diameter ratio and their non-dimensional interparticledistance/particle (SERS) of gold/silver scattering efﬁciency nanoislandarrangements A generalized relationship exponential between the surface-enhancedRaman Sensors and Actuators A: Physical

(1)

ContentslistsavailableatScienceDirect

Sensors and Actuators A: Physical

jo u r n al hom e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / s n a

A generalized exponential relationship between the surface-enhanced Raman scattering (SERS) efﬁciency of gold/silver nanoisland

arrangements and their non-dimensional interparticle distance/particle diameter ratio

P. Pal

^a

, A. Bonyár

^b

, M. Veres

^c

, L. Himics

^c

, L. Balázs

^a

, L. Juhász

^d

, I. Csarnovics

^a^,∗

aDepartmentofExperimentalPhysics,UniversityofDebrecen,Debrecen,Hungary

bDepartmentofElectronicsTechnology,BudapestUniversityofTechnologyandEconomics,Budapest,Hungary

cInstituteforSolidStatePhysicsandOptics,WignerResearchCentreforPhysics,Budapest,Hungary

dDepartmentofSolidStatePhysics,UniversityofDebrecen,Debrecen,Hungary

a r t i c l e i n f o

Articlehistory:

Received19April2020

Receivedinrevisedform27May2020 Accepted23July2020

Availableonline25July2020

Keywords:

Goldandsilvernanoislands Surface-enhancedRamanscattering Plasmonics

Photonicdevices Sensors

a b s t r a c t

Theoptimizationofthegeometricalpropertiesofgold/silvernanoparticlearrangementstomaximize theirsurface-enhancedRamanscattering(SERS)efﬁciencyisstudiedinthiswork.Forthispurpose,the metallicnanostructureswerecreatedbythermallyannealinggoldandsilverthinﬁlmlayersdeposited ontoglasssubstrates.TheSERScapabilitiesofthesampleswereevaluatedbymeasuringananalytesolu- tionofbenzophenonewiththreedifferentexcitationlaserwavelengths.Systematicinvestigationswere carriedoutondifferentgoldandsilvernanoislandsamplestodeterminehowtheSERSenhancement dependsonthegeometrical(particlediameter,interparticledistance)andopticalparameters(plasmon wavelength)ofthenanostructures,aswellasonthewavelengthoflaserexcitation.Theimportance ofmatchingtheexcitationwavelengthwiththeresonantplasmonabsorbancepropertiesofthesur- facewasproved.However,itwasalsoshownthattheoptimizationofthegeometricalpropertiesof thenanoislandarrangementsdominatesovertheselectionoftheexcitationwavelength.Ageneralized exponentialrelationshipbetweentheSERSenhancementandthenon-dimensionalinterparticledis- tance/particlediameterratiowasestablished.Optimaltechnologicalparametersforthefabricationof gold/silvernanoislandSERSsubstrateswereproposed.

1. Introduction

Ramanspectroscopy is an efficientvibrational spectroscopic techniquethatallowstheanalysisofthestructureandcomposition ofmaterialsinsolid,liquid,orgasstate[1].Thedisadvantageisthat Ramanscatteringhaslowscatteringefficiency.Surface-enhanced Ramanspectroscopy(SERS)isusedfordecadestoimprovethepro- cesssensitivity[2] TheSERSphenomenonwasfirstobservedin 1974byFleischmannetal.[3],whoobservedunexpectedlylarge Ramansignalswhenstudyingpyridineadsorbedonaroughened silverelectrode[4]

∗ Correspondingauthorat:Department ofExperimentalPhysics,Insituteof Physics,FacultyofScienceandTechnology,UniversityofDebrecen,4026Debrecen Bemsq18/a,Hungary.

E-mailaddress:csarnovics.istvan@science.unideb.hu(I.Csarnovics).

ThemagniﬁcationofRamansignalsbyseveralordersofmagni- tudeismadepossiblebythenanostructuredorroughenedmetal surface[5].SeveralstudieshaveaddressedtheSERSenhancement mechanism.Nowadays,itisacceptedthattwomajorfactorscon- tributetotheenhancementofRamansignals:electromagnetic[6]

andchemical[7,8],fromwhichtheelectromagneticmechanismis consideredtobethekeyone[6,9].SERSincorporatestheessential benefitsofRamanspectroscopysuchasfingerprintidentificationof molecules,non-destructiveanalysis,minimalsamplepreparation;

analysisofbiologicalsamples;thepossibilityofcarryingoutﬁeld analysisusingportabledevices[8],allwithhighsensitivity,which insomecasesevenallowsthedetectionofasinglemolecule[4,10].

Silverandgoldnanostructuresarethemostwidelyusedmate- rialsinSERSsubstratesduetotheirsurfaceplasmonresonance (LSPR)properties,which cover awide wavelengthrange in the visible and near-infrared regions where most Ramanmeasure- mentsaremade[1,11].Severalmethodshavebeendevelopedfor thefabricationofthesesubstrates,whichfollowoneoftwogen-

https://doi.org/10.1016/j.sna.2020.112225

0/).

(2)

SERSsubstrates[26,27] andthedevelopmentand fabricationof substrateswith optimal enhancement properties [4]. The SERS enhancementfactor(SERSEF)ofasubstratedependsonseveral factors,suchasthegeometrical propertiesof thenanoparticles (size,shape,interparticledistance,andgeneralarrangement)orthe wavelengthoftheexcitationsource[28].Thesefactorsareinter- twinedinanontrivialway,formingacomplexsystem,whichis essentialtobeoptimizedforthebestSERSperformance.

