Ligand-Dependent Colloidal Stability Controls the Growth of

Subscriber access provided by Iowa State University | Library

Article

Ligand-Dependent Colloidal Stability Controls the Growth of Aluminum Nanocrystals Benjamin D. Clark, Christopher J. DeSantis, Gang Wu, David Renard, Michael J. McClain, Luca Bursi, Ah-Lim Tsai, Peter Nordlander, and Naomi J. Halas J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b12255 • Publication Date (Web): 06 Jan 2019 Downloaded from http://pubs.acs.org on January 6, 2019

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.


Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,



the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.


Page 1 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Journal of the American Chemical Society

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 24

Ligand-Dependent Colloidal Stability Controls the Growth of Aluminum Nanocrystals Benjamin D. Clark,1,4 Christopher J. DeSantis,2,4 Gang Wu,5 David Renard,1,4 Michael J. McClain,1,4 Luca Bursi,3,4 Ah-Lim Tsai,5 Peter Nordlander,3,4 and Naomi J. Halas*,1,2,4 1

Department of Chemistry, 2Department of Electrical & Computer Engineering, 3Department of Physics & Astronomy, 4Laboratory for Nanophotonics, Rice University, 6100 Main Street, Houston, Texas 77005, United States 5 Division of Hematology, Department of Internal Medicine, The University of Texas McGovern Medical School, 6431 Fannin St, Houston, Texas 77030, United States

Abstract: The precise size- and shape-controlled synthesis of monodisperse Al nanocrystals remains an open challenge, limiting their utility for numerous applications that would take advantage of their size and shape-dependent optical properties. Here we pursue a molecular-level understanding of the formation of Al nanocrystals by titanium(IV) isopropoxide-catalyzed decomposition of AlH3 in Lewis base solvents. As determined by electron paramagnetic resonance spectroscopy of intermediates, the reaction begins with the formation of Ti3+-AlH3 complexes. Proton nuclear magnetic resonance spectroscopy indicates isopropoxy ligands are removed from Ti by Al, producing aluminum(III) isopropoxide and low-valent Ti3+ catalysts. These Ti3+ species catalyze elimination of H2 from AlH3 inducing the polymerization of AlH3 into colloidally unstable lowvalent aluminum hydride clusters. These clusters coalesce and grow while expelling H2 to form colloidally stable Al nanocrystals. The colloidal stability of the Al nanocrystals and their size are determined by the molecular structure and density of coordinating atoms in the reaction, which is controlled by choice of solvent composition.


1

Page 3 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction: Colloidally synthesized Al nanocrystals (NCs) have become promising alternatives to noble metal NCs for plasmonics as demonstrated by applications ranging from spectroscopic detection of unmodified DNA to plasmon-enhanced photocatalysis.1–3 Al NCs support localized surface plasmon resonances, or collective oscillations of conduction band electrons, that can be tuned from the ultraviolet (UV) to the infrared by controlling the size and shape of Al NCs. 4–6 To prevent oxidation of the aluminum(III) hydride (AlH3) precursors and metallic Al NCs, the synthesis of Al NCs requires air- and water-free conditions.4,7 The synthesis of Al NCs is distinct from those of noble metal NCs because a transition metal catalyst, typically titanium(IV) isopropoxide (Ti(OiPr)4), is generally required, making AlH3 decomposition analogous to polymerization of olefins using Ziegler-Natta catalysts. The challenges of studying this rigorously air- and water-free system provide new opportunities to broaden our understanding of shape- and size-controlled synthesis of main group, transition metal and alloyed NCs. In general, a variety of aspects, such as seed defects, supersaturation, ligands that bind preferentially to particular facets, and the colloidal stability of metal atoms, clusters and NCs, determine the size and shape of colloidally synthesized NCs.8–12 Recently, cumyl dithiobenzoate-terminated polystyrene ligands were used to control the shape of colloidally synthesized Al NCs by supersaturation and selectively binding to {100} facets of Al, though the underlying growth mechanism of Al NCs was not described in detail.13 Our understanding of the parameters that control NC growth can be advanced by the use of analytical techniques such as microscopy and spectroscopy, along with the theoretical insight provided by the analysis of spectroscopic signatures.14–16 These approaches also allow us to establish an understanding of the formation of complexes between precursors and ligand molecules that are responsible for determining the final particle morphology.17–21


2


Page 4 of 24

In this study, we use 1H nuclear magnetic resonance (NMR) and electron paramagnetic resonance (EPR) spectroscopies to probe reaction intermediates and byproducts produced during Ti(OiPr)4-catalyzed decomposition of AlH3 in Lewis base solvents. EPR spectroscopy is routinely used to provide structural information about organic radicals and transition metal complexes with an unpaired electron.22,23 The molecular structures of AlH3 radical anions, radicals with Al-Al bonds, and bimetallic titanium(III) aluminum hydrides have been investigated by EPR spectroscopy, providing a framework for these results.24–30 A primary characteristic of radicals is their g-factor, a measure of the magnetic moment of an unpaired electron in an applied magnetic field. The g-factor of main group radicals is close to the value for a free electron (g = 2.0023) but deviates for transition metal complexes due to spin-orbit coupling, serving as a spectroscopic signature for identification.24–31 Hyperfine splitting may be observed in EPR spectra that arises from coupling between unpaired electrons and nuclei, which provides information about the molecular structure of radicals.22–31 As most molecules with an unpaired electron are EPR active, the radical concentration can be determined by double integration of the EPR spectrum against a spin standard. In our case, deconvoluting the EPR absorption spectra based on the g-factor of each intermediate enabled determination of their contributions to the EPR signal during Al NC growth. Using this data we describe a molecular-level view of the growth of Al NCs from AlH3 and Ti(OiPr)4 that is consistent with ensemble growth kinetics and control of Al NC size by choice of solvent composition. Results and Discussion: Molecules of AlH3 are Lewis acids that form adducts with one or two aprotic Lewis bases. In these adducts, Al adopts pseudo-tetrahedral or pseudo-trigonal pyramidal molecular geometries (Fig. S1).32–35 With Al NCs synthesized using the Ti(OiPr)4-catalyzed approach (Fig. 1A), the


