Kathy Ehrig

Origin of the supergiant Olympic Dam Cu-U-Au-Ag deposit, South Australia: Was a sedimentary basin involved?

The supergiant Olympic Dam Cu-U-Au-Ag ore deposit of South Australia occurs in a tectonic-hydrothermal breccia complex that is surrounded by Mesoproterozoic granite. The breccia is composed mainly of granite clasts and minor amounts of Mesoproterozoic volcanic clasts. Very thick (>350 m) sections of bedded sedimentary facies that occur in the breccia complex include laminated to very thin planar mudstone beds, thin to medium internally graded sandstone beds, and thick conglomerate beds. The bedded sedimentary facies extend continuously across a 1.5 km × 0.9 km area and are not limited to small separate maar craters, as previously thought. Detrital chromite and volcanic quartz in the bedded sedimentary facies cannot be matched with local sources, and imply that the provenance extended beyond the area of Olympic Dam. The lateral continuity, provenance characteristics, great thickness, below-wave-base lithofacies, and intracontinental setting suggest that the bedded sedimentary facies are remnants of a sedimentary basin that was present at Olympic Dam prior to formation of the breccia complex. We conclude that the Olympic Dam hydrothermal system operated beneath and partly within an active sedimentary basin, was not confined to maar craters, and did not vent. This new view of the setting raises the possibility that sedimentary processes were involved in ore genesis.

Uraninite from the Olympic Dam IOCG-U-Ag deposit: Linking textural and compositional variation to temporal evolution

Abstract The Olympic Dam IOCG-U-Ag deposit, South Australia, the world’s largest known uranium (U) resource, contains three main U-minerals: uraninite, coffinite, and brannerite. Four main classes of uraninite have been identified. Uraninite occurring as single grains is characterized by high-Pb and ΣREE+Y (ΣREY) but based on textures can be classified into three of these classes, typically present in the same sample. Primary uraninite (Class 1) is smallest (10–50 mm), displays a cubic-euhedral habit, and both oscillatory and sectorial zoning. “Zoned” uraninite (Class 2) is coarser, sub-euhedral, and combines different styles of zonation in the same grain. “Cobweb” uraninite (Class 3) is coarser still, up to several hundred micrometers, has variable hexagonal-octagonal morphologies, varying degrees of rounding, and features rhythmic intergrowths with sulfide minerals. In contrast, the highest-grade U in the deposit is found as micrometer-sized grains to aphanitic varieties of uraninite that form larger aggregates (up to millimeter) and vein-fillings (massive, Class 4) and have lower Pb and ΣREY, but higher Ca. Nanoscale characterization of primary and cobweb uraninite shows these have defect-free fluorite structure. Both contain lattice-bound Pb+ΣREY, which for primary uraninite is concentrated within zones, and for cobweb uraninite is within high-Pb+ΣREY domains. Micro-fractured low-Pb+ΣREY domains, sometimes with different crystal orientation to the high-Pb+ΣREY domains in the same cobweb grain, contain nanoscale inclusions of galena, Cu-Fe-sulfides, and REY-minerals. The observed Pb zonation and presence of inclusions indicates solid-state trace-element mobility during the healing of radiogenic damage, and subsequent inclusion-nucleation + recrystallization during fS2-driven percolation of Cu-bearing fluid. Tetravalent, lattice-bound radiogenic Pb is proposed based on analogous evidence for U-bearing zircon. Calculating the crystal chemical formula to UO2 stoichiometry, the sum of cations (M*) is ~1 for most classes, but the presence of mono-, di-, and trivalent elements (ΣREY, Ca, etc.) drive stoichiometry toward hypostoichiometric M*O2–x. In the absence of measured O and constraints of hypostoichiometric fluorite-structure, charge-balance calculations showing O deficit in the range 0.15–0.36 apfu is compensated by assumption of mixed U oxidation states. Crystal structural formulas show up to 0.20 apfu Pb and 0.25 apfu ΣREY in Classes 1–3, while for Class 4, these are an order of magnitude less. Low-Pb and ΣREY subcategories of Classes 2 and 3 are similar to massive uraninite with ~0.2 apfu Ca. Other elements (Si, Na, Mn, As, Nb, etc.), show two distinct geochemical trends: (1) across Classes 1–3; and (2) Class 4, whereby low-Pb+ΣREY sub-populations of Classes 2 and 3 are part of trend 2 for certain elements. Plots of alteration factor (CaO+SiO2+Fe2O3) vs. Pb/U suggest two uraninite generations: early (high-Pb+ΣREY, Classes 1–3); and late (massive, Class 4). There is evidence of Pb loss from diffusion, leaching and/or recrystallization for Classes 2–3 (low-Pb+ΣREY domains). Micro-analytical data and petrographic observations reported here, including nanoscale characterization of individual uraninite grains, support the hypothesis for at least two main uraninite mineralizing events at Olympic Dam and multiple stages of U dissolution and reprecipitation. Early crystalline uraninite is only sparsely preserved, with the majority of uraninite represented by the massive-aphanitic products of post-1590 Ma dissolution, reprecipitation, and possibly addition of uranium into the system. Coupled dissolution-reprecipitation reactions are suggested for early uraninite evolution across Classes 1 to 3. The calculated oxidation state [U6+/(U4++U6+)] of the “early” and “late” populations point to different conditions at the time of formation (charge compensation for ΣREY-rich early fluids) rather than auto-oxidation of uraninite. Late uraninites may have formed hydrothermally at lower temperatures (T < 250 °C).

