pubs.acs.org/NanoLett
Plasmonic Nanoshell Synthesis in Microfluidic Composite Foams Suhanya Duraiswamy,† and Saif A. Khan*,†,‡ † ‡
Department of Chemical and Biomolecular Engineering, 4 Engineering Drive 4, E5-02-28, Singapore 117576 and Singapore-MIT Alliance, 4 Engineering Drive 3, E4-04-10, Singapore 117576 ABSTRACT The availability of robust, scalable, and automated nanoparticle manufacturing processes is crucial for the viability of emerging nanotechnologies. Metallic nanoparticles of diverse shape and composition are commonly manufactured by solution-phase colloidal chemistry methods, where rapid reaction kinetics and physical processes such as mixing are inextricably coupled, and scaleup often poses insurmountable problems. Here we present the first continuous flow process to synthesize thin gold “nanoshells” and “nanoislands” on colloidal silica surfaces, which are nanoparticle motifs of considerable interest in plasmonics-based applications. We assemble an ordered, flowing composite foam lattice in a simple microfluidic device, where the lattice cells are alternately aqueous drops containing reagents for nanoparticle synthesis or gas bubbles. Microfluidic foam generation enables precisely controlled reagent dispensing and mixing, and the ordered foam structure facilitates compartmentalized nanoparticle growth. This is a general method for aqueous colloidal synthesis, enabling continuous, inherently digital, scalable, and automated production processes for plasmonic nanomaterials. KEYWORDS Foams, emulsions, microfluidics, nanoshells, gold, nanoparticles, plasmonics
C
olloidal synthesis methods for nanometer-scale metallic particles have fueled spectacular advances in diverse areas of application including plasmonics, chemical catalysis, optoelectronics, and biological sensing.1-6 The solution-phase chemical reduction of aqueous ionic precursors in the presence of other molecular additives acting as stabilizing or capping agents is a central idea in colloidal synthesis.7 The associated ideas of nucleation and growth have, in recent decades, led to the discovery of new methods for the preparation of a large variety of “designer” nanoparticles of different shapes, sizes, and composition.8 The viability and eventual success of many emerging nanotechnologies now hinges on the availability of robust, reproducible, and scalable colloidal synthesis methods that yield larger quantities of material.9 However, innovations in scalable processing have thus far lagged behind rapid advances in proof-of-concept applications. One of the key reasons for this impasse is the fact that colloidal chemistry often involves rapid, autocatalytic, diffusion-controlled processes; reagent addition and mixing, which are method and operator dependent, are inextricably coupled to the chemistry.10-12 Reagent homogenization and chemical reaction proceed simultaneously in the reaction vessel, thus making it very difficult to tightly control the final outcome. The synthesis of colloidal nanomaterials via bulk solution-phase methods remains an art even after the discovery of suitable chemical ingredients, and all else remaining the same, there is much that is dependent on the “synthetic skills” of the chemist.
Confinement of reagents in dispersed water drops of a water-in-oil microemulsion has been extensively studied as an attractive alternative to bulk solution-phase synthesis.13 Such systems, where self-assembled surfactant micelles function as individual batch reactors, have been used for the synthesis of a broad range of colloidal nanomaterials.14 However, the scalability of such methods is limited by the emulsification procedure employed; it is commonly not possible to control reagent dispensing at the scale of a single micelle. Continuous flow-based microfluidic synthesis methods provide another alternative, promising superior control over reaction conditions while operating at steady-state, thus simultaneously ensuring reproducibility.15 Scaling to higher production rates is straightforward in principle and involves parallelization of multiple reactor units. Several microfluidic methodsforthesynthesisofsemiconducting,16,17 metallic,18,19 dielectric,20,21 magnetic,22 and core-shell23,24 nanoparticles have been demonstrated recently.25 A common