Polymer-Directed Growth of Plasmonic Aluminum Nanocrystals

Aug 28, 2018 - This work makes it clear that chemical synthesis of Al NCs with well-controlled shapes, sizes, and structures is practically feasible...
0 downloads 0 Views 6MB Size
Article pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Polymer-Directed Growth of Plasmonic Aluminum Nanocrystals Shaoyong Lu,† Hua Yu,† Samuel Gottheim,‡,§ Huimin Gao,∥ Christopher J. DeSantis,‡,§ Benjamin D. Clark,‡,§ Jian Yang,§,⊥ Christian R. Jacobson,‡,§ Zhongyuan Lu,*,†,∥ Peter Nordlander,*,§,⊥,#,∇ Naomi J. Halas,*,§,⊥,#,∇ and Kun Liu*,†

J. Am. Chem. Soc. Downloaded from pubs.acs.org by RMIT UNIV on 10/30/18. For personal use only.



State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012, P.R. China ‡ Department of Chemistry, Rice University, Houston, Texas 77005, United States § Laboratory for Nanophotonics, Rice University, Houston, Texas 77005, United States ∥ Institute of Theoretical Chemistry, Jilin University, Changchun 130021, P.R. China ⊥ Department of Physics & Astronomy, Rice University, Houston, Texas 77005, United States # Department of Material Science and Nanoengineering, Rice University, Houston, Texas 77005, United States ∇ Department of Electrical and Computer Engineering, Rice University, Houston, Texas 77005, United States S Supporting Information *

ABSTRACT: The challenge of controllable chemical synthesis of aluminum nanocrystals (Al NCs) has been met with only limited success. A major barrier is the absence of effective ligands to control the nucleation and growth of Al NCs. Here we demonstrate the size- and shape-controlled synthesis of monodisperse Al NCs using a polymer ligand, cumyl dithiobenzoate-terminated polystyrene (CDTB-PS). Density functional theory (DFT) calculations indicate that CDTBPS shows selective absorption on Al{100} facets, inducing the formation of nanocubes and trigonal bipyramids. An excess of CDTB-PS, however, decreases the supersaturation of Al atoms, leading to the formation of {111} facetterminated octahedral and triangular plates. The concentration of the catalyst, titanium (IV) isopropoxide, determines the size of Al NCs by controlling the number of seeds. Depending on nanoparticle size, the solutions of Al NCs possess distinct colors, a characteristic feature of plasmonic nanomaterials. This robust and controlled chemical synthesis of Al NCs lays a foundation for Al as a sustainable plasmonic material for current and future applications.



INTRODUCTION Aluminum, the most abundant metal on the earth’s surface, holds exceptional promise as a sustainable plasmonic material.1−3 In comparison with noble/coinage metal (Au and Ag) plasmonic nanomaterials, Al nanostructures possess a broader range of plasmon resonances from the ultraviolet (UV) to the near-infrared spectral range3−5 and, ultimately, greatly reduced relative cost. Al nanostructures have shown promising properties in many applications, such as surface-enhanced Raman spectroscopy,6,7 full-color displays,8,9 photocatalysis,10,11 and photocurrent enhancement in solar cells.12 Most of these Al nanostructures have been fabricated by top-down lithographic techniques; thus far, chemically synthesized Al nanoparticle progress has been greatly limited in comparison to this approach. In general, chemical synthesis of metallic nanocrystals (NCs) has developed into a mature chemistry over the past two decades, where impressive progress has been made toward the precise control of nanocrystal size, shape, and composition.13−22 In many nanoparticle syntheses, organic molecular ligands have been essential for size and shape control, typically © XXXX American Chemical Society

through facet-specific organic−inorganic interface absorption.23,24 For the chemical synthesis of Al NCs, however, significant efforts over the past two decades have yielded only limited control of size and shape.25−28 Lack of knowledge of ligand−Al interactions is one important factor limiting the development of Al NC synthesis with size and shape control. Unlike many other metal NCs that can be obtained by reduction of their metal ions, the high reduction potential of Al limits this synthetic strategy. It also limits exposure to H2O and O2 during synthesis, further restricting synthetic approaches. Al nanoparticle synthesis has been achieved by the catalytic or thermal decomposition of Al precursors, such as Al hydrides (AlH3)25−27 and alkyl aluminum compounds.28 These precursors react with many organic moieties, including proton-containing and carbonyl-containing functional groups.29 Many common ligands that have been successfully used for the synthesis of inorganic NCs are entirely incompatible with Al nanoparticle synthesis conditions. In Received: August 28, 2018

