Rietveld全图拟合法定量分析伊利石矿的精确度及误差来源

伊利石是一种重要的矿产资源，准确获得矿石中矿物组成和含量具有重要的理论和实际意义。Rietveld全图拟合法采用整个衍射图进行分析，具有较高的准确度。然而由于缺少纯的伊利石样品，目前对于该方法在分析伊利石矿样的精确度和误差来源尚不清楚。本文采用Rietveld全图拟合法对人工配制和天然的伊利石典型矿样进行了分析和计算，并与X射线荧光光谱(XRF)分析结果进行了对比，探讨了该方法的精确度及产生误差的因素。结果表明：人工配制的伊利石-石英二元混合物绝对误差范围为-0.9%~0.9%；伊利石-石英-钠长石三元混合物绝对误差范围为-1.9%~1.6%；伊利石-石英-钠长石-方解石-高岭石多元混合物绝对误差范围-2.3%~1.6%。天然伊利石矿样的Rietveld定量结果与XRF分析结果具有很好的一致性。表明该方法在分析伊利石矿样中具有较高的精确度，误差来源主要受伊利石结构模型、各物相中原子热振动参数Uiso取值合理性，以及择优取向等因素的影响。因此，在实际分析伊利石矿样时，建议根据实际情况选择合理的矿物结构模型，原子热振动Uiso取值应参考文献进行合理设定，制备试样应尽量减弱择优取向。

Accuracy and Error Sources of the Rietveld Full Pattern Fitting Method in Quantitative Analysis of Illite Ores

BACKGROUND: Illite is an important mineral resource. It is of great theoretical and practical significance to accurately obtain the mineral composition and content of illite ores. The Rietveld full pattern fitting method uses the whole diffraction pattern for analysis and shows high accuracy. However, due to the lack of pure illite samples, the accuracy and error sources of this method for analyzing the content of illite ores are not clearly known at present.OBJECTIVES: To understand the accuracy and error sources of the Rietveld full pattern fitting method in quantitative analysis of illite ores.METHODS: The Rietveld full pattern fitting method was used to quantitatively analyze artificial and natural illite ore samples. The results of X-ray fluorescence spectrometry (XRF) of natural illite ores were compared with the chemical compositions calculated from the mineral contents by the Rietveld method.RESULTS: The results showed that the maximum absolute error ranges of illite-quartz binary mixtures, illite-quartz-albite ternary mixtures, and illite-quartz-albite-calcite-kaolinite multiple mixtures were -0.9%-0.9%, -1.9%-1.6%, and -2.3%-1.6%, respectively. The chemical compositions of natural illite ores calculated by Rietveld method were in good agreement with the results of XRF. This indicated that the Rietveld method had high accuracy in the quantitative analysis of mineral contents of natural illite ore samples. The error sources were mainly affected by the illite structural model, atomic thermal displacement parameters U_iso, and preferred orientation.CONCLUSIONS: A reasonable structural model of illite should be chosen according to the actual samples. The values of atomic thermal vibration U_iso should be reasonably set according to the references. Preferred orientation should be decreased as much as possible during the sample preparation.