Homogeneous Silver-Coated Nanoparticle Substrates for Enhanced

Oct 28, 2006 - using a monolayer of FITC-conjugated human serum albumin (FITC-HSA) and tested using laser scanning microscopy at 488 nm excitation wav...
3 downloads 7 Views 439KB Size
Download PDF
J. Phys. Chem. B 2006, 110, 23085-23091

23085

Homogeneous Silver-Coated Nanoparticle Substrates for Enhanced Fluorescence Detection Fang Xie,† Mark S. Baker,‡ and Ewa M. Goldys*,† DiVision of Information and Communication Sciences, and Australian Proteome Analysis Facility Ltd. & Department of Chemistry and Biomolecular Science, Macquarie UniVersity, Sydney, NSW 2109, Australia ReceiVed: April 6, 2006; In Final Form: September 14, 2006

A simple method has been developed for the deposition of uniform silver-coated nanoparticles on glass substrates, with a homogeneous distribution shown by scanning electron microscopy (SEM). UV-visible spectroscopy and energy-dispersive X-ray analysis (EDX) have been used to characterize both the optical density and elemental content of the deposited nanoparticles. The fluorescence enhancement was investigated using a monolayer of FITC-conjugated human serum albumin (FITC-HSA) and tested using laser scanning microscopy at 488 nm excitation wavelength. The enhancement factor was calculated from individual spectra recorded with a Fluorolog-Tau-3 spectrofluorometer. We identified the nanoparticle growth regime which led to fluorescence enhancement. Such enhancement is detectable when Au core-Ag shell nanoparticles increased their size to 47 nm, in agreement with theoretical estimates. The origin of this enhancement for appropriate size nanoparticles is attributed to the effect of an increased excitation rate from the local field enhanced by the interaction of incident light with the nanoparticles and/or higher quantum yield from an increase of the intrinsic decay rate of the fluorophore. We thus demonstrated that the Au core-Ag shell nanostructures on glass surfaces are promising substrates for fluorescence enhancement with outstanding macroscopic homogeneity. This important feature will pave the way for the application of our substrates in biotechnology and life sciences such as imaging and sensing of biomolecules in proteomics.

Introduction Fluorescence applications are rapidly becoming a leading methodology in medical diagnostics and biotechnology,1 primarily because of their versatility, potential for multiplexing, ease of use, and remarkable sensitivity. Apart from fluorophore stability, the detection limit of a fluorophore is determined by the ratio of its signal to the background emission, often due to unavoidable sample autofluorescence. Therefore, high fluorophore brightness is a critical requirement for fluorescence detection of low levels of analytes, especially in the presence of any interfering background fluorescence. A range of methods have been developed to date to increase the sensitivity of fluorescence detection.2-8 While the studies of the effects of metallic surfaces on fluorescence can be backdated to 1970,9,10 fluorescence amplification by metal nanostructures is a relatively new methodology which has been explored extensively only over the past 5 years. The most recent advances in metal induced fluorescence enhancement (MIFE) and its applications in biotechnology were reviewed by Lakowicz et al.11-14 The phenomenon of metal induced fluorescence enhancement is proposed to be due to interactions of the excited fluorophores with surface plasmon resonances in metals, and this effect is particularly prominent in metal nanostructures. Such structures are able to produce desirable effects such as increased quantum yields, decreased lifetimes, increased photostability, and potential for improved energy transfer.14 However, at very close proximity the fluorescence quenching effect competes with all these favorable effects.15 It has been demonstrated that * Corresponding author. E-mail: [email protected]. † Division of Information and Communication Sciences. ‡ Australian Proteome Analysis Facility Ltd & Department of Chemistry and Biomolecular Science.

fluorophore quenches within 5 nm from the surface of metallic particles. At further distances the enhancement starts to override the quenching and reaches its maximum at about 10 nm from the metal surface.16,17 At larger metal-fluorophore separation the enhancement effect progressively declines. Various MIFE substrates have been reported, mostly based on silver nanoscale structures, including silver nanoparticles deposited on glass surfaces to form silver island films (SiFs).18 In a typical SiF preparation, silver salt is reduced to silver, which is then randomly deposited on the glass surface, upon the sequential addition of sodium hydroxide, ammonium hydroxide, and D-glucose solutions. This preparation method results in a relatively wide distribution of metal particle sizes, ranging from 30 to 80 nm, and displays the plasmon extinction maximum near 430 nm. The SiF substrates have been reported to produce enhancement factors of about 5 and 7 for Cy3-DNA and Cy5DNA, respectively.18 The main drawbacks of SiFs as substrates for fluorescence amplification are poor reproducibility and lack of spatial homogeneity leading to undesirable variations in the enhancement factor. More homogeneous and slightly smaller silver nanoparticles were also investigated as MIFE substrates by using covalent deposition of silver colloids on a glass surface.19,20 Due to improved control of experimental conditions, it has been reported that the brightness of fluorophore deposited on such substrates is 50% higher than on SiFs under the same optical excitation density. Recently, anisotropic particles such as silver nanorods21 and silver triangles22 have also been reported by the same group. In this method, silver nanorods or triangular nanoparticles were deposited directly on glass substrates using seed-mediated cetyltrimethylammonium bromide (CTAB)directed growth, with nanostructure sizes depending on the immersion time. The application of nanorods and triangular