TheeffectofnanoparticlesizeontheSERSperformancewas previouslystudied[29–32]withmixedconclusions.Itwasshown thatthehighestSERSEFcouldbereachedwithsphericalgoldand silvernanoparticleswithadiameterof50nm[29,31],whileoth- ersreported a broaderrange of optimalparticlesizesbetween 30–100nmfortheimprovementofSERS[30,32].Generally,when theparticlesaretoosmall,boththeactualconductivityandthe scatteringpropertiesarereduced,whichexpectedlyreducesthe SERSefﬁciency[30].Astheparticlesizeincreases,theSERSeffect increasessincetheyactaslargerscatteringcenters.However,in thesepapers,therewasnostudyconcerningtheinterparticledis- tance.

Besides,otherworks studiedthisparameter, asinterparticle distancealsoaffects theSERSenhancement.In previousworks, a relationship betweenthe interparticle distanceand theSERS enhancementwasdemonstrated[33,34].Theclosertheparticles aretoeachother,thehigherthespecificsurfacecoverageoftheir activesurfacesis, andthesurfacedensityofhot-spots(number ofareaswithincreasednearfieldintensitiesduetocoupledplas- monsbetweentheparticles)isalsohigher,whichaltogetherresults inanimprovedSERSefficiencyasit wasshowninourprevious work[35].Atveryshortdistances(e.g.,lessthan1nm),quantum mechanicalphenomenaplayarolethatlimitsthepossibleachiev- ablefieldstrengths[36].Mostpapersonlystudytheinfluenceof interparticledistanceand donottakeintoaccounttheeffectof thenanostructures’size,exceptforafew[35,37–39].Itcouldbe concludedthatseveralpastworksinvestigatedtheeffectofthese parametersindependently;theircomplexrelationshipisnotyet fullydescribed.

Theshape ofthenanoparticlescanalsobeafactor,itseffect onthe SERSenhancement was investigated by others [40–42].

Accordingtothese,anisotropicmetallicnanoparticleswithcom- plexshapes(e.g.,hexagonal,triangular,cubical,starshape),could bebeneﬁcialduetothemoreintensenear-ﬁeldsatthesharpedges.

However,thecomplicatedsynthesisofsomeoftheseparticlesand theirsubsequentsurfacechemistryarethemaindrawbackoftheir currentapplication.

Anothergeneralconsiderationis that thewavelength ofthe excitationsource shouldmatchthe plasmonicproperties (LSPR excitationrange)ofthesubstrateascloseaspossible.TheLSPR absorbance wavelength shifts to red with larger nanoparticle size and smaller interparticle distance (due to the plasmonic coupling) [43,44]. If the excitation wavelength is far from the plasmonexcitationofthesubstrate,theSERSenhancementcan

applicationofSERSforthedetectionofothermaterialssystems, somerecentreviewscanberecommended[46,47].

2. Materialsandmethods 2.1. PreparationofSERSsubstrates

ThesubstratesfortheSERSactivenanoparticleswerecutfrom microscopeslideswithahandheldglasscuttertoformglassplates.

TheywerecleanedinanEMMI-20HCultrasonicbathin96%ethanol andthenwipeddrywithasterilepaper.Theamountofmaterialto beevaporatedwasmeasuredonaSartoriusMicroM3psemi-micro analyticalbalance.Thethinmetallicfilmswithcontrolledthick- nesseswerepreparedbythermalevaporation.Thelayerthickness wasmeasuredbyusinganAmbiosXP-1profilometer.Thethick- nessesofthecreatedgoldlayerswere9and12nm,whileforthe silverlayer15 and25nm. Sampleswerethen heat-treatedin a quartzglasstubeplacedinanovenfilledwithAr:Hprecursorgas.

Theheattreatment(solid-statedewetting)tookplaceatdifferent temperatures(350^◦C,450^◦C,and550^◦C)fordifferentperiods(15, 30,60and120min),resultinginavarietyofgeometricalparam- eters (e.g.,particlediameter, interparticledistance). During the treatment,thehightemperatureinducesdiffusiononthesurface ofthesubstrate,causingthethinﬁlmﬁrsttoburst,thentoform nanoscaleislands.

2.2. CharacterizationofSERSsubstrates

Thefreshly preparedmetallicnanoislandswereexaminedby using a scanning electron microscope(SEM) and then an opti- calspectrophotometer.SEMimageswererecordedwithaHitachi S4300-CFEinstrument.Multiplelocationsandmagniﬁcationswere takenoneachsample.Theimageswereevaluatedbyusingthe National Instruments VisionAssistant software package, which determinedthemeannanoislanddiameter(deﬁnedastheequiv- alentdiameterofacirclehavingthesameprojectedareaasthe nanoisland),measurementuncertainty,anddiameterdistribution.

Tofurtherevaluatethediameterdistribution,weusedtheOrig- inPro9softwareandacustom-writtenMatlabscripttodetermine theaverageinterparticledistanceanditsstandarddeviation.The measurementuncertaintyofthediameteroftheproducednanos- tructuresandtheinterparticledistancewas10–15%,depending on the sample. The next step was to investigate the plasmon wavelengthofthecreatedmetallicnanostructures.Measurements weremadewithan OceanOpticsRedTideUSB650optical ﬁber spectrophotometer.Theopticaltransmittanceofthesampleswas measuredinair.

Figs.1and2illustratesSEMimages(withanaccelerationvoltage of5kVandmagniﬁcationof×20000or×30000)ofgoldandsilver nanoislandsandtheirtransmittancespectra.Thediameter,interparticledistance,andplasmonwavelengthscouldbecontrolledby

(3)

Fig.1. Illustrationofthepreparedgoldnanoislands:a)SEMimageofgoldsample1;b)SEMimageofgoldsample2;c)Thecorrespondingopticaltransmittancespectraof thesamples.ThetechnologicalparametersaregiveninTable1.ThemagniﬁcationofSEMimageswas×20000.