drawback in such methods is the dominant presence of microchannel walls, which often leads to rapid, irreversible particle deposition and thus degradation of product quality with time.26,27 In this paper, we demonstrate that self-assembly of ordered microscale foams is a simple and versatile solution for robust and reproducible continuous colloidal syntheses of plasmonic nanostructures. Figure 1a is a two-dimensional schematic illustrating the basic features of the foams used in this work. The foam is an ordered alternation of gas and liquid cells (hence the term “composite”28) dispersed in an immiscible oil phase flowing within a microchannel. The overall foam structure resembles the compact “bamboo” structures observed in confined cylindrical foams.29 Operation between the dry and wet foam limits leads to charac-
* To whom correspondence should be addressed. E-mail:
[email protected]. Fax: (65) 6779 1936. Tel.: (65) 6516 5133. Received for review: 07/15/2010 Published on Web: 08/23/2010 © 2010 American Chemical Society
3757
DOI: 10.1021/nl102478q | Nano Lett. 2010, 10, 3757–3763
FIGURE 1. (a) Two dimensional schematic of flowing composite foams showing an alternating train of gas and liquid foam cells flowing through a continuous oil stream, where the gas-aqueous interfaces are flattened with sharp Plateau border curvatures. (b) Stereomicroscopic images of the composite foam with an inset showing a magnified view of the gas-aqueous interface; the sharp plateau border curvature is clearly visible. (c-f) Series of stereomicroscope images showing the breakup and formation of composite foams at the T-junction. Nitrogen gas, water, and aqueous solutions of black dye are introduced into one arm of the microfluidic T-junction, and silicone oil is delivered into the other arm. At low volumetric oil flow rates relative to the aqueous streams, alternate gas and liquid cells are pinched off at the T-junction and flow through a thin oil film, thus forming the ordered foam lattice shown in (b). (g) Mixing within a foam cell, quantified by the normalized standard deviation σ*[I(y)] of dye intensity I(y); σ* ) σ[I(y)]/〈I(y)〉. Mixing is completed 10 mm downstream of the dispensing junction, corresponding to a mixing time of ∼0.6 s. (h) Cross-sectional intensity profiles (along the dotted lines of the inset stereomicroscope images) at two different axial positions (z ) 1 and 4 mm) along the microchannel. Nearly overlapping profiles for consecutive aqueous cells (plotted in different colors) demonstrate that mixing patterns display very little cell-to-cell variation, a crucial ingredient for reproducible synthesis processes. (i,j) Splitting of foam cells into equal-sized daughter cells at a bifurcation. All scale bars correspond to 300 µm.
ago33 and have since then been objects of intense fundamental and applied research.5,34 The interest in this class of particles derives from the fact that they have broadly tunable plasmon-derived optical resonances,35 making them promising candidates for biomedical imaging, biosensing, in vivo therapeutic applications, and more recently as building blocks of plasmonic metamaterials.36,37 This optical tunability is accomplished by varying core sizes and shell thicknesses and is well understood in terms of classical electromagnetic theory38 as also by recent theoretical approaches that describe optical behavior in terms of plasmon hybridization.39 The sensitive size-dependence of optical properties
teristic flattening of adjacent fluid interfaces and sharp Plateau border curvatures (Figure 1b). We have previously demonstrated the generation of such foams as a subset of a broader class of microfluidics-enabled multiphasic emulsion structures.30 Such three-phase microfluidic motifs were originally pioneered by Ismagilov and co-workers in the context of protein crystallization studies.31,32 We highlight the salient features of these foams by demonstrating the first continuous process for the manufacture of metallodielectric “nanoshells”, each comprising a silica nanoparticle core encased within a thin gold shell. These particles were first introduced by Halas and co-workers more than a decade © 2010 American Chemical Society
3758
DOI: 10.1021/nl102478q | Nano Lett. 2010, 10, 3757-–3763