A

DOI: 10.1021/jacs.8b08937 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

Figure 1. (a) Reaction scheme for synthesis of Al nanocrystals. (b) SEM images of Al NCs with an average size of 224.2 ± 16.7 nm. Inset: higher magnification SEM image of the same Al NCs. (c−e) High-resolution transmission electron microscopy (HR-TEM) images of an individual Al NC. The images of parts d and e were taken from the center and edge of the Al NC (c), respectively. (f) Selected area electron diffraction (SAED) pattern of the Al NC, indicating a single crystal structure. (g) Molecular structure of CDTB-PS. (h) Powder X-ray diffraction (PXRD) pattern of the same batch sample, confirming the formation of pure fcc aluminum. (i) Initial states of the reaction solutions 3 min after being injected with different concentrations of Ti(i-PrO)4: 0.40, 0.35, 0.30, 0.25, and 0.20 mM (from left to right). (j) Final states of the solutions after being reacted for 4 h. (k−o) Precipitates of Al NCs in the final solutions after centrifugation (shown in the top-right corner) and corresponding representative TEM images of Al NCs with different sizes: 134.4 ± 9.2 (k), 172.8 ± 10.5 (l), 193.2 ± 10.0 (m), 224.2 ± 16.7 (n), and 250.3 ± 19.5 nm (o). Scale bars: 1 μm (b), 200 nm (the inset of part b), 50 nm (c), 2 nm (d and e), and 200 nm (k−o).



RESULTS AND DISCUSSION Al NCs were synthesized by Ti(i-PrO)4-catalyzed decomposition of 1-methylpyrrolidine alane (H3Al(1-MP), Figure S1) at 50 °C in tetrahydrofuran (THF) (Figure 1a) under a N2 atmosphere. H3Al(1-MP) is a single source precursor for Al metal through hydride oxidations. The formation of H2 from H3Al(1-MP) is catalyzed by Ti(i-PrO)4 which produces lowvalent Al intermediates and metallic Al atoms for the growth of Al NCs. In searching for appropriate polymer ligands for the synthesis of Al NCs, we found that, among various polymer ligands, cumyl dithiobenzoate-polystyrene (CDTB-PS) (with a number-average molecular weight of Mn = 4.5 kg/mol and a polydispersity index of PDI = 1.09, Figure S2) yielded colloidally stable Al NCs with high size and shape homogeneity, as shown in the scanning electron microscope (SEM) image (Figure 1b). In contrast, only irregularly shaped NCs were obtained when we used CDTB or CDTB-PS with short PS chain length (Mn = 1.1 kg/mol) as the surface ligand (Figure S3). CDTB reacted with the AlH3 precursor much faster than its polymer counterparts, i.e., CDTB-PS (Figure S4). Although the reaction rates of the AlH3 precursor with CDTB-PS with different Mn were comparable, only polymer ligands with relatively longer chains (i.e., Mn of 4.5 and 21.5

addition, to grow Al NCs, the binding between the capping ligand and Al surface should not be too strong to block the addition of new Al atoms. In contrast, capping ligands (e.g., carboxyl and phosphate functional groups) tightly binding on the Al or Al oxide surface may not be appropriate for the growth of Al NCs, especially to control their size and shape. The lack of molecular ligands to control the Al surface during nanoparticle growth has been a critical limitation and represents a key chemical challenge. Herein, we report the size- and shape-controlled synthesis of monodisperse Al NCs using a polymer ligand, cumyl dithiobenzoate-terminated polystyrene (CDTB-PS). Inspired by the fact that a functional group at the end of a polymer chain is much less reactive with AlH3 than that of a small molecule,30 we have investigated the use of end-functionalized polymer ligands for the synthesis of Al NCs. The polymer chains can work as stealth ligands, hindering the highly reactive AlH3 precursor from reacting with their end-functional groups, which subsequently control the growth of Al NCs during the catalytic decomposition of AlH3. B

