Polymorphic Assembly from Beveled Gold Triangular Nanoprisms

Apr 26, 2017 - Wheeler High School, Marietta, Georgia 30068, United States. Nano Lett. , 2017, 17 (5), pp 3270–3275. DOI: 10.1021/acs.nanolett.7b009...
0 downloads 0 Views 3MB Size
Letter pubs.acs.org/NanoLett

Polymorphic Assembly from Beveled Gold Triangular Nanoprisms Juyeong Kim,†,‡ Xiaohui Song,†,‡ Fei Ji,∥ Binbin Luo,† Nicole F. Ice,⊥ Qipeng Liu,∥ Qiao Zhang,∥ and Qian Chen*,†,‡,§ †

Department of Materials Science and Engineering, ‡Frederick Seitz Materials Research Laboratory, and §Department of Chemistry, University of Illinois, Urbana, Illinois 61801, United States ∥ Institute of Functional Nano and Soft Materials (FUNSOM), Collaborative Innovation Center for Suzhou Nano Science and Technology (NANO-CIC), Jiangsu Key Laboratory for Carbon-Based Functional Materials and Devices, Soochow University, Suzhou, Jiangsu 215123, People’s Republic of China ⊥ Wheeler High School, Marietta, Georgia 30068, United States S Supporting Information *

ABSTRACT: The shape anisotropy of nanoparticle building blocks is of critical importance in determining their packing symmetry and assembly directionality. While there has been extensive research on the effect of their overall geometric shapes, the importance of nanometer morphology details is not well-recognized or understood. Here we draw on shapeanisotropic gold triangular nanoprism building blocks synthesized based on a method we recently developed; besides the “large-scale” triangular prism shape (79.8 nm in side length and 22.0 nm in thickness), the prisms are beveled with their sides convexly enclosed by two flat {100} facets. We engineer the balance between electrostatic repulsion and entropically driven depletion attraction in the system to generate self-assemblies without or with the effect of the nanoscale beveling detail. A conventional, planar honeycomb (p-honeycomb) lattice forms with the triangular basal planes packed on the same plane at low depletion attraction, whereas an unexpected interlocking honeycomb (i-honeycomb) lattice and its “supracrystal” forms are assembled with additional close-paralleling of side facets at high depletion attraction. The i-honeycomb lattice renders all the metallic surfaces in close proximity and leads to a surface-enhanced Raman scattering signal nearly 5-fold higher than that in the p-honeycomb lattice and high sensitivity for detecting the model molecule Rhodamine 6G at a concentration as low as 10−8 M. Our study can guide future work in both nanoparticle synthesis and self-assembly; nanoscale geometrical features in anisotropic nanoparticles can be used as an important handle to control directional interactions for nonconventional ordered assemblies and to enrich diversity in self-assembly structure and function. KEYWORDS: Gold triangular nanoprism, directional self-assembly, shape anisotropy, depletion attraction, morphology details, SERS

S

Selective chemical modifications can sometimes augment the role of small surfaces by rendering them strongly attractive,19−21 but there are limited chemistry options for precisely selective ligand coating. Here, we use highly monodisperse beveled gold triangular prisms (side = 79.8 ± 3.7 nm, thickness = 22.0 ± 1.7 nm) with two types of interaction surfaces, the basal plane and the beveled side, as the building blocks with high aspect ratio and nanometer morphology details (Figure 1a).22 These prisms are coated by positively charged cetyltrimethylammonium chloride (CTAC) ligands to render them electrostatically repulsive. They are geometrically intriguing with two protruded {100} facets intersecting at an angle of 120−126° on the sides (Figures 1a, S1, and S2). The two intersecting facets yield

hape-anisotropic nanoparticles are excellent building blocks for engineering collective properties through self-assembly, partially due to their unique functions from asymmetric quantum-confinement and partially due to their greatly enriched self-assembly structures from directional interactions.1−6 Shape anisotropy encodes intricate symmetries in the final assemblies. For example, faceted anisotropic nanoparticles often preferentially self-assemble to maximize facet− facet contact,7 such as the cubic lattice for nanocubes8,9 and Minkowski lattice for nanosized octahedra.10,11 Regarding nanoparticles of high aspect ratio, such as plate- and rodlike shapes, the dominantly large facets often determine the assembly symmetry. For instance, nanoplates regardless of composition have been shown to entropically favor selfassembly into columnar stacks in the face-to-face configuration,12−16 and nanorods assemble via side-by-side attachment.17,18 The facets with smaller surface areas are meanwhile not well-recognized or utilized in the self-assembly structures. © 2017 American Chemical Society

