English

• 1. Introduction

  Metal-organic cages (MOCs) are three-dimensional, discrete polyhedral structures formed through coordination bonds between central metal ions and structurally complementary organic ligands. These assemblies exhibit well-defined geometric shapes, precisely tailored cavities, and highly adaptable chemical properties [1–15]. Due to these distinctive features, MOCs have attracted significant attention in recent years, demonstrating considerable potential in a wide range of applications, including adsorption and separation [16–22], catalysis [23–33], cargo transport [34], and molecular sensing [35–39]. In recent years, stimulus-responsive MOCs have attracted significant research interest due to their ability to incorporate responsive features into their structures [40–44]. This capability provides a robust platform for dynamically modulating intermolecular interactions and functions. By integrating external stimuli such as light [45–48], pH [49–52], temperature [53,54], or chemical agents [55–61], the structural and functional properties of MOCs can be dynamically modulated, enhancing their versatility and broadening their potential applications. Among these stimuli, light offers a particularly attractive means of regulating MOCs due to its cost-effectiveness, operational efficiency, non-invasive nature, and minimal generation of byproducts. MOCs can be readily modified to exhibit light-responsive behavior by introducing photoresponsive units. For example, in 2007, Fujita successfully incorporated azobenzene into MOCs, which enabled the complexes to undergo light-induced transformations through ligand isomerization [62]. Building on this approach, Clever advanced the field in 2013 by integrating dithienylethene into the ligand backbone, thereby creating MOCs with reversible light-switchable structures [63]. More recently, in 2019, Feringa introduced molecular motors into a metal-organic cage, allowing it to display both light- and thermally induced transformation properties [64]. These studies highlight the significant potential of photoresponsive MOCs in applications such as smart materials, molecular recognition, and controlled catalysis.

  Azobenzene has been widely studied in photochemistry [65–69] and supramolecular chemistry [70–73] since its discovery by Mitscherlich in 1834 [74]. The widespread interest in azobenzene arises from its unique photoisomerization properties. Upon exposure to ultraviolet and visible light, azobenzene derivatives undergo rapid and reversible isomerization between their cis and trans configurations. These transitions induce significant changes in molecular geometry, polarity, and electron distribution [75,76], endowing azobenzene derivatives with considerable potential for applications in photoresponsive molecular switches, light-driven molecular machines, and smart materials.

  In recent years, azobenzene derivatives have been widely utilized as photoresponsive units in MOCs. Within these systems, azobenzene units are typically incorporated in two primary configurations: as pendant groups attached to the cage framework or as integral components of the cage backbone. This incorporation allows MOCs to undergo structural disassembly/reassembly, conformational changes, or alterations in chemical properties upon light irradiation, enabling precise photomodulation of cage functionality (Fig. 1). Despite the significant advancements in azobenzene-based MOCs, a comprehensive review summarizing these developments is still lacking. Such a review is crucial to guide the continued progress of photoresponsive MOCs. Therefore, this review aims to highlight the key research achievements in azobenzene-functionalized MOCs, offer theoretical insights for the design and functional optimization of photoresponsive MOCs, and explore their potential future applications.

  Figure 1

  

  Figure 1.  Schematic illustrations of photoresponsive azobenzene-containing MOCs.

  DownLoad: Full-Size Img PowerPoint

  2. Azobenzene units dangled on the cage

  Azobenzene units can be incorporated into metal-organic cages in two main ways: they can be attached either to the interior of the cavity or to the exterior of the cage framework. This configuration enables the integration of photoresponsive units into preexisting metal-organic cage structures, which significantly simplifies ligand design and synthesis. However, attaching the azobenzene units in this manner also limits the ability to modify the cage's backbone, as the overall structural framework remains largely unchanged. While this approach reduces synthetic complexity, it may restrict further customization of the cage architecture.

  In 2007, Fujita and colleagues reported an azobenzene-functionalized metal coordination cage, designated as C1 [62]. This structure self-assembles spontaneously into a spherical complex comprising 12 metal ions and 24 pyridine-based bent bridging ligands (Fig. 2). The azobenzene units are attached to the inner side of the bent ligands and are positioned within the cavity of C1 after assembly. The trans-azobenzene configuration in ligand L1a increases the hydrophobicity of the cavity, which allows cage C1a to encapsulate hydrophobic guests such as pyrene in a CH₃CN/H₂O (1:1) solution. In contrast, cage C1b, derived from ligand L1b, does not exhibit this encapsulation behavior due to its different structural features. The incorporation of azobenzene units within the cage allows for tunable hydrophobicity, as the cis-isomer of azobenzene exhibits higher polarity compared to its trans counterpart. Upon irradiation with 365 nm ultraviolet (UV) light, the azobenzene units in the cage undergo a trans-to-cis isomerization. Because the cis isomer is more polar than the trans isomer, this transformation decreases the hydrophobicity of the cavity, which can lead to the expulsion of hydrophobic guest molecules. Heating the system at 50 ℃ for 12 h induces the guests to re-enter the cavity, as the azobenzene units revert to their trans configuration. This unique photoresponsive behavior enables the cage to reversibly bind and release hydrophobic guests under light stimulation, without altering the structural integrity of the cage. This strategy of modulating the internal environment of MOCs via light irradiation offers a versatile approach for designing a variety of photoresponsive "molecular nanoparticles".