DOI: 10.1021/jacs.8b08937 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

the reaction solution (3 min) became darker (Figure 1i and Figure S8), indicating the formation of more Al seeds. Consequently, at the final growth stage of the Al NCs (4 h), solutions with different seed concentrations showed various distinct colors (Figure 1j) as a result of the formation of Al NCs with sizes tuned from 134.4 ± 9.2 to 250.3 ± 19.5 nm (Figure 1k−q). A similar color evolution was also observed during growth (Figure S9). Notably, the colors turned more striking when the Al NCs in solution were centrifuged into precipitates (top-right corner in Figure 1k−q). The sizedependent color of Al NCs is a characteristic feature of plasmonic nanomaterials. The plasmonic properties of Al NCs will be discussed later. The UV−visible spectra for the samples in Figure 1j are provided in Figure 5a. It is well-known that the shape of an fcc nanocrystal is mainly determined by the ratio of areas between {100} and {111} facets31,32 and the internal structures of the seed,33 for example, if they are single crystal, singly twinned, or multiply twinned.34 The formation of the singly twinned structure is due to the introduction of stacking faults, which disrupts the normal atom layer stacking sequence of face centered cubic (fcc) structure from ABCABCABC to ABCABABCA (Figure 2

kg/mol) can effectively prevent the Al NCs from coalescing during their growth (Figure S3). In addition, if the CDTB-PS with Mn of 4.5 kg/mol completely reacted with AlH3 before the addition of Ti(i-PrO)4 catalyst, only irregularly shaped NCs were obtained (Figure S5). These control experiments suggested that both the end-group dithioester and long inert polymer chains are necessary for the controlled growth of Al NCs under these reaction conditions. High resolution transmission electron microscopy (HRTEM) images of Figure 1c−e show a representative Al NC which possesses a cuboctahedral shape (bounded by six {100} and eight {111} facets). The Al NC is surrounded by an amorphous ∼4 nm thick Al oxide layer. The formation of a passivation oxide layer upon exposure to air is important for their potential applications because it can effectively protect the inner core of Al NCs from further oxidization and allow for surface modification and functionalization. It should be also emphasized that no oxide layer was formed during the growth of Al NCs. The lattice fringe of 2.05 Å of the NC is close to the lattice spacing of the (200) plane at 2.04 Å for the facecentered cubic (fcc) lattice structure of Al (Figure 1d). The square symmetry of the electron diffraction pattern in Figure 1g indicates that the NC is single crystalline with an fcc lattice structure (Figure 1e). The powder X-ray diffraction (PXRD) pattern in Figure 1h further confirms the well-defined fcc structure of the Al NCs. No peak corresponding to crystalline Al2O3 is found, further indicating its amorphous structure of the surface passviate layer. Additionally, no Ti content in the Al NCs was detected by energy-dispersive X-ray spectrometer spectroscopy and inductively coupled plasma measurements. To achieve Al NCs with narrow size distributions, we applied a size-distribution focus strategy,23 which requires a sufficiently high monomer concentration through either raising the temperature or increasing the concentration of H3Al(1MP). For example, with an initial H3Al(1-MP) concentration of 50 mM, when the temperature was raised from 40 to 50 °C, the size of Al NCs (defined as the diameter of their twodimensional projection in TEM images) decreased from 320.2 ± 61.4 to 224.2 ± 16.7 nm with its standard deviation (SD) greatly reducing from 0.19 to 0.07 (Figure S6). Deceasing the initial H3Al(1-MP) concentration from 50 to 15 mM resulted in smaller Al NCs (183.7 ± 37.9 nm) with much broader size distributions (SD of 0.21) (Figure S7). Additionally, further increasing the initial H3Al(1-MP) concentration from 50 to 80 mM led to severe coalescence and aggregation of the produced Al NCs (Figure S7c). The concentration of the catalyst, Ti(i-PrO)4, was found to play an important role in the control of the Al NC dimensions. Ti(i-PrO)4-catalyzed decomposition of H3Al(1-MP) can be viewed as an analogue of enzyme−substrate reactions, for which the reaction rate (v) follows Michealis−Menten kinetics (see the discussion in the Supporting Information). Briefly, with a sufficiently high concentration of substrate (i.e., H3Al(1MP)), the formation rate of Al atoms (vAl) is equal to the product of the catalytic rate constant (kcat) and the initial concentration of the catalyst Ti(i-PrO)4. Therefore, tuning the initial concentration of Ti(i-PrO)4 allows for control of the formation rate of free Al atoms, which subsequently determines the number of Al seeds formed in the initial stage of the reaction. Consequently, the ratio of Al seeds to Al precursor controls the dimensions of the produced Al NCs. For example, when the initial catalyst concentration was increased from 0.20 to 0.40 mM, the early stage yellow color of

