Broad Band Enhancement of Light Absorption in Photosystem I by

May 18, 2010 - ... Using Spinach Leaves Extract in Ethanol. Marc P. Y. Desmulliez , David E. Watson , Jose Marques-Hueso , Jack Hoy-Gig Ng. 2016,71-78...
24 downloads 0 Views 2MB Size
Download PDF
pubs.acs.org/NanoLett

Broad Band Enhancement of Light Absorption in Photosystem I by Metal Nanoparticle Antennas Itai Carmeli,† Itai Lieberman,† Leon Kraversky,‡ Zhiyuan Fan,§ Alexander O. Govorov,*,§ Gil Markovich,*,† and Shachar Richter*,† †

Center for NanoScience and Nanotechnology and School of Chemistry, Tel Aviv University, Tel Aviv, Israel 69978, Center for NanoScience and Nanotechnology and Department of Biochemistry, Tel Aviv University, Tel Aviv, Israel 69978, and § Department of Physics and Astronomy, Ohio University, Athens, Ohio 45701 ‡

ABSTRACT The photosystem I (PS I) protein is one of nature’s most efficient light harvesting complexes and exhibits outstanding optoelectronic properties. Here we demonstrate how metal nanoparticles which act as artificial antennas can enhance the light absorption of the protein. This hybrid system shows an increase in light absorption and of circular dichroism over the entire absorption band of the protein rather than at the specific plasmon resonance wavelength of spherical metal nanoparticles (NPs). This is explained by broad-resonant and nonresonant field enhancements caused by metal NP aggregates, by the high dielectric constant of the metal, and by NP-PS I-NP antenna junctions which effectively enhance light absorption in the PS I. KEYWORDS Photosystem I, plasmon enhancement

P

plasmonic metal nanostructures13-16 and have theoretically predicted enhanced light absorption of the PS I assisted by metal nanoparticles.15 However, up to date, direct experimental demonstration of the absorption enhancement in large biomolecular objects such as the PS I was not reported.

hotosynthetic proteins such as the photosystem I (PS I) are nature’s most efficient light harvesting complexes which exhibit outstanding optoelectronic properties.1 The PS I active reaction-center is a chlorophyll protein complex located in thylakoid membranes of chloroplasts and cyanobacteria (Figure 1) which mediates a lightinduced electron transfer through a series of redox reactions. This efficient and robust system exhibits a quantum yield close to 100%, with an energy yield around 58%.1 Furthermore, its nanoscale dimension and the generation of a 1 V photovoltage2 qualifies the PS I molecule as a promising unit for applications in molecular optoelectronics, such as a protein based solar cell, a nano energy source to power nano electronic circuits, and an efficient sensor for light.3-7 However, despite its excellent quantum yield and its large photopotential generation, a single PS I monolayer absorbs a relatively small portion (∼1%) of the sunlight, which limits the PS I applicability. One of the suggested solutions to this problem is the use of surface plasmons to enhance the photoinduced signal. Several photophysical phenomena are already known to be enhanced by surface plasmons, such as surface enhanced Raman scattering,8,9 surface enhanced infrared absorption,10 and surface enhanced fluorescence.11 Lieberman et al. have recently observed surface plasmon resonance enhanced absorption in small molecules attached to silver nanoparticles.12 We and others have investigated the enhancement of absorption in photovoltaic systems by

FIGURE 1. Molecular structure scheme of the PS I attached to metal colloids. PS I is composed of polypeptide chains (gray) in which chlorophyll (green) and carotenoids (orange) are imbedded. The chromophores which mediate the electron transfer are represented by the space fill model (magenta). The PS I covalently binds to the metal colloids through Cys mutations along the polypeptide backbone (space fill model, yellow). Calculated local NP enhancement factors in the PS I regime are presented as background descending color gradient from red to light blue (see text and Supporting Information for detailed information).