10.1021/jp062170p CCC: $33.50 © 2006 American Chemical Society Published on Web 10/28/2006

23086 J. Phys. Chem. B, Vol. 110, No. 46, 2006 nanoparticles in metal induced fluorescence enhancement shows that the enhancement factors for indocyanine green (ICG)-HSA monolayers are about 12 and 16, respectively, with similar optical excitation densities. Silver fractal-like nanostructures deposited on glass surfaces have also shown considerable fluorescence enhancement factors.23 In that study, silver foil electrodes were placed in deionized water about 10 mm apart and a constant current was applied for a short period; an enhancement factor of over 1000 using fluorescein isothiocyanate (FITC) conjugated human serum albumin (HSA) monolayers was reported.23 In addition to varying the particle shape described above, different techniques for nanoparticle deposition, such as electroplating and light-assisted deposition,23,24 have also been employed for metal induced fluorescence enhancement. We report here a new and simple approach to produce a MIFE substrate leading to good uniformity. In this method, a seed Au colloid is first covalently deposited on (3-aminopropyl)trimethoxysilane (APTMS) coated glass substrates. Subsequently, the coated glass is immersed in a silver enhancer solution to form Au core-Ag shell containing nanostructures. Depending on the time employed during the silver enhancing step, the size of these nanoparticles can range from 10 nm to several hundred nanometers. A monolayer of FITC-conjugated human serum albumin (FITC-HSA) is then formed both on a silver nanoparticle covered glass surface and on bare glass surface as a control. Laser scanning microscopy at 488 nm excitation wavelength reveals the enhancement of fluorescence emission in the areas containing the nanoparticles as well as the bare glass surface without the nanoparticles. The enhancement factor was determined as the ratio of fluorescence intensity on silver to fluorescence intensity on glass based on the fluorescence spectra. Scanning electron microscopic (SEM) images prove that the substrates thus produced are spatially homogeneous, in contrast to the competing methods, which is of significance for real technological applications. Experimental Section Materials. The following materials were purchased from Sigma-Aldrich and used as received: HAuCl4‚3H2O, trisodium citrate dehydrate, (3-aminopropyl)trimethoxysilane (APTMS), silver enhancing solution A, silver enhancing solution B (silver enhancing kit), and FITC conjugated human serum albumin (FITC-HSA). Concentrated HCl, HNO3, H2SO4, and methanol were obtained from J. T. Baker Inc., and 30% H2O2 was obtained from VWR. Glass microscope slides were obtained from Fisher Scientific. Nanopure water (>18.0 MΩ), purified using the Millipore Milli-Q gradient system, was used in all experiments. Au Colloid Synthesis in Aqueous Solution. The Au colloid was synthesized according to the Frens method with slight modifications.25 All glassware used in synthesis was thoroughly cleaned in aqua regia (HCl:HNO3 ) 3:1), rinsed with Milli-Q water, and oven-dried prior to use. The stock solutions of 1% HAuCl4 and 38.8 mM sodium citrate were prepared in advance. Other solutions were freshly made as needed. In a typical Au colloid preparation, 100 mL of 1 mM HAuCl4 was brought to a rolling boil with vigorous stirring in a 500 mL three-neck round-bottom flask equipped with a condenser. A 10 mL volume of 38.8 mM sodium citrate was rapidly added to the vortex of the solution, which resulted in a color change from pale yellow to red wine. Boiling was continued for about 10 min before the heating mantle was removed with vigorous stirring until the solution reached room temperature. The resulting colloidal solution was characterized by UV-visible spectroscopy which

