(±)-N-(3-Hydroxy-1,2-diphenylpropyl)-4-methylbenzenesulfonamide

In the title compound, C22H23NO3S, the relative stereochemistry of the two stereogenic centres is anti with respect to the H atoms. The molecular packing of the crystal shows a double-strand arrangement, consisting of one strand of (S*,S*) enantiomers and one strand of (R*,R*) enantiomers. Both strands lie parallel to each other along the a axis. Each strand is made up of dimers in which the molecules are connected to each other via an intermolecular O—H⋯O hydrogen bond between the hydroxyl groups and an O—H⋯π interaction with the aromatic ring. These units are then connected to neighbouring dimers via N—H⋯O hydrogen bonds and C—H⋯O interactions. Intramolecular C—H⋯O interactions are also observed.

In the title compound, C 22 H 23 NO 3 S, the relative stereochemistry of the two stereogenic centres is anti with respect to the H atoms. The molecular packing of the crystal shows a doublestrand arrangement, consisting of one strand of (S*,S*) enantiomers and one strand of (R*,R*) enantiomers. Both strands lie parallel to each other along the a axis. Each strand is made up of dimers in which the molecules are connected to each other via an intermolecular O-HÁ Á ÁO hydrogen bond between the hydroxyl groups and an O-HÁ Á Á interaction with the aromatic ring. These units are then connected to neighbouring dimers via N-HÁ Á ÁO hydrogen bonds and C-HÁ Á ÁO interactions. Intramolecular C-HÁ Á ÁO interactions are also observed.

Comment
The title racemic sulfonamide was obtained unintentionally as a product from the study of the organocatalytic α-oxidation of phenylacetaldehyde catalysed by (S)-proline. The relative stereochemistry of the two stereogenic centres was established by X-ray crystallography as anti with respect to the H atoms of C8 and C15 (Fig. 1).
The molecular packing of the crystal shows a double strand arrangement, which consists of one strand of (8S*,15S*) enantiomers and one strand of (8R*,15R*) enantiomers. Both strands lie parallel to each other along the a axis and a number of hydrogen bonds has been observed throughtout the crystal lattice.
After 1 h, sodium borohydride (151 mg, 4.00 mmol) was added to the mixture at 273 K and the mixture was stirred overnight.
The mixture was then poured onto a biphasic mixture of HCl (1 mol l -1 ) and EtOAc (1:1, 8 ml) at 273 K and vigorously stirred for 10 minutes. The organic phase was separated and the aqueous phase was extracted with EtOAc (8 ml x 4). The combined organic extracts were washed with brine, dried over MgSO 4 and concentrated in vacuo to afford a yellow oil.

Refinement
Hydrogen atoms attached to carbon and nitrogen atoms were placed in calculated positions and refined using the riding model (N-H = 0.86 Å & C-H 0.93-0.97 Å), with U iso (H) = 1.2 and 1.5U eq (parent atom) for the nonmethyl and methyl groups, respectively. The hydroxyl H-atom was disordered over two sites involved in either O-H···O hydrogen bonding to a neighboring alcohol or O-H···π interactions with a neighboring phenyl ring. In the final refinement these two hydrogen atoms were included, fixed in these two positions. After the final refinement a peak of electron density of 1.05 e Å -3 , distanced supplementary materials sup-2 0.82 Å from the sulfonamide oxygen O2, was observed. No evidence of disorder could be discerned. This peak was also present in an alternate refinement using data that had been corrected for absorption. This refinement was indistinguishable from structure presented here. Fig. 1. The molecular structure and atom numbering scheme of the title compound with displacement ellipsoids drawn at the 50% probability level for non-H atoms.   Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating Rfactors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.