Synthesis and characterization of macrocycle octamethyl cyclo[4](1,3-(4,6)-dimethylbenzene)[4]((4,6-benzene)(1,3-dicarboxylate) (OC-4)The macrocycle of this study, octamethyl cyclo[4](1,3-(4,6)-dimethylbenzene)[4]((4,6-benzene)(1,3-dicarboxylate) (OC-4), was generated via a combined fragment coupling and Suzuki-Miyaura reaction sequence (Fig. 2a). Various reaction conditions were evaluated sequentially (Supplementary Table 1). Cyclization in toluene at 373 K (condition i) gave C4v-OC-4 in a yield of 31% with little if any C2v-OC-4 being observed in this case. In acetonitrile at 333 K (condition ii), the atropisomers with C2v and C4v symmetry (i.e., C2v- and C4v-OC-4) were isolated from the reaction in 7% or 8% yield, respectively. Theoretical analyzes using semiempirical methods (PM7) were carried out. The results revealed that C4v-OC-4 is more stable than C2v-OC-4 (Supplementary Fig. 23 and Table 2), thus providing a rational preference for the isomer with C4v symmetry observed under both experimental conditions. Meanwhile, consider the corresponding cyclization Pd-catalyst contained intermediate (i.e., C4v-OC-4-im and C2v-OC-4-im), formation heat of C4v-OC-4-im was calculated and larger than that of C2v-OC-4-im. The finding implied that C2v-OC-4 as a kinetic product predominantly present under low-temperature condition (e.g., reaction condition ii).Fig. 2: Synthesis and characteristics of macrocycle OC-4.a Synthesis of macrocycle OC-4. Reaction conditions: (I) 1 (1 equiv.), 3 (1 equiv.), CsF (25 equiv.) and Pd(dppf)2Cl2• CH2Cl2 (0.20 equiv.) in toluene, 373 K, 12 h, C2v-OC-4, trace, C4v-OC-4, 31%; (ii) 1 (1 equiv.), 3 (1 equiv.), CsF (25 equiv.) and Pd(dppf)2Cl2• CH2Cl2 (0.20 equiv.) in acetonitrile, 333 K, 48 h, C2v-OC-4, 7%, C4v-OC-4, 8%. b Structures and conformational studies of macrocycle OC-4. (b1) Heating induced quantitative conformational change from C2v- to C4v-OC-4 in solid state; (b2) and (b3), or (b4) and (b5), top and side views of C2v- or C4v-OC-4 with a stick form in the single crystal of [C2v-OC-4•5CYH] or [C4v-OC-4•3CYH], respectively. The insert is the thin layer chromatography (TLC) analysis on silica gel plates (eluent: petroleum ether: ethyl acetate = 3:1, v/v) of C2v-, C4v-OC-4, and the conversion at 573 K. c1–c2 1H NMR spectra of 5.00 mM C4v-OC-4 (c1) and C2v-OC-4 (c2) recorded in TCE-d2 at 298 K (500 MHz). d Expanded view of the temperature-dependent 1H NMR spectra of C2v-OC-4 (5.00 mM) in TCE-d2 (500 MHz).Macrocycles C2v- and C4v-OC-4 were fully characterized in TCE-d2 solution at 298 K via 1H and 13C NMR spectroscopy, correlation spectroscopy (COSY) and nuclear Overhauser effect spectroscopy (NOESY) (Fig. 2c and Supplementary Figs. 11-14 and Figs. 17-20). Both OC-4 isomers gave rise to peaks when subjecting to matrix assisted laser desorption ionization-time of flight high resolution mass spectrometric analysis that were the same within error (MALDI-TOF HRMS) (m/z 1207.4080 or 1207.4083, respectively; calcd m/z 1207.4092) (Supplementary Figs. 16, 22).Macrocycles C2v- and C4v-OC-4 were further confirmed by single crystal X-ray diffraction analysis. In both cases, diffraction-grade crystals were grown by subjecting a CH2Cl2 and cyclohexane (CYH) (1:1, v/v) solution of the respective pure isomer to slow evaporation. The resulting structures revealed macrocyclic frameworks with “C2v-” or “hourglass-like” C4v- symmetry, respectively (Fig. 2b). In [C2v-OC-4•5CYH], the meso-dimethylbenzene units on C2v-OC-4 exist as parallel sets, while the ester groups are found in the four corners and point to the outside of the cavity (diameter = 7.8 Å) (Supplementary Figs. 31–33). The torsion angles between neighboring