ABSTRACT Sample preparation methods for energy‐dispersive x‐ray fluorescence (EDXRF) analysis of archaeological and geochemical silicic samples (e.g., pottery, rocks, and soil) were evaluated. This study focused on loose powder and pressed powder pellet specimens, which are relatively easy to prepare, rather than glass bead specimens that require more complex preparation. Six geochemical reference materials (RMs) were prepared as ten loose powder specimens and ten pressed powder pellets each, and 22 components (Na 2 O, MgO, Al 2 O 3 , SiO 2 , P 2 O 5 , K 2 O, CaO, TiO 2 , MnO, Fe 2 O 3 , V, Cr, Ni, Cu, Zn, Rb, Sr, Y, Zr, Nb, Ba, and Pb) were quantified using EDXRF. Based on the analytical values and relative standard deviations (RSDs), the accuracy and precision of the analytical results obtained using loose powder and pressed powder pellet preparation were compared. When the analytical values were compared with the recommended values for each geochemical RM, the degree of agreement was generally similar for both preparation methods. By contrast, comparison of the RSDs showed that the two methods were generally comparable for major oxides, whereas for minor elements, preparing powder pellets slightly reduced the RSDs. Nevertheless, even with pressed powder pellets, several analytical values exhibited relatively large RSDs. Therefore, for quantitative EDXRF analysis of the silicic samples, preparing multiple batches of loose powder (or performing multiple measurements with reloaded loose powder), which is easier and more convenient, may yield more reliable results than preparing a small number (e.g., a single) pressed powder pellets.
Ichikawa et al. (Mon,) studied this question.