aqueous fluid samples containing nanoparticles were analyzed with a UV-vis spectrometer (Shimadzu UV-2450). The samples were further purified by repeated centrifugation and redispersion (in DI water) steps. A drop of this sample was placed onto a 200 mesh formwar protected copper grid and allowed to dry overnight. The copper grid was then analyzed using either TEM (JEOL 2010, accelerating voltage 200 kV) or FESEM (JEOL JSM-6700f, accelerating voltage 4-25 kV). Several images were taken at different locations on the grid and the particle size was found by manually measuring the diameter of at least 250 particles from several electron microscopy images. Results and Discussion. Microfluidic Composite Foams: Salient Features. Flowing microscale composite foams possess a unique set of structural and functional features that make them attractive for nanoparticle processing. Aqueous reagents for colloidal synthesis can be controllably dispensed as liquid cells of identical size that serve as individual reaction “flasks” and are effectively isolated from other reagent-filled cells and the microchannel walls during their transit through the microchannel. Reagents are dispensed during the process of foam generation, as shown in Figure 1c-f (see also Movie M1, Supporting Information). Gas bubbles and reagent-filled aqueous drops are alternately formed at a microfluidic T-junction and subsequently assemble downstream into an ordered foam lattice (Movie M2 in Supporting Information).41,42 The injected gas periodically clears reagents from the T-junction, thereby preventing reagent buildup and nanoparticle deposition that occurs when laminar fluid streams are left undisturbed.26,27 These features are a significant advance over current droplet-based microfluidic synthesis methods, where droplet formation is sensitive to fluid properties such as interfacial tension, and downstream droplet coalescence is a frequent occurrence.18
necessitates the use of particle populations with tightly controlled size distributions in practical applications. Materials and Methods. Materials. Tetraethyl orthosilicate (TEOS, 99.999%), ammonium hydroxide (28% NH3 in water, 99.99+%), and hydroxylamine hydrochloride (trace metal basis, 99.99%) from Aldrich Chemical Co., Singapore, 3-aminopropyl tris (trimethylsiloxy) silane (99%), hydrogentetrachloroaurate(III) trihydrate (99.99%), tetrakis(hydroxymethyl) phosphonium chloride (THPC, 80% in water), sodium hydroxide (reagent grade, 97%), potassium carbonate (99.99%) from Sigma-Aldrich Co. Ltd., Singapore and ethanol (analytical reagent grade, absolute) from Fisher scientific Co., Singapore were all used as obtained without any further purification. Ultrapure water (18 MΩ-cm, ELGA, Singapore) and glassware washed in aqua regia and rinsed thoroughly in water were used for experiments. Microfabrication. Microfluidic device patterns were fabricated onto silicon wafers by standard photolithography using negative photoresist SU-8 2050. Devices were subsequently molded in poly(dimethyl siloxane) (PDMS) using the soft lithography technique.40 Briefly, PDMS was molded onto the SU-8 masters at 70 °C for 4 h, peeled, cut, and cleaned. Inlet and outlet holes (1/16-in. o.d.) were punched into the device. The microchannels were irreversibly bonded to a glass slide precoated with a thin layer of PDMS after a brief 35 s air plasma treatment. The microchannels have rectangular cross-section and are 300 µm wide, ∼155 µm deep, and 0.45 m long. Microfluidic Device Setup and Operation. Syringe pumps (Harvard, PHD 2000) were used to deliver silicone oil (Dow Corning DC50, viscosity 10 cSt), a mixture of gold-seeded silica particles and gold-plating solution (S) and reducing agent solution (R) to the microfluidic device while nitrogen gas was delivered from a cylinder equipped with a two-stage pressure regulator through circular PEEK tubings (60 µm, 1.5 m long) leading into the on-chip gas-inlet. The plating solution was aged (>24 h) gold hydroxide formed by the addition of 249 mg (1.8 mM) of potassium carbonate in 1 L (0.435 mM) of hydrogen tetrachloroaurate (III) trihydrate solution. The ionic gold concentration ([Au3+]) in the plating solution was varied from 0.43 to 0.29 mM, and volume fraction of gold-seeded silica particles (“fS”) in S was varied