Table1

Parametersofthetwodifferentgoldsamples.

AuSample1 AuSample2

Initiallayerthickness 9nm 12nm

Annealingtemperature 550^◦C 550^◦C

Annealingtime 15min 15min

Table2

Parametersofthetwodifferentsilversamples.

AgSample1 AgSample2

Initiallayerthickness 15nm 25nm

Annealingtemperature 350^◦C 350^◦C

Annealingtime 60min 60min

theinitialthicknesswhilekeepingtheannealingtemperatureand timeﬁxed.Fig.1and2illustratetwodifferentgold/silversamples withthefabricationparametersgiveninTables1and2,respectively.Theseﬁguresshowhowtheinitiallayerthicknessaffects

thediameteroftheformednanoislandsandtheinterparticledis- tanceatthesameannealingtemperatureandtime.Bystartingfrom athickerlayer,largernanoislandsareformedfartherapart,whilea thinnerﬁlmproducessmallernanoislandsthatareclosertogether.

Also,thedifferentgoldandsilverpatternsclearlyshowhowthe transmittanceofthesamplevarieswithdiameter,affectedbythe initiallayerthickness.

2.3. SamplepreparationforSERSmeasurements

TheanalytesolutionusedfortheRamanmeasurementswas 50mMbenzophenone,dissolvedinisopropylalcohol,andallSERS measurements wereperformedwitha solution takenfrom the same batch. During thepreparation, special care wasgiven to dissolveallthebenzophenonecrystalsandtohomogenizethesolu- tion.

Spincoaterwasusedtoplacethebenzophenonesolutionon theSERSsubstrates,which allowscreatinga uniformthin layer of theanalyte.The substrateswereplaced in thecenter of the

(4)

Fig.2.Illustrationofthepreparedsilvernanoislands:a)SEMimageofsilversample1;b)SEMimageofsilversample2;c)Thecorrespondingopticaltransmittancespectra ofthesamples.ThetechnologicalparametersaregiveninTable2.ThemagniﬁcationofSEMimageswas×30000.

device,whichwasﬁxedwithdouble-sidedadhesive.Then,using amicropipette,20␮Lofthesamplesolutionwasdroppedintothe centeroftheactivesurface,andthespincoaterwasstartedand rotatedat1000rpmfor1min.Thecentrifugalforceoccurringdur- inghigh-speedrotationuniformlydistributesthesolutiononthe substratesurface.

2.4. SERSmeasurements

ARenishawinVia, a Renishaw1000B,and a HoribaLabRam RamanspectrometerwereusedforRamanmeasurementsonthe samples.Duringthework,488,532,and633nmlaserswereusedas excitationsources,andthemeasurementtimeforeachsamplewas 10s.Theexcitationbeamwasfocusedontothesamplesurfacewith a50xlens.Ramanmeasurementswereperformedinseveralpoints onthesubstratesurface(5–7points),andthemeasurementhas beenrepeatedinthepointwheretheSERSenhancementproved tobethebest.ForthereferenceRamanspectrum,theanalytewas

placedonabareglassslideunderthesameconditionsandinthe sameamount.

Thecharacteristicspectraofbenzophenonewereobtainedin ourmeasurementsonapureglasssubstrateandmetallicnanopar- ticlesaswell.Themainbandsofbenzophenonecanbeobservedon bothwithdifferentintensities(seeFig.3).Basedonthese,theSERS enhancementwascalculatedfromtheintensityofthe1596cm⁻¹ peak,whichcorrespondstothevibrationoftheC Cbondsofthe phenylring[48].

Inthiswork,wecalculatedtheSERSenhancementbasedonthe enhancementfactor(EF)[9],whichcanbedeﬁnedasEq.(1):

EF=ISERS

⁄

N_SERS IRS

⁄

NRS

(1)

Inthisapproach,I_RSistheRamansignalundernon-SERScon- ditions(e.g.,onthesurface ofthereferencesample).Underthe sameexperimentalconditionsand thesamepreparationcondi- tions,theSERSsignalmeasuredontheSERSsubstratewillhave theintensityofISERS[9,30].Atthesametime,theNRS–istheaver-

(5)

Fig.3. ThemeasuredRamanandSERSspectraofthebenzophenone.

agenumberofmoleculesinthescatteringvolumefortheRaman (non-SERS)measurement,andNSERS–istheaveragenumberof adsorbedmoleculesinthescatteringvolumefortheSERSexper- iments.ThisEFcalculationapproachignoresthefactthatSERSis asurface-sensitivephenomenon:onlysurface-adsorbedmolecules contributetothehighestEMenhancement,andthesignalcontribu- tiondecaysexponentiallywithincreasingdistancetothesurface.

Sinceinourexperiments,thesamplewaspresentonlyintheform ofathinspin-coatedlayer(forbothSERSandreferencemeasure- ments)weassumedthatN_SERS∼=N_RS,thusEFcanbesimpliﬁedas theratioofthemeasuredintensities.

3. Resultsanddiscussion 3.1. Goldsubstrates

Forbothgoldandsilvernanoislands,measurementsweremade withthreelasersofdifferentwavelengths.Beforethat,theiroptical parameters,diameter,andinterparticledistanceweremeasured, andtheEFwasevaluatedforeachexcitationwavelength.

Asmentionedintheintroduction,therelationshipbetweenthe plasmon wavelength ofthe substrateand the excitation wave- lengthaffectstheSERSenhancement.Thethreedifferentexcitation wavelengthsinourstudyaremeanttoprobethisrelationship.The plasmonresonanceofnanoislandscanbeeffectivelytunedwiththe particlediameter,structure,andinterparticledistance[33,38,39].

Fig.5.Surface-enhancedRamanspectraofbenzophenonemeasuredonagold nanoislandsubstratewiththreedifferentexcitationwavelengths.