3

Page 5 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


formation of adducts between Lewis bases (Fig. 1B) and AlH3 from dimethylethylamine alane (DMEAA) influences the final size of the Al NCs (Fig. 1).4 Through transmission electron microscope (TEM) imaging and analysis (Fig. 1C), the particles synthesized in 1,4-dioxane were found to have an average size of 62 ± 11 nm (number measured (N) = 252), while the average size of the particles from tetrahydrofuran (THF) was 157 ± 36 nm (N = 255). This trend in size control was also observed when only a stoichiometric amount of tertiary amine ligand, either tetramethylethylenediamine (TMEDA) or N-methylpyrrolidone (NMP), in toluene was used as the reaction solvent. The Al NCs from the reaction with TMEDA were smaller than those with NMP, with average sizes of 101 ± 22 nm (N = 215) and 134 ± 60 nm (N = 228). The histograms associated with the size measurements for each reaction solvent are shown in Fig. S2. Despite control over NC size by choice of solvent composition, the resulting Al NCs exhibit different shapes, dominated by truncated octahedra but also including singly and multiply twinned particles (Fig. 1C).

Figure 1. A. Reaction scheme for the colloidal synthesis of Al NCs by the decomposition of DMEAA of with Ti(OiPr)4. B. Molecular structure of the Lewis base ligands used in this study:


4


Page 6 of 24

1,4-dioxane, THF, TMEDA, and NMP. C. Representative TEM images of Al NCs synthesized in dioxane (top left) and THF (top right), and with stoichiometric amounts of TMEDA (bottom left) and NMP (bottom right) at 40 C. The scale bars are 100 nm. D. 1H NMR spectrum of reagents and byproducts during the synthesis of Al NCs in a mixture of deuterated dioxane and THF. In experimental and theoretical investigations of the dissociative adsorption of H2 by metallic Al surfaces, which is conceptually the growth of Al NCs from AlH3 reversed, incorporation of Ti into Al surfaces catalyzes H2 splitting.36–40 These studies imply that Al NCs synthesized with Ti(OiPr)4 could be doped with metallic Ti atoms. In this situation, all four isopropoxy ligands would be removed from Ti(OiPr)4 by AlH3 to generate metallic Ti. However, energy-dispersive X-ray spectroscopy of an Al NC synthesized in THF found no Ti within the Al NC (Fig. S3). Examination of dried powder samples using X-ray powder diffraction confirmed the face-centered cubic lattice of metallic Al, while thermogravimetric analysis under O2 indicated that the percentage of metallic Al was ~89% using THF and ~75% with dioxane (Fig. S4). Analysis of the reaction intermediates by 1H NMR spectroscopy (Fig. 1D) revealed a small amount of Al(OiPr)3 was produced relative to the initial amount of Ti(OiPr)4.41 Notably, the hydrides of the AlH3 precursor had a broad 1H NMR signal at 3.05 ppm that decreased after the addition of Ti(OiPr)4, while the signal from dissolved H2 at 4.55 ppm increased (Fig. 1D). Detection of H2 during the formation of Al NCs verified AlH3 was reduced to metallic Al by Ti3+-catalyzed hydride oxidation.42 This data, coupled with the superb plasmonic properties of individual Al NCs, provides strong evidence that Ti is not doped into the lattice of Al NCs.4,5 We hypothesize that reduced Ti3+(OiPr)3 coordinates with AlH3 and catalyzes elimination of H2, which polymerizes AlH3 into low-valent aluminum hydride clusters that coalesce and grow into Al NCs.


5

Page 7 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. EPR spectroscopy of Al NCs synthesized in ethers. A. Experimental EPR spectra of Ti(OiPr)4 with increasing equivalents of AlH3 in dioxane. B. EPR spectra of Ti(OiPr)4 with increasing equivalents of AlH3 in THF. C. Simulations of the EPR spectrum as a function of the g-factor for a 1:1 ratio of AlH3 to Ti(OiPr)4 in dioxane (top) and THF (bottom). D. Proposed structures of the Ti3+(OiPr)4-AlH3 complexes stabilized by dioxane (top) and THF (bottom). The nuclei used to produce the simulated spectra are highlighted with colored spheres while the hyperfine splitting values used are presented in Table S1. Mixtures of Ti3+ and Al2+ complexes were responsible for the signal while reduced Ti3+-Al2+ complexes that had lost a molecule of H2 were EPR silent due to spin-exchange coupling. Initially, Ti(OiPr)4 was EPR inactive with Ti in the +4 oxidation state. After a stoichiometric amount of AlH3 was added, the formation of Ti3+(OiPr)4-AlH3 complexes was confirmed by EPR spectroscopy in dioxane (Fig. 2A) and THF (Fig. 2B). Simulations of the EPR spectra of the 1:1 ratio of Al to Ti (Fig. 2C), indicated that Ti3+(OiPr)4-Al3+H3 complexes having a g-factor of 1.952 constituted 90% of the radicals in dioxane and 85 % in THF (Table S1). The remaining fraction of the signals was from Ti4+(OiPr)4-Al2+H3 complexes with a g-factor of 1.966. The prominent nuclear hyperfine splitting (I = 5/2) of Al2+, in contrast to the lack of nuclear hyperfine of Ti3+, supports the presence of Al2+ among these intermediates (Fig. 2C). The


