English

• Fluorescent materials have been extensively studied over the past decades because of their wide applications in bioimaging [1, 2], chemosensing [3, 4], optoelectronics [5, 6], etc. In particular, fluorescence-based sensors have become one of the leading sensing technologies due to their fast response, easy visualization, and high sensitivity and selectivity [7, 8]. As a result, the development of novel fluorescence-based sensors has received ever-growing interests. Conventional small molecule fluorophores, conjugated organic/inorganic polymers and dendrimers have been widely developed so far with a special emphasis on fluorescent sensing purposes [9, 10]. Specifically, conjugated polymers (CPs) are composed of backbones possessing π-conjugated subunits, and thus benefit from electron delocalization. This endows CPs high sensitivity and selectivity toward fluorescent sensing [11-13]. Nevertheless, the preparation of CPs typically involves several synthetic steps and exhibits relatively inferior molecular organization, which greatly impedes their multitudinous applications. Substantial efforts have been made in addressing such issues, within which the coordination-driven self-assembly method outstands. Unlike the stepwise covalent synthetic protocols, this strategy is facile yet highly efficient due to its synthetic advantages, including high degree of predictability, inherent self-correction, fewer steps and traceability [14-19].

  Metallacycles, as an important type of metallosupramolecules, are constructed by coordination-driven self-assembly, and their topology and physicochemical properties can be finely tuned by judicious choices of building units with variable numbers, locations, relative orientations, and additional functional moieties [20-27]. For instance, perylene diimide (PDI) derivatives [28-34], as archetypal fluorophores with excellent optical and electrochemical properties, have been utilized as large rigid and planar scaffolds to construct luminescent metallacycles and metallacages. Such assemblies not only inherit the photophysical properties of their building fluorophore moieties but also exhibit novel features that are not observable for the single components, such as tunable emission wavelengths, markedly enhanced fluorescence efficiencies, and selective response toward specific molecules. Although significant progress has been made on the design and synthesis of emissive metallacycles, a universal strategy for constructing novel emissive metallacycles is highly desirable but remains absent.

  Herein, we develop a strategy to build a series of metallacycles 4a–4f by multicomponent coordination-driven self-assembly. Notably, the methodology relies mainly on the stoichiometry of the individual building units, as well as the geometry and length of the ligands. As shown in Scheme 1, these metallacycles include three parts: PDI-based tetrapyridyl ligand (1), dicarboxylic ligands (2a–2f) and cis-Pt(PEt₃)₂(OTf)₂ (3). Interestingly, these metallacycles show symmetrical, bowtie-like shapes in solid-state and bright emission in acetonitrile with fluorescence quantum yields (Φ_F) exceeding 98% for metallacycles 4a, 4c–4f. Moreover, these metallacycles display fluorescence decays in response to variable concentrations of picric acid (PA) with a low detection limit of 2.8 × 10⁻⁶ mol/L (S/N = 3), comparable to the reported values from literature. This study describes a versatile strategy for the design and preparation of highly emissive bowtie-shaped metallacycles as fluorescence-based sensors, which will certainly promote the development of metallacycles for chemosensing applications.

  Scheme 1

  

  Scheme 1.  Cartoon representations of the self-assembly of bowtie-shaped metallacycles 4a–4f.

