Previous   |    Next

doi:10.2204/iodp.proc.345.102.2014

Handheld X-ray fluorescence analyses

During Expedition 345, we performed semiquantitative X-ray fluorescence (XRF) chemical analyses of rock surfaces using a ThermoFisher Niton XL3t GOLDD+ handheld XRF analyzer to help identify minerals in selected samples from the working and archive halves. Chemical analysis by handheld XRF is a nondestructive technique that was used on igneous rocks during Expeditions 330 and 335 (Expedition 330 Scientists, 2011; Expedition 335 Scientists, 2012). Analytical procedures were adapted from those developed during Expedition 335 (see the evaluation of this device in XRF in Expedition 335 “Supplementary material”).

Analytical methods

The handheld XRF analyzer can measure elemental concentrations of as many as 33 elements simultaneously and is equipped with three excitation filters (high range, main range, and light range) for optimizing sensitivity. Calibrations and analytical procedures are preset by the manufacturer and are not adjustable. Analyses were carried out using the Cu/Zn procedure in mining mode. This setting was selected, as it is the recommended method for the elements of interest for petrological characterization. The analyzed elements and their corresponding filter positions are in Table T5. Counting time was 150 s for each analysis. An automatically adjusted helium flow was used to avoid interference with the ambient air for the light elements (Mg, Na, and K). For consistency, an 8 mm spot size was used for all measurements on samples and standards except for macroscopically heterogeneous samples, with which a 3 mm spot size was used.

Calibration and standardization

To test the accuracy of the analyses, pressed powders of four certified silicate standards, BHVO-2 (US Geological Survey [USGS] Geochemical Reference Standard-Basalt, Hawaiian Volcano Observatory), JB-2 (basalt, Geological Survey of Japan [GSJ]), JA-2 (andesite, GSJ) and BCR-2 (USGS Geochemical Reference Standard-Basalt, Columbia River), were analyzed (Table T6). The measured concentrations differed from published values by 20% or more, except for MnO, CaO, and Cu. The concentration of MgO and trace elements Ba, Cr, Co, and Ni are below detection limits except for BHVO-2.

A tentative effort was made to improve the accuracy of the analyses on major oxides by using a rock standard–based approach similar to that developed during Expedition 335. Calibration curves were defined by plotting the raw concentrations measured on three certified silicate rock standards, BHVO-2, JB-2, and JA-2, with an 8 mm spot size, as a function of their reference concentrations published in Govindaraju (1994) and downloaded from the GeoReM database (Table T7); BCR-2 was used as quality control. This approach allowed for improving the accuracy of the measured values for most elements but not enough to allow a quantitative estimate of the major element composition of silicate rocks. The accuracy results obtained for the quality control standard BCR-2 were better than 10% for SiO₂, TiO₂, FeO^(T), MnO, CaO, Zr, and Sr (Table T7).

Our study indicates that this instrument is not adequate for making quantitative measurements of the oxide contents or the trace element concentrations of igneous rocks. The data produced with this instrument are semiquantitative at best and therefore inappropriate for scientific reports. The handheld XRF analyzer is useful only for rapid qualitative assessment of the compositions of specific geological materials (e.g., identification of Mn crusts or oxides) before more accurate analyses of the samples can be carried out using either shipboard XRD and/or ICP-AES or shore-based analyses. However, the procedure for using and handling the instrument is tedious and time consuming and may not be worth the effort and time of the shipboard scientists.

Top of page   |    Previous   |    Next