ALBERT

Description: Mount Keith is a large (up to 500 m thick, and 〉2 km long) serpentinized and talc-carbonate-altered dunite body. The geologic features of Mount Keith are similar to many of the ultramafic bodies in the Agnew-Wiluna belt and other greenstone belts within the eastern Yilgarn craton. Mount Keith hosts a large disseminated nickel sulfide orebody that is presently exploited by open pit. There are currently two hypotheses that explain the site of formation of these dunite bodies: either kilometer-scale komatiite lava channels, or subvolcanic sills intruded into the greenstone sequence are responsible. The intrusive model is essentially based on two pieces of geologic evidence: the presence of a single 20-m-scale and single 20-cm-scale ultramafic apophysis in overlying dacitic rocks; and the presence of dacite inclusions within the contact zone of the upper part of the Mount Keith ultramafic unit. New geologic data show that the western part of the Mount Keith ultramafic unit and the overlying dacite sequence are allochthonous and that the contact zone between the two comprises multiple anastomosing layer-parallel fault slices of both lithologies at a scale of meters to tens of meters. Within some of the fault blocks, primary igneous contacts are present between the top of the Mount Keith ultramafic unit and the overlying dacite. New evidence shows that both pyroxene and minor olivine spinifex-textured rocks are present along much of the upper preserved part of the Mount Keith ultramafic unit and pseudomorphs after pyroxene grains exist on a scale of 20 to 200 μ m that formed during emplacement and survived the cooling of the 500-m-thick Mount Keith ultramafic unit. Moreover, the evidence for a large-scale ultramafic intrusive apophysis is unsubstantiated, as this apophysis is shown to be a slice within a brittle/ductile fault. The reinterpretation of the small-scale apophysis is consistent with this being an ultramafic enclave within a younger dacite flow. Finally, the dacite exposed within centimeters to meters of the western ultramafic contact commonly retains delicate igneous textures and shows no evidence of thermal effects from the Mount Keith ultramafic unit. Modeling of the heat budget assuming the dimensions of the Mount Keith ultramafic unit shows that an intrusion of this size should have produced wholesale melting of roof rocks. Our findings indicate that the Western Dacite sequence was not present above the Mount Keith ultramafic unit during its emplacement and cooling. A model is favored for the extrusive origin of the Mount Keith ultramafic unit with construction of the olivine cumulate pile at the floor of an approximately 2-km-wide, intermittently sulfide-bearing, turbulent lava pathway. At the terminal stages of eruption magma flow decreased and finally ceased, allowing the lava to drain and cause the upper crust to collapse onto the top of the crystal pile, producing in places sharp juxtaposition of chilled margin rocks with coarse-grained olivine cumulates. In most places rapid flow and turbulent convection resulted in resorption of the crust, but local formation of stagnant ponds allowed survival of an olivine spinifex-textural profile. More extensive ponding of the magma allowed in situ fractionation and formation of gabbroic and pyroxenitic rocks derived from highly fractionated komatiite magma. The Mount Keith ultramafic unit is believed to be the product of extrusive magmatism from which an olivine cumulate pile developed by upward accretion at the floor of a major lava pathway; these rocks were then extensively modified by deformation.

Description: A key step in the formation of many magmatic Ni-Cu-PGE sulfide deposits is the addition of crustal sulfur to mafic or ultramafic magmas. Sulfur addition has been proposed to take place via two different types of processes: (1) production of sulfurous fluids within the thermal aureole around an intrusion accompanying breakdown of sulfide- or sulfate-bearing minerals during devolatilization, followed by diffusive or advective transport of these fluids into the magma, and (2) by direct melting and assimilation of wall rock and xenoliths. We consider physical and chemical controls on the timescales of these processes and show that wall-rock and xenolith melting is by far the most efficient and quickest process for adding crustal sulfur, with melting processes taking place on a scale of minutes to years. In contrast, liberation of sulfur from a thermal aureole via diffusion is much slower and requires timescales of millions of years—two orders of magnitude longer than the time required for an intrusion to solidify by diffusion. We conclude that sulfur, which may be liberated in thermal aureoles (produced either via devolatilization reactions or dissolution involving hydrothermal fluids) and which must be transported via diffusional processes, has a negligible effect on the formation of magmatic Ni-Cu-PGE deposits.

