English

• Metal-organic cages have attracted huge attention because of their well-fined geometries and confined functional cavities, which are widely applied in drug delivery [1-4], catalysis [5-8], guest recognition [9-11] and separation [12, 13], and sensing [14-17]. Ligand exchange strategies are commonly used to synthesize and tune the properties of metal-organic cages [18, 19]. Palladium-based metal-organic cages with labile coordination bond can undergo ligand exchange [20] to achieve cage-to-cage transformations, which have been used to synthesize novel cages [21-25]. Compared with the direct coordination into cages, the cage-to-cage conversions, which are post-synthetic strategies originating from the ligand exchanging at the molecular level, can not only obtain higher-level self-assemblies [26-28], but also mimic the function of bio-systems [29, 30]. In the conversions of these Palladium-based cages, the basicity and geometric structure of the ligands are two key factors that affect the transformation efficiency. The basicity of the ligands on the initial and target cages determines the direction of transformation [31] and structural difference of ligands affects the process of cage-to-cage transformation [32]. When two ligands have large structural differences (especially the length), the initial homoleptic cage could be converted efficiently into another homoleptic cage [31, 33]. When the two ligands are of complementary structure [34-36], only a single of the heteroleptic cage will be obtained, which is described as integrative self-sorting. And when the two ligands have similar structures [37, 38], ligand scrambling would occur―the addition of the new ligand to the initial cage would generally lead to the formation of a series of heteroleptic cages (Fig. 1a). So far as we know, there are only a few reports of exploiting hydrogen bonding interactions and steric effects to avoid the ligand scrambling [35, 39, 40]. Therefore, it is of significant scientific value to develop new and universal strategies to achieve efficient interconversion of cages with similar structures.

  Figure 1

  

  Figure 1.  Cartoon illustration of (a) cage-to-cage transformation with similar structure; (b) step-by-step transformation of cage 2a to 2b by introduction of a mediated ligand Lm.

  DownLoad: Full-Size Img PowerPoint

  Herein, we report the stepwise transformation of a cage (2a) to another one (2b) with similar structure in an extremely high yield by introducing an intermediate ligand (Lm), as illustrated in Fig. 1b. The basicity of Lm is between the initial cage ligand (La) and the target cage ligand (Lb), and its structure is also very different from the two. As a consequence, the intermediate cage (2m) can be obtained by adding Lm to 2a in 96% yield, and then almost quantitatively (97%) transferred to 2b in the presence of excess Lb. By comparison, the direct 2a-to-2b transformation by adding Lb to cage 2a leads to the formation of multiple mixed cages because of ligand scrambling, which results in a low yield (49%) of 2b formation and makes it difficult to be purified. This mediated cage-to-cage transformation provides a universal strategy for effective preparation of molecular cages from their counterparts with weaker ligand basicity.

  The structures of La, Lm and Lb are shown in Fig. 2. Both La and Lb are V-shaped ligands while Lm adopts a quite different configuration with an azobenzene connecting with two meta-pyridines. The introduction of phenyloxy weak electron donating group on the meta position of Lm-pyridine was expected to make the basicity of Lm weaker than Lb (methoxy strong electron donating groups stay on the para position of Lb-pyridine) but stronger than La (no electron donating group on the La-pyridine).

  Figure 2

  

  Figure 2.  Structures of the ligands used in this study.

  DownLoad: Full-Size Img PowerPoint

  La [41] and Lb [31] were prepared as reported and Lm was synthesized by the reduction of 3-(4-nitrophenoxy)pyridine (Scheme S1 in Supporting information). The relative basicity investment was carried out by ¹H NMR trifluoroacetic acid titration experiments. Lm was found to be protonated more easily than La (Figs. S4 and S5 in Supporting information), and Lb was protonated more readily when compared to Lm (Figs. S6 and S7 in Supporting information). The basicity sequence La < Lm < Lb was thus confirmed.