benzene rings range from 57° to 63°. In [C4v-OC-4•3CYH], the aromatic rings on C4v-OC-4 adopt a 1,3-alternate-like conformation and the ester groups all point to the same side of the cavity. A near circular cavity with a diameter about 8.7 Å is seen (Supplementary Figs. 34–36), with inter-ring angles between neighboring aromatic rings ranging from 84° to 112° being observed.Thermal conversion of OC-4 with intermediate processThe fact that C2v-OC-4 was mainly obtained as a minor product under the lower temperature (333–353 K) reaction condition is consistent with it being a kinetic product and C4v-OC-4 being the thermodynamic product. Support for this supposition came from the theoretical analysis noted above, as well as the finding that C2v-OC-4 could be converted quantitatively to C4v-OC-4 in the solid state at high temperature (573 K) (Fig. 2b, Supplementary Fig. 39). Temperature-dependent 1H NMR spectroscopic studies carried out in TCE-d2 solution (5.00 × 10-3 M) from 233 K to 373 K revealed that both the C2v and C4v isomers of OC-4 were stable on the laboratory time scale at the temperatures lower than 343 K. However, evidence of conversion between C2v- and C4v-OC-4 was seen at temperatures ≥343 K (Fig. 2d, Supplementary Figs. 24, 25, 27, 28). Considering the close relationship between molecular conformations and temperatures/solvents, the stability of OC-4 isomers in toluene-d8 was investigated. Through temperature-dependent 1H NMR spectroscopic studies, it was observed that the transformation of C2v-OC-4 into C4v-OC-4 began at a relatively low temperature of 318 K (Supplementary Fig. 26), which is significantly lower than in TCE-d2 solution. Furthermore, at temperatures above 373 K, almost all of C2v-OC-4 converted into C4v-OC-4 and other forms of OC-4. These findings suggest that toluene-d8 promotes the conversion from C2v- to C4v-OC-4.Time-dependent 1H NMR spectroscopic analyzes of C2v-OC-4 (5.00 × 10−3 M) carried out at 393 K in TCE-d2 to study the isomerization process (Figs. 3a, b, Supplementary Fig. 40). After 2 h, the signals corresponding to C4v-OC-4 were integrated to ca. 71.4% of the total whereas those for C2v-OC-4 accounted for only 5.6% of the total (Fig. 3c)56. Moreover, signals ascribable to new species were observed. Their overall ratio increased to a maximum of 49.2% with time before decreasing to a final equilibrium value of ca. 23.0%. These results lead us to propose that one or more intermediates are involved in the conversion process. Further analysis of the NMR data provided support for the possible formation of three intermediates, namely isomers of OC-4 with C1-, Cs-, and C2-symmetry, respectively. Similarly, the transformation from C2v- to C4v-OC-4 was carried out at 373 K in toluene-d8 (note: the concentration was 1.00 × 10−3 M due to its low solubility in toluene-d8 (about 1.00 × 10−3 M)) (Supplementary Fig. 50). After 3 h, nearly all C2v-OC-4 are converted to C4v- (97.5%) and various forms of OC-4 (2.5%). The signals corresponding to three intermediates reached a maximum total proportion of 70.2% after 0.25 h, before dropping to a final equilibrium value of 2.5% (Supplementary Figs. 50, 51). For comparison purposes, the conversion process was also conducted in TCE-d2 using the same concentration (1.00 × 10−3 M) and temperature (373 K). In this case, the percent conversion of C2v-OC-4 was lower, reaching 86.9% after 3 h (Supplementary Figs. 53, 54). The decrease in conversion temperature to 373 K and the increase in percent conversion to more than 99% of C2v-OC-4 in toluene-d8 suggest that the transformation from C2v- to C4v-OC-4 may be affected by the choice of solvents.Fig. 3: Atropisomer conversion