from ∼0.003 to 0.01%. Freshly prepared hydroxyl amine hydrochloride (4 mM) was used as the reducing agent in all experiments. Flow rates of the individual streams were S, 8 µL/min; R, 2 µL/min each and the volumetric flow ratio of aqueous reagents (S + R) to oil was maintained at 2. The nitrogen gas pressure was maintained at 17 psig. Sample Collection and Analysis. The outlet from the reactor was connected via fluoropolymer (FEP) tubing to a 2 mL centrifuge tube. Approximately 1 mL of the aqueous sample was collected in the tube for every experimental condition. The oil formed a layer on the surface of the collected aqueous fluid while the gas simply escaped into the ambient. The oil layer was carefully decanted and © 2010 American Chemical Society
Relative motion between liquid-filled foam cells and the microchannel walls, mediated by the intervening oil, generates recirculating fluid motion within the cells. Minute differences in surface roughness between opposite microchannel walls (due to microfabrication tolerances) and small differences in curvature of the fluid interfaces break the symmetry of the velocity field within the cell and cause rapid mixing of the contents (Figure 1g).43 The mixing patterns within the aqueous foam cells are extremely robust and display little cell-to-cell variation at any given downstream location (Figure 1h). In situ foam generation therefore crucially enables controllable and reproducible reagent dispensing and rapid mixing. Furthermore, the intense mixing does not entail high shear rates, which can often lead to irreversible colloidal aggregation in macroscale methods involving mixing by agitation.44,45 A final attractive feature of such foams is the potential to perform elementary “digital” operations such as cell splitting and bifurcation in a microfluidic network, as illustrated in Figure 1i,j. Such features have attracted considerable attention lately for 3759
DOI: 10.1021/nl102478q | Nano Lett. 2010, 10, 3757-–3763
FIGURE 2. (a) A two-dimensional conceptual schematic of our nanoshell manufacturing method. (b) Diagrammatic representation of the experimental setup.
FIGURE 3. (a,b) TEM and SEM images of a population of complete smooth gold shells obtained with [Au3+] ∼ 0.42 mM while the volume fraction of gold-seeded silica particles (fS) and the shell thicknesses are (a) 1.5 × 10-3 % and ∼14 nm; (b) 0.6 × 10-3 % and ∼22 nm, respectively. The insets show magnified images of a single gold nanoshell. (c) Ensemble optical absorbance spectra for several different shell thicknesses from 14 to 40 nm. We observed that the shell thickness increases with increasing [Au3+]:fS ratios accompanied by a blue shift in the spectral peak from ∼670 to ∼635 nm. Scale bars correspond to (a) 100 nm, (b) 500 nm, and insets, 50 nm.
potential applications in autonomous logic-based multistep chemical processing operations.46,47 Electroless Plating of Dielectric Nanoparticles: Synthesis of Gold “Nanoshells” and Gold “Nanoislands”. The most commonly employed synthetic strategy for nanoshell fabrication involves aqueous-based electroless plating of nanometer-scale multicrystalline gold films onto silica nanoparticle surfaces (diameter d ) 50-200 nm) that have been preseeded with small gold nanoparticles (diameter d ) 2-5 nm).33,48 In this method, gold nanoparticle seeds on the silica particles serve as catalytic sites for the reduction of aqueous Au3+ to Au0 by reducing agents such as hydroxylamine.49 Electroless plating involves rapid autocatalytic reaction kinetics, and the reaction rate can be limited by diffusion of ionic species to the catalytic particle surfaces.50 An estimate for diffusionlimited linear growth rate (dr/dt) is obtained as dr/dt ∼ DCAuVM/ r, where D and CAu are respectively the diffusivity and concentration of Au3+, VM is the molar volume of gold, and r is the particle radius; this simple calculation yields dr/dt ∼ 200 nm/s for D ) 1 × 10-9 m2s-1, CAu ) 0.1 mM, r ) 5 nm, and VM ) 10 cm3/mol. Instantaneous linear growth rates can be as high as ∼200 nm/s, and nanoshell growth is typically completed within seconds.51 Growing uniformly thick shells on all particles in suspension therefore requires reagent dispensing and homogenization to be accomplished at time scales smaller than those for shell growth. This synthetic strategy is therefore extremely sensitive to the details of reagent dispensing in the reaction vessel. Processing in stirred batch vessels may also involve conflicting imperatives; for example when high stirring speeds required for rapid mixing can cause rapid shear-induced par© 2010 American Chemical Society