  DownLoad: Full-Size Img PowerPoint

  The formation of bowtie-shaped metallacycles 4a–4f was confirmed by multiple techniques including ³¹P{¹H}, ¹H NMR and electrospray ionization time-of-flight mass spectrometry (ESI-TOF-MS). As shown in Figs. 1a–f, the ³¹P{¹H} NMR spectra of 4a–4f split into two doublet peaks at 5.13 ppm and −0.47 ppm for 4a, 5.38 ppm and −0.42 ppm for 4b, 5.40 ppm and −0.14 ppm for 4c, 5.18 ppm and −0.35 ppm for 4d, 4.95 ppm and −0.76 ppm for 4e, and 5.51 ppm and −0.64 ppm for 4f, respectively. These two doublet peaks nearly share equal intensities with concomitant ¹⁹⁵Pt satellites, corresponding to different phosphorus environments. This result clearly indicates the formation of discrete symmetric metallacycles. The α-pyridyl protons H_a and β-pyridyl protons H_b (Figs. 1g–m) of ligand 1 in metallacycles 4a–f were noticed, showing obvious downfield chemical shifts compared to free ligand 1. Investigation of the ESI-TOF-MS provides further support for the stoichiometry of 4a–4f (Figs. S32, S36, S40, S44, S48, S52 in Supporting information). Accordingly, the observed peaks are consistent with the calculated ones. For instance, peaks at m/z 998.8642, 1024.8661, 1025.5089, 1036.1442, 1032.2203 and 1032.2203 were found, corresponding to [4a–3OTf]³⁺, [4b–3OTf]³⁺, [4c–3OTf]³⁺, [4d–3OTf]³⁺, [4e–3OTf]³⁺ and [4f–3OTf]³⁺, respectively. All these results evidenced the successful formation of the bowtie-shaped metallacycles 4a–4f.

  Figure 1

  

  Figure 1.  Partial ³¹P{¹H} NMR spectra (121.4 MHz, CD₃CN, 295 K) of 4a (a), 4b (b), 4c (c), 4d (d), 4e (e) and 4f (f). Partial ¹H NMR spectra (400 MHz, CD₃CN, 295 K) of 1 (g), 4a (h), 4b (i), 4c (j), 4d (k), 4e (l) and 4f (m).

  DownLoad: Full-Size Img PowerPoint

  Dark red single crystals 4a–4f suitable for X-ray diffraction analysis were successfully obtained by slow evaporation of dioxane into the DMF solution (vapors of dioxane are diffusing into DMF) of metallacycles. The crystal structures of metallacycles 4a–4f are shown in Figs. 2a–f, which unambiguously confirms their two-dimensional bowtie-shaped structures. X-ray crystallographic analyses suggest that these metallacycles share similar structures in which the pyridyl groups and carboxylic groups are linked by four platinum atoms, leading to the formation of a [1 + 2 + 4] structure. Based on the distances between the platinum atoms, the length and width of metallacycles 4a–4f are 1.68 × 0.97 nm, 1.66 × 1.07 nm, 1.64 × 1.06 nm, and 1.64 × 1.06 nm, 1.67 × 1.07 nm and 1.63 × 1.11 nm (Figs. 2a–f), respectively. Metallacycles 4a–4c, 4e and 4f exhibit the same packing mode (Fig. S1 in Supporting information), with the metallacycle units aligned linearly via intermolecular interactions. Dissimilarly, the 4d metallacycles reassemble into parallelepipeds, where two adjacent metallacycles are connected by two C—H···O interactions (d_{H···O plane} = 2.61 Å) between PDI and CF₃SO₃^– anions and four F···π interactions (d_{F···π plane} = 3.11–3.14 Å) between CF₃SO₃^– anions and PDI. Moreover, the neighboring metallacycle 4d belonging to different parallel pipelines is connected by C—H···O (d_{H···O plane} = 2.39–2.62 Å). This observation is in accordance with that of previous studies [35, 36]. In such a manner, the packing mode in 4d was finally formed and different from other metallacycles, which may be rationalized by their different growth environments.

  Figure 2

  

  Figure 2.  Crystal structures of 4a (a), 4b (b), 4c (c), 4d (d), 4e (e) and 4f (f). Hydrogen atoms, counterions, solvent molecules and triethylphosphine units are omitted for clarity.

