Symmetry breaking of highly symmetrical nanoclusters for triggering highly optical activity

Abstract

Developing new approaches to fulfill the enantioseparation of nanocluster racemates and construct cluster-based nanomaterials with optical activity remains highly desired in cluster science, because it is an essential prerequisite for fundamental research and extensive applications of these nanomaterials. We herein propose a strategy termed “active-site exposing and partly re-protecting” to trigger the symmetry breaking of highly symmetrical nanoclusters and to render cluster crystals optically active. The vertex PPh₃ of the symmetrical Ag₂₉(SSR)₁₂(PPh₃)₄ (SSR = 1, 3-benzenedithiol) nanocluster was firstly dissociated in the presence of counterions with large steric hindrance, and then the exposed Ag active sites of the obtained Ag₂₉(SSR)₁₂ nanocluster were partly re-protected by Ag⁺, yielding an Ag₂₉(SSR)₁₂-Ag₂ nanocluster with a symmetry-breaking construction. Ag₂₉(SSR)₁₂-Ag₂ followed a chiral crystallization mode, and its crystal displayed strong optical activity, derived from CD and CPL characterizations. Overall, this work presents a new approach (i.e., active-site exposing and partly re-protecting) for the symmetry breaking of highly symmetrical nanoclusters, the enantioseparation of nanocluster racemates, and the achievement of highly optical activity.

Graphical Abstract

.

1. Introduction

Chirality has long been one of the most important research topics since it is an amazing phenomenon ubiquitous in life, nature, and the universe ranging from nanoscale molecules (e.g., proteins, sugars, and DNA) to mater-scale substances (e.g., eyes, shells, and flowers), and to the vast galaxy [1], [2], [3]. As one of the most appealing characteristics of nanostructures, optical activity, including circular dichroism (CD), circular polarized luminescence (CPL), and vibrational circular dichroism (VCD), has attracted much attention and been considered as the prerequisite to exploit chirality-related applications [4,5]. The study of chirality dated back to 1811 when the optical activity was observed in asymmetric quartz crystal by François Arago, after which tremendously experimental and theoretical efforts were made [6], [7], [8]. In nanoscience, the optical activity is generally achieved either by constructing metal nanoparticle-based assemblies in chiral arrangements or by the conjugation of metal nanoparticles with peripheral chiral molecules [9], [10], [11]. Indeed, the large-sized nanoparticle itself without chiral stabilizers is almost optically inactive in light of its homogeneous and symmetrical packing in the metallic kernel. In vivid contrast, metal nanoclusters, routinely described as ultrasmall metal nanoparticles [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], have been exploited as ideal platforms to investigate the intrinsic chirality of metallic kernels, owing to their plentiful kernel constructions and kernel-surface bonding environments [24], [25], [26].

The chirality of nanoclusters mainly originates from three aspects: (i) chirality in the metallic kernel, (ii) chirality in the metal-ligand interface, and (iii) chirality in the peripheral carbon group [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38]. Among these aspects, the first aspect (i.e., chiral metallic kernel) has attracted much scientific interest since it not only represents the most intrinsic character in analyzing the origin of chirality in nanomaterials, but also exists in the other two aspects due to the transmission effect via intracluster interactions [39]. Besides, aside from nanoclusters with chiral peripheral ligands, most structurally asymmetrical nanoclusters with achiral ligands are racemic in their solutions and crystals [40,41]. In this context, the chiral separation of these racemic clusters is an essential prerequisite for their fundamental research and extensive applications. Although several approaches (e.g., high-performance liquid chromatography separation and chiral self-assembly) have been exploited to separate cluster enantiomers from their racemates [42], [43], [44], [45], [46], [47], the enantioseparation has only been accomplished in limited examples. New approaches to fulfill the chiral resolution of nanocluster racemates and construct cluster-based nanomaterials with optical activity remain highly desired in cluster science.