  The known 2a [41] and 2b [31] were then synthesized by the self-assembling of La and Lb with [Pd(CH₃CN)₄](BF₄)₂ respectively. The mediated cage 2m was obtained similarly by stirring Lm and [Pd(CH₃CN)₄](BF₄)₂ in 2:1 ratio in CD₃CN at 70 ℃ for 5 h. 2m was characterized by ¹H, ¹³C, DOSY and ¹H–¹H COSY NMR spectroscopy and high-resolution mass spectrometry. Further X-ray single crystal analysis reveals that 2m crystallizes in monoclinic crystal system with C 2/c space group, and each square-planar Pd²⁺ is coordinated to four N atoms on pyridine of different Lm, generating the lantern shaped M₂L₄ structure (Fig. S13 and Table S1 in Supporting information) [42-44], which is the same pattern as 2a and 2b.

  The direct cage-to-cage conversion was subsequently carried out by mixing 2a with Lb in different molar ratios (1.0: 2.0, 4.0, 7.0) in CD₃CN at room temperature for 24 h (Scheme S4 in Supporting information). The ¹H NMR results (Fig. S14 in Supporting information) indicate that 2a disappears while 2b comes into being after the addition of Lb. Except for those proton signals of 2b and the released free La and Lb, the rest peaks are difficult to be clearly assigned. Nevertheless, further high-resolution mass spectrometry (Fig. 3) shows that the reaction mixture also contains hybrid molecular cages, such as [Pd₂(La)(Lb)₃](BF₄)₄ (abbreviate as H1, m/z at 378.5711 for [H1–4BF₄]⁴⁺, 533.7686 for [H1–3BF₄]³⁺, 844.1517 for [H1–2BF₄]²⁺) and [Pd₂(La)₂(Lb)₂](BF₄)₄ (abbreviate as H2, m/z 363.5635 for [H2–4BF₄]⁴⁺, 513.7659 for [H2–3BF₄]³⁺, 814.1346 for [H2–2BF₄]²⁺). These results indicate that obvious ligand scrambling occurs even though the basicity of Lb is stronger than that of La. Even though 7 equiv. Lb was used, the conversion efficiency was determined as low as 49% according to the NMR integrations (Figs. S15 and S16 in Supporting information). It should be noted that we failed to obtain pure 2b after several purification attempts, such as recrystallization and washing with poor solvents.

  Figure 3

  

  Figure 3.  ESI-MS spectrum of reaction mixture of direct transformation from 2a to 2b by mixing 2a and Lb in a 1.0: 7.0 molar ratio in CH₃CN at room temperature.

  DownLoad: Full-Size Img PowerPoint

  The step-by-step conversions from cage 2a to 2m and then to 2b (Scheme S3 in Supporting information) were then conducted. The first cage conversion process was investigated by treating 2a with Lm at room temperature for 24 h in CD₃CN (Scheme S5 in Supporting information). As shown in Fig. S17 (Supporting information), a new set of ¹H NMR peaks corresponding to 2m appears with the addition of 2.0 equiv. of Lm, and its intensity increases with the further addition of Lm, while the peaks of 2a gradually disappear. The ¹H NMR results suggest that 2a was completely transferred into 2m after the addition of 7.0 equiv. of Lm (Fig. 4a). This conversion process was further confirmed by ESI-MS spectrometry of the mixtures of 2a and Lm. Both 2a and 2m peaks occur when the molar ratio of Lm to 2a is 2.0 (Fig. S19 in Supporting information), while those 2a ones disappear when the ratio come to 7.0 (Fig. S20 in Supporting information). The additional free La and Lm were removed by the simply addition of ethyl ether to the clear solution and 2m precipitated out. Further preparative experiment was conducted in CH₃CN, and pure 2m was obtained from ethyl ether in a high yield of 96% (Scheme S5 in Supporting information).

