Crystal structures of (R S)-N-[(1R,2S)-2-benzyloxy-1-(2,6-dimethylphenyl)propyl]-2-methylpropane-2-sulfinamide and (R S)-N-[(1S,2R)-2-benzyloxy-1-(2,4,6-trimethylphenyl)propyl]-2-methylpropane-2-sulfinamide: two related protected 1,2-amino alcohols

Molecules of two related 1,2-amino alcohols are linked via N—H⋯O=S hydrogen bonds, forming chains along [100] for the first compound and along [010] for the second compound.


Chemical context
1,2-Amino alcohols are found in a variety of pharmaceutically active compounds (Lee & Kang, 2004) and have been used extensively as components of chiral ligands and auxiliaries in asymmetric synthesis (Ager et al., 1996;Pu & Yu, 2001). In order to develop new chiral ligands and as part of an advanced undergraduate laboratory course, we sought to make a series of 2-aryl-1-methyl-1,2-amino alcohols. The most straightforward synthesis of these compounds was reported by Ellman (Tang et al., 2001;Evans & Ellman, 2003). The method relies upon the chiral ammonia equivalent, 2-methyl-2-propanesulfinamide (tert-butanesulfinamide), which is readily available from a variety of commercial sources or easily synthesized on scale (Weix et al., 2005). In the original Ellman report, the absolute configuration of the products was determined by deprotection of the amine and alcohol, cyclization to form the corresponding oxazolidinone, and correlation of the 1 H NMR spectra with the literature (Zietlow & Steckhan, 1994).
We report herein on the syntheses and structures of two different but related protected 1,2-amino alcohols, (1) and (2), from the addition of an arylmagnesium bromide to an N-tert-butanesulfinyl imine (Evans & Ellman, 2003). The reaction of imine (3a) with xylylmagnesium bromide, (4a), (see Fig. 1) resulted in a mixture of amino alcohol products from which the major product of the reaction, (1), was isolated in 27% yield after chromatographic separation of the diastereomers. The stereochemistry of this major product was confirmed by X-ray diffraction and the result is consistent with the sense of induction reported by Evans & Ellman (2003).
The analogous reaction with mesitylmagnesium bromide, (4b), also resulted in a mixture of products, from which the major product, (6), was isolated in 43% yield. A mixture of other diastereomers was also isolated, from which a crystal suitable for X-ray diffraction was grown. Unexpectedly, X-ray (Top) Reaction scheme depicting the synthesis of (1) and (5) from (3a), for which (1) is the major product of the reaction. (Bottom) Reaction scheme depicting the synthesis of (6) and (7) from (3a), and (8) and (2) from (3b), for which (6) is the major product of the reaction from (3a), and (8) is the major product from (3b).

Figure 2
The molecular structure of compound (1), with atom labelling. Displacement ellipsoids are drawn at the 50% probability level.

Figure 3
The molecular structure of compound (2), with atom labelling. Displacement ellipsoids are drawn at the 50% probability level. analysis showed this crystal to be (2), a product that could only have derived from a diastereomerically different isomer of (3a). Upon further investigation, we discovered that the starting material, which we had assumed was pure (3a), contained the minor diastereomer, (3b), in about 8% (determined by 1 H NMR; Fontenelle et al., 2014), which had formed due to racemization in the synthesis of (3a). Based on the work of Evans & Ellman (2003), it was deduced that (2) is the minor product expected from the reaction of (3b) with an arylmagnesium bromide. Although no further separations were performed on this mixture that contained (2), it follows that the other diastereomers present were (7), the minor product from the reaction with (3a), and (8), the major product from the reaction with the slight impurity of (3b).

Structural commentary
The molecular structures of compounds (1) and (2) are illustrated in Figs. 2 and 3, respectively. The essential difference in the conformation of the two compounds is that the phenyl ring (C5-C10) is inclined to the benzene ring (C11-C16) by 28.52 (7) in (1) and by 44.65 (19) in (2).

Figure 5
A partial view of the crystal packing of compound (2), illustrating the formation of the hydrogen-bonded chains along [010] (hydrogen bonds are shown as dashed lines; see Table 2 for details). Displacement ellipsoids are drawn at the 50% probability level.
toluene (20 ml) were added and the mixture was cooled to 195 K under nitrogen. The Grignard reagent (4a) or (4b) in toluene was placed under positive nitrogen pressure and was added to the Schlenk flask dropwise by cannula at 195 K. The reaction was stirred at 195 K and stopped when complete consumption of the imine was confirmed by thin-layer chromatography (30% ethyl acetate in hexanes, stained with ceric ammonium molybdate). The reaction was quenched with aqueous saturated sodium sulfate (1.5 ml), then the mixture was warmed to room temperature, dried over sodium sulfate, filtered through Celite, and the solvent was removed under reduced pressure. The ratio of diastereomers was determined by 1 H NMR of the crude material, specifically by examining the amine (N-H) proton resonances. The chemical shifts of anti diastereomers like (1) and (6) were found around = 3.78 p.p.m., while those for syn diastereomers were found slightly further upfield at = 3.61 (mixture, see below) and 3.66 (5) p.p.m.. The crude viscous yellow oil was purified by column chromatography. Crystals suitable for single-crystal X-ray diffraction were obtained from slow evaporation of methanol solutions.

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 3. For (1), the absolute configuration was determined using 4260 quotients, which gave a Flack parameter of 0.005 (12). The value obtained without D obs (h) as a restraint was À0.02 (3), calculated from 5203 Friedel pairs. For (2), the absolute configuration was determined using 1713 quotients, which gave a Flack parameter of 0.03 (6). The value obtained without D obs (h) as a restraint was À0.04 (8), calculated from 2882 Friedel pairs. In (2), the needle-shaped crystal diffracted weakly at higher angles. The cut-off resolution of 0.72 Å was chosen to maximize the number of enantiomerdetermining reflections, while limiting the inclusion of very weak high-angle data. The largest residual peak of 0.72 e Å À3 is located in the S1-C20 bond.
For both structures, the amine H atoms were located from difference Fourier maps and freely refined. The C-bound H atoms were placed geometrically and treated as riding with C-H = 0.95-1.00 Å and with U iso( H) = 1.5U eq (C) for methyl H atoms and = 1.2U eq (C) for other H atoms.

Special details
Experimental. Dry solvents were prepared from ACS grade, inhibitor free solvents by passage through activated molecular sieves in an Innovative Technology solvent purification system. CDCl 3 was purchased from Cambridge Isotope Laboratories, Inc., and dried over molecular sieves. 1 H and 13 C NMR spectra were recorded on an Avance 500 MHz spectrometer with residual protiated solvent as a reference. Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. Refinement. The amine H atom was found from the difference Fourier map and refined freely. All other H atoms were placed geometrically and treated as riding atoms: methine, C-H = 1.00 Å with U iso (H) = 1.2U eq (C), methylene, C-H = 0.99 Å with U iso( H) = 1.2U eq (C), methyl, C-H = 0.98 Å with U iso( H) = 1.5U eq (C), sp 2 , C-H = 0.95 Å with U iso( H) = 1.2U eq (C). The absolute configuration was deterimined using 4260 quotients, which gave a Flack parameter of 0.005 (12) (Parsons andFlack, 2004, Parsons et al., 2013). The value obtained without D obs (h) as a restraint was -0.02 (3), calculated from 5203 Friedel pairs (Flack, 1983).