ARTICLE pubs.acs.org/JPCC
Correlating Molecular Surface Coverage and Solution-Phase Nanoparticle Concentration to Surface-Enhanced Raman Scattering Intensities Marie Carmelle S. Pierre,† Prescott M. Mackie,† Maryuri Roca,‡ and Amanda J. Haes* Department of Chemistry, University of Iowa, Iowa City, Iowa 52242, United States
bS Supporting Information ABSTRACT: Control over the composition, shape, size, stability, and local dielectric environment of solution-phase metallic substrates is vital to consistent surface-enhanced Raman scattering (SERS) signals. Because of their inherent instability, solution-phase nanoparticles can undergo uncontrolled aggregation when target molecules are added. Here, we demonstrate that both molecular surface coverage of the Raman active molecule, 2-naphthalenethiol (2-NT), and nanoparticle concentration are critical parameters for obtaining reproducible SERS signals using solution-phase gold nanoparticles. Both gold nanoparticle and 2-naphthalenethiol concentrations are varied, and the extinction of the nanoparticle substrate and the SERS intensity of the target molecule are monitored as a function of time. These results indicate that extinction and SERS spectral intensities increase predictably below full monolayer surface coverage. When excess molecules are added, uncontrolled and irreproducible nanoparticle aggregation leads to optimal overlap between the plasmonic properties of the nanoparticles and the SERS excitation wavelength. Importantly, this is the first report which correlates solutionphase nanoparticle concentration and stability to molecular surface coverage for simultaneous localized surface plasmon resonance (LSPR) and SERS spectroscopic measurements. As a result, these data should facilitate the experimental design and use of solutionphase SERS substrates for more predictable molecular detection.
’ INTRODUCTION Gold nanoparticles exhibit unique extinction spectra which arise from their localized surface plasmon resonance (LSPR),1,2 an optical phenomenon not observed in bulk material.36 LSPR spectra are dependent on many parameters including nanoparticle shape, size, local dielectric environment, stability, and surface chemistry.79 Control over these parameters leads to predictable optical properties thereby allowing for their integration into many applications including: optical filters,10,11 photon energy transport,12,13 biological diagnostics,14,15 lasers,16,17 and surface-enhanced Raman scattering (SERS) detection.1821 The localized electromagnetic fields accompanying LSPR excitation are a key factor in the intense signals observed in all surface-enhanced spectroscopies including SERS.22,23 SERS, a spectroscopic technique where the intensity of the normal Raman effect is increased up to 9 orders of magnitude,2428 is used to improve detection. SERS is generally attributed to a combination of at least two major contributions. The largest factor of SERS enhancements depends on the LSPR or strong electromagnetic (EM) fields that extend several nanometers away from a nanostructure.35,26,2833 Although still debated,34,35 SERS enhancements can be maximized when the excitation wavelength is tuned to overlap with the LSPR.36,37 The second contribution r 2011 American Chemical Society
to SERS enhancements is a short-range chemical (CHEM) effect which arises from the electronic coupling of charge transfer between and adsorption of a molecule on a nanostructure.3840 Both mechanisms are hypothesized to broaden the molecular orbitals of the adsorbate which subsequently overlap with the LSPR of the nanostructures.35,41,42 Because SERS provides sensitive and specific molecular information in aqueous media, detection of biological43,44 and environmental45,46 species is often targeted. Despite their widespread use, applications which utilize solution-phase metal nanoparticles18,4749 for reproducible and quantifiable SERS responses remain limited because of unstable and unpredictable short and long-range interactions between nanoparticles in solution upon target molecule addition.50,51 Analysis is further complicated when solution-phase nanoparticles are used because large clusters form52 and uncontrolled nanoparticle aggregation occurs. Furthermore, varied surface charges can result when the pH, ionic strength, or molecular composition of the surrounding solution is modified.53,54 While quantitative Received: July 1, 2011 Revised: August 9, 2011 Published: August 21, 2011 18511
dx.doi.org/10.1021/jp206243y | J. Phys. Chem. C 2011, 115, 18511–18517
The Journal of Physical Chemistry C
ARTICLE
Figure 1. Structural and optical characterization of gold nanoparticles: (A) TEM image of the gold nanoparticles used (d = 13.2 ( 1.3 nm); (B) extinction and (C) corresponding SERS spectra for 4 nM gold nanoparticles incubated with 6 μM 2-NT are shown over a 1 h period; (D) comparison of the (black square) SERS intensity at 1064 cm1 and (red circle) extinction at 785 nm and are plotted vs time. SERS parameters: tint = 10 s, λex = 785 nm, P = 52 mW.