  DownLoad: Full-Size Img PowerPoint

  The UV-vis absorption and fluorescence emission spectra of ligand 1 and metallacycles 4a–4f are depicted Table 1, Fig. S52 and Table S3 (Supporting information). The UV-vis absorption spectra for both ligand 1 and metallacycles 4a–4f exhibit three absorption bands centered at 456 nm, 487 nm, and 522 nm, respectively. Metallacycles 4a–4f display absorption bands similar to those of ligand 1 because of the weak absorption of carboxylic ligands 2a–2f compared with PDI. Similarly, the fluorescence emission of all the metallacycles exhibit similar curves, with an intense peak centered at 540 nm. Noticeably, the fluorescence intensities of metallacycles 4a, 4c–4f are all higher than that of ligand 1, whereas the intensity of metallacycle 4b is much lower than that of ligand 1 under identical conditions. The fluorescence quantum yields of metallacycles 4a, 4c–4f in CH₃CN reached 99%, 99%, 98%, 99% and 99% (Table 1 and Figs. S2–S9 in Supporting information), respectively. However, owing to the introduction of an electron-donating group (-NH-) into metallacycle 4b, its Φ_F value is only 32%, because the photoinduced electron transfer (PET) from the NH groups to the fluorophore offers a non-radiative pathway to quench the emission [37].

  Table 1

  

  Table 1.  UV-vis absorption and fluorescence emission data.^a

  DownLoad: CSV

  Nitroaromatic compounds are widely used in firework manufacturing, chemical industry, leather, pharmaceutical, and dye industries. Leakage of nitroaromatic compounds during production and transportation not only pollutes the groundwater and soil environments, but also poses great threats to human beings. Therefore, the rapid detection of these compounds has always been an important task [38].

  To explore the potential application of the metallacycles in the detection of nitroaromatic compounds, we chose picric acid as a model compound to perform fluorescence titration tests. As shown in Fig. 3a, with the addition of picric acid, the emission intensity of metallacycle 4a gradually decreased. Once the addition of picric acid reached a ratio of 13 equiv., its emission was almost completely quenched. Correspondingly, the quenching constant was determined to be 5.2 × 10⁴ L/mol by nonlinear fitting (Fig. 3b), and the detection limit was 2.8 × 10⁻⁶ mol/L (S/N = 3), comparable with the reported values from literature [39-42]. This quenching process can be directly noticed by naked eyes. Similar quenching phenomena were also observed for other bowtie-shaped metallacycles 4b–4f (Figs. S53–S57 and Table S4 in Supporting information), indicating their potential use as chemical sensors for the detection of nitroaromatic compounds.

  Figure 3

  

  Figure 3.  Fluorescence spectra of 4a (a) with increased amounts of picric acid (c = 10 µmol/L in metallacycle concentration, λ_ex = 365 nm). Stern-Volmer plot I₀/I versus the concentration of picric acid (b). The inset denotes the fluorescent photographs before and after the addition of picric acid.

  DownLoad: Full-Size Img PowerPoint

  In summary, a series of PDI-based bowtie-shaped metallacycles were successfully constructed by coordination-driven multicomponent self-assembly. Their structures were well resolved by single crystal X-ray diffraction analysis. Moreover, most metallacycles showed remarkable fluorescence quantum yields in CH₃CN and all metallacycles experienced fluorescence quenching upon the addition of variable concentrations of picric acid. The detection limit can reach as low as 2.8 × 10⁻⁶ mol/L. These findings clearly indicate that such metallacycles can be used as fluorescent sensors for the detection of nitroaromatic compounds. This work justifies our strategy to systematically prepare metallacycles with tailor-made photophysical and chemical properties, which will promote the development of functional metallacycles and their applications in various scenarios.

  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.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 22171219), and the Fundamental Research Funds for the Central Universities (No. xzy022021004). We thank Dr. Gang Chang at the Instrument Analysis Center and Dr. Aqun Zheng and Junjie Zhang at the Experimental Chemistry Center of Xi'an Jiaotong University for NMR and fluorescence measurements.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.07.031.

  
  
• 1. [1]

     J.T. Hou, W.X. Ren, J.S. Kim, et al., Chem. Soc. Rev. 46 (2017) 2076–2090. doi: 10.1039/C6CS00719H

  2. [2]

     M. Collot, Mater. Horiz. 8 (2021) 501–514. doi: 10.1039/D0MH01186J

  3. [3]

     S. Shanmugaraju, P.S. Mukherjee, Chem. Commun. 51 (2015) 16014–16032. doi: 10.1039/C5CC07513K

  4. [4]

     T.L. Mako, J.M. Racicot, M. Levine, Chem. Rev. 119 (2019) 322–477. doi: 10.1021/acs.chemrev.8b00260