  Figure 2

  

  Figure 2.  Self-assembly of M₁₂L₂₄ spherical complexes C1 with 24 endohedral azobenzene groups (L1a) or 24 methoxy groups (L1b) and hydrophobic guest. Copied with permission [62]. Copyright 2007, Wiley Publishing Group.

  DownLoad: Full-Size Img PowerPoint

  Nitschke and colleagues employed a subcomponent self-assembly strategy to functionalize the outer vertices of MOC with azobenzene units, leading to the formation of a tetrahedral metal cage, designated as C2 (Fig. 3a). Upon 365 nm light irradiation, the azobenzene units undergo a transformation to the cis-isomer, which has a larger spatial footprint, increasing steric hindrance at the vertices of the cage. This added hindrance further influences host–guest interactions and the overall dynamic behavior of the system. This steric clash triggers the disintegration of the MOC, accompanied by the release of encapsulated guest molecules. When the azobenzene units revert to their trans-configuration upon irradiation with 500 nm light, the cage reassembles and recaptures the guest molecules. By exploiting this reversible process, the researchers successfully achieved the selective separation and purification of progesterone from a steroid mixture using the MOC (Fig. 3b). Notably, these cages exhibit excellent recyclability, as shown in Fig. 3c, offering an efficient and sustainable method for purifying high-value steroid compounds with potential applications in pharmaceuticals and materials science [77].

  Figure 3

  

  Figure 3.  (a) Self-assembly of photoresponsive cage C2. (b) Steroid guests investigated for binding within C2, progesterone represented as red circles and green, blue and orange circles represented mestranol, cholesterol and 7-dehydrocholesterol respectively. (c) Illustration of progesterone separation from mixture steroid guests. Copied with permission [77]. Copyright 2024, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  The trans-to-cis isomerization of azobenzene units can significantly impact the solubility of the metal-organic cage itself. Zhou and colleagues designed a metal-organic cage, C3, by incorporating 24 ligands functionalized with pendant azobenzene units (Fig. 4a), and 12 Cu₂ paddlewheel units which composed of two copper ions coordinated to four carboxylate ligands (Fig. 4b). This cage exhibits distinctive solubility behavior: it dissolves in chloroform but gradually precipitates under dark conditions, only to redissolve upon UV light irradiation [78]. This phenomenon can be attributed to two key factors. First, the cis-isomerization of the azobenzene ligands diminishes intermolecular interactions between cages, such as π-π stacking. Second, the cis-azobenzene configuration has a larger dipole moment, which enhances its solubility in polar solvents. This reversible solubility transition enables the binding and release of larger guest molecules that would otherwise be too large to fit within the cage cavity. In the dark, C3 gradually precipitates, and methylene blue (MB) is captured in the pockets formed between the trans-C3 cages due to strong intermolecular interactions. Upon UV light irradiation, the photoisomerization of azobenzene weakens these interactions and increases the solubility of the cage, resulting in the release of MB (Fig. 4c). Notably, this process can be reversibly controlled by alternating between blue and UV light irradiation. This work offers a novel strategy for reversible guest capture and release.

  Figure 4

  

  Figure 4.  (a) Light-induced trans/cis-isomerization of L3. (b) Structure of C3. (c) Schematic illustrations of reversible formation of insoluble trans-C3 and soluble cis-C3 with capture and release of MB. Copied with permission [78]. Copyright 2014, Wiley Publishing Group.

  DownLoad: Full-Size Img PowerPoint

  To overcome the challenges of aggregation and low isomerization efficiency, Zhou and colleagues confined the MOCs within mesoporous silica [79]. The practical application of photoresponsive MOCs has often been limited by their low photoresponsive efficiency in the solid state as shown in Figs. 5a and b. As such, developing strategies to enhance the photoresponsive performance of MOCs in solid-state environments is crucial. By encapsulating the MOCs within mesoporous silica, the aggregation typically observed in bulk MOCs was effectively prevented, allowing for freely reversible cis-trans isomerization under 365 nm/450 nm light irradiation (Fig. 5c). The confined C3 exhibited significantly higher adsorption differences for propene (48.2%) and brilliant blue G (43.9%) before and after isomerization compared to their bulk counterparts (11.2% and 7.8%, respectively). This difference is attributed to the partial blocking of active copper adsorption sites in the cis configuration and aggregated state. This strategy presents a novel method for unlocking the full potential of photoresponsive materials in future applications.

  Figure 5

  

  Figure 5.  (a) Schematic illustrations of light-induced reversible trans/cis-isomerization of C3. (b) Only a few C3 units at the edges exhibit photoreponsiveness in the solid state. (c) C3 dispersed in the nanoscaled spaces of mesoporous silica. Copied with permission [79]. Copyright 2019, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  3. Azobenzene units as an integral component of the cage backbone

  When azobenzene is incorporated as a linker within the backbone of MOCs, its photoactivity is often constrained by the rigid framework of the complex. This rigidity makes the trans-to-cis isomerization more difficult, which in turn extends the thermal half-life of the cis-azobenzene isomer after complex formation [80,81]. These factors collectively limit the ability of MOCs with azobenzene-based linkers to exhibit efficient light-driven responses.