* To whom correspondence should be addressed, [email protected], [email protected], and [email protected]. Received for review: 01/23/2010 Published on Web: 05/18/2010 © 2010 American Chemical Society

2069

DOI: 10.1021/nl100254j | Nano Lett. 2010, 10, 2069–2074

SCHEME 1. Illustration of the Chemical Route for the Synthesis of the PS I Metal Colloids Hybrid System and Its Detachment from the Metal Colloids by Mercaptopropionic Acid (MPA) Addition

Aldrich) aqueous solution, with final reaction volume of 2 mL. Several batches were prepared in a similar manner varying the rate of borohydride addition and PS I concentration to study the reproducibility of the method. A control sample of PS I suspension was prepared for comparison between the NP-PS I hybrid system and unbound PS I, by adding 2 µL of PS I stock solution to the buffer solution with a final volume of 2 mL. Figure 2 depicts typical transmission electron microscopy (TEM) images of the PS I capped silver and gold colloidal particles. The gold particles exhibited a broad size distribution ranging from 3 to 30 nm with the presence of aggregates (Figure 2a). The average size of the Au NPs was ∼20 nm. The Ag NP-PS I complexes formed a combination of small particles with an average size of ∼10 nm (Figure 2b) and unusually large aggregates (100-500 nm) with particularly interesting shapes (Figure 2c). The large aggregates exhibited a porous structure and ring-like shapes, where the voids were probably occupied by the PS I. Indeed, a closer examination of the TEM images (inset of Figure 2c) shows round cavities with the dimension of the PS I. These structures were probably formed by encapsulation the PS I during the metal reduction step (the PS I proteins themselves are transparent to TEM imaging). A control sample of metal nanopaticles produced in the absence of the PS I molecules immediately precipitated. This was an indication that adsorption of the PSI confined the growth of the colloidal particles and stabilized them in solution. A second control solution composed of PS I and borohydride (see Scheme 1) showed that the reducing agent had no observable affect on the absorption bands of the chlorophyll chromophores. In order to explore the optical properties of these hybrid systems, we have performed UV-vis absorption, CD, and fluorescence spectroscopy experiments. First, we have evaluated the PS I to metal binding efficiency using fluorescence measurements. The 680 nm emission band of a colloidal PS I and of the hybrid systems, excited at 400 nm, is displayed in Figure 3. It can be seen that binding of the PS I to the metal colloidal particles caused a substantial quenching of the emission by almost an order of magnitude. This observation indicates a high binding efficiency of the PS I to the metal nanoparticles as well as of an efficient energy transfer from

In addition to the enhancement of the effective absorption coefficient, it can be assumed that the circular dichroism (CD) signal, being proportional to the difference between the absorption coefficients of right- and left-handed circularly polarized light, should also be enhanced. To a first approximation, pure electromagnetic effects are involved in such a plasmon-assisted enhancement.12 However, molecular conformation changes might also be accompanied by CD enhancements that are not due to changes in the molecular absorption coefficient.17 In this work we demonstrate a technique to enhance light absorption of the PS I by the use of colloidal metal nanoparticles (NP) that act as optical antenna for the PS I. It is demonstrated that both gold and silver nanocrystals bound to PS I molecules are able to enhance the PS I light absorption over the entire absorption spectrum of the PS I. We postulate that this broad-band enhancement is due to efficient electromagnetic coupling between the PS I and NPs, and nonresonant field enhancement in NP-molecule-NP junctions. A model that supports the observed enhancement of the AuNP-PS I hybrid system is presented (a detailed description of the model can be found in the Supporting Information). Though direct covalent attachment of genetically engineered PS I to various surfaces has been achieved,3-5,18 the formation of PS I capped colloidal metal nanoparticles is more challenging due to requirements for colloidal stability. Therefore, in order to obtain a high yield attachment of the PS I to metal colloid particles, a new preparation method was developed. Instead of the conventional method of attaching the protein molecules to presynthesized metal nanoparticles, the metal capped PS I was prepared in a single synthesis step, where the metal ions were reduced in the presence of the protein molecules. Schematic drawings of the PS I metal colloid hybrid system and of its synthesis are shown in Figure 1 and Scheme 1, respectively. Specifically, to 20 mM tricine buffer solution, pH 7.7, bDM 5.0 × 10-5% by weight (n-dodecyl-β-D-maltoside, SigmaAldrich), 1 mM of AgNO3 or HAuCl4 (Sigma-Aldrich) was added. To this, 2 µL from ∼1 × 10-5 M PS I stock solution was introduced and the metal ions were reduced by slowly adding drops of 2 mM sodium borohydride (NaBH4, Sigma© 2010 American Chemical Society

