A geochemical application of the ITRAX scanner to a sediment core containing eastern Mediterranean sapropel units

The ITRAX micro-X-ray fluoresence (XRF) core scanner has been applied in a sediment geochemistry investigation. The core sections selected contain examples of the organic-rich sedimentary units (sapropels) that form periodically in the eastern Mediterranean basin. Sapropels are visually obvious from their dark coloration, but the ITRAX X-radiograph also reveals physical property changes that result mainly from the high pore-water content of sapropels. A consideration of wavelength-dispersive XRF data from discrete samples of the most recent sapropel (S1) was made along with the set of elements reported by the ITRAX instrument\textquoterights energy-dispersive XRF system over core sections containing S1. This allowed selection of a suite of eight inter-element ratios or element integrals through which characteristic features of sapropel development and geochemistry were revealed. While recognizing that the measured XRF element integrals from the ITRAX do not have an exact constant relationship with element concentration over changing sediment types, this combination of ratios provides significant information for geochemical interpretation. These include evidence for: (i) the presence of high C_org contents in the visual sapropel from Ba/Ti and Br/Cl ratios; (ii) a thinning of the original sapropel thickness by post-depositional oxidation from Mn/Ti and Cu/Ti ratios; (iii) pyrite authigenesis in the residual visual sapropel from Fe/Ti and S/Cl ratios and the As integral; and (iv) aragonite formation in and around the sapropel from the Sr/Ca ratio. These same ratios were then used to interpret ITRAX data from a deeper section of the same core containing the older sapropel S3, where the same characteristics, including the relict post-depositional oxidative thinning of the original unit, could be identified with only minor differences of detail. Directions of supply of Fe, As and Cu into the sapropels could be inferred from profile shapes.