Catenated Cages Mediated by Enthalpic Reaction Intermediates

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Jul 2021Catenated Cages Mediated by Enthalpic Reaction Intermediates Shijun Xu†, Pan Li†, Zi-Ying Li, Chunyang Yu, Xiaoyun Liu, Zhiqiang Liu and Shaodong Zhang Shijun Xu† Key Laboratory of Specially Functional Polymeric Materials and Related Technology, Ministry of Education, East China University of Science and Technology, Shanghai 200237 , Pan Li† School of Chemistry and Chemical Engineering, Frontiers Science Center for Transformative Molecules, Shanghai Key Laboratory of Electrical Insulation and Thermal Aging, Shanghai Jiao Tong University, Shanghai 200240 , Zi-Ying Li Key Laboratory of Specially Functional Polymeric Materials and Related Technology, Ministry of Education, East China University of Science and Technology, Shanghai 200237 , Chunyang Yu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] School of Chemistry and Chemical Engineering, Frontiers Science Center for Transformative Molecules, Shanghai Key Laboratory of Electrical Insulation and Thermal Aging, Shanghai Jiao Tong University, Shanghai 200240 , Xiaoyun Liu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Key Laboratory of Specially Functional Polymeric Materials and Related Technology, Ministry of Education, East China University of Science and Technology, Shanghai 200237 , Zhiqiang Liu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] The National and Local Joint Engineering Research Center for Biomanufacturing of Chiral Chemicals, Zhejiang University of Technology, Hangzhou 310014 and Shaodong Zhang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] School of Chemistry and Chemical Engineering, Frontiers Science Center for Transformative Molecules, Shanghai Key Laboratory of Electrical Insulation and Thermal Aging, Shanghai Jiao Tong University, Shanghai 200240 https://doi.org/10.31635/ccschem.020.202000360 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesTrack Citations ShareFacebookTwitterLinked InEmail Catenated cages are generally considered thermodynamically more stable than their constituent monomeric cages. However, the catenation mechanism is yet to be elucidated; it would require systematic investigation into the structural effects of the building blocks, their enthalpic and entropic contributions, and the effect of solvents. By inspecting these factors, we rationalized some design principles for the efficient construction of catenated cages. Our study revealed that a steric hindrance linker and a rigid panel led to the formation of an enthalpy-favored encapsulated intermediate before catenation occurred. The stability of this enthalpic intermediate was crucial for cage catenation, as the reactions were otherwise outcompeted by an entropy-favored intermediate. The formation of the latter was facilitated significantly by a flexible panel and solvent molecules that stably resided within the monomeric cage. This study provides a guideline for the elaboration of catenated cages with more sophisticated topologies, which could be extended to other complex supramolecular assemblies. Download figure Download PowerPoint Introduction Since the seminal works of Sauvage et al.1 and Stoddart et al.,2 mechanically interlocked structures with various topologies have been studied widely3–6 and used to construct molecular switches,7–9 motors,10–12 and machines13–16; also, it could be applied further to discover novel chemistry, materials, and medicines. Among these interlocked molecules, catenanes, which are composed of two or more monomeric cages, are appealing. Fujita et al.17–19 reported the first series of triply interpenetrated metal–organic cages. By applying a similar principle of ligand–metal complexation, several triply,20–23 quadruply,24–27 and sophisticated catenated coordination cages28–31 have been elaborated. Since 2010, catenanes consisting of covalent cages started attracting considerable attention, with some examples reported recently by Cooper, Mastalerz, Sessler, Li and Zhang,32–36 which were typically synthesized via dynamic covalent chemistry. It is believed that both coordination and covalent catenated cages are thermodynamically more stable than the constituent monomeric cages.17,24,32–35,37 However, this presumption was made with limited molecular models—very often with only one example—in each previous work. To rationalize the design of catenated cages, a systematic investigation of their underlying formation