Geometallurgical Modeling at Olympic Dam Mine, South Australia

Modeling of geometallurgical variables is becoming increasingly important for improved management of mineral resources. Mineral processing circuits are complex and depend on the interaction of a large number of properties of the ore feed. At the Olympic Dam mine in South Australia, plant performance variables of interest include the recovery of Cu and U3O8, acid consumption, net recovery, drop weight index, and bond mill work index. There are an insufficient number of pilot plant trials (841) to consider direct three-dimensional spatial modeling for the entire deposit. The more extensively sampled head grades, mineral associations, grain sizes, and mineralogy variables are modeled and used to predict plant performance. A two-stage linear regression model of the available data is developed and provides a predictive model with correlations to the plant performance variables ranging from 0.65–0.90. There are a total of 204 variables that have sufficient sampling to be considered in this regression model. After developing the relationships between the 204 input variables and the six performance variables, the input variables are simulated with sequential Gaussian simulation and used to generate models of recovery of Cu and U3O8, acid consumption, net recovery, drop weight index, and bond mill work index. These final models are suitable for mine and plant optimization.

Chemical zoning and lattice distortion in uraninite from Olympic Dam, South Australia

Abstract Compositionally zoned uraninite from the Olympic Dam iron oxide-copper-gold deposit is rarely preserved, but represents an early product of in situ transformation of primary uraninite. Electron backscatter diffraction data (inverse pole figure, image quality, and grain reference orientation deviation mapping) reveal formation of zoned uraninite to be the result of a sequence of superimposed effects rather than from primary growth mechanisms alone. This is the first known microstructural analysis of uraninite showing crystal-plastic deformation of uraninite via formation and migration of defects and dislocations into tilt boundaries. Defining grain-scale characteristics and microstructural features in radiogenically modified minerals like uraninite carries implications in better understanding the processes involved in their formation, highlights limitations in the use of uraninite for U-Pb chemical ages, as well as for constraining the incorporation and release of daughter radioisotopes, especially where zoning, porosity, fractures, and microstructures are present.

Matrix effects in Pb/U measurements during LA-ICP-MS analysis of the mineral apatite

U–Pb ages of several apatite reference materials, acquired by LA-ICP-MS over a 3.5 year period using the Otter Lake apatite as a primary standard, show systematic offsets (up to 3%) from reference ages obtained by isotope dilution mass spectrometry. The offsets are well outside known analytical errors from counting statistics, uncertainty in common Pb corrections or excess 206Pb in high-Th/U materials. The age offsets also do not correlate strongly with the concentrations of F, Cl, REE, Th and U, with accumulated radiation dose or with the depth of ablation pits. We suggest that the observed age offsets are related to elemental fractionation at the ablation site, but their direction and magnitude cannot at present be predicted from compositional differences in different apatite samples. Our results further suggest that Otter Lake apatite is too heterogeneous for use as a calibration standard. Two other apatite samples, OD306 (1596.7 ± 7.1 Ma) and the 401 apatite (530.3 ± 1.5 Ma) are introduced as potential U–Pb reference materials.