process of OC-4.a Conversion of C2v- to C4v-OC-4 at 393 K in TCE-d2. b Expanded view of time-dependent 1H NMR (700 MHz) spectra of C2v-OC-4 (5.00 × 10−3 M) in TCE-d2 after warming at 393 K. c Time-dependent speciation plots for C2v- (red dot and line), C1- (purple dot and line), Cs- (green dot and line), C2- (orange dot and line) and C4v-OC-4 (blue dot and line) after heating C2v-OC-4 (5.00 × 10−3 M) at 393 K. d Schematic representation of the method used to obtain single crystals of intermediate Cs-OC-4. e Top and side views of Cs-OC-4 of the single crystal structure of [Cs-OC-4•n-hexane] in stick form.Efforts were made to capture the presumed intermediate(s) generated during the C2v- to C4v-OC-4 isomerization process. Toward this end, C2v-OC-4 (10 mg) was dissolved in TCE-d2 (0.6 mL) and placed into a NMR tube (5 mm) (Fig. 3d). A relatively large integration for the presumed intermediates (a relative ratio of up to 49.2%) is seen after heating the solution at 393 K for 15 min. After allowing the mixture to cool to room temperature, the solution was transferred to a clean vial. A super silent adjustable air pump was then used to remove the TCE-d2 solvent and any other volatiles while keeping the temperature constant. The resulting solid residue was dissolved in a mixture of CH2Cl2 and CH3OH (1.5 mL; 1:1, v/v). After slow evaporation at 298 K for 1 day, diffraction-grade single crystals were obtained. An ensuing X-ray diffraction analysis revealed the presence of three sets of crystals corresponding to C2v-OC-4, C4v-OC-4, and a new OC-4 species with Cs symmetry (i.e., Cs-OC-4) (Fig. 3e and Supplementary Figs. 57–59). In Cs-OC-4, the benzene units adapt a “down-up-up-down-up” orientation. The torsion angles between adjacent benzene units range from 55° to 110° (Supplementary Fig. 58). To the best of our knowledge, Cs-OC-4 constitutes the limited structurally characterized intermediate to be captured during an atropisomerization process involving presumably cooperative rotations about single C-C σ bond.We further attempted to collect pure sample of Cs-OC-4 from the crystalline mixture of Cs-, C2v- and C4v-OC-4 for solution phase NMR spectral analysis. To ensure purity, the Cs-OC-4 crystals were separated mechanically and their integrity were checked by measuring the cell parameters crystal by crystal by X-ray diffraction analysis. A small amount (less than 0.1 mg) of Cs-OC-4 was collected in this way. The resulting sample was washed with petroleum ether and CYH three times, respectively, and dried at room temperature for 24 h to remove residue solvents (e.g., CH2Cl2, CYH and petroleum ether). The Cs-OC-4 samples were then dissolved in TCE-d2 and subjected to 1H, COSY and NOESY spectral analysis at 298 K (Supplementary Figs. 60–63). The main peak pattern proved consistent with the structure of Cs-OC-4 elucidated via the single crystal diffraction analysis. However, two small sets of signals were observed with the same integral ratios as 9.7% (Supplementary Fig. 61). On this basis we suggest that isomers with pseudo-C2 and C1 symmetry coexist with Cs-OC-4 in TCE-d2 solution. Support for this proposed coexistence came from theoretical calculation (cf. Fig. 4 and Supplementary Figs. 52, 55, 56).Fig. 4: Potential energy analysis of atropisomer conversion corresponding to OC-4.Potential energy diagram for the conversion of C2v-OC-4 to C4v-OC-4 from the time-dependent 1H NMR spectral studies of C2v-OC-4 (5.00 × 10−3 M) in TCE-d2 at 393 K. The transition states, as well as those for C1-OC-4 and C2-OC-4, are listed based on the theoretical calculations.When Cs-OC-4 was allowed to sit in TCE-d2 solution at room temperature for 12–40 h (Supplementary Fig. 64), relatively weak signals corresponding to both C2v- and C4v-OC-4 were observed in the 1H NMR spectrum. Time-dependent 1H NMR spectroscopic studies of Cs-OC-4 were then carried out at 393 K in TCE-d2 solution. As shown in Supplementary Fig. 66, the Cs-OC-4 ratio decreased to 15.8% at final equilibrium within 30 min. Meanwhile, the ratio of C2v-OC-4 increased to a maximum of 14.3% at 4 min, and then decreased to 4.1% at equilibrium. Over the course of these studies, the proportion of C1-OC-4 and C2-OC-4 stayed roughly identical with 1.9% at final equilibrium, while the population of C4v-OC-4 increased to 76.4% by the time equilibrium was reached. In a separate experiment, C4v-OC-4 in TCE-d2 solution was held at 393 K for 12 h (Supplementary Fig. 67). Based on 1H NMR spectral integrations, the initially pure sample was converted to a mixture containing the C2v-OC-4 (5.1%), Cs-OC-4 (17.5%), C1-OC-4 (2.9%) and C2-OC-4 (2.9%) forms, in addition to the thermodynamically favored C4v-OC-4 isomer (Supplementary Fig. 68). A similar final ratio of each OC-4 form was seen when C2v-OC-4 or Cs-OC-4 were warmed under identical conditions. However, there is no if little change observed when C4v-OC-4 is heated at 373 K for 12 h in a toluene-d8 solution (Supplementary Fig. 69). This suggests that the conversion from C2v- to C4v-OC-4 in the toluene-d8 is irreversible.With time-dependent 1H NMR spectroscopic detailed analyzes of C2v-OC-4 (5.00 × 10−3 M) carried out in TCE-d2 at 393 K (Fig. 3b), the relationship between the concentration ratio of each OC-4 form and time (t) could be expressed as below (for details see the Supplementary equations 1–25 and Figs. 41–48):$$[{{{{\boldsymbol{C}}}}}_{2{{{\boldsymbol{v}}}}}-{{{\bf{O}}}}{{{\bf{C}}}}-{{{\mathbf{\ 4}}}}]\%=5.6205+94.3795\times \exp \left(\frac{-0.142{{{\rm{t}}}}}{94.3795}\right)$$
(1)
$$[{{{{\boldsymbol{C}}}}}_{{{{\boldsymbol{s}}}}}-{{{\bf{O}}}}{{{\bf{C}}}}-{{{\mathbf{\ 4}}}}]\%=0.736\times \frac{94.37\times (1-{e}^{{{{\rm{\gamma }}}}})\times ({{{\rm{\alpha }}}}+{{{\rm{\beta }}}}+0.31754)}{{{{\rm{\alpha }}}}+{{{\rm{\beta }}}}+1.31754}$$
(2)
$$[{{{{\boldsymbol{C}}}}}_{{{{\mathbf{\ 1}}}}}-{{{\bf{O}}}}{{{\bf{C}}}}-{{{\mathbf{\ 4}}}}]\%=[{{{{\boldsymbol{C}}}}}_{{{\mathbf{\ 2}}}}-{{{\bf{O}}}}{{{\bf{C}}}}-{{{\mathbf{\ 4}}}}]\%=0.132\times \frac{94.37\times (1\,-\,{{{{\rm{e}}}}}^{{{{\rm{\gamma }}}}})\times ({{{\rm{\alpha }}}}+{{{\rm{\beta }}}}+0.31754)}{{{{\rm{\alpha }}}}+\,{{{\rm{\beta }}}}+1.31754}$$
(3)
$$[{{{\boldsymbol{C}}}}{{{\mathbf{\ 4}}}}_{{{\boldsymbol{v}}}}-{{{\boldsymbol{OC}}}}-{{{\mathbf{\ 4}}}}]\%=94.37\times (1-{{{{\rm{e}}}}}^{{{{\rm{\gamma }}}}})-[{{{{\boldsymbol{C}}}}}_{{{{\boldsymbol{s}}}}}-{{{\mathbf{\ OC}}}}-{{{\mathbf{\ 4}}}}+{{{{\boldsymbol{C}}}}}_{{{{\mathbf{\ 1}}}}}-{{{\mathbf{\ OC}}}}-{{{\mathbf{\ 4}}}}+{{{{\boldsymbol{C}}}}}_{{{{\mathbf{\ 2}}}}}-{{{\mathbf{\ OC}}}}-{{{\mathbf{\ 4}}}}]$$
(4)
Here, α, β and γ are functions of heating time (t) as below:$$\alpha=1.6161\times \exp \left(\frac{-{{{\rm{t}}}}}{1102.30398}\right)$$
(5)
$${{{\rm{\beta }}}}=19.55909\times {{{\rm{exp}}}}\left(\frac{-{{{\rm{t}}}}}{326.3319}\right)$$
(6)
$${{{\rm{\gamma }}}}=\frac{-0{{{\rm{.0014t}}}}}{0.9437}$$
(7)
