FEASIBILITY STUDY ON QUANTIFICATION OF SOIL MINERALS BY XRD-RIETVELD FULL-SPECTRUM FITTING METHOD-A CASE STUDY FROM COMMON PURPLISH SOIL IN SICHUAN BASIN
The solid framework of soil represents a complex mixture of crystalline and amorphous miner-als,whose precise mineral composition and quantification play a crucial role in evaluating soil properties,structure,and fertility levels.Traditional X-ray external scaling methods based on peak intensity(d-I/10)are hindered by the need for multiple reference intensity values selection,making it challenging to accu-rately reflect the true mineral content.Similarly,uncalibrated Rietveld full spectrum fitting methods still struggle with accuracy verification in quantifying complex natural soil minerals.In this study,we employ the Rietveld full spectrum fitting method to address these challenges.We select high purity single minerals with excellent crystallization and configure four sets of soil simulation samples with fixed proportions of 3-6 phases(quartz+hematite+kaolinite;quartz+hematite+kaolinite+illite;quartz+hematite+kaolin-ite+illite+calcite;quartz+hematite+kaolinite+illite+calcite+potassium feldspar).We comparatively discuss the quantitative results of the K value method,RIR value method,and Rietveld full spectrum fit-ting method.Furthermore,we conduct practical applications of the Rietveld full spectrum fitting method and quantitative determination of amorphous composition in the purplish soil widely distributed in the Si-chuan Basin.Our results demonstrate that the quantitative results of the Rietveld method(absolute devia-tion range=0.8％-4.8％,mean value=2.7％)and K value method(0.6％-8.8％,mean value=3.1％)outperform those of the RIR value method(absolute deviation range=1％-28％,mean value=11.2％)across the four groups of soil simulation samples.For low-content clay minerals,the Rietveld full spec-trum fitting method(absolute error range=1.8-3.8,mean value=2.3)exhibits greater sensitivity than the K-value method(absolute error range=1.4-5.5,mean value=3.0).The element distribution curve of the purplish soil sample,quantified using the Rietveld full spectrum fitting method,correlates highly with the chemical analysis(XRF)results,with a correlation of 99.78％,indicating that the quantitative results of the full spectrum fitting method can more accurately reflect the sample information.These find-ings highlight the potential of the unmarked XRD-Rietveld full spectrum fitting quantitative method to a-void analytical errors associated with conventional methods and effectively enhance the accuracy of quanti-tative results,thereby offering significant application prospects.
万方数据
