A Tautoleptic Approach to Chiral Hydrogen‐Bonded Supramolecular Tubular Polymers with Large Cavity

Abstract A new strategy towards tubular hydrogen‐bonded polymers based on the self‐assembly of isocytosine tautomers in orthogonal directions is proposed and experimentally verified, including by 1H fast magic‐angle spinning (MAS) solid‐state NMR. The molecular tubes obtained possess large internal diameter and tailor‐made outer functionalities rendering them potential candidates for a number of applications.


List of Figures
. S19

Materials and Methods
All chemicals were used as received from commercial suppliers. Compounds 1a [1] and 1b [2] were prepared according to reported procedures. All moisture sensitive reactions were carried out under an atmosphere of dry nitrogen or argon using oven-dried glassware. Anhydrous tetrahydrofuran was distilled from benzophenone-sodium, dichloromethane was distilled from calcium hydride and toluene was distilled from sodium. Yields refer to chromatographically and spectroscopically
Note: HMPA can be replaced with less toxic DMPU with c.a. 10% reduction in yield.

Synthesis of 14
Preparation of isocyanate 13. A mixture of acid X1 (315 mg, 0.57 mmol, 1.0 eq.), triethylamine (148 µl, 1.04 mmol, 2.0 eq.) and diphenylphosphoryl azide (133 µl, 0.61 mmol, 1.15 eq.) in dry toluene (7.0 ml) was heated at 40°C for 1.5 h. Then, the temperature was increased to 90°C and heating was continued for 4 h. After cooling to room temperature, the solvent was removed under reduced pressure and the residue was dried in high vacuum. The crude isocyanate 13 so-obtained was used in the next step without further purification.
Compound 12 (80 mg, 0.13 mmol, 1.0 eq.) was dissolved in dry DCM (5.0 ml) and TFA (0.5 ml) was added dropwise while cooling reaction mixture in an ice bath. After addition, the mixture was allowed to slowly reach the room temperature (without removing the ice bath). Then, the reaction mixture was stirred for an additional hour. The volatiles were removed under reduced pressure and the residue was dried under high vacuum. The crude salt obtained above was dissolved in THF (8.0 ml) followed by addition of triethylamine (0.2 ml). The solution obtained was added to isocyanate 13 (see the procedure above) and the mixture was heated at 50°C in a sealed vial overnight. The solvent was removed under reduced pressure and the residue was purified by flash chromatography (PE to DCM/MeOH 60 : 1) to afford 121 mg (62%) of 14 as a colourless oil.    In all experiments presented, 1 H and 13 C chemical shifts were referenced with respect to neat TMS using L-alanine as a secondary reference: 1.3 ppm for the CH 3 1 H resonance, corresponding to 1.85 ppm for adamantane, [3] , and 20.5 ppm for the CH 3 13 C resonance, corresponding to 38.5 ppm for adamantane [4] . 14 N chemical shifts were referenced relative to neat CH 3 NO 2 using the 14 N resonance of NH 4 Cl (powdered solid) at 341.2 ppm as an external reference [5] . To convert to the chemical shift scale frequently used in protein NMR, where the alternative IUPAC reference (see Appendix 1 of ref [6] ) is liquid ammonia at 50°C, it is necessary to add 379.5 ppm to the given values. [7] . 1 [8], [9] One rotor period of BABA recoupling [10], [11] was used for the excitation and reconversion of DQ coherence. For each of 256 t 1 FIDs, 16 transients were coadded with a recycle delay of 6 seconds. The F 1 = 2F 2 diagonal is indicated as a dashed black line.

S2. Solid State NMR
The base contour level is shown at 1% of the maximum peak intensity.
14 N -1 H HMQC experiments. [9], [12], [13] The spectrum was recorded using the R 3 recoupling scheme [14] for the recoupling of the 14 [16] with a pulse length of 6.0 µs ( 1 H - 13  spectrum was recorded with 11264 co-added transients and a recycle delay of 6 seconds. The 1 H -15 N spectra were recorded with 30720 (enantiopure) or 22528 (racemic) co-added transients and S15 a recycle delay of 3 seconds. S16  (Table S2) and better resolution of Ha and Hb protons of the isocytosine amino group. The presence of two tautomeric forms of isocytosine in racemic 1b is easily explained assuming an energetically favourable 3H-bonding interaction of isocytosine units between two enantiomers that enable the formation of 1D zig-zag polymeric structures. These polymers can further assemble in orthogonal directions to form insoluble corrugated sheets (Fig. S4).

