Characterisation of preferred orientation in crystalline materials by x-ray powder diffraction.

Abstract

Texture, i.e. preferred orientation, can cause large systematic errors in quantitative analysis of crystalline materials using x-ray powder diffraction (XRPD) data. Various mathematical forms have been proposed for the application of preferred orientation corrections. The most promising of these appears to be the single-parameter March (1932) model proposed by Dollase (1986).Li and O'Connor (1989) applied the March model to determine the level of preferred orientation in various gibbsites using two procedures. The first involved the Rietveld (1969) least squares pattern-fitting method. Each pattern was Rietveld-analysed in two ways, initially assuming random orientation of the crystallites and subsequently with the March model. The second procedure for preferred orientation analysis, described here as the line ratio method, determines preferred orientation factors according to the intensity ratios of carefully selected line pairs.In the thesis the procedures proposed by Li and O'Connor for texture analysis have been evaluated with XRPD data sets for molybdite, calcite and kaolinite. The results indicate that while the March formula improves agreement between the' calculated and measured patterns in Rietveld analysis, other forms of systematic error in the intensity data appear to limit the effectiveness of the March formula in general. It has been found also that the line ratio method improves agreement between the data sets, but less effectively than the Rietveld method. It is proposed that extinction is likely to be the most influential source of systematic error competing with texture.