ticle agglomeration.44,45 Results from batch synthesis, where typically multimodal particle size distributions are obtained, are provided as Supporting Information (Figure S1). The reproducibility and scalability of this process is quite limited and other processing alternatives are clearly needed. Figure 2a is a two-dimensional conceptual schematic of our method, and Figure 2b is a diagrammatic representation of the experimental setup. Nitrogen gas, an aqueous mixture of gold-seeded silica particles in gold-plating solution (labeled S) and aqueous reducing agent solution (labeled R) are delivered continuously to a microfluidic T-shaped junction, where the reagents are dispensed as aqueous cells within a composite foam lattice, as mentioned above (Figure 1). The plating solution and reducing agent used were originally proposed by Graf and van Blaaderen in a study exploring the controlled synthesis of nanoshells.49 The flowing foam cells spend ∼120 s in the long microchannel downstream of the T-junction, and subsequently enter a collection vial where the gas escapes, and the aqueous and oil phases spontaneously form two immiscible fluid layers. We conducted a series of nanoshell growth experiments where we systematically explored the parameter space 3760
DOI: 10.1021/nl102478q | Nano Lett. 2010, 10, 3757-–3763
FIGURE 4. Particle size distribution of (a) silica particle cores with the inset showing SEM image of the particle population. Average size of the silica particle population is 177 ( 16 nm. (b-e) Size distribution of complete smooth gold nanoshells with insets showing TEM images of the corresponding particle population with average sizes (b) 205 ( 14, (c) 210 ( 22, (d) 221 ( 18, and (e) 260 ( 19 nm. Scale bars correspond to 200 nm.
The minimum complete shell thickness depends on the average spacing between gold nanoparticle seeds attached to the silica nanoparticle surface. For the gold-seeded silica used in our experiments, we estimate this spacing to be ∼6 nm. (Assuming 25% area coverage of the silica core surface by 3 nm gold seeds (an assumption made on the basis of electron microscopy observations), the total number of seeds attached to the silica surface (n) ) 0.25 × (projected area of a 175 nm silica particle)/(projected area of a 3 nm seed) ∼850. Further assuming equally spaced seeds, the distance between adjacent seeds works out to ∼6 nm.) Ensemble UV-vis absorbance measurements (Figure 3c)
comprised by the volume fraction of gold-seeded silica particles fS (%) and ionic gold concentration [Au3+] (mM) delivered into the aqueous cells of the foam lattice, while keeping all other reagent concentrations and flow rates fixed. Uniformly sized particles with complete, smooth gold shells were obtained in a subset of the parameter space (Figure 3a,b), where we could tune shell thicknesses from ∼10 to 40 nm. Histograms are provided in Figure 4; tight, unimodal particle size distributions are obtained under all synthesis conditions - a remarkable improvement over the small-scale batch synthesis of Figure S1 in Supporting Information. © 2010 American Chemical Society
3761
DOI: 10.1021/nl102478q | Nano Lett. 2010, 10, 3757-–3763
FIGURE 5. (a) TEM image of gold-seeded silica particles showing ∼3 nm gold seeds (dark spots) on silica particles. (b-d) TEM images illustrating gold seeds grown from ∼3 nm to (b) ∼10 nm, (c) ∼35 nm, and (d) nearly coalesced islands. (e) Ensemble optical absorbance measurements. The particles exhibit a broad spectral peak between 600 and 900 nm due to the presence of anisotropic islands of different sizes. (f-g) SEM images corresponding to TEM images in (a-d). Ionic gold concentration [Au3+] and volume-fraction of gold-seeded silica (fS) in the plating solution are as follows: (b) 0.29 mM and 1.2 × 10-2%, (c) 0.39 mM and 3.8 × 10-3% and (d) 0.4 mM and 3.4 × 10-3%, respectively. All scale bars correspond to 100 nm except for SEM image (f) which corresponds to 1 µm.
