Insights into U-Th-Pb mobility during diagenesis from laser ablation U-Pb dating of apatite fossils

Abstract

We studied the feasibility of using laser ablation inductively coupled plasma mass spectrometry (LA–ICPMS) for U-Pb dating of vertebrate and invertebrate apatite fossils. Two specimens of Carcharodon Carcharias from Pleistocene (2.58–0.01 Ma) sediments in Baja California peninsula, Mexico, give Gelasian ages of 2.0 ± 0.4 Ma and 2.0 ± 0.6 Ma which are consistent with the palaeontologic record reported for this species suggesting that the U-Pb system in the shark teeth remained closed until the present. Dentine data from an Oligocene (33.9–23.0 Ma) shark tooth, Otodus angustidens, from El Cien Formation, Mexico, give an anomalously young age of 19 ± 0.2 Ma with a low 207Pb/206Pb value of 0.65. This age is largely controlled by a sub-population of anomalously radiogenic data with low concentrations of U and Pb suggesting remobilization of Pb from high to low U domains. Regressing relatively non-radiogenic data from high U and Pb domains using a fixed average crustal value for initial common 207Pb/206Pb gives an age of 28.6 ± 0.2 Ma, which agrees with U-Pb zircon ages from tuff horizons in the sequence. Enamel from the same shark tooth is much lower in U and Pb and shows evidence for substantial mixing with radiogenic Pb from the dentine. Apatite shells from a brachiopod specimen, Lingula cobourgensis, and conodont fossils contain high U concentrations and significant proportions of radiogenic Pb. Some conodont specimens show high Th/U ratios combined with evidence of significant excess 206Pb. The amount of excess 206Pb requires a minimum Th partition coefficient of about 16,000 and likely much higher suggesting enrichment by ion exchange rather than recrystallization. In some cases conodonts with low Th/U can be precisely dated. A specimen of I. simulator gives an age of 302 ± 4 Ma which is consistent with its well constrained Gzhelian range (303.7–298.9 Ma). We conclude that processing LA–ICPMS data at the single cycle level allows more precise absolute dating of fossils. In some cases, separate regression of totally reset and partially reset domains can resolve ages of early and late diagenesis.