Plasmonic Properties of Single Multispiked Gold Nanostars

Feb 22, 2012 - using FDTD solutions 7.0 (Lumerical Solutions, Inc.). During the ... medium in the top and side regions was taken as 1.0 (air) and that...
1 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Langmuir

Plasmonic Properties of Single Multispiked Gold Nanostars: Correlating Modeling with Experiments Lei Shao,† Andrei S. Susha,‡ Lap Shan Cheung,‡ Tapan K. Sau,§ Andrey L. Rogach,*,‡ and Jianfang Wang*,† †

Department of Physics, The Chinese University of Hong Kong, Shatin, Hong Kong SAR Department of Physics and Materials Science & Centre for Functional Photonics, City University of Hong Kong, Kowloon, Hong Kong SAR § International Institute of Information Technology-Hyderabad, Gachibowli, Hyderabad-500032, A. P., India ‡

S Supporting Information *

ABSTRACT: Gold nanostars, possessing multiple sharp spikes, have emerged as promising plasmonic particles in the field of ultrasensitive sensing. We have developed a water-based method for high-yield synthesis of size-tunable anisotropic gold nanoparticles with a varying number of spiky surface protrusions, and performed systematic experimental and theoretical analyses of the optical properties of the single gold nanostars by characterizing them simultaneously with scanning electron microscopy and dark-field scattering spectroscopy. The morphologies and corresponding scattering spectra of the individual gold nanostars have been compared with electromagnetic simulations of the plasmonic resonances utilizing the finite-difference time-domain (FDTD) method. The study provides a correlation between the experimental and calculated scattering spectra and charge distributions of the different plasmon modes in the individual gold nanostars with varying numbers and relative orientations of surface protrusions. Our results provide guidelines for choosing gold nanostars with a proper number of spikes and appropriate dimensions of the core and arms for particular plasmonic applications as well as for further developing preparation methods of multispiked metal nanoparticles.

■

INTRODUCTION There has been a substantial body of research on the preparation, characterization, and application of metal nanoparticles dating over many years.1−3 This has gained greater momentum due to the emerging field of “Plasmonics”.4−7 Synthesis of metal nanoparticles is gradually shifting from spherical or simple monolithic entities to increasingly complexshaped, anisotropic, and branched particles.3,8 Creation of different complex nanoforms allows one to generate new shapedependent properties from a given volume of a material and appears to be an effective strategy for tuning the optical properties of plasmonic metal (e.g., gold, silver) nanoparticles. Plasmonic metal nanoparticles are potential candidates for a variety of applications such as SERS substrates, sensors, biological (bioassay) labels, near-field optical microscopy sources, and so forth.9 Dispersed gold or silver nanoparticles exhibit brilliant colors, which result from intense light absorption and scattering by small particles due to the collective oscillations of the conduction electrons of the particles upon photoexcitation (localized surface plasmon resonance, LSPR).10 LSPR characteristics such as the peak position, line width, and so forth are highly sensitive to the change in the particle size, shape, and the properties of the surrounding media, in addition to the composition of the nanoparticle itself. Properties of individual particles are © 2012 American Chemical Society

modified, when a particle is present in an ensemble due to the distributions in the particle size and shape, and may also depend on the extent of the interparticle separation and the nature of interactions between particles. The internal spectral inhomogeneity caused by, for example, the nonspherical shape of a given nanoparticle is usually masked in the averaged signal in an ensemble measurement. It could however be retrieved by carrying out measurements on individual particles. Following the study of Klar et al.,11 there have been several reports of single-nanoparticle measurements in the past several years. Jin et al. and Mock et al. carried out single-particle LSPR studies on silver nanoparticles of spherical, triangular prism, and pentagonal shapes.12,13 They demonstrated that the scattering peaks of silver nanoparticles red-shift with the increase in the particle size, and blue-shift with the rounding or truncation of the corners of triangularly shaped particles. Nehl et al. reported from dark-field microscopy that individual (∼100 nm) starshaped gold nanoparticles support multiple plasmon resonances, which result in polarization-dependent multipeak scattering that is extremely sensitive to the local dielectric Special Issue: Colloidal Nanoplasmonics Received: December 6, 2011 Revised: February 3, 2012 Published: February 22, 2012 8979

dx.doi.org/10.1021/la2048097 | Langmuir 2012, 28, 8979−8984

Langmuir

Article

Figure 1. (a) Extinction spectra of an ensemble of gold nanostars in water, with the corresponding TEM image shown in the inset. (b, c) Normalized experimental scattering spectra and SEM images (insets) of two differently shaped single gold nanostars.