Fig.4aillustratestheeffectofthenanoislanddiameterontheplas- monwavelength.Themeasureddiameter,interparticledistance and plasmonwavelength ofthethree nanoislandarrangements shownin theﬁgurearethefollowing:sample1:50nm,45nm, 514nm,sample2:54nm,46nm,527nm,sample3:59nm,37nm, 561nm,respectively.Itcanbeseenthatbyincreasingthediam- eterofthegoldnanoislands,thepositionoftheplasmonpeakis shiftingtohigherwavelengths.Therelationshipbetweentheplas- monabsorbancepropertiesandtheexcitationwavelengthscanbe studiedbasedonFig.4b.Sincetheplasmonresonanceofthegold nanoislandsisinthe500−600nmrange,itcanbeexpectedthatthe laserwith532nmexcitationwavelengthwillyieldthehighestSERS enhancement.Fig.4bpartiallyconﬁrmthis,since,ingeneral,the 532nmexcitationresultedinthehighestaverageEF.Fig.5presents sampleSurface-enhancedRamanspectrameasuredwiththethree differentexcitationsources,whichalsoillustratethatthehighest enhancementwasobtainedbyusing532nmexcitation.Itshouldbe notedthattheRamanscatteringwillbenon-resonantforallthese excitationwavelengths.Theanalysisoftheabsorptionspectrumof benzophenone[49]showsthattheabsorptionbandsofthecom- poundareintheUVregion,wellbelowourlowestlaserwavelength of488nm.

However,there are somepeculiaritiesworth mentioning. In Fig.4b,thepresenteddataaregivenasafunctionofthenormalized absorbancemeasuredatthegivenexcitationwavelength.Ahigher

Fig.4.a)Normalizedabsorbancespectraillustratingtheeffectofdifferentgoldnanoislanddiametersontheplasmonicabsorbanceofthesubstrates(averagediameters:

#1:50nm,#2:54nm,#3:75nm);b)Dependenceoftheenhancementfactorontheexcitationwavelengthinthefunctionofthenormalizedabsorbanceatthatwavelength.

Forthis,thenormalizedabsorbanceofthesamples(asinFig.4a)wereevaluatedattheexcitationlaser’swavelength.

(6)

Fig.6. Enhancementfactorsobtainedfor532nmexcitation,inthefunctionofa)theaveragegoldnanoislanddiameter(D);b)theaverageinterparticledistance(r);andc) theinterparticledistance/nanoislanddiameterratio(r/D).

normalizedabsorbancemeansthatthenanoparticlearrangement hasastrongerplasmonicresonanceatthegivenwavelength.The factthattheEFdoesnot correlatedirectlywiththenormalized absorbance indicates that the excitation wavelength is impor- tantnotonlybecauseofitsrelationtoplasmonexcitation.E.g., insomeofthecases,a633nmexcitationwithsmallnormalized absorbance(around0.1,inotherwordsfarfromtheplasmonres- onancepeak) resulted in betterenhancement than the532nm excitationwithhighnormalizedabsorbance(around1,e.g.,inthe plasmonresonancepeak).Also,the488nmexcitationresultedin thesmallestaverageEFregardlessofthenormalizedabsorbance atthiswavelength.Theseobservationsindicatethatotherfactors determineorinﬂuencetheeffectofexcitationwavelengthonthe SERSEFbesidesitsrelationtotheplasmonicpropertiesofthesub- strate.

Toevaluatetheeffect of thegeometricalparameters onthe measuredSERS enhancementfactorsin moredetail, theresult- ingEFsobtainedwiththe532nmlaser excitationwere plotted asa functionofboth theaveragenanoislanddiameter(Fig.6a) and the average interparticle distance (Fig. 6b). As expected – accordingtothediscussionsintheintroduction–themeasured EFshowsastrongpositivelinearcorrelationwiththeparticlesize andalso decaysexponentiallywithincreasing interparticledistance.Thevariationofthedatadecreasessigniﬁcantlywhenthe obtainedEFs are plotted as a function of theinterparticle distance/nanoisland diameter ratio (r/D). The enhancement factor showsaclearexponentialrelationshipwiththisnon-dimensional parameter(equationgiveninFig.6c,inwhichtheIDistheinterpar-

ticledistance/nanoislanddiameterratio),whichmakesthedesign andoptimizationoftheSERSsubstratessimpler.

Based on the results, the highest SERS enhancement (∼6.5) wasachieved withgoldnanoislands preparedwiththe follow- ingfabricationparameters:initiallayerthickness:9nm,annealing temperature:500^◦C,annealingtime:15min.TheSERSsubstrate fabricatedwiththese conditions hasthefollowing parameters:

plasmonwavelengthpeak:528nm,particlediameter:58nm,interparticledistance:30nm.

3.2. Silversubstrates

Thesilver-basednanoislandSERSsubstrateswereinvestigated withthesamemethodology.Fig.7ashowsthethreetypicalspectra withthefollowinggeometricalandplasmonicproperties(diameter,interparticledistance,andplasmonwavelength):sample1:

34nm, 41nm, 422nm, sample 2: 37nm, 36nm, 436nm, sam- ple3:39nm,31nm,445nm.For thesilver nanoislandsamples, theplasmonabsorbancepeaksarebetween400−460nm,thus,as expected,thehighestaverageenhancementfactorswereachieved withthe488nm laserexcitation (seeFig. 7b).Thisisalso con- ﬁrmedbythesampleSurface-enhancedRamanspectrapresented inFig.8.However,similarlytowhatwesawinthecaseofthegold nanoislands,themeasuredEFdoesnotcorrelateclearlywiththe normalizedabsorbance,regardlessof theexcitationwavelength (Fig.7b).

