Ligand Effects on Intramolecular Configuration, Intermolecular Packing, and Optical Properties of Metal Nanoclusters

Surface modification has served as an efficient approach to dictate nanocluster structures and properties. In this work, based on an Ag22 nanocluster template, the effects of surface modification on intracluster constructions and intercluster packing modes, as well as the properties of nanoclusters or cluster-based crystallographic assemblies have been investigated. On the molecular level, the Ag22 nanocluster with larger surface steric hindrance was inclined to absorb more small-steric chlorine but less bulky thiol ligands on its surface. On the supramolecular level, the regulation of intramolecular and intermolecular interactions in nanocluster crystallographic assemblies rendered them CIEE (crystallization-induced emission enhancement)-active or -inactive nanomaterials. This study has some innovation in the molecular and intramolecular tailoring of metal nanoclusters, which is significant for the preparation of new cluster-based nanomaterials with customized structures and enhanced performances.

Herein, a new Ag 22 nanocluster, formulated as Ag 22 (S-Adm) 10 (DPPM) 4 Cl 6 (abbreviated as Ag 22 -L1, where S-Adm = 1-adamantanethiol and DPPM = bis(diphenylphosphino)methane), was synthesized and structure-determined by X-ray single-crystal diffraction. The combination of this Ag 22 nanocluster and a previously reported Ag 22 (SPhMe 2 ) 12 (DPPE) 4 Cl 4 (abbreviated as Ag 22 -L2, where SPhMe 2 = 2,5-dimethyl thiophenol and DPPE = 1,2-bis (diphenylphosphino)ethane) constructed a platform to investigate the effects of surface modification on intramolecular constructions and intermolecular packing modes, as well as the properties of nanoclusters or cluster-based crystallographic assemblies. On the molecular level, because of the larger surface steric hindrance of Ag 22 -L1 relative to Ag 22 -L2, the Ag 22 -L1 surface contained more small-steric chlorine but fewer bulky thiol ligands. On the supramolecular level, Ag 22 -L2 displayed intramolecular and intermolecular interactions in its crystallographic assembly, while these interactions were absent in the Ag 22 -L1 crystal. Ag 22 -L2 was CIEE (crystallization-induced emission enhancement) active while Ag 22 -L1 was CIEE inactive. The optical absorptions and emissions of these two Ag 22 nanoclusters were also compared.
Synthesis of Ag 22 (S-Adm) 10 (DPPM) 4 Cl 6 (Ag 22 -L1). Specifically, 60 mg of AgNO 3 (0.36 mmol) and 40 µL of H 2 PtCl 6 (0.2 g/mL; 0.015 mmol) were dissolved in 20 mL of CH 3 OH and 1 mL of CH 3 CN. Then, 40 mg of DPPM (0.1 mmol) and 30 mg of HS-Adm (0.18 mmol) were added. After stirring for 30 min, 100 mg of NaBCNH 3 (1.59 mmol; dissolved in 2 mL of MeOH) was added. The reaction was allowed to proceed for 5 h. After that, the mixture in the organic phase was rotavaporated under vacuum and washed several times by MeOH and Hex. Then, 10 mL of CH 2 Cl 2 was used to extract the obtained Ag 22 -L1 nanocluster. The yield is 30% based on the Ag element (calculated from AgNO 3 ). Of note, although Pt did not exist in the final Ag 22 -L1, the absence of Pt sources resulted in the failure of the nanocluster synthesis ( Figure S1). Such a phenomenon has also been observed in previous works [53].  4 Cl 4 ](SbF 6 ) 2 + 2Cl − . Nanoclusters were crystallized in a CH 2 Cl 2 /ether system with a vapor diffusion method (Table S1).

Results
The Ag 22 -L1 nanocluster was synthesized by directly reducing the Ag-SR-DPPM complexes by NaBCNH 3 (Scheme S1; see more details in Materials and Methods). The electrospray ionization mass spectrometry (ESI-MS) measurement was performed to verify the molecular composition and to determine the valence state of the Ag 22 -L1 nanocluster. As shown in Figure S2, the mass result of the nanocluster exhibited an intense peak at 2897.54 Da. The excellent match of the experimental and simulated isotope patterns illustrated that the measured formula was [Ag 22 (S-Adm) 10 (DPPM) 4 Cl 6 ] 2+ . The "+2" valence state of the nanocluster matched well with the existence of (SbF 6 ) − counterions in the crystal lattice, i.e., the molar ratio between the cluster and the counterion was 1:2, as depicted in Figure S3. According to the valence states of Ag 22 -L1, its nominal electron counts was determined as 4e [56], i.e., 22(Ag) − 10(SR) − 6(Cl) − 2(charge) = 4e, the same as that of Ag 2 -L2 [54]. Moreover, the chlorine ligands in Ag 22 -L1 were proposed to originate from the H 2 PtCl 6 or from the CH 2 Cl 2 solvent, which has also been discovered in previously determined nanoclusters [57][58][59][60].