Description: Interspinifex ores are developed where pools of sulfide liquid overlie thermally eroded komatiite flows, such that sulfide occupies the space between spinifex olivine plates. Microbeam X-ray fluorescence mapping and 3-D X-ray computed tomography have been used to investigate microstructures and chemical zonation within interspinifex ores from Coronet shoot, Kambalda. Sulfide compositions in the interspinifex space match the composition of the overlying sulfide pool. Aluminous silicate, interpreted as displaced silicate melt, forms a film at the silicate-sulfide interface, locally developing dome-like plumes. The film and plumes are characterized by fine skeletal chromite that diminishes in abundance over about a decimeter downward into the interspinifex zone. These relationships are strong evidence for a primary magmatic origin for this ore type, driven by the strong tendency of dense, inviscid sulfide liquid to infiltrate and melt underlying rocks. Such infiltration-melting interfaces may be a common feature at the base of massive sulfide ores, taking different forms depending on the lithological and fracturing characteristics of the footwall rocks.

Description: The remobilization of metals during postdeposition hydrothermal alteration of magmatic sulfide ores has the potential to result in large haloes, the recognition of which could potentially enlarge the detectable footprint of this ore type. The Miitel komatiite-hosted nickel sulfide deposit in Western Australia was used as a case study to investigate the nature and 3-D geometry of the geochemical halo created by the remobilization of base metals, gold, and platinum group elements (PGE) into the rocks surrounding the mineralization. At Miitel, anomalous metal enrichment is found in the country rocks surrounding the massive sulfides, up to 250 m away from the ore. This enrichment, detected using portable X-ray fluorescence (pXRF) and backed up by laboratory analyses, occurs in the Mount Edwards footwall basalt within decimeters of the contact with the overlying Widgiemooltha komatiites. It is associated with the presence of nickel arsenides. Gersdorffite and minor nickeline are concentrated within small quartz and carbonate veinlets, and are interpreted to form during the circulation of arsenic-rich hydrothermal fluids. Results of lead fire assay analyses and in situ laser ablation-inductively coupled plasma-mass spectrometer (LA-ICP-MS) analyses indicate high PGE concentrations (Pd and Pt) and minor gold associated with the observed nickel and arsenic enrichment. Results from a larger, regional-scale study, combined with this PGE enrichment, suggest that the massive nickel sulfides from the Miitel ore are the source of the remobilized nickel in the country rocks. The presence of Pd- and Pt-enriched trace arsenide phases in country rocks and shear zones may be a generally applicable proximity indicator for nickel sulfides in hydrothermally altered terranes.

Description: More than 390 chromite grains from komatiites and komatiitic basalts from the Yilgarn craton of Western Australia and the Finnish part of the Fennoscandian Shield were analyzed using in situ laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) to identify ruthenium (Ru) signatures in chromite associated with nickel sulfide-bearing rocks. Results indicate a potential method to discriminate mineralized and barren komatiite and komatiitic basalt units based on Ru concentrations in chromite and indicate potential for chromite to be used as a resistate indicator mineral in exploration for komatiite-associated nickel sulfide deposits. Chromites from barren komatiites and komatiitic basalts display Ru concentrations mostly between ~150 and 600 ppb. Chromites from mineralized units have distinctly lower Ru contents (〈150 ppb). These results can be interpreted in terms of the much higher partition coefficient for Ru into sulfide liquid compared to that of Ru into chromite, resulting in much lower Ru concentrations in chromite where both chromite and sulfide liquid are present and competing for Ru. As a result, the Ru content of chromite can be used to determine if a komatiite melt equilibrated with a sulfide liquid during crystallization, and therefore, if a system and/or sequence is prospective for nickel sulfide mineralization. The strength of this method compared to previous whole-rock exploration techniques derives from combining (1) the geochemical properties of a chalcophile element that records an ore-forming process while being strongly immobile during postmagmatic processes, with (2) the in situ analysis of a mineral that is generally preserved even in highly altered and mildly weathered komatiites and that is a common constituent of detrital heavy mineral samples. Chromite Ru content has potential as a prospectivity indicator, applicable to a wide range of media including bedrock, laterites, and detrital resistates heavy mineral samples.

