Surface-Enhanced Spectra on D-Gluconic Acid Coated SilverNanoparticles

IGOR O. OSORIO-ROMAN,* VICTORIA ORTEGA-VASQUEZ, VICTOR VARGAS C., andRICARDO F. AROCADepartamento de Quimica Inorganica, Facultad de Quimica, Pontificia Universidad Catolica de Chile, Chile (I.O.O.-R.); Departamento de

Quimica, Facultad de Ciencias, Universidad de Chile, Chile (V.O.-V., V.V.C.); and Department of Chemistry and Biochemistry, Faculty of

Sciences, University of Windsor, Windsor, Canada (R.F.A.)

Coated silver (Ag) colloids synthesized with D-glucose permit the

observation of surface-enhanced fluorescence (SEF) and surface-enhanced

resonance Raman scattering (SERRS) of the rhodamine B (RhB)

molecule. The organic coating formed during the synthesis of the Ag

nanostructures was identified by its surface-enhanced Raman scattering

(SERS) spectrum as D-gluconic acid. The RhB molecule is used to

exemplify the distance dependence of SEF and SERRS on the coated Ag

nanostructures. The fluorescence enhancement factor for RhB on D-

gluconic acid coated silver nanoparticles was determined experimentally

and estimated using a simple model. Further support for the plasmon

enhancement is obtained from the fact that the measured fluorescence

lifetime of RhB on the silver coated with D-gluconic acid is shorter than

that found on a glass surface. A very modest enhancement factor is

obtained, as expected for very short distance between RhB and the metal

surface. Given the very thin metal–fluorophore separation, estimated

from the size of the D-gluconic acid, the energy transfer or fluorescence

quenching is still efficient and the SEF enhancement is just overcoming

the energy transfer. Therefore, both SEF and SERRS are observed.

Notably, the aggregation of coated nanoparticles also increases the

enhancement factor for SEF.

Index Headings: D-gluconic acid; Coated nanoparticles; Silver nanopar-

ticles; Plasmon enhanced emission; Surface-enhanced resonance Raman

scattering; SERRS; Surface-enhanced fluorescence; SEF.

INTRODUCTION

Plasmon enhanced emission and plasmon enhanced scatter-ing1 reveal a fundamental difference in their enhancementdistance dependence.2,3 Surface-enhanced fluorescence (SEF)is maximized at a certain distance from the surface of thenanostructure, while surface-enhanced Raman scattering(SERS) or surface-enhanced resonance Raman scattering(SERRS) is highest for the first molecular layer (the ‘‘firstlayer effect’’) adsorbed onto the metal nanostructure. Becauseboth (SERS and SEF) are due to plasmon coupling to localizedsurface plasmon resonances (LSPR)4,5 in nanostructures ofdifferent size, shape, and composition, the fabrication of layer-protected metal nanoparticles (NPs)6–9 is a driving force forSEF applications. In particular, nanoparticles protected with anorganic layer are attractive substrates for SEF with potentialapplications as biosensors.10 The main objective of this workwas to explore the fabrication of coated silver (Ag) nano-structures with sugars. The synthesis of silver colloids using D-glucose as reducing agent offers silver nanostructures coatedwith an organic thin film of D-gluconic acid. The organic film

provides a thin overlayer that separates the adsorbedfluorophore from the metal surface. Coated silver nano-structures were characterized by scanning electron microscopy(SEM), atomic force microscopy (AFM), ultraviolet–visible(UV-Vis) absorption, and surface-enhanced Raman scattering.

MATERIALS AND METHODS

Sodium hydroxide (99.99%), 3-aminopropyltriethoxysilane(APTS), rhodamine B, D-glucose, and D-gluconic acid werepurchased from Sigma-Aldrich; silver nitrate (99.9%) wasobtained from Merck. Tri-distilled water (.18 mX) was usedfor the solution preparation. Glass slides were cleaned with‘‘piranha solutions’’ (3:1 concentrated sulfuric acid/30%hydrogen peroxide) for 24 hours. Deionized water was usedto rinse the glass substrates, which were then dried under awarm air stream. The clean glass slides were silanized bysoaking them in APTS at 0.2% in acetone solution for 12hours. The APTS-coated glass slides were rinsed withdeionized water and dried in air.