Herein, a strategy termed “active-site exposing and partly re-protecting” was proposed to trigger the symmetry breaking of highly symmetrical nanoclusters and to render cluster crystals optically active (Scheme 1). The Ag₂₉(SSR)₁₂(PPh₃)₄ (Ag₂₉-PPh₃ for short; SSR = 1,3-benzenedithiol) nanocluster was highly symmetrical with four Ag-PPh₃ vertex units. The introduction of counterions with large steric hindrance to the nanocluster system induced the dissociation of vertex PPh₃ and the generation of Ag₂₉(SSR)₁₂ (Ag₂₉ for short) with exposed surface Ag active sites. Furthermore, the Ag⁺ addition triggered the re-protection of partly exposed Ag sites on the Ag₂₉ nanocluster surface, yielding the Ag₂₉(SSR)₁₂-Ag₂ (Ag₂₉-Ag for short) nanocluster with a symmetry-breaking construction. The cluster crystals of both Ag₂₉-PPh₃ and Ag₂₉ clusters were racemic; by comparison, the crystallization of Ag₂₉-Ag followed a chiral mode, accomplishing the enantioseparation of nanocluster racemates. Accordingly, the crystal of the symmetry-breaking Ag₂₉-Ag cluster displayed strong optical activity, derived from CD and CPL characterizations.

Scheme 1.

Illustration of the “active-site exposing and partly re-protecting” strategy for triggering the symmetry breaking of highly symmetrical nanoclusters and rendering cluster crystals optically active.

The DTAB addition-induced PPh₃ dissociation rendered vertex active Ag of the Ag₂₉-PPh₃ nanocluster exploring, while the obtained Ag₂₉ cluster still adhered to the racemic crystallization. Besides, the Ag⁺ addition triggered the re-protection of partly exposed Ag active sites and the symmetry breaking of the nanocluster, and the crystallization of Ag₂₉-Ag cluster entities follows a chiral mode.

2. Experimental methods

2.1. Materials

All following reagents were purchased from Sigma-Aldrich and used without further purification, including silver nitrate (AgNO₃, 99%, metal basis), triphenylphosphine (PPh₃, 99%), 1,3-benzene dithiol (SSR, 99%), sodium borohydride (NaBH₄, 99%), dodecyltrimethylammonium bromide (CH₃(CH₂)₁₀CH₂-N-(CH₃)₃Br, DTAB, 98%), tetramethylammonium bromide ((CH₃)₄NBr, TMAB, 98%), tetramethylammonium bromide ((C₄H₉)₄NBr, TBAB, 98%), silver acetate (CH₃COOAg, 99%), methylene chloride (CH₂Cl₂, HPLC, Sigma-Aldrich), methanol (CH₃OH, HPLC grade), N,N-dimethylformamide (DMF, HPLC grade), and ethyl ether ((CH₃CH₂)₂O, HPLC grade).

2.2. Synthesis of Ag₂₉(SSR)₁₂(PPh₃)₄ (Ag₂₉-PPh₃)

The preparation of Ag₂₉-PPh₃ was based on the reported method of the Bakr group [48]. The counterion of Ag₂₉-PPh₃ was Na⁺.

2.3. Nanocluster transformation from Ag₂₉-PPh₃ to Ag₂₉(SSR)₁₂ (Ag₂₉)

30 mg of the Ag₂₉-PPh₃ crystal was dissolved in 10 mL of DMF, and 50 mg of DTAB was added under vigorous stirring. After 30 min, the organic layer was separated to produce the Ag₂₉ nanocluster, which was used for crystallization directly. The yield was 95% based on the Ag element (calculated from the Ag₂₉-PPh₃).

2.4. Nanocluster transformation from Ag₂₉ to Ag₂₉(SSR)₁₂-Ag₂ (Ag₂₉-Ag)

30 mg of the Ag₂₉ crystal was dissolved in 10 mL of DMF, and 3 mg of CH₃COOAg was added under vigorous stirring. After 30 min, the organic layer was separated and poured into 200 mL of CH₂Cl₂. The precipitate was collected to produce the Ag₂₉-Ag nanocluster. The yield was 90% based on the Ag element (calculated from the Ag₂₉). Of note, the crystal analysis demonstrated that the Ag₂₉-Ag crystal contained both Ag₂₉-Ag and Ag₂₉ nanoclusters. The counterions of both Ag₂₉ and Ag₂₉-Ag were DTAB.

2.5. Crystallization of the Ag₂₉ nanocluster series

Single crystals of these Ag₂₉ nanoclusters were cultivated at 15 °C by vapor diffusing the ethyl ether into a DMF solution of the cluster. After 2 weeks, red crystals were collected, and the structures of these Ag₂₉ nanoclusters were determined.