  Hardie and colleagues reported a switchable MOC featuring azobenzene linking units [82]. This structure, designated as C4, is an M₃L₂ cage composed of a functionalized tripodal cyclotriguaiacylene (L4) derivative, bearing pyridyl-azo-phenyl groups, which are coordinated to Ir(Ⅲ) centers (Fig. 6a). When azobenzene is integrated into the cage backbone, the isomerization efficiency is significantly reduced, with a maximum of only approximately 39%. Consequently, irradiation at 355 nm generates a mixture of complexes containing different isomers. As the azobenzene units undergo isomerization, the cage structure distorts but remains intact (Fig. 6b), demonstrating reversible light-responsive switching behavior. This behavior can be attributed to the high rotational flexibility of the ligand design, which allows the ligands to adapt to structural changes via single-bond rotations, maintaining a configuration suitable for cage formation. Additionally, replacing the chemically inert Ir(Ⅲ) centers with more labile Pd(Ⅱ) ions accelerates the self-assembly process. However, this substitution results in cage disassembly upon ligand isomerization [83], highlighting the crucial role of the metal center's chemical inertness. While the inert metal centers slow the self-assembly kinetics, they play an essential role in preventing the disintegration of the coordination cage. The combination of highly flexible ligands and inert metal centers provides a valuable strategy for designing photoresponsive MOCs with switchable properties, offering insights into the development of advanced stimuli-responsive materials.

  Figure 6

  

  Figure 6.  (a) Schematic illustrations of synthesis of C4 and energy-minimized structure. (b) Energy-minimized structure of C4 with differing numbers of cis-isomerization ligand arms. Copied with permission [82]. Copyright 2018, the Royal Society of Chemistry.

  DownLoad: Full-Size Img PowerPoint

  Liu and colleagues reported an example of an M₂L₄ MOC incorporating an azobenzene linker ligand [84], which undergoes reversible light-induced disassembly and reassembly. This cage, designated as C5, is composed of ligands containing two azobenzene units attached to their arms and coordinated to Pd(Ⅱ) centers (Fig. 7). Upon irradiation with 365 nm light, the azobenzene units undergo isomerization, inducing significant conformational changes in the ligand (L5). Consequently, the cis-isomer of the ligand loses the geometry necessary for coordination with Pd(Ⅱ), thus preventing the formation of the MOC. As a result, C5 disassembles under UV light irradiation. When the ligand reverts to its trans-isomer, the cage reassembles, enabling reversible light-driven assembly and disassembly of the MOC. This work demonstrates a robust strategy for dynamically controlling MOC formation through the use of photoresponsive linkers.

  Figure 7

  

  Figure 7.  Structures of C5 and L5, along with their light-induced self-assembly and disassembly. Reproduced with permission [84]. Copyright 2019, the Royal Society of Chemistry.

  DownLoad: Full-Size Img PowerPoint

  Ward and colleagues successfully integrated azobenzene moieties into a bis(pyrazolyl-pyridine) framework, synthesizing a novel bidentate bridging ligand, L6 (Fig. 8a). In the presence of Co(Ⅱ) ions, the trans-configuration of L6 (trans-L6) self-assembled into a helical coordination complex, trans-C6. Upon exposure to 340 nm ultraviolet light, trans-L6 underwent photoisomerization to its cis-configuration (cis-L6), leading to the formation of a tetrahedral coordination complex, cis-C6 (Fig. 8b). This system exhibited excellent reversibility, with white light irradiation facilitating the regeneration of trans-C6. Extensive cyclic light-switching experiments confirmed the remarkable fatigue resistance of this photoresponsive transformation. The ligand's inherent photochromic properties enabled precise and reversible light-driven interconversion between trans-C6 and cis-C6, with both forms maintaining identical chemical compositions but distinct structural architectures [85]. Notably, the cis-isomer complex displayed exceptional thermodynamic stability relative to its mononuclear counterparts, requiring approximately five weeks at ambient temperature to spontaneously revert to the trans-isomer.

  Figure 8

  

  Figure 8.  (a) Photoswitching between trans-L6 and cis-L6. (b) Photoswitching between helical trans-C6 and tetrahedral cis-C6. Copied with permission [85]. Copyright 2024, the Royal Society of Chemistry.

  DownLoad: Full-Size Img PowerPoint

  Beves and colleagues presented an example of light-induced switching between discrete structural assemblies [86]. A ligand functionalized with azobenzene-appended pyridyl groups (Fig. 9a) was combined with Pd(Ⅱ) to form two distinct structural motifs: a double-walled triangular assembly (C7-R) and a distorted tetrahedral assembly (C7-T) (Fig. 9b). Upon irradiation with 410 nm light, the tetrahedral structure disassembled, yielding the more kinetically stable triangular assembly, C7-R. This assembly slowly reverted to the tetrahedral form at 333 K. Interestingly, during these structural transformations, no well-defined assemblies of the cis-azobenzene isomer were observed. Both C7-R and C7-T exclusively contained the trans-azobenzene isomer. This behavior can be attributed to the photostationary state induced by 410 nm irradiation, in which approximately 80% of the ligands adopt the trans-isomer configuration. This isomeric composition favors the disassembly of the relatively unstable tetrahedral structure and promotes the thermodynamically stable triangular assembly. In contrast, irradiation with 530 nm light resulted in a mixture containing only about 20% trans-isomer ligands, which was insufficient to stabilize the triangular assembly, leading to an ill-defined mixture (Fig. 9c). At 333 K, the triangular assembly gradually reverted to the tetrahedral form, while the cis-isomer ligands failed to adopt well-defined structures. This study introduces a novel strategy for amplifying subtle changes in isomer distribution to induce significant compositional changes within the system, providing valuable insights into the design of light-responsive supramolecular architectures.