environment.14 Wang et al. showed that the plasmon resonances of bimetallic particles are broader than those of pure silver or gold particles and that the broadening is determined to be due to the frequency dependence of the dielectric functions of the particles.15 When a metal nanoparticle is excited at its resonance wavelength, in addition to the absorption and scattering of the incident radiation, a local electric field enhancement occurs.10 The local field on the metal surface increases to a maximum when the wavelength of the incident light coincides with the LSPR frequency of the nanoparticle. Furthermore, the maximum field enhancement factor is higher for metal nanoparticles with low symmetry.16 Therefore, the local electric field, which determines the signals of plasmon-enhanced spectroscopy, such as SERS, fluorescence, and so forth,17 also depends on the size and shape of metal nanoparticles. It is wellknown that the Raman cross sections of molecules can be enhanced by many orders of magnitude, when the molecules are adsorbed onto rough metal surfaces.18 Similarly, the fluorescence of dye molecules can be enhanced when the molecules are placed in the vicinity of metallic nanoparticles.19 These phenomena are partly attributed to the electromagnetic enhancement of the optical field at sharp edges of plasmonic metal nanoparticles.20 Thus, gold nanoparticles of complex shapes can focus light at nanometer length scales and are suitable for applications that require subwavelength resolution and/or amplification of light. Multispiked gold nanoparticles allow for the tuning of plasmon resonances through the visible to near-IR spectral regions, where tissues are relatively transparent. This tuning is especially important for biological applications.8 Mode splitting of coupled plasmons has been described as earlier as in 1981 by Kreibig et al.,4 and according to the calculations of Nordlander and co-workers, the core of star-shaped gold nanoparticles serves as a nanoscale antenna for the surface protrusions, and enhances the excitation cross section as well as the local electromagnetic fields of the tip plasmons.21 Such multispiked nanoparticles show great promise for a variety of fundamental studies and applications, and in particular for SERS.22−25 According to the calculations performed by Liz-Marzan and coworkers, SERS enhancement factors for molecules adsorbed near the tips of a gold nanostar could approach 10 orders of magnitude,26 and multispiked plasmonic nanoparticles offer enhanced local fields near the tips of the protruded spikes,27 which are relatively easily accessible to analyte molecules. We have reported that multipod, branched and multispiked

plasmonic nanoparticles that have sharp tips and edges show very high sensitivities toward local changes in the dielectric environment around the particles.28,29 Several improvements and modifications in the preparation of star-shaped gold nanoparticles have been made over recent years,22,26,30−33 allowing us to achieve better control of their LSPR properties and expand their range of potential applications. We have recently reported a water-based highyield synthesis method of size-tunable gold nanostars, delivering multispiked particles in large quantities for the purpose of various studies and applications.34 In this study, we used the advantage of this synthetic technique to produce multispiked gold nanoparticles with varying surface protrusion characteristics, which have been systematically analyzed by comparing the morphologies and scattering spectra of single gold nanostars with electromagnetic simulations of their plasmon resonances utilizing the finite-difference time-domain (FDTD) method. The results of our study complement and extend the previous work of Nordlander et al.21 and others35,36 as they provide direct correlation of experimental and calculated scattering spectra and charge distributions of different plasmon modes in individual gold nanostars. Our studies further reveal the dependence of the plasmon resonances on the number of protrusions and the relative orientations of the protrusions as well as the effects of the size and tip sharpness of the protrusions.