clearly indicated the presence of two spectral resonance peaks with a sharp lower wavelength peak at ∼650 nm that blue shifted with increasing shell thickness, and a broadshouldered peak at 900 nm. These peaks may be thought of as resulting from a hybridization of inner and outer shell plasmons.39,52 These ensemble measurements are in qualitative agreement with Mie theory calculations for light scattering from single nanoshells, made using measured average particle sizes (Supporting Information, Figure S2), thus complementing electron microscopy in indicating the presence of narrowly distributed particle populations. Dielectric nanoparticles decorated with small metallic islands (“nanoislands”) are also of considerable interest as substrates for a wide range of biological sensing applications.53 The optical properties of such particles depend on size and spacing of the metallic islands. Interactions between surface plasmons of neighboring islands can strongly affect the positions of the optical resonances.54 By operating in the goldlimited region of the fS-[Au3+] parameter space, we were able to synthesize tightly controlled populations of silica nanoparticles decorated with gold islands of various sizes (Figure 5b-d,f-h). Electron microscopy analysis of these particles also indicates a possible mechanism of nanoshell growth in which growing seed particles first coalesce into large, irregularly shaped islands that eventually coalesce to yield a complete shell ( Supporting Information, Figure S3). Further, we observe that island growth is anisotropic, being more pronounced in-plane along the silica surface than in the radial direction, leading to large flattened islands. The combined effect of anisotropy in island shape and localized surface plasmon interactions between islands is to red shift the absorption maximum (Figure © 2010 American Chemical Society
5e) with further peak broadening due to finite island size distributions.38 This is also seen in marked variations in the color of the nanoparticle suspensions (Supporting Information, Figure S4). This microfluidic method is simple to implement; reagent dispensing and rapid mixing is accomplished in robust, automated fashion with no operator intervention, and uniformly sized unaggregated particles requiring no postsynthesis treatment are obtained. There is no a priori limitation on the silica core sizes that can be used; this method can handle the full range of typical core sizes used for plasmonic nanoshell fabrication (50-500 nm). In general, for a given particle morphology (i.e., thin/thick shell or small/large nanoislands), different core sizes will necessitate device operation in different regions of the synthesis parameter space. The lower limit on shell thicknesses achievable by this method is set by the seed size and spacing as discussed above, and while we have fabricated shells up to 40 nm thick in this work, further parameter space exploration can yield even thicker shells. The method is not limited merely to electroless plating chemistries and can be readily adapted to a broad range of colloidal syntheses. As a further example, we have successfully synthesized gold nanorods involving seeded growth of small gold seed crystals in relatively concentrated surfactant solutions (Figure S5 of Supporting Information). In addition, the gas cells of the foam lattice may also carry a gas-phase reagent for colloidal chemistry, further broadening the spectrum of possibilities.55 However, we are currently limited to room-temperature aqueousbased colloidal chemistry due to the material used for device fabrication (PDMS). Extension to higher-temperatures and 3762
DOI: 10.1021/nl102478q | Nano Lett. 2010, 10, 3757-–3763
nonaqueous syntheses will require microfabrication in materials such as glass or silicon, and appropriate selection of the carrier “oil” phase, which has to be immiscible with the solvent(s) used, while preferentially wetting the microchannel surfaces. We are able to operate continuously at singledevice throughputs of ∼1 mL/h with silica volume fractions of up to 0.015% with the current generation of devices, and have tested device operation for up to 12 h per run. Even relatively modest parallelization can therefore yield preparative quantities of material. The inherently digital nature of foam-based nanoparticle processing also raises the possibility of autonomous logic-based processing for high-throughput applications such as, for example, rapid parameter-space exploration of colloidal syntheses. We thus present a new technological schema for continuous plasmonic nanoparticle synthesis, where we successfully bridge multiphase microfluidics and colloidal synthesis to make an important advance toward continuous sustainable nanomanufacturing processes.
