Geochemistry and Nd isotope systematics of Archaean granites of the Eastern Goldfields, Yilgarn Craton, Australia: implications for crustal growth processes

Abstract

The Archaean granites and granitic gneisses of the Eastern Goldfields in the Yilgarn Craton, Western Australia, can be divided into two major groups (high-Ca and low-Ca), and three minor groups (high-HFSE (high field strength elements), mafic, and syenitic). The high-Ca group (68–77% SiO₂) with high Al₂O₃, Na₂O and Sr, and low Y, shares many features with typical Archaean tonalite-trondhjemite suites, but has higher K₂O, Rb, and Th contents. The low-Ca group (70–76% SiO₂) differs from the high-Ca group in having lower Al₂O₃, CaO, and Na₂O, but higher K₂O, Rb, Th, Zr, Y, La and Ce contents. Granites of the high-HFSE, mafic and syenitic groups form a minor component (10–20%) of the Eastern Goldfields granites. The siliceous (74–77% SiO₂) high-HFSE granites are restricted to a narrow NNW-trending zone and are characterised by high TiO₂, total FeO, MgO, Y, Zr and Ce, but only moderate Rb, Th and Pb contents. The A-type syenites (50–68% SiO₂) are distinguished by their high total alkalies and mainly occur along tectonic lineaments. The mafic group (55–70% SiO₂) is lithologically diverse and exhibits a wide range of K₂O, Rb, Th, La and Ce contents. Granite emplacement occurred between 2.69 and 2.60 Ga, contemporaneous with, or post-dating, greenstone formation. ϵ_Nd values for the high-Ca, mafic and syenitic groups are similar (mostly 0.0 to +2.5), whereas those for the low-Ca group show a pronounced polarity, from −4.5 in the west to +2.0 in the east. Nd depleted-mantle model ages for the low-Ca granites range from 3.2 Ga in the west to 2.75 Ga in the east. Model ages for the high-Ca group (2.75–2.9 Ga) overlap with ages for inherited zircons from these granites. The high-Ca granites appear to have been derived at high pressures by partial melting of a mafic to intermediate source (subducted oceanic crust, crustal underplate or thickened crust), or remelting of an older tonalitic source derived by such a process. If the high-Ca granites were largely crust derived, then their extensive source rocks must have formed less than 250 Ma prior to granite formation. The low-Ca granites appear to be the products of crustal reworking, probably from a tonalitic to granodioritic protolith that was progressively younger towards the east, perhaps in the order of several hundred million years younger, or involved an additional younger component. The localisation of the crustally-derived high-HFSE granites suggests either geochemical zonation of the crust or specific tectonic processes, such as rifting. The mafic granites, and possibly the syenites, imply a significant mantle contribution and new crustal growth, although the extent is equivocal. The crust of the Eastern Goldfields appears to have been generated by both lateral and vertical accretion over a significant period from, or before, 3.0 Ga until the time of granite emplacement.