Mitsuhiro
Iwasaki
a,
Naoki
Kobayashi
a,
Yukatsu
Shichibu
ab and
Katsuaki
Konishi
*ab
aGraduate School of Environmental Science, Hokkaido University, North 10 West 5, Sapporo 060-0810, Japan
bFaculty of Environmental Earth Science, Hokkaido University, North 10 West 5, Sapporo 060-0810, Japan. E-mail: konishi@ees.hokudai.ac.jp; Fax: +81 11 7064538; Tel: +81 11 7064538
First published on 29th June 2016
In the recent development of structurally defined ligand-stabilized gold clusters, it has been revealed that not only the inorganic units but also the surrounding organic ligands substantially affect their electronic/optical properties. In this work, a series of core + exo type Au8 clusters decorated by dppp (Ph2P(CH2)3PPh2) and arylthiolate ligands ([Au8(dppp)4(SR)2]2+, 1–5) were synthesized, and their optical properties were studied in order to gain insights into the perturbation effects of the organic ligands. 1–5 showed visible absorption and photoluminescence emission bands at longer wavelengths compared to their chloro- and acetylide-modified analogues, suggesting the contribution of weak non-bonding interactions of the Au framework with the ligand heteroatoms. Upon acid treatment, 2- and 4-pyridinethiolate clusters (R = Py, 2 and 4) showed larger red shifts of the absorption and emission bands than the 3-pyridyl isomer (3), implying the involvement of the resonance structures of the SPy units. On the other hand, all regioisomers (2–4) showed large photoluminescence enhancements upon pyridine protonation. X-ray crystallographic and NMR analyses of 4 and its protonated form (4′) showed that the electron-deficient pyridinium rings of 4′ form π-stacks with neighbouring phenyl groups of dppp, suggesting that the orientation of the surface aromatics is a plausible factor governing the emission efficiency. These observations provide examples of successful modulation of optical properties of small gold clusters through the electronic and/or steric perturbation by the proximal organic ligands, highlighting the importance of the ligand design in the fine tuning of cluster properties directed for optical chemosensors and luminescent materials.
Recently, we reported the site-specific introduction of organic functionality on the core + exo type octagold cluster in the synthesis of [Au8(dppp)4(CCR′)2]2+, which accommodates four diphosphine (dppp = Ph2P(CH2)3PPh2) and two acetylide ligands.31 We also provided an example of the modulation of the optical properties through π-conjugated units attached to the gold cluster. For example, pyridylethynyl-modified clusters (R′ = Py) showed protonation-induced absorption and photoluminescence responses, which were critically dependent on the relative position of the pyridine nitrogen atom. Especially, the photoluminescence of the 3-pyridyl isomer was negligibly affected by the protonation event, which is in contrast to the significant quenching observed for the 2- and 4-pyridyl isomers. From these observations, we have proposed that the resonance structures of π-conjugated ligand units critically affect the electronic structure of the cluster moieties.
In the present paper, we used analogous Au8 clusters with arylthiolate ligands ([Au8(dppp)4(SR)2]2+, 1–5) to obtain insights into the factors affecting the electronic transitions of the Au8 moiety. First, we demonstrate that weak non-bonding interactions involving the ligand heteroatoms have substantial effects on the clusters' electronic properties through the study on a series of core + exo type Au8 clusters with different anionic ligands. We also investigated the optical response of the pyridinethiolate (SPy)-modified clusters (R = Py) to the pyridine protonation events, and show that the optical absorption (excitation) and photoluminescence emission energies were critically dependent on the π-resonance structures of the SPy units, but the emission intensity (quantum yield) was not. Based on the 1H NMR investigations coupled with X-ray structural analyses we show that the emission intensities are mainly governed by the structural factors rather than the electronic factors of the ligand environment. Although there are several recent reports of gold clusters that show ligand-dependent electronic properties,26–28 the present approach is advantageous since the overall geometrical feature of the gold skeleton (Au8) is essentially preserved, which allows the inspection of the subtle effects of the ligand environments through the systematic studies for clusters with a variety of ligands.
