Quantitative Analysis of Iron and Silicon Concentrations in Iron Ore Concentrate Using Portable X-ray Fluorescence (XRF)
Quantitative Analysis of Iron and Silicon Concentrations in Iron Ore Concentrate Using Portable X-ray Fluorescence (XRF)
复制标题
使用便携式 X 射线荧光 (XRF) 定量分析铁精矿中的铁和硅浓度
DOI：
10.1177/0003702819871627
复制
发表时间：
2020
影响因子：
3.5
通讯作者：
Xie Shuyun
中科院分区：
文献类型：
--
作者：
Zhou Shubin;Yuan Zhaoxian;Cheng Qiuming;Weindorf David C.;Zhang Zhenjie;Yang Jie;Zhang Xiaolong;Chen Guoxiong;Xie Shuyun
便携式X射线荧光光谱仪（pXRF）作为一种快速、无损、多元素分析技术，在矿产勘查、环境评价、考古分析等领域有着广泛的应用。然而，在冶炼工业中进行的应用很少，特别是在分析精矿样品中的金属浓度时。分析了pXRF法测定铁精矿中金属含量的有效性。对于这一概念验证研究，从东北大学矿物加工实验室（中国沈阳）收集了以Fe和Si为主的铁矿石样品，并使用pXRF、基于实验室的XRF和滴定方法直接进行分析。发现精矿的密实度（密度）对pXRF读数的影响很小。Fe和Si的pXRF读数与基于实验室的XRF结果相当。基于pXRF和XRF结果之间的强相关性（Fe：R2 > 0.99，Si：R2 > 0.96），采用线性校准来提高pXRF读数的准确性。根据21个铁精矿样品的XRF结果和pXRF结果之间的关系推导出线性回归方程，用于校准pXRF，然后对另外5个样品进行验证。结果表明，普通最小二乘法（OLS）显著提高了预测精度，尤其是对铁元素的预测，相对误差（REs）从4.26%~ 8.32%降低到0.03%~ 3.27%。因此，pXRF显示出快速，定量分析铁精矿中的铁浓度的强大前景。根据本研究获得的结果，需要进行更大规模、更全面的研究来确认所获得的结果。
As a technique capable of rapid, nondestructive, and multi-elemental analysis, portable X-ray fluorescence (pXRF) has applications to mineral exploration, environmental evaluation, and archaeological analysis. However, few applications have been conducted in the smelting industry especially when analyzing the metal concentration in ore concentrate samples. This research analyzed the effectiveness of using pXRF in determining the metal concentration in Fe concentrate. For this proof of concept study, Fe ore samples dominated by Fe and Si were collected from the Northeastern University Mineral Processing Laboratory (Shenyang, China) and directly analyzed using pXRF, laboratory-based XRF, and titration methods. The compactness (density) of the ore concentrate was found to have very little effect on pXRF readings. The pXRF readings for Fe and Si were comparative to laboratory-based XRF results. Based on the strong correlations between the pXRF and XRF results (Fe: R2 > 0.99, Si: R2 > 0.96), linear calibrations were adopted to improve the accuracy of pXRF readings. Linear regression equations derived from the relations between XRF results and pXRF results of 21 Fe ore concentrate samples were used to calibrate the pXRF, and then validation was performed on five additional samples. Results from this preliminary study suggest that ordinary least squares (OLS) regression improves the accuracy dramatically, especially for Fe with relative errors (REs) decreasing to 0.03%–3.27% from 4.26%–8.32%. Consequently, pXRF shows strong promise for rapid, quantitative analysis of Fe concentration in Fe ore concentrate. Based on the results obtained in this study, a larger, more comprehensive study is warranted to confirm the results obtained.