Xie et al. revealed the extinction maximum at 519 nm. Transmission electron microscopy (TEM) showed the particles have an average size of 9 ( 1.4 nm. Self-Assembled Monolayer (SAM) of Au Colloid. The microscope glass slide was first cleaned by piranha solution consisting of 4 parts of H2SO4 to 1 part of 30% H2O2 at 60 °C. (Caution: Piranha solution is extremely caustic. Use only with extreme care.) The glass surface then was derivatized with the organosilane APTMS by immersing it into 10 mM methanolic solution of this organosilane.26 After 2 h, the substrate was removed and rinsed with CH3OH to remove unbound monomers from the surface. Prior to derivatization with Au colloid, the glass substrate was rinsed thoroughly with water and then immersed in the Au colloid solution to form a monolayer of Au nanoparticles on the glass surface. Depending on the time and concentration of the Au colloid solution employed, Au nanoparticle layers with various interparticle distances were formed on the glass surface. A final rinse with water and drying with N2 concluded the derivatiation process. Silver Deposition on Au Nanoparticle Surface. Equal amounts of the silver enhancer solutions A and B were well mixed using a vortex. The glass substrate modified with Au colloid was then immersed in the mixed solution for a certain period. The silver deposition procedure was carried out in the dark as the silver enhancer kit is light sensitive. Extensive water washing was conducted with the glass substrate to terminate the reaction between nanoparticles and the enhancer solution. The glass substrate coated with Au core-Ag shell nanoparticles was then rinsed with water and dried in air for characterization and further experiments. Fluorophore-Protein Conjugation Monolayer Formation. We selected FITC-HSA as a model fluorophore-protein conjugate for this study, based on earlier reports that albumin proteins spontaneously bind to glass and silver surfaces to form a complete monolayer.27 Binding of FITC-HSA to the silvered surface, or, in control experiments, to the glass surface, was accomplished by incubating the surfaces in a 10 µM FITCHSA solution overnight at 4 °C, followed by rinsing with buffer to remove the unbound materials. Methods. As illustrated in Figure 1, the glass slide was cut to approximately 20 mm × 7 mm in size, to fit into a 2 mL Eppendorf tube for derivatization. After derivatization with APTMS, the glass surface was patterned by an adhesive tape cut with a V shape as shown in Figure 1. Following the Au colloid self-assembly and silver deposition steps, both the glass and Au core-Ag shell nanoparticle deposited surface were coated with FITC-HSA, which forms an about 4 nm thick protein monolayer.38 This fluorescent monolayer makes it possible to quantitatively compare the fluorescence intensity of fluorophore-protein conjugates with and without silver nanostructures. The enhancement factor obtained when using the silver was determined as the ratio of fluorescence intensity on silver to fluorescence intensity on glass, given both surfaces are known to have almost equal monolayer coverage.28 Characterization. Transmission electron microscopy was performed with a JEOL Model 1200 EXII instrument operating at 100 kV accelerating voltage. UV-visible extinction spectra were obtained using a Cary spectrophotometer (Cary 5000 UVvis-NIR spectrophotometer). The glass substrates were maintained in an upright position in the solid sample holder. Scanning electronic microscopic images were collected using a scanning electronic microscope (JEOL-Jem-1200 EX II electron microscope), equipped with an energy-dispersive X-ray analysis (EDX) system (Traktor TN-2000 energy-dispersive spectrom-

Homogeneous Silver-Coated Nanoparticle Substrates

J. Phys. Chem. B, Vol. 110, No. 46, 2006 23087

Figure 1. Schematic illustration of the deposition procedure for Au core-Ag shell nanoparticles on APTMS-modified glass surface (top view).

Figure 2. UV-visible spectrum of the synthesized 9 ( 1.4 nm Au colloid.

eter). Fluorescence images were captured by using a laser scanning microscopy SPM2 system (Leica Microsystems) with an electronically controlled acoustooptical beam splitter capable of a minimum bandwidth setting of 5 nm. The sample was excited with an Ar laser emitting at 488 nm with measured power of 0.90 µW, and the emission was collected over a range of 500-600 nm. Fluorescence emission spectra were collected by a Fluorolog-Tau-3 system from Jobin-Yvon-Horiba with 450 W Xe lamp excitation. The spectral width was set to 8 nm. All spectra were corrected for system response at room temperature. To minimize the contribution of scattering excitation light reflected from the Au/Ag nanoparticle surface, the sample was fixed in the sample holder with incident angle of 60°. The emission spectra were recorded over a range of 495-600 nm for the excitation wavelength of 470 nm. Results and Discussion Figure 2 shows the extinction spectrum of a typical Au colloid solution with the extinction maximum at 519 nm and full width at half-maximum of 10 nm, which indicates the absence of aggregation of the Au colloid. The TEM images of these particles were recorded and their size distributions were obtained from at least 200 individual particles using Scion Image beta release 2 (available at www.scioncorp.com). The Au colloid had an average diameter of 9 nm, with a standard deviation of 1.4 nm, implying that the Au colloid solution was practically monodisperse. Using such monodisperse Au colloid as seed for silver deposition gives better prospects for size uniformity after silver enhancement, because in conventional Ag colloid synthesis strong size variations have been reported.29