6


Page 8 of 24

molecular structures of the Ti(OiPr)4-AlH3 complexes in dioxane and THF are proposed to contain five-coordinated Al atoms (Fig. 2D) based on the simulated EPR spectra, prior EPR investigations, and structural analysis of bimetallic Ti3+2H2Al complexes.27–30,42,43 The Ti and Al atoms are bridged by two hydrides and an isopropoxy ligand based on the simulated EPR spectra (Fig. 2C) and the 1H NMR spectra of the reaction intermediates (Fig. S5A). However, the exact number of bridging and terminal hydrides cannot be determined by EPR because simulations with fewer hydrides (e.g. from loss of H2) match the experimental spectra. Consistent with other reports, the hydrides were not observed at room temperature by 1H NMR.42 In contrast to sterically stabilized Ti3+-AlH3 compounds with pseudo-tetrahedral molecular geometry around Al, a bridging isopropoxy ligand between Al and Ti increases the valency of Al, pushing the hydrides together and predisposing the complex to loss of H2.42,43 Without the addition of further AlH3, this complex is relatively stable, but reacts with AlH3 to facilitate the polymerization of Al atoms into Al NCs (Scheme 1). EPR analysis of the formation of Al NCs was accomplished by sequentially adding aliquots of DMEAA to the Ti(OiPr)4-AlH3 complex to produce samples with Al:Ti ratios of 2:1, 3:1, 4:1, 5:1 and 10:1. These spectra are snapshots of the paramagnetic intermediates present during Al NC growth. In both solvents, the EPR signal shifted from a narrow signal at ~3400 G (g = 1.952) to a broad feature at ~3330 G (g = 1.986) as the ratio of Al to Ti was increased, which allowed the reactions to proceed to completion. The formation of Ti3+(OiPr)3 is consistent with the appearance of the signal at ~3410 G (g = 1.947) during the later stages of the reactions with higher ratios of Al to Ti.


7

Page 9 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Scheme 1. A. Formation of Ti(OiPr)4-AlH3 complex followed by reductive elimination of H2 and abstraction of an isopropoxy ligand by Al, produces Ti3+(OiPr)3, the active catalyst for the growth of Al NCs. B. Representative catalytic cycle of the Ti3+(OiPr)3 with two molecules of AlH3 to produce metallic Al2 and three H2 molecules. The changes in the EPR spectra during the reaction and the observation of H 2 by 1H NMR indicate that both diamagnetic Ti4+ and Al3+ were reduced to Ti3+ and Al2+ by oxidation of two hydride ligands to form one molecule of H2. The proposed mechanism for the generation of the active Ti3+ catalyst (Scheme 1A) begins with the reaction of AlH3 with Ti(OiPr)4 (I) to form reduced Ti(OiPr)4-AlH3 complexes (II). After reductive elimination of H2 to form III, Al abstracts an isopropoxy ligand, generating the active Ti3+ catalyst (IV) for the polymerization of AlH3 and elimination of H2 to form metallic Al clusters (Scheme 1B). We propose that Ti3+(OiPr)3 (IV) reacts with AlH3 to form V, which releases H2 to form VI. Insertion of additional AlH3 releases H2 and generates low-valent titanium-aluminum-hydride (Ti-Al-H) clusters like VII that can release H2 and metallic Al while regenerating IV. These steps are consistent with the analogy to enzyme kinetics where Ti3+(OiPr)3 serves as the enzyme and AlH3 are the substrates that bind to form enzyme-substrate complexes (V-VII) that eventually dissociate into metallic Al clusters and


8


Page 10 of 24

H2 while regenerating Ti3+(OiPr)3.13 This catalytic cycle controls the rate of Al seeds formation and then likely occurs on the surface of growing Al NCs as the concentration of AlH3 is depleted. Variable temperature EPR spectroscopy down to ~125K of samples with g-factors of 1.952 (Ti3+(OiPr)4-AlH3 complex) and 1.986 (signal during Al NC growth) indicated the magnetic susceptibilities of each sample was inversely proportional to temperature (Fig. S6). For the Ti3+(OiPr)4-AlH3 complex, the signal doubled when cooled, while the EPR signal from Al NC growth quadrupled when cooled. We assign the broad signal at ~3330 G (g = 1.986) to the existence of low-valent aluminum hydride clusters with Ti3+(OiPr)3 catalysts attached to their surface through bridging hydrides and isopropoxy ligands. Though the exact molecular structure of these Ti-Al-H clusters is presently unknown, two potential structures contribute to their paramagnetic behavior: 1. Ti3+ species on the cluster surface, and 44 2. non-metallic clusters with an odd number of electrons.45,46 We rule out the possibility of conduction electron spin resonance (CESR) from metallic clusters contributing significantly to the observed spectra because of the temperature dependence of the signals (Fig. S6) and because CESR signals from Al metal have a g-factor of 1.996.47–49 Similarly, Al radicals have g-factors above 2.00, while bimetallic (C5H5)2Ti3+-2H2-AlH2 has a lower g-factor of ~1.990 from spin orbit coupling to the d-orbitals of Ti.25–29 Hence, based on these g-factors, the signals with g = 1.986 can only arise from Al clusters with Ti3+ catalysts on their surface or an odd number of electrons.44–49 During the later stages of the reaction in dioxane (higher ratios of Al to Ti), small hyperfine splitting of 3.0 – 3.4 G was observed that was absent from the corresponding spectra in THF. Because 14N has a nuclear spin of 1, reactions were performed using TMEDA and NMP (Fig. S7), bidentate and monodentate tertiary amine analogues of dioxane and THF, to investigate if this hyperfine splitting was from


9

Page 11 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


bidentate coordination of Ti-Al-H clusters. The EPR spectra (Fig. S7) confirm that bidentate coordination of Ti-Al-H clusters produces hyperfine splitting during Al NC growth.

