Mineralogy and geochemistry of iron-oxides in the Olympic Dam IOCG-deposit and adjacent prospects, South Australia

Abstract

Iron-oxide copper gold (IOCG) mineralisation is defined by an abundance of hematite and/or magnetite. The predominance of one Fe-oxide over the other and their textural relationships can, however, differ significantly within a single deposit or within the same metallogenic province, as a response to variations in the genetic conditions. Analysis of Fe-oxides, bridging scales of observation from deposit down to the nanoscale, highlights different formation conditions within deposits and prospects from the ~1.6 Ga Olympic Dam district, South Australia. In addition, Fe-oxides from the ~5.3-1.6 Ma El Laco Volcanic Complex (Chile), which are debated in terms of magmatic and/or hydrothermal origins, were analysed for comparison with those formed in IOCG systems. At Olympic Dam, a characteristic oscillatory-zoned hematite, containing up to wt% concentrations of U-W-Sn-Mo (‘granitophile’ elements), is the predominant Fe-oxide over the ~6 km-strike and ~2 km-depth of the breccia-hosted mineralisation. Complex textural and compositional zoning patterns within the hematite indicate overprinting and replacement of pre-existing minerals, including earlier hematite, during subsequent episodes of fluid-assisted brecciation and mineralisation. A diverse range of features at the micron- to nanoscale indicates that pseudomorphic replacement of hematite occurred via coupled dissolution and reprecipitation reactions leading to grain-scale (re)mobilisation of minor/trace elements. Primary geochemical signatures are nevertheless partially or selectively preserved. This allows the use of hematite as a reliable U-Pb geochronometer. Mineralogical-geochemical study of Fe-oxides from the outer shell, a weakly-mineralised domain between the host granite and the Olympic Dam orebody, allows for re-interpretation of deposit formation. Iron-oxide assemblages comprise oscillatory-zoned, silician magnetite, high field strength element-bearing hematite, and various interconversion products between the two oxides. Formation of such assemblages is associated with early alkali-calcic alteration (calc-silicate inclusions in silician magnetite), breakdown of magmatic Fe-Ti-oxides, and replacement of igneous magnetite by silician magnetite. Geochemical modelling at 400 °C suggests magnetite-replacement at pH/fO₂ conditions that coincide with stability shifts of K-feldspar → sericite, and ilmenite → rutile. Outer shell formation was initiated at the depth of granite emplacement (~6-8 km), following volatile release from fluids ponding at intrusion margins. Mineralisation continued during uplift to shallower depth, with cupola collapse following extensive fluid release that facilitated brecciation and orebody formation. Nanoscale analysis of Fe-oxides by scanning transmission electron microscopy using Z-contrast imaging and mapping provides further insights into ore-forming processes. Fingerprinting of fluid-mineral interactions during subsequent overprint of U-W-Sn-Mo-bearing hematite show that twins provide pathways for fluid percolation and trap elements exchanged during cycles of coupled dissolution-reprecipitation. Metal nanoparticles (NP) are also trapped within pores developed during transient porosity, which can be hosted by fluid inclusions. Unusual Cu-As-zoning at the micron-scale correlates with Cu-(As)-NPs along fluid inclusion trails and Si-Al-K-bearing twin planes linking such metal enrichment to hydrolytic alteration. W-Pb-enrichment occurs along 2-3 nm-wide twin crests and as W-(Pb)-bearing NPs in hematite displaying ‘granitophile element’-signature with preferential removal of W and preservation of Sn, as shown by grain-scale element patterns. A two-fold superstructure model with oxygen vacancies is constructed to explain W-release from hematite. Nanoscale features observed in relation to twins also allow prediction of disturbances to the U/Pb-systematics of hematite with implications for geochronology. Nanomineral inclusions in magnetite are valuable petrogenetic indicators that can clarify ore-forming processes. At Acropolis, vein titanomagnetite, within ~1.6 Ga volcanic rocks, features nanoscale inclusions with ulvöspinel-hercynite pairs and ilmenite-trellis exsolutions followed by subsequent overprinting (rutile replacement of Ti-phases). Spinel-group associations and ilmenite/magnetite oxythermobarometry supports hydrothermal-magnetite formation at ≥500 °C. Silician magnetite from Wirrda Well hosts nanoinclusions of the rare Al-amphibole, tschermakite. Considering the metamorphism of the host ~1.85 Ma granite, the presence of tschermakite could represents a metamorphic phase that crystallised prior to IOCG-style mineralisation. In contrast, magnetite from Fe-ores at El Laco contains nm-scale inclusions of paired clinopyroxenes (augite-pigeonite/clinoenstatite) with intergrowths indicative of rapid growth and exsolution at low pressures/high temperatures, typical of crystallisation from melts. Sulfur-bearing coatings on comparable but finer-grained pyroxene-bearing magnetite are indicative of ‘magnetite flotation’ via attachment to bubbles of vapour+fluid, allowing transport from a deep magma reservoir to surface. Iron-oxides show remarkable variation in compositional signatures and nanoscale heterogeneity. Detailed petrographic, geochemical and mineralogical analysis of Fe-oxides represents a rich and often untapped source of information that can assist in constraining ore formation conditions and contribute to improved genetic models.