SERS detection platforms were previously reported for gold nanoparticles,5558 nanoparticle extinction and SERS spectra were not simultaneously reported. As a result, microscopy,59 microfluidics,59 and internal standards56,60 were previously used to quantify SERS measurements. In this study, we demonstrate that the molecular surface coverage of the target molecule 2-naphthalenethiol (2-NT) and known gold nanoparticle concentrations are critical in generating correlated extinction and SERS signals using solutionphase nanostructures. Gold nanoparticle and 2-NT concentrations will be varied while nanoparticle extinction and SERS signal intensities of the molecule are monitored. Three-dimensional (3D) contour plots will demonstrate that the behavior of extinction and SERS intensities is correlated when molecular surface coverage is less than an estimated61 monolayer surface coverage. Above this value, large but less reproducible SERS signal intensities are observed. This report is the first study in which molecular surface coverage on solution-phase gold nanoparticles is correlated with simultaneous LSPR and SERS measurements and should facilitate the predictable detection of other target molecules when quantitative SERS detection is required.
’ EXPERIMENTAL METHODS Reagents and Chemicals. Gold(III) chloride trihydrate (HAuCl4 3 3H2O), sodium chloride (NaCl), and 2-NT,(99%) were purchased from Sigma (St. Louis, MO). All other chemicals were purchased from Fisher Scientific (Pittsburgh, PA). Water (18.2 MΩ cm1) was obtained using a Nanopure System from
Barnstead (Dubuque, IA). For all experiments, glassware was cleaned with aqua regia (3:1 HCl:HNO3), rinsed thoroughly with water, and oven-dried overnight before use. Gold Nanoparticle Synthesis. Gold nanoparticles were synthesized via citrate reduction.50,62 Briefly, a 1 mM HAuCl4 3 3H2O solution (50 mL) was refluxed and vigorously stirred for 15 min. Once a rolling boil was achieved, 5 mL of a 39 mM trisodium citrate solution was quickly added. The solution was refluxed for an additional 10 min and allowed to equilibrate to room temperature while stirring. The resulting average gold nanoparticle diameter (d) was 13.2 ( 1.3 nm as determined from transmission electron microscopy. Transmission Electron Microscopy (TEM). TEM was performed using a JEOL JEM-1230 microscope equipped with a Gatan CCD camera. Samples were prepared on 400 mesh copper grids coated with a thin film of Formvar and carbon (Ted Pella). The nanoparticle solution was diluted in a 50% waterethanol mixture. The solution (∼50 μL) was pipetted onto a grid and promptly drained using filter paper. Over 600 nanoparticles (N) were analyzed using Image Pro to estimate the mean nanoparticle diameter. Propagated error in the nanoparticle surface area is estimated from TEM data. Sample Preparation. Au nanoparticles (15.0 nM) and 2.65 mM 2-NT (in ethanol) stock solutions were prepared and then diluted to 50 μM in water. The citrate concentration was maintained at 0.25 mM for all measurements. Upon addition of 2-NT and nanoparticles into a cuvette, the mixture was briefly vortexed, and extinction and SERS spectra were collected every 10 s and 1 min, respectively, for 1 h. Gold nanoparticle 18512
dx.doi.org/10.1021/jp206243y |J. Phys. Chem. C 2011, 115, 18511–18517
The Journal of Physical Chemistry C
ARTICLE
concentration was estimated63 via extinction spectroscopy using an extinction coefficient (ε520nm) = 2 108 M1 3 cm1. Simultaneous Extinction and SERS Spectroscopy. Extinction and SERS measurements were performed simultaneously using a modified sample holder and 1 cm quartz cuvette. LSPR spectra were collected near the top of the cuvette using an UVvis spectrometer (HR4000 or USB4000, Ocean Optics). All spectra were baseline-adjusted at ∼450 nm to account for white light source variations. Raman scattering was monitored near the base of the cuvette with excitation occurring perpendicular to the extinction light source using an Examiner785 (DeltaNu) and the following parameters: excitation wavelength (λex) = 785 nm, integration time (tint) = 10 s, and laser power (P) = 52 mW. All spectra were background corrected using samples that contained only nanoparticles (no 2-NT). Random XYZ and Renka-Cline matrix conversion methods (Origin 7.5) were used to interpolate and generate 3D extinction and SERS plots. Only spectral intensities greater than 3 times the spectral noise were used in analysis. Spectral noise was estimated as the standard deviation of the baseline from 1900 to 1700 cm1.