2070

DOI: 10.1021/nl100254j | Nano Lett. 2010, 10, 2069-–2074

FIGURE 2. High-resolution TEM images of gold (a) and silver (b, c) PS I-NP hybrids. The Ag NPs complexes exhibit small particles with an average size of ∼10 nm (b) with the presence of unusual large aggregates (100-500 nm) with a porous structure (c).

FIGURE 3. Quenching of PS I fluorescence emission by colloidal metal NPs. Binding of the PS I to Ag (blue curve) and Au NPs (red curve) causes a decrease in emission relative to the pure PS I suspension (black curve) by factors of 5 and 7 for the Ag and Au PS I-NP hybrid systems, respectively. PS I concentration in all samples (NP-PS I hybrid systems and PS I suspension) was 1.5 nM.

the chlorophyll molecules distributed within the PS I to the metal particles. We have estimated the energy transfer time scales, τ, for the complexes, τAu ∼ τAg ∼ 0.5 ns, using a simple dipolar interaction model and assuming an average chromophoremetal particle separation for all chlorophyll molecules (for details, see Supporting Information). We see that these nonradiative lifetimes are shorter than the typical PS I fluorescence time scale (∼5 ns) by an order of magnitude.19 This estimation is consistent with the strong fluorescence quenching effect observed in our experiments on the metal NP-PS I system (Figure 3). Next, the absorption spectrum of the bare PS I was compared to that of the metal-SP I hybrid systems (Figure 4 parts a and b). In the case of the Au-NP, the featureless background absorption spectra resulting from NP Rayleigh scattering20,21 was subtracted from the spectra before estimating the enhancement factor by fitting a 1/λ4 function to the baseline of the absorption curves. In the case of the Ag-PS I hybrid system, no background was subtracted. © 2010 American Chemical Society

FIGURE 4. Light absorption enhancement of the PS I-Ag (a) and -Au (b) colloid hybrid system. Enhanced absorption of light by the hybrid system (black curves) is reduced (red curves) upon addition of MPA. Absorption spectra of a suspension of unbound PS I (blue curve) are drawn for comparison. PS I concentration in all samples (NP-PS I hybrid systems and PS I suspension) was 1.5 nM. (Inset) Absorption spectra of PS I suspension before (black curve) and after (red curve) addition of MPA (40 mM).

The typical absorption spectra of the chlorophyll molecules in the PS I consist of two distinct bands centered around 440 and 680 nm (Figure 4, blue curves).3 It can be seen that the binding of the PS I to the metal nanoparticles resulted in a clear enhanced absorption of light by the PS I. To confirm that the enhancement of light absorption was 2071

