π-Stacking Stopper-Macrocycle Stabilized Dynamically Interlocked [2]Rotaxanes

The synthesis of mechanically interlocked molecules is valuable due to their unique topologies. With π-stacking intercomponent interaction, e.g., phenanthroline and anthracene, novel [2]rotaxanes have been synthesized by dynamic imine clipping reaction. Their X-ray crystal structures indicate the π-stackings between the anthracene moiety (stopper) on the thread and the (hetero)aromatic rings at the macrocycle of the rotaxanes. Moreover, the length of glycol chains affects the extra π-stacking intercomponent interactions between the phenyl groups and the dimethoxy phenyl groups on the thread. Dynamic combinatorial library has shown at best 84% distribution of anthracene-threaded phenanthroline-based rotaxane, coinciding with the crystallography in that the additional π-stacking intercomponent interactions could increase the thermodynamic stability and selectivity of the rotaxanes.

Meanwhile, the synthetic complexity could raise if the number of recognition sites increases or the thread became unsymmetrical. Indeed, the co-conformational [2]rotaxane isomers should have good stability to avoid isomerization by molecular shuttling. Therefore, developing a selective synthesis of dynamic [2]rotaxane with only one of R 2 NH 2 + ions recognition site occupied by macrocycle is highly critical.
Herein, we report new phenanthroline-based rotaxanes by employing 1,10-phenanthroline-2,9-dicarbaldehyde (dialdehyde 1) to form dynamic imine macrocycle. In the presence of the anthracene moiety, additional π-stacking between the phenanthroline unit and anthracene unit is observed in the solid-state structures of [2]rotaxanes, indicating further stabilization of the interlocked structure. The π-stacking at the stopper offers a higher specificity on rotaxane synthesis, making anthracene threaded phenanthroline-based rotaxanes as the major products in dynamic combinatorial library competing with other threads.
To verify the intercomponent stabilization [52][53][54][55][56][57], rotaxanes with stronger π-stacking interaction were synthesized. Dialdehyde 1 was selected due to its conjugated aromatic structure. By mixing equimolar amounts of thread 5-H·PF 6 , dialdehyde 1 and diamine 3 in MeCN at room temperature (Scheme 1), the crude 1 H NMR spectrum (Figure 1b) of the reaction mixtures reveals total consumption of the starting materials to the respective [2]rotaxane, 8-H·PF 6 . The disappearance of the aldehyde proton -CHO resonance at ca. 10.2 ppm supports the complete consumption of the dialdehyde compound in the reaction. Meanwhile, the characteristic signal for the benzylic methylene group -CH 2 protons adjacent to the R 2 NH 2 + center was shifted and split from δ = 5.2 and 4.4 ppm in thread 5-H·PF 6 to δ = 5.75-5.64 and 4.72-4.60 ppm in rotaxane 8-H·PF 6 , indicating its encirclement by the crown ether rings (Table S1). Hence, the macrocycle formed in the reaction contains two more carbon atoms and one more nitrogen atom compared to that of dialdehyde 2, making 8-H·PF 6 a [27]crown-9 rotaxane. Moreover, by changing the diamine from 3 to 4, [24]crown-8 rotaxane 9-H·PF 6 was obtained in high yield. For dialdehyde 2, only [24]crown-8 rotaxane 10-H·PF 6 and 14-H·PF 6 were obtained possibly due to the weaker π-π stacking stabilization. To study the π-stacking intercomponent interaction in the phenanthroline-based rotaxanes, the rotaxanes were crystallized for X-ray crystallography analysis. Single crystals of 8-H·PF 6 and 9-H·PF 6 were obtained by slow evaporation from CH 2 Cl 2 -hexane solution and MeCN-i Pr 2 O solution, respectively. From the solid-state structures (Figure 2A-F, Table S2), [N + -H···N] interaction was observed between the dialkylammonium ion, diimine and the phenanthroline moiety in both rotaxanes. Interestingly, the ethylene glycol unit (−CH 2 CH 2 O−) did not interact with the dialkylammonium ion as [N + -H···O] in 8-H·PF 6 , 9-H·PF 6 and 10-H·PF 6 , which is an essential interaction in DB24C8-and B21C7based rotaxane was not observed. Moreover, π-stacking intercomponent interactions were observed between the anthracene and phenanthroline unit with center-to-center distance ranges from 3.437 to 3.652 Å in 8-H·PF 6 and from 3.500 to 3.816 Å in 9-H·PF 6 , respectively. The two phenyl moieties participate in the π-stacking intercomponent interaction with another stopper, the 3,5-dimethoxylphenyl unit. The center-to-center distance between the three phenyl rings for 8-H·PF 6 is 3.922 Å and 4.291 Å while the center-to-center distance between the two phenyl rings for 9-H·PF 6 is 3.882 Å. The solid-state structures suggest that the π-stacking between the anthracene and the phenanthroline possesses a stronger interaction which can further stabilize the [2]rotaxanes. Moreover, the phenanthroline unit is taking a more important role than the ethylene