Received: March 6, 2017 Revised: April 16, 2017 Published: April 26, 2017 3270

DOI: 10.1021/acs.nanolett.7b00958 Nano Lett. 2017, 17, 3270−3275

Letter

Nano Letters

Figure 1. Beveled gold triangular prism assembly via droplet evaporation. (a) Schematics of the assembly process. Left: A water droplet (light blue) on a silicon wafer consists of individual prisms (blue) and CTAC micelles (green sphere). The zoomed-in views show the beveled geometry of a prism, including a schematic on the left and an SEM image on the right, and the structure of a CTAC micelle. Right: Schematics of the depletion attraction-driven assembly process and the self-assembled i-honeycomb lattice. The top view schematic of the i-honeycomb lattice on the silicon wafer features the staggered alignment between two adjacent prisms by half of the single prism thickness. The side view schematic shows the closely packed configuration as side-by-side attachment. (b−d) SEM images of the i-honeycomb lattice (from left to right: a high-resolution image, a singlelayer, and multilayers). The orange lines highlight the beveled sides in the i-honeycomb lattice. Scale bars: 10 nm for (a), 40 nm for (b), and 100 nm for (c,d).

such as cCTAC, temperature, and surface ligand chemistry to elucidate the depletion-driven assembly mechanism. The nonconventional interlocking honeycomb (i-honeycomb) lattice is observed to form in a relatively wide window of cCTAC and prism concentration (cprism) (cCTAC = 5−10 mM and cprism with a peak value of 8−16 in extinction at 665 nm, Figure S3). As shown in Figure 1, each prism stands vertically on the substrate and is closely packed side-by-side (side view). The top view schematics and scanning electron microscopy (SEM) images clearly show the interlocked secondary stacking of columns (Figure 1b−d), where the beveled sides alternately stack. In this packing, even the tiny interstices between columns caused by the beveled sides are completely filled. The gap distance between adjacent prisms is measured as 5−6 nm, consistent with CTAC ligand thickness on the prism surface.13,27 At low magnifications (Figure 1d), we even see staircases of the interlocked layers, highlighting the threedimensional (3D) ordering of the i-honeycomb lattice. Because of the beveled sides, the i-honeycomb lattice renders a more efficient tessellation of 3D space than the commonly observed planar honeycomb lattice (p-honeycomb).24,28,29 Namely, the geometry packing density is 100% for the i-honeycomb lattice if not considering ligand layers and only 81% for the phoneycomb lattice where the beveled geometry gives rise to voids between prisms (Figure S4). We show that a seemingly trivial nanoscopic detail of beveled side matters to the packing geometry of the final assembled structure. We attribute the formation of the i-honeycomb lattice, a structure with maximized surface contact given the building block geometry, to an intricate interplay between depletion attraction (by CTAC micelles) and electrostatic repulsion. Depletion attraction has a purely entropic origin and arises by maximizing the translational entropy of small, nonadsorbing depletants, such as the CTAC micelles here, in the particle suspension.13,30 In such systems, assembling particles possess

additional interacting surfaces and directionalities on the prism side when interparticle interaction takes effect, which is distinct from triangular prisms with flat sides reported in literature.13,23,24 We use depletion attraction (CTAC micelles as the depletant) as the tunable driving force for self-assembly, which is a nonspecific entropic effect independent of particle surface chemistry. We observe that self-assembled structures are selected thermodynamically, based on whether the area of the beveled prism side is negligible or not, an effect that is not observed with thin triangular prisms with flat sides13 or equilateral polygonal shapes.11 The prisms are self-assembled on silicon substrates via a droplet evaporation method in the presence of depletants. Our assembly method enables facile control and use of the assembled architectures for later surface-enhanced Raman scattering (SERS) measurements. The conventional droplet evaporation method generates closely packed assemblies that are kinetically formed by capillary forces and convective liquid flow, which move particles in the droplet until they settle on substrates.25,26 In contrast, we optimize parameters to make the self-assembly process thermodynamically controlled. Specifically, a drop (10 μL) of the prism solution is placed on a silicon wafer. The prism solution contains CTAC molecules of various initial concentrations (CTAC concentration (cCTAC) = 1.5−10 mM); the CTAC molecules form into small micelles (diameter = 5−6 nm13) in water and serve as the depletants (Figure 1a). The droplet evaporates slowly while covered by a lid at room temperature. The CTAC micelle concentration and the depletion attraction strength are thus gradually increased during solvent drying to overrun decreasing electrostatic repulsion (due to more concentrated counterion screening), which produces self-assembly structures close-packed to different extent on the substrate. We investigate changes in the assembly configuration with respect to various parameters 3271

