First Principles Estimates of 238U/235U Fractionation in Solutions and Crystals

Abstract:

²³⁸U/²³⁵U variation in nature is of interest to geochemists as a paleoredox proxy, as U(IV) species are typically ~1‰ heavier than U(VI). This fractionation is mainly driven by the nuclear field shift effect. Previous theoretical calculations of equilibrium U-isotope fractionation [1] have focused on simple gas-phase species. This study aims to model a wider array of U-bearing species, focusing on those relevant to natural systems. The results are compared to experiments. We use the calibrated DFT-PAW method [2] to make more complex species tractable.

Our results agree with experimental determinations of the U(IV)-U(VI) fractionation in solution at ambient temperature [3] (1.4‰ calculated vs. 1.6‰ measured). We predict measurable speciation effects of 0.2-0.3 ‰ among the UO₂-CO₃^2- species predominant in oxic seawater. Common organic ligands also produce measurable fractionations. Between UO₂(catecholate), a functional group common in siderophores that also complex U [4], and UO₂(CO3)₃^4-, the calculated fractionation is ~0.3‰. Calculations for UO₂PO₄^-, an analog for phosphodiester bonds in extracellular polymeric substances that sorb U at low pH [5], result in a fractionation of ~0.4‰.

Previous theoretical work has focused on gas-phase analogues of aqueous species. We tested the importance of solvation to U-isotope fractionations, and found that both explicit and implicit solvation methods can produce changes in the coordination and geometry of the ligands surrounding the central U atom, relative to in vacuo models. These changes impact calculated field shift effects. As an example, for the species UO₂(OOCCOO)₂ a solvated model differs by ~0.3‰ from the unsolvated molecule. We do not find a useful correlation between uranyl U=O bond lengths in U(VI) species and the field shift effect. We also estimate nuclear volume fractionations between U(IV) in zircons and U(VI), U(V), and U(IV) in silicate melt analogues. The zircon-melt fractionation is predicted to be as large as 0.7‰ and 0.5‰ for U(VI)- and U(V)-dominated melts, respectively, at ~700ºC. Zircon fractionation relative to U(IV) in melts is not resolvable.

[1] Abe et al. (2008) J Chem Phys 129:164309 [2] Schauble (2013) PNAS 110:17714 [3] Wang et al. (2015) GCA 158:262–275 [4] Moll et al., (2009) Microbiol 25:157-166 [5] Kelly et al. (2002) GCA 66:3855-3871