Figure 2. (a, b) Atom layer stacking sequence of cuboctahedron (CO) and anticuboctahedron (ACO), respectively. (c, d) Cartoon models of parts a and b, respectively. (e, f) SEM images of an individual Al NC with CO and ACO structures, respectively. The corresponding cartoon models are shown at the top-right corner. (g, h) TEM images of an individual Al nanocrystal with CO and ACO structure, respectively. The corresponding SAED patterns are shown in the top-right corners. When {111} facets of a cuboctahedral or anticuboctahedral Al nanocrystal are parallel to the substrate, the TEM image (i) (the SAED pattern is shown in the top-right corner) and SEM image (j) (the cartoon model is shown in the top-right corner) are obtained, but one cannot verify whether it is a CO or ACO from this view. Scale bars: 100 nm.

and Figure S10). In the present case, Al NCs with both single crystal and singly twinned structures coexisted in the system (Figure 1b and Figure 2) with a total proportion of 94%. Both cuboctahedral (single crystalline) and anticuboctahedral (singly twinned) Al NCs were observed in the same sample. We observed that the concentration of CDTB-PS determined the growth rates along the ⟨100⟩ and ⟨111⟩ directions, resulting in Al NCs with different shapes. With a relatively low CDTB-PS concentration of 0.30 mM, (100) facet-terminated Al NCs with a slightly truncated cube shape (single crystal) and right bipyramid shape (singly twinned crystal) were observed with a population fraction of 46 and 48%, respectively (Figure 3a). Increase of the CDTB-PS concentration resulted in the Al NCs terminated with higher {111}/{100} area ratios. As shown in Figure 3a−d, when the CDTB-PS concentration was increased from 0.30 to 0.70 mM, the shape of the resulting single crystal (singly twinned) NCs C

DOI: 10.1021/jacs.8b08937 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

Figure 3. TEM images (a−d) and corresponding PXRD patterns (e−h) of Al NCs prepared with a series of CDTB-PS concentrations of 0.30, 0.40, 0.50, and 0.70 mM, respectively. Shown in the top-right corners of the TEM images are the shapes of Al NCs with single crystal and singly twinned structures: (a) slightly truncated cubes and right bipyramids, (b) cuboctahedron and anticuboctahedron, (c) truncated octahedron and truncated triangular plate, and (d) octahedron and triangular plate. Green denotes Al(100) and yellow denotes Al(111). Scale bars: 500 nm (a−d).

Figure 4. DFT calculated adsorption energy of CDTB-PS on Al(100), Al(111), and Al(110) facets (from left to right).

evolved from cube (right bipyramid), to cuboctahedron (anticuboctahedron), to truncated octahedron (truncated triangular plate), and eventually to octahedron (triangular plate), of which the surface was primarily {111} terminated. The PXRD patterns of the corresponding samples confirmed that the {111}/{100} peak intensity ratio increased with increasing CDTB-PS concentration (Figure 3e−h). For lower concentrations of CDTB-PS, however, Al NCs with coalescence and aggregation (0.10 mM) or with larger sizes (over 300 nm) and poor homogeneity (0.15 mM) became the main product (Figure S11). During growth, the Al NCs generally change from small spherical NCs to large more faceted ones as the final product (Figure S9). Without changing the ligand concentration, we did not observe the shape change from {100} terminated cubes to {111} terminated octahedrons at different extents of growth. Therefore, the mechanism of NC shape control could differ from the depletion of capping ligand during growth which was very well studied previously by Xia’s group.32 We used density functional theory (DFT) calculations to understand the configuration and facet selectivity of dithioester (the end group of CDTB-PS) absorbed on different pure Al