![]() | ||
Fig. 1 (a) Synthesis scheme for [Au8(dppp)4(SR)2]2+ (1–5) and (b) crystal structure of 4 with hydrogen atoms omitted for clarity. |
The crystal structure of 4 also reveals that the thiolate ligands are linked to the gold unit in a monodentate manner. This is in contrast to the binding patterns found in conventional all-thiolate protected clusters,1 whose thiolate ligands mostly have multidentate character to form protecting staple motifs around a polyhedral gold core. The two S-substituents are oriented along the long axis of the prolate-shaped gold units, reflecting the sp3 character of the S atom and steric hindrance with the neighbouring P–Ph groups. The molecular structure in the solid state appears to be virtually retained in solution. The 31P NMR spectrum of 4·(NO3)2 in CD3OD/CD2Cl2 showed signals at 36.1, 50.5 and 54.4 ppm with an integrated intensity ratio of 1:
2
:
1 (Fig. S4c, ESI†), which is in agreement with the X-ray structure. Such 31P NMR patterns were also observed for the other thiolate-modified clusters (see the ESI†).
![]() | ||
Fig. 3 Schematic illustration of possible weak interactions between the gold unit and the neighbouring (a) S or (b) N atoms. |
We have also reported previously that Au8 clusters (6 and 7) exhibit photoluminescence at ∼580 nm upon excitation of the HOMO–LUMO transition.6,31 Likewise, the thiolate-modified clusters (1–5) showed similar photoluminescence but the emission bands were observed at longer wavelengths, which are in accordance with the red shifts of the absorption bands (Table 1).
Entry | Cluster | X | Absorption | Photoluminescence | |||||
---|---|---|---|---|---|---|---|---|---|
λ f /nm | λ c /nm | Δλ/nm | λ f /nm | λ c /nm | Δλ/nm | I c/If | |||
a In MeOH at 20 °C (2 μM). b With MeSO3H (1.20 mM). | |||||||||
1 | 1·(NO3)2 | SPh | 530 | 530 | 0 | 621 | 621 | 0 | 1.0 |
2 | 2·(NO3)2 | S(2-Py) | 518 | 543 | 25 | 604 | 629 | 25 | 10.2 |
3 | 3·(NO3)2 | S(3-Py) | 528 | 532 | 4 | 612 | 618 | 6 | 4.7 |
4 | 4·(NO3)2 | S(4-Py) | 526 | 535 | 9 | 609 | 625 | 16 | 4.5 |
It should be noted that the degree of the protonation-induced red shifts varied significantly with the N position of the pyridine functionality (Fig. 2a and Table 2, entries 2–4). Similar trends have been reported in the optical responses of pyridylethynyl-ligated Au8 clusters ([Au8(dppp)4(CCPy)2]2+),31 for which the critical effects of the resonance structures of the π-conjugated ligand moiety were implicated. Also in the present thiolate system, the electronic effect of the π-resonance structure appeared to be a plausible main factor responsible for the observed spectral shifts. For instance, in the protonated form of the 3-pyridyl isomer (3), the location of the positive charge should be limited only within the pyridine ring (Fig. 4b). Consequently, the protonation would cause only minor electronic effects on the Au8 unit, and accordingly, the observed shifts of the absorption and emission bands were very small (Fig. 2a(ii); Table 2, entry 3). On the other hand, when 2 or 4 is protonated, a resonance contributor with the C
S+–Au fragment should be generated (Fig. 4a). The positive charge and the conjugated π-system attached to the exo Au atom would substantially affect the electronic structures of the Au8 unit, which may lead to large red shifts. Among the three isomers (2–4), the 2-pyridyl isomer (2) showed especially large band shifts (25 nm). As mentioned in the above section, the N atom of the free-base form of 2 likely interacts with the exo gold atom (Fig. 3b). The protonation would result in the breaking of such weak interaction, which may also contribute to the relatively large responses.
