English

• Over the past few decades, photosensitizers (PSs) have attracted significant attention due to their extensive applications in photocatalytic organic reactions [1-7], photodynamic therapy (PDT) [8-13], photon-driven water splitting for hydrogen production [14,15], and triplet-triplet annihilation upconversion [16-18]. Efficient photosensitization requires strong absorption of excitation light, a high intersystem crossing (ISC) yield to produce the triplet state, and a long triplet lifetime to enable reactions with reactants [19-21]. PSs activate molecular oxygen to generate reactive oxygen species (ROS) such as singlet oxygen (¹O₂), hydroxyl radicals (^•OH), hydrogen peroxide (H₂O₂), and superoxide anions (O₂^•−), which act as powerful oxidants in photocatalysis [22,23].

  Enhancing PS photosensitization efficiency involves increasing the ISC rate by decreasing the energy gap (ΔE_ST) between the lowest singlet and triplet states or increasing the spin-orbit coupling (SOC) constant [24-26]. The heavy-atom effect, utilizing elements like Pt, Ir, Ru, I, and Br, is a key approach [27,28]. For example, iodine-modified homoleptic bis(dipyrrinato)zinc(Ⅱ) complexes, synthesized by Gasser, Chao, and their collaborators, exhibit enhanced ISC rates and improved PDT properties [29]. Similarly, an unsymmetrical donor-acceptor PS with triphenylamine and pyrazine units designed by the Xing group demonstrated efficient photooxidation under blue light [30]. Furthermore, Zhang group showcased guest-regulated ROS generation in porphyrin-based metallacages, achieving enhanced photosensitization through ΔE_ST reduction [31]. These advances highlight the demand for novel PSs with superior efficiency and unique applications.

  Since the mid-1990s, coordination-driven self-assembly based on the formation of metal-ligand coordination bonds characterized by controllability, directionality, and reversibility has emerged as a powerful methodology for constructing discrete supramolecular coordination complexes (SCCs), including one-dimensional (1D) helices, two-dimensional (2D) polygons, and three-dimensional (3D) polyhedral [32-37]. The careful selection of coordination metal ions and multidentate organic building blocks allows for the generation of SCCs with well-defined topological architectures (sizes, shapes, and geometries) and unique physicochemical properties, as well as attractive functionalities across various research fields, including chemistry, biology, and materials science [38-49]. SCCs also provide platforms for PS systems with high photosensitization efficiency. For instance, Mukherjee et al. developed a water-soluble Pd₁₆L₈ metallacage functionalized with benzothiadiazole for selective aerobic oxidation of aryl sulfides [50]. Moreover, pre- and post-assembly modifications enable the integration of PSs into SCC skeletons, allowing precise control over PS number, distribution, and energy transfer interactions. Recently, our group successfully demonstrated ROS-increasing and ROS-switchable photosensitization processes by utilizing various PSs in Pt(Ⅱ)-based metallacycle platforms through energy transfer interactions [51]. Despite significant progress in SCC-based PSs with numerous applications, challenges persist. For example, the heavy-atom effect in Pt(Ⅱ)-based PS metallacycles may lead to noticeable fluorescence quenching and an increased triplet state lifetime compared to their coordination PS precursors, attributed to the substantial ISC caused by the SOC process. The underlying mechanisms of this phenomenon have yet to be thoroughly explored. Additionally, there is a relative lack of efficient and straightforward methods to adjust the photosensitization of SCC-based PSs.

  In this context, we designed and synthesized a BF₂-chelated dipyrromethene (BODIPY)-containing dipyridyl donor (L1) and four Pt(Ⅱ) acceptors (L2–L5) featuring different Pt atoms and organic moieties. When the dipyridyl building block was combined with these four Pt(Ⅱ) acceptors, a series of triangular metallacycles (M1–M4) was obtained via coordination-driven self-assembly with exceptionally high yields (Fig. 1). The coordination interactions between the pyridine nitrogen atoms and the Pt(Ⅱ) metal centers imparted distinct photophysical properties to the constructed BODIPY-based metallacycles, differentiating them from one another and from their BODIPY-containing precursors. Notably, the photosensitization efficiency of these metallacycles exhibited tunable properties, influenced by the varying Pt(Ⅱ) acceptors with different Pt atoms and organic moieties, as well as differing degrees of π-conjugation. Additionally, leveraging the excellent photosensitization efficiency of metallacycle M3, we successfully demonstrated its photocatalytic application in the visible-light-driven oxidative coupling of various amines to imines. This research not only presents a novel and straightforward approach for tuning the photosensitization efficiency and photocatalytic reactivity of SCC-based PSs through supramolecular coordination strategies, but also explores their potential in catalyzing chemical oxidation to mimic photosynthetic processes.