DOI: 10.1021/nl100254j | Nano Lett. 2010, 10, 2069-–2074

due to binding of the PS I to the NPs, large concentrations (40 mM) of mercaptopropionic acid (MPA) were added to the two samples. MPA was previously shown to partially replace other surfactant molecules attached to metal NPs (see Scheme 1); thus a substantial decrease of the metal enhanced absorption is expected.12 As can be seen in Figure 4, MPA addition caused the detachment of the PS I from the NPs which was accompanied by a clear decrease in light absorption for both Ag and Au colloid systems. In the case of the Au-PS I hybrid, the MPA stabilized the stripped Au NPs well and the Au plasmon band minimally changed after the MPA treatment, while in the case of the Ag NPs the MPA treatment always caused a substantial destabilization of the colloid and very large Ag NP aggregates formed, with a strong bulklike Ag absorption band around 310 nm. In a control experiment where MPA (40 mM) was added to a PS I suspension, no substantial influence on the absorption spectra was detected (inset of Figure 4). The decrease of the adsorption signal upon addition of the MPA indicates that indeed the spectral enhancement effect observed is due to interaction of the PS I with the metal colloids. Similar enhancement factors were observed for repeated experiments with small deviations due to varying size distributions and various aggregation stages of the nanoparticles. As the enhancement factor increased, a higher degree of similarity between the bare PS I spectra and that of the Ag colloid-PS I hybrid system was observed (Figure 5). For the Ag-PS I system, we have estimated enhancement factors of the order of ∼2-4 for the 680 nm peak at the three different samples (Figure 4a and Figure 5). In this system, the broad PS I absorption band at 440 nm overlaps with the expected Ag NP plasmon peak around 400 nm. Therefore, it is harder to separate the contributions of the PS I and Ag NPs for the 440 nm absorption band. However, variations between enhancement factors that were obtained for three different samples of the Ag-PS I hybrid systems (Figure 5) provide additional insight into the spectrally broad nature of the enhancement phenomenon. Each of these samples contained constant silver ion and reducing agent concentrations with varying PS I concentrations (variations within a factor of ∼3). These differences as well as random variations in reduction kinetics led to variations in size and aggregation distributions between the various samples. The absorption spectrum of pure PS I in solution (shown in the blue curves in Figure 5a-c) consist of a fine structure which corresponds to the chlorophyll’s electronic transitions (marked by the red arrows). The relative magnitude of these fine structure details, which originate solely from the enhanced PS I absorption, provides a way to compare the 440 nm band enhancement factor of the three samples. Thus, from the small magnitude of the fine structure details in Figure 5a, which are of the same order as the details in the pure PS I sample, it can be deduced that the strong absorption of sample I at the 440 nm band is mostly due to © 2010 American Chemical Society

FIGURE 5. Absorption spectra of unbound PS I suspension (blue curves) and Ag NP-PS I hybrid system (black curves) for three different samples (a-c). (d) The same curves as (c) but with the PS I suspension absorption of (c) multiplied by a factor of 5 in order to demonstrate the high degree of similarity between the absorption spectra of the PS I suspension and that of the hybrid system. The red arrows in the image mark the positions of the PS I fine absorption structures. PS I concentrations of the NP-PS I hybrid system and PS I suspension were 5 nM (sample I), 1.5 nM (sample II), and 2.3 nM (sample III). Note that for each sample the concentration of PS I in the hybrid system and in the PS I suspension is equal.

extinction of the Ag NPs. Note that a very small enhancement is observed in sample I also at the 680 nm band. On the other hand, the absorption of the hybrid systems in samples II and III (parts b and c of Figure 5, black curves), shows fine details at the 440 nm band enhanced by factors of roughly ∼2 and ∼4, for samples II and III, respectively. As a result, the spectrum of sample III is similar in shape to the absorption spectrum of the isolated PS I suspension 2072

