Enantiodiscrimination of Inherently Chiral Thiacalixarenes by Residual Dipolar Couplings

Inherently chiral compounds, such as calixarenes, are chiral due to a nonplanar three-dimensional (3D) structure. Determining their conformation is essential to understand their properties, with nuclear magnetic resonance (NMR) spectroscopy being one applicable method. Using alignment media to measure residual dipolar couplings (RDCs) to obtain structural information is advantageous when classical NMR parameters like the nuclear Overhauser effect (NOE) or J-couplings fail. Besides providing more accurate structural information, the alignment media can induce different orientations of enantiomers. In this study, we examined the ability of polyglutamates with different side-chain moieties—poly-γ-benzyl-l-glutamate (PBLG) and poly-γ-p-biphenylmethyl-l-glutamate (PBPMLG) —to enantiodifferentiate the inherently chiral phenoxathiin-based thiacalix[4]arenes. Both media, in combination with two solvents, allowed for enantiodiscrimination, which was, to the best of our knowledge, proved for the first time on inherently chiral compounds. Moreover, using the experimental RDCs, we investigated the calix[4]arenes conformational preferences in solution, quantitatively analyzed the differences in the alignment of the enantiomers, and discussed the pitfalls of the use of the RDC analysis.


Preparative separation of enantiomers
Preparative chiral separation of compounds was performed using an AutoPuritication system (Waters) equipped with an automated analytical control of the collected fractions.The analytical mode was utilized to screen the chromatographic conditions for the target preparative enantioseparation using a polysaccharide-based analytical column Chiral Art Amylose-SA (2504.6 mm i.d., 5 μm) (YMC, Germany).
After finding the best possible conditions for each substance, a chiral preparative column, Chiralpak IA (25020 mm i.d., 5 μm) from Daicel (Japan), was employed to perform their chiral resolution on multi-milligram scale.
While the separation of compounds II and III did not provide fully separated enantiomers, compounds 1 and 2 were successfully resolved.
Thus, the best conditions for chiral resolution of 2 consisted of a mobile phase comprising a heptane/propan-2ol (96/4, v/v) mixture with the flow rate set to 15 mL•min −1 .The corresponding retention times were 14.98 and 22.34 min.The separation was performed at room temperature (21-23 °C, air-conditioning).Injection volume was 0.5 mL and the sample concentration was 5 mg•mL −1 in a heptane/propan-2-ol (1/2, v/v) mixture.The detection wavelength was 235 nm, and the automatic fraction collection was set to ensure the highest possible amount of the collected substance and the highest possible purity of the respective fractions (Figure 4, main text).

NMR measurements
NMR spectra were recorded on spectrometer Bruker Avance III 600 MHz operating at 600.13 MHz for 1 H and 150.92 MHz for 13 C with 5 mm triple-resonance TCI cryoprobe.The anisotropic samples were measured using F1-coupled HSQC 4 spectra (2k data points in both the direct and the indirect dimension, NS = 8, scaling factor = 8).

Preparation and composition of anisotropic samples
The calculated amount of the alignment media was weighted directly in the NMR tube to which, subsequently, the analyte dissolved in the solvent was transferred.The amount of the alignment medium was derived from the volume of the solvent (the amount of the analyte was negligible) as well as from the molecular weight of the alignment medium itself.The weights of each part of the individual samples are in the Tables S1-S4.After mixing of all the components, the sample was left a few hours to dissolve spontaneously.Then, it was necessary to homogenize the solution due to its high viscosity.For that, manual centrifuging was used.Homogeneity of the sample was controlled by a handmade polarizer as well as by the measurement of 2 H NMR. This can reveal a possible inhomogeneity and/or residual signal of non-aligned solvent.If the latter is the case, further alignment medium needs to be added or part of the solvent evaporated.
Adequate alignment order can be also monitored by the measurement of quadrupolar splitting ΔQ.We took an advantage of using CDCl3 and THF-d8 as the solvents as reliable probes for examining the chiral properties of the sample.For isotropic samples, CDCl3 exhibits a singlet, while in an anisotropic sample, the singlet splits into a doublet.On the other hand, THF-d8 in 2 H NMR spectra has two doublets for the two non-equivalent methylenes in the achiral anisotropic sample, while in the chiral anisotropic sample, the signals split into four doublets.
Due to the alignment, a lock on the solvent might not be possible.Therefore, an external capillary with a different solvent is usually inserted into the sample.The choice of a suitable solvent is driven by the chemical shift of the solvent peaks in the 2 H experiment, as they should not interfere with the peaks of the solvent in which the alignment medium and analyte are dissolved.A DMSO-d6 capillary was selected for the combination PBLG + CDCl3 and PBLG + PBDG + CDCl3, 1,1,2,2-tetrachloroethane-d2 for PBLG + THF-d8 as well as for PBLG + PBDG + THF-d8 and PBPMLG + THF-d8 and an acetone capillary for PBPMLG + CDCl3.Table S3: Quantities of individual components for samples involving isolated enantiomers 1-E1 and 1-E2.

