Selenium Concentrations in Twenty‐Six Geological Reference Materials: New Determinations and Proposed Values
Selenium Concentrations in Twenty‐Six Geological Reference Materials: New Determinations and Proposed Values
复制标题
DOI：
10.1111/j.1751-908x.2009.00003.x
复制
发表时间：
2009-06
影响因子：
3.8
通讯作者：
D. Savard;L. Bédard;S. Barnes
中科院分区：
文献类型：
--
作者：
D. Savard;L. Bédard;S. Barnes
人们对整个岩石中硒含量的兴趣越来越大，部分原因是它是贱金属和贵金属勘探的有用工具。硒在全岩地球化学中常常被忽略，因为大多数实验室不能对这种元素进行可靠的测定。这些困难的后果是国际地质参考物质中硒的指定值或认证值很少，因此无法遵循Kane和Potts（2007年）提出的获得此类参考物质可靠值的“最佳做法”。为了解决这个问题，我们通过巯基棉纤维预浓缩，然后用INAA（Se/TCF-INAA技术）在26种国际地质参考材料和一种质量控制材料（KPT-1）中测定了Se。这些值与一组公布的值一起用于估计这27个参考样品的Se浓度。为7个RM开发了稳健的统计数据，标准差等于或小于使用Horwitz函数计算的精密度，因此可以提出共识值。对于三个RM，离群值的存在给出了不太稳健的结果，并提出了建议值。对于其中17个RM，仅提供了信息值，因为测定结果不充分或因为推导出的数据标准偏差较大。
The interest in selenium concentrations in whole rocks is growing, in part because it is a useful tool for base and precious metal exploration. Selenium is often neglected in whole rock geochemistry because of the inability of most laboratories to make reliable determinations of this element. A consequence of these difficulties is a paucity of assigned or certified values for Se in international geological reference materials, so that the “best practice” proposed by Kane and Potts (2007) to obtain robust values for such reference materials cannot be followed. In order to address this problem, we have determined Se by pre‐concentration on thiol‐cotton fibre followed by INAA (Se/TCF‐INAA technique) in twenty‐six international geological reference materials, and one quality control material (KPT‐1). These values were used, in conjunction with a set of published values, to estimate Se concentrations for these twenty‐seven reference samples. Robust statistics were developed for seven of the RMs, with standard deviations equal to or less than precisions calculated using the Horwitz function and so that consensus values could be proposed. For three of the RMs, the presence of outliers gave less robust results, and suggested values are proposed. For seventeen of the RMs, only information values are provided, because either insufficient determinations were available or because large standard deviations of the data were derived.