DOI: 10.1021/acs.nanolett.7b00958 Nano Lett. 2017, 17, 3270−3275

Letter

Nano Letters

Figure 2. CTAC concentration effects on depletion attraction-driven self-assembly. (a) SEM images of prism assembly structures under different initial cCTAC (from left to right: 0.1, 1.5, 3.5, 5, and 10 mM). The schematics represent the corresponding self-assembled structures (0.1 mM, disordered prisms; 1.5 mM, single-layer p-honeycomb; 3.5 mM, multilayer p-honeycomb; 5 mM, single-layer i-honeycomb; and 10 mM, multilayer ihoneycomb). The prisms stand vertically on the substrate with cCTAC > 3.5 mM. (b) Schematics of the volume overlap of exclusion layers between two prisms in different configurations. The dark purple shapes are the prisms, the light purple layer shows the exclusion layer surrounding a prism, and the middle purple layer shows the volume overlap. The prism has three elemental overlapping configurations for exclusion layers, the parallel side-by-side ΔVp−side, the interlocking side-by-side ΔVi−side and the face-to-face ΔVface. (c) A graph showing the relative volume overlap ΔV (red bar in arbitrary unit) as a function of configuration in different assembled structures. The ΔV values are normalized with that of the multilayer ihoneycomb lattice as 1. Scale bars: 100 nm.

trend matches with our experimental observations of the four assemblies following an increase in cCTAC and correspondingly in depletion attraction strength (Figure 2a), each in extended regions on the substrate. This consistency shows that our precise control of depletion attraction strength distinguishes different levels of close-packing in the final structures. To corroborate the assembly mechanism based on the interplay of solution-mediated interactions, we directly characterize the assemblies formed in the solution and eliminate possible effects from the substrate and the capillary effect during solvent drying. The hypothesis is as follows. In the droplet evaporation process, gradual solvent drying concentrates CTAC micelles (increasing n) so that depletion attraction overruns the decreasing electrostatic repulsion, which drives the self-assembly of the prisms. The increased mass of the assembled structures causes them to fall down on the substrate. To validate this hypothesis, we keep a prism suspension containing high cCTAC (100 mM) for a prolonged period (without solvent evaporation) at room temperature. This process allows individual prisms to self-assemble and reach thermodynamic equilibrium based on internanoparticle interactions, and generates eventual formation of black sediments at the bottom of the solution (Figure S7). The sediments are carefully transferred on a silicon wafer and flash-dried under vacuum. The SEM images of the sediments consist of micronsized 3D supracrystals with a few micrometers in size lying on the substrate (Figure 3a and see more examples in Figure S8). In essence, the enlarged view of the supracrystal (Figure 3a) clearly shows prisms closely packed and interlocked with each other, which is identical to the prism arrangements in the ihoneycomb lattice formed via the droplet evaporation (Figure 2a). The long-range order of the prisms is shown throughout the surface facets of the supracrystal in the long axis, and the tilted view also demonstrates the long-range periodicity in the interior of the supracrystal (Figures 2a and S8). In addition, an ensemble analysis of the supracrystals through small-angle X-