RDC data of racemic mixtures of 1 and 2 in PBLG/PBPMLG and CDCl 3 /THF-d 8
The assignment of the RDCs to E1 or E2 of the racemic mixtures of compounds 1 and 2 was performed by systematic variation of all possible assignments and subsequent selection of the combination with the best uncertainty-weighted quality factor q. 5 Due to the uncertainties in the conformational assignment, we fit the RDC data of the racemic mixtures also to the structures obtained by X-ray diffraction.** The measurements of the racemic mixture of compound 2 in the system PBPMLG/CDCl 3 did not provide any results.
Table S6.Weighted quality factors q for both enantiomers (E1 and E2) of the racemic mixture of 1 in PBLG and CDCl3.

Spectra of 1 and 2 in the racemic mixtures of PBLG-PBDG
All the spectra of compounds 1 and 2 in the racemic mixtures of PBLG-PBDG (1:1) in CDCl3 or THF-d8 were folded in the 13 C domain to measure smaller window, and, thus, obtain better resolution.

Crystallographic data for 2:
A slow evaporation of the solvent (CHCl3/EtOH mixture) was used to crystallize the product from its nearly saturated solution at room temperature.The structure of 2 was measured using a D8 VENTURE equipped with a Photon CMOS detector with Cu-Ka (λ=1.54178Å) radiation at 180 K.The structure was in a triclinic system, P 21/c space group with lattice parameters a=13.4176(3)Å, b=24.4722( 6
Small imaginary frequencies (always max.-20 cm -1 ) were found in the case of 2 (1,2-alternateAB with inverted rings A and B, 1,3-alternate, cone, partial cone with inverted ring A and partial cone with inverted ring D).These imaginary modes were found for the tert-butyl groups, sometimes with the ethoxy/methoxy groups present at the lower rim.The exact position of the rotation does not concern the stability of the optimized structure, nor the calculation of residual dipolar couplings, and, according to the literature, these could be neglected if the frequencies are less than tens of wavenumbers 9,12 .

Figure S3 :
Figure S3: 13 C-1 H HMBC NMR spectra of 2 in CDCl3 measured at room temperature at 600 MHz.

Figure S10 :
Figure S10: F1-coupled HSQC spectrum of 1 in the racemic mixture of PBLG-PBDG and CDCl3 measured at room temperature and 600 MHz.

Figure S11 :
Figure S11: F1-coupled HSQC spectrum of 1 in the racemic mixture of PBLG-PBDG and THF-d8 measured at room temperature and 600 MHz.

Figure S12 :
Figure S12: F1-coupled HSQC spectrum of 2 in the racemic mixture of PBLG-PBDG and CDCl3 measured at room temperature and 600 MHz.