facets without Al2O3 (as Al NCs were synthesized under oxygen and water-free conditions) (see more details in the Supporting Information). DFT calculations reveal that the most probable configuration for the CDTB-PS molecules absorbed on Al surfaces is the lying-down geometry with the S atom of the CS functional group occupying the hollow sites on Al facets (Figure 4 and Figures S12 and S13). Meanwhile, the Al atoms bonded to the adsorbed S atom protrude only slightly from the surface atomic layer, implying the strong chemisorption of CDTB-PS molecules on the Al surface. Remarkably, the adsorption energy of CDTB-PS molecules on Al(100) (ΔEads/Al(100) = −115.959 kcal/mol) is much lower than that for the Al(111) and Al(110) facets (ΔEads/Al(111) = −15.260 kcal/mol and ΔEads/Al(110) = −12.416 kcal/mol) (Tables S2 and S3). This suggests that CDTB-PS molecules possess unique crystallographic selectivity for Al{100} facets, hindering the deposition of Al atoms and eventually leading to the formation of {100} facet-terminated NCs such as nanocubes and right bipyramids. The DFT calculations also reveal that the breaking of the C−S bond of CDTB-PS needs to overcome a high energy barrier, 12.27 kcal/mol, although the resulting molecular fragments exhibit a stronger adsorption D

DOI: 10.1021/jacs.8b08937 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

dark-field scattering, combined with theoretical calculations (Figure 5). Reflectance measurements of the Al NC solutions

on Al{100}. Nevertheless, the experimental temperature of our reaction system is 323 K and the corresponding energy provided is U = kBT = 0.64 kcal/mol, which is too low to break the CS bond. Thus, the CS bond of CDTB-PS adsorbed or reacted on Al{100} could not be broken under the current experimental conditions (Figure S14). Experimentally, we observed that Al cubes and right bipyramids with {100} facets were obtained at relatively low CDTB-PS concentrations (0.30 mM), while Al octahedra and triangular plates with {111} facets were achieved at high CDTB-PS concentrations. Similar results have been reported in the synthesis of Au nanocubes with cetyltrimethylammonium bromide (CTAB) as a surface ligand that has a strong selective affinity to Au{100} facets.35 According to the Thomson−Gibbs equation, a higher concentration of stabilizing ligands (CDTB-PS in the present case) leads to smaller differences between the chemical potential of metal atoms in solution and solid crystals, resulting in the formation of crystals with lower surface energies such as the {111} facet-terminated octahedra and triangular plates in the present scenario (see discussion in the Supporting Information and Figure S15). X-ray photoelectron spectroscopy (XPS) and Fouriertransform infrared spectroscopy (FT-IR) studies show agreement with the DFT calculations for the configuration of CDTB-PS absorbed on Al surfaces. XPS analysis of Al NCs with CDTB-PS surface ligands reveal that the deconvolution of the XPS spectrum of the Al 2p orbital (Figure S15a and b) exhibits three peaks, including metallic Al at 72.00 eV, oxidized Al at 74.40 eV, and Al at 74.88 eV, the latter indicating the existence of AlS bonds (Table S4). The formation of the Al oxide passivation layer was unavoidable during the preparation of XPS measurement. The XPS analysis reveals the ratio of Al3+ to Al0 to be about 4:1 for the surface of the Al NCs upon exposure to air. Considering that the detection depth of XPS spectroscopy is about 5 nm, the XPS result indicates the thickness of the oxide shell is about 4 nm, which is consistent with the TEM result (Figure 1e). Compared to isolated CDTB-PS, the binding energy of S 2p in the CS bond of CDTB-PS absorbed on the Al NC surface was reduced by 0.30 eV (Figure S16c and d). This indicates the S atom in the CS bond possesses a higher electron density after absorbing onto the Al surface, implying a strong electronic interaction between the Al NC surface and the CDTB-PS ligand. In contrast, the shifting of the binding energy of S 2p in the CSC bond was negligible (Figure S16d and Table S4). FT-IR analysis further confirmed that absorption could mainly be attributed to the interaction of the S in CS with the Al surface (Figure S17). Comparing with the spectra of CDTB-PS before and after absorption, the peaks of CS (1045 cm−1) and CS (870 cm−1) were red shifted to 972 and 840 cm−1, respectively. These results are in good agreement with the DFT-calculated peak position of the CS stretching frequency before and after absorption (from 1071 to 970 cm−1) (Table S5). These obvious shifts of the peaks imply that the absorption of dithioester on Al NCs is mainly attributed to the interaction of the S in CS with the Al surface. Besides the CS vibrational feature, the peaks of aromatic CH bending (in-plane) at 1069 and 1028 cm−1 were slightly red shifted to 1065 and 1021 cm−1, respectively, and meanwhile became weaker and broader, which is due to the change of the chemical environment of the benzene ring adjacent to the CS group. The optical properties of the Al particles were characterized with ensemble reflectance measurements and single particle