Silver colloids were prepared by reducing silver ions with D-glucose. First, 3.6 mL of glucose (1.2 3 10�2 M) was added to30 mL of AgNO3 (1.0 3 10�3 M). The solution was warmed upto 85 8C and then 300 lL of NaOH (0.08 M) was added slowly.After 30 min, the colloid solution had a gray color. The silvercolloids were transferred to the APTS glass slides byimmersion in the colloidal silver solution for 6, 12, 24, and36 hours. Finally, the APTS–glass slides were immersed for 2min in an aqueous solution of RhB 1.0 3 10�8 M.

The number of RhB molecules per area (lm2) afterimmersion was calculated measuring the fluorescence of theRhB solution before and after immersion of the APTS–glassand APTS–glass–NPs. We assume that the fluorescence isdirectly proportional to the concentration of molecules;therefore, the quantity of molecules adsorbed onto the surfacecan be extracted from the decrease in the fluorescence emissionof the solution. Therefore, the number of fluorescent moleculestransferred to the APTS–glass and the APTS–glass–NPs wasestimated using the difference between the fluorescence signalof the solution before and after immersion. In Table I thesecond column shows the values of RhB per area (lm2).

Absorption spectra were recorded using a Cary 50 scan UV-Vis spectrophotometer. The fluorescence spectra were recordedin the ISS-PC1 photon-counting spectrophotometer. Fluores-cence and Raman spectra of the solid samples were recordedusing a Renishaw Research Raman microscope system RM2000. This system is equipped with a Peltier charge-coupleddevice (CCD) detector and Leica microscope. Fluorescencespectra were recorded using the 514.5 nm line of an Argon-ionlaser, and Raman and SERS spectra were recorded using the

Received 25 February 2011; accepted 10 May 2011.

* Author to whom correspondence should be sent. E-mail: iosorior@uc.cl.

DOI: 10.1366/11-06279

838 Volume 65, Number 8, 2011 APPLIED SPECTROSCOPY0003-7028/11/6508-0838$2.00/0

� 2011 Society for Applied Spectroscopy

632.5 nm line laser. Single-point spectra for fluorescence,Raman, and SERS spectra were recorded with 4 cm�1

resolution and 10 s accumulation times. Using the micro-Raman, all measurements were recorded in backscatteringgeometry using a 503 microscope objective with a numericalaperture value of 0.75, providing a scattering area ofapproximately 1 lm2. Finally, the spectral scanning conditionswere chosen to avoid sample degradation.

Time-resolved measurements were recorded using a K2Multifrequency Cross-Correlation Phase and ModulationFluorometer. The emission was observed through a Schott550 nm cut-on filter. The modulation frequency was in the 10to 250 MHz range. The multi-exponential analysis wasaccomplished by means of Global Unlimited software.11

The characterization of the shape and size of the nanoparticlesynthesis with D-glucose was performed using a JEOL-5410scanning electron microscope (SEM) with energy dispersive X-ray analysis (EDAX).

The AFM images were recorded using a Digital InstrumentsNanoScope IV, operating in non-contact tapping mode with annþ-silicon tip. Images were collected with high resolution (512lines per scan) at a scan rate of 0.5 Hz. The quantummechanical calculations were carried out using Gaussian ‘03for Unix. Geometry optimization, harmonic frequencies, andintensities were computed at the B3LYP/Lanl2dz, 6-311þþG(d,p) level,12 and the scaling factor for the frequencies was0.9713.13 The convergence criteria in the structure optimization

were (Opt¼ tight), used for molecular systems with very smallforce constants.12

RESULTS AND DISCUSSION

The plasmon absorption spectra of the coated silver colloidalsolution and that of nanoparticles deposited on the silanizedglass (after 36 hours) are presented in Fig. 1. The silver colloidspectrum presents a broad band with a maximum at 435 nm,consistent with a broad distribution of silver nanoparticlessurrounded by a medium with a high refractive index.14 Theplasmon absorption spectrum of silver colloids deposited onAPTS–glass slides shows a strong band center at 400 nm and abroad band at 780 nm, which is due to the formation ofcolloidal aggregates. The SEM images of silver–glucosecolloids on graphite and on the APTS–glass–NPs substrateare shown in Figs. 2A and 2B, respectively, along with thecorresponding histograms. SEM images show that both thecolloidal nanoparticles (dehydrated) on graphite and the coatednanoparticles deposited on glass consist mainly of spheroidsand ellipsoids. The histogram analysis allows us to concludethat approximately 65% of the particles in colloidal solutionhave diameters between 20 and 80 nm and about 27% of themare in the 100 to 160 nm range. Likewise, for the sliver NPsdeposited on the APTS–glass surface, approximately 71% ofthe particles have a diameter from 60 to 120 nm and 24% are inthe 160 to 220 nm range. The AFM images of the silver