2.6. Preparation of nanocluster crystalline films

The concentration of the DMF/CH₂Cl₂ (1:9 of the volume ratio) solution of these Ag₂₉ nanoclusters was set as 30 mg/mL and then the solution was filtered with a 0.2 µm syringe filter. The solutions were stored for 12 h before use. 50 µL of the solutions were dropped onto a quartz substrate, and spin-coated (using LAURELL WS-650MZ-23NPPB) at 1000 rpm for 60 s. The cluster-impregnated quartz substrate was dried in the air for 12 h before the optical property characterization.

2.7. Characterizations

The optical absorption spectra of nanoclusters were recorded using an Agilent 8453 diode array spectrometer.

The photoluminescence (PL) spectra were measured on a FL-4500 spectrofluorometer with the same optical density.

Electrospray ionization mass spectrometry (ESI-MS) measurements were performed by a Waters XEVO G2-XS QTof mass spectrometer. The sample was directly infused into the chamber at 5 μL/min. For preparing the ESI samples, nanoclusters were dissolved in DMF (1 mg/mL) and diluted (v/v = 1:1) by CH₃OH.

³¹P nuclear magnetic resonance (NMR) spectra were acquired using a Bruker 600 Avance III spectrometer equipped with a Bruker BBO multinuclear probe (BrukerBioSpin, Rheinstetten, Germany).

The circularly polarized luminescence (CPL) spectra of nanoclusters were recorded using a JASCO CPL-300 instrument. For CPL measurements, the parameters were set as follows: scanning speed of 200 nm/min, response time of 2 s, band width of 10 nm, accumulations of 6.

The circular dichroism (CD) spectra were measured on a JASCO J-1500 circular dichroism spectrophotometer. The solid samples were measured with a DRCD-574 solid samples accessories, using an integrating sphere to detect the diffuse reflectance of samples.

2.8. X-Ray crystallography

The data collection for single-crystal X-ray diffraction (SC-XRD) of all nanocluster crystal samples was carried out on Stoe Stadivari diffractometer under nitrogen flow, using graphite-monochromatized Cu Kα radiation (λ = 1.54186 Å). Data reductions and absorption corrections were performed using the SAINT and SADABS programs, respectively. The structure was solved by direct methods and refined with full-matrix least squares on F² using the SHELXTL software package. All non-hydrogen atoms were refined anisotropically, and all the hydrogen atoms were set in geometrically calculated positions and refined isotropically using a riding model. All crystal structures were treated with PLATON SQUEEZE. The diffuse electron densities from these residual solvent molecules were removed. The CCDC number of the racemic Ag₂₉(SSR)₁₂ nanocluster is 2071574. The CCDC number of the chiral Ag₂₉(SSR)₁₂-Ag₂ (S enantiomer) nanocluster is 2071575. The CCDC number of the chiral Ag₂₉(SSR)₁₂-Ag₂ (R enantiomer) nanocluster is 2071643. The CCDC number of Ag₂₉(SSR)₁₂(PPh₃)₄ in the presence of TMAB is 2150072. The CCDC number of Ag₂₉(SSR)₁₂(PPh₃)₄ in the presence of TBAB is 2150121.