mechanism is, therefore, required. Several studies have been conducted elegantly concerning controlling the interchange between monomeric cages and their interlocked dimers. For example, Kuroda group reported that the catenation of a quadruply metallohelicate could be modulated effectively through auxiliary anion templates. Clever et al.38 also demonstrated the bulky derivatization of ligands with templating anions posed dramatic influence on the catenation of coordination cages. By contrast, little information is available regarding the catenation mechanism of covalent cages. Our research group has been focusing on the elaboration of catenated cages,39 which has been applied as novel building blocks for the search of emerging hierarchical superstructures.40 During this endeavor, recently, we found the examples that seemingly contradicted this long-standing conjecture. Thus, it prompted us to examine carefully the hitherto underexplored mechanism of cage catenation,41 which required systematic investigation into a panoply of factors such as the structural effects of the building blocks, their enthalpic and entropic contributions, and solvent effect. Herein, we sought to elaborate the correlation between the structural features of the building blocks, that is, linkers and panels in our study, and the product distribution of covalent catenated and monomeric cages (Scheme 1). The steric effect of the linkers and the rigidity/flexibility of the panels would be elucidated. By considering their enthalpic and entropic contributions, and the impact of solvent molecules, we also thought of gaining a thermodynamic insight into the formation mechanism of catenated cages by combining computational and experimental approaches to explore the production distribution mediated by specific reaction intermediates. Scheme 1 | Summary of product distribution of CSCs and the corresponding MCs produced by cycloimination of diamine linkers, that is, linker 1 or linker 2 and trialdehyde panels, that is, TFB or TFPB in this study. CSC, catenanes with symmetric cage; MC, monomeric catenanes; linker 1, 2-[3,5-di(3-aminophenyl)phenyl]-1,3-dithiane; linker 2, 3-[3-(3-aminophenyl)phenyl]aniline; TFB, 1,3,5-triformylbenzene; TFPB, 1,3,5-tris(p-formylphenyl)benzene. Download figure Download PowerPoint Experimental Methods All the catenanes with symmetric cages ( CSCs) and monomeric catenanes ( MCs) were synthesized through cycloimination between diamine linkers (3 equiv) and trialdehyde panels (2 equiv), using trifluoroacetic acid (TFA; 0.1 equiv) as catalyst and chloroform (CHCl3) as a solvent. The yields of all products were calculated in situ using ethyl acetate (EtOAc) as an internal standard calibration ( Supporting Information Figures S1–S3), and their structures were characterized by nuclear magnetic resonance (NMR), matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and single-crystal X-ray diffraction (SC-XRD), which are shown in detail in Supporting Information. Theoretical calculation of all compounds was performed on the Gaussian09 programS7 at B3LYP/6-31G* without any symmetry restriction. The detailed methods are available in Supporting Information Section 10. Results and Discussion The effect of building blocks on the formation of CSCs First, we synthesized a catenane with two symmetric cages, namely CSC-1, according to our previously reported procedure.39 This was achieved by reversible cycloimination reaction between 2-[3,5-di(3-aminophenyl)phenyl]-1,3-dithiane (denoted diamine linker 1) and 1,3,5-triformylbenzene (denoted trialdehyde panel TFB) using 0.1 equimolar TFA in CHCl3 as a solvent mixture (Scheme 1). The reaction yielded a soluble mixture of predominating CSC-1 (63% yield), monomeric cage MC-1 (30% yield), and a small amount of precipitate that was presumably the corresponding oligo- or polyimines with ill-defined structures.42–44 Then we conducted the reaction of 3-[3-(3-aminophenyl)phenyl]aniline (linker 2) and TFB under the same conditions (Scheme 1 and Figures 1a–1c). Although linker 2 was less sterically hindered, compared with linker 1 with an extra dithiane moiety, it was mainly converted into irreversible polymeric precipitate (56 mol %). This was significantly different from the reaction with linker 1 (vide supra). The soluble catenane CSC-2 and monomeric cage MC-2 in the reaction solution were respectively, isolated and characterized by MALDI-TOF MS and NMR spectroscopy (Figures 1a–1c). Figure 1 | (a) Cycloimination reaction of