Figure 3. Analysis of integrated EPR spectra from the reactions in dioxane and THF. A. and B. Integrated EPR spectra for increasing ratios of Al to Ti as a function of g-factor for reactions in dioxane (left) and THF (right). C. and D. Relative abundance of the four paramagnetic species, Ti3+(OiPr)4-Al3+H3, Ti4+(OiPr)4-Al2+H3 intermediates, Ti-Al-H clusters, and Ti3+(OiPr)3 as a function of Al:Ti ratio. E. and F. g-factors for each of the components identified by deconvolution of the integrated EPR spectra as a function of the ratio of Al to Ti in dioxane and THF. The EPR data was further analyzed by considering the integrated EPR signal intensity as a function of g-factor to determine the contributions from Ti(OiPr)4-AlH3 complexes and Ti-Al-H


10


Page 12 of 24

clusters for reactions in dioxane (Fig. 3A) and THF (Fig. 3B). To evaluate the relative abundance of each intermediate during the reaction, the absorption EPR spectra were deconvoluted using Lorentzian or Gaussian functions centered on their respective g-factors (Fig. S8). Plotting the area of the deconvoluted EPR absorption spectra for each species as a function of the molar ratio of Al to Ti for reactions in dioxane (Fig. 3C) and THF (Fig. 3D) enabled determination of the relative abundance of these intermediates during each EPR snapshot of the formation of Al NCs. As with the initial analysis of the EPR spectra (Fig. 2), four paramagnetic species were identified based on their g-factors, which were the same in dioxane (Fig. 3E) and THF (Fig. 3F). The four radical intermediates detected were assigned as follows: 1. g = 1.956, Ti3+(OiPr)4-Al3+H3 complexes, 2. g = 1.966, Ti4+(OiPr)4-Al2+H3 complexes, 3. g = 1.986, Ti-Al-H clusters with odd number of electrons, and 4. g = 1.947, solvated Ti3+(OiPr)3 complexes. Note that Ti3+(OiPr)3, the active catalyst for the production of metallic Al, appeared after the Ti(OiPr)4-AlH3 complexes, consistent with the abstraction of an isopropoxy ligand from Ti(OiPr)4 to generate Ti3+(OiPr)3 (Scheme 1A). Comparison of the relative abundance of Ti-Al-H clusters as a function of the ratio of Al to Ti in dioxane and THF indicates that the clusters form more rapidly in dioxane than in THF. This data suggests that the colloidal stability of Ti-Al-H clusters is higher in dioxane than in THF, resulting in the formation of more Al seeds in dioxane compared to THF, which in turn dictates the final size of the Al NCs.


11

Page 13 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Kinetics of Al NC formation in ethers. A. UV-visible spectra acquired every five minutes during the synthesis of Al NCs in dioxane. B. Value of the extinction at 300 nm plotted as a function of time for reactions in dioxane (blue), THF (red), and a 50:50 mix (purple) of THF and dioxane. C. Representative TEM images of the Al NCs produced by the reaction in pure THF (left), pure dioxane (right) and a 50:50 mixture of THF and dioxane (center). The scale bars are all equal to 100 nm.

To further investigate the theory that colloidal stability of Ti-Al-H clusters in the solvent controls the growth of Al NCs, UV-visible absorption spectroscopy was used to compare the relative rates of Al NC formation in dioxane, THF, and a 50:50 mixture (Fig. 4). Initially, Ti(OiPr)4 was dissolved in the solvent and had absorption in the UV below 330 nm (Fig. 4A, the spectra at 0 and 5 min overlap). After an aliquot of DMEAA in toluene was injected into the cuvette, the reaction immediately turned golden brown and displayed broad absorbance across the visible and UV with a shoulder at ~350 nm. Time-dependent density functional theory (TDDFT) calculations


12


Page 14 of 24

of the electronic absorption spectrum of Ti3+(OiPr) 4-AlH3 in dioxane have absorption edge at ~350 nm (Fig. S9), while TDDFT calculations of the optical properties of monolayer-protected Al clusters indicate they have absorption edges around ~350 nm that increase as the Al clusters become larger.50 From these calculations, we conclude the initially observed broad spectra reflect the dynamic mixture of Ti3+(OiPr)4-AlH3 complexes and Ti-Al-H clusters present at the beginning of the reaction. After ~10 minutes, production of plasmonic Al NCs was evident from the increase in extinction at ~300 nm due to the growing number of light-scattering particles. The relative rate of Al NC growth in each solvent was evaluated by comparing the extinction at 300 nm during the reactions (Fig. 4B). The data from these reactions (Fig. S10) revealed that Al NC growth was faster in THF than dioxane while the average size of particles in THF (118 ± 22 nm) was two times larger than the size of the particles from dioxane (63 ± 25 nm), as indicated by TEM (Fig. 4C).

Scheme 2. Proposed four-step mechanism for the generation of Al NCs that is governed by colloidal stability from Lewis base ligands in the solvent. Brown and gray spheres are lowvalent Ti and Al respectively. Generation of H2 occurs during the growth of Al NCs but has been omitted for clarity.


13

Page 15 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The relative reaction rates and the final sizes of the Al NCs synthesized in dioxane and THF (Fig. 4) support the hypothesis that Ti-Al-H clusters have greater colloidal stability in dioxane than in THF. The colloidal stability of the clusters influences the time required for each stage of the proposed mechanism of Al NC growth (Scheme 2), which influences the amount of seeds produced and results in control of Al NC size by choice of solvent composition. After the generation of Ti3+(OiPr)3 catalysts, the growth of Al NCs can be described by four steps: 1) Formation and coalescence of colloidally unstable Ti-Al-H clusters. 2) Coalescence of colloidally unstable clusters into Al seeds with defined crystal structures. 3) The remaining colloidally unstable Ti-Al-H clusters and Al seeds coalesce with each other (Al seeds may also coalesce) and grow into colloidally stable Al NCs. 4) After Al NCs have become stable colloid in the reaction solvent and the unstable Ti-Al-H clusters have been depleted, Al NC growth ends. A self-limiting ~3-5 nm oxide shell forms after exposure of the Al NCs to ambient conditions, rendering them airstable.1,4 With ligands like dioxane and THF, the colloidal stability of Ti-Al-H clusters is largely determined by the formation of dative bonds between O and Al atoms. Taking account of their densities, dioxane (C4H8O2) has 1.89x more coordinating O atoms than THF (C4H8O) per mL of solvent which is inversely related to the average size of the particles produced in reactions with dioxane, THF or mixtures of the solvents (Fig. S11). Increasing the density of coordinating atoms in the reaction solution, and the ability to coordinate in bidentate manner increases the colloidal stability of Ti-Al-H clusters, producing more Al seeds. Since the initial amount of AlH3 was constant in the reactions, less Al was left after the formation of seeds in dioxane so smaller Al NCs were produced in dioxane than THF, which had fewer seeds and more Al left for growth. Density functional theory calculations of the thermodynamically favored structure of metallic Al clusters as a function of the number of atoms suggest that Al clusters are not crystalline Al seeds until they