Thus, the thermodynamic and kinetic parameters of the conversion could be calculated in accord with the equilibria shown below (equation 8). The equilibrium constants (K), rate constants (k), Gibbs free energies (∆Gθ) and Gibbs activation energy (-∆G≠) between each other are given in Table 1. Setting the formation energy of C2v-OC-4 at 0.0 kcal, an energy diagram for the interconversion between the C2v- and C4v-OC-4 forms in TCE-d2 at 393 K could be constructed. It is shown in Fig. 4.Table 1 Reaction equilibrium constants (K)a calculated from the experimental data, forward rate constants (k), Gibbs free energies (∆Gθ)a, Gibbs activation free energies (-∆G≠)a; and theoretical computationalb,c formation energy values (E and ΔE) for the proposed equilibria involved in the conversion of C2v- to C4v-OC-4 at 393 KFurther support for the observed conformational conversions came from theoretical studies involving PM7 analyzes. These analyzes revealed that 1) Cs-OC-4 and C4v-OC-4 are energetically more stable than C2v-OC-4 and 2) a small energy barrier (≤10 kcal mol−1) persists between Cs-OC-4 and C2v-OC-4 or C4v-OC-4 (Supplementary Fig. 56).The calculations also confirmed that C1- and C2-OC-4, whose existence as minor intermediates was inferred from the spectral studies, are less stable than C2v-OC-4. Furthermore, the conformation optimization provided the possible structures for C1-OC-4, C2-OC-4 and other OC-4 transient state forms. These simulated structures and the results calculated from the experimental data shown above allowed the profiles corresponding to the proposed conversion of C2v-OC-4 to C4v-OC-4 to be obtained (Fig. 4).Conversion of C
4v
– to C
2v
-OC-4 or C
s
-OC-4 promoted by chemical reactionsThe reversible interconversion between OC-4 forms led us test whether C2v-OC-4 could be produced from C4v-OC-4 more efficiently than by simple heating in TCE-d2 solution (Fig. 5). With this goal in mind, pure C4v-OC-4 was subjected to hydrolysis to give the corresponding acid, cyclo[4](1,3-(4,6)-dimethylbenzene)[4]((4,6-benzene)(1,3-dicarboxylic acid) (CA-4). It was found that the acid product produced in this way included several forms (i.e., C4v-, C2v-CA-4, etc.) (Fig. 5a and Supplementary Figs. 70–72). Herein, different MOH (M = Li, Na, K, Rb or Cs) salts and tetra-n-butylammonium hydroxide (TBAH) were tested in the initial hydrolysis step prior to reesterification. Significant countercation effects were observed (Supplementary Fig. 70). The generated CA-4 mixture without further purification was then treated with tetra-n-butylammonium fluoride (TBAF) and iodomethane at room temperature for 48 h57. Specifically, the final C2v-, Cs- and C4v-OC-4 ratio proved highly dependent on the countercation species employed (Fig. 5b). The overall conversion ratio of C4v-OC-4 was relatively high in the case of NaOH (70.5%), KOH (63.1%) or RbOH (65.1%) (Fig. 5d). The tested organic base TBAH used in hydrolysis promoted largest conversion ratio of C4v-OC-4 as 90.0% (including C2v- and Cs-OC-4, 33.6% and 50.3% yield, respectively). The total conversion rates of C4v-OC-4 are much higher when compared to direct heating the TCE-d2 solution of C4v-OC-4 at 393 K (conversion rate as 28.4%). This suggests that the chemical reactions may facilitate the reversible conversion of C4v-OC-4 into different forms.Fig. 5: Chemical reaction promoted reversible atropisomer transformation of OC-4.a Schematic representation of molecular structure conversion from C4v- to C2v-OC-4 via subjecting to hydrolysis and esterification with different alkali hydroxide salts; b Expanded view of 1H NMR spectra of OC-4 recorded in TCE-d2 after esterification following treatment of CA-4 with different alkali hydroxide salts (500 MHz); c Structure of [C2v-CA-4-3H+]3-•Na+•K+•3Cl-•DMF and C4v-CA-4 shown in stick form obtained from a single crystal structural analysis of [[C2v-CA-4-3H+]3-•Na+•K+•3Cl-•DMF]•4H3O+•9.5H2O•5DMF and [C4v-CA-4•2DMF•H2O], respectively; d Isomer fraction of various OC-4 forms (red column: C2v-OC-4; purple column: C1-OC-4; green column: Cs-OC-4; orange column: C2-OC-4; blue column: C4v-OC-4) after reesterification following C4v-OC-4 