an exclusion layer wrapping their surface with a thickness equal to the radius of depletants. The total overlapping volume of exclusion layers through particle self-assembly lowers the free energy of the system by making more space available to the depletants in the solution following Edep = −nkBTΔV, where n is the number density of the depletant, kB is the Boltzmann constant, T is the temperature, and ΔV is the volume gained by the overlap of exclusion layers. From this relation, depletion attraction favors directional attachments of flat facets, more so as it overruns the competing facet−facet electrostatic repulsion. This clean geometry preference makes depletion attraction especially suitable for controlling faceted nanoparticle selfassembly.12−14 Still, the use of depletion in previous nanoparticle self-assembly studies is different from ours in that the effect of depletion attraction on utilizing nanoscopic morphology details is not well demonstrated. The prism self-assembly is shown to indeed depend on the initial cCTAC and the consequent depletant concentration in the droplet (Figure S5). As cCTAC increases, the depletion attraction increases and gradually wins over the interparticle electrostatic repulsion; the former favors facet−facet packing to maximize exclusion layer overlap while the latter tends to minimize facet−facet packing.16 As shown in Figure 2a, given other conditions the same, disordered structures are collected on the substrate at a very low cCTAC (0.1 mM), at which the CTAC molecules do not even form into depletant micelles to begin with (critical micelle concentration = 1 mM31). Electrostatic repulsion thus dominates and leads to no assemblies but random arrangements sitting on the substrate. At a higher cCTAC where depletion attraction comes into play, assemblies with increasing facet−facet packing are favored. We calculate the volume overlap ΔV of exclusion layers for four possible selfassembled structures (Figure 2b,c). The derived ΔV diagram (Figures 2c and S6) shows that ΔV increases sequentially from the case of single-layer p-honeycomb, multilayer p-honeycomb, single-layer i-honeycomb, to multilayer i-honeycomb. This 3272

DOI: 10.1021/acs.nanolett.7b00958 Nano Lett. 2017, 17, 3270−3275

Letter

Nano Letters

as the temperature increases, in that the entropic effect is weighted higher in the final free energy. To illustrate this effect, two droplets with the same cCTAC and cprism are dried under different temperatures given the same drying speed (Figure 3b). The one at room temperature (25 °C) generates the phoneycomb lattice (cCTAC = 3.5 mM), while the other kept in a closed oven at 60 °C gives rise to the i-honeycomb lattice. This observation is consistent with the more favored depletion attraction at higher temperature according to literature.13 Specifically, based on Edep = −nkBTΔV, the temperature term alone contributes to 11% increase in depletion attraction (from 25 to 60 °C). In addition, it has been reported that the micelle concentration n also increases up to 44% with this much temperature increase, which further enhances the role of depletion attraction.13 Depletion attraction does not concern surface chemistry of nanoparticles, which can serve as a versatile control handle in nanoparticle self-assembly. We demonstrate this feature by using prisms of different surface chemistry. The native CTAC ligands due to synthesis are exchanged with thiol ligands (2(boc-amino)ethanethiol) (see more details in the SI), which are shorter than the CTAC ligands while also being positively charged. Uniform distribution of sulfur elements on the prism basal plane in the EDS elemental mapping (Figure 3c) shows that the ligand exchange is successful. It is remarkable that the thiol-modified prisms also self-assemble into the i-honeycomb lattice under the same condition as that for the CTAC-coated prisms in the presence of the same concentration of depletants (Figure 3c, and see detailed assembly conditions in the SI). Besides 2-(boc-amino)ethanethiol, we find that the prisms ligand-exchanged with 11-mercaptoundecanoic acid, a representative negatively charged ligand, are also assembled into the i-honeycomb lattice (Figure S10), supporting the versatility of our method. Note that the excess CTAC ligands added after ligand exchange do not trigger further ligand exchange with the thiols already bound on the prism surface due to the much stronger bonding between sulfur and gold atoms. This observation is distinctive from conventional droplet evaporation-driven self-assembly, where an exchange in ligands can lead to very different final self-assembled structures due to differences in ligand length, charge density, and packing geometry.20,24,34,35 The different packing density results in distinctive SERS response; the i-honeycomb lattice has a 5-fold increase in SERS performance compared to the p-honeycomb lattice (Figure 4). The two superlattice configurations with multilayers are prepared separately on the silicon wafer, and their SERS behaviors are investigated under an incident laser source with 632 nm wavelength, which is close to the extinction wavelength for a single prism in solution (Figure S3). The R6G molecules as a probe are added to the prism solution and adsorbed naturally onto prism surfaces during the droplet evaporation self-assembly. R6G has characteristic features for carbon skeleton stretching modes at 1180, 1314, 1364, 1510, 1575, and 1650 cm−1.22 We hypothesize the SERS signals originate from R6G on ordered prisms on the top layer in that the multilayers below the top layer do not affect the SERS. Large assembled areas of both configurations are scanned using a confocal Raman spectrometer and show uniform distribution of scattering intensity (insets in Figure 4a,b), indicating homogeneously distributed hot spots and R6G molecules (see SEM images of the assemblies in Figure S11 and consistent signal enhancement on other large areas of the i-

