The Rietveld profile‐analysis refinement procedure [Rietveld, J. Appl. Cryst. (1969), 2, 65–71] is here applied to microdensitometer‐measured Guinier–Hägg X‐ray powder film data (strictly monochromated Cu Kα1 radiation). Various functions for describing the profile form of an individual reflection are tested; the most satisfactory fit is obtained for a modified and asymmetrized Lorentz function. The crystal structure of α‐Bi2O3 has been refined and no significant deviations are observed from the results of an earlier single‐crystal study [Malmros, Acta Chem. Scand. (1970), 24, 384–396]. An RF value [roughly comparable with the conventional R(F) value of single‐crystal work] of 0.068 was obtained, with standard deviations on the refined positional parameters on average three times larger than those obtained from the single‐crystal study. Further examples of the refinement technique are also given which illustrate both its reproducibility and range of application.

No abstract

No abstract

Full-profile analysis of microdensitometer-measured Guinier-H~igg X-ray powder film data (strictly monochromatized Cu K~I radiation) [see Malmros & Thomas (1977). J. Appl. Cryst. 10,[7][8][9][10][11] is here extended to treat a multicomponent system where two of the components are refinable. The method is demonstrated for the case of: (a) ~-Bi203/KC1 (standard), and (b) CaF2/SiO2 (s-quartz) together with a Mg(OH)2 matrix. In (a), a weight ratio BizO3/KCI of 1:1 gave an experimental result of 0-96(9):1. In (b), a number of admixtures have been studied in the range 1-60 wt% SiO2 all with 10 wt % CaF2. The agreement between weight ratios and profile refinement results suggests that the formal temperature factors function as effective absorption correction parameters. It is thus shown that approximate quantitative analysis is feasible from Guinier-H~igg diffraction data without any previously determined calibration curve. Moreover, the method permits the use of all available overlapping data in the refinement of a structure without needing to exclude data from known impurities [see Waltersson, Werner & Wilhelmi (1977). Cryst. Struct. Commun. 6,[231][232][233][234][235].This feature is likely to prove the most common application of the method.

No abstract