The perturbation effects coupled with the π-resonance structures of the SPy ligand were also observed in the shifts of the photoluminescence emission bands, where 2 and 4 showed definitely larger red shifts (25 and 16 nm, respectively) than 3 (6 nm) (Fig. 2b and Table 2). In contrast, such trends were not found for the emission intensities. The protonated forms of the three isomers all showed much higher emission intensities than the free-base forms by factors of 4–10 (Table 2). This is in contrast to the trend observed for ([Au8(dppp)4(CCPy)2]2+)31 upon protonation; 2- and 4-pyridyl isomers showed considerable losses of the emission (quenching), while the 3-isomer was virtually inactive. The comparable increases of the emission intensities of 3 and 4 upon pyridine protonation indicated that the resonance structures of the SPy unit are not involved in the ‘turn-on’ photoluminescence response. So we have to consider different mechanisms in order to account for this behaviour.
![]() | ||
Fig. 5 (a) Crystal structures of cationic moieties of 4·(NO3)2 and 4′·(BF4)4. (b) Structures with the orientation of a pyridine ring and the neighbouring Ph2 ring highlighted. |
Based on these observations, we next recorded the 1H and 31P NMR spectra of 4·(NO3)2 in the presence and absence of methanesulfonic acid. The aromatic region of the 1H NMR spectrum of 4·(NO3)2 at 293 K (Fig. 6a) reveals signals due to the pyridyl protons and four sets of Ph groups (Ph1, Ph2, Ph3, and Ph4), which were assigned with the aid of the 1H–1H COSY spectrum (Table S6 and Fig. S7, ESI†). Variable temperature studies showed that the broadened signals of the Ph2 and Ph3 protons are due to the slow exchange under the NMR time scale (Fig. S8, ESI†). Fig. 6b shows the spectrum of the protonated form (4′), where the pyridine protons underwent downfield shifts reflecting the decrease of the electron density of the aromatic ring. If the pyridine protonation does not affect the orientation and structures of the phosphine ligand moieties, the Ph signals would appear at the same positions. However, the signals due to some phenyl groups showed notable shifts upon pyridine protonation. Especially, the o-protons of Ph2 nearby the pyridine ring showed a marked upfield shift (Δδ = 0.12 ppm). Definite shifts were also observed for some protons of Ph1 and Ph2, whereas some other Ph protons (such as of Ph4) remained almost unchanged. Therefore, the protonation event caused the orientation change of the Ph2 and adjacent Ph1 units. Accordingly, in the 31P NMR at 293 K, the P atoms attached to the Ph1 (P1, Fig. S9, ESI†) and Ph2/Ph3 rings (P2) showed downfield shifts upon pyridine protonation.
![]() | ||
Fig. 6 Aromatic region of the 1H NMR spectra of 4 in CD3OD (1.5 mM) at 293 K (a) before and (b) after the addition of MeSO3H (2 molar eq.). For the numbering of the Ph groups, see Fig. 5a. |
The above different NMR profiles of 4 and its protonated form (4′) seem to reflect the conformational difference found in the crystalline state. As mentioned, the crystal structure of 4′ revealed that the Ph2 units favour to form π-stacked complexes with the electron-deficient pyridinium moiety. Such interligand π–π interaction arising from the electron-deficient character would also occur in solution, which may be associated with the observed shifts of the NMR signals. Indeed, in the 1H NMR spectrum of 5, which bears electron-withdrawing nitro substituents on the thiolate ligand of 1, the o-Ph2 protons appeared evidently upfield (δ = 7.63) from those of 1 (δ = 7.76) (Fig. S10, ESI†). This is very similar to the trend found in the spectra of 4 and 4′ (Fig. 6). Thus, the orientation of the surface aromatic arrays can be altered by switching of the weak interligand interaction through the control of π-electron density, which may lead to the observed emission enhancements.