DOI: 10.1021/nl100254j | Nano Lett. 2010, 10, 2069-–2074

(Figure 5d). Interestingly, the 680 nm absorption band of the three samples follows similar trends of absorption enhancement as of the 440 nm absorption band, an indication for a broad band enhancement mechanism. In the case of the Au-PS I system, ∼1.9- and ∼2.5-fold enhancements in light absorption were estimated for the 440 and 680 nm bands, respectively (Figure 4b). The enhancement factors of the two bands were estimated by dividing the absorption bands of the Au-PS I hybrid system by those of the pure PS I suspension. To evaluate the shape of the Au plasmon band, the absorption spectra of the pure PS I suspension multiplied by the enhancement factor 1.9 (the enhancement factor of the 440 nm band) was subtracted from that of the hybrid system (see Supporting Information). This procedure reveals a broad plasmon band that slightly overlaps the 680 nm band of the PS I but has no contribution around 440 nm. As can be seen in Figure 2a, this plasmon resonance comes from sub-100-nm roughly spherical Au NPs and small aggregates. Presented here is a simple model that accounts for the observed broad enhancement effect in the Au-PS I system (see Supporting Information for details). The model takes into account both on resonance and off resonance contributions. For this, we employ two simple versions shown in Figure S1 in the Supporting Information. In the first configuration, a PS I is attached to the surface of a single NP and approximated as a cylinder. The second configuration includes PS I sandwiched between two NPs. Then, using the empirical dielectric constants of metals, we simulate the local field enhancement factors (P(λ)) within the PS I molecule (Figure 1 and Supporting Information). The gold plasmon-dielectric enhancement,P(λ), is maximum and minimum at the positions shown in Figure 2Sb (Supporting Information). The calculated field enhancement of the PS I-Au NP complex at the PS I absorption wavelengths is found in the range of 1.06 < P (680 nm) < 5.24 and 1.04 < P (440 nm) < 3.54. The enhancement ranges represent the distance distribution of the chlorophyll chromophores within the PS I from the metal NP surfaces. In the NP-PS I-NP model, the numbers are overall larger: 1.62 < P (680 nm) < 7.87 and 1.26 < P (440 nm) < 4.46. Then, we calculate an enhancement factor averaged over the volume of the PS, P¯, and obtain: P¯ (λ ) 680 nm) ∼ 1.55 and 2.8 for PS I-NP and NP-PS I-NP, respectively. For the other important wavelength, P¯ (λ ) 440 nm) ∼ 1.33 and 1.9 in the two models. This averaging takes into account the strongly nonlinear nature of the distance dependence of the enhancement (almost ∼1/R6). Namely, enhancement of chlorophylls located closer to the metal surface is significantly larger. Importantly, these simple simulations show that the offplasmon enhancement can be substantial. Indeed, this is what we have observed in our experiments (Figure 4) for the Au complex, the experimental factors ∼1.9 (440 nm) and ∼2.5 (680 nm) are in good agreement with the calculated enhancement factors P ∼ 1.9 and 2.8 for the two bands © 2010 American Chemical Society

in the second (NP-PS I-NP) model. Since the Au-PS I complexes form aggregates, the electromagnetic field is strongly enhanced in the nanoparticle-nanoparticle junctions and our NP-PS I-NP model describes this situation. Indeed, we see that this model gives a better correlation between theory and experiment. From our simple theory applied to the Au-PS I complexes, we can draw the following important qualitative conclusions: (1) There is a substantial effect of off-resonance enhancement due to the high dielectric constant of the metal also at the wavelengths far from the plasmon resonance, (2) the NP-PS I-NP complex exhibits larger amplification factors than the NP-PS I which points that the off-plasmon enhancement in our experiments probably originates from “hot-spot” interparticle junctions, and (3) single Au NPs of that size cannot provide substantial enhancement for the PS I which are large proteins, of the same size scale as the Au NPs. The Ag-PS I system is more complex and cannot be described by the simple models with monodisperse spherical NPs, due to its large size and its complex and diverse shapes. Note that the TEM images revealed both small and very large Ag NPs (Figure 2b,c). In this ensemble, a significant higher enhancement is expected for larger Ag NPs than in the smaller ones.15 We conclude that the enhancement in this system must originate from the larger porous particles with rough surfaces, since the calculations indicate that the small Ag NPs cannot provide the observed enhancements (see Supporting Information). These particles should exhibit higher enhancement factors due to their large size. Moreover, for this system an additional enhancement factor is expected due to the confinement of the PS I proteins inside the nanocavities of the Ag aggregates. This should result in a larger enhancement in the Ag system compared to the Au system. One should note that the overall enhancement factor is an average of all Ag-PS I hybrid particles sizes and shapes. Therefore, the relatively modest enhancement factors observed for the Ag-PS I are due to the averaging of the large NPs strong enhancements with the low enhancements of PS I attached to small (∼10 nm) Ag NPs. To further evaluate the absorption enhancement effect of the metal particles, CD measurements were performed. In contrast to absorption spectroscopy, where the metal particles significantly contribute to the spectra, in CD spectroscopy the signal is expected to originate only from the chiral PS I contributions. CD spectra of the bare PS I, Ag, and Au-PS I hybrid samples are presented in Figure 6. It can be clearly seen that the coupling of the PS I to Ag and Au NPs caused a strong increase (a factor of ∼9) in the CD signal. The observation of a CD enhancement is particularly useful in the present work. Deducing quantitative information about the enhancement from the absorption spectra is difficult due to the overlap of plasmon and chlorophyll absorption bands, in particular in the case of metal nanoparticle aggregation which broadens the NP absorption spectrum. In the case of the CD spectra, since the multiple (96) chlorophyll chro2073

