Petrology, geochemistry and geochronology of mafic lithologies at the Olympic Dam iron oxide Cu-U-Au-Ag deposit : implications for tectonic settings and ore-forming processes

The Olympic Dam deposit, South Australia, is a supergiant iron oxide Cu-U-Au- Ag deposit currently containing the world’s largest uranium, fifth largest copper, and third largest gold resource (10,100 Mt at 0.78% Cu, 0.25 kg/t U₃O8, 0.30 g/t Au, 1 g/t Ag, BHP Billiton 2015 Annual Report). The exploration model leading to the discovery of the deposit in 1975 was source-oriented: altered mafic lithologies were considered as the potential source of copper and early exploration focused on defining and searching for these lithologies. Not surprisingly, drilling and underground mining at Olympic Dam have revealed the occurrence of more than one generation of mafic lithologies. Previous studies have considered these mafic lithologies to be closely related to brecciation, hydrothermal circulation (by providing heat), and mineralization at Olympic Dam. One suite of mafic rocks, inferred to belong to the ca. 1590 Ma Gawler Range Volcanics (GRV), has been proposed to be a major source of Cu (~50% of the contained Cu) found in the deposit. However, at over forty years since discovery of the deposit, questions remain as to which groups the various mafic lithologies at Olympic Dam belong to regionally, as to what kinds of tectonic settings there were when the mafic lithologies were emplaced, and as to how the emplacement and alteration of the mafic lithologies were related to the ore-forming processes at Olympic Dam. This study endeavours to provide answers to these issues. The presence of two generations of mafic lithologies at Olympic Dam has been confirmed by primary accessory apatite U-Pb dating. The first group is found to be correlated with the ca. 1590 Ma Gawler Range Volcanics and the Gawler silicic large igneous province (SLIP), consisting of intensely altered olivine-phyric basalt, and other mafic dykes of various textures (aphanitic, porphyritic, and doleritic). The second group comprises basaltic to mainly doleritic dykes (named the Olympic Dam dolerite), proved to belong to the ca. 820 Ma Gairdner Dyke Swarm and the Gairdner large igneous province (LIP). The ca. 1590 Ma olivine-phyric basalt at Olympic Dam typically contains a higher abundance of former olivine phenocrysts (~20 vol.%) and is more intensely altered thanequivalent mafic GRV outside the deposit at Kokatha, Mount Gunson and Wirrda Well. Therefore, this suite of basalt represents the most ultramafic component in the GRV recognized thus far. Compositions of a large number of Cr-spinel inclusions within olivine pseudomorphs and olivine-phyric basalt samples of mafic GRV (including Olympic Dam samples) reveal different magma types and indicate derivation of mafic GRV from a heterogeneous mantle source that may have been modified by subduction, implying a setting proximally to a continental margin or in a back-arc. This inference is also compatible with the previous proposal that the formation of the Gawler SLIP was associated with the assembly of the Laurentian supercontinent. The ca. 820 Ma Olympic Dam dolerite shows similar petrographic features and comparable compositional variations to the regional Gairdner Dyke Swarm. The correlation of the Olympic Dam dolerite with the Gairdner Dyke Swarm thus extends the spatial distribution and the compositional spectrum of the latter. Geochemical comparisons among LIP and mafic suites in South Australia (including the Gairdner Dyke Swarm), South China and North America, associated with the break-up of the supercontinent Rodinia are in support of the “missing-link” model in which South China was situated on top of a mantle plume and South Australia (including Olympic Dam) was affected by the plume-induced rift magmatism at ca. 820 Ma. In conclusion, from the perspective of mafic magmatism, the evolution of the supergiant Olympic Dam deposit turned out to be linked to two supercontinent cycles: the assembly of Laurentia at ca. 1590 Ma, and the break-up of Rodinia at ca. 820 Ma. Investigations of the alteration of the two generations of mafic lithologies at Olympic Dam have shown that they contain similar major secondary minerals (chlorite, sericite, and carbonate). Both are characterized by a mineral assemblage of magnetiteapatite ± chlorite ± quartz that is strikingly similar to the inferred early reduced iron oxide alteration present in the periphery and at depth of the Olympic Dam Breccia Complex that is the immediate host to ore. Secondary spongy apatite in the ca. 1590 Ma olivine-phyric basalt yielded an age that is broadly coeval with the basalt’s emplacement. Sericite-altered basalt produced a post-primary Rb-Sr isochron age of ca. 1180 Ma, likely to indicate the most recent significant sericite alteration of these rocks. Extreme concentrations of some components (e.g. up to 26 wt.% CO₂, 50 wt.% of Fe₂O₃, and 6,000ppm of Cr) and extraordinary near linear positive correlations between Cr andhigh field strength elements (e.g. Ti, Nb, and Zr) revealed in drill core assays of the basalt indicate significant whole-rock mass and/or volume loss due to hydrothermal alteration, in accordance with its previously proposed role as a major Cu source. Results obtained on the secondary apatite and titanite in the ca. 820 Ma Olympic Dam dolerite also confirmed hydrothermal activities associated with the emplacement of the younger dolerite. Pb isotope compositions of the dolerite as well as chalcopyrite and galena within the dolerite indicate derivation of radiogenic Pb from hydrothermal fluids circulating through the Olympic Dam Breccia Complex, implying that the dolerite was a part of the active hydrothermal system at Olympic Dam at ca. 820 Ma. Elevated Zn, Pb and depleted Cu concentrations of the ca. 820 Ma Olympic Dam dolerite compared to dolerite worldwide suggest that the dolerite can even be an additional Cu source to the Olympic Dam deposit. Moreover, the younger and less altered dolerite provides a better opportunity to envisage Cu depletion processes that may also be anticipated for the older olivine-phyric basalt, where such processes are no longer recognizable due to superimposition of multiple hydrothermal events on the basalt. At last and most importantly, magmatic-hydrothermal activities (ca. 1590 Ma, 1180 Ma, 820 Ma) recorded in both generations of mafic rocks can be correlated with ages (spanning from ca. 1590 Ma to 570 Ma) obtained on the Olympic Dam Breccia Complex. This advocates the existence of an all-encompassing, multi-stage, hydrothermal system at the supergiant Olympic Dam deposit