glycol units on formation of the [2]rotaxanes. Perspective view of (A) solid-state structures of dynamic rotaxane 8-H·PF 6 (CCDC code: 1993136); (B) the offset face-to-face π-π stacking intercomponent interaction between the anthracene and phenanthroline in 8-H·PF 6 ; (C) the offset face-to-face π-π stacking intercomponent interaction between the phenyl moieties in 8-H·PF 6 ; (D) solid-state structures of dynamic rotaxane 9-H·PF 6 (CCDC code: 1993139); (E) the offset face-to-face π-π stacking intercomponent interaction between the anthracene and phenanthroline in 9-H·PF 6 ; (F) the offset face-to-face π-π stacking intercomponent interaction between the phenyl moieties in 9-H·PF 6 ; (G) solid-state structures of dynamic rotaxane 10-H·PF 6 (CCDC code: 1993140); (H) the offset face-to-face π-π stacking intercomponent interaction between the anthracene and pyridine in 10-H·PF 6 . All hydrogen atoms and PF 6 anions are omitted for clarity.
By slow evaporation of the MeCN solution of 10-H·PF 6 , single crystal was obtained and analyzed by X-ray crystallography. In addition to the [N + -H···N] interaction, the solid-state structure of 10-H·PF 6 ( Figure 2G,H) shows aromatic π-π stacking interactions between the anthracene and phenyl unit with center to center π-π stacking interaction distance: 3.636 Å. With the π-stacking intercomponent interactions between the stopper and the macrocycle, the anthracene thread can be stabilized the rotaxane formation and allows the isolation of rotaxanes.
In order to test the compatibility of the formation of dynamic rotaxanes, other stoppers for clipping were used. In particular, threads 6-H·PF 6 and 7-H·PF 6 were used to investigate the effect of the π-stacking interactions, while clipping with thread 7-H·PF 6 gave symmetrical rotaxanes 12-H·PF 6 and 13-H·PF 6 in high yields. 1 H NMR signals of the phenanthroline unit in 12-H·PF 6 and 13-H·PF 6 have a significant difference in comparison to the signals in 8-H·PF 6 and 9-H·PF 6 ( Figure 1 and Figure S17, Table S1). For 8-H·PF 6 , H a , H b , H c and H d were upfield-shifted from δ = 8.01 to 7.73 ppm, δ = 8.10 to 7.44 ppm, δ = 8.58 to 8.08 ppm and δ = 8.68 to 8.40 ppm, respectively, compared to 12-H·PF 6 . The upfield movement of phenanthroline aromatic proton signals of 8-H·PF 6 and 9-H·PF 6 can be due to the electron sharing from the electron-rich anthracene moiety. Moreover, phenanthroline aromatic proton was located in anisotropic-induced magnetic field of the anthracene moiety, so it is upfield shifted because of shielding effect. Both support the existence of the π-stacking interactions. For thread 6-H·PF 6 , only rotaxane 11-H·PF 6 was obtained in low yield; changing the components to dialdehyde 2 or diamine 4 did not show a reasonable structure of rotaxane. This might be due to the strong intramolecular π-stacking between the two anthracene stoppers, so thread 6-H·PF 6 was folded and not reactive to form dynamic rotaxane. As a result, dialdehyde 1 allows more comprehensive combination to form dynamic rotaxane compared to dialdehyde 2, showing the potential of using phenanthroline for dynamic rotaxane synthesis.
After the successful synthesis of five new phenanthroline-based rotaxanes, dynamic combinatorial library (DCL) experiments were conducted to study the effect of π-stacking stabilization (Scheme 2). Threads 5-H·PF 6 and 7-H·PF 6 in 1:1 ratio were mixed with different dialdehydes and diamines in the reaction. As expected, there should be a competition on forming either anthracene-based or symmetrical rotaxane. The dominant product in each experimental entry should be more thermodynamically stable. Moreover, the temperature was set to 50 • C to maximize the thermodynamic selectivity of rotaxanes. After reacting for 18 h, ratios of rotaxanes were determined by integrating the proton signals in the 1 H NMR spectra (Figure 1, Figures S17 and S18). As a result, 1 H NMR spectrum of dynamic combinatorial library experiment indicates that the thread 5-H·PF 6 was more favorable than 7-H·PF 6 to undergo rotaxane clipping with dialdehyde 1. 8-H·PF 6 and 9-H·PF 6 were preferentially formed as the dominant products, which have an 84:16 and 73:27 NMR yield ratios compared to the 12-H·PF 6 and 13-H·PF 6 , respectively (Table 1). Moreover, the better selectivity in Entry A compared to Entry B may be due to the enhanced effect of the π-stacking interaction between the dimethyoxyphenyl moiety on stopper and the phenyl moiety on the macrocycle. While there was no preference for the synthesis of 10-H·PF 6 and 14-H·PF 6 , 50:50 NMR yield ratio was obtained in entry C using the pyridine dialdehyde 2. The reason is that the weak π-stacking interaction did not affect the stability of rotaxanes significantly, which further gives evidence of the phenanthroline unit in providing additional π-stacking interaction between the thread's stopper and the macrocycle.