  Figure 1

  

  Figure 1.  Schematic illustration of the coordination-driven self-assembly of triangular metallacycles M1–M4 and the chemical structures of building blocks L1–L5.

  DownLoad: Full-Size Img PowerPoint

  The dipyridyl donor L1, modified with the BODIPY group, was synthesized in three steps with a moderate yield (Scheme S1 in Supporting information). Notably, the BODIPY fluorophore was selected as the photosensitive unit due to its unique photophysical properties, including strong absorption of visible light, ease of functionalization, and excellent photostability. The Pt(Ⅱ) acceptors L2–L5, featuring different numbers of Pt atoms and varying degrees of organic π-conjugated moieties, were synthesized as previously described. This design was intended to enable simple and efficient regulation of the photosensitization process in the resulting BODIPY-containing metallacycles.

  By combining the dipyridyl donor L1 with four different Pt(Ⅱ) acceptor ligands (L2–L5), we prepared a family of triangular metallacycles (M1–M4) via a coordination-driven self-assembly strategy with high yields. Stirring mixtures of Pt(Ⅱ) acceptor ligands L2 and L3 with dipyridyl donor L1 in a 1:1 ratio in dichloromethane (DCM)/N,N-dimethylacetamide (DMAC) at 60 ℃ for 24 h led to the formation of discrete triangular metallacycles M1 and M2, respectively (Schemes S2 and S3 in Supporting information). Notably, triangular metallacycles M3 and M4 were synthesized by mixing di-Pt(Ⅱ) acceptors L4 and L5 with dipyridyl donor L1 in CH₂Cl₂ at room temperature for 2 h (Schemes S4 and S5 in Supporting information). Subsequently, comprehensive characterizations including multinuclear nuclear magnetic resonance spectroscopy (NMR) analysis (¹H, ¹³C, ³¹P, 2D diffusion-ordered spectroscopy (DOSY), ¹H–¹H correlation spectroscopy (COSY), and nuclear Overhauser effect spectroscopy (NOESY), electrospray ionization time-of-flight mass spectrometry (ESI-TOF-MS), and theoretical simulations were conducted to confirm the reliable construction of discrete and highly symmetric 2D metallacycles (Figs. S10–S38 in Supporting information).

  The obtained metallacycles M1–M4 were initially characterized using ³¹P{¹H} and ¹H multinuclear NMR spectroscopy. For example, in the ¹H NMR spectrum of metallacycle M3 (Fig. 2a), the H_a and H_b protons of the pyridine rings exhibited downfield shifts due to the coordination of the pyridine-N atoms with the Pt(Ⅱ) metal center, resulting in a loss of electron density compared to the dipyridyl donor L1. As shown in the ³¹P{¹H} NMR spectrum (Fig. 2b), a notable shift of approximately 5.84 ppm was observed for the sharp singlet of metallacycle M3 compared to the starting Pt(Ⅱ) acceptor L4. Additionally, there was a decrease in the coupling of the flanking ¹⁹⁵Pt satellites by approximately ΔJ_PPt = 181.8 Hz, attributed to back-donation from the platinum atoms. The constructed metallacycles M1-M4 were further characterized using 2D DOSY, ¹H-¹H COSY, and NOESY NMR techniques. For instance, the observation of cross peaks between the H_a and H_b signals of the pyridine rings and the PEt₃ protons (-CH₂ and –CH₃) corroborated the formation of the discrete triangular scaffold based on Pt-N bond coordination (Figs. S14, S20, S28 and S34 in Supporting information). Additionally, the 2D DOSY spectra of the prepared metallacycles M1–M4 revealed only one set of signals, indicating the formation of isolated species with high symmetry (Figs. S15, S21, S29 and S35 in Supporting information). ESI-TOF-MS investigation provided further support for the construction of discrete triangular metallacycles M1–M4. As illustrated in Fig. 2c, two main peaks at m/z = 1165.9733 and 902.9434 were observed in the mass spectrum of metallacycle M3, corresponding to the different charge states [M − 4OTf^-] and [M − 5OTf^-], respectively, resulting from the loss of trifluoromethanesulfonate counterions. Importantly, the isotopic resolution of these peaks aligned well with their theoretical distribution, suggesting the successful formation of the pre-designed triangular metallacycle. For the other metallacycles, ESI-TOF-MS also provided satisfactory mass data, as shown in Figs. S23 and S36 (Supporting information), confirming the formation of the corresponding metallacycles. To simulate the topological structure of the obtained metallacycles, the extended semiempirical tight-binding GFN2-xTB method (i.e., the geometry, frequency, noncovalent, extended tight binding semiempirical model) developed by Grimme and colleagues was employed due to its high computational efficiency for calculating structures and noncovalent interactions in large assembly systems [52-54]. As shown in Fig. S38 (Supporting information), molecular simulations indicated that metallacycles M1–M4 all exhibited a similar, roughly planar triangular structure with varying internal diameters. Additionally, the edge lengths of metallacycles M1–M4 were determined to be 17, 16, 27, and 24 Å, respectively.