■

EXPERIMENTAL SECTION

Sample Preparation. Gold nanostars were prepared in a one-pot synthesis in aqueous solutions using cetyltrimethylammonium bromide as the capping agent and ascorbic acid as the reducing agent.34 Nucleation and growth of multispiked gold nanoparticles were triggered by the addition of a small quantity of AgNO3.34 The number percentage of spherical (not multispiked) nanoparticles in the resulting ensemble was reproducibly below 5%, while the amount of surface protrusions (spikes) was controlled by the reaction time.34 Characterization Methods. We characterized the individual gold nanostars with scanning electron microscopy (SEM) and dark-field scattering spectroscopy. The Au nanostars were deposited on conductive indium tin oxide substrates by the drop-casting technique. The scattering spectrum of each nanostar was correlated with its geometrical structure by utilizing a pattern-matching method.37,38 SEM images were taken on an FEI Quanta 400 FEG microscope. Scattering spectra of the individual Au nanostars were measured on a dark-field optical microscope (Olympus BX60) integrated with a quartz-tungsten-halogen lamp (100 W), a monochromator (Acton SpectraPro 2300i), and a charge-coupled device camera (Princeton 8980

dx.doi.org/10.1021/la2048097 | Langmuir 2012, 28, 8979−8984

Langmuir

Article

Instruments Pixis 512B). The camera was thermoelectrically cooled to −70 °C during the measurements. A dark-field objective (100×, numerical aperture = 0.80) was employed for both illuminating the ITO substrate with the white excitation light and collecting the scattered light. Scattering spectra from the individual nanostars were corrected by first subtracting the background spectra taken from the adjacent regions without Au nanostars and then dividing them with the calibrated response curve of the entire optical system. Electromagnetic Simulations. The electromagnetic simulations were carried out to model the plasmonic responses of the single Au nanostars, utilizing a FDTD method. The calculations were performed using FDTD solutions 7.0 (Lumerical Solutions, Inc.). During the calculations, an electromagnetic pulse in the wavelength range from 450−900 nm was launched into a box containing the target Au nanostar to simulate a propagating plane wave interacting with the nanostructure. The Au nanostar and its surrounding space were divided into 1 nm meshes. The dielectric function of gold was represented with a combination of the Drude and Lorentz model, with parameters chosen to match the experimental dielectric data as close as possible:

εDL(ω) = ε∞ −

ω D2 2

ω + i γDω

−

Figure 2. (a) Normalized experimental scattering spectrum and SEM image (inset) of a one-spiked gold nanostar. (b) Calculated scattering spectra of the one-spiked nanostar obtained from the FDTD simulations. The blue, red, and green lines represent the scattering of light under different excitation polarizations. The black line is their sum. The in-plane x (blue) and y (red) directions are along and perpendicular to the symmetry axis of the nanostar, respectively. The z (green) axis is perpendicular to the substrate. (c) Charge distributions of the different plasmon modes in the one-spiked nanostar on its central cross section.

(Δε)ΩL 2 2

(ω − ΩL 2) + i ΓLω

(1)

two spheroids are measured from the SEM image to be at 41/ 32 and 61/13 nm, respectively. The FDTD-calculated scattering spectrum of a one-spiked Au nanostar matches well with the experimentally measured one (Figure 2b). When the excitation light is polarized along the symmetry axis (x direction), two plasmon modes centered at 530 and 825 nm can be excited. When the excitation light is polarized within the substrate plane and perpendicular to the symmetry axis (y direction), only one plasmon mode centered at 529 nm can be excited. The excitation light polarized perpendicular to the substrate can excite a plasmon mode at 513 nm. The 825 nm strong scattering peak agrees perfectly with the experimentally measured 824 nm one. The ∼530 nm plasmon modes excited by the x-, y-, and z-polarized light all contribute to the weaker scattering peak in the shorterwavelength region. The deviation of the calculated peak positions from the experiment results can be ascribed to the imperfect modeling of the core with a prolate spheroid. The smoothing of the sharp corners of the nanostar core would lead to blue shifts of the plasmon modes excited by the light polarized along the y and z axes.38 The calculated charge distributions of the nanostar at the different in-plane plasmon modes are shown in Figure 2c to make a clear inspection of the plasmon resonance peaks. We also provide the electric field intensity enhancement contours in Figure S1 in the Supporting Information. According to the charge contours, the two scattering peaks at 825 and 530 nm under the excitation of the x-polarized light correspond to a low-energy dipolar and a high-energy multipolar plasmon mode of the Au nanostar, respectively. The dipolar mode is a longitudinal plasmon resonance polarized along the x axis. The 529 nm scattering peak excited by the y-polarized light, on the other hand, corresponds to a second dipolar plasmon resonance which is transversely polarized (along the y-axis). The plasmonic response of the nanostar can also be interpreted from the plasmon hybridization concept. Under the excitation of the x-polarized light, the core and the tip exhibit their plasmon resonance at 544 nm (2.278 eV) and 620 nm (2.000 eV), respectively (Supporting Information Figure S2). The core dipolar plasmon mode and the tip dipolar plasmon mode couple strongly and hybridize to form bonding