Figure 5. (a) Experimental reflectance spectra of NC solutions with average sizes of 130 nm (gray), 170 nm (orange), 190 nm (purple), 220 nm (dark blue), and 250 nm (green). (b) Scattering spectra calculated using FDTD, corresponding to the experimental spectra. (c) The corresponding NC solutions for parts a and b with distinct colors. (d) Experimental single particle dark-field scattering spectra of cuboctahedral Al NCs. (e) Calculated single particle scattering spectra. (f) SEM images of the particles measured in part e with edge lengths of 122 nm (green), 160 nm (red), 207 nm (blue), and 240 nm (purple). Calculated S-polarized scattering response (g) and charge density plots (h) of the first three plasmonic modes for the 240 nm particle. Calculated S-polarized scattering response (i) and charge density plots (j) of the first three plasmonic modes for the 122 nm particle. Scale bars: 100 nm. Note: The energy of the octupolar mode in the small particle is slightly greater than 5.5 eV.

(Figure 5a) show clear peaks which red-shift with increasing size, and the peak positions are in agreement with theoretical scattering spectra based on the emsemble average of particle size and shapes (Figure 5b). The theoretical spectra are calculated using the finite difference time domain (FDTD) method (Lumerical), with orientational averaging used to reproduce an ensemble spectrum (Figures S18 and S19). The peaks in the reflectance spectra correspond to the colors observed in the particle solutions (Figure 5c). Single particle dark-field scattering measurements were obtained for cuboctahedra of various sizes (Figure 5d) using a custom-built UV−vis dark-field microspectrometer: these spectra are compared with finite element method (FEM) calculations (Figure 5e) using the geometry of the measured particles (Figure 5f). Calculations show that the optical spectrum is sensitive to the degree of corner truncation (Figure S20). This observed sensitivity is likely due, at least partially, to the decrease in E

DOI: 10.1021/jacs.8b08937 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society volume with increased truncation for a given particle edge length. The ensemble measurements are reasonably similar to the single particle measurements of similar sizes, particularly for particles below 200 nm, indicating a high degree of homogeneity of the synthesized particle solutions. Discrepancies between the single particle and ensemble measurements are due to the additional shapes present in the ensemble solutions and also to orientational effects in single particle measurements that may occur for particles in this size range. The calculated scattering under S-polarized light is used to perform mode analysis and identify the spectral features observed for the 240 nm (Figure 5g,h) and 122 nm (Figure 5i,j) particles. The S-polarized spectra show a clearer mode separation than the P-polarized spectra, an observation that agree with experimental measurements (Figure S21). Additionally, the theoretical S-polarized spectra give rise to a stronger scattering response. The spectral features observed in the visible region of the spectrum are primarily quadrupolar and higher order modes. The dipole peaks of all but the smallest particle solutions are either on the red end of the visible spectrum or in the near-infrared. In contrast, for the smallest particles, only the dipole and quadrupole resonances are observed. These results indicate that the observed colors of the Al NCs are primarily produced by higher order plasmon modes.

Naomi J. Halas: 0000-0002-8461-8494 Kun Liu: 0000-0003-2940-9814 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work is supported by the National Natural Science Foundation of China (21534004, 21474040, 21674042). The Rice part of the project was supported by the Robert A. Welch Foundation under grants C-1220 (N.J.H.) and C-1222 (P.N.).



