Trace metal analysis of sediment cores using a novel X-ray fluorescence core scanning method

2018 
Abstract The trace-metal composition of sediments provides important information on (past) environmental conditions, such as bottom water oxygenation, marine productivity, sediment provenance, and pollution. Whereas major and minor elements are often routinely analyzed using X-Ray Fluorescence (XRF) core scanning, analysis of trace metals with the same method is not yet established as a routine procedure. Here, we used a recently developed state-of-the-art XRF detector with a core scanner (Avaatech) to examine the optimal settings for analyzing a suite of trace metals (V, Cr, Ni, Cu, Zn, Mo, and U). Settings were optimized for fast analyses of sediment cores by extensive testing of primary energy settings, filters, and exposure times on two eastern Mediterranean Sea sapropel layers that archive episodes of past sea-floor anoxia. We reveal the following most advantageous (i.e., optimized for analytical accuracy and time efficiency) settings: (1) V, Cr, Ni at 20 kV with aluminum primary beam filter, (2) Cu, Zn, and U at 30 kV with ‘thick’ (125 μm) palladium primary beam filter, and (3) Mo at 50 kV with copper primary beam filter. For these trace elements, generally, ≥30 s of measurement are required for obtaining reliable data. Synthetic mixtures show that matrix effects, which are inherent to XRF analyses, are of particular importance for V. A correction for these matrix effects on V (e.g., using Compton scattering) may be necessary for samples with a large variability in carbonate content. XRF core-scanning measurements on synthetically, laminated sediments show that trace metals with contrasting atomic weights and related XRF penetration depths (V and Mo) can be determined at sub-mm resolution. We show that intensity results from the new XRF detector can be converted into concentrations using multivariate log-ratio calibration, allowing a fast quantitative prediction of sedimentary trace metal content using XRF core scanning.
    • Correction
    • Source
    • Cite
    • Save
    37
    References
    12
    Citations
    NaN
    KQI
    []
    Baidu
    map