Results
The Ag22-L1 nanocluster was synthesized by directly reducing the Ag-SR-DPPM complexes by NaBCNH3 (Scheme S1; see more details in Materials and Methods). The electrospray ionization mass spectrometry (ESI-MS) measurement was performed to verify the molecular composition and to determine the valence state of the Ag22-L1 nanocluster. As shown in Figure S2, the mass result of the nanocluster exhibited an intense peak at 2897.54 Da. The excellent match of the experimental and simulated isotope patterns illustrated that the measured formula was [Ag22(S-Adm)10(DPPM)4Cl6] 2+ . The "+2" valence state of the nanocluster matched well with the existence of (SbF6) − counterions in the crystal lattice, i.e., the molar ratio between the cluster and the counterion was 1:2, as depicted in Figure S3. According to the valence states of Ag22-L1, its nominal electron counts was determined as 4e [56], i.e., 22(Ag) − 10(SR) − 6(Cl) − 2(charge) = 4e, the same as that of Ag2-L2 [54]. Moreover, the chlorine ligands in Ag22-L1 were proposed to originate from the H2PtCl6 or from the CH2Cl2 solvent, which has also been discovered in previously determined nanoclusters [57][58][59][60].
Structurally, the Ag22-L1 nanocluster contained an Ag10 kernel which comprised two distorted trigonal bipyramidal Ag5 units via an edge-edge vertical assembling mode (   The overall constructions of Ag 22 -L1 and Ag 22 -L2 nanoclusters were almost the same. However, because of the different steric hindrances of ligands in these two nanoclusters (i.e., S-Adm and DPPM in Ag 22 -L1; S-PhMe 2 and DPPE in Ag 22 -L2), these two nanoclusters displayed some structural differences: (i) For the kernel structure: the average Ag-Ag bond length in bipyramidal Ag 5 of Ag 22 -L1 was 2.824 Å, much shorter than that in Ag 22 -L2 (i.e., 2.933 Å). In addition, the average Ag-Ag bond lengths between these two Ag 5 bipyramids were 2.870 and 2.937 Å in Ag 22 -L1 and Ag 22 -L2, respectively. In this context, due to the larger surface steric hindrance of Ag 22 -L1 relative to Ag 22 -L2, the Ag 10 kernel of the former nanocluster was compressed.
(ii) For the surface environment: the biggest structural difference between the two Ag 22 nanoclusters lay in their surface ligand environments in terms of the proportion of the chlorine in peripheral ligands. Specifically, the Ag 22 -L1 nanocluster contained 10 thiol and 6 chlorine ligands, while Ag 22 -L2 included 12 thiol and 4 chlorine ligands ( Figure 2). As shown in Figure 2A,B, a thiol ligand at the specific location on the Ag 22 -L2 surface was substituted by a chlorine ligand in Ag 22 -L1. Another thiol ligand at the symmetrical position was also replaced by chlorine. Such a substitution from bulky thiol to small-steric chlorine was reasonable by considering that the more compact surface environment on Ag 22 -L1, resulting from the bulkier DPPM and S-Adm ligands relative to DPPE and S-PhMe 2 , was unable to host as many bulky thiol ligands as Ag 22 -L2 ( Figure 2C,D). Moreover, several intramolecular noncovalent C-H···π and π···π interactions were observed in the Ag 22 -L2 structure, which was advantageous to the compact packing of its surface ligands [54]. By comparison, none of such noncovalent interactions was observed in Ag 22 -L1, which might be another reason that more small-steric chlorine but fewer bulky thiol ligands were arranged on the Ag 22 -L1 nanocluster surface. The overall constructions of Ag22-L1 and Ag22-L2 nanoclusters were almost the same. However, because of the different steric hindrances of ligands in these two nanoclusters (i.e., S-Adm and DPPM in Ag22-L1; S-PhMe2 and DPPE in Ag22-L2), these two nanoclusters displayed some structural differences: (i) For the kernel structure: the average Ag-Ag bond length in bipyramidal Ag5 of Ag22-L1 was 2.824 Å, much shorter than that in Ag22-L2 (i.e., 2.933 Å). In addition, the average Ag-Ag bond lengths between these two Ag5 bipyramids were 2.870 and 2.937 Å in Ag22-L1 and Ag22-L2, respectively. In this context, due to the larger surface steric hindrance of Ag22-L1 relative to Ag22-L2, the Ag10 kernel of the former nanocluster was compressed.
