In six melted CAIs, uranium sits in pyroxene edges, and mixing minerals can fake a million-year age gap

A preprint on arXiv by Merel E. Onderwater and colleagues maps major and trace elements in six coarse-grained calcium-aluminium-rich inclusions from CV3 chondrites with femtosecond laser ablation ICP-MS. These inclusions are the oldest dated Solar System solids, and their uranium-lead ages anchor the disk timeline. The maps are not a spacecraft-materials study. They ask which minerals hold uranium after melting, because a bulk age assumes uranium and lead came from the same reservoir.

In inclusions lacking perovskite or hibonite, Al,Ti-pyroxene is the dominant host of uranium, thorium, and most rare earth elements, with strong enrichments toward pyroxene grain edges. Light rare earths sit mainly in melilite. The authors read that zoning as igneous fractional crystallization, then melt evaporation that depleted ytterbium, europium, cerium, and uranium in outer regions and rims. Using neodymium and thorium maps as proxies for extinct curium-247, they model mineral-scale uranium-isotope shifts. That is a proxy model, not a direct curium measurement.

The chronology stakes are real but narrow. Brennecka and colleagues reported in Science in 2010 that meteoritic uranium isotope ratios vary, consistent with live curium-247 decaying to uranium-235. Connelly and coauthors later placed CAI formation near 4567 million years with high-precision lead-lead dating. This preprint predicts the largest uranium-235 excesses in melilite and apparent offsets of about 0.1 to 1.0 million years if isotopes and radiogenic lead come from different minerals. Bulk uranium compositions generally keep that bias under 0.15 million years. A satellite-polymer reading confuses disk chronology with spacecraft materials.

What would strengthen this is mineral-by-mineral uranium isotope measurements on the same inclusions, paired with lead from those phases, not another elemental map alone. Six CV3 objects cannot speak for all CAI types, and inclusions with perovskite or hibonite were set aside as a different host regime. Femtosecond ablation limits thermal fractionation relative to nanosecond lasers, but proxying curium with neodymium and thorium imports partitioning assumptions the maps do not test. Until those isotopes are measured, a 1-million-year mineral offset remains a model bound, not a revised age of the Solar System.