  Figure 2

  

  Figure 2.  (a) The partial ¹H NMR spectra (500 MHz, CD₂Cl₂, 298 K) of the ligand L4 (up), the self-assembled metallacycle M3 (middle) and ligand L1 (bottom). (b) The ³¹P NMR spectra (202 MHz, CD₂Cl₂, 298 K) of metallacycle M3 (up) and ligand L4 (bottom). (c) Theoretical (top) and experimental (bottom) ESI-TOF-MS of metallacycle M3.

  DownLoad: Full-Size Img PowerPoint

  The photophysical properties of metallacycles M1–M4, along with their BODIPY-based precursor L1, were investigated. The absorption spectra of ligand L1 and metallacycles M1–M4 were recorded in dichloromethane (Fig. 3, Figs. S1–S3 in Supporting information). The spectra displayed similar absorption bands at approximately 517 nm, characterized by relatively high molar absorption coefficients, attributed to the π–π^* transition of the BODIPY units (Fig. 3a). As illustrated in Figs. 3b, f and Table 1, ligand L1, as well as metallacycles M1 and M2, exhibited strong fluorescence emission at around 549, 560, and 560 nm, corresponding to light-green, yellow, and yellow colors in dichloromethane, respectively. In contrast, metallacycles M3 and M4 showed significant fluorescence quenching. When comparing the data from ligand L1 and metallacycles M1–M4 in the 3D excitation-emission matrix (3D EEMs) measurements (Fig. S2 in Supporting information), the formation of metallacycle M4 led to a noticeable change in the EEM peaks. Notably, the fluorescence quantum yield (Φ_F) and fluorescence lifetime (τ) of ligand L1 were determined to be 83.2% and 5.69 ns, respectively. Under the same conditions, the Φ_F of metallacycles M1 and M2 was calculated to be 75.6% and 69.5%, respectively, indicating a slight decrease compared to their BODIPY-based precursor L1. Interestingly, the Φ_F and τ of metallacycle M3 were evaluated to be 5.5% and 1.36 ns, respectively, reflecting substantial decreases compared to ligand L1. For metallacycle M4, the fluorescence emission was completely quenched, and the τ decreased to approximately 4.17 ns (Figs. 3c and d). Furthermore, both the apparent color and the fluorescence of ligand L1 and metallacycles M1–M4 also exhibited significant changes (Figs. 3e and f). These changes may be attributed to the enhanced ISC process arising from the heavy-atom effect associated with the formation of Pt-N bonds within the metallacycle scaffold.

  Figure 3

  

  Figure 3.  Absorption spectra (a) and emission spectra (b) of metallacycles M1–M4 (5 µmol/L) and ligand L1 (15 µmol/L) in dichloromethane. Comparison of quantum yields (c) and fluorescence lifetimes (d) of metallacycles M1–M4 (5 µmol/L) and ligand L1 (15 µmol/L) in dichloromethane. Photographs of L1 (15 µmol/L) and metallacycles M1–M4 (5 µmol/L) in natural light (e) and 365 nm (f) in dichloromethane.

  DownLoad: Full-Size Img PowerPoint

  Table 1

  

  Table 1.  The photophysical parameters of metallacycles M1–M4 and ligand L1 in dichloromethane.^a

  DownLoad: CSV

  Compound	λabs (nm)	λem (nm)	ΦF (%)	τ (ns)
  L1	517	549	83.2	5.69
  M1	515	560	75.6	5.71
  M2	517	560	69.5	4.91
  M3	518	555	5.5	1.36
  M4	518	556	< 1	4.17
  a The photophysical parameters of L1 were collected at a concentration of 15 µmol/L, while those for M1– M4 were collected at a concentration of 5 µmol/L, using slit settings of (5, 5).

