Pre-regulation of the planar chirality of pillar[5]arenes for preparing discrete chiral nanotubes

Regulating the chirality of macrocyclic host molecules and supramolecular assemblies is crucial because chirality often plays a role in governing the properties of these systems. Herein, we describe pillar[5]arene-based chiral nanotube formation via pre-regulation of the building blocks' chirality, which is different from frequently used post-regulation strategies. The planar chirality of rim-differentiated pillar[5]arenes is initially regulated by chiral awakening and further induction/inversion through stepwise achiral external stimuli. The pre-regulated chiral information is well stored in discrete nanotubes by interacting with a per-alkylamino-substituted pillar[5]arene. Such pre-regulation is more efficient than post-regulating the chirality of nanotubes.


General
All commercially available reagents and solvents were used as received. 1 H NMR and 13 C NMR spectra were recorded on JEOL JNM-ECS400, JNM-ECZ500R and JNM-ECA600P spectrometers at room temperature. Chemical shifts were reported in ppm versus tetramethylsilane. COSY and NOESY experiments were performed on JEOL JNM-ECA600P spectrometer at 30 ℃. UV-vis absorption spectra and circular dichroism (CD) spectra were recorded at 20 ℃ on JASCO V-750 spectrophotometer and JASCO J-1500 CD spectrometer, respectively. 2 mm quartz cuvets were used for both UV-vis and CD measurements. Highresolution ESI-MS was recorded on Thermo Fisher Scientific Exactive Plus mass spectrometer equipped with UltiMate 3000 HPLC. DFT calculations for structures of 1 with and without intramolecular hydrogen bond network (HBN) were carried out at ωB97XD/6-31G(d,p) level of theory using Gaussian 16 Rev C.01. 1 The initial geometry was generated according to the single crystal X-ray diffraction analysis of a pillar[5]arene having benzoic acid moieties and nbutyloxyl chains and the addition of water molecules. 2

Syntheses
Synthesis of peralkylamino-substituted pillar[5]arene 2 2 and compound 5 3 was followed reported procedure with optimization. added to quench the reaction. After concentrated under reduced pressure, the residue was chromatographed on a silica gel column using a mixture of n-hexane and acetone (10:1, v/v) as the mobile phase. The obtained white solid was further purified by thoroughly washing with a mixture of n-hexane and 2-propanol (99:1, v/v). Compound 4 was obtained as white solid (107 mg, 0.063 mmol, 7%). 1  To confirm that the chirality information of 1 observed on CD spectra was ascribed to planar chirality, monomeric compound 10 was synthesized by following the procedure shown in Scheme S2. Synthesis of the precursor 8 was followed literature. Compound 9. To a stirred solution of mono-substituted dihydroxybenzene 8 (102 mg, 0.57 mmol) in acetonitrile (10 mL) was added methyl 4-(bromomethyl) benzoate (156 mg, 0.68 mmol) and potassium carbonate (94 mg, 0.68 mmol). The reaction mixture was stirred under reflux for 2 h. After cooled down to room temperature, the solid was filtered off. The filtrate was concentrated under reduced pressure. The residue was chromatographed on a silica gel column using a mixture of n-hexane and acetone (from 10:1 to 5:1, v/v) as the mobile phase. Compound 9 was obtained as yellow oil (86.0 mg, 0.26 mmol, 48%). 1  To a stirred solution of di-substituted dihydroxybenzene 9 (80 mg, 0.244 mmol) in tetrahydrofuran (5 mL), sodium hydroxide (8 mg, 0.300 mmol) in water (5 mL) was added dropwise. The mixture was refluxed for 24 h. After the solvent was removed under reduced pressure, the solution was poured into aqueous HCl solution (1 M, 5 mL). The precipitate was collected by filtration, and washed with water (50 mL). After drying in vacuum, the product was obtained as white solid (67 mg, 0.212 mmol, 88%). 1        , only one set of peaks was observed, indicating that the free rotation of the units was fast on NMR time scale. The HBN was completely broken. In nonpolar solvent (i.e., dichloromethane-d 2 ), the 1 H NMR spectrum was similar to that in CDCl 3 , due to the slow unit rotation. The total volume of CDCl 3 was 500 μL in all cases. Upon addition of methanol-d 4 , the four sets of resonance coalesced gradually, because methanol is efficient to cut the HBN by forming hydrogen bonds with benzoic acid groups (i.e., solvation). 2 When ca. 2% of methanol-d 4 was added, the four sets of peaks coalesced to one set of broad peaks, indicating that the free rotation of the units in 1 was initiated.   (Fig. S12).