Figure S13 :
Figure S13: F1-coupled HSQC spectrum of 2 in the racemic mixture of PBLG-PBDG and THF-d8 measured at room temperature and 600 MHz.
) Å, c=16.4261(4)Å, α=90° β=103.8661(11)°γ=90°, Z=2, V=2660.5(5)Å 3 , Dc=1.296 g/cm3 , μ(Cu-K α)=2.846mm -1 .The data reduction and absorption correction were done with the Apex3 software.The structure was solved by charge-flipping methods using the Superflip software and refined by full matrix least squares on F squared value using Crystals software to the final values R=0.0579 and wR=0.1431using 9581 independent reflections (Θmax=68.244°),680 parameters and 88 restraints.The MCE software was used for the visualization of residual electron density maps.According to common practice, the hydrogen atoms attached to carbon atoms were placed geometrically with Uiso(H) in the range 1.2-1.5 Ueq of the parent atom (C).The disordered functional groups were refined with restrained geometry and occupancy constrained to full for each atomic position.The crystal was partially solvated (0.487(2)) with chloroform.The structure was deposited into Cambridge Structural Database under number CCDC 2039755.

Table S1 :
Quantities of individual components for all the samples involving compound 1.

Table S2 :
Quantities of individual components for all the samples involving compound 2.

Table S5 .
Weighted quality factors q for both enantiomers (E1 and E2) of racemic mixtures 1 and 2 within PBLG and PBPMLG alignment media in CDCl3 or THF-d8.* All q factors were very similar with values around 0.1, therefore, we were not able to determine the correct conformation.

Table S7 .
One bond residual dipolar couplings 1 D(C-H) or 1 D(C-C) for both enantiomers (E1 and E2) of the racemic mixture of 1 in PBLG/CDCl3 at 300 K extracted from F1-coupled HSQC spectra.

Table S8 .
Weighted quality factors q for both enantiomers (E1 and E2) of the racemic mixture of 1 in PBLG and THF-d8.

Table S10 .
Weighted quality factors q for both enantiomers (E1 and E2) of the racemic mixture of 1 in PBPMLG and CDCl3.

Table S11 .
One bond residual dipolar couplings 1 D(C-H) or 1 D(C-C) for both enantiomers (E1 and E2) of the racemic mixture of 1 in PBPMLG/CDCl3 at 300 K extracted from F1-coupled HSQC spectra.

Table S12 .
Weighted quality factors q for both enantiomers (E1 and E2) of the racemic mixture of 1 in PBPMLG and THF-d8.

Table S14 .
Weighted quality factors q for both enantiomers (E1 and E2) of the racemic mixture of 2 in PBLG and CDCl3.

Table S15 .
One bond residual dipolar couplings 1 D(C-H) or 1 D(C-C) for both enantiomers (E1 and E2) of the racemic mixture of 2 in PBLG/CDCl3 at 300 K extracted from F1-coupled HSQC spectra.

Table S16 .
Weighted quality factors q for both enantiomers (E1 and E2) of the racemic mixture of 2 in PBLG and THF-d8.

Table S18 .
Weighted quality factors q for both enantiomers (E1 and E2) of the racemic mixture of 2 in PBPMLG and THF-d8.

Table S31 .
One bond residual dipolar couplings 1 D(C-H) and 1 D(C-C) of racemic mixture 1 in the racemic mixture of PBLG-PBDG and CDCl3 at 300 K extracted from F1-coupled HSQC spectra.

Table S32 .
Weighted quality factors q of the racemic mixture 1 in the racemic mixture of PBLG-PBDG and THF-d8.

Table S33 .
One bond residual dipolar couplings 1 D(C-H) of racemic mixture 1 in the racemic mixture of PBLG-PBDG and THF-d8 at 300 K extracted from F1-coupled HSQC spectra.

Table S34 .
Weighted quality factors q of the racemic mixture 2 in the racemic mixture of PBLG-PBDG and CDCl3.

Table S35 .
One bond residual dipolar couplings 1 D(C-H) and 1 D(C-C) of racemic mixture 2 in the racemic mixture of PBLG-PBDG and CDCl3 at 300 K extracted from F1-coupled HSQC spectra.

Table S36 .
Weighted quality factors q of the racemic mixture 2 in the racemic mixture of PBLG-PBDG and THF-d8.

Table S37 .
One bond residual dipolar couplings 1 D(C-H) and 1 D(C-C) of racemic mixture 2 in the racemic mixture of PBLG-PBDG and THF-d8 at 300 K extracted from F1-coupled HSQC spectra.