14


Page 16 of 24

contain more than 65 atoms.51 For Al seeds and NCs, the most stable isomers were truncated octahedra with the face-centered cubic lattice of Al metal, while a minor fraction consisted of singly and multiply twinned particles, consistent with the final shapes of colloidally synthesized Al NCs (Fig. 1).51 To demonstrate the generality of transition metal catalyzed synthesis of Al NCs, reactions performed with vanadium(V) isopropoxide (VO(OiPr)3) and titanium(IV) tetrachloride (TiCl4) were examined by EPR spectroscopy (Fig. S12). VO(OiPr)3 was selected owing to 99.75% natural abundance of 51V nuclei with a nuclear spin of 7/2, while TiCl4 was chosen because both 37Cl and 35

Cl have nuclear spins of 3/2 and account for almost 100% of the natural abundance. The EPR

spectra of reactions performed with VO(OiPr)3 in dioxane and TiCl4 in toluene show the formation of V4+ and Ti3+ radicals after the addition of AlH3. As the ratio of AlH3 to the VO(OiPr)3 or TiCl4 increased, the spectra shifted and broadened in the same manner as the reactions with Ti(OiPr)4. Although Al NCs produced with Ti(OiPr)4, VO(OiPr)3, and TiCl4 in dioxane at 40 C, all had sizes of ~60 nm based on TEM imaging, the reactions with TiCl4 were ~4x faster while those with VO(OiPr)3 were ~10x slower than Ti(OiPr)4, suggesting that the ligands and the transition metal can be selected to tune the activity of the catalyst towards the formation of Al NCs. Notably, the polydispersity of the particles was greater with TiCl4 while slower catalysts like VO(OiPr)3 produced less polydisperse particles with superior plasmonic properties. Using VO(OiPr)3 slows the reaction enough that different sized Al NCs can be removed from the reaction as they grow, analogous to the synthesis of quantum dots with controlled sizes. As recently demonstrated, the ratio of AlH3 to Ti(OiPr)4 is an important parameter that enables control over Al NC size by controlling the rate of Al seed formation.13 This molecular-level understanding of the synthesis of


15

Page 17 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Al NCs by Ti3+(OiPr)3 catalyzed decomposition of AlH3 provides new opportunities for the rational synthesis of Al NCs of controlled size, shape, and composition. Conclusion: Through careful analysis of the colloidal synthesis of Al NCs through EPR and 1H NMR spectroscopies, a mechanism for the reactions by with Ti(OiPr)4 mediates the polymerization of AlH3 into Al NCs has been elucidated. AlH3 is as a single-source precursor for Al metal with hydride oxidation into H2, catalyzed by Ti3+(OiPr)3, providing the electrons required to produce metallic Al clusters. These clusters are colloidally unstable and coalesce and grow until they reach sufficiently large size to become colloidally stable. Ensemble measurements of particle growth indicated Al NCs grow more quickly in THF than dioxane owing to the lower colloidal stability of Ti-Al-H clusters in THF compared to dioxane, in agreement with the molecular-level mechanism of Al NC growth described in this work. By changing the solvent from THF to dioxane, the number of coordinating atoms available doubles, improving the colloidal stability of Ti-Al-H clusters and producing more Al seeds which results in smaller particles. This research provides a conceptual framework for the rational synthesis of Al NCs with controlled size that may help pave the route toward shape-controlled synthesis of Al NCs. Methods: Materials All chemicals were purchased from Sigma Aldrich and used as received unless otherwise noted. TiCl4 was obtained as a 1.0 M solution in toluene while 0.1 M stock solutions of Ti(OiPr)4 and VO(OiPr)3 were prepared in toluene. A 0.5 M solution of DMEAA in toluene was the molecular Al precursor for all reactions in this work. Anhydrous toluene, 1,4-dioxane, THF, TMEDA, and NMP were sparged with N2 before use and stored over molecular sieves. All


16


Page 18 of 24

glassware, including EPR/NMR tubes, cuvettes, vials, and syringes, was dried at 130 °C overnight before use. Synthesis of Al NCs Typical reactions were performed using a Schlenk line under N2 and standard techniques for handling air-sensitive materials. For these reactions, the temperature was maintained at 40 °C using an oil bath and hotplate with a thermocouple. Generally, 2 mL of DMEAA (0.5 M stock solution in toluene) was injected into 8 mL of the appropriate solvent to make a 100 mM solution of AlH3 and 0.1 mL of a 100 mM catalyst stock solution was injected once the solution reached the desired temperature. Reaction temperatures of 40 °C have been found to produce Al NCs with optimal size distributions, presumably because dimethylethylamine, which has a boiling point of ~37 °C, volatilizes out of the reaction, leaving only the ethers to direct the growth of the Al NCs. The reaction was allowed to proceed for two hours with Ti(OiPr)4, 30 min with TiCl4 and 24 hours with VO(OiPr)3. To adequately characterize the size and shape of the samples by TEM, the Al NCs were washed by adding 20 mL of 50 mM dibutyl phosphate in anhydrous cyclohexane at the end of the reaction. Dibutyl phosphate was used to quench unreacted AlH3 in the reactions, limiting the formation of extraneous oxide upon exposure to ambient conditions. The aluminum(III) dibutyl phosphate byproducts are soluble in hydrocarbons and cyclohexane has minimal absorption in the UV so residual cyclohexane molecules do not interfere with optical measurements of the Al NC colloids. To wash the Al NCs, they were centrifuged at 1000 ×g for ~5 min and re-suspended in ~25 mL of IPA followed by centrifugation at 10,000 ×g for ~10 min and re-suspension in IPA twice more.


