The mineral-scale chemistry that can shift the clock on the early Solar System

Calcium-aluminium-rich inclusions, or CAIs, are refractory components found in some meteorites. They formed very early in the Solar System, and uranium–lead dating of these objects helps scientists estimate when its first solids appeared. The method depends in part on uranium isotopes and the lead produced as uranium decays. If different minerals in one inclusion have different uranium isotope ratios, combining measurements from different mineral reservoirs could complicate an age calculation.

A team led by Merel E. Onderwater mapped major and trace elements in six coarse-grained CAIs from two CV3 chondrites. The inclusions had experienced melting and crystallization, processes that can sort elements among minerals. The researchers used femtosecond laser-ablation inductively coupled plasma mass spectrometry (fs-LA-ICP-MS): a laser sampled narrow lines across polished meteorite sections, and a mass spectrometer measured the elements released. The resulting maps show where elements occur, rather than only giving an average for each inclusion.

Uranium gathers at pyroxene grain edges

In the studied inclusions without perovskite or hibonite, aluminium-titanium pyroxene was generally the main host for uranium, thorium and many rare-earth elements. Uranium and thorium were especially concentrated toward pyroxene grain edges. In one example, uranium in pyroxene rose from about 0.2–0.3 parts per million in grain cores to 1–3 parts per million near the edges. Melilite, another abundant mineral, hosted more of the light rare-earth elements, including neodymium.

The authors interpret the pyroxene pattern as a record of fractional crystallization. As minerals crystallized from a melt, elements that did not readily enter early-forming crystals became concentrated in the remaining melt. Later-forming pyroxene, particularly its edges, could then acquire more uranium and other trace elements. The maps also show differences between inclusion interiors and outer regions. The outer regions and rims were relatively enriched in some refractory elements but depleted in more volatile rare-earth elements—including ytterbium, europium and cerium—and uranium. The researchers interpret this pattern as evidence that high-temperature melt evaporation modified the inclusions.

A modeled effect on uranium–lead ages

The study also examined curium-247, a short-lived isotope that decayed into uranium-235 early in Solar System history. Because curium no longer exists in these samples, the team could not map it directly. Instead, they used measured neodymium and thorium distributions as proxies for curium and modeled how differing curium-to-uranium proportions among minerals could affect their uranium isotope compositions.

In the model, melilite often developed the largest uranium-235 excesses, while anorthite could show deficits relative to the inclusion as a whole. These are predictions based on proxy elements and assumptions about curium’s behavior, not direct measurements of uranium isotope ratios in each mineral. The authors then tested hypothetical age calculations in which uranium isotope data and radiogenic lead came from different mineral reservoirs.

Those mismatches could yield apparent age offsets, reaching around 0.1 to 1.0 million years in some modeled combinations. But the authors distinguish these combinations from the scenario they consider most consistent with common dating practice: bulk uranium isotope measurements paired with radiogenic lead mainly associated with pyroxene. In that case, most modeled offsets were small, from zero to 0.03 million years, while two inclusions produced offsets of 0.06 and 0.14 million years. The latter is comparable to the cited analytical uncertainty of 0.16 million years for the canonical CAI formation age. The study therefore does not show that existing ages are generally wrong; it identifies conditions under which mineral-scale differences may matter.

What remains uncertain

Six inclusions provide detailed maps, but they cannot represent every kind of CAI. The laser’s 20-micrometre spot size suits coarse grains but cannot cleanly resolve very small minerals, including fine perovskite grains. The curium-related results are model estimates, and the inferred isotope differences depend on whether neodymium or thorium is the better proxy. Those proxies do not give identical predictions: neodymium generally yielded larger variations.

The paper text states that the manuscript was accepted for publication in Geochimica et Cosmochimica Acta. This article is based on the supplied paper text, not an independent evaluation of the research or its review process. The study’s central contribution is a more spatially detailed picture of uranium’s mineral hosts and zoning. The authors suggest that, where enough material is available, measuring uranium and lead in the same mineral—particularly pyroxene—or accounting for internal variation could make some high-precision ages more robust.