(ii) For the surface environment: the biggest structural difference between the two Ag22 nanoclusters lay in their surface ligand environments in terms of the proportion of the chlorine in peripheral ligands. Specifically, the Ag22-L1 nanocluster contained 10 thiol and 6 chlorine ligands, while Ag22-L2 included 12 thiol and 4 chlorine ligands ( Figure 2). As shown in Figure 2A,B, a thiol ligand at the specific location on the Ag22-L2 surface was substituted by a chlorine ligand in Ag22-L1. Another thiol ligand at the symmetrical position was also replaced by chlorine. Such a substitution from bulky thiol to small-steric chlorine was reasonable by considering that the more compact surface environment on Ag22-L1, resulting from the bulkier DPPM and S-Adm ligands relative to DPPE and S-PhMe2, was unable to host as many bulky thiol ligands as Ag22-L2 ( Figure 2C,D). Moreover, several intramolecular noncovalent C-H···π and π···π interactions were observed in the Ag22-L2 structure, which was advantageous to the compact packing of its surface ligands [54]. By comparison, none of such noncovalent interactions was observed in Ag22-L1, which might be another reason that more small-steric chlorine but fewer bulky thiol ligands were arranged on the Ag22-L1 nanocluster surface.  The Ag 22 -L1 cluster entities were crystallized in a triclinic crystal system with a P-1 space group, whereas the Ag 22 -L2 cluster entities were crystallized in a tetragonal crystal system with an I4 1 /a space group. Both nanoclusters followed a lamellar eutectic packing pattern between R-nanocluster and S-nanocluster enantiomers in the crystal lattice; however, due to their distinct crystal systems, the interlayer distances were different: 26.561 Å of Ag 22 -L1, and 28.957 Å of Ag 22 -L2 (Figure 3 and Figure S5). Of note, there are equal Rnanocluster and S-nanocluster enantiomers in the crystal lattice, and the crystalline material of the nanocluster was racemic. Furthermore, owing to the existence of several benzenerings in the Ag 22 -L2 nanoclusters, strong intracluster and intercluster interactions occurred, including C-H···π interaction and π-π stacking [54]. In vivid contrast, these interactions were absent within the Ag 22 -L1 nanocluster or among Ag 22 -L1 cluster entities ( Figure S6). sphere, Ag; red sphere, S; magenta sphere, P; green sphere, Cl; grey sphere, C; pink/white sphere, H.
The Ag22-L1 cluster entities were crystallized in a triclinic crystal system with a P-1 space group, whereas the Ag22-L2 cluster entities were crystallized in a tetragonal crystal system with an I41/a space group. Both nanoclusters followed a lamellar eutectic packing pattern between R-nanocluster and S-nanocluster enantiomers in the crystal lattice; however, due to their distinct crystal systems, the interlayer distances were different: 26.561 Å of Ag22-L1, and 28.957 Å of Ag22-L2 (Figures 3 and S5). Of note, there are equal R-nanocluster and S-nanocluster enantiomers in the crystal lattice, and the crystalline material of the nanocluster was racemic. Furthermore, owing to the existence of several benzene-rings in the Ag22-L2 nanoclusters, strong intracluster and intercluster interactions occurred, including C-H···π interaction and π-π stacking [54]. In vivid contrast, these interactions were absent within the Ag22-L1 nanocluster or among Ag22-L1 cluster entities ( Figure S6). The Ag22-L1 nanocluster (dissolved in CH2Cl2) exhibited three intense absorptions centered at 368, 494, and 635 nm ( Figure 4A). By comparison, the UV-vis spectrum of Ag22-L2 displayed several peaks at 445, 512, and 670 nm ( Figure 4A). The blue shifts in the optical absorptions of Ag22-L1 relative to Ag22-L2 resulted from the different electronic structures of the two Ag22 nanoclusters. The CH2Cl2 solution of Ag22-L1 emitted at 650 nm, while the emission of Ag22-L2 was located around 670 nm ( Figure 4B). The 20 nm blueshift and 1.2-fold enhancement of the emission of Ag22-L1 relative to that of Ag22-L2 resulted from their different electronic structures. Indeed, these two nanoclusters displayed different optical absorptions, demonstrating their distinguishable electronic excitations and HOMO-LUMO energy gaps (HOMO: the highest occupied molecular orbital; LUMO: the lowest unoccupied molecular orbital). In addition, the different electronic excitations endowed these two nanoclusters with distinct emissions.  The Ag22-L2 nanocluster was CIEE active owing to the presence of extensive intramolecular and intermolecular interactions in its crystal lattice [54]. In this context, the emission intensity of Ag22-L2 in the crystalline state was remarkably higher than that of the nanocluster in the solution or the amorphous state. By comparison, the Ag22-L1 was CIEE inactive since no significant enhancement in emission intensity was observed (Figure 4C). Actually, the Ag22-L1 in the amorphous or crystalline state was almost non-emissive. Such a striking contrast was reasonable considering that the intramolecular and intermolecular interactions were absent in the crystal lattice of Ag22-L1, as mentioned above. The investigation of the Ag22 nanocluster system promoted the understanding of the crystalline packing mode and the CIEE of cluster-based nanomaterials.