  BODIPY-based fluorophores show significant potential as PSs for generating ROS, particularly ¹O₂, through a photosensitization process. This capability is crucial for various applications, including PDT and photooxidation [55-59]. The incorporation of multiple BODIPY units into the scaffold of Pt(Ⅱ)-based metallacycles provides an excellent platform for ¹O₂ generation. Therefore, we examined the ¹O₂ generation efficiency of the obtained metallacycles M1–M4 and their BODIPY-based precursor L1 in dichloromethane, using the commercially available ¹O₂ indicator DMA. DMA reacts with the generated ¹O₂ to form nonfluorescent endoperoxide, leading to fluorescence quenching in the range of 400–500 nm. As shown in Figs. 4a and b and Fig. S4 (Supporting information), upon irradiation with white light (6 W) for 0–300 s, the fluorescence emission of DMA exhibited a rapid decrease, indicating the successful generation of ¹O₂ by these metallacycle-based PSs. By plotting Ln(A₀/A_t) against irradiation time, we quantitatively calculated the ¹O₂ generation efficiency rates. Based on the data fitting results (Figs. 4c and d, Table S1 in Supporting information), the ¹O₂ generation rates for metallacycles M1–M4 and ligand L1 were 0.28, 0.6, 2.76, 0.21, and 0.32 min⁻¹, respectively. Notably, metallacycles M1 and M4 exhibited similar ¹O₂ generation efficiencies compared to precursor L1 under the same concentration of BODIPY. Metallacycle M2 was approximately twice as efficient as ligand L1 in generating ¹O₂. Surprisingly, metallacycle M3 demonstrated the highest ¹O₂ generation efficiency, about 8.6 times greater than that of ligand L1. These results suggest that integrating BODIPY-containing PSs into Pt(Ⅱ)-based metallacycles allows for straightforward tuning of photosensitization efficiency by modifying pre-designed Pt(Ⅱ) acceptors through a coordination-driven self-assembly strategy, making it a strong candidate for diverse applications.

  Figure 4

  

  Figure 4.  Emission spectra of DMA in dichloromethane in the presence of (a) L1 (3 µmol/L) and (b) M3 (1 µmol/L). (c) The decomposition of DMA by L1, M1-M4. (d) The histogram of ¹O₂ generation rate of L1, M1–M4.

  DownLoad: Full-Size Img PowerPoint

  To gain a deeper understanding of the distinctly different photophysical properties related to ¹O₂ generation efficiency between ligand L1 and metallacycles M1–M4, density functional theory (DFT) calculations were employed to investigate the underlying mechanisms [60-62]. Due to the large number of atoms in these constructed metallacycles, we extracted coordination fragments as the initial model structures for our calculations rather than using the complete molecules. According to perturbation theory, both the ΔE_ST between singlet and triplet states and the SOC must be considered when researching ISC, as a reduction in ΔE_ST or an increase in SOC enhances the ISC rate constant (k_ISC). As shown in Fig. S39 (Supporting information), metallacycles M3 and M4 exhibited much lower ΔE_ST values compared to metallacycles M1 and M2, indicating that the introduction of di-Pt(Ⅱ) acceptors with larger organic π-conjugated moieties into the metallacycle skeleton can significantly improve the photosensitization process. Subsequently, we further investigated the SOC for metallacycles M1–M4, which plays a crucial role in the ISC process. The calculated SOC values for metallacycle M3 between S₁ and T₁/T₂/T₃ were greatly enhanced compared to those of the other metallacycles, which may be a key factor contributing to its highest photosensitization and ¹O₂ generation efficiency (Fig. 5a). Although metallacycle M4 has a small ΔE_ST, its calculated SOC values between S₁ and T₁/T₂ are relatively low, resulting in the lowest photosensitization efficiency. In addition, we found that, for M3 and M4, the excitation is primarily localized on the BODIPY group, whereas for M1 and M2, the excitation involves the aromatic rings, which leads to a decrease in the energy levels (Fig. S40 in Supporting information). Collectively, the simple introduction of different Pt(Ⅱ) acceptors to construct BODIPY-based metallacycles resulted in tunable photophysical properties related to the photosensitization process.