  Figure 4

  

  Figure 4.  ¹H NMR spectra (400 MHz, CD₃CN) of mixtures after stirring at room temperature for 24 h of (a) 2a and Lm in a 1.0: 7.0 molar ratio, (b) 2m and Lb in a 1.0: 7.0 molar ratio.

  DownLoad: Full-Size Img PowerPoint

  The transformation of cage 2m to 2b was finally carried out. Similarly, the ¹H NMR spectra were recorded after the progressive addition of Lb (2.0, 4.0 and 7.0 equiv.) to the CD₃CN solution of 2m at room temperature for 24 h (Scheme S6 and Fig. S21 in Supporting information). It can be found that the proton signals of 2b enhance with the addition of Lb, and those of 2m disappear entirely when the molar ratio of Lb to 2m reaches to 7.0 (Fig. 4b). Pure 2b was obtained after being precipitated from ethyl acetate in 97% yield in the preparative experiment (Scheme S6).

  Further experiments of the mixing of cage 2b with different amounts of Lm (2.0, 4.0 and 7.0 equiv.) were conducted to investigate the details of the back-transformation of 2b to 2m. According to the integrations of ¹H NMR spectrometry, the conversions are 7%, 12% and 19% respectively with the increasing addition of Lm (Fig. S23 in Supporting information) and the peaks of 2m, 2b, Lm and Lb all appear in the ESI-MS of the mixture of 2b with 2.0 and 7.0 equiv. of Lm (Figs. S24 and S25 in Supporting information). These results indicate that the cage-to-cage conversion is a dynamic equilibrium process.

  It can be seen that both direct and stepwise cage-to-cage transformations can be performed at room temperature. Compared with the low efficiency (49%) of the former, the latter has one more reaction step, but the yield has been greatly improved to 93% (i.e., 96% × 97%) after simple work-up. In addition, this cage transformation method of introducing a mediated ligand with large structural differences can well solve the problem of ligand scrambling without requiring additional modifications to the initial and final cages.

  In summary, the direct conversion of cage 2a to 2b results in the formation of heteroleptic cages because of serious ligand scrambling between La and Lb, making the conversion process complex. To avoid this ligand scrambling, an intermediate ligand Lm with different shape and basicity from La/ Lb was introduced and 2a was indirectly converted to 2b via the intermediate cage 2m. The latter mediated method possesses the advantage of high yield and simple operation, providing a new idea for preparing molecular cages from those ones with similar structure. This research further supports that the geometry and basicity of ligand are key factors among the cage-to-cage transformation. And furthermore, such a strategy of achieving effective cage-to-cage transformation in a controlled manner has good potential application prospects in the constructions of molecular machines and drug delivery systems.

  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 financially supported by the National Natural Science Foundation of China (No. 21572063), Shanghai Municipal Science and Technology Major Project (No. 2018SHZDZX03) and the Fundamental Research Funds for the Central Universities.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.07.036.

  
  
• 1. [1]

     J.E.M. Lewis, E.L. Gavey, S.A. Cameron, J.D. Crowley, Chem. Sci.3 (2012) 778–784. doi: 10.1039/C2SC00899H

  2. [2]

     N. Ahmad, H.A. Younus, A.H. Chughtai, F. Verpoort, Chem. Soc. Rev. 44 (2015) 9–25. doi: 10.1039/C4CS00222A

  3. [3]

     W. Zhu, J. Guo, Y. Ju, et al., Adv. Mater. 31 (2019) 1806774. doi: 10.1002/adma.201806774

  4. [4]

     Y. Liu, B. Shi, H. Wang, et al., Macromol. Rapid Commun. 39 (2018) 1800655. doi: 10.1002/marc.201800655

  5. [5]

     X. Qiu, Y. Zhang, Y. Zhu, et al., Adv. Mater. 33 (2021) 2001731. doi: 10.1002/adma.202001731

  6. [6]

     V. Marti-Centelles, A.L. Lawrence, P.J. Lusby, J. Am. Chem. Soc. 140 (2018) 2862–2868. doi: 10.1021/jacs.7b12146

  7. [7]