Asinthecaseofgoldnanoislands,theeffectofthegeomet- ricalparameterswasstudiedindetail.ThechartsofFig.9show

(7)

Fig.7. a)Normalizedabsorbancespectraillustratingtheeffectofdifferentsilvernanoislanddiametersontheplasmonicabsorbanceofthesubstrates(averagediameters:

#1:34nm,#2:37nm,#3:39nm);b)Dependenceoftheenhancementfactorontheexcitationwavelengthinthefunctionofthenormalizedabsorbanceatthatwavelength.

Forthis,thenormalizedabsorbanceofthesamples(asinFig.7a)wereevaluatedattheexcitationlaser’swavelength.

Fig.8.Surface-enhancedRamanspectraofbenzophenonemeasuredonasilver nanoislandsubstratewiththreedifferentexcitationwavelengths.

thesamelinearandexponentialtendenciesbetweenthemeasured EFandthenanoislanddiameterandinterparticledistance,respectively(forthedataobtainedwiththe488nmexcitation).TheEFalso showsaclearexponentialrelationshipwiththenon-dimensional (r/D)parameter(equationinFig.9c,whereIDistheinterparticle distance/nanoislanddiameterratio).Ithastobenotedthatsilver nanoislandsshowahigher respectiveSERSenhancementinthe functionofr/D(bycomparingtheequationsinFigs.6cand9c).

ThehighestSERSenhancement(∼27)wasachievedwithsilver nanoislands prepared with the following fabrication parameters:initiallayerthickness:25nm,annealingtemperature:350^◦C, annealingtime:60min.TheSERSsubstratefabricatedwiththese conditions has the following parameters: plasmon wavelength peak: 438nm, particle diameter: 61nm, interparticle distance:

22nm.

4. Conclusions

Bycomparingtheresultsobtainedonthegoldandsilvernanois- landarrangements,wecandrawthefollowingconclusions:

1)ThehighestaverageandmaximumobtainedSERSanalytical enhancementfactorswereachievedwiththe532nmlaserirradia- tionforgoldnanoislandsandwiththe488nmexcitationforsilver nanoislands.Thisprovesthattherelationshipbetweentheexci- tationwavelengthandtheplasmonabsorbancepropertiesofthe

surfaceisimportantandshouldbeconsideredwhenselectinga SERSsubstrate.

2)Ourﬁndingspartiallycontradictthefactthatforbothofthe cases,theobtainedEFsdidnotshowaclearcorrelationwiththe normalizedabsorbancepeaksconsideringallinvestigatedexcita- tionwavelengths.Thissuggeststhatotherfactors–suchasthe geometricalpropertiesofthenanoislandarrangements–dominate overtheexcitationwavelengthtoareasonableextent.

3)TheSERSEF showeda positivelinear correlationwiththe nanoparticle size and a negative exponential relation with the interparticledistanceforbothtypesofnanomaterials.Thiscanbe attributedtothehigherscatteringefﬁciency(biggerparticlesize) andthehigherdensityofnear-ﬁeldhotspotsonthesurfaceinthe caseoftightlypackedparticles.

4)A precise exponential relationship could be established betweenthemeasuredEFandthenon-dimensionalr/Dparame- ter,whichistheratiooftheinterparticledistanceandtheparticle diameter,forbothmaterialtypes.

5)Silvernanoislands(measuredat488nm)showedasteeper exponentialincreaseintheEFwithdecreasingr/Dcomparedtothe goldnanoislands(measuredat532nmexcitation).

6)Theseresultssuggestthatlargerandcloserpackedmetallic nanostructureswillresultinbetterSERSenhancement.Inthecase ofgold,thiscanbeachievedwithathinnerstartinglayer(6nm), lowerannealingtemperature(450^◦C)andlongerannealingtime (30−60min),orathickerlayer(9nm),highertemperature(500^◦C), andshortertime(15min).Inthecaseofsilver,largeranddenser nanoislandswereachievedwithathickerstartinglayer(25nm), lowannealingtemperature(350^◦C),andlongertime(60−120min).

7)Thesesetsoftechnologicalparametersresultinsmalleraver- ager/DparametersforthesilvernanoislandsandthushigherSERS enhancementfactors,comparedtothegoldnanoislands.

DeclarationofCompetingInterest

Pleasecheckthefollowingasappropriate:

•Allauthorshaveparticipated in(a)conceptionand design,or analysisand interpretationofthedata;(b)draftingthearticle orrevisingitcriticallyforimportantintellectualcontent;and(c) approvaloftheﬁnalversion.

•Thismanuscripthasnotbeensubmittedto,norisunderreview at,anotherjournalorotherpublishingvenue.

(8)

Fig.9.Enhancementfactorsobtainedfor488nmexcitation,inthefunctionofa)theaveragesilvernanoislanddiameter(D);b)theaverageinterparticledistance(r);andc) theinterparticledistance/nanoislanddiameterratio(r/D).

•Theauthorshavenoafﬁliationwithanyorganizationwithadirect orindirectﬁnancialinterestinthesubjectmatterdiscussedinthe manuscript

•Thefollowingauthorshaveafﬁliationswithorganizationswith directorindirectﬁnancialinterestinthesubjectmatterdiscussed inthemanuscript:

DeclarationofCompetingInterest

Theauthorsreportnodeclarationsofinterest.