Description: Portable X-Ray Fluorescence (pXRF) analysers allow on-site geochemical analysis of rock powders and drill core. The main advantages of pXRF analysis over conventional laboratory analysis are the speed of data collection and the low cost of the analyses, permitting the collection of extensive, spatially representative datasets. However, these factors only become useful if the quality of the data meets the requirements needed for the purposes of the study. Here, we evaluate the possible use of portable XRF to determine element concentrations and ratios used in exploration for komatiite-hosted nickel sulphides. A portable XRF analyser was used to measure a series of chalcophile and lithophile element concentrations (Si, S, K, Ca, Ti, Cr, Fe, Ni, Cu, Zn, As, Sr, and Zr) of 75 samples from three komatiite units associated with nickel sulphide ores in the Yilgarn Craton, Western Australia. Crucial steps in the study were the development of a strict calibration process as well as numerous data quality checks. The 670 analyses collected in this study were compared with conventional laboratory XRF data on discriminant diagrams commonly utilized in exploration for komatiite-hosted nickel sulphides (Cr vs Ni and Ni/Ti vs Ni/Cr). After comparing the results obtained with pXRF during this study with the laboratory values, we can conclude that portable XRF analyses can be used for rapid assessment of the nickel sulphide prospectivity of komatiites provided that strict control protocols are followed. Supplementary Material: is available at http://www.geolsoc.org.uk/SUP18706

Description: Sulfide droplets from fresh Mid-Ocean-Ridge Basalt (MORB) glasses show different textures. Some are fine-grained droplets consist of Monosulfide Solid Solution (Mss) and Intermediate Solid Solution (Iss) micrometric intergrowths with pentlandite at the Mss-Iss interface and disseminated Fe-oxide grains; other droplets display a characteristic "zoned" texture consisting of segregated massive grains of Mss and Iss, with euhedral Fe-oxides and pentlandite occuring as equant grains and as flame-shaped domains in the Mss formed by exsolutions. The difference in the textures implies a difference in the crystallization history of the sulfide droplets. These different textures are observed in droplets that are only millimeters apart in the same sample, and thus had an identical cooling history. Therefore, some other factors controlled the textural development. There is relationship between the size and the texture of the droplets. The larger sulfide droplets tend to have zoned textures and the smaller ones fine-grained textures. We propose that the latter have experienced greater undercooling before crystallization. The reason for the delay in crystallization could be that, in the small sulfide droplets, large stable grains with low surface to volume ratio cannot form, which results in higher effective solubility of the Mss. Due to the high degree of undercooling in the small droplets, there were numerous nucleation sites and the diffusion rates of the crystal components in the liquid were lower, leading to fine-grained Mss-Iss intergrowths. In contrast, larger droplets with lower effective solubility of Mss began to crystallize at higher temperature, and thus had fewer nucleation sites, higher diffusion rates, and more time for sulfide differentiation.

Description: The detectable footprints of komatiite-hosted nickel sulfide deposits are typically very small, but can potentially be enlarged by identifying subtle geochemical variations related to ore-forming processes in the host rocks. This study examines the spatial variability of whole-rock concentrations of platinum group elements (PGEs) within the host flow to massive nickel sulfide mineralization at the Long-Victor deposit, Kambalda dome (Western Australia), where a series of ore shoots occupy two subparallel channels, over a strike length of approximately 3,000 m. The basal komatiite flow unit at Long-Victor contains a wide range of platinum group element concentrations and PGE/Ti ratios in S-poor rocks outside the ore shoots. About a third of the samples analyzed show evidence for either enrichment or depletion in PGEs, as estimated from mantle-normalized ratios of Pt/Ti, Pd/Ti, and Rh/Ti, relative to background values typical of those found in Neoarchean Munro-type komatiites worldwide. The very strong correlations observed between Pt/Ti, Pd/Ti, and Rh/Ti testify to a primary magmatic origin of this signal. Depletion signatures are largely restricted to samples in the flanking environment within the basal flow, and are found both in spinifex-textured A-zone and cumulate B-zone samples. The strongest depletion signatures are preserved in the uppermost portions of the A-zone and decrease in magnitude with increasing depth from the stratigraphic top of the spinifex horizon downward. This is interpreted as the result of progressive flushing of the flow channel by PGE-undepleted lava subsequent to ore formation. Enriched signatures are largely restricted to cumulate rock types, and are found within both channels and flanks. The halo of anomalous PGE/Ti ratios, both depleted and enriched, extends more than 400 m from the cutoff of 0.4% Ni that defines the limit of disseminated ores, and is much more extensive and marked than anomalies defined by Ni concentrations, Ni/Cr ratios, or Ni/Ti ratios, which extend no more than 20 m beyond the disseminated ores themselves. The PGE enrichment halo is recognizable in rocks having no visible sulfide and having Ni values falling within the silicate background, and is attributed to the accumulation of small proportions of PGE-rich disseminated sulfide liquid formed at high R factors, with subsequent loss of S during hydrothermal alteration. Mapping of PGE/Ti ratios provides an effective and sensitive method for vectoring toward ore during mine-scale and prospect-scale exploration, and is potentially applicable to mafic systems as well as to komatiites.

