Rietveld Refinement of 2-theta Split Ranges - a Method for Reducing Analysis Time

This paper presents a method of reducing the X-ray diffraction data collection time, while maintaining high data quality, suitable for Rietveld refinement. The approach taken here is that phases in general show peak broadening as well as a loss of intensity with increasing 2-theta. Thus, by splitting the full analyzed 2-theta range into regions and adjusting the step-size and counting time to an appropriate level for each 2-theta range (i.e., small step-size and lower counting time at the lower angles, and larger step-size and larger counting time at higher angles) the total analysis time can be substantially shortened. In these combined refinements, refineable parameters should be coupled across the data sets. The use of this method is illustrated based on analyses of two NIST standards (LaB6 SRM660 and Si SRM640c), and a laboratory synthesized cuspidine (Ca4Si2O7F2) representing a more complicated phase. The proposed method cannot cut down the analysis time, to levels as those promised by never equipments, but it will reduce the time to half or less. INTRODUCTION New types of X-ray diffractometers are available on the market which promise magnificent data collected in a fraction of the time needed on older equipment. So what can we do if we cannot afford to purchase one of these wonder machines, but have to make good with what we have in the lab? Due to the effect of several contributing factors the intensity of peaks in diffraction patterns will decrease with increasing angle. For further description of these factors and their influence on the diffraction pattern, as well as the impact this may have on Rietveld refinements, please see Madsen and Hill [1]. Madsen and Hill [1] suggested using a data collection method which they named variable-counting-time (VCT). The VCT approach was based on gradually increasing the counting time with increasing angle. However, not all Rietveld refinement programs can handle this type of data making it impossible to use this approach. Thus, the approach taken here is to divide the analyzed 2-theta range into regions and adjusting the stepsize and counting time to an appropriate level for each 2-theta range (i.e., small step-size and lower counting time at the lower angles, and larger step-size and larger counting time at higher angles). Copyright©JCPDS International Centre for Diffraction Data 2003, Advances in X-ray Analysis, Volume 46. 220