DOI: 10.1021/nl100254j | Nano Lett. 2010, 10, 2069-–2074

Interaction (S.R.) and by the National Science Foundation project number: CBET-0933415 (S.G). We gratefully thank Professor Chanoch Carmeli and Dr. Michael Gozin for their contribution to this project. Supporting Information Available. A detailed model which includes modeling of the enhancement factor, numerical results for PS I-Au and -Ag hybrid systems and energy transfer rate, and estimation of the metal plasmon band shape. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES AND NOTES (1) (2)

FIGURE 6. CD spectra of the Ag- and Au-PS I hybrids. CD of the unbound PSI (black curve) is enhanced by a factor ∼9 when the PSI binds to Ag (blue curve) and Au (red curve) colloids. The PS I concentration in all samples (NP-PS I hybrid systems and PS I suspension) was 2.3 nM.

(3) (4)

mophores are well protected within the PS I protein, a CD enhancement due to collective conformation changes is unlikely and therefore a pure electromagnetic enhancement is assumed, without the disturbance of the NP absorption. Still, the apparent discrepancy between the enhancement factors estimated for the absorption (∼2-4) and the one observed for the CD (∼9) is intriguing. Possible factors which may resolve the observed discrepancy are associated with the additional magnetic component observed in nanometric metal colloids coated by organic monolayers,23 with aggregates which may be strongly chiral,24 or with the effects coming from chiral dissipative surface currents induced by molecular dipoles in the metal NPs.25 These factors may contribute to the enhanced CD spectra of the metal-PS I. In summary, it was shown that metal NPs act as broad band optical antennas to enhance the absorption of light over the entire absorption band of the PS I rather than at the specific plasmon resonance wavelength corresponding to spherical metal NPs. An off resonance enhancement mechanism further contributes to the broad band enhancement. We assign the off resonance enhancement effect observed in the Au NPs hybrid system to the high dielectric constant of the metal also at wavelengths far from the plasmon resonance and to the presence of NP-PS I-NP aggregates in solution, where high field hot spots in interparticle junctions were formed. The enhancement mechanisms are given further significance in the Ag NPs system due to the large size of the NP aggregates and their porous structure. Enhancement of light absorption by the photosynthetic proteins is important when considering the excellent optoelectronic properties of the PS I for device applications. The off resonance enhancement observed in this study results in an effective light absorption over the entire absorption band of the PS I which may contribute to improved performance of PS I base photovoltaic devices.

(5) (6)

(7)

(8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22)

(23) (24)

Acknowledgment. This work was supported in part by the ISF Converging Technologies program, Grant No. 1714/ 07 (G.M.), by the James Frank program of Light-Matter © 2010 American Chemical Society

(25)