  Figure 9

  

  Figure 9.  (a) Photoswitching between trans-L7 and cis-L7. (b) Proposed structures of the double-walled triangle C7-R and tetrahedron C7-T. (c) Schematic illustrations of photoresponsive structural changes of C7-R and C7-T. Copied with permission [86]. Copyright 2022, Wiley Publishing Group.

  DownLoad: Full-Size Img PowerPoint

  Furthermore, Beves and colleagues demonstrated photoswitchable catalysis MOCs [87]. By utilizing azophenyl ligand L8-A in combination with the shape-complementary ligand L8-B, the heteroleptic cage C8-AB was successfully assembled. Alternatively, C8-AB can also be obtained by mixing and heating the pre-formed homoleptic cages C8-A and C8-B1/C8-B2. Due to specific electrostatic interactions between the substrate and the cavity of the heteroleptic cage, only C8-AB exhibited effective substrate binding and subsequent catalytic activity. In contrast, homoleptic cages constructed from identical ligands (C8-A and C8-B1/C8-B2) showed neither efficient substrate binding nor catalytic activity. Notably, the heteroleptic cage preserved the intrinsic photoresponsive characteristics of the azobenzene moieties: upon irradiation with 530 nm light, the cage disassembled through photoisomerization, resulting in a concomitant reduction in catalytic efficiency. Re-irradiation with 405 nm light for 5 min promoted cage reassembly, restoring catalytic efficiency to levels comparable to those before 530 nm irradiation (Fig. 10). This phenomenon conclusively demonstrates that catalytic activity can be precisely and reversibly modulated between "on" and "off" states through optical manipulation. These findings introduce a novel paradigm for regulating catalytic activity within MOC cavities using non-destructive visible light irradiation.

  Figure 10

  

  Figure 10.  Schematic illustrations of self-assembly of homoleptic cages C8-A, C8-B1 and C8-B2, and photoswitchable heteroleptic cage catalyst C8-AB. Copied with permission [87]. Copyright 2024, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  Diazocines represent a novel class of azophenyl-based photoswitches with distinct characteristics compared to conventional azobenzene derivatives [88,89]. Unlike traditional azobenzenes, the cis-isomer of diazocines is thermodynamically more stable. Upon irradiation at 385 nm, diazocines undergo photoisomerization to form a metastable trans-isomer, which exhibits a thermal lifetime of several hours at ambient temperature [90]. Moreover, diazocines feature a more rigid molecular architecture and exhibit more precise photoresponsive behavior than their azobenzene counterparts.

  In a significant advancement, Clever and colleagues reported an example of an M₂L₄-type MOC (C9) [91], constructed from a diazocine-incorporated ligand (L9, Fig. 11a) and Pd(Ⅱ) ions. The cage structure, C9, retains the intrinsic photoswitching capabilities of the diazocine unit, transforming into a less stable trans-isomer upon UV irradiation. Notably, the inherent rigidity of the cyclic diazocine structure restricts the extent of conformational change, enabling the coordination geometry with Pd(Ⅱ) ions to remain intact even after photoisomerization. While the cage undergoes structural modification, it remains intact without disassembly, with its cavity expanding as a result of the photochemical transformation, thereby facilitating guest molecule encapsulation. Under dark conditions, the cage undergoes thermal relaxation, reverting to its original configuration and subsequently releasing the encapsulated guest molecules (Fig. 11b). This formation of a transient host-guest complex exemplifies dissipative self-assembly, as the metastable state is strictly dependent on continuous UV irradiation for its maintenance.

  Figure 11

  

  Figure 11.  (a) Photoswitching between cis-L9 and trans-L9. (b) Capture and release of guest molecules via isomerization between cis-C9 and trans-C9. Copied with permission [91]. Copyright 2022, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  Herges and colleagues [92] utilized diazocine as a structural framework, strategically modifying coordination vectors to synthesize ligands L10 and L11, which initially exhibited bite angles of 61° and 63°, respectively. Following cis-to-trans isomerization, these angles increased significantly to 137° and 110°, respectively (Fig. 12a). Notably, cis-L10 and trans-L11 were capable of coordinating with Co(Ⅱ) ions to form distinct M₂L₃ MOCs, cis-C10 and trans-C11, respectively (Fig. 12b). Under ambient conditions without UV irradiation, both ligands predominantly adopted the cis-configuration, facilitating the formation of the cage cis-C10 from cis-L10. Upon exposure to 385 nm light, both ligands underwent photoisomerization to their metastable trans-states, enabling the assembly of cage trans-C11 from trans-L11. Subsequent irradiation with 520 nm light completely restored the system to its pre-385 nm irradiation state (Fig. 12c). Remarkably, the structural interconversion between these two states exhibited exceptional fatigue resistance, maintaining stability over at least 20 switching cycles. This work demonstrates an example of reversible, light-mediated transformation between two distinct MOC architectures derived from different ligand configurations.

  Figure 12

  

  Figure 12.  Schematic illustrations of (a) photoswitchable diazocine-based ligands L10 and L11, (b) structures of cis-C10 and trans-C11, and (c) reversible assembly and disassembly of cis-C10 and trans-C11. Copied with permission [92]. Copyright 2022, Wiley Publishing Group.