linker 2 and TFB to yield monomeric cage MC-2 and catenated cage CSC-2. (b) MALDI-TOF MS (left) and 1H NMR (right, 500 MHz, CDCl3, 298 K) spectra of MC-2. (c) MALDI-TOF MS (left) and 1H NMR (right, 500 MHz, CDCl3, 298 K) spectra of CSC-2. Linker 2, 3-[3-(3-aminophenyl)phenyl]aniline; TFB, 1,3,5-triformylbenzene; MALDI-TOF MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry; 1H NMR, proton nuclear magnetic resonance. Download figure Download PowerPoint It is worth noting that even though the in situ yields of MC-2 (28% yield) was similar to that of MC-1 (30% yield), the reaction with linker 2 was less efficient in yielding catenated species CSC-2 (16% yield) (see Supporting Information Figure S1). This is in line with the large amount of precipitate formed by linker 2 (56 mol %), as the precipitate prevents its reversible and dynamic self-correction into the target catenated product, which is generally considered as a kinetic trap.41 By comparing the two structures of linkers 1 and 2, we expected that linker 2 tended to aggregate by intermolecular stacking, while the introduction of the 1,3-dithiane substituent would prevent such aggregation and improve the solubility of corresponding intermediates. These results, therefore, demonstrated the structural effect of the linkers on the product distribution of monomeric and catenated cages. Also, we attempted to synthesize catenated cages by performing the reaction with a more flexible trialdehyde panel, 1,3,5-tris(p-formylphenyl)benzene ( TFPB) with linker 1 and linker 2, respectively (Figures 2a–2c). However, the formation of catenated dimers CSC-3 and CSC-4 (see their structures in Scheme 1) was negligible for >100 h, and they could only be identified by MALDI-TOF MS analysis of the reaction solution (see Supporting Information Figures S7 and S8), but not by NMR. These results were seemly different from the general belief that catenated structures are thermodynamically more stable over their monomeric counterpart cages,17–19,32–36 the reason for which would be elucidated in the following sections. Figure 2 | (a) Cycloimination reaction of TFPB with linker 1 or linker 2 to yield trace amount of CSC-3 or CSC-4, and the predominant product MC-3 or MC-4 that was subsequently reduced to MC-3 r or MC-4 r so as to facilitate the isolation and characterization processes. (b) MALDI-TOF MS (left) and 1H NMR (right, 500 MHz, CDCl3, 298 K) spectra of reduced monomeric cage MC-3 r. (c) MALDI-TOF MS (left) and 1H NMR (right, 500 MHz, CDCl3, 298 K) spectra of reduced monomeric cage MC-4 r. TFPB, 1,3,5-tris(p-formylphenyl)benzene; linker 1, 2-[3,5-di(3-aminophenyl)phenyl]-1,3-dithiane; linker 2, 3-[3-(3-aminophenyl)phenyl]aniline; MALDI-TOF MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry; 1H NMR, proton nuclear magnetic resonance. Download figure Download PowerPoint The reaction with linker 1 was completed within 24 h, yielding MC-3 (62% yield) as the only meaningful cage species (see Supporting Information Figure S2), concomitant with precipitate formed by 38 mol % of linker 1. On the other hand, the reaction with linker 2 was incomplete over 24 h (49 mol % of linker 2 remained). Thereafter, it was quenched to afford MC-4 (29% yield, see Supporting Information Figure S3), with 23 mol % of linker 2 converted to the precipitate. We tried to isolate MC-3 and MC-4 from their reaction solution, but our efforts were in vain; therefore, we subjected them to a reduction reaction using NaBH(OAc)3. The reduced monomeric cages, MC-3 r and MC-4 r, were isolated and characterized unambiguously by MALDI-TOF MS and NMR spectroscopy analyses (Figures 2a–2c). Previous reports have proven that the flexibility of building blocks is likely to lower the yield of shape-persistent organic cages.45 Such impact is seemingly more pronounced with the panel TFPB in this study, where two prismatic cages interlocked to form a catenated dimer. The comparison of the opposite outcomes resulting from rigid TFB and flexible TFPB indeed demonstrated the different structural effects of the panels on the formation possibility and presumably the formation mechanism of catenated cages. Conversion of monomeric to catenated cages, followed by kinetic study As discussed above, the cycloimine reaction between rigid TFB and sterically hindered linker 1 yielded both monomeric cage MC-1 and catenated dimer CSC-1.39 The kinetic study of the reaction revealed that 80 mol % of linker 1 was consumed in 30 min, yielding mainly MC-1 (69% yield), with