²³⁸U/²³⁵U isotope ratios of crustal material, rivers and products of hydrothermal alteration: new insights on the oceanic U isotope mass balance

<p>In this study, the U isotope composition, <i>n</i>(²³⁸U)/<i>n</i>(²³⁵U), of major components of the upper continental crust, including granitic rocks of different age and post-Archaean shales, as well as that of rivers (the major U source to the oceans) was investigated. Furthermore, U isotope fractionation during the removal of U at mid-ocean ridges, an important sink for U from the oceans, was investigated by the analyses of hydrothermal water samples (including low- and high-temperature fluids), low-temperature altered basalts and calcium carbonate veins. All analysed rock samples from the continental crust fall into a limited range of δ²³⁸U between −0.45 and −0.21 ‰ (relative to NBL CRM 112-A), with an average of −0.30 ± 0.15 ‰ (2 SD, <i>N</i> = 11). Despite differences in catchment lithologies, all major rivers define a relatively narrow range between −0.31 and −0.13 ‰, with a weighted mean isotope composition of −0.27 ‰, which is indistinguishable from the estimate for the upper continental crust (−0.30 ‰). Only some tributary rivers from the Swiss Alps display a slightly larger range in δ²³⁸U (−0.29 to +0.01 ‰) and lower U concentrations (0.87–3.08 nmol/kg) compared to the investigated major rivers (5.19–11.69 nmol/kg). These findings indicate that only minor net U isotope fractionation occurs during weathering and transport of material from the continental crust to the oceans. Altered basalts display moderately enriched U concentrations (by a factor of 3–18) compared to those typically observed for normal mid-ocean ridge basalts. These, and carbonate veins within altered basalts, show large U isotope fractionation towards both heavy and light U isotope compositions (ranging from −0.63 to +0.27 ‰). Hydrothermal water samples display low U concentrations (0.3–1 nmol/kg) and only limited variations in their U isotope composition (−0.43 ± 0.25 ‰) around the seawater value. Nevertheless, two of the investigated fluids display significantly lower δ²³⁸U (−0.55 and −0.59 ‰) than seawater (−0.38 ‰). These findings, together with the heavier U isotope composition observed for some altered basalts and carbonate veins support a model, in which redox processes mostly drive U isotope fractionation. This may result in a slightly heavier U isotope composition of U that is removed from seawater during hydrothermal seafloor alteration compared to that of seawater. Using the estimated isotope compositions of rivers and all U sinks from the ocean (of this study and the literature) for modelling of the isotopic U mass balance, this gives reasonable results for recent estimates of the oceanic U budget. It furthermore provides additional constraints on the relative size of the diverse U sinks and respective net isotope fractionation during U removal.