Figure 3. Versatility of the depletion attraction-driven assembly. (a) SEM images of supracrystals that are assembled in the solution phase containing 100 mM of CTAC. Different magnifications (high to low) show the structural details of supracrystals from nanoscale to micronscale. (b) Schematics of the drying condition at different temperatures (T = 60 °C and room temperature), and SEM images of i-honeycomb lattices (T = 60 °C) and p-honeycomb lattices (room temperature). (c) Schematic of ligand exchange from CTAC to 2-(boc-amino)ethanethiol on the prism surface (top) and energy dispersive spectroscopy (EDS) maps of a single prism after the ligand exchange (bottom: green for sulfur and red for gold). The SEM image shows that the i-honeycomb lattice is obtained using the thiolated prisms. Scale bars: 100 nm, 200 nm, and 1 μm for a (from left to right), 100 nm for b, and 50 nm for EDS and 100 nm for the SEM image in c.

ray scattering displays 1D lamellar structure consistent with the face-to-face stacking of the prisms (Figure S9). The signal from lateral prism ordering is not detected clearly, presumably due to relatively smaller occurrence of the lateral ordering than the face-to-face stacking direction. The formation of the supracrystal in high cCTAC indicates the i-honeycomb lattice is a thermodynamically stable configuration, consistent with our hypothesis. The supracrystals adopt their own surface “facets” as elongated hexagonal prisms in two types about micron in size, a rectangular facet with the beveled prism sides and tips exposed, and a hexagonal facet with the planar prism faces. The majority of the intact supracrystals are measured as ∼8 μm in length and ∼2 μm in width. To the best of our knowledge, this is the first supracrystal composed of triangular-shaped components and with micron-sized facets in high aspect ratio. The free-standing supracrystals are likely to have unique plasmonic properties because they are an ensemble of individual nanoparticle building blocks (hot spots) in 3D, and the mesoporosity between particles may enable their use for delivering small molecules.32,33 Unlike most enthalpic interactions that are weakened relatively as temperature increases, depletion is more prominent 3273

DOI: 10.1021/acs.nanolett.7b00958 Nano Lett. 2017, 17, 3270−3275

Letter

Nano Letters

gold triangular prisms. These morphological details can give rise to a unique localized electric field distribution and plasmonic coupling. Concentration of the electric field occurs more intensely on the tip areas than the side areas within a single prism (Figure 4e) and the electric field intensity increases dramatically with the ordered lattices (7.8 for the single prism, 15.4 for the p-honeycomb lattice, and 63.1 for the i-honeycomb lattice). Specifically, the electric field intensity is four times stronger with the i-honeycomb lattice than that of the phoneycomb lattice, which is in good agreement with our experimental observations in SERS. The sharp tips and bevels in the superlattices promote high intensity electric fields as hot spots.37,38 It appears that the stronger electric field enhancement with the i-honeycomb lattice originates from two plausible factors: (i) higher density of ordered hot spots that contribute to a plasmonic antenna effect39−41 and (ii) beveled sides that produce intense electromagnetic radiation through a “lightning rod” effect.42−44 In this work, we show using the examples of beveled gold triangular prisms whose nanoscopic morphology details can induce unexpected self-assembly structures of superior functions for applications. We see both an emphasis on shape details and a depletion-based control of geometric packing as applicable to other nanoparticles with different morphological details45−48 and elemental compositions such as silver,49,50 which will broaden diversity of anisotropic nanoparticle selfassembly structures and functions. In addition, our approach can be applied to other assembly methodologies as a control handle for the assembly structure, such as pressure-directed self-assembly,51,52 droplet evaporation on patterned substrates53,54 and field-directed assembly.55 Beyond nanoparticle self-assembly, our study draws interesting parallels with the effects of shape complementarity observed in the interactions of “living” matter, where adaptive nanoscopic morphology details can modulate protein−protein or protein−DNA interactions.56−58

