Development of Efficient Trace Element Quantification Methods in Carbonate Rocks

This study evaluates the feasibility of using a faster handheld x-ray fluorescence (XRF) method with little to no sample preparation to facilitate the analysis of several important trace elements in carbonate rocks when dealing with large sample quantities. The use of handheld energy-dispersive x-ray fluorescence (ED-XRF) has become popular because it offers nondestructive quantifiable elemental analyses of samples. ED-XRF has been used by geologists in rock core analyses, and has been shown to be capable of producing robust quantifiable results when compared with wavelength-dispersive x-ray fluorescence (WD-XRF) for many elements in pelitic rocks. However, there has been little to no published studies on the use of XRF on carbonate rocks, which may be partly due to the scarcity of internationally accepted trace element reference standards. Trace element abundances of several marine carbonate and carbonatite rocks are presented which have been determined in this study by multiple XRF and inductively coupled plasma mass spectrometry (ICP-MS) methods. Four carbonate standards in development, provided by the USGS, have been characterized with the use of ICP-MS and WD-XRF. Analyses of all samples were performed using different methods by two independent labs at Utah State University and Washington State University. These standards offer the wide elemental ranges necessary to properly quantify geochemical data within the unique matrix of carbonate rocks. A comparison of pressed powder pellets was made between ED-XRF and WD-XRF on carbonate rock samples from the Ordovician Garden City Formation and Pogonip Group of northern and west-central Utah respectively, to determine differences in results between the two methods. Additionally, slabbed hand samples as well as loose powdered samples prepared from the same rock samples were also analyzed and compared with the pressed powder pellets with ED-XRF, to determine if sample preparation had significant effects on sample analyses.