In the next step we employed the established technique of self-assembly of Au colloids on glass surfaces, investigated earlier in surface enhanced Raman scatttering applications.26 An alkoxysilane, APTMS, was first chemically bonded to hydroxyl groups on the glass substrate surface. The particle attachment was made possible through the alkoxysilane containing amino functional groups. As shown in Figure 3A, the Au colloid was dispersed in a single layer on the glass surface. Individual particles in this layer were physically separated, but small interparticle distance (approximately 4 ( 1.3 nm) ensured the strength of electromagnetic interaction. The self-assembled Au monolayers thus prepared were stable owing to the following factors. First, the Au colloid tends to tightly bind to the glass surface as numerous bonds are formed between Au colloid and substrate surface. The density of these bonds is reported to be as high as 4.5 hydroxyl groups/nm2 of the glass surface.30 For a 9 nm diameter nanoparticle with an approximately 5 × 5 nm effective contact area,26 this produces about 120 linkages leading to strong binding of the nanoparticles to glass. Second, the thermodynamic stability of these surfaces is very high; thus the exchange with molecules in solution containing the same functional group practically does not occur.26 Strong covalent bonds to the glass substrate reduce the surface mobility of the nanoparticles and thus prevent their spontaneous coalescence. Therefore, in the silver deposition step, the reduced silver only covers the Au colloid surface and does not change the location of bound colloidal Au nanoparticles. The silver deposition was done by applying the silver enhancer solution to Au colloids attached to the glass surface. The thickness of silver layer coating was controlled by the length of the silver enhancing step, as this reaction could be simply interrupted by extensive rinsing with water. Depending on the interparticle distance of the Au colloid precursor, the maximum silver enhancing time was varied. Early experiments showed that when silver deposition was conducted for up to 5 min, the coated Au core-Ag shell nanoparticles effectively aggregated to form a bulk film that could be peeled from the glass surface. During the enhancement process, silver was not deposited on the bare glass surface between the Au nanoparticles. This was separately verified by putting a bare glass slide in a silver enhancement solution for 3 min, the same time as our other silver enhancements. Visual inspection showed no silver deposition on the glass surface. This is in agreement with earlier reports which stated that only gold, silver, sulfide, and selenide of zinc and mercury can be enhanced by the silver enhancer solution.31 The process of silver enhancement can be understood as follows. A silver enhancer solution generally consists of silver ions and a reducing agent, buffered at an acidic pH. The reaction between silver ions and the reducing agent, which is silver nitrate

23088 J. Phys. Chem. B, Vol. 110, No. 46, 2006

Xie et al.

Figure 3. SEM images of (A) 9 nm Au colloid, (B) Au core-Ag shell nanoparticles after 1 min silver enhancing step (∼19 nm), and (C) Au core-Ag shell nanoparticles after 3 min silver enhancing step (∼47 nm). Inserts show the morphology of individual particles.

Figure 4. SEM images of Au core-Ag shell nanoparticles with silver deposition for (A) 10 s, (B) 1.5 min, (C) 3 min, (D) 4 min, and (E) 5 min.

and hydroquinone, respectively, for example, could be explained by the following reaction: 2AgNO3 + C6H4(OH)2 f CO(CHCH)2CO + 2HNO3 + 2AgV When the Au colloids are placed in a solution containing reducing molecules, for example, hydroquinone, and silver ions, they will attract both the reducing molecules and silver ions.31 After attaching to the Au crystal, the hydroquinone molecule releases two electrons into the valence cloud of the Au crystal lattice. These extra electrons will eventually cause include two silver ions to be included into the lattice by reducing them to metallic silver atoms. At the same time, hydroquinone molecules are oxidized to quinine and physically released from the growing crystal. In our method, we used normal tape to pattern the APTMSmodified glass surface. Hence the effect of the tape on the amount of FITC-HSA binding should be considered. We have done the test before applying such a configuration (Figure 1) in our experiment. In the test, half of the APTMS-modified glass slide was covered by tape and then removed. The other half of the surface was kept untouched. Both covered, later removed and untouched surfaces were incubated in the FITC-HSA solution overnight at 4 °C for laser scanning microscope characterization with excitation at 488 nm. The image (not shown) shows identical surfaces with respect to fluorescence intensity, suggesting the same amount of protein-dye conjugation and hence the validity of our method. The scanning electron micrographs (Figure 3B,C) show not only the size but also the surface morphology of the Au coreAg shell nanoparticle. They are approximately spherical, especially at short silver enhancing times. Upon close examination (Figure 4), the formed nanoparticles reveal a more complex shape, especially at longer silver enhancing times, which we will refer to as a nanoflower instead of a smooth spherical surface. Figure 4 clearly shows the nanoflower shape of these nanoparticles, with increasing silver enhancing time. The silver coating is believed to be initially homogeneous over the Au core and with increasing enhancing time forms the nanoflowershaped particles. The average particle sizes, calculated from 100 particles, with 1 and 3 min silver enhancing time are about 19 ( 2.7 nm and 47 ( 7.1 nm, respectively.