3. Results and discussion

The Ag₂₉-PPh₃ nanocluster was prepared using the previously reported procedure [48]. The introduction of DTAB (N,N,N-trimethyl-1-dodecanaminium bromide) induced the transformation from Ag₂₉-PPh₃ to Ag₂₉, and the Ag₂₉-Ag nanocluster was obtained via anchoring Ag⁺ onto the Ag₂₉ nanocluster surface (see Experimental Method for more details). Only two Ag⁺ ions could be introduced onto the nanocluster surface, while the Ag₂₉(SSR)₁₂-Ag_x nanoclusters with x = 1, 3, 4 were absent. Such a tendency might result from the tunable chemical reactivity of the Ag₂₉(SSR)₁₂ framework in reacting with the introduced Ag⁺ ions. The bare Ag₂₉(SSR)₁₂ could react with Ag⁺ with a high degree of activity, while the Ag₂₉-Ag showed no activity to further react with Ag⁺. That is, the addition of Ag⁺ onto the Ag₂₉(SSR)₁₂ framework might passivate the corresponding nanocluster. ESI-MS measurement was performed to determine the molecular formula of these Ag₂₉ nanoclusters (Fig. S1). The ESI-MS spectrum of Ag₂₉-PPh₃ exhibited five separated signals, corresponding to the [Ag₂₉(SSR)₁₂(PPh₃)_n]³⁻ where n ranged from 0 to 4 (Fig. S1a,b). The existence of these five peaks was in agreement with the previously reported “dissociation-aggregation pattern” of the PPh₃ ligands in the nanocluster [49]. The mass spectrum of Ag₂₉ showed an intense mass peak that matched the [Ag₂₉(SSR)₁₂]³⁻. All PPh₃ ligands were dissociated from the Ag₂₉ surface after the DTAB addition since no peak of [Ag₂₉(SSR)₁₂(PPh₃)_n]³⁻ (n = 1-4) was observed (Fig. S1c). The Ag₂₉-Ag displayed two mass peaks, as shown in Fig. S1d, and the excellent match of the experimental and simulated isotope patterns demonstrated that these two peaks matched the [Ag₂₉(SSR)₁₂]³⁻ and [Ag₂₉(SSR)₁₂-Ag]²⁻, respectively. The mass signal of [Ag₂₉(SSR)₁₂-Ag]²⁻ verified the capture of Ag⁺ onto the Ag₂₉(SSR)₁₂ surface. However, the mass peak of [Ag₂₉(SSR)₁₂-Ag₂]¹⁻ was absent, resulting from the weak interactions between the Ag₂₉(SSR)₁₂ framework and Ag⁺ ions. Besides, the PPh₃ dissociation among the conversion from Ag₂₉-PPh₃ to Ag₂₉ and Ag₂₉-Ag was further verified by the ³¹P NMR measurement, where the 26.20 ppm signal of Ag₂₉-PPh₃ disappeared in the latter two nanoclusters (Fig. S2).

The structural comparisons of these Ag₂₉ nanoclusters are shown in Fig. 1 and S2–S4. The Ag₂₉-PPh₃ contained an icosahedral Ag₁₃ kernel that was stabilized by four Ag₃(SR*)₆, where two SR* make up a SSR, to generate an Ag₂₅(SSR)₁₂ framework. The four terminals of Ag₂₅(SSR)₁₂ were further capped by Ag-PPh₃ units, giving rise to the overall structure of the highly symmetrical Ag₂₉-PPh₃ (Fig. 1a and S3). Upon the DTAB-addition induced conversion from Ag₂₉-PPh₃ to Ag₂₉, the PPh₃ ligands on the Ag₂₉-PPh₃ surface were dissociated while the overall configuration of nanocluster remained highly symmetrical (Fig. 1b and S4). Such a PPh₃ dissociation was proposed to result from the competition effect between PPh₃ and DTAB ― the PPh₃ ligands followed a “dissociation-aggregation pattern” on the Ag₂₉ nanocluster surface [49], while the presence of DTAB with a long carbon chain enabled the nanocluster surface to be fully covered and the dissociated PPh₃ could no longer be re-anchored onto the nanocluster vertex. Besides, the large steric hindrance of DTAB might also cause the PPh₃ dissociation in light of the steric effect in the nanocluster crystal lattice. Indeed, the Ag₂₉ nanocluster would maintain its framework, and no PPh₃ ligand was peeled off from the cluster surface when TMAB and TBAB surfactants with short carbon chains were introduced. Of note, such a bare Ag₂₉ structure has also been discovered in the presence of C₆₀ with a large steric (i.e., Ag₂₉(SSR)₁₂(C₆₀)_n) [50].

Fig. 1.

Atomically precise structures of the Ag₂₉ nanocluster series. (a) The racemic Ag₂₉ nanocluster enantiomers. (b) The racemic Ag₂₉ nanocluster enantiomers. (c) The chiral Ag₂₉-Ag nanoclusters (R enantiomer). (d) The chiral Ag₂₉-Ag nanoclusters (S enantiomer). Color legends: blue/light blue sphere, Ag; red spher e, S; magenta sphere, P; grey sphere, C. For clarity, all H atoms are omitted.