’ RESULTS AND DISCUSSION Time-Dependent Extinction and SERS Spectral Responses.
The SERS phenomenon originates from interactions between Raman-active vibrational modes of a molecule and the chemical and electromagnetic properties of noble metal nanoparticles.22,23,34,35,58 Typically, SERS signals increase as nanoparticle aggregation occurs which induces heightened electromagnetic fields in gaps between nanostructures.64 Solution-phase nanostructures are intrinsically unstable in the presence of target molecules which can be considered as either a positive effect given increased SERS signals or a negative effect as nanoparticle aggregation is difficult to control thereby limiting the reproducibility and/or quantitative capabilities of SERS intensities for signal transduction. Both of these experimental observations are related to and observed via time-dependent changes in the electromagnetic properties of the nanoparticles. An example of these variations is shown in Figure 1. A 4 nM gold nanoparticle (d = 13.2 ( 1.3 nm) solution incubated with 6 μM 2-NT exhibits time-dependent extinction (in arbitrary units (AU), Figure 1B) and SERS (in arbitrary density units (ADU) per milliwatt per second, Figure 1C) spectral responses. The extinction maximum wavelength (λmax) is centered at 520.4 nm then red-shifts and decreases in magnitude as a new lower energy extinction band develops with increasing time (Figure 1B) which indicates molecule-induced nanoparticle aggregation. As this occurs, the SERS signals from 2-NT increases with increasing time (Figure 1C). Because 2-NT quickly displaces the stabilizing citrate molecules from the nanoparticle surfaces before the plasmonic properties of the nanoparticles overlap with the Raman excitation wavelength, citrate is not observed in any of the SERS spectra. Correlated trends are observed between low energy extinction and SERS vibrational bands. Both extinction and SERS signals are dynamic and increase steadily over the first ∼15 min of the nanoparticlemolecule incubation period. For instance, time dependent changes in extinction at 785 nm and the 2-NT vibrational mode at νvib = 1064 cm1 increase then slightly decrease (Figure 1D). ̅ These plasmonic trends are attributed to nanoparticle aggregation and sedimentation, respectively. 2-NT
Figure 2. 2-NT concentration and time-dependent spectral changes for 4 nM gold nanoparticles. (A) Extinction at 785 nm and (B) SERS intensity at 1064 cm1 for (black square) 0.5 μM, (green circle) 2 μM, (dark blue up triangle) 4 μM, (light blue down triangle) 6 μM, (red diamond) 8 μM, and (yellow left triangle) 11 μM 2-NT are included.
displaces the stabilizing citrate molecules on the nanoparticle surface thereby destabilizing the nanoparticles via double layer disruption which facilitates additional (1) 2-NT binding to the nanoparticle surface, (2) nanoparticle aggregation, and consequently (3) an increase in both low energy extinction and SERS intensities. Finally, after the extinction and SERS signals maximize, both signals slightly decrease with increasing time. After ∼2 h, the magnitude of the signals decrease dramatically, and nanoparticle aggregates are visibly observed at the bottom of the sample cell (data not shown). To evaluate how extinction and SERS intensities vary with target molecule concentration, a 4 nM gold nanoparticle solution is mixed with 2-NT concentrations varying from 0.5 to 11 μM. Changes in extinction at the initial LSPR maximum wavelength, the Raman excitation wavelength, and at an energy between the Raman excitation and molecular vibrational energies are tracked at λ = 520 (Figure S1, Supporting Information), 785 (Figures 2A and 3B), and 821 nm (Figure S2, Supporting Information), respectively. Comparisons between these wavelengths and SERS intensities at 1064 cm1 (Figure 2B) as a function of both time and 2-NT concentration were subsequently evaluated. As expected, extinction at 520 nm decreases with increasing time for all 2-NT concentrations. In contrast, extinction at both 785 and 821 nm increases with increasing 2-NT concentrations for a fixed incubation time. Because extinction trends are similar at the two lower energy wavelengths, extinction at 785 nm was chosen for evaluation in all subsequent studies. Importantly, extinction at 785 nm and SERS intensities systematically increase and saturate more quickly than similar experiments performed with lower 2-NT concentrations. Notably, these trends are reproducible when 4 nM gold nanoparticle solutions are incubated with less than 6 μM 2-NT concentrations. Above 6 μM 2-NT concentrations, similar trends in extinction and SERS intensities are 18513
dx.doi.org/10.1021/jp206243y |J. Phys. Chem. C 2011, 115, 18511–18517
The Journal of Physical Chemistry C
ARTICLE
Figure 3. 3D contour maps of extinction at λ = 785 nm vs time for (A) 2, (B) 4, (C) 6, and (D) 8 nM gold nanoparticle concentrations. The black line represents theoretically calculated 100% 2-NT surface coverage on gold nanoparticles.
