Benchtop μXRF as a tool for speleothem trace elemental analysis : Validation, limitations and application on an Eemian to early Weichselian (125–97 ka) stalagmite from Belgium

Abstract

Variations of trace element (e.g. Mg, Sr, Ba, Fe, Zn etc.) concentrations along a speleothem's growth axis constitute important paleoclimate proxies. The use of laboratory micro X-ray fluorescence spectrometry as a fast and cheap alternative for conventional mass spectrometry techniques for trace element analysis on speleothems has been explored in the past and yielded satisfactory results. However, within the speleothem community there is need for an in-depth investigation of the full potential of this analytical technique. Compared to other types of paleoclimate archives, benchtop (μ)XRF analysis on speleothems is analytically more challenging because of the high-crystalline speleothem matrix and the low abundance of the elements of interest. In this study, several speleothem samples with differences in mineralogy (calcite versus aragonite) and composition are investigated. Various instrumental parameters are tested and recommendations are made for future studies applying (μ)XRF analysis to speleothems. Quantification based on a multiple standard calibration and an assessment of the error is carried out. Through validation with mass spectrometry techniques, it is confirmed that benchtop μXRF devises are able to generate speleothem trace element records. Successful results were obtained for Sr, Mg and Fe, while Zn and Ba were quantified in samples characterized by high concentrations. Nevertheless, caution has to be taken when interpreting the results, due to the presence of diffraction caused by the crystallinity of the samples. The elements which provide reliable results are sample specific and depend on the type of matrix and elemental abundance. These findings are applied on an Eemian to early Weichselian stalagmite from the Han-sur-Lesse Cave, Belgium. Time series were constructed for Mg and Sr, creating a multiproxy dataset together with previously obtained stable isotope (δ13C and δ18O) ratios, growth-rate and stalagmite morphology. It appears that Mg and Sr are not primarily controlled by prior calcite precipitation, but rather by changes in vegetation activity above the cave.