Furthermore, the Ag⁺ addition to the “bare” Ag₂₉ triggered the re-protection of partly exposed Ag sites (2/4; the 50% re-occupation) on the nanocluster surface, yielding the Ag₂₉-Ag nanocluster (Fig. 1c and S5). Significantly, because of the presence of the combined Ag⁺, Ag₂₉-Ag displayed a symmetry-breaking construction. The Ag(cluster vertex)•••Ag(combination) bond lengths were all around 2.77 Å, demonstrating its strong binding ability given that such lengths were close to the Ag(kernel)•••Ag(icosahedral surface) bond lengths in the Ag₁₃ kernel of the nanocluster. The corresponding bond lengths among Ag₂₉-PPh₃, Ag₂₉, and Ag₂₉-Ag nanoclusters were further compared (Fig. S6). The average bond lengths of Ag(kernel)•••Ag(icosahedral surface) in Ag₂₉ and Ag₂₉-Ag were both lengthened relative to that of Ag₂₉-PPh₃ with 0.98% and 0.76%, respectively (Fig. S6a). Besides, the average Ag(icosahedral surface)•••Ag(icosahedral surface) bond lengths in Ag₂₉-PPh₃ were increased by 2.06% and 0.76%, respectively, to Ag₂₉ and Ag₂₉-Ag (Fig. S6b). In addition, the average Ag(icosahedral surface)•••Ag(motif) bond length displayed a 1.03% elongation in both Ag₂₉ and Ag₂₉-Ag compared with that of the Ag₂₉-PPh₃ (Fig. S6c). Accordingly, both the icosahedral Ag₁₃ kernel and the Ag₂₅(SSR)₁₂ framework were extended with the conversion from Ag₂₉-PPh₃ to Ag₂₉ and Ag₂₉-Ag. As for the interactions between the Ag vertex and the icosahedral Ag₁₃, the average bond lengths in Ag₂₉ and Ag₂₉-Ag were 3.058 and 3.151 Å, respectively (Fig. S6d, solid lines). However, no analogous interaction was perceived in Ag₂₉-PPh₃ ― distances between them (Fig. S6d, dotted lines) ranged from 3.493 to 3.643 Å (averagely, 3.523 Å). Consequently, the vertex Ag atoms became closer to the icosahedral Ag₁₃ kernel upon the conversion from Ag₂₉-PPh₃ to both Ag₂₉ and Ag₂₉-Ag, and the newly generated Ag₄ pyramid-like structures rendered the Ag₂₉ framework more robust [51,52].

Ag₂₉-PPh₃ cluster compounds can be crystallized with cubic and trigonal systems, i.e., Ag₂₉-PPh₃-cubic and Ag₂₉-PPh₃-trigonal (Fig. S7a,b) [48,53]. Although the crystals of Ag₂₉ and Ag₂₉-Ag appeared to be cubes, the same as those of Ag₂₉-PPh₃ crystals, the crystalline systems of both Ag₂₉ and Ag₂₉-Ag were orthorhombic with space groups of Pbcn and C222₁, respectively (Fig. S7c–e). The comparisons of cell parameters of different Ag₂₉ crystal lattices were shown in Fig. S7f. Each crystal lattice of Ag₂₉-Ag was composed of four cluster units, i.e., 4*{[Ag₂₉(SSR)₁₂-Ag₂]₂[Ag₂₉(SSR)₁₂]₁}, and the three cluster entities in a unit were packed with a “Λ-shape” mode (Fig. S8).

The innermost icosahedral Ag₁₃ in the Ag₂₉ nanocluster was highly symmetrical, whereas the asymmetric arrangement of the surface "triskelion"-like Ag₄(SR)₆ architectures rendered the chiral torsion of the overall cluster framework (Fig. 1). For Ag₂₉-PPh₃, each crystal lattice was composed of eight Ag₂₉ cluster compounds, half of which were R-enantiomers while another half were S-enantiomers (Fig. 1a). In this context, the Ag₂₉-PPh₃ crystal was racemic (Fig. 2a). The same phenomenon was observed in the Ag₂₉ crystal lattice (Fig. 1b and 2b). Consequently, although both Ag₂₉-PPh₃ and Ag₂₉ nanocluster crystals were highly emissive, their crystals were CD and CPL silent and optically inactive (Fig. 2e,f, black and red lines).

Fig. 2.

Packing of Ag₂₉ nanocluster entities in the crystal lattice and the corresponding optical activity. (a) Crystal lattice of the racemic Ag₂₉-PPh₃ nanoclusters with no optical activity. (b) Crystal lattice of the racemic Ag₂₉ nanoclusters with no optical activity. (c) Crystal lattice of chiral Ag₂₉-Ag nanoclusters (R enantiomer) with optical activity. (d) Crystal lattice of chiral Ag₂₉-Ag nanoclusters (S enantiomers) with optical activity. The orange and blue labels of Ag₂₉ cluster entities represent their R and S enantiomerism, respectively. (e) CD spectra of different crystals of the Ag₂₉ series. (f) CPL spectra of different crystals of the Ag₂₉ series.