Acknowledgments

ThisworkwasfinanciallysupportedbythegrantGINOP-2.3.2- 15-2016-00041.Theprojectisco-financedbytheEuropeanUnion and the EuropeanRegional Development Fund. This workwas supportedbytheVEKOP-2.3.2-16-2016-00011grant,whichisco- financedbytheEuropeanUnionand EuropeanSocialFund.The researchreported inthis paperwas partiallysupported bythe HigherEducationExcellenceProgramoftheMinistryofHuman CapacitiesintheframeofNanotechnologyandMaterialsScience (BMEFIKP-NAT)andalsoBiotechnology(BME-FIKP-BIO)research areasofBudapestUniversityofTechnologyandEconomics.Istvan CsarnovicsisgratefulforthesupportoftheJánosBólyaiResearch ScholarshipoftheHungarianAcademyofSciences.Thesupport

throughtheNewNationalExcellenceProgramoftheMinistryof HumanCapacitiesisacknowledgedaswell.

AppendixA. Supplementarydata

Supplementarymaterial relatedto this articlecanbe found, in the online version, at doi:https://doi.org/10.1016/j.sna.2020.

112225.

References

[1]L.T.Hoang,H.V.Pham,M.T.T.Nguyen,Investigationofthefactorsinﬂuencing thesurface-enhancedramanscatteringactivityofsilvernanoparticles,J.

Electron.Mater.(2019)1–8,http://dx.doi.org/10.1007/s11664-019-07870-8.

[2]C.L.Haynes,C.R.Yonzom,X.Zang,R.P.VanDuyne,Surface-enhancedRaman sensors:earlyhistoryandthedevelopmentofsensorsforquantitative biowarfareagentandglucosedetection,J.RamanSpectrosc.36(2005) 471–484,http://dx.doi.org/10.1002/jrs.1376.

[3]M.Fleischmann,P.J.Hendra,A.J.McQuillan,Ramanspectraofpyridine adsorbedatasilverelectrode,ChemR.Phys.Lett.26(1974)163–166,http://

dx.doi.org/10.1016/0009-2614(74)85388-1.

[4]R.Pilot,R.Signorini,C.Durante,L.Orian,M.Bhamidipati,L.Fabris,Areview onsurface-enhancedramanscattering,Biosensors.9(2019)1–100,http://dx.

doi.org/10.3390/bios9020057.

[5]J.Choi,J.-H.Kim,J.-W.Oh,J.-M.Nam,Surface-enhancedraman scattering-baseddetectionofhazardouschemicalsinvariousphasesand matriceswithplasmonicnanostructures,Nanoscale.11(2019)20379–20391, http://dx.doi.org/10.1039/c9nr07439b.

[6]S.-Y.Ding,E.-M.You,Z.-Q.Tian,M.Moskovits,Electromagnetictheoriesof surface-enhancedRamanspectroscopy,Chem.Soc.Rev.46(2017) 4042–4076,http://dx.doi.org/10.1039/c7cs00238f.

(9)

[7]A.Otto,The“chemical”(electronic)contributiontosurface-enhancedRaman scattering,J.RamanSpectrosc.36(2005)497–509,http://dx.doi.org/10.1002/

jrs.1355.

[8]A.Otto,ChargetransferinﬁrstlayerenhancedRamanscatteringandsurface resistance,Q.Phys.Rev.3(2017)1–14.

[9]H.K.Lee,Y.H.Lee,C.S.L.Koh,G.C.Phan-Quang,X.Han,C.L.Lay,X.Y.Ling,H.Y.F.

Sim,Y.-C.Kao,Q.An,Designingsurface-enhancedRamanscattering(SERS) platformsbeyondhot-spotengineering:emergingopportunitiesinanalyte manipulationsandhybridmaterials,Chem.Soc.Rev.48(2019)731–756, http://dx.doi.org/10.1039/c7cs00786h.

[10]H.Chu,S.Song,C.Li,D.Gibson,Surfaceenhancedramanscatteringsubstrates madebyobliqueangledeposition:methodsandapplications,Coatings7 (2017)1–23,http://dx.doi.org/10.3390/coatings7020026.

[11]P.A.Mosier-Boss,ReviewofSERSsubstratesforchemicalsensing, Nanomaterials.7(2017)142.

[12]S.-Y.Ding,X.-M.Zhang,B.Ren,Z.-Q.Tian,Surface-enhancedraman spectroscopy(SERS):generalintroduction,EncyclopediaofAnalytical Chemistry(2014)1–34,http://dx.doi.org/10.1002/9780470027318.a9276.

[13]Z.-Q.Tian,B.Ren,Infraredandramanspectroscopyinanalysisofsurfaces, EncyclopediaofAnalyticalChemistry(2006)1–40,http://dx.doi.org/10.1002/

9780470027318.a2516.

[14]J.Theiss,P.Pavaskar,P.M.Echternach,R.E.Muller,S.B.Cronin,Plasmonic nanoparticlearrayswithnanometerseparationforhigh-performanceSERS substrates,NanoLett.10(2010)2749–2754,http://dx.doi.org/10.1021/

nl904170g.

[15]X.Y.Zhang,J.Zhao,A.V.Whitney,J.W.Elam,R.P.VanDuyne,Ultrastable substratesforsurfaceenhancedRamanspectroscopy:Al2O3overlayers fabricatedbyatomiclayerdepositionyieldimprovedanthraxbiomarker detection,J.Am.Chem.Soc.128(2006)10304–10309,http://dx.doi.org/10.

1021/ja0638760.

[16]J.T.Bahns,A.Imre,V.K.Vlasko-Vlasov,J.Pearson,J.M.Hiller,L.H.Chen,U.

Welp,EnhancedRamanscatteringfromfocusedsurfaceplasmons,Appl.Phys.

Lett.91(2007),081104,http://dx.doi.org/10.1063/1.2759985.

[17]J.N.Anker,W.P.Hall,O.Lyandres,N.C.Shah,J.Zhao,R.P.VanDuyne, Biosensingwithplasmonicnanosensors,Nat.Mater.7(2008)442–453, http://dx.doi.org/10.1038/nmat2162.