17

Page 19 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TEM imaging TEM samples were prepared by drop casting a few µL of Al NCs in IPA onto the carbon coated grid. Structural characterization of the Al NCs was performed by TEM with a JOEL 1230 operating at 80 kV. The average Al NC size was determined from measurements of TEM images made using ImageJ software. EPR spectroscopy While inside of an argon filled glovebox, a 7” screw-capped Suprasil EPR tube with an outer diameter of 5 mm (Wilmad) was filled with 0.5 mL of a 10 mM solution of the transition metal catalyst in the appropriate solvent and then sealed using a screw cap with a silicon septum. The sealed vessels were transported to the X-band EPR spectrometer (Bruker EMX) and a 100 µL gas-tight syringe was used to inject aliquots of DMEAA (10 – 50 µLs) into the vessel to initiate the production of radical intermediates and the growth of Al NCs. For the EPR measurements, a single scan at 1 mW microwave power was acquired at each ratio of AlH3 to transition metal compound at room temperature using non-saturating conditions with a modulation amplitude of 0.1 G and a modulation frequency of 100 kHz. Liquid N2 was used to cool the samples enabling measurement of the EPR spectra at various temperatures. Simulations of select EPR spectra were performed using Simfonia. Parameters used to for the simulation of experimental spectra are presented in Table S1. Details on the deconvolution of the absorption EPR spectra are described in Fig. S8. 1

H NMR spectroscopy Samples for 1H NMR spectroscopy were prepared in a similar manner as those for EPR

analysis using the same tubes but using C4D8O (containing tetramethylsilane) and C4D8O2 as the solvents and a 20 mM concentration of Ti(OiPr)4. The 1H NMR spectra were acquired using a


18


Page 20 of 24

Bruker 500 MHz NMR at room temperature. A total of 8 scans were collected for each sample. Approximately 15 minutes elapsed between scans. UV-Visible spectroscopy For the analysis of relative reaction kinetics by optical spectroscopy, 10 mL of a 0.1 mM solution of Ti(OiPr)4 was prepared by serial dilution in either THF, dioxane, or a 50:50 mixture of the two in an argon filled glovebox and 3 mL was added to a UV-grade quartz cuvette with 10 mm path length that was sealed with an open-top screwcap and a silicon septum. The sealed cuvette was transferred to a Cary 5000 UV-Vis spectrometer with a temperature controller set to 40 °C. After 8 min for thermal equilibration, 30 µL of DMEAA (5 mM) was injected into the cuvette to initiate the formation of Al NCs and extinction spectrum was recorded every five minutes for an hour. The reactions were given four hours to proceed to completion before exposure to ambient conditions and washing with dibutyl phosphate stock. X-ray spectroscopies Energy dispersive X-ray spectroscopy was performed with a JOEL 2100F TEM operating at 200 kV. Powder X-ray diffraction spectra of dried Al NCs were measured using Cu K-alpha radiation from a Rigaku D/MAX-2100 spectrometer. Thermogravimetric analysis Dried samples (~1 mg) were heated to ~1100 C in pure oxygen and converted to Al2O3 with a Q600 Simultaneous TGA/DSC (TA Instruments). The mass gained during the experiment was used to determine the initial fraction of metallic Al from reactions in THF and dioxane.


19

Page 21 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References: (1)

(2)

(3)

(4)

(5) (6) (7)

(8)

(9)

(10)

(11)

(12) (13)

(14)

(15)

(16)

Tian, S.; Neumann, O.; Mcclain, M. J.; Yang, X.; Zhou, L.; Zhang, C.; Nordlander, P.; Halas, N. J. Aluminum Nanocrystals: A Sustainable Substrate for Quantitative SERSBased DNA Detection. Nano Lett. 2017, 17 (8), 5071–5077. Swearer, D. F.; Zhao, H.; Zhou, L.; Zhang, C.; Robatjazi, H.; Martirez, J. M. P.; Krauter, C. M.; Yazdi, S.; McClain, M. J.; Ringe, E.; et al. Heterometallic Antenna−reactor Complexes for Photocatalysis. Proc. Natl. Acad. Sci. 2016, 113 (32), 8916–8920. Smith, K. J.; Cheng, Y.; Arinze, E. S.; Kim, N. E.; Bragg, A. E.; Thon, S. M. Dynamics of Energy Transfer in Large Plasmonic Aluminum Nanoparticles. ACS Photonics 2018, 5 (3), 805–813. McClain, M. J.; Schlather, A. E.; Ringe, E.; King, N. S.; Liu, L.; Manjavacas, A.; Knight, M. W.; Kumar, I.; Whitmire, K. H.; Everitt, H. O.; et al. Aluminum Nanocrystals. Nano Lett. 2015, 15 (4), 2751–2755. DeSantis, C. J.; McClain, M. J.; Halas, N. J. Walking the Walk: A Giant Step toward Sustainable Plasmonics. ACS Nano 2016, 10, 9772–9775. Clark, B.; Jacobson, C. R.; Lou, M.; Yang, J.; Zhou, L.; Gottheim, S.; DeSantis, C. J.; Nordlander, P.; Halas, N. J. Aluminum Nanorods. Nano Lett. 2018, 18 (2), 1234. Haber, J. A.; Buhro, W. E. Kinetic Instability of Nanocrystalline Aluminum Prepared by Chemical Synthesis; Facile Room-Temperature Grain Growth. J. Am. Chem. Soc. 1998, 120 (42), 10847–10855. Sánchez-Iglesias, A.; Winckelmans, N.; Altantzis, T.; Bals, S.; Grzelczak, M.; LizMarzán, L. M. High-Yield Seeded Growth of Monodisperse Pentatwinned Gold Nanoparticles through Thermally Induced Seed Twinning. J. Am. Chem. Soc. 2017, 139 (1), 107–110. Chiu, C. Y.; Li, Y.; Ruan, L.; Ye, X.; Murray, C. B.; Huang, Y. Platinum Nanocrystals Selectively Shaped Using Facet-Specific Peptide Sequences. Nat. Chem. 2011, 3 (5), 393– 399. Lin, H. X.; Lei, Z. C.; Jiang, Z. Y.; Hou, C. P.; Liu, D. Y.; Xu, M. M.; Tian, Z. Q.; Xie, Z. X. Supersaturation-Dependent Surface Structure Evolution: From Ionic, Molecular to Metallic Micro/Nanocrystals. J. Am. Chem. Soc. 2013, 135 (25), 9311–9314. Chiu, C.-Y.; Wu, H.; Yao, Z.; Zhou, F.; Zhang, H.; Ozolins, V.; Huang, Y. FacetSelective Adsorption on Noble Metal Crystals Guided by Electrostatic Potential Surfaces of Aromatic Molecules. J. Am. Chem. Soc. 2013, 135 (41), 15489–15500. Polte, J. Fundamental Growth Principles of Colloidal Metal Nanoparticles - a New Perspective. CrystEngComm 2015, 17 (36), 6809–6830. Lu, S.; Yu, H.; Gottheim, S.; Gao, H.; DeSantis, C. J.; Clark, B. D.; Yang, J.; Jacobson, C. R.; Lu, Z.; Nordlander, P.; et al. Polymer-Directed Growth of Plasmonic Aluminum Nanocrystals. J. Am. Chem. Soc. 2018, jacs.8b08937. Loh, D.; Sen, S.; Bosman, M.; Tan, S. F.; Zhong, J.; Nijhuis, C. A.; Kral, P.; Matsudaira, P.; Mirsaidov, U. Multi-Step Nucleation of Nanocrystals in Aqueous Solution. Nat. Chem. 2017, 9, 77–82. Polte, J.; Ahner, T. T.; Delissen, F.; Sokolov, S.; Emmerling, F.; Thunermann, A.; Kraehnert, R. Mechanism of Gold Nanoparticle Formation in the Classical Citrate Synthesis Method Derived from Coupled In Situ XANES and SAXS Evaluation. J. Am. Chem. Soc. 2010, 132 (4), 1296–1301. Wang, F.; Richards, V. N.; Shields, S. P.; Buhro, W. E. Kinetics and Mechanisms of