TABLE I. Two-component lifetime analysis of RhB and intensity fraction of RhB on glass and on silver substrates.

Surface f1a s1b f2 s2 v2,c ,s.d

Glass 0.042 6 0.006 14.77 6 3.0 0.958 3.82 6 0.03 1.3 4.27Ag (6H) 0.162 6 0.005 10.35 6 0.48 0.838 2.14 6 0.02 1.0 3.47Ag (12H) 0.204 6 0.006 8.38 6 0.30 0.796 2.24 6 0.02 0.9 3.49Ag (24H) 0.219 6 0.006 7.46 6 0.25 0.781 2.36 6 0.02 1.4 3.48Ag (36H) 0.205 6 0.006 7.89 6 0.28 0.788 2.21 6 0.02 1.3 3.36Ag–citrate 0.047 6 0.004 9.13 6 1.67 0.956 0.54 6 0.01 1.2 0.95

a s1, s2 lifetimes in nanoseconds.b f1, f2 intensity fractions.c v2, reduced v-square; calculated using 0.28 for the phase standard deviation.d ,s. ¼ f1 3 s1þ f2 3 s2 (ns).

FIG. 1. UV-Vis extinction spectra of silver colloid and the thin silver film onthe APTS silanized glass (after 36 hours).

FIG. 2. SEM images and histograms of the size distribution of the silver NPs.(A) Silver–glucose colloid solution and (B) silver deposited on APTS glass.

APPLIED SPECTROSCOPY 839

nanostructures formed after dipping the APTS–glass substratesin the colloidal solution for 6, 12, 24, and 36 hours are given inFig. 3. From the AFM images it can be seen that increasing theimmersion time improves the concentration of NPs adsorbedonto the APTS–glass substrate, as well as the formation ofaggregates.

The Raman spectrum of the APTS–glass–NPs, presented inFig. 4A, is in fact the SERS spectrum of D-gluconic acid, asubproduct of the redox reaction of the silver nitrate with D-glucose. Moreover, D-gluconic acid is chemically adsorbedonto the silver nanoparticles through its carboxylic group. TheRaman spectrum of the D-gluconic acid in solid state is givenin Fig. 4B. The characteristic vibrational Raman modes of theD-gluconic are the mC–H stretching vibrations in the region of2800–3100 cm�1, mC–C vibrations in the region above 1000

cm�1, and the band at 1606 cm�1 assigned to mCO in COOHobserved with very low relative intensity. This SERS spectrumof the chemisorbed species is obviously quite different in thepattern of relative intensities. The carboxylic band is seen withhigh relative intensity. The vibrational modes mC–H, mC–C,and mCO in COOH are shifted towards lower wavenumbers.The SERS of D-gluconic acid is also an indication that theorganic acid acts as a capping agent for the silver nanoparticles.

To support the assumption that the D-gluconic acid acts ascapping agent of the silver nanoparticles, we performed adensity functional theory (DFT) calculation of the vibrationalRaman spectrum of D-gluconic acid forming a Ag2 complex,and the results are shown in Fig. 5 (top). It is likely to findgluconate coating the Ag colloids, and consequently, wecalculated the amount of D-gluconic acid using its Ka (pKa ¼3.86) under the reaction conditions in the synthesis. It wasestimated that the amount of D-gluconic acid is three timeshigher (concentration) than the gluconate. In addition, the pHof the colloid was kept below 5.0, bringing the acid–baseequilibrium towards D-gluconic acid. The result for the silvercomplex of D-gluconic acid seems to confirm the assignmentof the mC–O in –COOH mode in the SERS as the vibrationalband with high relative intensity. The trend observed in thecalculated spectra presents the same behavior seen in theexperimental spectra, that is, an increase of intensity of theband and a shift to lower frequency, with respect to thespectrum of the D-gluconic acid molecule. A vibrational bandcalculated at 187 cm�1 and assigned to mO––Ag2 couldcorrespond to the experimental band recorded at 240 cm�1. Insummary, the experimental and theoretical results seem tosupport the assumption that the silver nanoparticle is coatedwith D-gluconic acid. The acid interacts with the silver surfacethrough the carboxylic group forming a silver complex orientedperpendicular to the metal surface. Therefore, the SERSspectrum may have contributions from both the gluconic acidand small fraction of gluconate in the first layer of adsorbates.