Figure 5. UV-visible spectra for the Au colloid dominated by absorption (A); silvered Au colloid with 1 min silver enhancing time showing strong absorption and scattering (B); and silvered Au colloid after 3 min silver enhancing time, dominated by scattering (C).

The composition of the nanostructure modified surface was tested by energy-dispersive X-ray analysis (EDX) (results not shown). In the range within 4 keV, the characteristic peaks at 2.195 keV (M) and 2.984 keV (KR) of gold and silver, respectively, were observed. For the 9 nm Au colloid, in addition to Si (KR, 1.739 keV) and C (KR, 0.277 keV) peaks, originating from the glass substrate and APTMS monolayer, a characteristic peak attributed to Au appears at 2.195 keV. After the silver enhancing step, the characteristic peak for silver at 2.984 keV can be seen and its peak intensity increases with increasing silver enhancing time, which suggests that the silver content in the nanoparticles has increased. This is consistent with the fact that the thickness of silver shell increases and the particle grows. In addition to SEM and EDX characterizations, UV-visible spectroscopy was used to record optical properties of the fabricated nanoparticles. The Au core-Ag shell nanoparticles immobilized on glass surface showed characteristic surface plasmon peaks, as shown in Figure 5. In contrast to the Au monolayer, which has a characteristic absorption peak at 519 nm (see also Figure 2), the Au core-silver shell nanoparticles on glass substrate have two distinctive peaks at 394 and 595 nm for 19 nm nanoparticles, and 394 and 635 nm for 47 nm nanoparticles, respectively. With increasing time of the silver enhancing step, and particle size, the extinction at these wavelengths is increased, with a red shift of the long wavelength maximum. The fluorescence images (Figure 6) corresponding to each of the UV-visible spectra have also been recorded for each sample. As mentioned in the Experimental Section, the laser scanning microscopy measurements were carried out after the

Homogeneous Silver-Coated Nanoparticle Substrates

J. Phys. Chem. B, Vol. 110, No. 46, 2006 23089

Figure 6. Laser scanning microscopy images of FITC-HSA monolayer on both nanoparticles and glass surface (excitation 488 nm, emission collection 520-560 nm): (A) 9 nm Au colloid as MIFE substrate, quench; (B) Au core-Ag shell nanoparticles after 1 min silver enhancing step (nanoparticle size 19 nm) as a MIFE substrate, almost no effect; (C) Au core-Ag shell nanoparticles after 3 min Silver enhancing step (nanoparticles size 47 nm) as a MIFE substrate, clear enhancement.

samples were incubated with FITC-HSA to form a complete monolayer in the “confined” area, which included both derivatized metal and clean glass surfaces. Figure 6A shows that the fluorescence intensity from FITC-HSA in the triangular area covered with the Au colloid monolayer is less than that of the surrounding glass surface used as control. As shown in Figure 6B, the Au core-Ag shell nanoparticles produced with 1 min silver enhancing step, the fluorescence signal from the silvered surface is comparable with that from the glass surface. When the silver enhancing time has been increased to 3 min and hence the size of the nanoparticles increased to 47 nm, the fluorescence signal observed from the silvered area is much greater than that from the glass surface, demonstrating fluorescence enhancement. This result unambiguously shows that, depending on the size of the nanoparticles, the fluorescence of FITC was either quenched, minimally affected, or enhanced in comparison to the control area, where fluorescence was due to FITC-HSA monolayer on the bare glass surface. Such an effect of nanoparticles on fluorescence is attributed to the interplay of two principal factors: the change of excitation rate caused by the local electromagnetic field enhanced by interaction of incident light with the metal nanoparticles and the change of quantum yield due to the effect of metals on the intrinsic decay rate of the fluorophore.11 The fluorescence enhancement is known to correlate with the extinction properties of metal particles which are due to two contributions, from absorption (CA) and scattering (CS).32-36 The scattering component of the extinction relates to the extent by which the plasmons can radiate energy as a far-field propagating wave. When a metal particle is spherical with size comparable to the incident wavelength, its optical properties agree reasonably well with Mie theory. We note that our nanoparticles are less than 100 nm in size, where Mie theory reaches its limits, but despite that it has been widely used to describe properties of such a colloid.14 Based on this theory, the extinction cross section for a particle with a dielectric constant 1 is given by

CE ) CA + CS ) k1 Im(R) + k14 |R|2/6π

(i)

where k1 ) 2πn1/λ0 is the wavevector of the incident light in medium 1 and R is the polarizability of the sphere of radius r:

R ) r3(m - 1)/(m + 21)