Chemical and textural interpretation of late-stage coffinite and brannerite from the Olympic Dam IOCG-Ag-U deposit

Abstract The Olympic Dam iron-oxide copper-gold-silver-uranium deposit, South Australia, contains three dominant U minerals: uraninite; coffinite; and brannerite. Microanalytical and petrographic observations provide evidence for an interpretation in which brannerite and coffinite essentially represent the products of U mineralizing events after initial deposit formation at 1.6 Ga. Marked compositional and textural differences between the various types of brannerite and coffinite highlight the role of multiple stages of U dissolution and reprecipitation. On the basis of petrography (size, habit, textures and mineral associations) and compositional variation, brannerites are divided into four distinct groups (brannerite-A, -B, -C and -D), and coffinite into three groups (coffinite-A, -B and -C). Brannerite-A ranges in composition from what is effectively uraniferous rutile to stoichiometric brannerite, and has elevated (Mg +Mn + Na + K) and (Fe + Al) compared to other brannerite types. It displays the most diverse range of morphologies, including complex irregular-shaped aggregates, replacement bands, and discrete elongate seams. The internal structure of brannerite-A consists of randomly-oriented hair-like needles and blades. Brannerite-B (<5 μm in size) is generally prismatic and typically associated with baryte and REY minerals (REE+Y= REY). Brannerite-C and -D are both associated with Cu-(Fe)-sulfides and are typically composed of irregular masses and blebs (10-50 μm in size) with a more uniform or massive internal structure. Brannerite-D is distinct from -C and always contains inclusions of galena. Brannerite-B to -D all contain elevated ΣREY, with brannerite-B and -C having elevated As, and brannerite-D having elevated Nb. All coffinite is typically globular (each globule is 2-10 μm in size) to collomorphic in appearance. Coffinite-A ranges from discrete globules to collomorphic bands completely encompassing quartz. Coffinite-B is always found with uraninite, and includes collomorph coffinite enveloped by massive uraninite, as well as aureoles of coffinite on the margins of uraninite crystals. Coffinite-C is associated with brannerite and REY minerals. The majority of coffinite is heterogeneous. Brannerite and coffinite have probably precipitated as part of a late-stage hydrothermal U-event, which might have involved the dissolution and/or reprecipitation of earlier precipitated uraninite, or could represent the products of a later U mineralizing event. Evidence which supports formation of late-stage coffinite and brannerite includes: (1) low-Pb contents of both minerals; (2) coffinite is commonly found on the edges of uraninite, implying later deposition; and (3) coffinite is often found on the edge of brannerite aggregates, suggestive of brannerite precipitation occurred before coffinite. Moreover, there are many features (e.g. banding, scalloped edges, alteration rinds, variable compositions etc.) indicative of hydrothermal alteration processes.

Alteration at the Olympic Dam IOCG–U deposit: insights into distal to proximal feldspar and phyllosilicate chemistry from infrared reflectance spectroscopy

ABSTRACT Twenty thousand metres of diamond drill core representing a 14 km cross-section from weakly to intensely altered Roxby Downs Granite through the Olympic Dam Breccia Complex, host to the Olympic Dam iron-oxide–copper–gold–uranium deposit in South Australia, was analysed using the HyLogger-3 spectral scanner. Thermal and shortwave infrared spectroscopy results from 30 drill holes provide insight into the spatial relationships between quartz, orthoclase–microcline, albite–oligoclase and progressively changing sericite and chlorite compositions. The relative proportions of quartz, feldspars and phyllosilicates were mapped with thermal infrared spectroscopy. Variations in the chemistry of sericite and chlorite were extracted by proxy from their shortwave infrared spectral response, together with their relative spatial distribution. HyLogger scanning has revealed four deposit-scale mineralogical trends, progressing from least-altered Roxby Downs Granite into mineralisation where most of the feldspar has been replaced by sericite + hematite + quartz: (1) a progressive Al–OH wavelength shift of 2205 nm to 2210 nm for sericite, followed by a spatially rapid reversal corresponding to lower phengite/muscovite abundance ratios; (2) progressive Mg/Fe–OH wavelength shift of 2248 nm to 2252 nm reflecting an increase in the Fe:Mg ratio of chlorite; (3) increasing ratio of microcline to orthoclase followed by a rapid decrease; and (4) slightly decreasing ratio of albite to oligoclase followed by plagioclase destruction prior to albite replacement by sericite. The HyLogger feldspar results support recent petrographic evidence for hydrothermal albite and K-feldspar at the Olympic Dam deposit, not previously reported. The spectral results from continuous HyLogger scans also show that the microscopic observations and proposed feldspar replacement reactions are not locally isolated phenomena, but are applicable at the deposit and regional-scale. A modified quartz–K-feldspar–plagioclase ternary diagram utilising mineralogy interpreted from HyLogger thermal infrared spectra (QAPTIR) diagram along with supporting data on the abundance ratios of orthoclase/microcline and albite/plagioclase, and the wavelength shifts in characteristic absorption features for sericite and chlorite, can be used as empirical vectors towards mineralisation within the Olympic Dam mineral system, with potential application to other IOCG ore-forming systems. Intrusion of Gairdner Dyke Swarm dolerite dykes into sericite ± hematite altered Roxby Downs Granite results in retrograde albite–chlorite–magnetite alteration envelopes (up to tens of metres thick) overprinting the original sericite ± hematite alteration zone and needs to be carefully evaluated to ensure that such areas are not falsely downgraded during exploration.