Materials and Methods
Instrumentation and General experimental. Unless otherwise stated, all reagents and anhydrous solvents were purchased from commercial sources and used without further purification. Acetonitrile was distilled over calcium hydride. Dried acetonitrile was stored over 3 Å molecular sieves. Flash column chromatography was performed using silica gel or aluminum oxide. Analytical TLC was performed on pre-coated silica gel plates (0.25 mm thick, 60F254, Merck, Darmstadt, Germany) and pre-coated ALUGRAM ® (0.20 mm thick, F254, MN, Dueren, Germany). Spots were observed under UV light. 1 H and 13 C NMR spectra were recorded on a Bruker Advance-III spectrometer (at 400 and 101 MHz, respectively). Chemical shifts are reported in parts per million from low to high field and referenced to residual solvent (CDCl 3 : 1 H, 7.26 ppm; 13 C, 77.16 ppm. CD 3 CN: 1 H, 1.94 ppm; 13 C, 1.32 ppm). NMR data were processed using MestReNova software (Mestrelab). Coupling constants are reported in Hertz. Standard abbreviations indicating multiplicity were used as follows: m = multiplet, quint = quintet, q = quartet, t = triplet, d = doublet, s = singlet, br = broad. High-resolution mass spectra were recorded on a Bruker Autoflex mass spectrometer (MALDI-TOF) and a Thermo Fisher Scientific UPLC-Q exactive focus hybrid quadrupole-orbitrap mass spectrometer in positive ion mode (ESI-MS). Selected crystals were used for intensity data collection on a Bruker AXS Kappa Apex II Duo diffractometer at 173K using frames of oscillation range 0.3 • , with 2 • < θ < 28 • .
General procedure for synthesis of [2]rotaxanes. A solution of thread (0.1 mmol) in MeCN (15 mL) was added diamine (0.1 mmol) and dialdehyde (0.1 mmol). The yellow mixture was stirred for overnight at room temperature. The solvents were evaporated under reduced pressure. Product is isolated by recrystallization (MeCN/Et 2 O) or column chromatography (neutral alumina, CH 2 Cl 2 /n-hexane).
General procedure for dynamic combinatorial library entries A−C. A solution of thread 5-H·PF 6 (0.1 mmol) and thread 7-H·PF 6 (0.1 mmol) in MeCN (15 mL) was added a diamine (0.1 mmol), dialdehyde (0.1 mmol) and 0.1 g of 3 Å molecular sieve. The yellow mixture was stirred for 18 h at 50 • C. The solvents were evaporated under reduced pressure. The ratio of rotaxanes in mixture was measured by 1 H NMR spectroscopy.

Conclusions
In conclusion, we had successfully synthesized new [2]rotaxanes by imine clipping reaction to form the macrocycle. π-Stacking intercomponent interactions between the macrocycle and the stopper of [2]rotaxanes 8-H·PF 6 , 9-H·PF 6 and 10-H·PF 6 were confirmed from their X-ray structural analysis. Intercomponent interactions were deduced and also corroborated in solution, besides solid state. Dynamic combinatorial library experiments of rotaxane formation resulted in phenanthroline-based rotaxanes with anthracene unit in threads as the major products, demonstrating the importance of π-stacking intercomponent interaction between the stopper and the macrocycle on the selective synthesis of the dynamic rotaxanes or related imine-based organic-inorganic hybrid materials [58].

Data Availability Statement:
The data presented in this study are available on request from the corresponding authors.