(ii)

where m is the complex dielectric constant of the metal. For the simple core-shell structure of the nanoparticles, eq iii should

be used for calculation of the polarizability:39

R ) R3

(s - m)(c + 2s) + (1 - g) (c - s)(m + 2s) (s + 2m)(c + 2s) + (1 - g)(2s - 2m)(c - s) (iii)

in which R is the radius of the core-shell nanoparticles, c and s are the complex dielectric constant of the core and shell materials, respectively, and g is the volume fraction of the shell layer which is defined as g ) 1 - (R13/R3) (R1 is the radius of core material). Both monometallic (Au and Ag) and bimetallic nanoparticles are able to enhance fluorescence which is assisted by the modification of the exciting electric field by polarized nanoparticles. Relative polarizabilites for core-only and coreshell geometries can easily be calculated base on eqs ii and iii, respectively.39 The polarizability ratio of 47 nm diameter Au nanoparticles with respect to Au core-Ag shell nanoparticles with 9 nm diameter Au core, and 47 nm outer Ag diameter at the wavelength of 488 nm used in this work, is close to unity (0.96). The use of core-shell Au-Ag nanoparticles for fluorescence enhancement offers the potential to exploit the good homogeneity of Au colloids with small additional gains in polarizability at 488 nm. In eq i, the first term corresponds to the cross section due to absorption and the second term represents the cross section due to scattering. Literature suggests that the absorption term CA is linked to quenching while the scattering term CS is linked to enhancement in surface enhanced fluorescence.11 Equations i, ii, and iii show that CA increases as r3 whereas CS increases as r6, which implies that small metal particles are expected to quench fluorescence because absorption dominates over the scattering, while larger metal particles are expected to enhance fluorescence because the scattering component is dominant over the absorption. On the basis of these considerations, the effect of silvered Au colloid on fluorescence can be correlated with the scattering and absorption components of the extinction spectra (Figure 5). A typical extinction spectrum with little or no scattering is similar to that shown in our Figure 2; this is because, based on theoretical calculations, the scattering efficiency for 9 nm Au colloid is almost 0.35 The emergence of a scattering component is manifested in the spectra as a second broad feature which is red-shifted with respect to the characteristic peak for Ag colloid at 394 nm.37 For small silver-coated Au colloids of similar size the absence of the scattering component and the extinction spectrum dominated by absorption with maximum at 394 nm37 is consistent with its quenching effect on fluorescence. After a

23090 J. Phys. Chem. B, Vol. 110, No. 46, 2006

Xie et al.

Figure 7. (left) Image of sample 3 which was already shown in Figure 6C. (right) Distance dependence of enhancement factor plotted for the confined area shown in left. It shows an average of enhancement factor of about 10.

Figure 8. (left) Fluorescence emission intensity of FITC-HSA (from top to bottom): on substrate with 3 min silver enhancing time; on substrate with 1 min silver enhancing time; clean glass surface; Au colloid monolayer. (right) Normalized emission spectra of FITC-HSA on glass and on Au core-Ag shell nanoparticles with 3 min silver enhancing time.

1 min silver enhancing step, the extinction spectrum shows that, in addition to the absorption peak centered at 394 nm, a new peak at 595 nm has appeared at a position consistent with a scattering component. This agrees with theoretical estimates that the scattering efficiency for 19 nm Ag colloid is nonnegligible, at about 2.1.35 In this region it is not possible to predict whether the nanoparticles will act as a quencher or enhancer of fluorescence, as both absorption and scattering components coexist. With increasing size of silvered Au colloid, the scattering component becomes dominant over the absorption component (Figure 5C). The theoretical calculations provide evidence of the scattering efficiency increased to 6,35 for nanoparticles of this size. These observations correlate with the enhancement of fluorescence for such nanoparticles, which was confirmed by the fluorescence image shown in Figure 6C. These results are in good agreement with those found in other references.14,35,37 The enhancement factor was determined as the ratio of fluorescence intensity on silvered surface to the fluorescence intensity on glass surface, given both surfaces are known to have complete monolayer coverage.28 As shown in Figure 7, the enhancement factor for the silvered sample with 3 min silver enhancing time is about 10 based on image analysis of sample 3. Figure 8 (left) shows the fluorescence spectra from FITCHSA monolayer on samples with Au colloid monolayer, 1 min silver enhancing time, and 3 min silver enhancing time, respectively, as well as a control sample of the clean glass

surface. The enhancement factor for the silvered sample with 3 min silver enhancing time is 13.06 (Figure 8, left). Figure 8 (right) shows the normalized emission intensities of FITC-HSA on glass and on silver nanoparticles, indications that the spectral characteristics of FITC are the same. The observed increase of fluorescence intensity can be due to geometric effects such as an increased surface area. We note most other reports state the same fluorophore coverage on both silvered and glass surfaces.19-22 We have accounted for such geometric factors by calculating the total radiant flux emitted by the fluorophore layer covering our nanostructures. This is a more exact approach compared with the calculation of the surface area. The derivation is given in the Supporting Information, and only a summary is given here. First we calculated the radiant flux emitted by a sphere (modeling a single nanoflower petal) covered with a uniform fluorescent layer. This layer is illuminated by a parallel exciting light beam, and consequently the sides of the sphere are not as bright as its top, which has been taken into account for appropriate integration. We used a Lambertian source approximation consistent with typical observation conditions. By comparing the radiant flux from such a sphere with the radiant flux emitted by a flat surface in the absence of that sphere, we found that the sphere (despite larger surface area) emits only two-thirds the light that would be emitted by the area of the flat surface occupied by that sphere (equal to its cross section). Second, we provided an upper estimate of the radiant flux from the excited nanoflower. The emission from the nanoflower was modeled as the emission from identical

Homogeneous Silver-Coated Nanoparticle Substrates spheres forming its petals. The number of petals was estimated by evaluating the number of spheres that can be located over the flat surface area intersecting the centers of these spheres. Finally, we estimated the total flux from all petals by multiplying the flux from each by the number of petals. This was an upper estimate as an accounting was not made of different petals obscuring one another. The final comparison was made between the flux from the modeled nanoflower compared with the flux from the flat square surface area which is required to accommodate this nanoflower, where the factor of 1.9 was obtained. This means that the presence of nanoflowers on the surface can only account for about 90% increase in fluorescence intensity, with the remaining factor of about 6 attributable to fluorescence enhancement. Conclusion A fast and simple method has been proposed for producing homogeneous silvered nanostructures on a glass surface as potential substrates for fluorescence enhancement assays. Silver was deposited on a Au colloid surface which was immobilized on a glass surface by covalent binding. By varying the silver enhancing time, the sizes of these nanoparticles could be varied from about 10 nm to several hundred nanometers, while maintaining homogeneous (i.e., uniform) distribution. A monolayer of a model protein, FITC-conjugated human serum albumin (FITC-HSA), was formed on silver nanoparticle covered glass surfaces or bare glass surfaces used as controls. Depending on the size of these nanoparticles, the fluorescence of the fluorophore (FITC) was either quenched, unaffected, or enhanced in comparison to the controls, where fluorescence was due solely to the FITC-HSA monolayer located on the bare glass surfaces. The Au core-Ag shell nanostructures produced after 3 min silver enhancing time produced an enhancement factor of 13-fold, suggestive that this system can be employed to produce technologically promising substrates for fluorescence enhancement assays. It was shown that the enhancement factor is much higher than the predictions based on the radiative flux from a monopatterned surface alone. The origin of this enhancement obtained using appropriate size Ag nanoparticles is attributed to the effect of increased excitation rate from the locally enhanced electromagnetic field by interaction of incident light with the nanoparticles and a higher quantum yield from an increase of the intrinsic decay rate of the fluorophore. One of the important advantages of substrates prepared in this way was macroscopic uniformity of the surface, unlike in many competing methods, due to monodisperse Au seed on the glass surface and isotopic growth of silver on the Au colloid surface. These important features will pave the way for application of such substrates in many areas of life sciences and biotechnology such as imaging and sensing applications based on enhanced fluorescence detection. Homogeneous core-shell nanoparticles presented here could significantly improve the sensitivity of fluorescence-based biological assays, and also could be utilized in high-throughput screening and drug discovery.14 Acknowledgment. F.X. would like to acknowledge financial support from International Macquarie University Research Scholarship and the Macquarie University Biotechnology Research Institute (MUBRI). The assistance of Ms. Katie McBean, University of Technology, Sydney, in SEM and EDX characterization is greatly appreciated.

J. Phys. Chem. B, Vol. 110, No. 46, 2006 23091 Supporting Information Available: Calculation of the total radiant flux emitted by the fluorophore layer covering our nanostructures. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Lakowicz, J. R. Probe Design and Chemical Sensing, Lakowicz, J. R., Ed.; Plenum Press: New York, 1994; Vol. 4. (2) Kronick, M. N. J. Immunol. Methods 1986, 92, 1. (3) Gosling, J. P. A. Cell 1990, 36, 1408. (4) Lo¨vgen, T.; Pettersson, K. Fluorescence Immunoassay and Molecular Applications; Van Dyke, K., Van Dyke, R., Eds.; CRC Press: New York, 1990; pp 234-250. (5) Casay, G. A.; Shealy, D. B.; Patonay, G. Probe Design and Chemical Sensing; Lakowicz, J. R., Ed.; Plenum Press: New York, 1994; Vol. 4. (6) He, L.; Musick, M. D.; Nicewarner, S. R.; Salinas, F. G.; Benkovic, S. J.; Natan, M. J.; Keating, C. D. J. Am. Chem. Soc. 2000, 122, 9071. (7) Near-Infrared Dyes for High Technology Applications; Daehne, S., Resch-Genger, U., Wolfbeis, O. S., Eds.; Kluwer Academic Publishers: New York, 1998. (8) Walker, N. J. Science 2002, 296, 557. (9) Drexhage, K. H. J. Lumin. 1970, 12, 693. (10) Ford, G. W.; Weber, W. H. Surf. Sci. 1981, 109, 451. (11) Lakowicz, J. R.; Geddes, C. D.; Gryczynski, I.; Malicka, J.; Gryczynski, Z.; Aslan, K.; Lukomska, J.; Matveeva, E.; Zhang, J.; Badugu, R.; Huang, J. J. Fluoresc. 2004, 14, 425. (12) Aslan, K.; Lakowicz, J. R.; Geddes, C. D. Anal. Bioanal. Chem. 2005, 382, 926. (13) Alsan, K.; Gryczynski, I.; Malicka, J.; Matveeva, E.; Lakowicz, J. R.; Geddes, C. D. Curr. Opin. Biotechnol. 2005, 16, 55. (14) Lakowicz, J. R. Anal. Biochem. 2005, 337, 171. (15) Kerker, M.; Blatchford, C. G. Phys. ReV. B 1982, 26, 4082. (16) Gersten, J. I.; Nitzan, A. Surf. Sci. 1985, 158, 165. (17) Lakowicz, J. R. Anal. Biochem. 2001, 298, 1. (18) Lakowicz, J. R.; Shen, Y.; D’Auria, S.; Malicka, J.; Fang, J.; Gryczynski, Z.; Gryczynski, I. Anal. Biochem. 2002, 301, 261. (19) Sokolov, K.; Chumanov, G.; Cotton, T. M. Anal. Chem. 1998, 70, 3898. (20) Geddes, C. D.; Cao, H.; Gryczynski, I.; Gryczynski, Z.; Fang, J.; Lakowicz, J. R. J. Phys. Chem. A 2003, 107, 3443. (21) Aslan, K.; Leonenko, Z.; Lakowicz, J. R.; Geddes, C. D. J. Phys. Chem. B 2005, 109, 3157. (22) Aslan, K.; Lakowicz, J. R.; Geddes, C. D. J. Phys. Chem. B 2005, 109, 6247. (23) Geddes, C. D.; Parfenov, A.; Roll, D.; Fang, J.; Lalowicz, J. R. Langmuir 2003, 19, 6236. (24) Geddes, C. D.; Parfenov, A.; Lakowicz, J. R. Appl. Spectrosc. 2003, 57, 526. (25) Frens, G. Nat. Phys. Sci. 1973, 241, 20. (26) Grabar, K. C.; Freeman, R. G.; Hommer, M. B.; Natan, M. J. Anal. Chem. 1995, 67, 735. (27) Doron, A.; Katz, E.; Willner, I. Langmuir 1995, 11, 1313. (28) Malicka, J.; Gryczynski, I.; Geddes, C. D.; Lakowicz, J. R. J. Biomed. Opt. 2003, 8 (3), 472. (29) Huang, T.; Murray, R. W. J. Phys. Chem. B 2001, 105, 12498. (30) Kruger, A. A. Surface and Near-Surface Chemistry of Oxide Materials; Nowotny, J., Dufour, L. C., Eds.; Elsevier Science Publishers: Amsterdam, 1988; pp 413-448. (31) Immunogold-silVer Staining: Principles, Methods, and Applications; Hayat, M. A., Ed.; CRC Press: Boca Raton, 1995. (32) Kreirib, U.; Vollmer, M. Optical Properties of Metal Clusters; Springer: Berlin, 1995; p 532. (33) Feldheim, D. L.; Foss, C. A. Metal Nanoparticles: Synthesis, Characterization and Applications; Marcel Dekker Inc.: New York, 2002; p 338. (34) Kerker, M.; Blatchford, C. G. Phys. ReV. B 1982, 26 (8), 4052. (35) Ygurabide, J.; Ygurabide, E. Anal. Biochem. 1998, 262, 137. (36) Ygurabide, J.; Ygurabide, E. Anal. Biochem. 1998, 262, 157. (37) Messinger, B. J.; Raben, U.; Chang, R. K.; Barber, P. W. Phys. ReV. B 1981, 24 (2), 649. (38) Lakowicz, J. R.; Malicka, J.; D’Auria, S.; Gryczynski, I. Anal. Biochem. 2003, 320, 13. (39) Moskovits, M.; Srnova-Sloufova, I.; Vickova, B. J. Chem. Phys. 2002, 116, 23.