Figure 4. SERS measurements for p-honeycomb and i-honeycomb lattices. (a,b) Optical microscopy images of p-honeycomb lattices and i-honeycomb lattices on silicon wafer substrates, respectively (see SEM images in Figure S11). The bright gray areas correspond to locations of assembled prisms. The dark gray areas are bare silicon wafer. The inset images show Raman signal map of the black boxed area measured by a confocal Raman spectrometer (excited source wavelength = 632 nm, [R6G] = 10−4 M). (c) Raman spectra of different assembly structures prepared on a silicon wafer separately with [R6G] = 10−4 M (red curve, i-honeycomb lattice; blue curve, p-honeycomb lattice; gray curve, scattered single prisms). (d) Raman spectra of the i-honeycomb lattice with different dye concentrations ([R6G] = 10−4 M (red), 10−5 M (brown), 10−6 M (yellow), 10−7 M (green), and 10−8 M (blue)). The inset graph shows the rescaled signal with [R6G] = 10−8 M, which is detectable over the level of the noise. (e) Electric field distribution of a single prism, p-honeycomb lattice, and i-honeycomb lattice calculated by finite-difference time-domain (FDTD) simulations under an incident laser source with 632 nm wavelength. The configurations are shown from the top view. Scale bars: 10 μm.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.7b00958. Materials and Methods, Table S1, and Figures S1−S13 (PDF)

honeycomb lattice in Figure S12). The corresponding spectra clearly show the intensity difference at 1510 cm−1, 5.5 × 104 counts for the i-honeycomb lattice and 1.1 × 104 counts for the p-honeycomb lattice (Figure 4c). Although both configurations share similar interparticle spacing (∼5 nm) owing to the presence of CTAC surface layers, the degree of plasmonic coupling between adjacent prism tips/edges differs from each other. To further investigate the enhancement property of the ihoneycomb lattice, we decrease the probe molecule concentration from 10−4 to 10−8 M (Figure 4d). The i-honeycomb lattice detects concentrations of R6G as low as 10−8 M. In addition, the i-honeycomb lattice shows polarization-dependent SERS activity (Figure S13), which may be inherent in the hot spots from prism tips and beveled edges ordered in the ihoneycomb lattice.36 We attribute the superior sensitivity to enriched hot spots from the interlocked prism tips and bevels. From our FDTD simulations (see details in the SI), the beveled prisms retain six beveled side facets containing additional edges and three tips relatively sharpened by the bevels, compared with traditional



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Qiao Zhang: 0000-0001-9682-3295 Qian Chen: 0000-0002-1968-441X Author Contributions

J.K. and X.S. contributed equally. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Cong Xu at the University of Illinois for help with the EDS element mapping. The nanoparticle synthesis, 3274