Steps to developing iron-oxide U-Pb geochronology for robust temporal insights into IOCG and BIF mineralisation

distinct zones of Cu and Pb–Zn mineralisation. The textural and structural relationships associated with the mineralised lodes suggests that they formed through oblique-slip faulting within a transtensional regime. The geochemical and economic significance of the Copper Coast mineralisation is not yet resolved, although the structural and petrographic evidence, revealed by this study, indicate that the system may represent remobilization from a proximal VMS deposit (Figures 1–3).

Ore Minerals Down to Nanoscale: Petrogenetic Implications

Like most minerals, sulphides and oxides are compositionally heterogeneous at the smallest scale. The application of a variety of different techniques is necessary to fully understand their crystal-chemistry, each with its own inherent advantages and dis-advantages, and each with limited ranges of obser-vational scale. Focussed Ion Beam Scanning Electron Microscopy (FIB-SEM) opens up new avenues for in-situ sampling of small volumes of material which have been compositionally characterised by other techniques, including SEM, EPMA and LA-ICPMS. Slices lifted from the surface of a polished section, typically 40 x 20 µm in size and a few µm thick, can be imaged and prepared for Transmission Electron Microscope (TEM) analysis. This procedure allows for bridging micro- to nanoscale observations on a site which is of petrogenetic interest (e.g., Ciobanu et al., 2011). One of the main themes of inquiry is the distribution of trace elements in ores relative to subtle changes in mineral speciation, nanoparticle nucleation or crystal-structural modifications. Here we show examples of the application of FIB-SEM and TEM studies to such topics from across ore geology. 2 Homology in Mixed-Layer Chalcogenides Bismuth minerals are common accessory minerals in gold deposits, and include Pb-Bi-sulphosalts and – tellurides. They form homologous or polysomatic series where small chemical variation from one species to another is linked to modifications of crystal-structural layers (modules or blocks) in a predictive manner. These minerals can thus record subtle variations in the chemistry of an evolving ore system via crystallisation processes at the smallest scale. One example is the tetradymite group, where the crystalchemical formula nBi2·mBi2X3 (X = Te, Se, S; e.g., Cook et al., 2007) indicates that all the phases can be derived from archetypal layers of fixed width, but of variable numbers (n, m) in the stacking sequence. Based on TEM studies that show monotonic decrease of crystal modulation vectors with increase in Bi, the homology in the group has been

Discrimination and Variance Structure of Trace Element Signatures in Fe-Oxides: A Case Study of BIF-Mineralisation from the Middleback Ranges, South Australia

Uni- and multivariate statistical analyses of trace element laser-ablation inductively coupled plasma mass spectrometry data for Fe-oxides from banded iron formation (BIF) and BIF-hosted ores in 13 deposits/prospects of the Middleback Ranges, South Australia are presented. The obtained trace element signatures were considered within a petrographic-textural framework of iron-oxide evolution from magnetite through clean martite and porous martite to platy hematite, to evaluate changes in trace element concentrations with respect to the ore enrichment processes. Statistically valid distinctions among different hematite textures were indicated for most trace elements by linear mixed-effects models. Furthermore, the hematite data showed significant intra-class correlations between spot-analyses within individual polished blocks and correlations between polished blocks within individual deposits. The data are thus aggregated within their hierarchical levels. Two linear discriminant function analyses were performed to determine the combinations of trace elements that can distinguish hematite by textures and by location within the Middleback Ranges. Tin, a significant discriminator element in both models, reflects the regional influence of granite-affiliated hydrothermal fluids on the clean martite. This granitic signature, therefore, postdates formation of magnetite BIFs and potentially represents a supergene ore enrichment stage. The combination of Ni, Co, Ti and Nb was discovered to be uniquely attributed to discrimination of the Northern and Southern Middleback Ranges, indicating very specific local settings unrelated to hematite textures. Both local and regional settings impacting on the trace element signatures of Fe-oxides throughout iron ore formation are recognised, suggesting distinct ore enrichment conditions within various segments of the belt.