For the factors affecting the HOMO/LUMO features, geometrical structures of the gold frameworks and electronic perturbations of the organic ligand moieties are considered. Inspections of the X-ray crystallographic structures available (4, 4′, 6, 7) showed minor differences in the geometrical parameters of the Au8 frameworks (e.g., Au–Au lengths/angles, symmetry) (Tables S3–S5, ESI†). However, there is no clear trend to support the correlation of the cluster geometry and the absorption and emission energies. For instance, the Auexo–Auedge distances of 4′ and 6 were almost similar to each other (Table S3, ESI†), but the absorption band positions were much different (Tables 1 and 2). Although we have to take account of the difference of the structures in solution and packed crystals, electronic rather than geometrical factors of the ligands are likely to affect the absorption and photoluminescence band energies. Thus, for the thiolate-modified clusters (1–5), the weak interaction of the S atom and Auedge atoms may result in the formation of new orbitals. Among them, the 2-pyridinethiolated cluster (2) shows a band at an exceptionally short wavelength (518 nm) when compared with the other thiolate-modified clusters (526–531 nm), which may be correlated with the additional interaction of the pyridine N atom with Auexo, as illustrated in Fig. 3.
Meanwhile, the absorption and emission bands of 2 and 4 considerably red-shifted upon pyridine protonation, for which a critical role of the π-resonance structures of the SPy units was suggested (Fig. 4a). As mentioned, a resonance contributor with a positive charge on the S atom (e.g., 4′b, Fig. 7) is considered primarily responsible for the red shifts of the absorption/emission bands. The generation of such a resonance structure would lead to the cleavage of the S⋯Auedge weak interaction in the free base form. Indeed, in the crystal structures, the shortest S–Auedge distances of 4′ (3.437 Å) were much longer than those of 4 (3.166 Å) (Table S3, ESI†), and were comparable to the sum of van der Waals radii (3.46 Å). Thus, upon protonation, they would behave similarly to the non-thiolate type clusters (6, 7, Table 2), and would show blue shifts of the absorption band to ∼510 nm. However, 2 and 4 both showed opposite trends (red shifts) (Table 2 and Fig. 2). From these observations, the electronic coupling of the Au8 unit with a π-resonance structure of the SPy+ units of 2 or 4 may be involved in the observed red shifts (Fig. 7). The difference in the red shifts between 2 and 4 (Table 2, entries 2 and 4) suggests that the degree of the cluster–π interaction depends on the probability to take the effective resonance form having a positively charged S atom (e.g., 4′b). It should be also noted that [Au8(dppp)4(CCPy)2]2+ exhibits similar red shifts when the protonation induces an extended charged resonance structure.31 Therefore, the electronic coupling between the gold unit and the π-system appears to be a general phenomenon, which will expand the scope of cluster compounds. Such cluster–π electronic interactions have been recently reported in an alkynyl-ligated Au13 cluster system.29
From these considerations, it is clear that ligand moieties can cause direct perturbation effects on the electronic properties of the Au8 framework through weak coordinative non-bonding interactions with heteroatoms (S or N) and/or electronic coupling with the π-system. These electronic effects are likely the results of the involvement of the atomic orbitals of the heteroatoms or π-system in the molecular orbitals (e.g., HOMO and LUMO), so the difference of the absorption/emission energies may reflect the degree of the contribution of the ligand-derived moieties to the molecular orbitals. Further theoretical studies are required for the comprehensive understanding of the electronic perturbations of the ligands. Nevertheless, the above findings demonstrate the utility of the organic ligand environments for the fine-tuning of the electronic structures/properties of cluster compounds.
In this regard, recent papers on the ligand effects of emissive metal complexes showed that high emission intensities are obtained when the steric hindrances of the ligand environments are increased.36,37 It is claimed that the small structural difference between the ground and excited states, which arises from the enhanced rigidity of the hindered complexes, leads to the decrease of the probability of the non-radiative path. We think that such a steric-based mechanism can reasonably account for our observation. Thus, the π-stack formation upon pyridine protonation may rigidify the encapsulating environment to suppress the conformational freedom. Consequently, the structural change upon excitation would be restricted, leading to high emission intensity. Thus, the emission efficiency of the present Au8 cluster system may be controlled by the switching of weak non-bonding interligand interaction around the inorganic moiety.
Footnote |
† Electronic supplementary information (ESI) available: Details of synthesis and characterisation data, and supporting spectroscopic and crystal data. CCDC 1477150 for 4·(NO3)2 and 1477151 for 4′·(BF4)4. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6cp03129c |
This journal is © the Owner Societies 2016 |