  Figure 5

  

  Figure 5.  (a) Working mechanism of PSs M3. (b) Generation of ¹O₂ and the photooxidation reaction of M3. (c) Visible-light-driven oxidative coupling of various amines to imines catalyzed by M3 (reaction conditions: amine (0.1 mmol), catalyst M3 (2.0 mol%), MeOD/CD₂Cl₂ (800 µL, v/v = 2/1), air, white light (6 W). Conversion rate was determined in situ by ¹H NMR).

  DownLoad: Full-Size Img PowerPoint

  Since metallacycle M3 exhibits satisfactory photosensitization efficiency and an extremely high ¹O₂ generation rate of 2.76 min⁻¹ in dichloromethane, we conducted a visible-light-driven oxidative coupling of various amines to imines, catalyzed by metallacycle M3, through a photosensitization process involving the irradiation of a mixture of M3 and molecular oxygen (Fig. 5b). The progress of the oxidation reaction was monitored using ¹H NMR spectroscopy (Figs. S5–S9 in Supporting information). As shown in Fig. 5c, upon white light (6 W) irradiation for over 20 h, five different amines were converted to imines with outstanding conversion rates (exceeding 90%) via a chemical oxidation catalyzed by metallacycle M3 (2.0 mol% relative to amines) as the PS. The high photooxidation activity of metallacycle M3 can be attributed to its enhanced photosensitization process and excellent ¹O₂ generation efficiency.

  In summary, we have developed an efficient and straightforward approach to tune the photosensitization process of PSs based on Pt(Ⅱ) SCCs by modifying pre-designed Pt(Ⅱ) acceptors. We designed and synthesized a family of triangular metallacycles M1–M4, each decorated with three BODIPY units and various Pt(Ⅱ) acceptors, using a coordination-driven self-assembly strategy. The resulting metallacycles M1–M4 exhibited significantly different photophysical properties, particularly in photosensitization efficiency, compared to their BODIPY-based building block L1. Notably, metallacycle M3 demonstrated approximately 8.6-fold higher ¹O₂ generation efficiency than its corresponding precursor L1. This tunable photosensitization efficiency in the constructed metallacycles is attributed to the incorporation of different Pt(Ⅱ) acceptors with varying Pt atoms and degrees of π-conjugated organic moieties into the BODIPY-containing metallacycle scaffold. This research not only presents a unique method for fabricating efficient PSs with tunable photosensitization efficiency and excellent photocatalytic reactivity via a supramolecular coordination strategy but also significantly enhances the application of PSs in catalyzing chemical oxidation to mimic photosynthetic chemistry.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Pei-Pei Jia: Writing – original draft, Investigation, Data curation, Conceptualization. Yi-Xiong Hu: Writing – original draft, Conceptualization. Zhen-Chen Lou: Methodology. Xuelei Ji: Conceptualization. Xing-Dong Xu: Methodology, Conceptualization. Haitao Sun: Methodology. Tongxia Jin: Methodology. Feng Zheng: Methodology, Conceptualization. Lin Xu: Writing – review & editing, Methodology, Conceptualization.

  Acknowledgments

  The authors acknowledge the financial support by the National Natural Science Foundation of China (Nos. 22301081, 22301269 and 22401096), China Postdoctoral Science Foundation (No. 2023M731095), the Shanghai Frontiers Science Center for Molecular Intelligent Syntheses, and the Fundamental Research Funds for the Central Universities, Young Talent Fund of Association for Science and Technology in Shaanxi, China (No. 20240628), Scientific Research Program Funded by Education Department of Shaanxi Provincial Government (No. 23JK0752), Foundation of Yulin Association for Science and Technology (No. 20230512).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110835.

  
  
• 1. [1]

     R. Banerjee, D. Chakraborty, W.T. Jhang, Y.T. Chan, P.S. Mukherjee, Angew. Chem. Int. Ed. 62 (2023) e202305338. doi: 10.1002/anie.202305338

  2. [2]

     M. Lu, S.B. Zhang, M.Y. Yang, et al., Angew. Chem. Int. Ed. 62 (2023) e202307632. doi: 10.1002/anie.202307632

  3. [3]

     L.J. Chen, S. Chen, Y. Qin, et al., J. Am. Chem. Soc. 140 (2018) 5049–5052. doi: 10.1021/jacs.8b02386

  4. [4]

     D. Zhang, W. Yu, S. Li, et al., J. Am. Chem. Soc. 143 (2021) 1313–1317. doi: 10.1021/jacs.0c12522

  5. [5]

     F. Mei, H. Lin, et al., Smart Mol. 1 (2023) e20220001. doi: 10.1002/smo.20220001

  6. [6]