where the used parameters are the high-frequency dielectric constant ε∞ = 5.8421, the plasma frequency ωD = 13518 THz, the Drude damping constant γD = 79.513 MHz, the Lorentz weighting factor Δε = 1.7072, the Lorentz oscillator strength ΩL = 4343.7 THz, and the Lorentz line width ΓL = 403.08 THz. The refractive index of the medium in the top and side regions was taken as 1.0 (air) and that in the bottom of a Au nanostar was set at 1.9 to simulate the ITO substrate. The scattering signal from a single nanostar was collected at all angles and then integrated.

■

RESULTS AND DISCUSSION Figure 1 shows the difference in the optical properties of an ensemble of the gold nanostars and of single multispiked Au nanoparticles of varying morphology. The aqueous colloidal suspension of the Au nanostars (Figure 1a) shows an inhomogeneously broadened extinction spectrum extending from the visible to NIR, with the relative intensities of the peaks depending on the sample preparation conditions.34 At the same time, the individual nanostars (Figure 1b,c) show well-defined maxima in their scattering spectra, with the number and spectral positions of peaks strongly depending on the number of surface protrusions and their orientation. In the following, we present the results of the correlation analysis of the experimental and calculated scattering spectra of gold nanostars with an increasing number of surface protrusions. Figure 2a shows the experimentally measured scattering spectrum and corresponding SEM image of a representative one-spiked nanostar. Both a weak scattering peak in the shorter-wavelength region and a strong scattering peak in the longer-wavelength region can be clearly observed. The two scattering peaks are positioned at 573 and 824 nm, respectively. According to the plasmon hybridization model,21 the two observed plasmon modes of the one-spiked nanostar can be understood by considering interactions between the plasmons associated with the core and the arm. To elucidate how the two scattering peaks are formed, we performed FDTD calculations to pick up the different plasmon resonance modes. The one-spiked nanostar is modeled by a prolate spheroid core with a protruding arm of a more slender prolate spheroid shape. The semimajor axes of the two spheroids lie on the same line parallel to the substrate. The semimajor/semiminor axes of the 8981

dx.doi.org/10.1021/la2048097 | Langmuir 2012, 28, 8979−8984

Langmuir

Article

two strong plasmon resonances centered at 697 and 734 nm and a weak plasmon resonance around 515 nm. We obtained the electric field intensity enhancement contours at different resonance peaks (Figure S1, Supporting Information). The charge contours of each plasmon modes were calculated accordingly (Figure 3c). At the two strong, longer-wavelength plasmon resonances excited by the x-polarized light, the entire nanostar displays a dipolar pattern. At the 697 nm plasmon resonance wavelength, the charges in the two arms are attractive to those in the core but repulsive with respect to the two arms. At the 734 nm plasmon resonance wavelength, the charge in the smaller arm has the same sign as that in the core but is opposite to that in the larger arm. The plasmon resonance is more localized in the larger arm. The x-polarized light can also excite a weak high-order plasmon mode at around 515 nm. Under the excitation of the y-polarized light, two plasmon modes centered at 538 and 728 nm can be clearly observed. At the 728 nm mode, the entire nanostar exhibits a dipolar pattern. While at 538 nm, the three components of the nanostar (the core and the two arms) all exhibit dipolar patterns. The three dipoles point to the same direction. The dipoles in the two arms are attractive to each other and repulsive to that on the core. The calculated scattering spectrum of the double-spiked Au nanostar agrees qualitatively with the experimental one except that the experimentally observed 672 nm peak cannot find its counterpart in the FDTD-calculated result. The disappearance of the scattering peak in the calculated result could originate from the use of spheroids to model the nanostar arms. Some high-order plasmon modes that are excitable for sharp tips cannot be observed for smoother tips.39 Our simulation results for Au nanostars with sharp tips show that the sharp tip enables the excitation of additional high-order plasmon modes centered at around 650 nm, which will be presented in the following. The orientations of the arms can also affect the plasmonic response of a single Au nanostar. We therefore investigated by numerical simulation how the angle formed by the two arms of the double-spiked nanostar will affect its scattering properties. The nanostar was modeled by a spherical core with two identical protruding conical tips, as schematically shown in the inset in Figure 4a. The radius of the spherical core is 37 nm. The conical tips are right circular cones with their height set at 65 nm and aperture at 23°. The calculated scattering spectra are slightly different from those in Figure 3. The two splitting strong peaks excited by the x-polarized light shown in Figure 3b coalesce into one for the nanostar with two identical arms. The coalescence is believed to result from the absence of the symmetry breaking between the two tips. Additionally, a new scattering peak around 650 nm can be excited by the xpolarized light. The peak around 515 nm is maintained. The wavelengths of all the three peaks are seen to fluctuate slightly as the angle between the two conical tips is increased (Figure 4a and b). Under the excitation of the y-polarized light, in addition to the two peaks shown in Figure 3b, a new scattering peak around 660 nm can be observed. Interestingly, the peak wavelengths change only slightly with the tip angle (Figure 4d and e). The three plasmon modes are slightly red-shifted in comparison to their counterparts excited by the x-polarized light. We attribute the appearance of the new plasmon peaks at 650−660 nm to the excitation of some high-order plasmon modes under the x- and y-polarized light due to the sharp arms. Figure S3 in the Supporting Information shows the calculated charge distributions of the Au nanostar with the intertip angle