  DownLoad: Full-Size Img PowerPoint

  4. Conclusion

  In this review, we provide a comprehensive summary of significant advancements in the field of azophenyl-based photoresponsive MOCs. Through the strategic incorporation of photoresponsive functional groups, these MOCs can be precisely and efficiently modulated by light irradiation, offering a clean and non-invasive external stimulus. These photoresponsive systems show significant potential for a wide range of applications. For instance, light-mediated modulation of the internal chemical environment and solubility in photoresponsive 'molecular nanoparticles' could enable the development of advanced drug delivery systems with enhanced precision and control. By harnessing the reversible structural and functional changes induced by light, these systems can achieve targeted, on-demand release of therapeutic agents and may be integrated with polymer films or hydrogels to create composite materials with tailored properties. Furthermore, the combination of selective guest binding within the cage architecture and the ability to control encapsulation and release through non-destructive photochemical methods provides an efficient strategy for the separation and purification of valuable compounds. In addition, employing distinct coordination cages to catalyze specific chemical transformations, alongside light-mediated regulation, offers a novel strategy for designing sequential catalytic systems. By using light to adjust the properties and structure of the cage cavity, it becomes possible to perform multiple catalytic reactions in sequence, enabling the stepwise conversion of starting materials to final products without intermediate isolation. These advances open promising avenues in organic synthesis and green chemistry and are expected to drive further innovation in the design and application of photoresponsive MOCs.

  However, azophenyl-based MOCs still face several significant challenges, the most notable of which is their relatively low photoisomerization efficiency. Incomplete photoisomerization introduces uncertainties in structural transformations and compromises the efficacy of guest binding and release processes, thereby directly limiting their practical applicability. Furthermore, the fatigue resistance and recyclability of these photoresponsive systems remain critical factors that require thorough investigation and optimization. In conclusion, the integration of photoresponsive ligands into metal-organic cage architectures has greatly expanded their structural diversity and functional adaptability, demonstrating substantial potential for both fundamental research and practical applications. Addressing the current challenges through in-depth investigations is essential for advancing this field, and further research is crucial for translating fundamental discoveries into viable practical applications.

  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

  Xin Zhang: Writing – review & editing, Writing – original draft, Conceptualization. Tongxia Jin: Conceptualization. Changyin Yang: Conceptualization. Dezheng Xu: Writing – review & editing. Haidong Jia: Writing – review & editing. Lin Xu: Writing – review & editing, Conceptualization.

  Acknowledgments

  This work was supported by the National Nature Science Foundation of China (No. 22401096) and the Fundamental Research Funds for the Central Universities (East China Normal University).

  
  
• 1. [1]

     D. Fujita, Y. Ueda, S. Sato, et al., Nature 540 (2016) 563–566. doi: 10.1038/nature20771

  2. [2]

     A.M. Castilla, W.J. Ramsay, J.R. Nitschke, Acc. Chem. Res. 47 (2014) 2063–2073. doi: 10.1021/ar5000924

  3. [3]

     R. Chakrabarty, P.S. Mukherjee, P.J. Stang, Chem. Rev. 111 (2011) 6810–6918. doi: 10.1021/cr200077m

  4. [4]

     B. Sun, M. Wang, Z. Lou, et al., J. Am. Chem. Soc. 137 (2015) 1556–1564. doi: 10.1021/ja511443p

  5. [5]

     T.R. Cook, Y.R. Zheng, P.J. Stang, Chem. Rev. 113 (2013) 734–777. doi: 10.1021/cr3002824

  6. [6]

     R. Cué-Sampedro, J.A. Sánchez-Fernández, Molecules 30 (2025) 462. doi: 10.3390/molecules30030462

  7. [7]

     M. Yoshizawa, J.K. Klosterman, M. Fujita, Angew. Chem. Int. Ed. 48 (2009) 3418–3438. doi: 10.1002/anie.200805340

  8. [8]

     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

  9. [9]

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

  10. [10]

      S.C. Li, L.X. Cai, M. Hong, Q. Chen, Q.F. Sun, Angew. Chem. Int. Ed. 61 (2022) e202204732. doi: 10.1002/anie.202204732

  11. [11]

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

  12. [12]

      J. Zhao, K. Zeng, W.T. Dou, et al., Adv. Funct. Mater. 35 (2025) 2413920. doi: 10.1002/adfm.202413920

  13. [13]

      H. Liu, C. Guo, Y. Huang, et al., Chem. Sci. 15 (2024) 14352–14357. doi: 10.1039/d4sc02736a

  14. [14]

      Y. Wang, Y. Qin, X. Zhao, et al., Chin. Chem. Lett. 34 (2023) 107576. doi: 10.1016/j.cclet.2022.05.090

  15. [15]

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

  16. [16]

      Z. Wang, S. Furukawa, Acc. Chem. Res. 57 (2024) 327–337. doi: 10.1021/acs.accounts.3c00655

  17. [17]

      G.R. Lorzing, B.A. Trump, C.M. Brown, E.D. Bloch, Chem. Mater. 29 (2017) 8583–8587. doi: 10.1021/acs.chemmater.7b03361

  18. [18]

      Q.Q. Wang, S. Gonell, S.H.A.M. Leenders, et al., Nat. Chem. 8 (2016) 225–230. doi: 10.1038/nchem.2425

  19. [19]

      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

  20. [20]

      Q.W. Zeng, L. Hu, Y. Niu, et al., Chem. Commun. 60 (2024) 3469–3483. doi: 10.1039/d3cc05935a