     Q. Liu, M. Zuo, K. Wang, X.Y. Hu, et al., Chem. Commun. 59 (2023) 13707–13710. doi: 10.1039/D3CC04040B

  7. [7]

     M. Zuo, W. Qian, M. Hao, et al., Chin. Chem. Lett. 32 (2021) 1381–1384. doi: 10.1016/j.cclet.2020.09.033

  8. [8]

     Y. Qin, X. Chen, Y. Gui, et al., J. Am. Chem. Soc. 144 (2022) 12825–12833. doi: 10.1021/jacs.2c03895

  9. [9]

     K.X. Teng, D. Zhang, B.K. Liu, Angew. Chem. Int. Ed. 63 (2024) e202318783. doi: 10.1002/anie.202318783

  10. [10]

      Y. Li, F. Huang, P.J. Stang, S. Yin, Acc. Chem. Res. 57 (2024) 1174–1187. doi: 10.1021/acs.accounts.4c00031

  11. [11]

      V.N. Nguyen, Y. Yan, J. Zhao, J. Yoon, Acc. Chem. Res. 54 (2021) 207–220. doi: 10.1021/acs.accounts.0c00606

  12. [12]

      T.C. Pham, V.N. Nguyen, Y. Choi, et al., Chem. Rev. 121 (2021) 13454–13619. doi: 10.1021/acs.chemrev.1c00381

  13. [13]

      C. Li, P.P. Jia, Y.L. Xu, et al., Sci. China Chem. 64 (2021) 134–142. doi: 10.1007/s11426-020-9856-7

  14. [14]

      R. Gueret, L. Poulard, M. Oshinowo, et al., ACS Catal. 8 (2018) 3792–3802. doi: 10.1021/acscatal.7b04000

  15. [15]

      Y.S. Chen, P.V. Kamat, J. Am. Chem. Soc. 136 (2014) 6075–6082. doi: 10.1021/ja5017365

  16. [16]

      C. Fan, W. Wu, J.J. Chruma, et al., J. Am. Chem. Soc. 138 (2016) 15405–15412. doi: 10.1021/jacs.6b07946

  17. [17]

      X. Zheng, R.K. Kankala, C.G. Liu, et al., Coord. Chem. Rev. 438 (2021) 213870. doi: 10.1016/j.ccr.2021.213870

  18. [18]

      M. Xue, C. Cao, X. Zhou, et al., Inorg. Chem. 58 (2019) 14490–14497. doi: 10.1021/acs.inorgchem.9b02081

  19. [19]

      T.J. Penfold, E. Gindensperger, C. Daniel, C.M. Marian, Chem. Rev. 118 (2018) 6975–7025. doi: 10.1021/acs.chemrev.7b00617

  20. [20]

      X. Ma, J. Wang, H. Tian, Acc. Chem. Res. 52 (2019) 738–748. doi: 10.1021/acs.accounts.8b00620

  21. [21]

      M.S. Kwon, Y. Yu, C. Coburn, et al., Nat. Commun. 6 (2015) 8947. doi: 10.1038/ncomms9947

  22. [22]

      X. Li, S. Lee, J. Yoon, Chem. Soc. Rev. 47 (2018) 1174–1188. doi: 10.1039/C7CS00594F

  23. [23]

      B. Yang, Y. Chen, J. Shi, Chem. Rev. 119 (2019) 4881–4985. doi: 10.1021/acs.chemrev.8b00626

  24. [24]

      J. Zhao, K. Xu, W. Yang, et al., Chem. Soc. Rev. 44 (2015) 8904–8939. doi: 10.1039/C5CS00364D

  25. [25]

      W. Wu, X. Shao, J. Zhao, M. Wu, Adv. Sci. 4 (2017) 1700113. doi: 10.1002/advs.201700113

  26. [26]

      W.J. Li, Z. Hu, L. Xu, et al., J. Am. Chem. Soc. 142 (2020) 16748–16756. doi: 10.1021/jacs.0c07292

  27. [27]

      K.X. Teng, L.Y. Niu, N. Xie, Q.Z. Yang, Nat. Commun. 13 (2022) 6179. doi: 10.1038/s41467-022-33924-3

  28. [28]

      P.P. Jia, L. Xu, Y.X. Hu, et al., J. Am. Chem. Soc. 143 (2021) 399–408. doi: 10.1021/jacs.0c11370

  29. [29]