(17) Yen, B. K. H.; Gunther, A.; Schmidt, M. A.; Jensen, K. F.; Bawendi, M. G. Angew. Chem., Int. Ed. 2005, 44 (34), 5447–5451. (18) Duraiswamy, S.; Khan, S. A. Small 2009, 5 (24), 2828–2834. (19) Wagner, J.; Kirner, T.; Mayer, G.; Albert, J.; Kohler, J. M. Chem. Eng. J. 2004, 101 (1), 251–260. (20) Carroll, N. J.; Rathod, S. B.; Derbins, E.; Mendez, S.; Weitz, D. A.; Petsev, D. N. Langmuir 2008, 24 (3), 658–661. (21) Khan, S. A.; Gunther, A.; Schmidt Martin, A.; Jensen Klavs, F. Langmuir 2004, 20 (20), 8604–8611. (22) Hassan, A. A.; Sandre, O.; Cabuila, V.; Tabeling, P. Chem. Commun. 2008, 1783–1785. (23) Kohler, J. M.; Romanus, H.; Hubner, U.; Wagner, J. J. Nanomater. 2007, . (24) Khan, S. A.; Jensen, K. F. Adv. Mater. 2007, 19 (18), 2556–2560. (25) Song, Y.; Hormes, J.; Kumar, C. S. S. R. Small 2008, 4 (6), 698–711. (26) Boleininger, J.; Kurz, A.; Reuss, V.; Sonnichsen, C. Phys. Chem. Chem. Phys. 2006, 8 (33), 3824–3827. (27) Kohler, J. M.; Wagner, J.; Albert, J. J. Mater. Chem. 2005, 15 (19), 1924–1930. (28) Hashimoto, M.; Garsteck, P.; Whitesides, G. M. Small 2007, 3 (10), 1792–1802. (29) Weaire, D.; Hutzler, S. The physics of foams; Oxford University Press: New York, 1999. (30) Khan, S. A.; Duraiswamy, S. Lab Chip 2009, 9, 1840–1842. (31) Chen, D. L.; Li, L.; Reyes, S.; Adamson, D. N.; Ismagilov, R. F. Langmuir 2007, 23, 2255–2260. (32) Zheng, B.; Ismagilov, R. F. Angew. Chem., Int. Ed. 2005, 44 (17), 2520–2523. (33) Oldenburg, S. J.; Averitt, R. D.; Westcott, S. L.; Halas, N. J. Chem. Phys. Lett. 1998, 288, 243–247. (34) Nehl, C. L.; Grady, N. K.; Goodrich, G. P.; Tam, F.; Halas, N. J.; Hafner, J. H. Nano Lett. 2004, 4 (12), 2355–2359. (35) Oldenburg, S. J.; Jackson, J. B.; Westcott, S. L.; Halas, N. J. Appl. Phys. Lett. 1999, 75 (19), 2897–2899. (36) Hirsch, L. R.; Stafford, R. J.; Bankson, J. A.; Sershen, S. R.; Rivera, B.; Price, R. E.; Hazle, J. D.; Halas, N. J.; West, J. L. Proc. Natl. Acad. Sci. U.S.A. 2003, 100 (23), 13549–13554. (37) Cortie, M.; Ford, M. Nanotechnology 2007, 18, 235704. (38) Bohren, C. F.; Huffman, D. R. Absorption and Scattering of Light by Small Particles; Wiley: New York, 1983. (39) Prodan, E.; Radloff, C.; Halas, N. J.; Nordlander, P. Science 2003, 302, 419–422. (40) Beh, W. S.; Kim, I. T.; Qin, D.; Xia, Y.; Whitesides, G. M. Adv. Mater. 1999, 11 (12), 1038–1041. (41) Engl, W.; Ohata, K.; Guillot, P.; Colin, A.; Panizza, P. Phys. Rev. Lett. 2006, 96, 134505-1–5. (42) Christopher, G. F.; Anna, S. L. J. Phys. D: Appl. Phys. 2007, 40, R319–R336. (43) Ottino, J. M. The Kinematics of Mixing: Stretching, Chaos, and Transport; Cambridge University Press: Cambridge, 1989. (44) Torres, F. E.; Russel, W. B.; Schowalter, W. R. J. Colloid Interface Sci. 1991, 145 (1), 51–73. (45) Torres, F. E.; Russel, W. B.; Schowalter, W. R. J. Colloid Interface Sci. 1991, 142 (2), 554–574. (46) Drenckhan, W.; Cox, S. J.; Delaney, G.; Holste, H.; Weaire, D.; Kern, N. Colloid. Surf., A 2005, 263, 52–64. (47) Prakash, M.; Gershenfeld, N. Science 2007, 315, 832–835. (48) Westcott, S. L.; Oldenburg, S. J.; Lee, T. R.; Halas, N. J. Langmuir 1998, 14, 5396–5401. (49) Graf, C.; Blaaderen, A. v. Langmuir 2002, 18, 524–534. (50) Stremsdoerfer, G.; Perrot, H.; Martin, J. R.; Clechet, P. J. Electrochem. Soc. 1988, 135 (11), 2881–2886. (51) Weiss, G. H. J. Stat. Phys. 1986, 42 (1/2), 3–36. (52) Prodan, E.; Nordlander, P. J. Chem. Phys. 2004, 120 (11), 5444– 5454. (53) Anker, J. N.; Hall, W. P.; Lyandres, O.; Shah, N. C.; Zhao, J.; Duyne, R. P. V. Nat. Mater. 2008, 7, 442–453. (54) Wang, H.; Goodrich, G. P.; Tam, F.; Oubre, C.; Nordlander, P.; Halas, N. J. J. Phys. Chem. B 2005, 109, 11083–11087. (55) Brinson, B. E.; Lassiter, J. B.; Levin, C. S.; Bardhan, R.; Mirin, N.; Halas, N. J. Langmuir 2008, 24, 14166–14171. (56) Nikoobakht, B.; El-Sayed, M. A. Chem. Mater. 2003, 15, 1957– 1962.