2074

Brettel, K.; Leibl, W. BBA Bioenergetics 2001, 1507, 100–114. Lee, I.; Lee, J. W.; Stubna, A.; Greenbaum, E. J. Phys. Chem. B 2000, 104, 2439–2443. Carmeli, I.; Mangold, M.; Frolov, L.; Zebli, B.; Carmeli, C.; Richter, S.; Holleitner, A. W. Adv. Mater. 2007, 19, 3901–3905. Sepunaru, L.; Tsimberov, I.; Forolov, L.; Carmeli, C.; Carmeli, I.; Rosenwaks, Y. Nano Lett. 2009, 9, 2751–2755. Kaniber, S. M.; Simmel, F. C.; Holleitner, A. W.; Carmeli, I. Nanotechnology 2009, 20. Terasaki, N.; Yamamoto, N.; Tamada, K.; Hattori, M.; Hiraga, T.; Tohri, A.; Sato, I.; Iwai, M.; Iwai, M.; Taguchi, S.; Enami, I.; Inoue, Y.; Yamanoi, Y.; Yonezawa, T.; Mizuno, K.; Murata, M.; Nishihara, H.; Yoneyama, S.; Minakata, M.; Ohmori, T.; Sakai, M.; Fujii, M. BBA Bioenergetics 2007, 1767, 653–659. Das, R.; Kiley, P. J.; Segal, M.; Norville, J.; Yu, A. A.; Wang, L. Y.; Trammell, S. A.; Reddick, L. E.; Kumar, R.; Stellacci, F.; Lebedev, N.; Schnur, J.; Bruce, B. D.; Zhang, S. G.; Baldo, M. Nano Lett. 2004, 4, 1079–1083. Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Chem. Rev. 1999, 99, 2957–2975. Cotton, T. M.; Schultz, S. G.; Vanduyne, R. P. J. Am. Chem. Soc. 1980, 102, 7960–7962. Osawa, M. Near-Field Opt. Surf. Plasmon Polaritons 2001, 81, 163–187. Lakowicz, J. R. Anal. Biochem. 2005, 337, 171–194. Lieberman, I.; Shemer, G.; Fried, T.; Kosower, E. M.; Markovich, G. Angew. Chem., Int. Ed. 2008, 47, 4855–4857. Nakayama, K.; Tanabe, K.; Atwater, H. A. Appl. Phys. Lett. 2008, 93, 121904. Westphalen, M.; Kreibig, U.; Rostalski, J.; Luth, H.; Meissner, D. Sol. Energy Mater. Sol. Cells 2000, 61, 97–105. Govorov, A. O.; Carmeli, I. Nano Lett. 2007, 7, 620–625. Mackowski, S.; Wo¨rmke, S.; Maier, A.; Brotosudarmo, T.; Harutyunyan, H.; Hartschuh, A.; Govorov, A. O.; Scheer, H.; Bra¨uchle, C. Nano Lett. 2008, 8, 558–564. Berova, N.; Di Bari, L.; Pescitelli, G. Chem. Soc. Rev. 2007, 36, 914–931. Carmeli, I.; Frolov, L.; Carmeli, C.; Richter, S. J. Am. Chem. Soc. 2007, 129, 12352–12353. Gobets, B.; van Stokkum, I. H. M.; Rogner, M.; Kruip, J.; Schlodder, E.; Karapetyan, N. V.; Dekker, J. P.; van Grondelle, R. Biophys. J. 2001, 81, 407–424. Gryczynski, Z.; Lukomska, J.; Lakowicz, J. R.; Matveeva, E. G.; Gryczynski, I. Cheml. Phys. Lett. 2006, 421, 189–192. Roll, D.; Malicka, R. D.; Gryczynski, J.; Gryczynski, I.; Lakowicz, J. R. Anal. Chem. 2003, 75, 3440–3445. (a) Jiang, J.; Bosnick, K.; Maillard, M.; Brus, L. J. Phys. Chem. B 2003, 107, 9964–9972. (b) Prodan, E.; Radloff, C.; Halas, N. J.; Nordlander, P. Science 2003, 302, 419–422. (c) Romero, I.; Aizpurua, J.; Bryant, G. W.; Garcı´a De Abajo, F. J. Opt. Express 2006, 14, 9988–9999. Crespo, P.; Litran, R.; Rojas, T. C.; Multigner, M.; de la Fuente, J. M.; Sanchez-Lopez, J. C.; Garcia, M. A.; Hernando, A.; Penades, S.; Fernandez, A. Phys. Rev. Lett. 2004, 93, No. 0872041. Drachev, V. P.; Bragg, W. D.; Podolskiy, V. A.; Safonov, V. P.; Kim, W. T.; Ying, Z. C.; Armstrong, R. L.; Shalaev, V. M. J. Opt Soc. Am. B 2001, 18, 1896–1903. Govorov, A. O.; Fan, Z.; Hernandez, P.; Slocik, J. M.; Naik, R. R. Nano Lett. 2010, 10, 1374–1382.

DOI: 10.1021/nl100254j | Nano Lett. 2010, 10, 2069-–2074