Significantly, the Ag₂₉-Ag nanocluster entities underwent chiral self-assembly with the crystallization process, which was reminiscent of the behavior of tartaric acids. Although the crystal lattice of Ag₂₉-Ag contained both Ag₂₉-Ag and Ag₂₉ cluster molecules, all these molecules were R-type (or S-type) enantiomers in the R-Ag₂₉-Ag crystal (or S-Ag₂₉-Ag crystal), as depicted in Fig. 1c,d and 2c,d. In this context, the crystals of R-Ag₂₉-Ag and S-Ag₂₉-Ag displayed intense CD and CPL signals and were highly optically active. Luo et al. reported that the addition of ligands to superatom structures could activate or passivate a nanocluster [58]. Herein, the combination of active-Ag site exposing induced by DCTB addition and the partly re-protecting induced by Ag⁺ addition might activate the Ag₂₉(SSR)₁₂ cluster framework to follow an asymmetrically crystallographic packing and display highly optical activities. As shown in Fig. 2e and S9, R-Ag₂₉-Ag and S-Ag₂₉-Ag crystals exhibited mirror-image CD signals in the same wavelength region (i.e., at about 330 and 530 nm) with a dissymmetry factor of |g| = 7.0 × 10⁻⁴. By comparison, the enantio-separated Ag₂₉-PPh₃ nanocluster solutions displayed mirror image CD spectra at 460 nm with a dissymmetry factor of 1.5 × 10⁻³, much higher than that of the crystal of R-Ag₂₉-Ag and S-Ag₂₉-Ag [47]. The CPL results of R-Ag₂₉-Ag and S-Ag₂₉-Ag crystals showed a single signal at 795 nm (Fig. 2f). The maximum |g_lum| value of R-Ag₂₉-Ag or S-Ag₂₉-Ag crystals was determined to be approximately 5 × 10⁻² (Fig. S10), much higher than those of the reported metal nanoclusters [54], [55], [56], demonstrating the high optical activity of these nanocluster crystals. By comparison, the chiral Ag₂₉ nanoclusters protected by DHLA (dihydrolipoic acid) exhibited mirrored CPL signals at 660 nm with |g_lum| value of 2 × 10⁻³ [57]. The differences of the CD signals between enantio-separated Ag₂₉-PPh₃ solutions and chiral Ag₂₉-Ag crystals as well as the CPL signals between chiral Ag₂₉(DHLA)₁₂ solutions and chiral Ag₂₉-Ag crystals might stem from two aspects: (i, on the molecular level) their different molecular structures and surface chemistry and (ii, on the supramolecular level) their different packing states and intercluster interactions.

Upon the dissolution of the Ag₂₉-Ag crystal, the CPL signal was disappeared (Fig. S11a), suggesting that the chirality tautomerism occurred in equilibrium. Inversely, the re-crystallization of the Ag₂₉-Ag solution induced the chiral crystallographic self-assembly of cluster compounds that rendered the nanocluster CPL active again. Accordingly, the reversible transformation between the CPL-off solution and the CPL-on crystal of clusters has been accomplished (Fig. S11b). Because all solutions of Ag₂₉-PPh₃, Ag₂₉, and Ag₂₉-Ag were optically inactive, the generation of CPL was irrelevant to the PPh₃ dissociation, but was related to the Ag⁺ combination. Besides, considering that the Ag⁺ combination on the Ag₂₉(SSR)₁₂ surface was not that robust (as evidenced by ESI-MS), we defined that the chiral self-assembly was mainly triggered by the Ag⁺ combination among the cluster crystallization. Of note, the Nakashima group has reported the enantiomers of the Ag₂₉-PPh₃ nanocluster by using high-performance liquid chromatography (HPLC) [47]. These Ag₂₉-PPh₃ cluster molecules maintained their structures and compositions during the HPLC process. Compared with the HPLC technology, the crystallization-induced enantioseparation in this work was assigned to a chemical approach wherein the composition and configuration of the nanocluster were altered to a certain extent.