observed but the magnitude and rate of these changes are inconsistent (data not shown). Extinction and SERS Spectral Responses as a Function of Surface Coverage. To evaluate how nanoparticle concentration impacts these measurements, both 2-NT concentration and time dependent extinction and SERS changes were monitored. Nanoparticle concentration was estimated using the method developed by Haiss.63 Next, both gold nanoparticle and 2-NT concentrations are varied systematically and monitored as a function of time. For each nanoparticle concentration, 2-NT concentrations were selected to provide varying degrees of citrate displacement from the nanoparticle surface. Assuming all 2-NT molecules bind to the nanoparticle surface, molecular surface coverage is estimated using the equation % surface coverage ¼
SA 2-NT C2-NT 100% SA NP CNP
ð1Þ
where SA2-NT is the surface area occupied by a 2-NT molecule (0.244 nm2/molecule),61 SANP is the available surface area per nanoparticle (i.e., SANP = 4π(13.2 nm/2)2 = 547 ( 107 nm2/ nanoparticle), and C2-NT and CNP are 2-NT and nanoparticle concentrations, respectively. It should be noted that the 2-NT footprint will vary with orientation and packing density and that each nanoparticle exhibits a unique crystal structure and geometry (i.e., distinct surface area). As a result, this molecular coverage estimation represents only an average molecular surface coverage on all nanoparticles in solution. As a result, each 2-NT concentration yields an estimated molecular surface coverage if the nanoparticle concentration is known.
Time-dependent trends in extinction at 785 nm and SERS intensities at 1064 cm1 for each nanoparticle concentration (A = 2 nM, B = 4 nM, C = 6 nM, and D = 8 nM) are found in Figures 3 and 4, respectively. These information-rich plots summarize how spectral signals at a particular energy vary with 2-NT concentration and time. Specifically, 3D contour maps are included where the x axis represents time, the y axes represent molecular concentration (left) and estimated surface coverage (right), and the colors represent either extinction at λ = 785 nm or SERS intensity at νvib = 1064 cm1. Each (fixed nanoparticle ̅ concentration) map represents spectral evaluation of at least seven concentrations of 2-NT collected for 60 min. In all contour maps, a solid black line is included which represents a monolayer coverage of 2-NT and depends on nanoparticle concentration. As a result, the nanoparticle surface area will saturate when 4.5, 9.0, 13.5, and 18.0 μM 2-NT is added to and equilibrated with 2, 4, 6, and 8 nM gold nanoparticle solutions, respectively. Several trends are observed in both extinction (Figure 3) and SERS intensities (Figure 4). First, spectral intensities increase more rapidly over the first 510 min period as 2-NT concentration increases (for each fixed nanoparticle concentration) vs later times. Below the theoretical 2-NT monolayer coverage on the gold nanoparticles, spectral intensities slowly increase then saturate for 2 nM gold nanoparticle concentrations (Figures 3A and 4A). Additional increases in 2-NT concentration yield similar behaviors as a function of time. Likewise, the extinction and SERS intensities from 4 to 8 nM gold nanoparticle concentrations (Figure 3BD and Figure 4BD) saturate in ∼10 min for all 2-NT concentrations. Spectral maxima are observed at 100% 2-NT surface coverage for both extinction and SERS measurements for all nanoparticle 18514
dx.doi.org/10.1021/jp206243y |J. Phys. Chem. C 2011, 115, 18511–18517
The Journal of Physical Chemistry C
ARTICLE
Figure 4. 3D contour maps of SERS intensities at νvib = 1064 cm1 vs time for (A) 2, (B) 4, (C) 6, and (D) 8 nM gold nanoparticle concentrations. The black line represents theoretically calculated 100% ̅ 2-NT surface coverage on gold nanoparticles.
