The age of the earliest continental crust and onset of plate tectonics

Résumé: I will present new bulk Hf and Pb isotope data for 63 single zircons from Jack Hills (JH), Western Australia, which were obtained using solution chemistry and, respectively, MC-ICP-MS and ICP-MS, as opposed to in situ techniques. With the exception of one young zircon at 3.32 Ga, the remainder of the selected grains were previously dated at >3.9 Ga by ion-microprobe. This work extends and complements the solution chemistry data of Harrison et al. (2005) but uses bulk rather than in situ Pb-Pb ages to interpret the Hf isotope data. This larger data set is used to explore whether the host rocks of the JH zircons formed as a succession of pulses or rather as a single event, and to calculate the age and assess the nature of their crustal protolith. We find that the parent granites of the JH zircons analyzed here formed during a single pulse 4.1 ± 0.1 Gy ago by the remelting of a 4.30-4.36 Ga old protolith. Monte Carlo modeling of the inferred low Lu/Hf ratio distribution of this material (176Lu/177Hf <0.01) does not match either modern-type oceanic crust, island arc rocks, or subduction-zone type granites but rather fits early "KREEPy" crust produced by melting of the last products of the terrestrial magma ocean. We speculate that this primordial crust may have been eliminated by impacts or buried in the deep mantle, thus potentially accounting for the so-called hidden reservoir required most notably by 176Hf-143Nd and 142Nd systematics. -------------------------------------------------------------------------------- Séminaires généraux de l'Institut de Physique du Globe de Paris --------------------------------------------------------------------------------