基于实验室间协作实验评估土壤中重金属能量色散X射线荧光光谱分析方法性能

为考察土壤、水系沉积物中重金属元素能量色散X射线荧光光谱(ED-XRF)分析方法性能,组织了14家检测实验室利用统一提供的5个土壤和6个水系沉积物标准样品,在6家国内外仪器厂商生产的能谱仪(落地式、台式或便携式)上对As、V、Cr、Mn、Ni、Pb、Cu、Zn、Co、Mo等重金属元素进行了定量测试。方法检出限采用重复测定低含量重金属样品测试结果的标准偏差数据计算。各家实验室数据用稳健统计法处理后的结果,乘以系数1.6得到检出限限值。对数据进行统计分析前,基于土壤样品重金属常规检测质控要求和曼德尔h/k检验法对测试数据进行了筛选。各元素测试数据采用率(%)分别为:Zn(93.3～97.0),Pb(95.2～96.4),Mn(92.3～96.2),Cu(89.4～93.2),As(90.1～92.3),Ni(86.5～89.3),Cr(84.6～88.3),V(81.4～84.8)。采用率反映了目前检测实验室重金属测试结果准确度高低的现状,其中V、Cr采用率不高的主要原因可能与部分实验室没有有效扣除重叠峰干扰有关。采用谱图拟合获取强度的应用程序采用率优于区间积分法。相关实验室需要进一步优化...

Evaluation of analytical method performance for determination of heavy metals in soils by energy dispersive X-ray fluorescence spectrometry based on inter-laboratory collaborative experiments

In order to investigate the performance of analytical method for determination of heavy metals in soil and sediment by energy dispersive X-ray fluorescence spectrometry (ED-XRF), 14 testing laboratories were organized. Five soil and six sediment standard samples were provided uniformly. Quantitative tests on heavy metal elements including As, V, Cr, Mn, Ni, Pb, Cu, Zn, Co, and Mo were performed on spectrometers (desktop, floor-standing or portable) made by 6 domestic and foreign instrument manufacturers. The limit of detection of the method was calculated using the standard deviation data of repeated test results of samples with low content of heavy metals. The limit value was obtained using the treated results in each laboratory by robust statistical method to multiply by a coefficient of 1.6. Before performing statistical analysis of data, the measurement results were screened in advance based on the quality control requirements of conventional heavy metal analysis and Mandel test (h/k test method). The data adoption rate (%) of each element was listed as follows: Zn (93.3-97.0), Pb (95.2-96.4), Mn (92.3-96.2), Cu (89.4-93.2), As (90.1-92.3), Ni (86.5-88.9), Cr (84.6-88.3) and V (81.4-84.8). The adoption rate reflected the status of the accuracy of heavy metal test results for all laboratories involved. The main reason for the lower adoption rate of V and Cr was possibly related to the fact that the interference of overlapping peaks was not effectively deducted in some laboratories. The adoption rate of application program using spectrum fitting to obtain the intensity was better than that of interval integration method, and the and the relevant laboratories needed to further optimize and improve the application program. The measurement of heavy metals including As, Cr, Mn, Ni, Cu, Zn, Pb, and V in soil-sediment samples by ED-XRF in most laboratories could meet the requirements of quantitative analysis. Since the content range of Co in standard samples was relatively narrow, the precision and accuracy should be further evaluated. Only when the mass fraction of Mo in sample was higher than 1mg/kg, the requirements of quantitative analysis could be met.