and antibonding nanostar plasmon modes at 825 nm (1.503 eV) and 530 nm (2.340 eV), respectively. When the excitation light is y-polarized, the plasmon resonances of the core and the tip are centered at 528 nm (2.348 eV) and 530 nm (2.340 eV). The coupling between them is quite weak and the entire nanostar exhibits its plasmon resonance at 529 nm (2.344 eV). The weak plasmon coupling results from the small transverse plasmonic dipole moment of the tip. For the multispiked Au nanostars, the plasmon modes of the stars are more complicated. Figure 3 shows the plasmonic

Figure 3. (a) Normalized experimental scattering spectrum and SEM image (inset) of a double-spiked gold nanostar. (b) Scattering spectra of the double-spiked nanostar obtained from the FDTD simulations. The blue, red, and green lines represent the scattering of light under different excitation polarizations. The black line is their sum. The inplane x (blue) and y (red) directions are along and perpendicular to the bisector of the angle formed by the two arms, respectively. The z (green) axis is perpendicular to the substrate. (c) Charge distributions of the different plasmon modes in the nanostar on its central cross section.

response of a representative double-spiked nanostar. Four different scattering peaks can be unambiguously observed from the measured scattering spectrum. The experimentally obtained peak wavelengths are 567, 672, 744, and 776 nm, respectively. The scattering of the double-spiked Au nanostar was simulated using the FDTD method as well. The double-spiked nanostar was modeled by a spherical core with two protruding arms of a slender prolate spheroid shape. For simplicity, the centers of the core and the arms are placed in the plane parallel to the substrate and the two arms are aligned parallel to the substrate. The positions of the two arms and the angle between them were adjusted slightly to make the calculated spectral shape match the experimentally observed one. The radius of the core was set at 49 nm and the semimajor/semiminor axes of the two arms were 68/16 and 62/18 nm, respectively. The calculated scattering spectrum of the double-spiked nanostar exhibit peaks at 542, 700, and 731 nm. The two strong peaks in the longerwavelength range are contributed by the plasmon modes excited by the x- and y-polarized light (the in-plane x and y direction are along and perpendicular to the bisector of the angle formed by the two arms). The 542 nm scattering peak consists of the contributions from both the in-plane plasmon modes and the plasmon modes excited by light perpendicular to the substrate (z direction). The x-polarized light can excite 8982

dx.doi.org/10.1021/la2048097 | Langmuir 2012, 28, 8979−8984

Langmuir

Article

Figure 5. (a) Normalized experimental scattering spectrum and SEM image (inset) of a three-spiked gold nanostar. (b) Scattering spectra of the three-spiked nanostar obtained from the FDTD calculations. The blue, red, and green lines represent the scattering of light under different excitation polarizations. The black line is their sum. The x (blue) and y (red) directions are in-plane parallel and perpendicular to the line connecting the centers of the core and the central arm, respectively. The z (green) axis is perpendicular to the substrate. (c) Charge distributions of the different plasmon modes in the nanostar on the central cross sections of the core spheroid.

