On enantiomorph-polarity estimation

The behaviour of Rogers's η parameter for enantiomorph-polarity estimation is examined theoretically and experimentally on simulated intensity data for seven well-assorted compounds. An alternative parameter x, based on incoherent scattering from twin components related by a centre of symmetry, is also considered. It is found that both parameters are very well adapted to implementation in a least-squares program and converge well. The η parameter can give false and over-precise indications of chirality-polarity for structures which are nearly centrosymmetric, whereas the x parameter does not have this fault and converges more rapidly than η.