Correlated molybdenum and uranium isotope signatures in modern anoxic sediments: Implications for their use as paleo-redox proxy

Redox-sensitive trace metals and their isotopes have emerged as important tools that are used to reconstruct the redox-evolution of the ocean-atmosphere system. However, reliability of such reconstructions ultimately depends on a solid understanding of the proxies in the present-day oceanic system and their archival potential in sediments. This study compares isotope fractionation of molybdenum (Mo) and uranium (U) during their removal from seawater and deposition into sediments by investigating sites at various depths of the presently two largest restricted anoxic oceanic basins: The Black Sea and the Cariaco Basin. In support of previous investigations, our data indicate that Mo scavenging and isotope fractionation are mainly controlled by water column sulfide levels. In contrast to Mo, U reduction and immobilization appears to occur mainly at the sediment-water interface and within the uppermost few cm of the sediment pile in both basins. In the Black Sea, decreasing delta U-238 of surface sediments with increasing water depth correlate with trends for water column delta U-238, implying constant U isotope fractionation between water and sediment. However, increasing U concentrations and delta U-238 within the uppermost few cm of the sediment pile of both basins indicate additional U reduction with depth. Despite the different mechanisms for Mo and U removal and associated isotope fractionations, a similar inverse correlation between delta Mo-98 and delta U-238 is observed for sediments of both basins, which translates in a positive correlation of Mo and U isotope fractionation between the sediments and open seawater. The correlation of delta Mo-98 and delta U-238 indicates a similar response of isotope fractionation to the efficiency of Mo and U removal that is mainly controlled by sulfate reduction rates. High dissolved sulfide concentrations and sulfate reduction rates are responsible for very effective Mo and U removal and corresponding minor Mo and U isotope fractionation relative to seawater. Further, high dissolved sulfide concentrations also correlate positively with deep water renewal times, resulting in an isotopically fractionated water column with low delta U-238 (and somewhat higher delta Mo-98) in restricted basins with sluggish ventilations, such as the Black Sea. Both mechanisms result in negatively correlated delta Mo-98 and delta U-238 with high delta Mo-98 and low delta U-238 in sediments under strong euxinic conditions. The particularly strong correlation observed for Cariaco Basin sediments may indicate that its water column was variably stratified in the past. The observed delta Mo-98 and delta U-238 correlation of both basins can be reproduced in a simple coupled water column and sediment reactive transport model. Different slopes in delta Mo-98 and delta U-238 trends can be linked to varying degree of basin restriction, sulfate reduction rates, and isotope compositions of the respective water columns. The offset towards lower delta Mo-98 (and 6238-r u r), observed for Cariaco Basin sediments compared to those from the Black Sea, may be the result of inefficient Mo reduction with high Mo isotope fractionation or isotopically light Mo from a particulate Fe-Mn oxide shuttle. The results of this study will help to interpret sedimentary Mo and U isotope values, while showing that coupling of delta U-238 and delta U-238 in sedimentary archives may be useful for paleo-reconstruction work. (C) 2019 Published by Elsevier Ltd.