concentrations studied. At or above monolayer 2-NT surface coverages, the maximized extinction and SERS signals slowly decrease as time progresses. These effects are most apparent spectrally and visibly at the highest nanoparticle concentration (Figures 3D and 4D) as nanoparticles settle out of solution. Of note, extinction signals reach maximum values more slowly and persist for longer time periods than that for the SERS data thereby indicating that the SERS signals are likely dominated by large nanoparticle aggregates that form quickly and then settle out of solution while extinction results from the average optical response by all nanoparticles in solution. Nanoparticle collisions, and as a result, nanoparticle aggregates are more likely to form when the nanoparticle concentration increases and the double layer is disrupted by the thiolated molecules. Clearly, both extinction and SERS intensities increase with time and 2-NT concentration and then saturate at ∼100% surface coverage before decreasing. At a fixed 2-NT concentration and varying nanoparticle concentration, extinction and SERS data reveal several distinct trends. First, Figures 3 and 4 show that when a fixed 2-NT concentration is incubated with varying nanoparticle concentrations, extinction and SERS data (at a fixed time) systematically respond. For instance, 4 μM 2-NT causes the extinction at 785 nm for 2, 4, 6, and 8 nM nanoparticle concentrations to saturate at ∼0.40, 0.30, 0.28, and 0.24 AU, respectively. Noticeably, the saturated extinction values are inversely proportional to nanoparticle concentration. Importantly, this response is directly proportional to 2-NT surface coverage, which is ∼90, 45, 30, and 22% for the respective nanoparticle concentrations. If nanoparticle aggregation is promoted when 2-NT binds to the gold surfaces, the probability of
nanoparticle aggregation and spectral intensities should increase as molecular surface coverage increases. Decreasing nanoparticle concentration for a fixed 2-NT concentration does just this. As a result, spectral intensities decrease as nanoparticle concentration increases for a fixed 2-NT concentration. For instance, 4 μM 2-NT produces SERS signal intensities of ∼80, 55, 45, and 40 ADU 3 mW1 3 s1 for 2 (90%), 4 (45%), 6 (30%), and 8 (22%) nM nanoparticle concentrations (surface coverages), respectively. The probability of double layer disruption decreases as nanoparticle concentration increases for a fixed 2-NT concentration. This reduces the likelihood of nanoparticle aggregation which limits satisfaction of the mechanisms required for SERS. In contrast, the largest SERS signals are observed at 130, 155, 195, and 220 ADU 3 mW1 3 s1 for 2, 4, 6, and 8 nM nanoparticle concentrations, respectively. In each case, this maximum value corresponds to 2-NT concentrations which should form a full 2-NT monolayer because of the strong binding affinity between sulfur and the gold nanoparticle surfaces. When additional 2-NT is incubated with each nanoparticle solution, trends in SERS signal changes are inconsistent. In summary, SERS signals systematically vary when low molecular surface coverages are used thereby improving the usefulness of solution-phase nanoparticles in quantitative SERS assays. Correlated Three-Dimensional Spectral Maps. 3D maps allow for straightforward comparisons and correlations between parameters that affect spectral signals. When sufficient 2-NT is added to saturate the nanoparticle surface (100% surface coverage), an important result from these data reveals that extinction values maximize in ∼5, 10, 15, and 20 min while SERS intensities peak after ∼7, 5, 4, and 2 min for 2, 4, 6, and 8 nM nanoparticle 18515
dx.doi.org/10.1021/jp206243y |J. Phys. Chem. C 2011, 115, 18511–18517
The Journal of Physical Chemistry C
ARTICLE
instance, the extinction and SERS responses collected at 5 min for 6 nM gold nanoparticles increase with increasing 2-NT concentration. Once monolayer surface coverage is exceeded, both spectral responses decrease. Similar behavior is observed for all nanoparticle concentrations. This response demonstrates that as 2-NT surface coverage increases, citrate molecules are quickly displaced and signal changes are dominated by 2-NT surface coverage and/or slight plasmonic variations. Above monolayer surface coverages of 2-NT, both extinction and SERS signals maximize then decrease suggesting electromagnetic coupling between the solution-phase nanostructures which can settle out of solution. Furthermore, trends in nanoparticle concentration are also apparent. For a particular submonolayer 2-NT surface coverage addition, both extinction and SERS data are largely independent of nanoparticle concentration. At higher 2-NT concentrations, both extinction and SERS data increase systematically thereby once again suggesting that nanostructures are undergoing aggregation which facilitates electromagnetic coupling between nanoparticles.
