Reworking of Earth's first crust: Constraints from Hf isotopes in Archean zircons from Mt. Narryer, Australia

Abstract

Discoveries of >4. Ga old zircon grains in the northwest Yilgarn of Western Australia led to the conclusion that evolved crust formed on the Earth within the first few 100. Ma after accretion. Little is known, however, about the fate of the first crust that shaped early Earth's surface. Here we report combined solution and laser-ablation Lu-Hf-U-Pb isotope analyses of early Archean and Hadean detrital zircon grains from different rocks of the Narryer Gneiss Complex (NGC), Yilgarn Craton, Western Australia. The zircons show two distinct groups with separate evolutionary trends in their Hf isotopes. The majority of the zircon grains point to separation from a depleted mantle reservoir at ~3.8-3.9. Ga. The second Hf isotope trend implies reworking of older Hadean zircon grains. The major trend starting at 3.8-3.9. Ga defined by the Hf isotopes corresponds to a Lu/Hf that is characteristic for felsic crust and consequently, the primary sources for these zircons presumably had a chemical composition characteristic of continental crust. Reworked Hadean crust appears to have evolved with a similar low Lu/Hf, such that the early crust was probably evolved with respect to Lu-Hf distributions. The co-variation of Hf isotopes vs. age in zircon grains from Mt. Narryer and Jack Hills zircon grains implies a similar crustal source for both sediments in a single, major crustal domain. Age spectra and associated Hf isotopes in the zircon grains strongly argue for ongoing magmatic reworking over hundreds of millions of years of the felsic crustal domain in which the zircon grains formed. Late-stage metamorphic zircon grains from the Meeberrie Gneiss unit yield a mean U-Pb age of 3294.5 ± 3.2. Ma with initial Hf isotopes that correspond to the evolutionary trend defined by older NGC zircon grains and overlap with other detrital zircon grains, proving their genetic relationship. This 'Meeberrie event' is interpret here as the last reworking event in the precursor domain before final deposition. The continuous magmatic activity in one crustal domain during the Archean is recorded by the U-Pb ages and Hf isotope systematics of zircon grains and implies reworking of existing crust. We suspect that the most likely driving force for such reworking of crustal material is ongoing crustal collision and subduction. A comparison of Hf isotope signatures of zircon grains from other Archean terranes shows that similar trends are recognised within all sampled Archean domains. This implies either a global trend in crustal growth and reworking, or a genetic connection of Archean terranes in close paleo-proximity to each other. Notably, the Archean Acasta gneiss (Canada) shows a similar reworking patterns to the Yilgarn Craton of Hadean samples implying either a common Hadean source or amalgamation at the Hadean-Archean transition.