(1) Powell, C. J.; Swan, J. B. Phys. Rev. 1959, 115, 869. (2) West, P. R.; Ishii, S.; Naik, G. V.; Emani, N. K.; Shalaev, V. M.; Boltasseva, A. Laser Photonics Rev. 2010, 4, 795. (3) Knight, M. W.; King, N. S.; Liu, L.; Everitt, H. O.; Nordlander, P.; Halas, N. J. ACS Nano 2014, 8, 834. (4) Mcmahon, J. M.; Schatz, G. C.; Gray, S. K. Phys. Chem. Chem. Phys. 2013, 15, 5415. (5) Lecarme, O.; Sun, Q.; Ueon, K.; Misawa, H. ACS Photonics 2014, 1, 538. (6) Sharma, B.; Cardinal, M. F.; Ross, M. B.; Zrimsek, A. B.; Bykov, S. V.; Punihaole, D.; Asher, S. A.; Schatz, G. C.; Van Duyne, R. P. Nano Lett. 2016, 16, 7968. (7) Tian, S.; Neumann, O.; McClain, M. J.; Yang, X.; Zhou, L. N.; Zhang, C.; Nordlander, P.; Halas, N. Nano Lett. 2017, 17, 5071. (8) Kristensen, A.; Yang, J. K. W.; Bozhevolnyi, S. I.; Link, S.; Nordlander, P.; Halas, N. J.; Mortensen, N. A. Nat. Rev. Mater. 2017, 2, 16088. (9) Zhu, X.; Vannahme, C.; Højlund-Nielsen, E.; Mortensen, N. A.; Kristensen, A. Nat. Nanotechnol. 2016, 11, 325. (10) Zhou, L. N.; Zhang, C.; McClain, M. J.; Manjavacas, A.; Krauter, C. M.; Tian, S.; Berg, F.; Everitt, H. O.; Carter, E. A.; Nordlander, P.; Halas, N. J. Nano Lett. 2016, 16, 1478. (11) Swearer, D. F.; Zhao, H.; Zhou, L. N.; Zhang, C.; Robatjazi, H.; Martirez, J. M. P.; Krauter, C. M.; Yazdi, S.; McClain, M. J.; Ringe, E.; Carter, E. A.; Nordlander, P.; Halas, N. J. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 8916. (12) Hylton, N. P.; Li, X. F.; Giannini, V.; Lee, K.-L.; Ekins-Daukes, N. J.; Loo, J.; Vercruysse, D.; Dorpe, P. V.; Sodabanlu, H.; Sugiyama, M.; Maier, S. A. Sci. Rep. 2013, 3, 2874. (13) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Annu. Rev. Mater. Sci. 2000, 30, 546. (14) Sun, Y. G.; Xia, Y. N. Science 2002, 298, 2176. (15) Manna, L.; Milliron, D. J.; Meisel, A.; Scher, E. C.; Alivisatos, A. P. Nat. Mater. 2003, 2, 382. (16) Park, J.; An, K.; Hwang, Y.; Park, J.-G.; Noh, H.-J.; Kim, J.-Y.; Park, J.-H.; Hwang, N.-M.; Hyeon, T. Nat. Mater. 2004, 3, 891. (17) Habas, S. E.; Lee, H.; Radmilovic, V.; Somorjai, G. A.; Yang, P. D. Nat. Mater. 2007, 6, 692. (18) Grzelczak, M.; Pérez-Juste, J.; Mulvaney, P.; Liz-Marzán, L. M. Chem. Soc. Rev. 2008, 37, 1783. (19) Xia, Y. N.; Xiong, Y. J.; Lim, B.; Skrabalak, S. E. Angew. Chem., Int. Ed. 2009, 48, 60. (20) Langille, M. R.; Personick, M. L.; Zhang, J.; Mirkin, C. A. J. Am. Chem. Soc. 2012, 134, 14542. (21) Chen, H.; Shao, L.; Li, Q.; Wang, J. Chem. Soc. Rev. 2013, 42, 2679. (22) Wang, Y.; He, J.; Liu, C.; Chong, W. H.; Chen, H. Angew. Chem., Int. Ed. 2015, 54, 2022. (23) Yin, Y.; Alivisatos, A. P. Nature 2005, 437, 664. (24) Boles, M. A.; Ling, D.; Hyeon, T.; Talapin, D. V. Nat. Mater. 2016, 15, 141. (25) Haber, J. H.; Buhro, W. E. J. Am. Chem. Soc. 1998, 120, 10847. (26) Meziani, M. J.; Bunker, C. E.; Lu, F.; Li, H.; Wang, W.; Guliants, E. A.; Quinn, R. A.; Sun, Y.-P. ACS Appl. Mater. Interfaces 2009, 1, 703.



