Making low concentration in-house pressed pellet trace element standards for carbonate rock analyses by WD-XRF

Abstract

We describe a simple method for making in-house, low concentration (0–100ppm) pressed pellet standards for analyzing Ni, Cu, Mo, V and U in carbonate lithologies by wavelength dispersive X-ray fluorescence (WD-XRF) spectrometry. This method can easily be modified to include other elements of interest and to encompass a different range of concentrations. The purpose here was to circumvent the general lack of adequate geological certified reference materials (CRMs) having the necessary concentrations of the above five elements that are commonly used in paleoceanographic studies as proxies for changes in bioproductivity and paleoredox conditions in the water column and near the sediment–water interface. Standards were prepared by spiking 99.999% pure calcium carbonate powder with single element, inductively coupled plasma mass spectrometry (ICP-MS) standard stock solutions to yield final multi-element mixtures containing 0, 1, 2, 5, 10, 20, 50 and 100ppm of Ni, Cu, Mo, V and U, independently verified using acid digestion ICP-MS and PROTrace XRF trace element analyses. The level of in-house standard homogeneity was determined statistically using ANOVA tests (analyses of variance) on count rate data generated for each analyte from three surface transects (50×500μm) ablated on each standard using laser-ablation ICP-MS, demonstrating that all are statistically distinguishable and sufficiently homogenous for the intended purposes of larger spatial scale XRF measurements. The in-house standards were then used to refine preexisting XRF calibrations based on a broad suite of non-carbonate geological CRMs. The refined calibrations yield close agreement with ICP-MS and PROTrace XRF-derived values for the suite of in-house standards when analyzed repeatedly as ‘unknown samples’ by the XRF analytical protocol. The precision of sample preparation and analysis was assessed statistically by ANOVA tests on the results of repeated analyses of analyte concentrations for replicate XRF samples prepared from a representative carbonate sample, yielding reproducibility on the order of 1) <2.5% at the 50–100ppm level, 2) within ±5–10% or better at the 10–20ppm level, 3) within ±5–15% at the 5ppm level, and 4) ±10–55% at the 1–2ppm level which approaches analytical detection limits. Analytical detection limits and residual calibration backgrounds of the XRF protocol were constrained empirically by repeated analyses of ultrapure calcite blanks.