Vezinet, A. ; Pearson, D.G. ; Thomassot, E.
Lithos, 2021, 106040
Voir en ligne : https://doi.org/10.1016/j.lithos.2021.106040
Abstract :
Radiometric decay systems have played a crucial role in developing our understanding of the evolution of the early Earth. There are two main types of protocols for isotope measurements in geological materials : (i) bulk dissolution of rocks, or whole-grains and (ii) spatially resolved techniques (laser-ablation or ion-beam). These two approaches have sometimes led to results that are not easily reconciled for early Earth crustal rocks (≥ 3.6 Ga). While initial radiogenic isotope signatures (e.g. initial 176Hf/177Hf or initial 143Nd/144Nd) obtained from whole-rock protocols are significantly above chondritic values, indicative of extensive chemical differentiation of the mantle before 3.6 Ga, data from spatially resolved analysis of individual mineral growth domains point toward much less dramatic differentiation. This is indicated by the majority of data falling close to models of Earth’s mantle that had not experienced major silicate melt removal into the crust. These data show chondritic or sub-chondritic signatures. Interpretations of whole rock isochrons are built on assumptions about the history and relationship of a number of different samples to each other. At the heart of these assumptions, the effects of secondary process–such as metasomatism–on isotopic compositions and consequently on the age and initial ratio of isochrons, are often considered negligible. In order to evaluate the possible effects of metasomatism and metamorphism on co-genetic igneous suites we modelled the impact of contamination by an external component on both the isochron slope (the apparent age) and the isochron intercept (the initial radiogenic isotope signature). A significant outcome is that the age significance of some of the modified isochron arrays remains to a large extent within uncertainty of the original crystallisation age of the igneous suite. In other words, the original age signature is preserved, but with lower precision. The intercept of the isochron, from which the initial isotope ratio is calculated, however is often significantly modified, which has consequences for the interpretation of these signatures. Our results provide an explanation for the discrepancy between whole-rock and spatially-resolved results observed in early Earth material. Lastly, our results, applied to studies of ancient crustal rocks, are interpreted as indicative of no significantly depleted mantle domains before 3.6 Ga, and no Hfsingle bondNd isotopes decoupling at that time.