  21. [21]

      L.J. Wang, S. Bai, Y.F. Han, J. Am. Chem. Soc. 144 (2022) 16191–16198. doi: 10.1021/jacs.2c07586

  22. [22]

      L.J. Wang, Z.E. Zhang, Y.Z. Zhang, Y.F. Han, Angew. Chem. Int. Ed. 63 (2024) e202407278. doi: 10.1002/anie.202407278

  23. [23]

      C. García-Simón, M. Garcia-Borràs, L. Gómez, et al., Nat. Commun. 5 (2014) 5557. doi: 10.1038/ncomms6557

  24. [24]

      M. Morimoto, S.M. Bierschenk, K.T. Xia, et al., Nat. Catal. 3 (2020) 969–984. doi: 10.1038/s41929-020-00528-3

  25. [25]

      Z. Ashbridge, J.N.H. Reek, Nat. Synth. 3 (2024) 1197–1207. doi: 10.1038/s44160-024-00606-5

  26. [26]

      M. Yoshizawa, M. Tamura, M. Fujita, Science 312 (2006) 251–254. doi: 10.1126/science.1124985

  27. [27]

      M.D. Pluth, R.G. Bergman, K.N. Raymond, Science 316 (2007) 85–88. doi: 10.1126/science.1138748

  28. [28]

      Y. Ueda, H. Ito, D. Fujita, M. Fujita, J. Am. Chem. Soc. 139 (2017) 6090–6093. doi: 10.1021/jacs.7b02745

  29. [29]

      Y. Liu, L. Huang, L. Qin, T. Zhou, ChemCatChem 17 (2025) e202401487. doi: 10.1002/cctc.202401487

  30. [30]

      M.D. Ward, Chem. Commun. 60 (2024) 10464–10475. doi: 10.1039/d4cc03678f

  31. [31]

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

  32. [32]

      R. Saha, B. Mondal, P.S. Mukherjee, Chem. Rev. 122 (2022) 12244–12307. doi: 10.1021/acs.chemrev.1c00811

  33. [33]

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

  34. [34]

      Q.H. Ling, Z.C. Lou, L. Zhang, et al., Chem. Soc. Rev. 53 (2024) 6042–6067. doi: 10.1039/d3cs01081c

  35. [35]

      P. Abhik, R. Subhadip, Curr. Org. Chem. 28 (2024) 941–958. doi: 10.2174/0113852728292501240301062823

  36. [36]

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

  37. [37]

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

  38. [38]

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

  39. [39]

      B. Song, Z. Zhang, W. Dou, et al., Angew. Chem. Int. Ed. 64 (2025) e202414089. doi: 10.1002/anie.202414089

  40. [40]

      A.J. McConnell, C.S. Wood, P.P. Neelakandan, J.R. Nitschke, Chem. Rev. 115 (2015) 7729–7793. doi: 10.1021/cr500632f

  41. [41]

      P.J. Lusby, P. Müller, S.J. Pike, A.M.Z. Slawin, J. Am. Chem. Soc. 131 (2009) 16398–16400. doi: 10.1021/ja907297z

  42. [42]

      T.Y. Kim, R.A.S. Vasdev, D. Preston, J.D. Crowley, Chem. Eur. J. 24 (2018) 14878–14890. doi: 10.1002/chem.201802081

  43. [43]

      W. Wang, Y.X. Wang, H.B. Yang, Chem. Soc. Rev. 45 (2016) 2656–2693. doi: 10.1039/C5CS00301F

  44. [44]

      J. Zhao, Z. Zhou, G. Li, P.J. Stang, X. Yan, Natl. Sci. Rev. 8 (2021) nwab045. doi: 10.1093/nsr/nwab045

  45. [45]

      Y. Gu, E.A. Alt, H. Wang, et al., Nature 560 (2018) 65–69. doi: 10.1038/s41586-018-0339-0

  46. [46]

      D.H. Qu, Q.C. Wang, Q.W. Zhang, X. Ma, H. Tian, Chem. Rev. 115 (2015) 7543–7588. doi: 10.1021/cr5006342

  47. [47]

      R.J. Li, J. Tessarolo, H. Lee, G.H. Clever, J. Am. Chem. Soc. 143 (2021) 3865–3873. doi: 10.1021/jacs.0c12188

  48. [48]

      R.J. Li, J.J. Holstein, W.G. Hiller, J. Andréasson, G.H. Clever, J. Am. Chem. Soc. 141 (2019) 2097–2103. doi: 10.1021/jacs.8b11872

  49. [49]

      P. Mal, D. Schultz, K. Beyeh, K. Rissanen, J.R. Nitschke, Angew. Chem. Int. Ed. 47 (2008) 8297–8301. doi: 10.1002/anie.200803066

  50. [50]

      L. Xu, D. Zhang, T.K. Ronson, J.R. Nitschke, Angew. Chem. Int. Ed. 59 (2020) 7435–7438. doi: 10.1002/anie.202001059

  51. [51]

      W. Cullen, S. Turega, C.A. Hunter, M.D. Ward, Chem. Sci. 6 (2015) 625–631. doi: 10.1039/C4SC02090A

  52. [52]

      S.M. Jansze, G. Cecot, K. Severin, Chem. Sci. 9 (2018) 4253–4257. doi: 10.1039/c8sc01108g

  53. [53]