[18]K.C.Grabar,R.G.Freeman,M.B.Hommer,M.J.Natan,Preparationand characterizationofAucolloidmonolayers,Anal.Chem.67(1995)735–743, http://dx.doi.org/10.1021/ac00100a008.

[19]A.Wei,B.Kim,B.Sadtler,S.L.Tripp,Tunablesurface-enhancedRaman scatteringfromlargegoldnanoparticlearrays,Chemphyschem.12(2001) 743–745,

doi:10.1002/1439-7641(20011217)2:12<743::aid-cphc743>3.0.co;2-1.

[20]P.Z.EI-Khoury,E.Khon,Y.Gong,A.G.Joly,P.Abellan,J.E.Evans,N.D.

Browning,D.HuM.Zamkov,Electricﬁeldenhancementinaself-assembled 2Darrayofsilvernanospheres,J.Chem.Phys.141(2014)214308,http://dx.

doi.org/10.1063/1.4902905.

[21]H.Chen,Z.H.Sun,W.H.Ni,K.C.Woo,H.Q.Lin,L.D.Sun,C.H.Yan,J.F.Wang, Plasmoncouplinginclusterscomposedoftwo-dimensionallyorderedgold nanocubes,Small.5(2009)2111–2119,http://dx.doi.org/10.1002/smll.

200900256.

[22]M.Rycenga,C.M.Cobley,J.Zeng,W.Y.Li,C.H.Moran,Q.Zhang,D.Qin,Y.N.

Xia,Controllingthesynthesisandassemblyofsilvernanostructuresfor plasmonicapplications,Chem.Rev.111(2011)3669–3712,http://dx.doi.org/

10.1021/cr100275d.

[23]J.Kumar,K.G.Thomas,Surface-enhancedRamanspectroscopy:investigations atthenanorodedgesanddimerjunctions,J.Phys.Chem.Lett.2(2011) 610–615,http://dx.doi.org/10.1021/jz2000613.

[24]L.Zhang,C.Guan,Y.Wang,J.Liao,HighlyeffectiveanduniformSERS substratesfabricatedbyetchingmulti-layeredgoldnanoparticlearrays, Nanoscale.8(2016)5928–5937,http://dx.doi.org/10.1039/c6nr00502k.

[25]N.D.Israelsen,C.Hanson,E.Vargis,Nanoparticlepropertiesandsynthesis effectsonsurface-enhancedramanscatteringenhancementfactor:an introduction,TheScientiﬁcWorldJournal.(2015)1–12,http://dx.doi.org/10.

1155/2015/124582.

[26]S.L.Kleinman,B.Sharma,M.G.Blaber,A.I.Henry,N.Valley,R.G.Freeman,M.J.

Natan,G.C.Schatz,R.P.VanDuyne,Structureenhancementfactor relationshipsinsinglegoldnanoantennasbysurface-enhancedraman excitationspectroscopy,J.Am.Chem.Soc.135(2013)301–308,http://dx.doi.

org/10.1021/ja309300d.

[27]E.C.LeRu,J.Grand,N.Félidj,J.Aubard,G.Lévi,A.Hohenau,J.R.Krenn,E.

Blackie,P.G.Etchegoin,ExperimentalveriﬁcationoftheSERSelectromagnetic modelbeyondtheE4approximation:polarizationeffects,J.Phys.Chem.C Lett.112(2008)8117–8121,http://dx.doi.org/10.1021/jp802219c.

[28]M.Fan,G.F.S.Andrade,A.G.Brolo,Areviewonthefabricationofsubstratesfor surfaceenhancedRamanspectroscopyandtheirapplicationsinanalytical chemistry,Anal.Chim.Acta693(2011)7–25,http://dx.doi.org/10.1016/j.aca.

2011.03.002.

[29]S.Hong,X.Li,Optimalsizeofgoldnanoparticlesforsurface-enhancedRaman spectroscopyunderdifferentconditions,J.Nanomater.(2013)1–9,http://dx.

doi.org/10.1155/2013/790323.

[30]M.Moskovits,Surface-enhancedRamanspectroscopy:abriefretrospective,J.

RamanSpectrosc.36(2005)485–496,http://dx.doi.org/10.1002/jrs.1362.

[31]K.G.Stamplecoskie,J.C.Scaiano,V.S.Tiwari,H.Anis,Optimalsizeofsilver nanoparticlesforsurface-enhancedramanspectroscopy,J.Phys.Chem.C115 (2011)1403–1409,http://dx.doi.org/10.1021/jp106666t.

[32]P.N.Njoki,I.-I.S.Lim,D.Mott,etal.,Sizecorrelationofopticaland spectroscopicpropertiesforgoldnanoparticles,J.Phys.Chem.C111(2007) 14664–14669,http://dx.doi.org/10.1021/jp074902z.

[33]S.S.Masango,R.A.Hackler,N.Large,A.I.Henry,M.O.McAnally,G.C.Schatz, P.C.Stair,R.P.VanDuyne,High-resolutiondistancedependencestudyof surface-enhancedramanscatteringenabledbyatomiclayerdeposition,Nano Lett.16(2016)4251–4259,http://dx.doi.org/10.1021/acs.nanolett.6b01276.

[34]V.T.N.Linh,J.Moon,C.W.Mun,V.Devaraj,J.-W.Oh,S.-Gy.Park,D.-H.Kim,J.

Choo,Y.-I.Lee,H.S.Jung,Afacilelow-costpaper-basedSERSsubstratefor label-freemoleculardetection,Sens.ActuatorsBChem.291(2019)369–377, http://dx.doi.org/10.1016/j.snb.2019.04.077.

[35]A.Bonyár,I.Csarnovics,M.Veres,L.Himics,A.Csik,J.Kámán,B.Balázs,S.