Description: The partition coefficients of platinum group elements (PGE) and chalcophile elements Au, Re, Ag, Se, Bi, Te, and Sb, between arsenide and sulfide phases (D As/sulf ) have been estimated by measuring in situ concentrations of these elements using laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) in coexisting arsenide and sulfide minerals from the Beni Bousera Cr-Ni mineralization (North Morocco). Previous experimental studies and observations on the distribution of PGE in a number of As-rich, Ni-Cu-PGE ore deposits have shown that arsenide minerals may play an important role controlling the distribution of these metals in magmatic sulfide systems. However to date, there is no comprehensive study quantifying the partitioning behavior of these elements when arsenide minerals crystallize either directly from a sulfide melt or from an arsenide melt previously segregated by immiscibility from a sulfide melt. The Beni Bousera mineralization represents an excellent natural laboratory to evaluate these partition coefficients because maucherite (Ni 11 As 8 ) coexists in equilibrium with pyrrhotite, pentlandite, and chalcopyrite in form of globules mostly associated with pyrrhotite, and arsenide and sulfide minerals account for the bulk of the PGE (with the exception of Pt) and chalcophile elements in the samples. The laser ablation analyses reveal that maucherite is strongly enriched in all chalcophile elements, except Se, relative to sulfide minerals. The calculated D PGE As/sulf are the following: D Ir As/sulf = 920 D Rh As/sulf = 620, D Pt As/sulf = 330, D Pd As/sulf = 250, D Os As/sulf = 140, and D Ru As/sulf = 50. For the rest of elements, the obtained values are the following: D Sb As/sulf = 890, D Te As/sulf = 190, D Bi As/sulf = 50, D Re As/sulf = 6, D Au As/sulf = 310, D Ag As/sulf = 4, and D Se As/sulf = 0.6. These results clearly highlight the strong affinity of PGE for arsenide phases and the importance of these phases as potential carriers of PGE in Ni-Cu-PGE ore deposits.

Description: The Kunene Complex of Namibia-Angola is one of the largest anorthosite massifs on Earth (up to 18,000 km 2 ), consisting of several distinct anorthosite and leucotroctolite intrusions. The Namibian portion of the Kunene Complex measures ~80 x 50 km, ~4,000 km 2 , and is dominated by the Zebra Mountain lobe, a ~16-km-thick dome-like mass of interlayered, relatively unaltered dark leucotroctolite with relatively altered, "white," anorthosite. Past studies and the present work have found evidence for intrusion of two distinct phases of dark leucotroctolite into the white anorthosite, namely a relatively early, deformed, phase dated at 1363 ± 17 Ma (U-Pb in baddelyite), and a relatively later and undeformed phase whose absolute age remains unknown. The Kunene leucotroctolites are among the least evolved troctolites known from anorthosite complexes, with olivine containing 59 to 77 mol % forsterite and up to 1,700 ppm Ni, and plagioclase containing 56 to 69 mol % anorthosite. Our isotope data from the troctolites indicate a relatively small crustal component ( 18 O, ~5.3–7.3; 34 S, 0.5–1; and Nd T , 0.9–1.8), whereas Nd and oxygen isotope data from the white anorthosites, published by other workers, showed a slightly larger crustal component (e.g., Nd T as low as –3; 18 O up to 7.5). In the periphery of the Kunene Complex are several, relatively small (〈10 km 2 ), mafic-ultramafic intrusions comprising peridotite, pyroxenite, gabbro, troctolite, and anorthosite. Some of these bodies are Ni-Cu-PGE mineralized, including the Ohamaremba troctolite, the Oncocua pyroxenite, and the Ombuku peridotite-gabbronorite. The latter additionally contains a massive chromitite layer. A new U-Pb baddelyite age of 1220 ± 15 Ma for Ohamaremba indicates that the latter postdates the main Kunene Complex by ~140 Ma. The relative enrichment in MgO, Cr, and Ni, and the O, Nd, and S isotope characteristics of Kunene magmatism suggest that the primary magmas were predominantly mantle-derived picrites or basalts. The massif-type anorthosites formed through ascent of feldspathic slurries followed by downward draining of residual liquid. Subsequent magma pulses formed troctolitic sills within the anorthosite plutons and mafic-ultramafic satellite intrusions in the periphery of the anorthosites. The recurring nature of Kunene mafic-ultramafic magmatism results from several successive mantle upwellings. Partial mantle melts ascended through reactivated translithospheric lineaments along the southern margin of the Congo craton.