Fluorescence Enhancement Factor of Rhodamine B. Thefluorescence of RhB has been widely studied under differentexperimental conditions, and its quantum yield has beendetermined in water solution (Qy ¼ 0.65)15 and in solid state

FIG. 3. AFM images of silver-coated glass at 6, 12, 24, and 36 hours ofdipping the APTS glass in the silver colloidal solution.

FIG. 4. (A) SERS spectra of D-gluconic acid adsorbed on silver NPs glasssubstrate, and (B) the Raman spectra of solid D-gluconic acid.

FIG. 5. (A) DFT calculation of D-gluconic acid–Ag cluster complex, and (B)D-gluconic acid.

840 Volume 65, Number 8, 2011

(Qy ¼ 0.80),16 and its natural lifetime (sN ¼ 5.36 ns) wasestimated from the intensity of the appropriate absorption bandusing the Strickler and Berg relation.17 These values are used inthe calculation of the SEF enhancement factor. Figures 6A and6B display the absorption and fluorescence spectra of RhB inaqueous solution and in the solid phase deposited on theAPTS–glass surface, respectively. The pronounced shoulder atapproximately 515 nm is attributed to H-aggregates.16,18–20

The fluorescence spectrum in the solid phase shows a clearincrease in bandwidth compared with the spectrum in solution.A deconvolution of the fluorescence spectra using Lorentziansis also shown in Figs. 6A and 6B. The deconvolutionfluorescence spectrum of Fig. 6A shows a secondary maximumat 620 nm, a band associated with the fluorescence of the RhBaggregates. The relative intensity is approximately 1/5 themaximum intensity at 550 nm. In the fluorescence spectrum ofthe solid (see Fig. 6B), there is a secondary maximum at 625.5nm; its relative intensity increases about four times with respectto the solution spectra, indicating an increase of the amount ofRhB aggregates.

Figure 7 shows the enhanced fluorescence of RhB on coatedAg nanostructures formed on the APTS–glass substrate. Thesesurfaces (see Fig. 3) correspond to different immersion times ofthe APTS glass in the colloidal silver solution: 6, 12, 24, and36 hours. Fluorescence spectra were recorded on differentpoints of the sample (ten spectra on each surface), and thespectra shown in the Fig. 7 correspond to the average. Thesurface with the higher concentration of nanostructures (surfaceobtained after 36 hours of silver deposition) gives the highestSEF intensity of RhB (15 times larger than the intensity of RhBon the uncoated surface). Since the coating is very thin, thesurface-enhanced resonance Raman spectra scattering spectrumis also observed, as can be seen in Fig. 7, which displays anexpansion of scale in the wavelength region of 520 to 570 nm.The wavenumbers are assigned to aromatic ring vibrations ofRhB, in agreement with previously reported data.21 Both thefluorescence and the SERRS spectra intensities increase withthe addition of nanoparticles deposited on the APTS–glass (seeFig. 7). The very thin organic coating (;2 nm) permits theobservation of SERRS at 514.5 nm, as has been demonstratedrecently with silica-coated nanoparticles.22 SERS of thefluorophore is also attained with the 632.8 nm and with the785 nm laser line. The better signal-to-noise ratio obtained withthe 632.8 nm laser line (not shown) can be explained as acomplementary contribution of the pre-resonant Raman effect.In addition, the D-gluconic acid coating seems to be compactenough and does not allow the fluorescent molecules to becloser to the metal surface, avoiding quenching and explainingthe observation of SEF.

Calculation of the Fluorescence Enhancement Factor.The enhancement factor (EF) of the fluorescent signal for afluorophore interacting with a metal substrate is usuallycalculated from the fluorescence intensity ratio in the presenceand absence of the metal substrate, both measured underidentical experimental conditions. We use a simple modelincluding the most important experimental factors: fluores-cence dynamics, continuous flow fluorescence, and theconcentration of molecules on silver substrates.