Kökényesi,Investigationoftheperformanceofthermallygeneratedgold nanoislandsforLSPRandSERSapplications,Sens.ActuatorsBChem.255 (2018)433–439,http://dx.doi.org/10.1016/j.snb.2017.08.063.

[36]Y.Huang,Q.Zhou,M.Hou,L.Ma,Z.Zhang,Nanogapeffectsonnear-and far-ﬁeldplasmonicbehaviorsofmetallicnanoparticledimers,Phys.Chem.

Chem.Phys.17(2015)29293–29298,http://dx.doi.org/10.1039/c5cp04460j.

[37]Z.Skeete,H.-W.Cheng,Q.M.Ngo,Ch.Salazar,W.Sun,J.Luo,Ch-J.Zhong, Squeezed’interparticlepropertiesforplasmoniccouplingandSERS characteristicsofduplexDNAconjugated/linkedgoldnanoparticlesof homo/hetero-sizes,Nanotechnology27(2016)325706,http://dx.doi.org/10.

1088/0957-4484/27/32/325706.

[38]H.Xu,E.J.Bjerneld,J.Aizpurua,P.Apell,L.Gunnarsson,S.Petronis,B.Kasemo, Ch.Larsson,F.Hook,M.Kall,Interparticlecouplingeffectsin

surface-enhancedramanscatteringProc.SPIE4258,Nanoparticlesand NanostructuredSurfaces:NovelReportersWithBiologicalApplications,4258, 2001,pp.35–42,http://dx.doi.org/10.1117/12.430771.

[39]Z.Zhu,T.Zhu,Zh.Liu,Ramanscatteringenhancementcontributedfrom individualgoldnanoparticlesandinterparticlecoupling,Nanotechnology15 (2004)357–364,http://dx.doi.org/10.1088/0957-4484/15/3/022.

[40]K.Nehra,S.K.Pandian,M.S.S.Bharati,V.R.Soma,EnhancedcatalyticandSERS performanceofshape/sizecontrolledanisotropicgoldnanostructures,NewJ.

Chem.43(2019)3835–3847,http://dx.doi.org/10.1039/C8NJ06206D.

[41]H.Yockell-Leliévre,F.Lussier,J.-F.Masson,Inﬂuenceoftheparticleshapeand densityofself-assembledgoldnanoparticlesensorsonLSPRandSERS,J.Phys.

Chem.C119(2015)28577–28585,http://dx.doi.org/10.1021/acs.jpcc.

5b09570.

[42]Ch.J.Orendorff,A.Gole,T.K.Sau,C.J.Murphy,Surface-enhancedraman spectroscopyofself-assembledmonolayers:sandwicharchitectureand nanoparticleshapedependence,Anal.Chem.77(2005)3261–3266,http://dx.

doi.org/10.1021/ac048176x.

[43]NanoComposix,Tools,NanoComposix,2014(accessed30Jan2020http://

nanocomposix.com/pages/tools.

[44]M.Quinten,OpticalPropertiesofNanoparticleSystems:MieandBeyond, Wiley-VCH,Weinheim,Germany,2011.

[45]Y.-H.Kwon,R.Ossig,F.Hubenthal,H.-D.Kronfeldt,Inﬂuenceofsurface plasmonresonancewavelengthonSERSactivityofnaturallygrownsilver nanoparticleensemble,J.RamanSpectrosc.43(2012)1385–1391,http://dx.

doi.org/10.1002/jrs.4093.

[46]B.Sharma,R.R.Frontiera,A.-I.Henry,E.Ringe,R.P.VanDuyne,SERS:

materials,applications,andthefuture,Mater.Today15(2012)16–25,http://

dx.doi.org/10.1016/S1369-7021(12)70017-2.

[47]R.Pilot,R.Signorini,C.Durante,L.Orian,M.Bhamidipati,L.Fabris,Areview onsurface-enhancedramanscattering,Biosensors9(2019)57,http://dx.doi.

org/10.3390/bios9020057.

[48]Y.Fleger,Y.Mastai,M.Rosenbluh,D.H.Dressler,SurfaceenhancedRaman spectroscopyofaromaticcompoundsonsilvernanoclusters,Surf.Sci.603 (2009)788–793,http://dx.doi.org/10.1016/j.susc.2009.01.020.

[49]BenzophenonedatasheetintheNISTChemistryWebBook,https://webbook.

nist.gov/cgi/cbook.cgi?ID=C119619&Mask=400#UV-Vis-Spec.

Biographies

PetraPálisa1styearPh.D.studentattheDepartmentofExperimentalPhysics, FacultyofScienceandTechnology,UniversityofDebrecen.ShehasanM.Sc.degreein materialscience.Shehas3yearsofexperienceinRamanspectroscopyandSurface- enhancedRamanScattering.Shedidhermasterthesisinthisﬁeldandnowcontinue theworkduringherPh.D.studies.

AttilaBonyárisanassociateprofessorattheDepartmentofElectronicsTechnology atBudapestUniversityofTechnologyandEconomics.HehastwoM.Sc.degreesin electricalengineeringandbiomedicalengineeringandaPh.D.inelectricalengineer- ing.Hehas14yearsofexperienceinthedevelopmentofopticalandelectrochemical, afﬁnity-typebiosensors,utilizinglow-dimensionalnanomaterials,plasmonics,and nanometrology(AFM).

MiklósVeresistheheadoftheDepartmentofAppliedandNonlinearOpticsatthe InstituteforSolidStatePhysicsandOptics,WignerResearchCentreforPhysics.He hasanM.Sc.degreeinphysicsandaPh.D.inphysicsintheﬁeldofinvestigation ofcarbonmaterialswithRamanspectroscopy.Hehas20yearsofexperiencein RamanSpectroscopy,developingtheSRStechnique.Heisleadingaprojectbased onnanostructuresandappliedspectroscopy.

(10)