S3. Dynamic Light Scattering
Dynamic light scattering (DLS) experiments were carried out on a Zetasizer Nano Z (Malvern) instrument at 298K. Samples were prepared by dissolving the corresponding amount of monomers 1b-3 in toluene, filtering through PTFE membrane filter (AcroDisc, 0.12 micron) and aging for 24 hours. Measurements were at least duplicated and the data with good quality correlograms were used. It should be noted that for compound 2, much smaller aggregates with R H < 10 nm were obtained in chloroform solution as opposed to large polymeric assemblies in toluene (Fig. S6). This observation is in line with significantly weaker H-bonds in chlorinated solvents as compared to aromatic ones. Smaller degree of polymerization in this solvent was also evident from more resolved 1 H NMR spectrum as opposed to very broad spectrum in toluene (see Fig. S47). Figure S6: Concentration dependence of hydrodynamic radii of (2 4 ) n aggregates in chloroform solution.
Concentrations and mean sizes of aggregates are indicated on the graph S21

S4. Viscosimetry
Viscosity was measured on AMVn Automated Micro Viscosimeter (Antor Paar) using 2.5 mm diameter gold coated ball at 298 K except for monomer 2, viscosity of which was probed at three different temperatures (298K, 338K and 258K). The corresponding solutions of monomers 2 and 3 were made in toluene, whereas chloroform was used for 1bdue to solubility reasons. Two distinct regimes can be identified in double logarithmic plot of 2 in toluene (Fig. S7). The low-slope regime can be attributed to a solution dominated by cyclic tetramers, whereas higherslope dependence is caused by supramolecular polymerization of cyclic tetramers. In case of monomer 1b in chloroform, single-slope regime is observed, starting at very low concentration, which indicates very efficient supramolecular polymerization (Fig. S8). Figure S8: log-log plot of specific viscosity vs concentration for 1b in chloroform (298K).

S22
High cooperativity of aggregation of monomer 1b in chloroform was demonstrated using viscosity measurements in the presence of simple isocytosine derivative ChS (Fig. S9), which might serve as a chain stopper for both tautomeric forms of the isocytosine. No change in viscosity was observed even in the presence of 10 mol % (solubility limit of ChS) of chain stopper.

S5. Preparation of Gels
Known amount of monomers 1b or 2 were added into 4.0 ml glass vial containing a stirring bar followed by solvent (methanol-free CHCl 3 or toluene). The vial was screw-capped with a plastic cap containing heat resistant PFET membrane. The vial was immersed into preheated oil bath (80°C for CHCl 3 and 100°C for toluene) and stirred until clear solution was obtained. The sol produced was cooled to ambient temperature and left undisturbed for 48 hr. The gelation concentration corresponds to the lowest concentration of monomer at which self-supporting gel is obtained.

S6. Atomic Force Microscopy
AFM imaging was carried out in ambient conditions using the JPK NanoWizard 3 AFM in intermmitent contact mode using the RTESPA (Bruker), Scout 350 (Nu Nano) and Tap300AI-G (Budget-Sensors) probes. The samples were prepared by drop casting the solution on the freshly cleaved muscovite mica and drying in nitrogen stream. The sample in manuscript Fig. 4a was dried 30s after drop casting. The sample in manuscript Fig. 4b was prepared by ultrasonicating the solution prior drop casting and drying in nitrogen stream after 10 s. Fibril structures were also observed in AFM images of (1b 4 ) n and (3 4 ) n from toluene solutions (Fig. S11). Figure S11: AFM images (toluene solution) of (1b 4 ) n (a-b), phase-contrast AFM image of (1b 4 ) n (c) and (3 4 ) n (d).
The AFM images of molecular tubes at lower concentrations suggest that the mica substrate has a templating effect on the fibrils. In case where individual fibrils are observed, the preferential alignment of fibrils along the mica crystallographic axes is seen. In case of entangled network of fibrils at higher concentration, the preferential angles seem to be determined by a combined action of surface templation and the fibril-fibril interaction, resulting in broader distribution of S25 preferred alignment angles. These preferences were revealed by applying histogram of oriented gradients [17] analysis (Fig. S12)

S7. Host-guest chemistry
The complexation of fullerenes by nanotubes (1b 4 ) n was assessed by using UV spectroscopy. The solid sample of the gels C 60 @(1b 4 ) n and C 70 @(1b 4 ) n , prepared at 1:4 molar ratio, were diluted in toluene and UV spectra were collected (Fig. S14). Although only marginal shift of absorbance maximum (1-2 nm) was observed, the clear change in relative intensity of the absorption bands and the appearance of isobestic points was noted using both, C 60 and C 70 . The very small change in absorbance maximum is in line with the results from our previous studies on fullerene complexation with structurally related tetrameric supramolecular host. [2] Figure S14: Normalized UV spectra of C 60 @(1b 4 ) n (a) and C 70 @(1b 4 ) n (b) in toluene.

S8. Molecular modelling
The geometry of tetramer 2 4 was first calculated at semi-empirical level of theory (PM3, as implemented in Spartan 10) (Fig. S15). [18] The optimized structure of tetramer 2 4 was then used to construct an octameric fragment of the nanotube (2 4 ) n . Due to the very large number of atoms, the resulting structure was optimized using molecular mechanics. The results of the molecular modelling show that despite the large size of solubilizing chains, a stable tubular polymer, potentially benefiting from favourable π-π interactions, can form (Fig. S16).