a smaller fraction of CSC-1 (9% yield). Both precursors were consumed utterly within 1 h, producing a 63% yield of MC-1 and 19% yield of CSC-1. The monomeric cage MC-1 was then converted steadily to CSC-1 until an equilibrium is reached at 96 h, with minimal and maximal amounts of MC-1 (30% yield) and CSC-1 (63% yield) are produced (Figure 3a). Notably, a minute precipitate was formed throughout the reaction. Figure 3 | The conversions from monomeric cages to catenated cages, determined by in situ 1H NMR spectroscopy at different times given in Supporting Information Figures S14 and S15. The conversion profile of (a) MC-1 to CSC-1 and (b) MC-2 to CSC-2, respectively. The structures of MC-1, MC-2, CSC-1, and CSC-2 are present for reference. 1H NMR, proton nuclear magnetic resonance. Download figure Download PowerPoint Furthermore, we monitored the conversion from MC-2 to CSC-2 (Figure 3b). The reaction of less bulky linker 2 with TFB proceeded much faster than that with linker 1, as linker 2 was consumed within 10 min, producing the major product of 44% yield of MC-2 and only 3% yield of CSC-2, with 48 mol % of linker 2 being converted to the precipitate. Then MC-2 was gradually transformed into CSC-2 at equilibrium, with a 28% yield of MC-2 and 16% yield of CSC-2 within 2 h, during which the precipitate (56 mol %) was slightly increased by 8 mol %. This result revealed that irreversible oligo- or polyimines were mainly formed at the early stage of the reaction, and their conversion to the cage species was somewhat precluded. Alternatively, the kinetic evolutions of the reactions involving TFPB exhibited no conversion of MC-3 or MC-4 to the corresponding catenated dimers. By comparing the imination reactions of TFB with linker 1 and linker 2, we found that linker 1 favored the formation of a catenated dimer, as the yield ratio obtained for CSC-1/ MC-1 was 63%/30% at equilibrium, with the formation of a little precipitate, while linker 2 yielded more monomeric cage, as the ratio of CSC-2/ MC-2 was 16%/28%, and 56 mol % of linker 2 was irreversibly converted into precipitate. Regarding the effect of the panels, the outcome was also dramatically different. Apart from the incapacity of producing the catenated cages, the reactions with flexible TFPB were considerably slower than those with rigid TFB. The consumption of TFPB needed no less than 24 h (24 h for linker 1, and more than 24 h for linker 2), while TFB was consumed entirely within no more than 1 h (1 h for linker 1 and 10 min for linker 2). This kinetic difference was attributable to the lower electrophilicity of TFPB than that of TFB. Identification of reaction intermediates during cage formation The results mentioned above revealed different structural impacts of the linkers and panels on the formation possibility of the CSCs and the product distribution of MCs/ CSCs. We hypothesized that these distinct outcomes were caused by adverse competitive reactions during the cycloimination with different linkers and panels (Scheme 1), which could be probed by identifying the key reaction intermediates.39,46–48 However, as the intermediates involved in the imination condensations are dynamic and reversible, they are not directly accessible by synthesis. Therefore, we addressed this challenge by adding extra panels at the equilibrium of each reaction, hoping the cage entities could be broken or converted into the corresponding intermediates, and thus, the technique was denoted a cage-breaking experiment. These experiments allowed us to capture the key intermediates, and they were proved to mediate the product distribution of monomeric and catenated cages, demonstrated as follows. It was previously assumed that the inclusion of TFB into the inner cavity of MC-1 yielded the encapsulated species TFB⊂ MC-1, which might provide a template that facilitated the formation of CSC-1.39 Thereafter, the formation mechanism of CSC-1 was examined by adding more TFB, so as to vary the ratios of TFB/linker 1, at the equilibrium of the initial stoichiometric reaction ( TFB/linker 1 = 2/3, Figure 4a). When extra TFB was added into the reaction mixture, both CSC-1 and MC-1 declined gradually, while the encapsulated intermediate TFB⊂ MC-1 increased steadily as the dominating species (Figure 4b). This tendency continued until TFB/linker 1 = 12/3, when the trace of CSC-1 and MC-1 became undetectable by 1H NMR, but only by MALDI-TOF, with a peak at m/z 1514.413, referring to [M+H]+ for TFB⊂ MC-1 (Figure 