238U/235U measurement in single-zircon crystals: Implications for the Hadean environment, magmatic differentiation and geochronology

Abstract

Owing to the challengingly small amounts of uranium (U) they contain, the isotopic composition (²³⁸U/²³⁵U) of single zircon grains has never been measured. Leveraging methods we designed for analysis of small sample amounts and modern MC-ICPMS instruments, we show that precise (±0.04 to ±0.25‰) single-zircon ²³⁸U/²³⁵U measurements are now possible. We report data for 31 single grains from the Jack Hills conglomerate, and 3 reference zircon localities (FC-1, R33 and Temora). Consistent with the reducing conditions implied by the small Ce anomalies of many Hadean zircon, Jack Hills grains display only small δ²³⁸U variations (from -0.60 to -0.12‰). The distribution is centered on the average chondritic and bulk continental crust value, arguing against the widespread existence of Oklo-type reactors in the early Earth. The subtle δ²³⁸U variations in Jack Hills zircons are more plausibly explained by a small (∼0.10%) mass-dependent equilibrium isotope fractionation between at least one U-bearing accessory mineral and silicate melts, during magmatic differentiation under reducing conditions. In contrast, the large δ²³⁸U difference between pooled titanite and pooled zircon fractions from the Fish Canyon Tuff sample suggests larger isotope effects during igneous fractional crystallization under oxidizing conditions (∼QFM+2), with preferential removal of ²³⁵U from the melt and into zircon, and/or other accessory phases. We estimate that ∼50% of zircon dated by the CA-ID-TIMS method would be amenable to single-grain U isotope measurements, making this method widely applicable to future studies. This would enable (i) improvements in precision and accuracy of U-Pb and Pb-Pb dates, (ii) accurate investigation of U-series disequilibrium contribution to U-Pb discordance, and (iii) accurate re-evaluation of U decay constants.