G.Ertas,E.Ozbay,Concentricring structuresasefficientSERSsubstrates,IEEEJ.Sel.Top.Quantum Electron.19(2013)4601605(May–Jun).
[25]J.M. Steele, Z. Liu,Y.Wang,X.Zhang,Resonant and non-resonant generation and focusing of surface plasmons with circulargratings,Opt.Express14(2006)5664–5670.
[26]W.B.Chen,D.C.Abeysinghe,R.L.Nelson,Q.W.Zhan, Plas-moniclensmadeofmultipleconcentricmetallicringsunder radi-allypolarizedillumination,NanoLett.9(2009)4320–4325(Dec).
[27]D.P. Fromm, A. Sundaramurthy, P.J. Schuck, G. Kino, W. Moerner, Gap-dependent optical coupling of single bowtie nanoantennas resonant in the visible, Nano Lett. 4 (2004) 957–961.
[28]W.Ding,R.Bachelot,S.Kostcheev,P.Royer,R.E.deLamaestre, Surfaceplasmonresonancesinsilverbowtienanoantennaswith variedbowangles,J.Appl.Phys.108(2010)124314.
[29]G.Bi,L.Wang,L.Ling,Y.Yokota,Y.Nishijima,K.Ueno,etal., Opticalpropertiesofgoldnano-bowtiestructures,Opt.Commun. 294(2013)213–217.
[30]L.Wang,X.Xu,Spectralresonanceofnanoscalebowtieapertures invisiblewavelength,Appl.Phys.A89(2007)293–297.
[31]R.Murali,E.Walters,F.Zaman,C.Tabor,W.Huang,M. El-Sayed, et al., Fabrication of bow-tie nano-gap structures by electronbeamlithography,in:NanoScience andTechnology Institute(NSTI)NanotechnologyConferenceandTrade Show TechnicalProceedings,2007,258–260.
[32]S.Cakmakyapan,N.A. Cinel,A.O.Cakmak,E. Ozbay, Vali-dation of electromagnetic field enhancement in near-infrared throughSierpinskifractalnanoantennas,Opt.Express22(2014) 19504–19512.
[33]S.Mahajan,M.Abdelsalam,Y.Suguwara,S.Cintra,A.Russell,J. Baumberg,etal.,Tuningplasmonsonnano-structuredsubstrates forNIR-SERS,Phys.Chem.Chem.Phys.9(2007)104–109.
[34]K.Kneipp,R.R.Dasari,Y.Wang,Near-infraredsurface-enhanced Ramanscattering(NIRSERS)oncolloidalsilverandgold,Appl. Spectrosc.48(1994)951–955.