4c and Supporting Information Figure S9). Thus, this observation showed that when TFB was excessive, the erstwhile mixture of CSC-1 and MC-1 shifted toward TFB⊂ MC-1. This equilibrium shift was a thermodynamically driven process, as the formation of TFB⊂ MC-1 was favored enthalpically due to the host–guest interaction through π–π stacking and the disassembly of CSC-1 into two monomeric MC-1, which were also entropically favored, thereby confirming a retrospective template effect. Otherwise, spontaneous conversion of TFB⊂ MC-1 to CSC-1 was hampered when the internal cavity of every preformed MC-1 was occupied by excessive TFB. Figure 4 | Cage-breaking experiment for probing the reaction species by adding excessive TFB at the equilibrium of the initial stoichiometric reaction (TFB/linker 1 = 2/3). (a) The schematic illustrates the encapsulated intermediate TFB⊂MC-1 was formed as the dominating species by adding excessive TFB, concomitant with the suppression of MC-1 and CSC-1. (b) Evolution of CSC-1, MC-1, and TFB⊂MC-1 with the ratio of TFB/linker 1 changed from 2/3 to 12/3, determined by 1H NMR spectroscopy given in Supporting Information Figure S16. (d) MALDI-TOF MS spectra of the reaction mixtures with a ratio of TFB/linker 1 at 2/3 and 12/3, respectively. CSC-1, MC-1, and TFB⊂MC-1 with the corresponding m/z values are highlighted by the arrows with the corresponding color codes. Linker 1, 2-[3,5-di(3-aminophenyl)phenyl]-1,3-dithiane; 1H NMR, proton nuclear magnetic resonance; MALDI-TOF MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Download figure Download PowerPoint When we introduced excessive TFB to the stoichiometric reaction of TFB and the less hindered linker 2 until the reaction reached equilibrium, we also examined the evolution of its reaction species (Figure 5a–5c). CSC-2 (16% yield), MC-2 (28% yield), and a trace amount of encapsulated intermediate TFB⊂ MC-2 were detected at the equilibrium of the original stoichiometric reaction by NMR and MALDI-TOF MS ( see Supporting Information Figures S6 and S10). To our surprise, when extra TFB was introduced to this mixture, TFB⊂ MC-2 vanished immediately, and both CSC-2 and MC-2 decreased gradually until their complete disappearance at TFB/linker 2 = 7/3. Meanwhile, the intermediate S1 with disordered structure evolved progressively as the only meaningful product, and its structure was confirmed further by 1H NMR spectroscopy (see Supporting Information Figure S19). This was distinct from the reaction with more sterically hindered linker 1, along with the different product distribution of CSC-2/ MC-2 = 16%/28%, compared with CSC-1/ MC-2 = 63%/30%. Hence, the results indicated that the formation preference of CSC-2 was hampered to some extent, compared with CSC-1. So far, even though it is challenging to clarify the role of the dominating intermediate S1 by experiment, it is reasonable to assume that S1 was preferentially converted to the corresponding oligo- or polyimines, as a significantly larger quantity of precipitate (54 mol %) was formed in the reaction of TFB and linker 2, which we elucidated further in the following section, with the help of computational study, employing density functional theory (DFT). Figure 5 | Cage-breaking experiment for probing the reaction species by adding excessive TFB at the equilibrium of the initial stoichiometric reaction (TFB/linker 2 = 2/3). (a) The schematic illustration of the intermediate S1 with highly disordered structure was formed as the dominating species by adding excessive TFB, concomitant with the suppression of MC-2 and CSC-2, and the formation of polyimine precipitate. (b) Evolution of MC-2, CSC-2, and TFB⊂MC-2 with the ratio of TFB/linker 1 changed from 2/3 to 7/3, determined by 1H NMR spectroscopy given in Supporting Information Figure S18. (c) MALDI-TOF MS spectra of the reaction mixtures with different ratios of TFB/linker 2. CSC-2, MC-2, and intermediate S1 with the corresponding m/z values are highlighted by the arrows with the corresponding color codes. Linker 1, 2-[3,5-di(3-aminophenyl)phenyl]-1,3-dithiane; linker 2, 3-[3-(3-aminophenyl)phenyl]aniline; 1H NMR, proton nuclear magnetic resonance; MALDI-TOF MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Download figure Download PowerPoint As mentioned above, compared with the reactions involving TFB, those using TFPB favored the formation of monomeric cages, and only a negligible amount of catenated species was formed with the latter. Similar cage-breaking