  5. [5]

     V.W. Yam, V.K. Au, S.Y. Leung, Chem. Rev. 115 (2015) 7589–7728. doi: 10.1021/acs.chemrev.5b00074

  6. [6]

     M.P. Duffy, W. Delaunay, P.A. Bouit, M. Hissler, Chem. Soc. Rev. 45 (2016) 5296–5310. doi: 10.1039/C6CS00257A

  7. [7]

     M. Jiang, X. Gu, B.Z. Tang, et al., Chem. Sci. 8 (2017) 5440–5446. doi: 10.1039/C7SC01400G

  8. [8]

     S. Cui, Y. Wu, J. Wang, et al., Chin. Chem. Lett. 31 (2020) 487–493. doi: 10.1016/j.cclet.2019.04.014

  9. [9]

     O.S. Wolfbeis, Chem. Soc. Rev. 44 (2015) 4743–4768. doi: 10.1039/C4CS00392F

  10. [10]

      M. Gao, B.Z. Tang, ACS Sens. 2 (2017) 1382–1399. doi: 10.1021/acssensors.7b00551

  11. [11]

      T. Repenko, A. Rix, A.J.C. Kuehne, et al., Nat. Commun. 8 (2017) 470. doi: 10.1038/s41467-017-00545-0

  12. [12]

      W. Wu, G.C. Bazan, B. Liu, Nat. Chem. 2 (2017) 760–790. doi: 10.1016/j.chempr.2017.05.002

  13. [13]

      T.F. Abelha, C.A. Dreiss, M.A. Green, L.A. Dailey, J. Mater. Chem. B 8 (2020) 592–606. doi: 10.1039/C9TB02582K

  14. [14]

      K. Harris, D. Fujita, M. Fujita, Chem. Commun. 49 (2013) 6703–6712. doi: 10.1039/c3cc43191f

  15. [15]

      L. Xu, Y.X. Wang, L.J. Chen, H.B. Yang, Chem. Soc. Rev. 44 (2015) 2148–2167. doi: 10.1039/C5CS00022J

  16. [16]

      T.R. Cook, P.J. Stang, Chem. Rev. 115 (2015) 7001–7045. doi: 10.1021/cr5005666

  17. [17]

      G.H. Clever, P. Punt, Acc. Chem. Res. 50 (2017) 2233–2243. doi: 10.1021/acs.accounts.7b00231

  18. [18]

      M. Pan, K. Wu, J.H. Zhang, C.Y. Su, Coord. Chem. Rev. 378 (2019) 333–349. doi: 10.1016/j.ccr.2017.10.031

  19. [19]

      S. Tashiro, M. Shionoya, Acc. Chem. Res. 53 (2020) 632–643. doi: 10.1021/acs.accounts.9b00566

  20. [20]

      Y.X. Hu, X. Zhang, L. Xu, H.B. Yang, Isr. J. Chem. 59 (2018) 184–196.

  21. [21]

      G.Y. Wu, L.J. Chen, H.B. Yang, et al., Coord. Chem. Rev. 369 (2018) 39–75. doi: 10.1016/j.ccr.2018.05.009

  22. [22]

      L. Ma, T. Yang, M. Zhang, et al., Chin. Chem. Lett. 30 (2019) 1942–1946. doi: 10.1016/j.cclet.2019.07.043

  23. [23]

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

  24. [24]

      E.G. Percástegui, V. Jancik, Coord. Chem. Rev. 407 (2020) 213165. doi: 10.1016/j.ccr.2019.213165

  25. [25]

      Z. Guo, J. Zhao, X. Yan, et al., Chin. Chem. Lett. 32 (2021) 1691–1695. doi: 10.1016/j.cclet.2020.12.028

  26. [26]

      H. Liu, Z. Zhang, M. Zhang, et al., Angew. Chem. Int. Ed. 61 (2022) e202207289.

  27. [27]

      Z. Zhang, L. Ma, M. Zhang, et al., JACS Au 2 (2022) 1479–1487. doi: 10.1021/jacsau.2c00245

  28. [28]