Dynamics of Fluorescence on Silver Substrate. In order toanalyze the effect on the amount of silver deposited on glasscoated with APTS in the relaxation dynamics of the firstexcited electronic state, the lifetime of RhB deposited on

metallic substrates obtained after 6, 12, 24, and 36 hours ofimmersion in the solution of colloidal silver was measured. Thelifetime of RhB deposited on a substrate of silver producedfrom a solution of citrate using colloidal silver was alsomeasured.

The results of discrete exponential analysis decay are givenin Table I. In all cases, the data and phase modulation weresuccessfully adjusted to a bi-exponential model decay. Theseresults show that the lifetime of RhB on the silver surfacecoated with D-gluconic acid is lower than that obtained on aglass surface without silver NPs.23 On the other hand, it is wellknown that the dynamics of the relaxation of the first excitedelectronic state of a molecule near a metal surface is a functionof the fluorophore metallic surface distance.23,24 The resultsshow that the lifetime of RhB does not depend on the amount

FIG. 6. Absorption and fluorescence spectra of RhB. (A) Water solution, and(B) solid sample on APTS–glass substrate. Dashed line represents Lorentziandeconvolution of the fluorescence spectra.

FIG. 7. Fluorescence spectra of RhB, in solid phase on silver APTS–glasssubstrates, prepared with 6, 12, 24, and 36 hours of immersion and RhBSERRS spectra in the wavenumber expansion scale.

APPLIED SPECTROSCOPY 841

of silver deposited on the APTS–glass, confirming that thesurface–fluorophore distance remains practically constant.When the silver colloids obtained with citrate as reducingagent are used, the lifetime of RhB decreases significantly. Thecitrate does not form a strong spacer layer on the metal surface,allowing the fluorophore to be located directly on the surface.When the fluorophore is close to the metal surface, thefluorescence is quenched and lifetime decreases rapidly, i.e.,the nonradiative energy transfer process predominates.

Static Fluorescence and Dynamic Fluorescence on SilverSubstrate. The fluorescence is directly proportional to theconcentration of molecules; therefore, we calculate the totalnumber of molecules of RhB deposited on the glass and oneach surface. The number of molecules per square micrometer(RhB/lm2) was obtained from the difference intensity spectra[DI(k) ¼ Ibe(k) � Iaf(k)]; Ibe(k) is the intensity fluorescencespectrum of RhB recorded before and Iaf(k) is that after thesubstrate has been immersed in the RhB 1.0 3 10�8 M aqueoussolution. The intensity of the difference spectra is proportionalto the number of fluorophore molecules on the surface. TheIbe(k) and Iaf(k) spectra are registered in a conventionalspectrofluorometer under high signal-to-noise ratio conditions.The average enhancement factor (EF� IT) was calculated fromthe fluorescence intensity ratio (IT).

If we analyze column seven of Table II, we can determinethat the greater the amount of silver nanoparticles, the greaterthe value of the average EF � IT, and it seems to be incorrespondence with the AFM images of the surfaces in Fig. 3.Further, Fig. 3 shows the presence of large aggregates(diameter . 160 nm) on the APTS–glass surfaces that wereimmersed for a longer time in the colloidal solution (24 and 36h). The EF � IT trend is consistent with the increasingcontribution of aggregates of the nanoparticles producinggreater average value for the enhancement.

The contribution of the fluorescence dynamics or kineticfactor (KF) was calculated with the values in Table I, quantumyields, and natural lifetime of the RhB. The quantum yield ofemission of the molecules on glass is defined by Eq. 1, whilethe emission performance on silver substrates is defined by Eq.2:25

Q0 ¼C0

rad

C0rad þ Cint

no rad

ð1Þ

Qme ¼Crad

Crad þ Cintno rad þ Cme

no rad

ð2Þ

where Q0 is the emission yield of the molecules on glass, C0rad

is the molecular radiative deactivation, Cintno rad is the molecular

nonradiative deactivation, Qme is the emission yield of themolecules on the silver substrate, Crad is the molecular radiativedeactivation on the silver substrate, Cint

no rad is the molecularnonradiative deactivation, and Cme

no rad is the molecular non-radiative deactivation on the silver substrate. Our kinetic factorwas calculated as KF¼Qme/Q0 and, as shown by the values inTable II, these values are constants as the separation betweenthe metal–fluorophores is the same and corresponds to D-gluconic acid on the metal surface.