CONCLUSION Although the fundamental basis of shape-controlled synthesis of Al NCs has yet to be fully understood, the steric hindrance effect of long polymer chains and the selective adsorption of their dithioester end-groups on the Al{100} facet play a major role in determining the nucleation and growth of Al NCs, as well as their morphologies. This work makes it clear that chemical synthesis of Al NCs with well-controlled shapes, sizes, and structures is practically feasible. The realization of the synthesis of colorful monodisperse Al NCs with low cost significantly broadens the scope of large scale applications for plasmonic NCs, such as plasmonic enhanced solar cells,12 plasmonic color displays,36 plasmonic holograms,37 surface enhanced Raman scattering,6,7 plasmonic enhanced chiral catalysts and sensing,38 as well as security tagging and cryptography.39 The synthesis of Al NCs with well-defined crystallographic facets and sharp corners and edges also paves the path to study their roles for photocatalysis.11



ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.8b08937. Additional information, figures, tables, and equations related to the materials, synthesis of Al NCs, and characterization (PDF)



REFERENCES

AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] *[email protected] *[email protected] ORCID

Zhongyuan Lu: 0000-0001-7884-0091 Peter Nordlander: 0000-0002-1633-2937 F

DOI: 10.1021/jacs.8b08937 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society (27) McClain, M. J.; Schlather, A. E.; Ringe, E.; King, N. S.; Liu, L.; Manjavacas, A.; Knight, M. W.; Kumar, I.; Kenton, H. W.; Everitt, H. O.; Nordlander, P.; Halas, N. J. Nano Lett. 2015, 15, 2751. (28) Clark, B. D.; Jacobson, C. R.; Lou, M.; Yang, J.; Zhou, L. N.; Gottheim, S.; Desantis, C. J.; Nordlander, P.; Halas, N. J. Nano Lett. 2018, 18, 1234. (29) Brown, H. C.; Yoon, N. M. J. Am. Chem. Soc. 1966, 88, 1464. (30) Odian, G. Principles of Polymerization, 4th ed.; John Wiley & Sons: Hoboken, NJ, 2004; pp 42, 729. (31) Wang, Z. L. J. Phys. Chem. B 2000, 104, 1153. (32) Xia, X.; Zeng, J.; Oetjen, L. K.; Li, Q.; Xia, Y. J. Am. Chem. Soc. 2012, 134, 1793. (33) O’Brien, M. N.; Jones, M. R.; Brown, K. A.; Mirkin, C. A. J. Am. Chem. Soc. 2014, 136, 7603. (34) Xia, Y. N.; Xia, X. H.; Peng, H. C. J. Am. Chem. Soc. 2015, 137, 7947. (35) Lin, H. X.; Lei, Z. C.; Jiang, Z. Y.; Hou, C. P.; Liu, D. Y.; Xu, M. M.; Tian, Z. Q.; Xie, Z. X. J. Am. Chem. Soc. 2013, 135, 9311. (36) Duan, X. Y.; Kamin, S.; Liu, N. Nat. Commun. 2017, 8, 14606. (37) Montelongo, Y.; Tenorio-Pearl, J. O.; Williams, C.; Zhang, S.; Milne, W. I.; Wilkinson, T. D. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 12679. (38) Ma, W.; Xu, L. G.; de Moura, A. F.; Wu, X. L.; Kuang, H.; Xu, C. L.; Kotov, N. A. Chem. Rev. 2017, 117, 8041. (39) Cui, Y.; Hegde, R. S.; Phang, I. Y.; Lee, H. K.; Ling, X. Y. Nanoscale 2014, 6, 282.

G

DOI: 10.1021/jacs.8b08937 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX