English

• Biosystems are able to sense external stimuli, analyze the information, and respond to their surroundings correspondingly [1]. Inspired by many smart mechano-active systems with adaptable properties in nature, artificial materials are designed to sense their environment and respond to chemical or physical triggers resulting in changes in chemical structures or motion of relative components [2,3]. Rotaxanes [4], considered as a typical type of mechanically interlocked molecules (MIMs) [5-7], are crucial structural elements for the fabrication of advanced artificial materials [8,9]. They have been widely used for the construction of molecular shuttles and switches to study molecular motions induced by external stimuli [10,11] such as acid/base [12-15], redox [16-19], solvent [20-22], light [23-29], temperature [30,31] and ions [32-34]. Although many well-defined MIMs responsive to a single stimulus and one mode of mechanical motion have been developed over the past decades, there are still very limited examples of multistable rotaxanes that can be driven by three or more different stimuli owing to the lack of well-designed switchable host-guest systems [35-38]. A few systems that work under multi-stimuli conditions are only concerned with [2]rotaxane, while the research of multiple stimuli-responsive behaviors with higher order rotaxanes is still unavailable. Studying the response of such complex rotaxane systems responsive to external environments can contribute to the design and function of intelligent materials in the future [39-42], such as catalysis [43], imaging [44-49], surface mounted machines [50], and wearable devices [51].

  Despite wide applications of utilizing flexible crown ether and three-dimensional macrocycles for creating MIMs, hydrogen-bonded (H-bonded) aramide macrocycles [52] have been largely overlooked so far in this regard. These H-bonded macrocycles stand out due to their unique structural features, including a shape-persistent skeleton with a near-planar conformation, electron-rich cavity, and modifiable periphery, thus leading to diverse host-guest (H-G) interactions-based applications [53]. They have demonstrated significant potential in use for self-assembly [54,55], recognition [56], extraction [57], organic catalysis [58,59], and control of liquid crystal phases change [60], and construction of MIMs [61-64]. The azo aramide macrocycle containing azo groups in the molecular skeleton is particularly noteworthy as it is photoresponsive and shape-changeable, thereby making it possible to achieve precise encapsulation/release of guest molecules [65] and construction of autonomous molecular machines [66].

  In this work, we report a dynamic multistable [3]rotaxane whose motion can be driven by acid-base, thermal, solvent, and light. Such a multi-mode-driven [3]rotaxane [3]R-H features a H-bonded azo-macrocycle 1 as wheel component mechanically interlocked onto a non-photoactive thread with two different recognition sites, stoppered by equivalent ends (Scheme 1). A bipyridinium residue (BP) and two symmetric ammonium residues (AM) serve as primary (K_BP1 = (1.21 ± 0.07) × 10⁵, K_BP2 = (4.97 ± 0.98) × 10⁶) and secondary (K_AM = (1.29 ± 0.04) × 10⁴) electron-deficient binding sites for macrocycle 1, respectively, thanks to their ability to recognize hydrogen bond acceptors which are demonstrated in host-guest experiments (Figs. S3-S6 in Supporting information). Self-assembling with threads containing such motifs in nonpolar organic solvents allows the ready construction of rotaxanes. Leveraging this recognition process, rotaxane [3]R-Boc was synthesized in about 32% yield through a one-pot reaction between azo-macrocycle 1, 4'4-bipyridine, and the stopper S6. The target rotaxane [3]R-H was obtained by deprotection of the Boc group and acidification of the NH group using HPF₆. Rotaxane [3]R-H was fully characterized using ¹H and ¹³C NMR (Figs. S40 and S41 in Supporting information), HRESI-MS (Fig. S2 in Supporting information), 2D NOESY (Fig. S9 in Supporting information) and UV-vis absorption spectroscopies (Fig. S26 in Supporting information).

  Scheme 1

  

  Scheme 1.  Synthetic route of [3]rotaxanes. Charges are counterbalanced by PF₆⁻ counterions, which are omitted for clarity.

  DownLoad: Full-Size Img PowerPoint

  The 2D nuclear Overhauser effect (NOE) spectroscopy experiment was used to provide evidence for the threading of the axle component through the cavity of macrocycle 1 and thus the formation of the mechanical bond in [3]R-Boc (Fig. 1). Through-space cross-correlations were observed between protons H1, H2 and H7 at the BP binding site and the inner proton Hc of 1. Notably, proton H2 was shifted downfield more compared to proton H1, suggesting a larger deshielding effect after superposition than that of H1. This can be ascribed to the fact that both rings are located at the BP site and are very close in distance, resulting in H2 being able to form hydrogen bonds with both rings in the rotaxane [3]R-Boc simultaneously.

  Figure 1

  

  Figure 1.  Expanded 2D NOESY spectrum of [3]R-Boc (600 MHz, CDCl₃, 298 K, mixing time = 0.4 s).

  DownLoad: Full-Size Img PowerPoint

  The shuttling of protonated rotaxane [3]R-H was explored by ¹H NMR technique through comparing chemical shifts and signal splitting of protons on the other two control rotaxanes [3]R-Boc and [3]R-NH. The NMR spectrum of rotaxane [3]R-NH was very similar to that of [3]R-Boc (Fig. 2 and Fig. S10 in Supporting information). The chemical shifts of the protons in the macrocyclic skeleton and the BP site remain unchanged before and after deprotection. Since it was unlikely for each ring in [3]R-Boc to move towards its adjacent stopper by striding Boc groups on the axle, the high similarity in signal patterns between [3]R-Boc and [3]R-NH indicated that the macrocycle in [3]R-NH is located preferentially at the BP site. However, for [3]R-H, the macrocycle exhibited shuttling behavior between BP and AM sites. This was supported by the observation that protons H1 and H2 at the BP site in protonated rotaxane [3]R-H resonated with an upfield shift of 0.41 and 1.32 ppm, respectively, with respect to [3]R-Boc and [3]R-NH as a result of decreased C-H…O hydrogen bonding interactions with the ring. Interestingly, the resonance of the methylene proton H7 at BP site in [3]R-H underwent a pronounced deshielding effect, indicating that the C-H…O hydrogen bond between the macrocycle and H7 was enhanced. In addition, the splitting of signals pertaining to the AM station was observed for methylene protons (H9 and H9') adjacent to the ammonium site in [3]R-H. Taken together, along with the consideration that there are two recognition sites in [3]R-H, the macrocycle is supposed to shuttle between the two recognition sites. Given that a set of clear signals appear in the NMR spectrum at this time, it was speculated that the shuttle occurs on a time scale faster than the NMR time scale. To confirm the occurrence of shuttling, we performed a 2D NOESY analysis on rotaxane [3]R-H (Figs. S9-S11 in Supporting information). Coherently with the hypothesis of rapid shuttling, both protons H1, H2, and H7 at the BP binding site and protons H9/H9' at the AM binding site displayed cross-correlations with the inner protons of the ring. Notably, the cross peak (Hc, H9) presented stronger signal intensity than the cross peak (Hc, H9'), indicating that proton H9 is on average closer to the BP site concerning H9'. Proton H9, located between the BP site and the AM site, so within the shuttling range — resonated at a lower field, likely owing to the deshielding effect from hydrogen bonding between carbonyl oxygen atoms and proton H9. In contrast, proton H9', located outside the shuttling range of the ring, was shielded.

  Figure 2

  

  Figure 2.  Partial ¹H NMR spectra (400 MHz, 298K, CDCl₃) of [3]rotaxane [3]R-Boc (a), [3]rotaxane [3]R-NH (b), [3]rotaxane [3]R-H (c) and H-bonded azo-macrocycle 1 (d). (e) Schematic representation of [3]R-Boc, [3]R-NH, and [3]R-H. Charges are counterbalanced by PF₆⁻ counterions, which are omitted for clarity.

  DownLoad: Full-Size Img PowerPoint

  We attempted to estimate the shuttling rate by variable temperature NMR (Fig. S12 in Supporting information). However, no coalescence of signals in the ¹H NMR spectra of [3]R-H was observed in CDCl₃ even at temperatures of as low as 238 K, indicating that even at such a low temperature the shuttling occurs on a time scale faster than the NMR time scale (Fig. 2e). Meanwhile, as the temperature decreased, the signals of H1 and H2 protons at the BP site gradually moved toward higher fields, while H7 underwent further deshielding and the signal of it shifts towards lower fields, indicating a change in the average position of 1 on the axle. We speculate that for rotaxane [3]R-H, the attraction of BP and AM sites to the macrocycle causes macrocycle 1 to shuttle between these two sites. Changing the relative attraction of these two sites to the macrocycle can change its average position on the axle. Compared with the BP site, the AM site is more susceptible to environmental influences due to the presence of active hydrogen, which can alter the ionization degree of NH₂⁺. As the temperature decreases, the ionization of NH₂⁺ is suppressed, and the attraction to the macrocycle is enhanced, resulting in the shift of the relevant signal peaks in the variable temperature NMR.

  Different equivalents of trifluoroacetic acid (TFA) were added to [3]R-NH for NMR titration (Fig. S13 in Supporting information). As the acid equivalent gradually increased, the signal peaks of protons H1 and H2 at the BP site gradually shifted towards higher fields, indicating that the macrocycle was gradually decomplexation with the BP site. In addition, NH was protonated into NH₂⁺, and due to its positive charge, the signal peak of its adjacent characteristic methylene proton H9/H9′ gradually shifted towards a lower field. The increase in excess acid equivalent further inhibited the ionization of NH₂⁺, and the attraction of the AM site to the macrocycle gradually increased, causing the shuttling of 1 between BP and AM sites and the signal peaks of protons H9 and H9' to split. As the quantity of the acid increased, the difference between the resonances of H9 and H9' becomes larger. This phenomenon indicated that with the increase of acid, the average position of the macrocycle on the axle would gradually move towards the AM site. The determination of NOE distance for additions of different equivalent acids also supported this result (Figs. S14-S16 in Supporting information). The strength of the NOE signal is inversely proportional to the sixth power of distance. Based on this formula, it is possible to calculate the spatial distance between protons. Even if there is a two-fold error between the measured signal strengths, it is only about 10% reflected in the final calculated values [67]. Compared to adding 6 times the equivalent of acid to [3]R-NH, adding 20 times the equivalent of [3]R-NH increased the distance between the intra protons Ha/Hc of macrocycle 1 and the H1/H2 of the BP site, while the distance to the H9/H9' of the AM site decreased (Table S1 and Fig. S17 in Supporting information).

  Considering the correlation between the ionization degree of compounds and solvent polarity, a polarity change experiment was conducted. To this end, we changed the volume ratio of CDCl₃/DMSO-d and CDCl₃/cyclehexane-d₁₂ to adjust the solvent polarity and control the molecular motion of the ring on the axle (Fig. 3). In Fig. 3, we found that as the solvent polarity increased due to the stepwise addition of DMSO-d₆ to the fleshly prepared CDCl₃ solution of [3]R-H, the resonances of protons H1 and H2, which were located at the BP station, are both moved downfield. By contrast, the resonances of protons near the AM station, such as H9 and H9', shifted upfield. For example, the resonance of proton H1 moved downfield from 9.28 ppm (CDCl₃, ε = 4.81) to 9.38 ppm (CDCl₃/DMSO-d₆ = 100:1, ε = 5.28), 9.55 ppm (CDCl₃/DMSO-d₆ = 100:5, ε = 7.07), 9.81 ppm (CDCl₃/DMSO-d₆ = 100:40, ε = 17.98). On the flip side, the resonance of proton H9 moves upfield from 3.03 ppm (CDCl₃, ε = 4.81) to 2.85 ppm (CDCl₃/DMSO-d₆ = 100:1, ε = 5.28), 2.70 ppm (CDCl₃/DMSO-d = 100:5, ε = 7.07), 2.48 ppm (CDCl₃/DMSO-d₆ = 100: 40, ε = 17.98). Similar chemical shift changed also occur for protons H2 and H9' (Fig. S18 in Supporting information). It showed opposite effects as the solvent polarity decreases due to the stepwise addition of cyclohexane-d₁₂ to the fleshly prepared chloroform-d solution of [3]R-H (Fig. S21 in Supporting information). In these cases, it was confirmed that the BP site gradually moved out of the cavity of the rings; meanwhile, the ring moved onto the AM site. By altering the solvent polarity, the rings are moved to different positions on the axle on average. These phenomena infer that with the change of solvent polarity, [3]rotaxane [3]R-H exhibits the corresponding molecular motion and acts as a kind of intelligent molecule capable of sensing external changes. What is more, the results of the 2D NOESY experiment also effectively supported this conclusion (Figs. S19, S20, S22-S24 in Supporting information).

  Figure 3

  

  Figure 3.  Partial polar variable ¹H NMR spectra of [3]R-H (800 MHz, CDCl₃/DMSO-d₆, CDCl₃ and CDCl₃/cyclohexane-d₁₂, 298 K).

  DownLoad: Full-Size Img PowerPoint

  Considering that chloroform is unstable under long-term UV irradiation, we chose acetonitrile as the solution for photoresponsive studies. Following 30 min of UV irradiation at 365 nm wavelength, the ¹H NMR spectra with a freshly prepared 10 mmol/L [3]R-H solution in CD₃CN revealed new resonances (Figs. 4a and b). Specifically, protons H1 and H2 at the BP site displayed free, unbound signals (H1* and H2*), suggesting that macrocycle 1 was dissociated from the BP site after exposure to UV light. Concurrently, protons H9 and H9' at the AM site showed new signals at a lower field, indicating a stronger binding of the isomerized macrocycle to the AM site, which is also supported by host-guest experiments (Figs. S7 and S8 in Supporting information). ¹H NMR spectra revealed that approximately 36% of [3]R-H underwent photoisomerization. Reversing this process, 20 min of blue light (450 nm) irradiation restored the ¹H NMR spectrum to its original one (Fig. 4c). These observations indicate that UV irradiation facilitates the photoisomerization of [3]R-H, prompting the macrocycles to shift from the BP to the AM site on the axle (Fig. S25 in Supporting information).

  Figure 4

  

  Figure 4.  Partial ¹H NMR spectra of [3]R-H (400 MHz, CD₃CN, 298 K) (a) before (b) after 30 min 365 nm irradiation, and (c) after 20 min 450 nm irradiation.

  DownLoad: Full-Size Img PowerPoint

  Upon UV light irradiation, the π-π^* absorption band of trans-azobenzene in [3]R-H at approximately 340 nm gradually decreased, while the n-π^* absorption band of cis-azobenzene at around 430 nm increased, signifying an E→Z isomerization of azobenzene units. After approximately 3 min of irradiation, the system reached the photostationary state (PSS). Irradiating [3]R-H with blue light reversed this process, and the absorption spectrum returned to its initial state after about 1 minute (Fig. S26 in Supporting information). As shown in Fig. 5, the system exhibited no fatigue over at least 10 cycles of alternating 365 nm and 450 nm irradiation.

  Figure 5

  

  Figure 5.  Fatigue experiment of [3]R-H (10 µmol/L in CH₃CN, 298 K) interconversion upon irradiation with 365 and 450 nm in an alternating sequence; absorbance values measured at λ = 340 nm.

  DownLoad: Full-Size Img PowerPoint

  In conclusion, this study, to our knowledge, represents the first example of a multistable [3]rotaxane that can respond to multiple stimuli. Compared with the BP site in the rotaxane [3]R-H, the ionization of the AM site is more susceptible to the environment. External stimuli can open (enhance)/close (weaken) the interaction between the AM site and the macrocycle, causing the average position of the macrocycle on the axle to change. Stimuli such as acid/base, temperature, and polarity can change the relative binding strength between the [3]R-H axle site and the macrocycle. Light irradiations alter the shape of the macrocycle in [3]R-H and thus change the site selectivity. Such multi-stimuli responsive higher order rotaxane was achieved by using the approaches above, which is inspiring for the design of future intelligent materials.

  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

  Zhiyao Yang: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Kuirong Fu: Methodology, Investigation. Wentao Yu: Software, Formal analysis. Along Jia: Formal analysis, Data curation. Xinnan Chen: Visualization. Yimin Cai: Writing – review & editing, Validation. Xiaowei Li: Writing – review & editing, Writing – original draft, Supervision, Funding acquisition, Formal analysis, Conceptualization. Wen Feng: Writing – review & editing, Resources. Lihua Yuan: Writing – review & editing, Supervision, Funding acquisition, Formal analysis, Conceptualization.

  Acknowledgments

  This work is supported by the National Natural Science Foundation of China (No. 22271202 to L. Yuan, No. 22201193 to X. Li), the Sichuan Science and Technology Program (No. 2023NSFSC0109 to X. Li), the Fundamental Research Funds for the Central Universities and the Hundred Talent Program of Sichuan University (No. YJ2021158 to X. Li) and Sichuan University Interdisciplinary Innovation Fund (X. Li), Open Project of State Key Laboratory of Supramolecular Structure and Materials (No. SKLSSM2024037). We thank Dr. Dongyan Deng from College of Chemistry, and Dr. Pengchi Deng and Dr. Chen Yuan from Analytical & Testing Center, Sichuan University, for analytic testing and valuable help.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110842.

  
  
• 1. [1]

     T.R. Kelly, Molecular Machines, Springer-Verlag, Berlin, Heidelberg, 2005.

  2. [2]

     M.W. Urban, Stimuli-Responsive Materials, Royal Society of Chemistry, Cambridge, UK, 2016.

  3. [3]

     E. Moulin, L. Faour, C.C. Carmona-Vargas, N. Giuseppone, Adv. Mater. 32 (2020) 1906036. doi: 10.1002/adma.201906036

  4. [4]

     M. Xue, Y. Yang, X. Chi, Chem. Rev. 115 (2015) 7398–7501. doi: 10.1021/cr5005869

  5. [5]

     A.R. Pease, J.O. Jeppesen, J.F. Stoddart, et al., Acc. Chem. Res. 34 (2001) 433–444. doi: 10.1021/ar000178q

  6. [6]

     J.P. Sauvage, Angew. Chem. Int. Ed. 56 (2017) 11080–11093. doi: 10.1002/anie.201702992

  7. [7]

     J.F. Stoddart, Angew. Chem. Int. Ed. 56 (2017) 110994–111125.

  8. [8]

     S. Erbas-Cakmak, D.A. Leigh, C.T. McTernan, A.L. Nussbaumer, Chem. Rev. 115 (2015) 10081–10206. doi: 10.1021/acs.chemrev.5b00146

  9. [9]

     A.W. Heard, S.M. Goldup, ACS Cent. Sci. 6 (2020) 117–128. doi: 10.1021/acscentsci.9b01185

  10. [10]

      C. Bruns, J.F. Stoddart, The Nature of the Mechanical Bond: From Molecules to Machines, John Wiley & Sons, Inc., Hoboken, New Jersey, 2016.

  11. [11]

      X. Yan, F. Wang, B. Zheng, F. Huang, Chem. Soc. Rev. 41 (2012) 6042–6065. doi: 10.1039/c2cs35091b

  12. [12]

      C. Yu, X. Wang, C.X. Zhao, et al., Chin. Chem. Lett. 33 (2022) 4904–4907. doi: 10.1016/j.cclet.2022.03.004

  13. [13]

      Q. Zhao, Y. Chen, B. Sun, et al., Eur. J. Org. Chem. 2019 (2019) 3396–3400.

  14. [14]

      J. Ye, R. Zhang, W. Yang, et al., Chin. Chem. Lett. 31 (2020) 1550–1553. doi: 10.1016/j.cclet.2019.11.041

  15. [15]

      H.Y. Zhou, Y. Han, C.F. Chen, Mater. Chem. Front. 4 (2020) 12–28. doi: 10.1039/C9QM00546C

  16. [16]

      R.A. Bissell, E. Córdova, A.E. Kaifer, J.F. Stoddart, Nature 369 (1994) 133–137. doi: 10.1038/369133a0

  17. [17]

      H.V. Schröder, C.A. Schalley, Chem. Sci. 10 (2019) 9626–9639. doi: 10.1039/C9SC04118D

  18. [18]

      X.Y. Chen, H. Chen, J.F. Stoddart, Angew. Chem. Int. Ed. 135 (2023) e202211387. doi: 10.1002/ange.202211387

  19. [19]

      A. Bhadani, M. Kathiresan, Org. Chem. Front. 11 (2024) 2954–2980. doi: 10.1039/D4QO00061G

  20. [20]

      X.Y. Hu, X. Wu, Q. Duan, et al., Org. Lett. 14 (2012) 4826–4829. doi: 10.1021/ol302149t

  21. [21]

      S. Dong, J. Yuan, F. Huang, Chem. Sci. 5 (2014) 247–252. doi: 10.1039/C3SC52481G

  22. [22]

      Y. Yao, P. Zhang, D. Zhou, et al., Chin. Chem. Lett. 35 (2024) 108712. doi: 10.1016/j.cclet.2023.108712

  23. [23]

      H. Chong, C. Nie, L. Wang, et al., Chin. Chem. Lett. 32 (2017) 57–61.

  24. [24]

      G. Liu, C. Tian, X. Fan, et al., Org. Lett. 25 (2023) 8761–8765. doi: 10.1021/acs.orglett.3c03804

  25. [25]

      X.H. Gu, J.X. Yang, L.J. Liu, et al., New J. Chem. 47 (2023) 19767–19774. doi: 10.1039/D3NJ04164F

  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]

      Y. Zhang, Y. Chen, J.Q. Li, et al., Adv. Sci. 11 (2024) 230777.

  28. [28]

      W.J. Li, W.T. Xu, X.Q. Wang, et al., J. Am. Chem. Soc. 14 (2023) 14498.

  29. [29]

      T. Xiao, J. Wang, Y. Shen, et al., Chin. Chem. Lett. 32 (2021) 1377–1380. doi: 10.1016/j.cclet.2020.10.037

  30. [30]

      R. Liu, Y. Zhang, W. Wu, et al., Chin. Chem. Lett. 30 (2019) 577–581. doi: 10.1016/j.cclet.2018.12.002

  31. [31]

      J. Puigcerver, M. Alajarin, A. Martinez-Cuezva, J. Berna, Org. Biomol. Chem. 221 (2023) 9070–9075.

  32. [32]

      C. Ren, F. Chen, R. Ye, et al., Angew. Chem. Int. Ed. 58 (2019) 8034–8038. doi: 10.1002/anie.201901833

  33. [33]

      T.G. Johnson, M.J. Langton, J. Am. Chem. Soc. 145 (2023) 27167–27184. doi: 10.1021/jacs.3c08877

  34. [34]

      A. Arun, H.M. Tay, P.D. Beer, Chem. Commun. 60 (2024) 11849–11863. doi: 10.1039/D4CC03916E

  35. [35]

      H. Murakami, A. Kawabuchi, R. Matsumoto, et al., J. Am. Chem. Soc. 127 (2005) 15891–15899. doi: 10.1021/ja053690l

  36. [36]

      C.F. Lin, C.C. Lai, Y.H. Liu, et al., Chem. Eur. J. 13 (2007) 4350–4355. doi: 10.1002/chem.200601432

  37. [37]

      H.Y. Zhou, Y. Han, Q. Shi, C.F. Chen, Eur. J. Org. Chem. 2019 (2019) 3406–3411. doi: 10.1002/ejoc.201801785

  38. [38]

      Q. Zhang, H. Qian, T. Xiao, et al., Chin. Chem. Lett. 34 (2023) 108365. doi: 10.1016/j.cclet.2023.108365

  39. [39]

      X. Zhang, L.F. Chen, K.H. Lim, et al., Adv. Mater. 31 (2019) 1804540. doi: 10.1002/adma.201804540

  40. [40]

      I. Aprahamian, ACS Cent. Sci. 6 (2020) 347–358. doi: 10.1021/acscentsci.0c00064

  41. [41]

      B. Wang, Y. Lu, Nano-Micro Lett. 16 (2024) 155. doi: 10.1007/s40820-024-01379-4

  42. [42]

      T. Xiao, L. Zhou, X.Q. Sun, et al., Chin. Chem. Lett. 31 (2020) 1–9. doi: 10.1016/j.cclet.2019.05.011

  43. [43]

      R.J.M. Nolte, J.A.A.W. Elemans, Chem. Eur. J. 30 (2024) e202304230. doi: 10.1002/chem.202304230

  44. [44]

      D. Li, J. Wang, X. Ma, Adv. Opt. Mater. 6 (2018) 1800273. doi: 10.1002/adom.201800273

  45. [45]

      C.H. Wu, P.Q. Nhien, T.T.K. Cuc, et al., Top Curr. Chem. 318 (2023) 2.

  46. [46]

      Y. Zhang, Y. Wang, T. Chen, et al., Chem. Commun. 59 (2023) 8266–8269. doi: 10.1039/D3CC01929B

  47. [47]

      T. Chen, J. Wang, R. Tang, et al., Chin. Chem. Lett. 34 (2023) 108088. doi: 10.1016/j.cclet.2022.108088

  48. [48]

      L. Fang, Y. Dai, Y. Bai, et al., Chem. Commun. 60 (2024) 7646–7649. doi: 10.1039/D4CC01987C

  49. [49]

      F. Cui, Q. Zhang, T. Xiao, et al., Chin. Chem. Lett. 35 (2024) 110061. doi: 10.1016/j.cclet.2024.110061

  50. [50]

      R.D.S. Rodrigues, K.M. Mullen, ChemPlusChem 82 (2017) 814. doi: 10.1002/cplu.201700065

  51. [51]

      X. Xiong, Y. Chen, Z. Wang, et al., Nat. Commun. 14 (2023) 1331. doi: 10.1038/s41467-023-36920-3

  52. [52]

      Z. Liu, Y. Zhou, L. Yuan, Org. Biomol. Chem. 20 (2022) 9023–9051. doi: 10.1039/D2OB01263D

  53. [53]

      S. Huang, X. Li, Y. Cai, et al., Chem. Eur. J. 30 (2024) e202303394. doi: 10.1002/chem.202303394

  54. [54]

      Y. Yang, W. Feng, J. Hu, et al., J. Am. Chem. Soc. 133 (2011) 18590–18593. doi: 10.1021/ja208548b

  55. [55]

      Z. Wang, L. Mei, C. Guo, et al., Angew. Chem. Int. Ed. 135 (2023) e202216690. doi: 10.1002/ange.202216690

  56. [56]

      Y. He, M. Xu, R. Gao, et al., Angew. Chem. Int. Ed. 53 (2014) 11834–11839. doi: 10.1002/anie.201407092

  57. [57]

      Q. Jiang, L.T. He, S.Z. Luo, et al., Chin. Chem. Lett. 24 (2013) 881–884. doi: 10.1016/j.cclet.2013.05.025

  58. [58]

      K. Kang, J.A. Lohrman, S. Nagarajan, et al., Org. Lett. 21 (2019) 652–655. doi: 10.1021/acs.orglett.8b03778

  59. [59]

      J. Wu, Y. Luo, L. Chen, et al., Chem. Commun. 58 (2022) 12867–12870. doi: 10.1039/D2CC05027G

  60. [60]

      X. Li, B. Li, L. Chen, Angew. Chem. Int. Ed. 54 (2015) 11147–11152. doi: 10.1002/anie.201505278

  61. [61]

      X. Li, X. Yuan, P. Deng, et al., Chem. Sci. 8 (2017) 2091–2100. doi: 10.1039/C6SC04714A

  62. [62]

      Z. Ye, J. Wang, S.S.K. Kothapalli, et al., Chem. Commun. 56 (2020) 1066–1069. doi: 10.1039/C9CC08253K

  63. [63]

      Y. Zhou, J. Wu, Z. Liu, et al., Chem. Commun. 57 (2021) 13506–13509. doi: 10.1039/D1CC05501A

  64. [64]

      S. Huang, Z. Wang, J. Wu, et al., Chem. Commun. 60 (2024) 5622–5625. doi: 10.1039/D4CC00178H

  65. [65]

      Z. Ye, Z. Yang, L. Wang, et al., Angew. Chem. Int. Ed. 58 (2019) 12519–12523. doi: 10.1002/anie.201906912

  66. [66]

      Z. Yang, X. Wang, E. Penocchio, et al., Angew. Chem. Int. Ed. 64 (2025) e202414072. doi: 10.1002/anie.202414072

  67. [67]

      T.D.W. Claridge, Correlations through space: the nuclear Overhauser effect, Tetrahedron Organ. Chem. Ser. 27 (2009) 247–302.

• 

  Scheme 1  Synthetic route of [3]rotaxanes. Charges are counterbalanced by PF₆⁻ counterions, which are omitted for clarity.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  Expanded 2D NOESY spectrum of [3]R-Boc (600 MHz, CDCl₃, 298 K, mixing time = 0.4 s).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Partial ¹H NMR spectra (400 MHz, 298K, CDCl₃) of [3]rotaxane [3]R-Boc (a), [3]rotaxane [3]R-NH (b), [3]rotaxane [3]R-H (c) and H-bonded azo-macrocycle 1 (d). (e) Schematic representation of [3]R-Boc, [3]R-NH, and [3]R-H. Charges are counterbalanced by PF₆⁻ counterions, which are omitted for clarity.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Partial polar variable ¹H NMR spectra of [3]R-H (800 MHz, CDCl₃/DMSO-d₆, CDCl₃ and CDCl₃/cyclohexane-d₁₂, 298 K).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Partial ¹H NMR spectra of [3]R-H (400 MHz, CD₃CN, 298 K) (a) before (b) after 30 min 365 nm irradiation, and (c) after 20 min 450 nm irradiation.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Fatigue experiment of [3]R-H (10 µmol/L in CH₃CN, 298 K) interconversion upon irradiation with 365 and 450 nm in an alternating sequence; absorbance values measured at λ = 340 nm.

  下载: 全尺寸图片 幻灯片
•