      L.X. Cai, D.N. Yan, P.M. Cheng, et al., J. Am. Chem. Soc. 143 (2021) 2016–2024. doi: 10.1021/jacs.0c12064

  54. [54]

      D. Zhang, T.K. Ronson, S. Güryel, et al., J. Am. Chem. Soc. 141 (2019) 14534–14538. doi: 10.1021/jacs.9b07307

  55. [55]

      I.A. Riddell, M.M.J. Smulders, J.K. Clegg, et al., Nat. Chem. 4 (2012) 751–756. doi: 10.1038/nchem.1407

  56. [56]

      D. Zhang, T.K. Ronson, L. Xu, J.R. Nitschke, J. Am. Chem. Soc. 142 (2020) 9152–9157. doi: 10.1021/jacs.0c03798

  57. [57]

      D. Preston, A. Fox-Charles, W.K.C. Lo, J.D. Crowley, Chem. Commun. 51 (2015) 9042–9045. doi: 10.1039/C5CC02226F

  58. [58]

      J.K. Clegg, J. Cremers, A.J. Hogben, et al., Chem. Sci. 4 (2013) 68–76. doi: 10.1039/C2SC21486E

  59. [59]

      S.P. Zheng, Y.W. Xu, P.Y. Su, et al., Chin. Chem. Lett. 35 (2024) 108477. doi: 10.1016/j.cclet.2023.108477

  60. [60]

      J. Zhao, L. Cheng, K. Liu, et al., Chem. Commun. 56 (2020) 8031–8034. doi: 10.1039/d0cc01205j

  61. [61]

      J. Liang, B. Wu, C. Jia, X.J. Yang, CrystEngComm 11 (2009) 975–977. doi: 10.1039/b820640f

  62. [62]

      T. Murase, S. Sato, M. Fujita, Angew. Chem. Int. Ed. 46 (2007) 5133–5136. doi: 10.1002/anie.200700793

  63. [63]

      M. Han, R. Michel, B. He, et al., Angew. Chem. Int. Ed. 52 (2013) 1319–1323. doi: 10.1002/anie.201207373

  64. [64]

      C. Stuckhardt, D. Roke, W. Danowski, et al., Beilstein J. Org. Chem. 15 (2019) 2767–2773. doi: 10.3762/bjoc.15.268

  65. [65]

      H. Wang, C.N. Zhu, H. Zeng, et al., Adv. Mater. 31 (2019) 1807328. doi: 10.1002/adma.201807328

  66. [66]

      S. Wu, H.J. Butt, Macromol. Rapid Commun. 41 (2020) 1900413. doi: 10.1002/marc.201900413

  67. [67]

      P. Weis, S. Wu, Macromol. Rapid Commun. 39 (2018) 1700220. doi: 10.1002/marc.201700220

  68. [68]

      W.C. Xu, S. Sun, S. Wu, Angew. Chem. Int. Ed. 58 (2019) 9712–9740. doi: 10.1002/anie.201814441

  69. [69]

      W.T. Dou, Q.W. Zeng, Y. Kang, et al., Chin. Chem. Lett. 36 (2025) 109995. doi: 10.1016/j.cclet.2024.109995

  70. [70]

      L. Zhang, H.X. Wang, S. Li, M. Liu, Chem. Soc. Rev. 49 (2020) 9095–9120. doi: 10.1039/d0cs00191k

  71. [71]

      M. Younis, S. Ahmad, A. Atiq, et al., Chem. Rec. 23 (2023) e202300126. doi: 10.1002/tcr.202300126

  72. [72]

      H. Ren, P. Yang, H. Yu, Molecules 27 (2022) 3977. doi: 10.3390/molecules27133977

  73. [73]

      H. Yao, Z. Yang, X. Fan, et al., Mater. Chem. Front. 3 (2019) 1168–1173. doi: 10.1039/c9qm00141g

  74. [74]

      Demselben, Ann. Pharm. 12 (1834) 311–314. doi: 10.1002/jlac.18340120282

  75. [75]

      G.S. Kumar, D.C. Neckers, Chem. Rev. 89 (1989) 1915–1925. doi: 10.1021/cr00098a012

  76. [76]

      H.M.D. Bandara, S.C. Burdette, Chem. Soc. Rev. 41 (2012) 1809–1825. doi: 10.1039/C1CS15179G

  77. [77]

      A. Ghosh, J. Pruchyathamkorn, C. Fuertes Espinosa, J.R. Nitschke, J. Am. Chem. Soc. 146 (2024) 2568–2573. doi: 10.1021/jacs.3c11005

  78. [78]

      J. Park, L.B. Sun, Y.P. Chen, Z. Perry, H.C. Zhou, Angew. Chem. Int. Ed. 53 (2014) 5842–5846. doi: 10.1002/anie.201310211

  79. [79]

      Y. Jiang, J. Park, P. Tan, et al., J. Am. Chem. Soc. 141 (2019) 8221–8227. doi: 10.1021/jacs.9b01380

  80. [80]

      R.G. DiNardi, A.O. Douglas, R. Tian, et al., Angew. Chem. Int. Ed. 61 (2022) e202205701. doi: 10.1002/anie.202205701

  81. [81]

      P. Cecot, A. Walczak, G. Markiewicz, A.R. Stefankiewicz, Inorg. Chem. Front. 8 (2021) 5195–5200. doi: 10.1039/d1qi01063h

  82. [82]

