Spectral X-ray computed tomography for the chemical identification of critical minerals

Abstract

Differentiating minerals using high-resolution X-ray tomography (µCT) relies on distinct differences in the attenuation coefficient µ. The µ value depends on an interplay between the material density ρ and the effective atomic number Z_eff of a mineral phase. Difficulties in identifying mineral phases arise when this interplay gives similar µ values and thus limited contrast within µCT images. Untangling these two dependencies is essential to improve the three-dimensional chemical identification of critical minerals. Lab-based methods and techniques often incorporate different measures, but only show a limited application potential on multiphase geological samples. Using high-Z spectral laboratory-based µCT we studied the potential of directly identifying chemical elements within the practical margins of high-Z spectral detectors. This paper compares the results from three mineral deposits using two spectral µCT setups. Chemical elements with a Z higher than molybdenum and a concentration of at least some weight percentage were correctly identified using K-edge imaging. The suitability of the different high-Z spectral detectors depends largely on the availability of prior knowledge of the sample composition. Quantifying elemental concentrations is element- and sample specific and currently does not allow for optimal automated mineralogy solutions. Improving the three-dimensional identification of minerals can be achieved with dedicated analyses of the energy-dependent µ curve and therefore will remain the focus of future work.