Conclusions
In summary, a new Ag22 nanocluster, formulated as Ag22(S-Adm)10(DPPM)4Cl6, has been synthesized and structurally determined, which constituted an Ag22 cluster system together with the previously reported Ag22(S-PhMe2)12(DPPE)4Cl4. Based on this Ag22 cluster system, the effects of surface modification on intracluster constructions and intercluster packing modes, as well as the properties of nanoclusters or cluster-based crystallographic assemblies were investigated. The Ag22 nanocluster with larger surface steric hindrance was inclined to load more small-steric chlorine but fewer bulky thiol ligands on its surface. Moreover, the Ag22 nanocluster, which embodied several intramolecular and intermolecular interactions in cluster crystallographic assemblies, was CIEE active; by comparison, the Ag22 nanocluster without such interactions was CIEE inactive. This work provides new insight into the surface modification of metal nanoclusters and its effects on intramolecular configuration, intermolecular packing, and optical properties.

Supplementary
Materials: The following are available online at www.mdpi.com/article/10.3390/nano11102655/s1, Scheme S1. Synthetic procedure of the nanocluster; Figure S1. Comparison of optical absorptions of the nanocluster synthesis; Figure   The Ag 22 -L2 nanocluster was CIEE active owing to the presence of extensive intramolecular and intermolecular interactions in its crystal lattice [54]. In this context, the emission intensity of Ag 22 -L2 in the crystalline state was remarkably higher than that of the nanocluster in the solution or the amorphous state. By comparison, the Ag 22 -L1 was CIEE inactive since no significant enhancement in emission intensity was observed ( Figure 4C). Actually, the Ag 22 -L1 in the amorphous or crystalline state was almost non-emissive. Such a striking contrast was reasonable considering that the intramolecular and intermolecular interactions were absent in the crystal lattice of Ag 22 -L1, as mentioned above. The investigation of the Ag 22 nanocluster system promoted the understanding of the crystalline packing mode and the CIEE of cluster-based nanomaterials.

Conclusions
In summary, a new Ag 22 nanocluster, formulated as Ag 22 (S-Adm) 10 (DPPM) 4 Cl 6 , has been synthesized and structurally determined, which constituted an Ag 22 cluster system together with the previously reported Ag 22 (S-PhMe 2 ) 12 (DPPE) 4 Cl 4 . Based on this Ag 22 cluster system, the effects of surface modification on intracluster constructions and intercluster packing modes, as well as the properties of nanoclusters or cluster-based crystallographic assemblies were investigated. The Ag 22 nanocluster with larger surface steric hindrance was inclined to load more small-steric chlorine but fewer bulky thiol ligands on its surface. Moreover, the Ag 22 nanocluster, which embodied several intramolecular and intermolecular interactions in cluster crystallographic assemblies, was CIEE active; by comparison, the Ag 22 nanocluster without such interactions was CIEE inactive. This work provides new insight into the surface modification of metal nanoclusters and its effects on intramolecular configuration, intermolecular packing, and optical properties.
Author Contributions: X.K. and M.Z. designed the study; S.W., X.W., H.L., H.S. and J.H. performed the experiments and analyzed the data. All authors discussed the results and commented on the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding:
We acknowledge the financial support by NSFC (21631001 and 21871001), the Ministry of Education, and the University Synergy Innovation Program of Anhui Province (GXXT-2020-053).

Data Availability Statement:
The X-ray crystallographic coordinates for structures reported in this work have been deposited at the Cambridge Crystallographic Data Center (CCDC), under deposition numbers CCDC-2106804. These data can be obtained free of charge from the Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif, which has been mentioned in the article.

Conflicts of Interest:
The authors declare no conflict of interest.