Figure 4. (a, d) Calculated scattering spectra of double-spiked nanostars with their two identical tips forming different angles. The light for excitation is polarized along the symmetry axis of the nanostar (x axis) in (a) and perpendicular to it but parallel to the center cross section (y axis) in (d). The inset in (a) shows the schematic of a double-spiked nanostar with tunable intertip angle θ. (b, e) Plots of the peak wavelengths of the three modes versus the tip angle. (c, f) Peak intensities of the different modes under the excitation of x- and ypolarized light versus the tip angle, respectively.

oriented parallel to the substrate during the simulations for simplicity. The four spheroids have their semimajor/semiminor axes at 42/38 (core), 38/15 (upper arm), 30/10 (central arm), and 65/18 (lower arm) nm, respectively. The semimajor axis of the core is along the x axis. The three arms form angles of 56 (upper arm), 0 (central arm), and −45° (lower arm) relative to the x axis, respectively. The calculations indicate that the xpolarized light can excite two plasmon modes at 620 and 734 nm. The nanostar displays a multipolar pattern and a dipolar pattern pointing to the direction of the largest arm, respectively (Figure 5c). The y-polarized light can excite a strong plasmon mode centered at 734 nm and a much weaker peak around 630 nm. The plasmon resonance on the nanostar at the 734 nm wavelength also exhibits a dipolar pattern toward the direction of the largest arm. The charge distributions of the nanostar under the excitation of x- and y-polarized light are similar to each other at both 620 and 734 nm (Figure 5c). We therefore consider that the plasmon modes excited by differently polarized light at these wavelengths have the same nature. Both the x- and y-polarized light can also excite a very weak bump around 510 nm. They and the strong plasmon mode centered at 513 nm excited by the z-polarized light contribute to the experimentally observed scattering peak at 548 nm. The electric field intensity enhancement contours at different resonance peaks are presented in Figure S1 of the Supporting Information. We also provide the scattering spectra of a fivespiked Au nanostar in the Supporting Information (Figure S4) as yet another example of a complex plasmonic response of a multispiked gold nanoparticle.

of 30° at the wavelengths of 509, 649, and 769 nm. At 649 nm, the charges are concentrated in the tip area, leading to the formation of high-order plasmon modes. The charge distributions clearly indicate that the sharp tip enables the excitation of additional high-order plasmon modes. The scattering intensities of the different plasmon modes vary with the tip angle, as revealed in Figure 4c and f. The most apparent feature is that the intensity of the lowest-energy plasmon resonance peak decreases/increases sharply with the increase of the tip angle when the nanostar is excited by the x/ y-polarized light, respectively. When the nanostar is excited by the x-polarized light, the plasmon resonances in the tips are more efficiently excited at small tip angles. The dipole moment of the entire star along the light polarization direction is larger at smaller tip angles, leading to the higher intensity of the corresponding scattering peak. On the other hand, when the nanostar is excited by the y-polarized light, a larger tip angle will lead to a more efficient excitation of the tip plasmons and therefore a larger dipole moment of the entire nanostar. As a result, the scattering intensity of the lowest energy peak increases notably with the tip angle. When the arm number is larger than two, the multispiked nanostars exhibit much richer plasmonic properties. Figure 5 presents the plasmonic response of a three-spiked Au nanostar. Two distinct peaks at 670 and 798 nm and a weak bump at 548 nm are observed from the measured scattering spectrum (Figure 5a). FDTD calculations reveal that the two strong lowenergy scattering peak are excited by both the x- and ypolarized light and the high-energy peak is mostly excited by the z-polarized light (x and y directions are in-plane parallel and perpendicular to the line connecting the centers of the core and the central arm; z direction is perpendicular to the substrate). In the simulations, the core and the three arms of the nanostar were modeled by four prolate spheroids and all their centers were in a plane parallel to the substrate. The three arms are