DOI: 10.1021/acs.nanolett.7b00958 Nano Lett. 2017, 17, 3270−3275

Letter

Nano Letters

(30) Rossi, L.; Sacanna, S.; Irvine, W. T. M.; Chaikin, P. M.; Pine, D. J.; Philipse, A. P. Soft Matter 2011, 7, 4139−4142. (31) Imae, T.; Kamiya, R.; Ikeda, S. J. Colloid Interface Sci. 1985, 108, 215−225. (32) Wang, T.; Zhuang, J.; Lynch, J.; Chen, O.; Wang, Z.; Wang, X.; LaMontagne, D.; Wu, H.; Wang, Z.; Cao, Y. C. Science 2012, 338, 358. (33) Huang, M. H.; Thoka, S. Nano Today 2015, 10, 81−92. (34) Choi, J. J.; Bealing, C. R.; Bian, K.; Hughes, K. J.; Zhang, W.; Smilgies, D.-M.; Hennig, R. G.; Engstrom, J. R.; Hanrath, T. J. Am. Chem. Soc. 2011, 133, 3131−3138. (35) Gao, B.; Arya, G.; Tao, A. R. Nat. Nanotechnol. 2012, 7, 433− 437. (36) Liu, K.-K.; Tadepalli, S.; Kumari, G.; Banerjee, P.; Tian, L.; Jain, P. K.; Singamaneni, S. J. Phys. Chem. C 2016, 120, 16899−16906. (37) Lee, Y. H.; Shi, W.; Lee, H. K.; Jiang, R.; Phang, I. Y.; Cui, Y.; Isa, L.; Yang, Y.; Wang, J.; Li, S.; Ling, X. Y. Nat. Commun. 2015, 6, 6990. (38) Shi, Q.; Si, K. J.; Sikdar, D.; Yap, L. W.; Premaratne, M.; Cheng, W. ACS Nano 2016, 10, 967−976. (39) Giannini, V.; Fernández-Domínguez, A. I.; Heck, S. C.; Maier, S. A. Chem. Rev. 2011, 111, 3888−3912. (40) Alonso-González, P.; Albella, P.; Schnell, M.; Chen, J.; Huth, F.; García-Etxarri, A.; Casanova, F.; Golmar, F.; Arzubiaga, L.; Hueso, L. E.; Aizpurua, J.; Hillenbrand, R. Nat. Commun. 2012, 3, 684. (41) Rosen, D. A.; Tao, A. R. ACS Appl. Mater. Interfaces 2014, 6, 4134−4142. (42) Gersten, J.; Nitzan, A. J. Chem. Phys. 1980, 73, 3023−3037. (43) Orendorff, C. J.; Gole, A.; Sau, T. K.; Murphy, C. J. Anal. Chem. 2005, 77, 3261−3266. (44) Saute, B.; Narayanan, R. Analyst 2011, 136, 527−532. (45) Ye, X.; Gao, Y.; Chen, J.; Reifsnyder, D. C.; Zheng, C.; Murray, C. B. Nano Lett. 2013, 13, 2163−2171. (46) Zhong, Y.; Wang, J.; Zhang, R.; Wei, W.; Wang, H.; Lü, X.; Bai, F.; Wu, H.; Haddad, R.; Fan, H. Nano Lett. 2014, 14, 7175−7179. (47) Leary, R. K.; Kumar, A.; Straney, P. J.; Collins, S. M.; Yazdi, S.; Dunin-Borkowski, R. E.; Midgley, P. A.; Millstone, J. E.; Ringe, E. J. Phys. Chem. C 2016, 120, 20843−20851. (48) Wu, X.-J.; Chen, J.; Tan, C.; Zhu, Y.; Han, Y.; Zhang, H. Nat. Chem. 2016, 8, 470−475. (49) Zhang, Q.; Hu, Y.; Guo, S.; Goebl, J.; Yin, Y. Nano Lett. 2010, 10, 5037−5042. (50) Zeng, J.; Xia, X.; Rycenga, M.; Henneghan, P.; Li, Q.; Xia, Y. Angew. Chem., Int. Ed. 2011, 50, 244−249. (51) Bai, F.; Bian, K.; Li, B.; Wu, H.; Alarid, L. J.; Schunk, H. C.; Clem, P. G.; Fan, H. MRS Bull. 2015, 40, 961−970. (52) Li, B.; Bian, K.; Lane, J. M. D.; Salerno, K. M.; Grest, G. S.; Ao, T.; Hickman, R.; Wise, J.; Wang, Z.; Fan, H. Nat. Commun. 2017, 8, 14778. (53) Wen, T.; Zhang, D.; Wen, Q.; Zhang, H.; Liao, Y.; Li, Q.; Yang, Q.; Bai, F.; Zhong, Z. Nanoscale 2015, 7, 4906−4911. (54) Paik, T.; Yun, H.; Fleury, B.; Hong, S.-H.; Jo, P. S.; Wu, Y.; Oh, S.-J.; Cargnello, M.; Yang, H.; Murray, C. B.; Kagan, C. R. Nano Lett. 2017, 17, 1387−1394. (55) Singh, G.; Chan, H.; Baskin, A.; Gelman, E.; Repnin, N.; Král, P.; Klajn, R. Science 2014, 345, 1149. (56) Li, Y.; Zhang, X.; Cao, D. Sci. Rep. 2013, 3, 3271. (57) King, N. P.; Bale, J. B.; Sheffler, W.; McNamara, D. E.; Gonen, S.; Gonen, T.; Yeates, T. O.; Baker, D. Nature 2014, 510, 103−108. (58) Bai, Y.; Luo, Q.; Liu, J. Chem. Soc. Rev. 2016, 45, 2756−2767.

characterization (TEM, SEM, and SAXS), and self-assembly were supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering under award no. DE-FG02-07ER46471, through the Frederick Seitz Materials Research Laboratory at the University of Illinois. The SERS measurement and the FDTD calculations were supported by National Science Foundation, Grant NSF CHE 13-03757. N.F.I. acknowledges the National Science Foundation under Grant NSF EEC 14-07194 RET as part of the nano@illinois Research Experiences for Teachers (RET) project.