     N. Xu, K. Su, E.S.M. El-Sayed, et al., Chem. Sci. 13 (2022) 3582–3588. doi: 10.1039/D2SC00395C

  8. [8]

     K. Wu, K. Li, S. Chen, et al., Angew. Chem. Int. Ed. 59 (2020) 2639–2643. doi: 10.1002/anie.201913303

  9. [9]

     L. Yang, D. Zhang, H. Yan, et al., Inorg. Chem. 58 (2019) 4067–4070. doi: 10.1021/acs.inorgchem.9b00148

  10. [10]

      T.R. Schulte, J.J. Holstein, G.H. Clever, Angew. Chem. Int. Ed. 58 (2019) 5562–5566. doi: 10.1002/anie.201812926

  11. [11]

      A. Jimenez, R.A. Bilbeisi, T.K. Ronson, et al., Angew. Chem. Int. Ed. 53 (2014) 4556–4560. doi: 10.1002/anie.201400541

  12. [12]

      D. Zhang, T.K. Ronson, Y.Q. Zou, J.R. Nitschke, Nat. Rev. Chem. 5 (2021) 168–182. doi: 10.1038/s41570-020-00246-1

  13. [13]

      W. Fan, S.B. Peh, Z. Zhang, et al., Angew. Chem. Int. Ed. 60 (2021) 17338–17343. doi: 10.1002/anie.202102585

  14. [14]

      J. Dong, Y. Zhou, F. Zhang, Y. Cui, Chem. Eur. J. 20 (2014) 6455–6461. doi: 10.1002/chem.201304606

  15. [15]

      M. Zhang, M.L. Saha, M. Wang, et al., J. Am. Chem. Soc. 139 (2017) 5067–5074. doi: 10.1021/jacs.6b12536

  16. [16]

      Y. Cui, Z.M. Chen, X.F. Jiang, et al., Dalton Trans. 46 (2017) 5801–5805. doi: 10.1039/C7DT00179G

  17. [17]

      T. Feng, X. Li, J. Wu, et al., Chin. Chem. Lett. 31 (2020) 95–98. doi: 10.1016/j.cclet.2019.04.059

  18. [18]

      X. Kang, M. Zhu, Chem. Mater. 31 (2019) 9939–9969. doi: 10.1021/acs.chemmater.9b03674

  19. [19]

      M.J. Hardie, Chem. Lett. 45 (2016) 1336–1346. doi: 10.1246/cl.160780

  20. [20]

      C.H. Li, J.L. Zuo, Adv. Mater. 32 (2020) 1903762.

  21. [21]

      A. Kumar, P.S. Mukherjee, Chem. Eur. J. 26 (2020) 4842–4849. doi: 10.1002/chem.202000122

  22. [22]

      W.M. Bloch, G.H. Clever, Chem. Commun. 53 (2017) 8506–8516. doi: 10.1039/C7CC03379F

  23. [23]

      S. Bandi, D.K. Chand, Chem. Eur. J. 22 (2016) 10330–10335. doi: 10.1002/chem.201602039

  24. [24]

      K. Wu, B. Zhang, C. Drechsler, et al., Angew. Chem. Int. Ed. 60 (2020) 6403–6407.

  25. [25]

      S. Sudan, R.J. Li, S.M. Jansze, et al., J. Am. Chem. Soc. 143 (2021) 1773–1778. doi: 10.1021/jacs.0c12793

  26. [26]

      D.K. Chand, R. Manivannan, H.S. Sahoo, K. Jeyakumar, Eur. J. Inorg. Chem. 2005 (2005) 3346–3352. doi: 10.1002/ejic.200500192

  27. [27]

      J.J. Henkelis, C.J. Carruthers, S.E. Chambers, et al., J. Am. Chem. Soc. 136 (2014) 14393–14396. doi: 10.1021/ja508502u

  28. [28]

