BACKGROUND Serpentine is an important group of minerals, widely used in the chemical and building industries, agriculture, metallurgy, environmental protection, and other fields. Due to the high requirements for the mining grade of serpentine ores in China, many abandoned tailings have been formed, resulting in a large amount of waste causing environmental pollution. According to the different content of serpentine in the waste tailings, different industrial products can be produced, which can improve the natural environment and diversify serpentine enterprises. These works are based on the ability to determine serpentine contents quickly and accurately. At present, the determination method of serpentine ores is based on the industry standard HG/T 3575—2006 Analysis Method of Serpentine Ores. The content of magnesium is measured first, and the content of serpentine mineral is obtained after conversion. The analysis steps are tedious, and the analysis time is long. The existing detection methods obviously cannot meet the urgent needs of the market and enterprises. For a long time, the precision of the quantitative analysis of serpentine by X-ray diffraction has not been controlled quantitatively. The precision comparison between laboratories also lacks a corresponding basis.

OBJECTIVES It is urgent to improve the detection level of X-ray diffraction and further improve the detection technology of serpentine mineral content. The purpose of this study is to lay a foundation for the establishment of the standard method for the quantitative analysis of serpentine phase in the future and provide a basis for the evaluation of analytical methods and the monitoring of analytical quality.

METHODS This study refers to the method widely used internationally to express precision with repeatability limit and reproducibility limit, and to judge the reliability of analysis results in practice. According to the requirements of GB/T 6379.2—2004 standard, 10 samples with different serpentine content are selected and prepared for uniformity test. The serpentine content ranges from 5.00% to 84.78%, which basically covers the content range of serpentine deposits in China, meeting the requirements of the research program. The purpose of the uniformity test is to reduce the error of quantitative analysis, especially in the case of low content, because the uniformity of the sample has a great impact on the accuracy of the test results. The measured value of F_measured is less than the critical value of F_critical, indicating that the analyzed samples are uniform. Eight laboratories are selected to use the qualified samples for the precision collaborative test. According to the requirements of the quantitative limit and the optimal measurement range, and the results of the precision, accuracy and recovery test, the standard curve method has a small error, accurate test data, high recovery rate and stronger applicability for samples with serpentine mineral content of less than 20.00%. For high-content samples with serpentine mineral content of 20% or more, the parameters of the K value method all meet the requirements, considering the fast and convenient determination, it is more suitable for the application of the K value method. The laboratories participating in the collaborative test use the uniformly supplied analytical pure corundum reagent as the reference material, and use the X-ray diffraction standard curve method (provide the calibration curve) for samples with serpentine mineral < 20%, while use the X-ray diffraction K value method (provide K value) for samples with serpentine mineral ≥20% under the same test conditions. Only the data passing the test of Grubbs and Cochran can participate in the calculation of precision parameters, and the mathematical function relationship between the precision and content of X-ray diffraction standard curve method and K-value method is obtained.

RESULTS The experimental results show that the critical value of F_critical under 95% confidence is 3.50 for the collaborative test, while the measured value of F_measured after statistical calculation is less than 3.50, which proves that the 10 samples are relatively uniform and are qualified as the precision collaborative sample. Moreover, the data of the collaborative test are all less than 5% of the critical value, and all of them pass the Grubbs test and Cochran test. All of the data can be included in the calculation of precision parameters, which also proves the high consistency of the test results of different instruments in different laboratories.

CONCLUSIONS The calculation results of precision parameters show that the precision parameters of X-ray diffraction standard curve method and K-value method are positively correlated with the content of serpentine minerals, that is, the higher the content of serpentine minerals, the greater the repeatability limit and reproducibility limit, and vice versa. The sample with the lowest content of serpentine has the lowest repeatability limit and reproducibility limit, which also verifies that the error of X-ray diffraction for phase quantitative analysis mainly comes from the uniformity error of the sample. It confirms that among all factors affecting the results of X-ray diffraction quantitative analysis, the homogeneity of the sample is the most critical factor, and the influence of other factors can be overcome by changing the test conditions and other corresponding methods. If the sample is uniform enough, the error of quantitative analysis can be controlled. The research on the precision of the quantitative analysis method of serpentine phase provides technical support and basis for the establishment of the standard method of the quantitative analysis of serpentine phase, the evaluation of the analysis method and the monitoring of the analysis quality in the future.