■

CONCLUSIONS We have performed a combined and comparative experimental and theoretical study of how the optical properties of spiked (surface protruded) gold nanoparticles vary with the spike (surface protrusion) number, size, and orientation. We have systematically investigated the optical properties of single8983

dx.doi.org/10.1021/la2048097 | Langmuir 2012, 28, 8979−8984

Langmuir

Article

(6) Burda, C.; Chen, X.; Narayanan, R.; El-Sayed, M. A. Chem. Rev. 2005, 105, 1025. (7) Halas, N. J. Nano Lett. 2010, 10, 3816. (8) Guerrero-Martinez, A.; Barbosa, S.; Pastoriza-Santos, I.; LizMarzan, L. Curr. Opin. Colloid Interface Sci. 2011, 16, 118. (9) Sau, T. K.; Rogach, A. L.; Jäckel, F.; Klar, T. A.; Feldmann, J. Adv. Mater. 2010, 22, 1805. (10) Kreibig, U.; Vollmer, M. Optical Properties of Metal Clusters; Springer: Berlin, 1995. (11) Klar, T.; Perner, M.; Grosse, S.; von Plessen, G.; Spirkl, W.; Feldmann, J. Phys. Rev. Lett. 1998, 80, 4249. (12) Jin, R.; Cao, Y. W.; Mirkin, C. A.; Kelly, K. L.; Schatz, G. C.; Zheng, J. G. Science 2001, 294, 1901. (13) Mock, J. J.; Barbic, M.; Smith, D. R.; Schultz, D. A.; Schultz, S. J. Chem. Phys. 2002, 116, 6755. (14) Nehl, C. L.; Liao, H.; Hafner, J. H. Nano Lett. 2006, 6, 683. (15) Wang, X.; Zhang, Z.; Hartland, G. V. J. Phys. Chem. B 2005, 109, 20324. (16) Vo-Dinh, T. Trends Anal. Chem. 1998, 17, 557. (17) Nie, S.; Emory, S. R. Science 1997, 275, 1102. (18) Lee, P. C.; Meisel, D. J. Phys. Chem. 1982, 86, 3391. (19) Liu, Y.; Blair, S. Opt. Lett. 2003, 28, 507. (20) Barnes, W. L. J. Mod. Opt. 1998, 45, 661. (21) Hao, F.; Nehl, C. L.; Hafner, J. H.; Nordlander, P. Nano Lett. 2007, 7, 729−732. (22) Khoury, C. G.; Vo-Dinh, T. J. Phys. Chem. C 2008, 112, 18849. (23) Hrelescu, C.; Sau, T. K.; Rogach, A. L.; Jäckel, F.; Feldmann, J. Appl. Phys. Lett. 2009, 94, 153113. (24) Rodrıguez-Lorenzo, L.; Alvarez-Puebla, R. A.; Pastoriza-Santos, I.; Mazzucco, S.; Stephan, O.; Kociak, M.; Liz-Marzan, L. M.; Abajo, F. J. G. D. J. Am. Chem. Soc. 2009, 131, 4616. (25) Pazos-Perez, N.; Barbosa, S.; Rodriguez-Lorenzo, L.; Aldeanueva-Potel, P.; Perez-Juste, J.; Pastoriza-Santos, I.; AlvarezPuebla, R. A.; Liz-Marzan, L. M. J. Phys. Chem. Lett. 2010, 1, 24. (26) Kumar, P. S.; Pastoriza-Santos, I.; Rodriguez-Gonzalez, B.; Garcia de Abajo, F. J.; Liz-Marzan, L. M. Nanotechnology 2008, 19, 015606. (27) Hrelescu, C.; Sau., T. K.; Rogach, A. L.; Jaekel, F.; Laurent, G.; Douillard, L.; Charra, F. Nano Lett. 2011, 11, 402. (28) Chen, H.; Kou, X.; Yang, Z.; Ni, W.; Wang, J. Langmuir 2008, 24, 5233. (29) Dondapati, S. K.; Sau, T. K.; Hrelescu, C.; Klar, T. A.; Stefani, F. D.; Feldmann, J. ACS Nano 2010, 4, 6318. (30) Liao, H.-G.; Jiang, Y.-X.; Zhou, Z.-Y.; Chen, S.-P.; Sun, S.-G. Angew. Chem. Int. Ed. 2008, 47, 9100. (31) Wu, H.-L.; Chen, C.-H.; Huang, M. H. Chem. Mater. 2009, 21, 110. (32) Barbosa, S.; Agrawal, A.; Rodriguez-Lorenzo, L.; Alvarez-Puebla, R. A.; Kornowski, A.; Weller, H.; Liz-Marzan, L. M. Langmuir 2010, 26, 14943. (33) Trigari, S.; Rindi, A.; Margheri, G.; Sottini, S.; Dellepiane, G.; Giorgetti, E. J. Mater. Chem. 2011, 21, 6531. (34) Sau, T. K.; Rogach, A. L.; Döblinger, M.; Feldmann, J. Small 2011, 7, 2188. (35) Ma, W. Y.; Yang, H.; Hilton, J. P.; Lin, Q.; Liu, J. Y.; Huang, L. X.; Yao, J. Opt. Express 2010, 18, 843. (36) Mazzucco, S.; Stephan, O’; Colliex, C.; Pastoriza-Santos, I.; LizMarzan, L. M.; Garcia de Abajo, J.; Kociak, M. Eur. Phys. J.: Appl. Phys. 2011, 54, 33512. (37) Chen, H. J.; Sun, Z. H.; Ni, W. H.; Woo, K. C.; Lin, H.-Q.; Sun, L. D.; Yan, C. H.; Wang, J. F. Small 2009, 5, 2111. (38) Shao, L.; Woo, K. C.; Chen, H. J.; Jin, Z.; Wang, J. F.; Lin, H.-Q. ACS Nano 2010, 4, 3053. (39) Grillet, N.; Manchon, D.; Bertorelle, F.; Bonnet, C.; Broyer, M.; Cottancin, E.; Lermé, J.; Hillenkamp, H.; Pellarin, M. ACS Nano 2011, 5, 9450.