The experimental average enhancement factor represents aproperty of a fluorophore on a metallic substrate and, in ourexperiments, depends on the amount of silver nanostructuresand fluorophore deposited on the surface probed by the laser

beam. Surface-enhanced fluorescence depends on two factors:increasing electromagnetic field and the kinetic factor:

EF� IT ¼El

E0

� �Qme

Q0

� �ð3Þ

To define the kinetic factor, this equation can be written withEqs. 1 and 2, as follows:

EF ¼ Mloc

Mrad

MT þ Q�10 � 1

ð4Þ

where Mloc corresponds to amplification of the electromagneticfield defined by Mloc¼El/E0, where El is the electric field at thenanoparticles and E0 is the electric field of the excitation light.The kinetic factor in Eq. 4 is defined by Mrad/(MTþQ�1

0 � 1),where Mrad ¼ Crad / C0

rad and MT ¼ Crad(me) / C0rad þMrad.

We calculated the amplification factor considering theconcentration of molecules, and the new Eq. 5 can be written as

EF ¼ Mloc

Mrad

MT þ Q�10 � 1

CF ð5Þ

where CF corresponds to the ratio between molecules on silverand molecules on glass, which is the concentration factorshown in Table II for each of the surfaces. Moreover, we canestimate the average value of Mloc (see Table II). The KFvalues are higher than those of Mloc, and this could beexplained due to the close proximity to the metal surface,where nonradiative deactivation is important.

Finally, the EF calculated with Eq. 5 is in agreement withthe enhancement factors (EF� IT) that were measured directlyfrom the ratio of fluorescence intensities. The low valueobtained here can be explained in terms of the short distancebetween the surface and RhB. The ;2 nm separation,estimated from the size of the D-gluconic acid molecule, isnot sufficient to achieve the maximum value of enhancement.At this distance, the energy transfer process, quenching thefluorescence, is still efficient. The optimal dye–metal surfacedistance should be greater than 5 nm.2,26

CONCLUSION

Organic coated silver colloids have been prepared using D-glucose as reducing agent of silver nitrate, forming Agnanoparticles coated with D-gluconic acid. A thin and compactcoating is formed that allows the observation of surface-

TABLE II. Factors that determine the amplification of fluorescence insubstrate silver–D-gluconic acid.

SubstrateRhB/lm2

3 10�5,a CFb KFc Mlocd EF � IT

e

Glass 6.616 hours 11.3 1.70 2.24 0.72 312 hours 16.9 2.56 2.24 1.12 624 hours 22.6 3.42 2.23 1.34 1036 hours 26.7 4.04 2.22 1.69 15

a Concentration calculated from DI(k) ¼ Ibe(k) � Iaf(k).b Increased concentration of molecules, calculated by the ratio between

molecules on silver and molecules on glass.c Kinetic factor calculated with the values of Table I.d Amplification of the electromagnetic field.e The enhancement factor (EF� IT) from the fluorescence intensity ratio.

842 Volume 65, Number 8, 2011

enhanced resonance Raman scattering and surface-enhancedfluorescence. The average SEF enhancement factor wasdetermined using surfaces with variable concentrations ofcoated nanostructures and aggregates. An average enhance-ment factor of 15 was obtained for the most concentrated filmof the coated nanoparticles transferred to the glass surfacetreated with 3-aminopropyltriethoxysilane. In addition, thelifetime measurements indicate that the metal–fluorophoreseparation is constant for samples with different nanoparticleconcentrations on the silinized glass. The SERRS observed onthe fluorescent background is consistent with the spectrum ofthe dye. The SERS of the coated silver nanostructures confirmsthe characterization of D-gluconic acid as the main coatingcomponent.

ACKNOWLEDGMENT

Dr. Igor O. Osorio-Roman acknowledges FONDECYT Initiation ResearchGrand N811100067.