20


(17)

(18) (19)

(20)

(21)

(22) (23) (24) (25) (26)

(27)

(28)

(29)

(30) (31) (32)

(33)

Page 22 of 24

Aggregative Nanocrystal Growth. Chem. Mater. 2014, 26 (1), 5–21. Liu, H.; Owen, J. S.; Alivisatos, A. P. Mechanistic Study of Precursor Evolution in Colloidal Group II-VI Semiconductor Nanocrystal Synthesis. J. Am. Chem. Soc. 2007, 129 (2), 305–312. Goulet, P. J. G.; Lennox, R. B. New Insights into Brust−Schiffrin Metal Nanoparticle Synthesis. J. Am. Chem. Soc. 2010, 132 (28), 9582–9584. Polte, J.; Tuaev, X.; Wuithschick, M.; Fischer, A.; Thuenemann, A. F.; Rademann, K.; Kraehnert, R.; Emmerling, F. Formation Mechanism of Colloidal Silver Nanoparticles: Analogies and Differences to the Growth of Gold Nanoparticles. ACS Nano 2012, 6 (7), 5791–5802. Straney, P. J.; Marbella, L. E.; Andolina, C. M.; Nuhfer, N. T.; Millstone, J. E. Decoupling Mechanisms of Platinum Deposition on Colloidal Gold Nanoparticle Substrates. J. Am. Chem. Soc. 2014, 136 (22), 7873–7876. Yin, X.; Shi, M.; Wu, J.; Pan, Y. T.; Gray, D. L.; Bertke, J. A.; Yang, H. Quantitative Analysis of Different Formation Modes of Platinum Nanocrystals Controlled by Ligand Chemistry. Nano Lett. 2017, 17 (10), 6146–6150. Bunce, N. J. Introduction to the Interpretation of Electron Spin Resonance Spectra of Organic Radicals. J. Chem. Educ. 1987, 64 (11), 907. Basu, P. Use of EPR Spectroscopy in Elucidating Electronic Structures of Paramagnetic Transition Metal Complexes. J. Chem. Educ. 2001, 78 (5), 666. Giles, J. R. M.; Roberts, B. P. E.S.R. Spectra and Reactivity of Alane Radical Anions in Solution. J.C.S Chem. Comm. 1981, 0, 1167–1168. Pluta, C.; Porschke, K.-R.; Kriiger, C.; Hildenbrand, K. An Al-Al One-Electron π Bond. Angew. Chem. Int. Ed. Engl. 1993, 32 (3), 388–390. Wiberg, N.; Blank, T.; Kaim, W.; Schwederski, B.; Linti, G. Tri(Supersilyl)Dialanyl (TBu3Si)3Al2 • and Tetra(Supersilyl)Cyclotrialanyl (TBu3Si)4Al3 • - New Stable Radicals of a Group 13 Elemenet from Thermolysis of (TBu3Si)4Al2. Eur. J. Inorg. Chem. 2000, 1475–1481. Henrici-Olivé, G.; Olivé, S. EPR Investigation of Transition Metal Complexes. XIII. EPR Detection of the Complex (C5H5)2TiH2AlH2. J. Organomet. Chem. 1970, 23 (1), 155– 157. Kenworthy, J. G.; Myatt, J.; Symons, M. C. R. Electron Spin Resonance of a Series of Biscyclopentadienyltitanium(III) Hydrides. Deductions from Solid-State Spectra. J. Chem Soc. A 1971, 1020–1024. Mach, K.; Antropiosova, H. The Catalytic System [(C5H5)2TiCl]2/LiAlH4 in Aromatic Solvents I. Formation of Titanocene Hydride Complexes; An ESR Study. J. Organomet. Chem. 1983, 248, 287–298. Barron, A. R.; Wilkinson, G. Transition-Metal Aluminohydride Complexes. Polyhedron 1986, 5 (12), 1897–1915. Krusic, P. J.; Tebbe, F. N. ESR Study of a Paramagnetic Titanium Methylene Complex and of Its Reactions with Olefins. Inorg. Chem. 1982, 21 (7), 2900–2902. Bennett, F. R.; Elms, F. M.; Gardiner, M. G.; Koutsantonis, G. A.; Raston, C. L.; Roberts, N. K. Stable Tertiary Phosphine Adducts of Alane. Organometallics 1992, 11 (4), 1457– 1459. Atwood, J. L.; Bennett, F. R.; Jones, C.; Koutsantonis, G. a.; Raston, C. L.; Robinson, K. D. Polydentate Tertiary Amine Aluminium Hydride Adducts: Monomeric versus


