Abstract

In an earlier treatment Destro & Marsh (1987). Acta Cryst. A43, 711-718, an attempt was made to predict the shapes of high-angle -2 scan profiles by convoluting a low-angle profile with the presumedly known spectral distribution function for the incident (crystal-monochromated Mo K) radiation but it was found necessary to introduce a third component, an 'aberration function', that varied with the Bragg angle . It is shown here that the primary purpose of the aberration function is to correct for defects in the spectral-distribution function. In particular, the effective intensity ratio between the K2 and K1 spectral lines can apparently deviate greatly (by more than 10%) from the theoretical value of 0.499, depending upon the alignment of the monochromator crystal, and an appreciable amount of white radiation may also be present. By a suitable modification of the spectral-distribution function, high-angle scan profiles can be predicted from accurate measurements of a low-angle profile; as a result, scan-truncation losses can, for the most part, be removed. However, modeling the spectral distribution function appropriate to a particular experiment remains a difficult empirical procedure.