      J. Karges, U. Basu, O. Blacque, H. Chao, G. Gasser, Angew. Chem. Int. Ed. 58 (2019) 14334–14340. doi: 10.1002/anie.201907856

  30. [30]

      R.Z. Dong, X.H. Shi, H. Liu, et al., J. Colloid Interf. Sci. 665 (2024) 871–878. doi: 10.1016/j.jcis.2024.03.197

  31. [31]

      K. Gao, Y. Cheng, Z. Zhang, et al., Angew. Chem. Int. Ed. 63 (2024) e202319488. doi: 10.1002/anie.202319488

  32. [32]

      C. Guo, F. Su, P. Su, X. Yu, X. Li, Sci. China Chem. 66 (2023) 1940–1962. doi: 10.1007/s11426-023-1601-0

  33. [33]

      D. Bokotial, K. Acharyya, A. Chowdhury, P.S. Mukherjee, Angew. Chem. Int. Ed. 63 (2024) e202401136. doi: 10.1002/anie.202401136

  34. [34]

      Y. Jiang, Z. Zeng, J. Yao, et al., Chin. Chem. Lett. 34 (2023) 107576. doi: 10.1016/j.cclet.2022.05.090

  35. [35]

      K.E. Ebbert, E. Benchimol, A. Platzek, Angew. Chem. Int. Ed. 63 (2024) e202413323. doi: 10.1002/anie.202413323

  36. [36]

      T. Jin, K. Zeng, X. Zhang, et al., Angew. Chem. Int. Ed. 63 (2024) e202409878. doi: 10.1002/anie.202409878

  37. [37]

      L. Zhou, C. Yang, W. Dou, et al., Chin. Chem. Lett. 35 (2024) 108669. doi: 10.1016/j.cclet.2023.108669

  38. [38]

      Y. Sun, C. Chen, J. Liu, P. J Stang, Chem. Soc. Rev. 49 (2020) 3889–3919. doi: 10.1039/D0CS00038H

  39. [39]

      J. Gemen, Michał J. Białek, M. Kazes, Chem 8 (2022) 2362–2379. doi: 10.1016/j.chempr.2022.05.008

  40. [40]

      D. Zhang, M. Li, B. Jiang, et al., J. Colloid Interface Sci. 652 (2023) 1494–1502. doi: 10.1016/j.jcis.2023.08.184

  41. [41]

      H.Y. Lin, L.Y. Zhou, F. Mei, et al., Angew. Chem. Int. Ed. 62 (2023) e202301900. doi: 10.1002/anie.202301900

  42. [42]

      H. Duan, T. Yang, Q. Li, et al., Chin. Chem. Lett. 35 (2024) 108878. doi: 10.1016/j.cclet.2023.108878

  43. [43]

      Q. Ling, T. Cheng, S. Tan, J. Huang, L. Xu, Chin. Chem. Lett. 31 (2020) 2884–2890. doi: 10.1016/j.cclet.2020.08.020

  44. [44]

      Q. Feng, N. Li, Z. Zhang, et al., Chin. Chem. Lett. 34 (2023) 108439. doi: 10.1016/j.cclet.2023.108439

  45. [45]

      Y.X. Hu, W.J. Li, P.P. Jia, et al., Adv. Optical Mater. 8 (2020) 2000265. doi: 10.1002/adom.202000265

  46. [46]

      Y. Wang, Y. Qin, X. Zhao, et al., Chin. Chem. Lett. 34 (2023) 107966. doi: 10.1016/j.cclet.2022.107966

  47. [47]

      H. Dobashi, L. Catti, Y. Tanaka, M. Akita, M. Yoshizawa, Angew. Chem. Int. Ed. 59 (2020) 11881–11885. doi: 10.1002/anie.202004168

  48. [48]

      Y. Li, Y. Dong, L. Cheng, et al., J. Am. Chem. Soc. 141 (2019) 8412–8415. doi: 10.1021/jacs.9b02617

  49. [49]

      Z. Guo, G. Li, H. Wang, et al., J. Am. Chem. Soc. 143 (2021) 9215–9221. doi: 10.1021/jacs.1c04288

  50. [50]

      P.K. Maitra, S. Bhattacharyya, N. Hickey, P.S. Mukherjee, J. Am. Chem. Soc. 146 (2024) 15301–15308. doi: 10.1021/jacs.4c02956

  51. [51]

      P. Jia, Y. Hu, Z. Zeng, et al., Chin. Chem. Lett. 34 (2023) 107511. doi: 10.1016/j.cclet.2022.05.025

  52. [52]