Acknowledgment. S.D. acknowledges a graduate scholarship from the National University of Singapore. We gratefully acknowledge research funding from NUS Grant R-279-000220-112/133 and the Chemical and Pharmaceutical Engineering Programme of the Singapore-MIT Alliance (SMA). We thank R. B. L. Chia and J. W. Chui for assistance in initial experiments. Finally, we thank Dr. R. Akkipedi at the A*STAR Institute of Materials Research and Engineering, Singapore for kindly facilitating access to cleanroom facilities. Supporting Information Available. (1) Batch synthesis of gold nanoshells. (2) Mie theory calculation for complete, smooth gold shells. (3) Anisotropic nanoisland growth. (4) Gold-seeded silica suspension, gold shells, and gold nanoislands. (5) Synthesis of rod-shaped gold nanocrystals in composite foams. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES AND NOTES (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)
Alivisatos, P. Nat. Biotechnol. 2004, 22 (1), 47–52. Engheta, N. Science 2007, 317, 1698–1702. Jain, P. K.; El-Sayed, M. A. Nano Lett. 2008, 8 (12), 4347–4352. Lassiter, J. B.; Knight, M. W.; Mirin, N. A.; Halas, N. J. Nano Lett. 2009, 9 (12), 4326–4332. Loo, C.; Lowery, A.; Halas, N.; West, J.; Drezek, R. Nano Lett. 2005, 5 (4), 709–711. Narayanan, R.; El-Sayed, M. A. Nano Lett. 2004, 4 (7), 1343–1348. Feldheim, D. L.; Colby, A.; Foss, J. Metal nanoparticles: Synthesis, characterization and Applications; Marcel Dekker, Inc.: New York, 2002. Lu, X.; Rycenga, M.; Skrabalak, S. E.; Wiley, B.; Xia, Y. Annu. Rev. Phys. Chem. 2009, 60, 167–92. Masala, O.; Seshadri, R. Annu. Rev. Mater. Res. 2004, 24, 41–81. Park, J.; Privman, V.; Matijevic, E. J. Phys. Chem. B 2001, 105, 11630–11635. Overbeek, J. T. G. Adv. Colloid Interface Sci. 1982, 15, 251–277. Kimling, J.; Maier, M.; Okenve, B.; Kotaidis, V.; Ballot, H.; Plech, A. J. Phys. Chem. B 2006, 110, 15700–15707. Pileni, M. P. J. Phys. Chem. 1993, 97, 6961–6973. Pileni, M. P. Nat. Mater. 2003, 2, 146–150. deMello, J.; deMello, A. Lab Chip 2004, 4 (2), 11N–15N. Nakamura, H.; Tashiro, A.; Yamaguchi, Y.; Miyazaki, M.; Watari, T.; Shimizu, H.; Maeda, H. Lab Chip 2004, 4 (3), 237–240. © 2010 American Chemical Society
3763
DOI: 10.1021/nl102478q | Nano Lett. 2010, 10, 3757-–3763