hydrolysis with different alkali hydroxide salts, error bars correspond to S.D.Deeper insight into the countercation effects on the above conversion came from isolation of a putative intermediate. Here, C4v-OC-4 was subjected to hydrolysis with NaOH, followed by protonation with HCl (10 equiv.), and then redissolved in KOH solution (DMF/H2O (1: 1, v/v)) or dissolved directly in DMF/H2O (1: 1, v/v). The resulting products were allowed to undergo slow evaporation at room temperature. After two weeks, diffraction-grade single crystals were obtained. One structure is [C4v-OC-4•2DMF•H2O]. The X-ray diffraction data of another structure revealed a ratio of 1:1:1:3 between C2v-CA-4, Na+, K+, and Cl-. When considering the charge balance and the formation of carboxylate anions in metal complexation, it is recommended that four protons should distribute among the solvent water molecules. This is because water is a stronger Brønsted-Lowry base compared to Cl-. However, it should be noted that the protons on hydronium ions cannot be precisely located due to their disorder and technical limitations. Finally, this single crystal structure was suggested as a charge-neutral ionic form with [[C2v-CA-4-3H+]3-•Na+•K+•3Cl-•DMF]•4H3O+•9.5H2O•5DMF, rather than the conventional form as C2v-CA-4•NaCl•KCl•HCl•13.5H2O•6DMF. This is because the former representation provides more detailed information about the structure and coordination (Fig. 5c). In [[C2v-CA-4-3H+]3-•Na+•K+•3Cl-•DMF]•4H3O+•9.5H2O•5DMF, the resulting structure showed that K+ could effectively bridge the O atoms of neighboring carboxyl groups on C2v-CA-4. The associated stabilization was by theoretical calculations (for details see the Supplementary Tables 9–16). The putative stabilization of other alkali metal cations was studied by theory and, as expected, revealed differing degrees of thermodynamic stability for each cation-CA-4 isomer pair that as a general rule were consistent with the relative ratios seen by experiment. This finding could be advantageous in the development of molecular machines, devices, as well as stimuli-responsive smart materials.Host–guest properties of OC-4The different shapes seen for C4v-OC-4 and C2v-OC-4 led us to explore whether differences in their molecular recognition features might be observed. Initial studies involved probing the interactions between these two OC-4 forms and the linear guests 1,8-dibromooctane (G1), octane-1,8-di-thiol (G2) and 1,9-decadiyne (G3). For this study, mixtures containing C2v-OC-4 (20 mM) and 1 molar equal of the guest in question, either G1, G2 or G3, were studied via 1H NMR spectroscopy in CDCl3/CD3OD (1:2, v/v) (note: In all of the tested cases, the discussed solution system showed minimal solvent effects, meaning that the host–guest response was at its highest level.) (Supplementary Figs. 80–82). Little or no changes in the proton signals, either of the guests or the macrocycle, were seen in the case of C2v-OC-4. In contrast, notable shifts in key signals were seen in the case of C4v-OC-4 (Supplementary Figs. 83, 86, 89). 1H NMR spectroscopic Job-plot analyzes provided support for a 1:1 ([H]/[G]) binding stoichiometry for all three substrates (Supplementary Figs. 85, 88, 91). Further evidence for the formation of 1:1 complexes, C4v-OC-4⊃guest (where guest = G1, G2 or G3), came from electrospray ionization high-resolution mass spectrometric (ESI-HRMS) analyzes, the expected peaks were observed (Supplementary Figs. 92–94). Association constants (Ka) of (4.7 ± 0.5) M−1, (6.2 ± 0.6) × 10 M−1 or (6.0 ± 0.6) × 10 M−1 were seen for C4v-OC-4 and guests G1, G2 or G3, respectively.The above pseudo-rotaxane complexes were further