应用便携式X射线荧光光谱仪(PXRF)分别在原位和实验室条件下对53个土壤样品中的Cu, Pb, As, Cr, Ni和Zn等重金属进行测定, 并与原子吸收/原子荧光法测定值进行对比, 建立一元线性回归模型分析PXRF数据质量。 通过测定土壤样品原位含水量并选取部分样品进行室内水分定量实验, 分析土壤水分对于PXRF测定结果的影响。 结果表明, PXRF检出限分别为Cu: 10.6 mg·kg-1, Pb: 8.1 mg·kg-1, As: 5.7 mg·kg-1, Cr: 22.5 mg·kg-1, Ni: 21.6 mg·kg-1, Zn: 10.4 mg·kg-1; 原位测定时Pb, Cr, Ni和Zn可以达到定量水平; 经过风干磨细处理, Cu, Pb, Cr, Ni和Zn在实验室条件下可以达到定量水平, 说明X射线荧光光谱法适用于土壤重金属的快速测定与评价。 水分对于PXRF测定结果具有“稀释”作用, 原位条件下土壤含水量<15%时与>25%时样品的平均相对误差分别为-17%与-31%; 实验室条件下土壤含水量从风干土水平提高到30%, 测定的平均相对误差由10%变为-24%。 土壤水分升高可能会导致数据质量和准确性降低, 建议原位测定时控制土壤含水量在25%以内。

Total concentrations of Cu, Pb, As, Cr, Ni and Zn were determined for 53 soil samples using portable X-ray fluorescence (PXRF) system in in-situ and ex-situ (Lab.)conditions. PXRF metal concentrations were statistically compared with analytical results from traditional AAS/AFS analysis. The ability of PXRF instrument to produce comparable analytical results to the reference method was assessed by linear regression. To investgate the effects of soil moisture on PXRF, the in-situ moisture content of all soil samples was quantified and the metal concentrations of selected samples with known moisture contents were measured too. The results showed that the detection limits of PXRF for Cu, Pb, As, Cr, Ni and Zn were 10.6, 8.1, 5.7, 22.5, 21.6 and 10.4 mg·kg-1, respectively. A good degree of linearity was found for Pb, Cr, Ni and Zn in in-situ condition. While in ex-situ condition, quantitative level data were achieved across the entire range of samples tested for Cu, Pb, Cr, Ni and Zn. X-ray fluorescence spectrometry was shown to be an effective tool for quantification and rapid assessment of heavy metals in soils. Soil moisture content did affected the performance of PXRF, the mean percent difference for soil samples in-situ with moisture content less than 15% and higher than 25% was -17% and -31% respectively. In ex-situ condition, as the soil moisture content increased from air dried level to 30%, the mean percent difference decreased from 10% to -24%. The dilution effect of moisture in soils may cause discrepancies with conventional analytical results and induce worse data quality, and it should be controlled within 0～25% in in-situ condition.