spiked to multispiked nanostar particles. Dark-field scattering spectroscopy has been used to collect the scattering spectrum of single nanoparticles, which has been correlated with its geometrical structure by employing a pattern-matching method using SEM and dark-field microscopy. For theoretical simulation, we have used the FDTD method along with the plasmon hybridization model. The FDTD simulations reveal that the plasmonic properties of the multispiked gold nanostars are highly sensitive to excitation polarization. The observed plasmon peaks are a combination of the various plasmon modes excited by unpolarized white light. In general, the number of the plasmon modes increases with the number of protrusions. As the number of protrusions is increased, higher-order plasmon modes can generally be observed. The higher-order plasmon modes are also affected by the symmetry of the entire nanostar and relative sizes of the protrusions. The plasmonic properties of gold nanostars are also found to be mainly determined by the core and the largest arm. Moreover, for twospiked gold nanostars, the resonance strengths of the plasmon modes are strongly dependent on the angle formed by the two arms, while the peak wavelengths are relatively insensitive to the change of the arm angle. Such dependences on the arm angle are expected to be maintained for gold nanostars with more than two arms, although the plasmon hybridization and geometry are much more complicated. The FDTD method along with the plasmon hybridization model satisfactorily correlates the observed plasmonic properties of single to multispiked individual gold nanoparticles. Such studies are useful in predicting the appropriate structure of gold nanostars with suitable plasmon resonance properties for their desired applications.

■

ASSOCIATED CONTENT

* Supporting Information S

Charge distribution of different plasmon modes in a doublespiked Au nanostar; the plasmon hybridization diagram for the one-spiked gold nanostar; the scattering spectrum and the SEM image of a five-spiked nanostar. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (A.L.R); jfwang@phy. cuhk.edu.hk (J.F.W.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by NSFC/RGC Joint Scheme (Reference No.: N_CUHK465/09, Project Code: 2900339), and by City University of Hong Kong. The FDTD simulations in this work were conducted in the High Performance Cluster Computing Centre, Hong Kong Baptist University, which is supported by Hong Kong RGC and Hong Kong Baptist University.

■

REFERENCES

(1) Skrabalak, S. E.; Xia, Y. ACS Nano 2009, 3, 10. (2) Murphy, C. J. ACS Nano 2009, 3, 770. (3) Sau, T. K.; Rogach, A. L. Adv. Mater. 2010, 22, 1781. (4) Kreibig, U.; Althoff, U.; Pressmann, H. Surf. Sci. 1981, 106, 308. (5) Kelly, K. L.; Coronado, E.; Zhao, L. L.; Schatz, G. C. J. Phys. Chem. B 2003, 107, 668. 8984

dx.doi.org/10.1021/la2048097 | Langmuir 2012, 28, 8979−8984