      J.J. Henkelis, J. Fisher, S.L. Warriner, M.J. Hardie, Chem. Eur. J. 20 (2014) 4117–4125. doi: 10.1002/chem.201304437

  29. [29]

      P.M. Cheng, L.X. Cai, S.C. Li, et al., Angew. Chem. Int. Ed. 59 (2020) 23569–23573. doi: 10.1002/anie.202011474

  30. [30]

      T. Sawada, H. Hisada, M. Fujita, J. Am. Chem. Soc. 136 (2014) 4449–4451. doi: 10.1021/ja500376x

  31. [31]

      S.M. Jansze, K. Severin, J. Am. Chem. Soc. 141 (2019) 815–819. doi: 10.1021/jacs.8b12738

  32. [32]

      S. Pullen, G.H. Clever. Acc. Chem. Res. 51 (2018) 3052–3064. doi: 10.1021/acs.accounts.8b00415

  33. [33]

      D. Samanta, P.S. Mukherjee, Chem. Eur. J. 20 (2014) 12483–12492. doi: 10.1002/chem.201402553

  34. [34]

      R.J. Li, F. Fadaei-Tirani, R. Scopelliti, K. Severin, Chem. Eur. J. 27 (2021) 9439–9445. doi: 10.1002/chem.202101057

  35. [35]

      D. Preston, J.E. Barnsley, K.C. Gordon, J.D. Crowley, J. Am. Chem. Soc. 138 (2016) 10578–10585. doi: 10.1021/jacs.6b05629

  36. [36]

      W.M. Bloch, J.J. Holstein, W. Hiller, G.H. Clever. Angew. Chem. 129 (2017) 8399–8404. doi: 10.1002/ange.201702573

  37. [37]

      A.M. Johnson, R.J. Hooley. Inorg. Chem. 50 (2011) 4671–4673. doi: 10.1021/ic2001688

  38. [38]

      M. Yamashina, T. Yuki, Y. Sei, et al., Chem. Eur. J. 21 (2015) 4200–4204. doi: 10.1002/chem.201406445

  39. [39]

      R. Zhu, W.M. Bloch, J.J. Holstein, et al., Chem. Eur. J. 24 (2018) 12976–12982. doi: 10.1002/chem.201802188

  40. [40]

      B. Chen, J.J. Holstein, A. Platzek, et al., Chem. Sci. 13 (2022) 1829–1834. doi: 10.1039/d1sc06931d

  41. [41]

      P. Liao, B.W. Langloss, A.M. Johnson, et al., Chem. Commun. 46 (2010) 4932–4934. doi: 10.1039/c0cc00234h

  42. [42]

      V. Croué, S. Krykun, M. Allain, et al., New J. Chem. 41 (2017) 3238–3241. doi: 10.1039/C7NJ00062F

  43. [43]

      D.A. McMorran, P.J. Steel, Angew. Chem. Int. Ed. 37 (1998) 3295–3297. doi: 10.1002/(SICI)1521-3773(19981217)37:23<3295::AID-ANIE3295>3.0.CO;2-5

  44. [44]

      L.P. Zhou, Q.F. Sun, Chem. Commun. 51 (2015) 16767–16770. doi: 10.1039/C5CC07306E

• 

  Figure 1  Cartoon illustration of (a) cage-to-cage transformation with similar structure; (b) step-by-step transformation of cage 2a to 2b by introduction of a mediated ligand Lm.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Structures of the ligands used in this study.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  ESI-MS spectrum of reaction mixture of direct transformation from 2a to 2b by mixing 2a and Lb in a 1.0: 7.0 molar ratio in CH₃CN at room temperature.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  ¹H NMR spectra (400 MHz, CD₃CN) of mixtures after stirring at room temperature for 24 h of (a) 2a and Lm in a 1.0: 7.0 molar ratio, (b) 2m and Lb in a 1.0: 7.0 molar ratio.

  下载: 全尺寸图片 幻灯片
•