同步辐射X射线荧光光谱测定沉积物中元素含量的归一方法研究——以四海龙湾纹层沉积物为例

RESEARCH ON NORMALIZATION METHOD FOR ELEMENT ANALYSIS OF SEDIMENT WITH SYNCHROTRON RADIATION X-RAY FLUORESCENCE(SRXRF)——AN EXAMPLE OF VARVED SEDIMENT IN LAKE SIHAILONGWAN,NORTHEAST CHINA

同步辐射X射线荧光光谱分析(SRXRF)是一种基于同步加速器光源与X射线荧光光谱分析(XRF)相结合的原位分析方法,可以直接对沉积物进行扫描获得每个点的能谱图,从而计算元素含量的相对变化。但是由于荧光光谱定性和定量分析中存在的问题,国内发表文献中各种XRF原位分析数据均以相对含量或者计数表示。本文通过对各种标准物质(土壤、沉积物和岩石)的SRXRF分析,结合标准样品的定量数据,对同步辐射X射线荧光光谱的定量方法进行了研究。鉴于元素质量数和谱线强度的差异,不同元素需要运用不同的归一方法。对于K和Ca元素,采用ROI_(0-195)背景归一;Ti元素采用ROI_(293-303)背景归一;Mn元素采用ROI_(214-222)背景归一;而较高序数的元素Fe采用拟合康普顿散射峰归一。通过上述归一方法,我们获得了K、Ca、Ti、Mn和Fe元素的定量标准曲线,相关系数分别为0.89、0.98、0.96、0.93和0.98。运用标准样品对各元素的定量标准曲线进行检验,结果表明:Fe元素含量预测的平均相对偏差小于10%,K、Ca和Ti的平均相对偏差小于20%,Mn的相对偏差较大,为35.9%。从总体上看,SRXRF获得的数据可以基本满足湖泊沉积物中元素分析的定量化需求。为验证SRXRF方法能否揭示元素含量的季节性变化,我们对四海龙湾纹层沉积物进行了分析,按选定的归一方法计算样品中元素特征谱线的强度,利用各元素的标准曲线计算岩芯沉积物中元素含量。K、Ca、Ti、Mn和Fe元素含量具有明显的季节韵律变化,浅色碎屑微层中Ti和Ca含量较高,而Mn和Fe丰度的高值对应于暗色有机质-碎屑混合层。初步研究结果表明同步辐射X射线荧光光谱分析可以揭示湖泊中沉积的元素的季节性变化,对研究过去季节-年尺度的古气候、古环境变化具有重要的科学意义。

Synchrotron radiation X-ray fluorescence （SRXRF） is an in situ analysis method based on source of synchrotron accelerator and X-ray fluorescence.Relative changes of element content can be acquired through scanning the sediments to gain spectrum of every point.However,almost all published XRF data based on the in situ analysis are expressed in relative content or counts since the problems existed in qualitative and quantitative analysis.Therefore,it is necessary to develop a quantitative method.In this paper,we try to obtain a quantitative method for SRXRF through analyzing the SRXRF data of standard substance （soil,sediment,rock） and combining with quantitative data of standard samples.It is necessary to develop different normalization methods for different elements because of the diversities of mass numbers and intensities of spectral lines.ROI0-195 normalization is best for K and Ca elements,while ROI293-303 and ROI214-222 is useful for Ti and Mn,respectively.Considering Fe with higher mass number,the normalization of Compton scattering is best.Based on the above methods,the correlations for K,Ca,Ti,Mn and Fe are 0.89,0.98,O.96,0.93,and 0.98,respectively.Standard sample has been used to verify the quantitative standard curves of all elements.The results show that average RSD of prediction is below 10％ for Fe,below 20％ for K,Ca and Ti,while the Mn RSD is larger with a value of 35.9％.Generally,the method in this study can be used for quantitatively analyzing elements in lacustrine sediment.In order to test whether the result can capture seasonal elemental variation or not,SRXRF is used to measure the varve sediment from Lake Sihailongwan.Quantitative element contents are calculated by the standard curves.The result shows the variation of K,Ca,Ti,Mn and Fe are rhythmic following with varve structure.Element Ti and Ca rich in the light clastic layers,while Mn and Fe rich in the dark layers consisted of organic matters and clastics.Our preliminary result suggests that the SRXRF is valuable for paleoclimatic and paleoenvironmental research at seasonal to yearly resolution.