      S. Oldknow, D.R. Martir, V.E. Pritchard, et al., Chem. Sci. 9 (2018) 8150–8159. doi: 10.1039/c8sc03499k

  83. [83]

      E. Britton, R.J. Ansell, M.J. Howard, M.J. Hardie, Inorg. Chem. 60 (2021) 12912–12923. doi: 10.1021/acs.inorgchem.1c01297

  84. [84]

      S. Fu, Q. Luo, M. Zang, et al., Mater. Chem. Front. 3 (2019) 1238–1243. doi: 10.1039/c9qm00160c

  85. [85]

      M.B. Tipping, L. Pruñonosa Lara, A.B. Solea, L.K.S. von Krbek, M.D. Ward, Chem. Sci. 15 (2024) 8488–8499. doi: 10.1039/d4sc01575d

  86. [86]

      A.D.W. Kennedy, R.G. DiNardi, L.L. Fillbrook, W.A. Donald, J.E. Beves, Chem. Eur. J. 28 (2022) e202104461. doi: 10.1002/chem.202104461

  87. [87]

      R.G. DiNardi, S. Rasheed, S.S. Capomolla, et al., J. Am. Chem. Soc. 146 (2024) 21196–21202. doi: 10.1021/jacs.4c04846

  88. [88]

      W. Moormann, D. Langbehn, R. Herges, Beilstein J. Org. Chem. 15 (2019) 727–732. doi: 10.3762/bjoc.15.68

  89. [89]

      A. Mukherjee, M.D. Seyfried, B.J. Ravoo, Angew. Chem. Int. Ed. 62 (2023) e202304437. doi: 10.1002/anie.202304437

  90. [90]

      R. Siewertsen, H. Neumann, B. Buchheim-Stehn, et al., J. Am. Chem. Soc. 131 (2009) 15594–15595. doi: 10.1021/ja906547d

  91. [91]

      H. Lee, J. Tessarolo, D. Langbehn, et al., J. Am. Chem. Soc. 144 (2022) 3099–3105. doi: 10.1021/jacs.1c12011

  92. [92]

      D. Hugenbusch, M. Lehr, J.S. von Glasenapp, A.J. McConnell, R. Herges, Angew. Chem. Int. Ed. 62 (2023) e202212571. doi: 10.1002/anie.202212571

• 

  Figure 1  Schematic illustrations of photoresponsive azobenzene-containing MOCs.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Self-assembly of M₁₂L₂₄ spherical complexes C1 with 24 endohedral azobenzene groups (L1a) or 24 methoxy groups (L1b) and hydrophobic guest. Copied with permission [62]. Copyright 2007, Wiley Publishing Group.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Self-assembly of photoresponsive cage C2. (b) Steroid guests investigated for binding within C2, progesterone represented as red circles and green, blue and orange circles represented mestranol, cholesterol and 7-dehydrocholesterol respectively. (c) Illustration of progesterone separation from mixture steroid guests. Copied with permission [77]. Copyright 2024, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Light-induced trans/cis-isomerization of L3. (b) Structure of C3. (c) Schematic illustrations of reversible formation of insoluble trans-C3 and soluble cis-C3 with capture and release of MB. Copied with permission [78]. Copyright 2014, Wiley Publishing Group.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) Schematic illustrations of light-induced reversible trans/cis-isomerization of C3. (b) Only a few C3 units at the edges exhibit photoreponsiveness in the solid state. (c) C3 dispersed in the nanoscaled spaces of mesoporous silica. Copied with permission [79]. Copyright 2019, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) Schematic illustrations of synthesis of C4 and energy-minimized structure. (b) Energy-minimized structure of C4 with differing numbers of cis-isomerization ligand arms. Copied with permission [82]. Copyright 2018, the Royal Society of Chemistry.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Structures of C5 and L5, along with their light-induced self-assembly and disassembly. Reproduced with permission [84]. Copyright 2019, the Royal Society of Chemistry.

  下载: 全尺寸图片 幻灯片
  

  Figure 8  (a) Photoswitching between trans-L6 and cis-L6. (b) Photoswitching between helical trans-C6 and tetrahedral cis-C6. Copied with permission [85]. Copyright 2024, the Royal Society of Chemistry.

  下载: 全尺寸图片 幻灯片
  

  Figure 9  (a) Photoswitching between trans-L7 and cis-L7. (b) Proposed structures of the double-walled triangle C7-R and tetrahedron C7-T. (c) Schematic illustrations of photoresponsive structural changes of C7-R and C7-T. Copied with permission [86]. Copyright 2022, Wiley Publishing Group.

  下载: 全尺寸图片 幻灯片
  

  Figure 10  Schematic illustrations of self-assembly of homoleptic cages C8-A, C8-B1 and C8-B2, and photoswitchable heteroleptic cage catalyst C8-AB. Copied with permission [87]. Copyright 2024, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 11  (a) Photoswitching between cis-L9 and trans-L9. (b) Capture and release of guest molecules via isomerization between cis-C9 and trans-C9. Copied with permission [91]. Copyright 2022, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 12  Schematic illustrations of (a) photoswitchable diazocine-based ligands L10 and L11, (b) structures of cis-C10 and trans-C11, and (c) reversible assembly and disassembly of cis-C10 and trans-C11. Copied with permission [92]. Copyright 2022, Wiley Publishing Group.

  下载: 全尺寸图片 幻灯片
•