characterized via single crystal X-ray diffraction analyzes (Fig. 6a). Diffraction-grade single crystals were obtained by slow evaporation of mixtures containing C4v-OC-4 and guests G1, G2 or G3 in CH2Cl2/CH3OH (1:2, v/v), respectively. In the solid state, these linear guest thread through the cavity of C4v-OC-4 (Supplementary Figs. 95–97). A particularly notable [3]pseudo-rotaxane complex, 2C4v-OC-4⊃n-eicosane, was obtained using n-eicosane (G4) as guest. Short separations (around 3.8 Å) between the carbon atoms of the linear guests and the carbon atoms located on the benzene units of C4v-OC-4 were observed. These findings are consistent with the pseudo-rotaxane structures being stabilized via C–H···π interactions.Fig. 6: Single crystal X-ray structures of the complexes involving C4v-OC-4 and linear guests or fullerenes.a Single crystal X-ray molecular structure of the complexes formed from C4v-OC-4 and 1,8-dibromooctane (G1), octane-1,8-dithiol (G2), 1,9-decadiyne (G3), and n-eicosane (G4). These species crystallize as [C4v-OC-4•1,8-dibromooctane], [C4v-OC-4•octane-1,8-dithiol], [C4v-OC-4•1,9-decadiyne], and [2C4v-OC-4⊃n-eicosane], respectively. b Single crystal X-ray diffraction structure of C4v-OC-4כC60 and C4v-OC-4כC70, species that crystalize as [C4v-OC-4•C60•3toluene•2THF] and [C4v-OC-4•C70•2toluene], respectively.In a further study, explored whether C2v-OC-4 or C4v-OC-4 would act as a fullerene receptor. In fact, 1H NMR spectral responses were observed after adding 1 molar equiv. of C60 or C70 to TCE-d2 solutions of C4v-OC-4 (note: the TCE-d2 was used due to its preferable solubility for OC-4 and fullerenes) (Supplementary Fig. 100). In contrast, little or no response was seen in the case of C2v-OC-4 (Supplementary Fig. 99). We rationalize this difference in binding propensity to the fact that C4v-OC-4 has a more open cavity and is thus better able to act as an endoreceptor for the fullerene guest. In TCE-d2, 1H NMR spectral titrations provided support for a 1:1 binding stochiometry and Ka values of (5.9 ± 0.6) × 103 M−1 and (5.2 ± 0.5) × 103 M−1 for the interaction between C4v-OC-4 and C60 and C70 (Supplementary Figs. 102, 103), respectively. These results basically coincided with the calculated (Ka) values (2.4 ± 0.2) × 103 M−1 and (2.2 ± 0.2) × 103 M−1 between C4v-OC-4 and C60 or C70 via UV-vis titrations (Supplementary Figs. 104, 105). Further evidences for the formation of 1:1 complex between C4v-OC-4 and C60 or C70 came from MALDI-TOF HRMS analysis (Supplementary Figs. 106, 107). Single crystal structures of C4v-OC-4כC60 and C4v-OC-4כC70 were determined via single crystal X-ray diffraction analysis (Fig. 6b and Supplementary Figs. 108–111). Both structures revealed 1:1 composition similar to what was inferred from the solution phase studies. Based on the metric parameters (i.e., the ≤3.8 Å separation between the carbon atoms of fullerenes C60 or C70 and those of C4v-OC-4), we suggest that complex formation is driven in part by stabilizing π···π donor-receptor interactions.Theoretical calculations were further performed to obtain additional insight into the interactions between C4v-OC-4 and representative linear guests, C60, and C70 (Supplementary Tables. 19-22). The lowest energies of the limiting outside and interpenetrated binding modes involving C4v-OC-4 and each guest were considered in vacuum using the MM+ methods included in the HyperChem 8.0 program or the PM7 method available in the MOPAC program. The ΔEin−out values provide support for the notion that the insert mode is more stable than various hypothetical outside binding modes.
Controlled interconversion of macrocyclic atropisomers via defined intermediates