1. R. Aroca, Surface-Enhanced Vibrational Spectroscopy (John Wiley andSons, Chichester, 2006), Chap. 3, pp. 73–106.

2. R. Aroca, G. J. Kovacs, C. A. Jennings, R. O. Loutfy, and P. S. Vincett,Langmuir 4, 518 (1988).

3. A. Wokaun, H. P. Lutz, A. P. King, U. P. Wild, and R. R. Ernst, J. Chem.Phys. 79, 509 (1983).

4. E. C. Le Ru, P. G. Etchegoin, J. Grand, N. Felidj, J. Aubard, and J. G.Levi, J. Phys. Chem. C 111, 16076 (2007).

5. K. A. Willets and R. P. Van Duyne, Annu. Rev. Phys. Chem. 58, 276(2007).

6. A. R. Guerrero and R. F. Aroca, Angew. Chem. Int. Ed. 50, 665 (2011).7. Y. Zhang, K. Aslan, M. J. Previte, and C. D. Geddes, Appl. Phys. Lett. 90,

053107/1 (2007).8. J. Zhang, J. Malicka, I. Gryczynski, and J. R. Lakowicz, J. Phys. Chem. B

109, 7643 (2005).9. J. R. Lakowicz, C. D. Geddes, I. Gryczynski, J. Malicka, Z. Gryczynski, K.

Aslan, J. Lukomska, E. Matveeva, J. Zhang, R. Badugu, and J. Huang, J.Fluoresc. 14, 425 (2004).

10. C. D. Geddes, K. Aslan, I. Gryczynski, and J. R. Lakowicz, ‘‘Metal-Enhanced Fluorescence Sensing’’, in Fluorescence Sensors and Biosen-sors, R. B. Thompson, Ed. (Taylor and Francis Group, Boca Raton, FL,2006), Chap. 7, pp. 121–181.

11. J. M. Beechem, E. Graton, and W. W. Mantulin, Global Lifetime DataAnalysis (University of Illinois, 1990).

12. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J.R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery, Jr., R. E. Stratmann,J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C.Strain, O. Farkas, J. B. V. Tomasi, M. Cossi, R. Cammi, B. Mennucci, C.Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala,Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B.Foresman, J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov, G. L.A. Liu, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T.Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P.M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres, C.Gonzalez, M. Head-Gordon, E. S. Replogle, and J. A. Pople, Gaussian 03,Revision A.1 (Gaussian, Inc., Pittsburgh, 2003).

13. A. P. Scott and L. Radom, J. Phys. Chem. 100, 16502 (1996).14. J. C. Noguez, J. Phys. Chem. C 111, 3806 (2007).15. R. F. Kubin and A. N. Fletcher, J. Luminesc. 27, 455 (1982).16. M. Kawasaki and S. Mine, J. Phys. Chem. B 109, 17254 (2005).17. S. J. Strickler and R. A. Berg, J. Chem. Phys. 37, 814 (1962).18. F. del Monte and D. Levy, J. Phys. Chem. B 102, 8036 (1998).19. D. S. dos Santos and R. F. Aroca, Analyst 132, 450 (2007).20. T. Itoh, V. Biju, M. Ishikawa, Y. Kikkawa, K. Hashimoto, A. Ikehata, and

Y. Ozaki, J. Phys. Chem. B 110, 21536 (2006).21. J. Zhang, X. Li, X. Sun, and Y. Li, J. Phys. Chem. B 109, 12544 (2005).22. J. F. Li, Y. F. Huang, Y. Ding, Z. L. Yang, S. B. Li, X. S. Zhou, F. R. Fan,

W. Zhang, Z. Y. Zhou, D. Y. Wu, B. Ren, Z. L. Wang, and Z. Q. Tian,Nature (Washington, D.C.) 464, 392 (2010).

23. R. R. Chance, A. Prock, and R. Silbey, Adv. Chem. Phys. 37, 1 (1978).24. A. Wokaun, H. P. Lutz, A. P. King, U. P. Wild, and R. R. Ernst, J. Chem.

Phys. 79, 509 (1983).25. E. C. Le Ru and P. G. Etchegoin, Cornell University, Natl. Lab. Arch.

Phys. http://arxiv.org/abs/physics/0509154 (2005).26. H. Szmacinski, K. Ray, and J. R. Lakowicz, Anal. Biochem. 385, 358

(2009).

APPLIED SPECTROSCOPY 843