21

Page 23 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(34)

(35) (36)

(37)

(38)

(39) (40) (41)

(42)

(43) (44) (45) (46)

(47)

(48)

(49) (50)

(51)

Polymeric Species. J. Chem. Soc. Chem. Commun. 1992, No. 7, 541. Dautel, V. R.; Ziel, W. Ultrarotspeltroskopische Untersuchungen an Koordinationsverbindungen de Aluminiumwasserstoffs. Berichte der Bunsengesellschaft fur Phys. Chemie 1960, 1234–1244. Aldridge, S.; Downs, A. J. Hydrides of the Main-Group Metals: New Variations on an Old Theme. Chem. Rev. 2001, 101 (11), 3305–3365. Chopra, I. S.; Chaudhuri, S.; Veyan, J. F.; Chabal, Y. J. Turning Aluminum into a NobleMetal-like Catalyst for Low-Temperature Activation of Molecular Hydrogen. Nat. Mater. 2011, 10 (11), 884–889. Al-mahboob, A.; Muller, E.; Karim, A.; Muckerman, J. T.; Ciobanu, C. V; Sutter, P. SiteDependent Activity of Atomic Ti Catalysts in Al-Based Hydrogen Storage Materials. J. Am. Chem. Soc. 2012, 134, 10381–10384. Zheng, M. M.; Ren, T. Q.; Chen, G.; Kawazoe, Y. Improved Interaction of Hydrogen on Transition-Metal-Doped Al(100) Stepped Surface. J. Phys. Chem. C 2014, 118 (14), 7442–7450. Boruah, B.; Kalita, B. Exploring Enhanced Hydrogen Adsorption on Ti Doped Al Nanoclusters: A DFT Study. Chem. Phys. 2018, 25–30. Frankcombe, T. J. Proposed Mechanisms for the Catalytic Activity of Ti in NaAlH4. Chem. Rev. 2012, 112 (4), 2164–2178. Maryam, K. M.; Andrus, M. B.; Burt, S. R.; Woodfield, B. F. Generalized Preparation Method and Characterization of Aluminum Isopropoxide, Aluminum Phenoxide, and Aluminum n-Hexyloxide. Polyhedron 2013, 62, 18–25. Brown, A. C.; Altman, A. B.; Lohrey, T. D.; Hohloch, S.; Arnold, J. Hydride Oxidation from a Titanium-Aluminum Bimetallic Complex: Insertion, Thermal and Electrochemical Reactivity. Chem. Sci. 2017, 8, 5153–5160. Butler, M. J.; Crimmin, M. R. Magnesium, Zinc, Aluminium and Gallium Hydride Complexes of the Transition Metals. Chem. Commun. 2017, 53, 1348–1365. Liu, J.; Yu, J.; Ge, Q. Transition-Metal-Doped Aluminium Hydrides as Building Blocks for Supramolecular Assemblies. J. Phys. Chem. A 2010, 114 (46), 12318–12322. Cox, D. M.; Trevor, D. J.; Whetten, R. L.; Rohlfing, E. A.; Kaldor, A. Aluminum Clusters: Magnetic Properties. J. Chem. Phys. 1986, 84 (8), 4651–4656. Hua, Y.; Liu, Y.; Jiang, G.; Du, J.; Chen, J. Geometric Transition and Electronic Properties of Titanium-Doped Aluminum Clusters: AlnTi (n = 2-24). J. Phys. Chem. A 2013, 117 (12), 2590–2597. Cirri, A.; Silakov, A.; Lear, B. J. Ligand Control over the Electronic Properties within the Metallic Core of Gold Nanoparticles. Angew. Chemie - Int. Ed. 2015, 54 (40), 11750– 11753. Agrachev, M.; Antonello, S.; Dainese, T.; Ruzzi, M.; Zoleo, A.; Aprà, E.; Govind, N.; Fortunelli, A.; Sementa, L.; Maran, F. Magnetic Ordering in Gold Nanoclusters. ACS Omega 2017, 2 (6), 2607–2617. Sambles, J. R.; Cousins, J. E. Sample Size Effects on the Spin Resonance of Conduction Electrons in Aluminium. Phys. Stat. Sol. 1976, 75, 155–161. Lindgren, J.; Clayborne, A.; Lehtovaara, L. Optical Properties of Monolayer-Protected Aluminum Clusters: Time-Dependent Density Functional Theory Study. J. Phys. Chem. C 2015, 119 (33), 19539–19547. Shao, X.; Wu, X.; Cai, W. Growth Pattern of Truncated Octahedra in Al N (N < 310)


22


Page 24 of 24

Clusters. J. Phys. Chem. A 2009, 114 (1), 29–36. Acknowledgements: This research was financially supported by the Army Research Office (MURI W911NF-12-1 0407), the Air Force Office of Scientific Research Multidisciplinary Research Program of the University Research Initiative (MURI FA9550-15-1-0022), the Defense Threat Reduction Agency (HDTRA1-16-1-0042), the National Institute of Health (NS094535, A.T.) and the Welch Foundation under Grant C-1220 (N.J.H.) and C-1222 (P.N.) Author Information: Corresponding Author *E-mail: [email protected]. ORCID Benjamin D. Clark: 0000-0003-0390-9538 David Renard: 0000-0002-1917-679X Luca Bursi: 0000-0002-4530-0424 Naomi J. Halas: 0000-0002-8461-8494 Peter Nordlander: 0000-0002-1633-2937


23

Ligand-Dependent Colloidal Stability Controls the Growth of

Recommend Documents