Figure 5. 3D contour maps correlating (A) extinction at 785 nm and (B) SERS intensity at 1064 cm1 for 2-NT as a function of both gold nanoparticle and 2-NT concentrations (incubation time = 5 min). The solid black lines represent estimated monolayer surface coverage of 2-NT on the surface of gold nanoparticles. The error in nanoparticle diameter was used to propagate the uncertainty of these values (range shown with dashed lines). On average, surface coverage exceeds a monolayer to the right and below the solid line.
concentrations, respectively. We hypothesize that trends in extinction data are dictated by molecular (2-NT) concentration, and as a result, the average distance between and double layer thickness surrounding the nanoparticles while SERS signals is dominated by quickly forming nanoparticle aggregates which efficiently settle out of solution. To test our hypothesis, extinction and SERS intensities at a fixed time (5 min) were plotted in 3D contour maps for discrete energies in extinction and SERS spectral responses (represented by the color scale) (panels A and B of Figure 5, respectively) to evaluate correlated trends between nanoparticle (y axis) and molecule (x axis) concentrations. These plots were generated using at least seven concentrations of 2-NT (x axis) vs four nanoparticle concentrations (y axis) with colors representing spectral magnitudes. A 5 min incubation time was selected as it correlates with the average time required to observe maximized SERS intensities. In both contour maps, the ∼100% surface coverage (solid line) and propagated error in nanoparticle surface area (dashed line) are included. Surface coverage exceeds a monolayer surface coverage to the right and below the solid line. Figure 5 shows that both extinction at λ = 785 nm and SERS intensities at νvib = 1064 cm1 increase linearly with increasing ̅ 2-NT concentration (fixed nanoparticle concentration) when the theoretical surface coverage is less than a monolayer. For
’ CONCLUSIONS Solution-phase nanoparticles are commonly used as SERS substrates. Despite their extensive use in various applications, the ability to use solution-phase nanoparticles as quantifiable SERS substrates is limited by the inherent instability of the nanoparticles, the homogeneity (or lack thereof) of nanoparticle morphology, and/or unknown/unreported nanoparticle concentrations. Herein, we demonstrated that nanoparticle aggregation and as a result extinction and SERS spectral intensities are impacted by molecular surface coverage, a variable which depends on nanoparticlemolecule interactions as well as on both nanoparticle surface area and plasmonic changes. By plotting extinction and SERS intensities as a function of both nanoparticle and 2-NT concentrations, correlated extinction and SERS intensities were observed at submonolayer surface coverages. Above this value, nanoparticles aggregated rapidly and uncontrollably when the molecule interacted with the nanoparticle surface which induced electromagnetic coupling between nanostructures—a result which thereby led to relatively larger but unpredictable spectral intensities. These data indicate that relative nanoparticle surface area and molecular surface coverages are key parameters in evaluating both plasmonic and quantitative SERS assays. Further exploitation of these phenomena should lead to more predictable SERS signal for quantifiable molecular detection. ’ ASSOCIATED CONTENT
bS
Supporting Information. Two- and three-dimensional extinction data at 520 and 821 nm for 4 nM gold nanoparticles incubated with varying 2-NT concentrations. This information is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. Present Addresses ‡
Lawrence University.
Author Contributions †
Contributed equally.
18516
dx.doi.org/10.1021/jp206243y |J. Phys. Chem. C 2011, 115, 18511–18517
The Journal of Physical Chemistry C
’ ACKNOWLEDGMENT The authors gratefully acknowledge financial support from the Department of the Navy, Office of Naval Research (YIP Award #N00014-07-1-0827). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the Office of Naval Research. ’ REFERENCES (1) McDonnell, J. M. Curr. Opin. Chem. Biol. 2001, 5, 572–577. (2) Lyon, L. A.; Musick, M. D.; Natan, M. J. Anal. Chem. 1998, 70, 5177–5183. (3) Haynes, C. L.; Van Duyne, R. P. J. Phys. Chem. B 2001, 105, 5599–5611. (4) Link, S.; El-Sayed, M. A. J. Phys. Chem. B 1999, 103, 8410–8426. (5) Kreibig, U.; Gartz, M.; Hilger, A.; Hovel, H. Optical Investigations of Surfaces and Interfaces of Metal Clusters. In Advances in Metal and Semiconductor Clusters; Duncan, M. A., Ed.; JAI Press Inc.: Stamford, CT, 1998. (6) Schatz, G. C.; Young, M. A.; Van Duyne, R. P. Top. Appl. Phys. 2006, 103, 19–46. (7) Hutter, E.; Fendler, J. H. Adv. Mater. 2004, 16, 1685–1706. (8) Ivanov, M. R.; Bednar, H. R.; Haes, A. J. ACS Nano 2009, 3, 386–394. (9) Haes, A. J.; Van Duyne, R. P. Anal. Bioanal. Chem. 2004, 379, 920–930. (10) Flaugh, P. L.; O’Donnell, S. E.; Asher, S. A. Appl. Spectrosc. 1984, 38, 847–850. (11) Spry, R. J.; Kosan, D. J. Appl. Spectrosc. 1986, 40, 782–784. (12) Merschdorf, M.; Kennerknecht, C.; Willig, K.; Pfeiffer, W. New J. Phys. 2002, 4, 1–15. (13) Liao, C.-Y.; Wang, H. P.; Chen, F. L.; Huang, C. H.; Fukushima, Y. Adv. Material Res. 2008, 4750, 11211124. (14) El-Sayed, I.; Huang, X.; El-Sayed, M. Nano Lett. 2005, 5, 829–834. (15) Huang, X.; Jain, P. K.; El-Sayed, I. H.; El-Sayed, M. A. Nanomedicine 2007, 2, 681–693. (16) Taira, S.; Kitajima, K.; Katayanagi, H.; Ichiishi, E.; Ichiyanagi, Y. Sci. Technol. Adv. Mater. 2009, 10, 1–6. (17) Pustovalov, V. K.; Babenko, V. A. Laser Phys. Lett. 2004, 1, 516–520. (18) He, X.; Zhao, X.; Chen, Y.; Feng, J. J. Mater. Charact. 2008, 59, 380–384. (19) Lu, L.; Sun, G.; Zhang, H.; Wang, H.; Xi, S.; Hu, J.; Tian, Z.; Chen, R. J. Mater. Chem. 2004, 14, 1005–1009. (20) Yan, B.; Thubagere, A.; Premasiri, W. R.; Ziegler, L. D.; Dal Negro, L.; Reinhard, B. M. ACS Nano 2009, 3, 1190–1202. (21) Sun, L.; Song, Y.; Wang, L.; Guo, C.; Sun, Y.; Liu, Z.; Li, Z. J. Phys. Chem. C 2008, 112, 1415–1422. (22) Schatz, G. C.; Van Duyne, R. P. Electromagnetic Mechanism of Surface-Enhanced Spectroscopy; Wiley: New York, 2002. (23) Haes, A. J.; Haynes, C. L.; McFarland, A. D.; Schatz, G. C.; Van Duyne, R. P.; Zou, S. MRS Bull. 2005, 30, 368–375. (24) Jeanmaire, D. L.; Van Duyne, R. P. J. Electroanal. Chem. Interfacial Electrochem. 1977, 84, 1–20. (25) Nie, S.; Emory, S. R. Science 1997, 275, 1102–1106. (26) Xu, H.; Aizpurua, J.; Kall, M.; Apell, P. Phys. Rev. E 2000, 62, 4318–4324. (27) Chaney, S. B.; Shanmukh, S.; Dluhy, R. A.; Zhao, Y. P. Appl. Phys. Lett. 2005, 87, 031908–031909. (28) Fang, J. X.; Yi, Y.; Ding, B. J.; Song, X. P. Appl. Phys. Lett. 2008, 92, 131113. (29) Mulvaney, P. MRS Bull. 2001, 26, 1009–1014. (30) El-Sayed, M. A. Acc. Chem. Res. 2001, 34, 257–264. (31) Mulvaney, P. Langmuir 1996, 12, 788–800.
ARTICLE
(32) Kreibig, U. Optics of Nanosized Metals. In Handbook of Optical Properties; Hummel, R. E., Wissmann, P., Ed.; CRC Press: Boca Raton, FL, 1997. (33) Hulteen, J. C.; Treichel, D. A.; Smith, M. T.; Duval, M. L.; Jensen, T. R.; Van Duyne, R. P. J. Phys. Chem. B 1999, 103, 3854–3863. (34) Michaels, A. M.; Nirmal, M.; Brus, L. E. J. Am. Chem. Soc. 1999, 121, 9932–9939. (35) Kinnan, M. K.; Chumanov, G. J. Phys. Chem. C 2007, 111, 18010–18017. (36) Haynes, C. L.; Van Duyne, R. P. Mater. Res. Soc. Symp. Proc. 2002, 728, 217–222. (37) McFarland, A. D.; Young, M. A.; Dieringer, J. A.; VanDuyne, R. P. J. Phys. Chem. B 2005, 109, 11279–11285. (38) Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Chem. Rev. 1999, 99, 2957–2976. (39) Liu, Y. C. Langmuir 2002, 18, 174–181. (40) Fromm, D. P.; Sundaramurthy, A.; Kinkhabwala, A.; Schuck, P. J.; Kino, G. S.; Moerner, W. E. J. Chem. Phys. 2006, 124, 061101. (41) Campion, A.; Kambhampati, P. Chem. Soc. Rev. 1998, 27, 241–250. (42) Daniels, J. K.; Chumanov, G. J. Phys. Chem. B 2005, 109, 17936–17942. (43) Hudson, S. D.; Chumanov, G. Anal. Bioanal. Chem. 2009, 394, 679–686. (44) Porter, M. D.; Lipert, R. J.; Siperko, L. M.; Wang, G.; Narayanana, R. Chem. Soc. Rev. 2008, 37, 1001–1011. (45) Lee, D.; Lee, S.; Seong, G. H.; Choo, J.; Lee, E. K.; Gweon, D. G. Appl. Spectrosc. 2006, 60, 373–377. (46) De Jesus, M. A.; Giesfeldt, K. S.; Sepaniak, M. J. J. Raman Spectrosc. 2004, 35, 895–904. (47) Zhang, J.; Li, W.; Cheng, Y.; Zhang, X.; Huang, S.; Ma, H. Mater. Chem. Phys. 2010, 119, 188–194. (48) Jin, R.; Cao, Y. C.; Hao, E.; Metraux, G. S.; Schatz, G. C.; Mirkin, C. A. Nature 2003, 425, 487–490. (49) Okada, N.; Hamanaka, Y.; Nakamura, A.; Pastoriza-Santos, I.; Liz-Marzan, L. M. J. Phys. Chem. 2004, 108, 8751–8755. (50) Roca, M.; Haes, A. J. J. Am. Chem. Soc. 2008, 130, 14273–14279. (51) Schwartzberg, A. M.; Grant, C. D.; Wolcott, A.; Talley, C. E.; Huser, T. R.; Bogomolni, R.; Zhang, J. Z. J. Phys. Chem. B 2004, 108, 19191–19197. (52) Yoon, J. H.; Park, J. S.; Yoon, S. Langmuir 2009, 25, 12475–12480. (53) Saleh, N.; Kim, H.-J.; Phenrat, T.; Matyjaszewski, K.; Tilton, R. D.; Lowry, G. V. Environ. Sci. Technol. 2008, 42, 3349–3355. (54) Jolivet, J. P.; Vayssieres, L.; Chaneac, C.; Tronc, E. MRS Proc. 1997, 432, 145–150. (55) Stokes, R. J.; McBride, E.; Wilson, C. G.; Girkin, J. M.; Smith, W. E.; Graham, D. Appl. Spectrosc. 2008, 62, 371–376. (56) Bell, S. E. J.; Mackle, J. N.; Sirimuthu, N. M. S. Analyst 2005, 130, 545–549. (57) Ryu, K.; Haes, A. J.; Park, H.-Y. P.; Nah, S. N.; Kim, J. K.; Chung, H.; Yoon, M.-Y.; Han, S.-H. J. Raman Spectrosc. 2010, 41, 121–124. (58) Creighton, J. A.; Blatchford, C. G.; Albrecht, M. G. J. Chem. Soc., Faraday Trans. 2 1979, 75, 790–798. (59) Chan, J.; Fore, S.; Wachsman-Hogiu, S.; Huser, T. Laser Photonics Rev. 2008, 2, 325–349. (60) Bell, S. E. J.; Sirimuthu, N. M. S. Chem. Soc. Rev. 2008, 37, 1012–1024. (61) Kolega, R. R.; Schlenoff, J. B. Langmuir 1998, 14, 5469–5478. (62) Grabar, K. C.; Freeman, R. G.; Hommer, M. B.; Natan, M. J. Anal. Chem. 1995, 67, 735–743. (63) Haiss, W.; Thanh, N. T. K.; Aveyard, J.; Fernig, D. G. Anal. Chem. 2007, 79, 4215–4221. (64) McMahon, J. M.; Henry, A. I.; Wustholz, K. L.; Natan, M. J.; Freeman, R. G.; Van Duyne, R. P.; Schatz, G. C. Anal. Bioanal. Chem. 2009, 394, 1819–1825. 18517
dx.doi.org/10.1021/jp206243y |J. Phys. Chem. C 2011, 115, 18511–18517