The DMF solutions of all Ag₂₉ nanoclusters exhibited almost the same optical absorptions with an intense peak at 445 nm and several shoulder bands at 320, 365, and 510 nm (Fig. 3a, solid lines); such similar absorptions demonstrated that the electronic orbits of clusters were mainly constituted by the Ag₂₉(SSR)₁₂ framework, but were irrelevant to the surface PPh₃ or Ag stabilizers, which was reminiscent of the electronic properties of the Ag₂₉(DHLA)₁₂ nanocluster [59]. Besides, the Ag₂₉-PPh₃ and Ag₂₉-Ag cluster solutions emitted at 640 nm, while the Ag₂₉ displayed photoluminescence (PL) at 633 nm (Fig. 3a, dotted lines). The Ag₂₉-PPh₃ showed the strongest PL intensity among all Ag₂₉ clusters, and 3% and 12% reductions were detected by comparing the emission intensities of Ag₂₉ and Ag₂₉-Ag, respectively, with that of Ag₂₉-PPh₃.

Fig. 3.

Optical properties of different Ag₂₉ nanoclusters. (a) Comparison of optical absorptions and PL emissions (nanoclusters were dissolved in DMF) of Ag₂₉-PPh₃ (black lines), Ag₂₉ (red lines), and Ag₂₉-Ag (blue lines) nanoclusters. (b) Comparison of optical absorptions and PL emissions (nanoclusters were in a crystallized film) of Ag₂₉-PPh₃ (black lines), Ag₂₉ (red lines), and Ag₂₉-Ag (blue lines) nanoclusters.

The optical absorptions and emissions of cluster crystalline films were further compared (Fig. 3b). The optical absorptions of all Ag₂₉ films were similar, whereas the 525 nm peak of Ag₂₉ was more intensive than those of other clusters (Fig. 3b, solid lines). The more pronounced absorption feature at 525 nm of the Ag₂₉ crystalline film might arise due to the intercluster close packing enhancing the excitations between the Ag₁₃ subunit and surface ligands [59]. Besides, a 10 nm red-shift was observed (i.e., 455 nm versus 445 nm) by comparing these absorptions with those of the cluster solutions. The emission wavelengths of cluster crystallized films displayed remarkable red-shifts relative to those of cluster solutions ― Ag₂₉-PPh₃ film emitted at 700 nm, Ag₂₉ film emitted at 707 nm, and Ag₂₉-Ag film emitted at 652 nm (Fig. 3b, dotted lines). The significant shift in emissions between different forms was expected to result from a combined effect of the electronic coupling and of lattice-origin, non-radiative decay pathways occurring through electron-phonon interactions [48,60,61]. In reference to the red-shift of emissions from the 700 nm of the Ag₂₉-PPh₃ film to the 707 nm of the Ag₂₉ film, or the blue-shift to the 652 nm of the Ag₂₉-Ag film, in addition to the aforementioned explanations, such a shift was also rationalized in terms of the different surface structures and crystalline packing modes among different Ag₂₉ nanoclusters [51,52,62].

4. Conclusion

In summary, a strategy termed “active-site exposing and partly re-protecting” was presented to trigger the symmetry breaking of highly symmetrical nanoclusters and the chiral self-assembly of cluster molecules in crystals, and to render these crystals highly optically active. The introduction of counterions with large steric hindrance dissociated the PPh₃ from the symmetrical Ag₂₉(SSR)₁₂(PPh₃)₄ nanocluster, and the vertex exposed Ag active sites of the nanoclusters was re-protected by Ag⁺, yielding an Ag₂₉(SSR)₁₂-Ag₂ nanocluster with a symmetry-breaking construction. The obtained Ag₂₉(SSR)₁₂-Ag₂ underwent chiral self-assembly with the crystallization process, and its crystal displayed strong optical activity, determined by both CD and CPL characterizations. Our work presents a new strategy for breaking the symmetry of highly symmetrical nanoclusters, inducing the enantioseparation of nanocluster racemates, and achieving strong optical activity of cluster-based nanomaterials.

Declaration of competing interest

The authors declare that they have no conflicts of interest in this work.

Biographies

Manzhou Zhu(BRID: 09776.00.26551) is currently the Changjiang Chair Professor of Chemistry at Anhui University. He received his Ph.D. in Chemistry from the University of Science and Technology of China (USTC) in 2000. He then conducted postdoctoral research at USTC and Carnegie Mellon University (CMU, Pittsburgh, USA). He joined the chemistry faculty of Anhui University in 2010. His current research interests include atomically precise nanoclusters, structure-property correlation of nanoclusters, and applications.

Xiao Wei received his M.S. in Chemistry from Anhui University and is currently a Ph.D. candidate under the supervision of Prof. Manzhou Zhu. His research interests include the structure and property of metal nanoclusters.