• DOI:10.1107/S0108768199003948
• Corpus ID: 6050431

Absolute configurations of Emycin D, E and F; mimicry of centrosymmetric space groups by mixtures of chiral stereoisomers.

@article{Walker1999AbsoluteCO,
  title={Absolute configurations of Emycin D, E and F; mimicry of centrosymmetric space groups by mixtures of chiral stereoisomers.},
  author={Walker and Pohl and Herbst-Irmer and Gerlitz and Rohr and Sheldrick},
  journal={Acta crystallographica. Section B, Structural science},
  year={1999},
  volume={55 Pt 4},
  pages={
          607-616
        }
}

• Walker, Pohl, +3 authors Sheldrick
• Published 1 August 1999
• Chemistry
• Acta crystallographica. Section B, Structural science

The crystal structures of Emycin E (1), di-o-bromobenzoyl-Emycin F (2) and o-bromobenzoyl-Emycin D (3) have been determined by X-ray analysis at low temperature. Emycin E and o-bromobenzoyl-Emycin D both crystallize with two molecules in a triclinic unit cell. These two structures can be solved and refined either in the centrosymmetric space group P1;, with apparent disorder localized at or around the expected chiral centre, or in the non-centrosymmetric space group P1 as mixtures of two diastereomers without disorder. Only the latter interpretation is consistent with the chemical and spectroscopic evidence. Refinements in the centrosymmetric and non-centrosymmetric space groups are compared in this paper and are shown to favour the chemically correct interpretation, more decisively so in the case of the bromo derivative as a result of the anomalous dispersion of bromine. Structures (1) and (3) provide a dramatic warning of the dangers inherent in the conventional wisdom that if a structure can be refined satisfactorarily in both centrosymmetric and non-centrosymmetric space groups, the former should always be chosen. In these two cases, despite apparently acceptable intensity statistics and R factors (5.87 and 3.55%), the choice of the centrosymmetric space group leads to the serious chemical error that the triclinic unit cell contains a racemate rather than two chiral diastereomers! The weakest reflections are shown to be most sensitive to the correct choice of space group, underlining the importance of refining against all data rather than against intensities greater than a specified threshold. The use of similar-distance restraints is shown to be beneficial in both P1 refinements. Di-o-bromobenzoyl-Emycin F crystallizes in the monoclinic space group P2(1) with one molecule in the asymmetric unit and so does not give rise to these problems of interpretation. The absolute configuration of the two bromo derivatives, and hence the Emycins in general, was determined unambiguously as S at the chiral centre C3.