experiments were conducted to probe the key reaction intermediates of TFPB/linker 1 (Figure 6a) and TFPB/linker 2 (see Supporting Information Figure S12), respectively. Figure 6 | Cage-breaking experiment for probing the reaction species by adding excessive TFPB at the equilibrium of the initial stoichiometric reaction (TFPB/linker 1 = 2/3). (a) The schematic illustrates MC-3, a trace amount of CSC-3, encapsulated intermediate TFB⊂MC-3, and the reaction intermediate S3, which was mainly converted to precipitate. A trace amount of S2 was detectable throughout the reaction. (b) An evolution of the only meaningful product MC-3 with the ratio of TFB/linker 1 changed from 2/3 to 7/3, determined by 1H NMR spectroscopy given in Supporting Information Figure S20. (c) MALDI-TOF MS spectra of the reaction mixtures with different ratios of TFPB/linker 1. Reaction species with the corresponding m/z values are highlighted by the arrows with the corresponding color codes. Linker 1, 2-[3,5-di(3-aminophenyl)phenyl]-1,3-dithiane; 1H NMR, proton nuclear magnetic resonance; MALDI-TOF MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Download figure Download PowerPoint First, we observed an apparent decrease in MC-3 (Figure 6b) and MC-4 (see Supporting Information Figure S21), which were converted mainly to the corresponding precipitates. Although all reaction species could not be identified by 1H NMR spectroscopy (see Supporting Information Figures S20 and S21), the evolution profiles of reaction intermediates were unambiguously monitored by MALDI-TOF MS. In the case of TFPB/linker 1, only a trace amount of encapsulated intermediate TFPB⊂ MC-3 was detected at the equilibrium of the stoichiometric reaction, which vanished quickly upon the addition of excessive TFPB (Figure 6c), and the of m/z corresponding to reaction intermediate S2 while that of became In contrast, the of TFPB⊂ MC-4 was in the case of TFPB/linker 2, but the of the intermediate increased and that of was present (see Supporting Information Figure These were in line with the negligible production of CSC-3 and CSC-4 when TFPB was Therefore, the experiments in Figure 6 and Supporting Information Figures confirmed that the flexible TFPB the encapsulated intermediates, but facilitated the formation of reaction intermediates with disordered We assumed that these intermediates were preferentially converted to oligo- or polyimine and suppression of encapsulated intermediate TFPB⊂ MC-3 or TFPB⊂ MC-4 was seemingly for the formation of only a amount of the catenated on the stability of encapsulated intermediates by calculation we all the encapsulated intermediates by with the kinetic study and cage-breaking these allowed us to clarify and the correlation between the key reaction intermediates and product of monomeric and catenated cages. For MC-1 and MC-2, the of TFB into their internal the host–guest TFB⊂ MC-1 and TFB⊂ MC-2 by an decrease of and respectively (Figures and structures showed that such enthalpic were achieved by a and stacking between the TFB and MC-1 and MC-2, respectively (see detailed in Supporting Information Figure Thus, it both encapsulated intermediates, TFB⊂ MC-1 and TFB⊂ MC-2, more thermodynamic stability than the monomeric cages, in facilitated their catenation to CSC-1 and CSC-2. TFB⊂ MC-1 was more stable than TFB⊂ MC-2 by (Figures and The formation of intermediates TFPB⊂ MC-3 (Figure and TFPB⊂ MC-4 (see Supporting Information Figure also showed an decrease of and respectively. However, these were only of the values for TFB⊂ MC-1 and TFB⊂ MC-2. This was due to the of flexible TFPB (see the in Supporting Information Figures and which in the of π–π stacking and of compared with the rigid panel TFB. These results seemingly the experimental observation of the negligible formation of the catenated cages CSC-3 and Notably, the influence of solvent molecules might more of a when the host–guest inclusion within a much larger cavity of MC-3 or Hence, we subsequently the impact of CHCl3 as the solvent molecules during the formation of each of the encapsulated intermediate. As for MC-1 with smaller cavity see detailed calculation in Supporting Information Figure we first tried to the of CHCl3 molecules, which could with the formation of TFB⊂ MC-1 (see Supporting Information for computational The calculation revealed that the of the inclusion was by only one CHCl3 within a cage (see Supporting Information Figure When both TFB and CHCl3 molecules were the of TFB to the MC-1 cavity that of