      C. Bannwarth, S. Ehlert, S. Grimme, J. Chem. Theory Comput. 15 (2019) 1652–1671. doi: 10.1021/acs.jctc.8b01176

  53. [53]

      M. Bursch, H. Neugebauer, S. Grimme, Angew. Chem. Int. Ed. 58 (2019) 11078–11087. doi: 10.1002/anie.201904021

  54. [54]

      W. Humphrey, A. Dalke, K. Schulten, J. Mol. Graph. 14 (1996) 33–38. doi: 10.1016/0263-7855(96)00018-5

  55. [55]

      Y. Qin, X. Liu, P.P. Jia, L. Xu, H.B. Yang, Chem. Soc. Rev. 49 (2020) 5678–5703. doi: 10.1039/C9CS00797K

  56. [56]

      D. Zhang, M.C. Xiong, L.Y. Niu, Q.Z. Yang, Chem. Commun. 58 (2022) 10060–10063. doi: 10.1039/D2CC04191J

  57. [57]

      Y. Hou, Q. Liu, J. Zhao, Chem. Commun. 56 (2020) 1721–1724. doi: 10.1039/C9CC09058D

  58. [58]

      J. Zhou, Y. Zhang, G. Yu, et al., J. Am. Chem. Soc. 140 (2018) 7730–7736. doi: 10.1021/jacs.8b04929

  59. [59]

      X.L. Li, N. Han, R.Z. Zhang, et al., ACS Appl. Mater. Interfaces 15 (2023) 55803–55812. doi: 10.1021/acsami.3c12827

  60. [60]

      F. Neese, WIREs-Comp. Mol. Sci. 12 (2022) e1606. doi: 10.1002/wcms.1606

  61. [61]

      M.J. Frisch, G.W. Trucks, H.B. Schlegel, et al., Gaussian 16, Revision A. 03, Gaussian, Inc., Wallingford CT, 2016.

  62. [62]

      F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 7 (2005) 3297–3305. doi: 10.1039/b508541a

• 

  Figure 1  Schematic illustration of the coordination-driven self-assembly of triangular metallacycles M1–M4 and the chemical structures of building blocks L1–L5.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) The partial ¹H NMR spectra (500 MHz, CD₂Cl₂, 298 K) of the ligand L4 (up), the self-assembled metallacycle M3 (middle) and ligand L1 (bottom). (b) The ³¹P NMR spectra (202 MHz, CD₂Cl₂, 298 K) of metallacycle M3 (up) and ligand L4 (bottom). (c) Theoretical (top) and experimental (bottom) ESI-TOF-MS of metallacycle M3.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Absorption spectra (a) and emission spectra (b) of metallacycles M1–M4 (5 µmol/L) and ligand L1 (15 µmol/L) in dichloromethane. Comparison of quantum yields (c) and fluorescence lifetimes (d) of metallacycles M1–M4 (5 µmol/L) and ligand L1 (15 µmol/L) in dichloromethane. Photographs of L1 (15 µmol/L) and metallacycles M1–M4 (5 µmol/L) in natural light (e) and 365 nm (f) in dichloromethane.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Emission spectra of DMA in dichloromethane in the presence of (a) L1 (3 µmol/L) and (b) M3 (1 µmol/L). (c) The decomposition of DMA by L1, M1-M4. (d) The histogram of ¹O₂ generation rate of L1, M1–M4.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) Working mechanism of PSs M3. (b) Generation of ¹O₂ and the photooxidation reaction of M3. (c) Visible-light-driven oxidative coupling of various amines to imines catalyzed by M3 (reaction conditions: amine (0.1 mmol), catalyst M3 (2.0 mol%), MeOD/CD₂Cl₂ (800 µL, v/v = 2/1), air, white light (6 W). Conversion rate was determined in situ by ¹H NMR).

  下载: 全尺寸图片 幻灯片

  Table 1.  The photophysical parameters of metallacycles M1–M4 and ligand L1 in dichloromethane.^a

  Compound	λabs (nm)	λem (nm)	ΦF (%)	τ (ns)
  L1	517	549	83.2	5.69
  M1	515	560	75.6	5.71
  M2	517	560	69.5	4.91
  M3	518	555	5.5	1.36
  M4	518	556	< 1	4.17
  a The photophysical parameters of L1 were collected at a concentration of 15 µmol/L, while those for M1– M4 were collected at a concentration of 5 µmol/L, using slit settings of (5, 5).

  下载: 导出CSV
•