REFERENCES

(1) Quan, Z.; Fang, J. Nano Today 2010, 5, 390−411. (2) Sau, T. K.; Rogach, A. L.; Jäckel, F.; Klar, T. A.; Feldmann, J. Adv. Mater. 2010, 22, 1805−1825. (3) Lu, F.; Yager, K. G.; Zhang, Y.; Xin, H.; Gang, O. Nat. Commun. 2015, 6, 6912. (4) Thorkelsson, K.; Bai, P.; Xu, T. Nano Today 2015, 10, 48−66. (5) Yang, P.; Zheng, J.; Xu, Y.; Zhang, Q.; Jiang, L. Adv. Mater. 2016, 28, 10508−10517. (6) Smith, A. F.; Weiner, R. G.; Skrabalak, S. E. J. Phys. Chem. C 2016, 120, 20563−20571. (7) Torquato, S.; Jiao, Y. Nature 2009, 460, 876−879. (8) Rycenga, M.; McLellan, J. M.; Xia, Y. Adv. Mater. 2008, 20, 2416−2420. (9) Li, Z.; Okasinski, J. S.; Gosztola, D. J.; Ren, Y.; Sun, Y. J. Mater. Chem. C 2015, 3, 58−65. (10) Damasceno, P. F.; Engel, M.; Glotzer, S. C. ACS Nano 2012, 6, 609−614. (11) Liao, C.-W.; Lin, Y.-S.; Chanda, K.; Song, Y.-F.; Huang, M. H. J. Am. Chem. Soc. 2013, 135, 2684−2693. (12) Saunders, A. E.; Ghezelbash, A.; Smilgies, D.-M.; Sigman, M. B.; Korgel, B. A. Nano Lett. 2006, 6, 2959−2963. (13) Young, K. L.; Jones, M. R.; Zhang, J.; Macfarlane, R. J.; EsquivelSirvent, R.; Nap, R. J.; Wu, J.; Schatz, G. C.; Lee, B.; Mirkin, C. A. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 2240−2245. (14) Zhang, J.; Lang, P. R.; Meyer, M.; Dhont, J. K. G. Langmuir 2013, 29, 4679−4687. (15) Jana, S.; Phan, T. N. T.; Bouet, C.; Tessier, M. D.; Davidson, P.; Dubertret, B.; Abécassis, B. Langmuir 2015, 31, 10532−10539. (16) Kim, J.; Jones, M. R.; Ou, Z.; Chen, Q. ACS Nano 2016, 10, 9801−9808. (17) Li, L.-s.; Walda, J.; Manna, L.; Alivisatos, A. P. Nano Lett. 2002, 2, 557−560. (18) Ye, X.; Jin, L.; Caglayan, H.; Chen, J.; Xing, G.; Zheng, C.; Doan-Nguyen, V.; Kang, Y.; Engheta, N.; Kagan, C. R.; Murray, C. B. ACS Nano 2012, 6, 2804−2817. (19) Caswell, K. K.; Wilson, J. N.; Bunz, U. H. F.; Murphy, C. J. J. Am. Chem. Soc. 2003, 125, 13914−13915. (20) Nie, Z.; Fava, D.; Kumacheva, E.; Zou, S.; Walker, G. C.; Rubinstein, M. Nat. Mater. 2007, 6, 609−614. (21) Chakrabortty, S.; Guchhait, A.; Ong, X.; Mishra, N.; Wu, W.-Y.; Jhon, M. H.; Chan, Y. Nano Lett. 2016, 16, 6431−6436. (22) Chen, L.; Ji, F.; Xu, Y.; He, L.; Mi, Y.; Bao, F.; Sun, B.; Zhang, X.; Zhang, Q. Nano Lett. 2014, 14, 7201−7206. (23) Millstone, J. E.; Park, S.; Shuford, K. L.; Qin, L.; Schatz, G. C.; Mirkin, C. A. J. Am. Chem. Soc. 2005, 127, 5312−5313. (24) Walker, D. A.; Browne, K. P.; Kowalczyk, B.; Grzybowski, B. A. Angew. Chem., Int. Ed. 2010, 49, 6760−6763. (25) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R.; Witten, T. A. Nature 1997, 389, 827−829. (26) Hu, H.; Larson, R. G. Langmuir 2005, 21, 3963−3971. (27) Alkilany, A. M.; Frey, R. L.; Ferry, J. L.; Murphy, C. J. Langmuir 2008, 24, 10235−10239. (28) Agarwal, U.; Escobedo, F. A. Nat. Mater. 2011, 10, 230−235. (29) Fu, Q.; Ran, G.; Xu, W. Nano Res. 2016, 9, 3247−3256. 3275

DOI: 10.1021/acs.nanolett.7b00958 Nano Lett. 2017, 17, 3270−3275