What Hf isotopes in zircon tell us about crust-mantle evolution

The 176Lu-176Hf radioactive decay system has been widely used to study planetary crust-mantle differentiation. Of considerable utility in this regard is zircon, a resistant mineral that can be precisely dated by the U-Pb chronometer and record its initial Hf isotope composition due to having low Lu/Hf. Here we review zircon U-Pb age and Hf isotopic data mainly obtained over the last two decades and discuss their contributions to our current understanding of crust-mantle evolution, with emphasis on the Lu-Hf isotope composition of the bulk silicate Earth (BSE), early differentiation of the silicate Earth, and the evolution of the continental crust over geologic history. Meteorite zircon encapsulates the most primitive Hf isotope composition of our solar system, which was used to identify chondritic meteorites best representative of the BSE (176Hf/177Hf = 0.282793 ± 0.000011; 176Lu/177Hf = 0.0338 ± 0.0001). Hadean-Eoarchean detrital zircons yield highly unradiogenic Hf isotope compositions relative to the BSE, providing evidence for the development of a geochemically enriched silicate reservoir as early as 4.5 Ga. By combining the Hf and O isotope systematics, we propose that the early enriched silicate reservoir has resided at depth within the Earth rather than near the surface and may represent a fractionated residuum of a magma ocean underlying the proto-crust, like urKREEP beneath the anorthositic crust on the Moon. Detrital zircons from world major rivers potentially provide the most robust Hf isotope record of the preserved granitoid crust on a continental scale, whereas mafic rocks with various emplacement ages offer an opportunity to trace the Hf isotope evolution of juvenile continental crust (from εHf[4.5 Ga] = 0 to εHf[present] = + 13). The river zircon data as compared to the juvenile crust composition highlight that the supercontinent cycle has controlled the evolution of the continental crust by regulating the rates of crustal generation and intra-crustal reworking processes and the preservation potential of granitoid crust. We use the data to explore the timing of generation of the preserved continental crust. Taking into account the crustal residence times of continental crust recycled back into the mantle, we further propose a model of net continental growth that stable continental crust was firstly established in the Paleo- and Mesoarchean and significantly grew in the Paleoproterozoic.