      S. Chen, P. Slattum, C. Wang, L. Zang, Chem. Rev. 115 (2015) 11967–11998. doi: 10.1021/acs.chemrev.5b00312

  29. [29]

      P.D. Frischmann, V. Kunz, F. Wurthner, Angew. Chem. Int. Ed. 54 (2015) 7285–7289. doi: 10.1002/anie.201501670

  30. [30]

      F. Wurthner, C.R. Saha-Moller, D. Schmidt, et al., Chem. Rev. 116 (2016) 962–1052. doi: 10.1021/acs.chemrev.5b00188

  31. [31]

      Y. Hou, Z. Zhang, M. Zhang, et al., J. Am. Chem. Soc. 142 (2020) 18763–18768. doi: 10.1021/jacs.0c09904

  32. [32]

      Q.H. Ling, J.L. Zhu, Y. Qin, L. Xu, Mater. Chem. Front. 4 (2020) 3176–3189. doi: 10.1039/D0QM00540A

  33. [33]

      X. Chang, S. Lin, P.J. Stang, et al., J. Am. Chem. Soc. 142 (2020) 15950–15960. doi: 10.1021/jacs.0c06623

  34. [34]

      Y. Hou, Z. Zhang, M. Zhang, et al., CCS Chem. 3 (2021) 3153–3160.

  35. [35]

      Z. Yang, Y. Wang, P.J. Stang, et al., J. Am. Chem. Soc. 142 (2020) 13689–13694. doi: 10.1021/jacs.0c06666

  36. [36]

      H. Duan, Y. Li, L. Cao, et al., Angew. Chem. Int. Ed. 59 (2020) 10101–10110. doi: 10.1002/anie.201912730

  37. [37]

      J.L. Zhu, L. Xu, H.B. Yang, et al., Nat. Commun. 10 (2019) 4285. doi: 10.1038/s41467-019-12204-7

  38. [38]

      K.S. Ju, R.E. Parales, Microbiol. Mol. Biol. Rev. 74 (2010) 250–272. doi: 10.1128/MMBR.00006-10

  39. [39]

      Y. Hu, M. Ding, X.Q. Liu, L.B. Sun, H.L. Jiang, Chem. Commun. 52 (2016) 5734–5737. doi: 10.1039/C6CC01597B

  40. [40]

      L. Zhang, Y. Sun, Z. Yao, et al., Chin. Chem. Lett. 31 (2020) 2428–2432. doi: 10.1016/j.cclet.2020.04.003

  41. [41]

      Y. Hou, S. Li, Z. Zhang, L. Chen, M. Zhang, Polym. Chem. 11 (2020) 254–258. doi: 10.1039/C9PY00895K

  42. [42]

      Z.Y. Li, Z.Q. Yao, X.H. Bu, et al., Chin. Chem. Lett. 32 (2021) 3095–3098. doi: 10.1016/j.cclet.2021.03.008

• 

  Scheme 1  Cartoon representations of the self-assembly of bowtie-shaped metallacycles 4a–4f.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  Partial ³¹P{¹H} NMR spectra (121.4 MHz, CD₃CN, 295 K) of 4a (a), 4b (b), 4c (c), 4d (d), 4e (e) and 4f (f). Partial ¹H NMR spectra (400 MHz, CD₃CN, 295 K) of 1 (g), 4a (h), 4b (i), 4c (j), 4d (k), 4e (l) and 4f (m).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Crystal structures of 4a (a), 4b (b), 4c (c), 4d (d), 4e (e) and 4f (f). Hydrogen atoms, counterions, solvent molecules and triethylphosphine units are omitted for clarity.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Fluorescence spectra of 4a (a) with increased amounts of picric acid (c = 10 µmol/L in metallacycle concentration, λ_ex = 365 nm). Stern-Volmer plot I₀/I versus the concentration of picric acid (b). The inset denotes the fluorescent photographs before and after the addition of picric acid.

  下载: 全尺寸图片 幻灯片

  Table 1.  UV-vis absorption and fluorescence emission data.^a

  下载: 导出CSV
•