Discussion on complexation of 1 with DBB and DCE
We tried to cultivate single crystals of the complexes between 1 with DBB and DCE to understand the interaction details in these complexes. Unfortunately, we failed to obtain highquality single crystal available to analyze by X-ray diffraction. However, our recent work (Chem. Commun., 2020, 56, 8424−8427), 4 in which we have observed the achiral solventdependent CD-signal change of planar chiral pillar[5]arene 13 (Fig. S12), can help us to understand the mechanism of the planar chiral further induction and inversion of 1.
Pillar[5]arene 13 has the same chiral substituents as 1 on two rims, and showed positive and negative Cotton effects in DCE and DBB, respectively (Chem. Commun., 2020, 56, 8424−8427). 4 These observations are same as that in our current research. In our current research, addition of DCE in the CD-awakened 1 resulted in positive CD signal of 1, while addition of DBB in CD-awakened 1 caused further negative CD signal (Fig. 3 in the main text).
In Fig. S12, we showed the single crystalline structures of the 1:1 host-guest complex of 13 and DBB, and 13 and DCE.
In the complex of 13 and DBB, the DBB molecule is long enough to repel the third carbon atom (i.e., carbons γ and ε) on each substituent ( Figure S12c). Thus, the bulky ethyl branch (carbons γ and δ in cyan color) on aliphatic chains are located out of the cavity to minimize the steric hindrance between the bromine atoms of DBB and aliphatic chains, which resulted in a tendency of pS induction of 13. Thus, the complex of 13 and DBB in crystalline state only shows pS chirality. In other words, the (S,S,S,S,S,pS,pS,pS,pS,pS)-form of 13 are more stable than (S,S,S,S,S,pR,pR,pR,pR,pR)-form when complexing with DBB. Because of the same chiral substituents in 1 and 13, we believe that complexation of DBB and 1 also leads a pS-favoring rotation of 1.
In contrast, with short length, DCE is fully buried into the central part of the pillar[5]arene's cavity without interrupting the substituents on rims ( Figure S12d). As a consequence, enough space is made in the cavity for the ethyl branch (carbons γ and δ in cyan color) on aliphatic chains, leading a (S,S,S,S,S,pR,pR,pR,pR,pR)-form of 13. Similarly, we believe that complexation with DCE also leads a pR-favoring rotation of 1, which make the CD conversion of 1.

Discussion on chirality regulation of 1 by DBB and DCE in absence of MeOH.
Different from the cases of planar chirality awakening 1, the planar chiral induction effect was very weak when guest molecules were directly added to the solution of 1 in chloroform without awakening the chirality in advance (Fig. 3b). Addition of DBB induced ca. 30% of the CD intensity of chirality-silenced 1 compared with the chirality-awakened 1. DCE failed to induce the planar chirality of 1 in absence of MeOH.
The weak chiral induction of 1 on addition of DBB was caused by the dimerization of 1 as evidenced by the appearance of the dimer peaks on 1 H NMR spectra (Fig. S13-S14). To form dimers, the intramolecular HBNs would be broken and rearranged to intermolecular HBNs, which provided a chance for the unit rotation of 1, thus induced its planar chirality to some extent. In contrast, DCE cannot trigger the dimerization of 1 even in a much excess amount (Fig. S14). The different effect on dimerization of 1 on addition of DBB and DCE is also ascribed to the length-effect of the guest molecules (Fig. S12). DCE may be fully buried into the central part of the pillar[5]arene's cavity without interrupting the intramolecular HBN on rims, while DBB is long enough to influence the groups on rims. Nevertheless, awakening the planar chirality of 1 via MeOH in advance was important to launch an effective chiral regulation.

Fig. S13
Partial 1 H NMR spectra (600 MHz, CDCl 3 ) of 1 at various concentrations. At diluted concentrations (0.1-0.5 mM), no change was observed on 1 H NMR spectra. As the concentration increased to 1.5 mM and higher, a new set of peaks emerged (H a (dimer) and H b (dimer)), which implied that self-assemblies formed as concentration was increased. Similar behavior has been observed in our previous work, 2 where evidences demonstrated the concentration-dependent dimerization of compounds similar to 1 via intermolecular hydrogen bonds. Therefore, we assigned the emerging peaks as dimers of 1.  1 mM), all monomeric 1 was dimerized. By contrast, even 35 equiv. of DCE was not able to trigger the dimerization of 1. The different effect on dimerization of 1 on addition of DBB and DCE is ascribed to the length-effect of the guest molecules. As mentioned above (Fig. S12), the DBB molecule is long enough to repel the third carbon atom on each substituent. This will surely interrupt the intramolecular HBN on the other rim of 1. As a consequence, the stability of the intramolecular HBN decrease, which gives a chance for dimerization of 1 by forming intermolecular hydrogen bonds. In contrast, DCE is fully buried into the central part of the pillar[5]arene's cavity without interrupting the intramolecular HBN on rims.              Fig. 4 and S22). (b) CD spectra of pR-rich nanotubes constructed from 1 (0.1 mM), 2 (0.5 equiv.), MeOH (10% volume fraction of the solution), and DCE (35 equiv.) in different sequences (Fig. S23-26).

Chirality storage of the nanotubes after removing awakener and regulators
By removing the solvent and heating the pre-regulated nanotubes at 120 ℃ for 5 days in vacuum, we tried to remove all external factors (i.e., methanol, DBB and DCE) that was used to regulate the chirality of 1. By analyzing the 1 H NMR spectra before and after heating, we found that all the external factors (methanol and DBB) in the pS-rich nanotubes prepared by path A were removed (Fig. S28). The pR-rich nanotubes prepared by pre-regulation show weak CD signals. After removal of DCE and methanol, the CD changes were difficult to monitor. Therefore, we only checked the stability of the aforementioned pS-rich nanotubes after removing external factors in solution by CD measurement (Fig. S29).

Fig. S28
Partial 1 H NMR spectra of pS-rich nanotubes (0.05 mM, CDCl 3 ) prepared by path A (top) and after heating the sample in vacuum at 120 ℃ for 5 days (bottom). DBB was completely removed as evidenced by the disappearance of the DBB proton signals after heating.

Fig. S29
Time-dependent normalized CD spectra (chloroform, 20 ℃) of the pS-rich nanotubes prepared by path A after removing DBB and methanol. Fig. S30 Optimization of the charge amount of DBB. The CD spectra (CHCl 3 ) of 1 (0.1 mM) upon addition of DBB (0 -10 equiv.) were recorded at 20 ℃. The CD intensity slightly changed upon addition of DBB from 1 to 5 equiv.; on further addition of DBB, the CD changes were negligible. Thus, 5 equiv. of DBB was chosen for all other experiments in this research.