A comparative in situ LA-ICP MS trace-element study on pyrite from three different, variably auriferous, Archaean to Palaeoproterozoic palaeoplacer deposits in the Ouro Fino Syncline (Quadrilatero Ferrifero; Brazil), the Elliot Lake area north of Lake Huron (SE Canada) and several deposits within the Witwatersrand Basin (South Africa) revealed systematic differences between morphologically different pyrite types and between the various palaeoplacer deposits. Especially the Ni and Au concentrations as well as Co/Ni and Mo/Ni ratios were found to be systematically different in detrital compact, detrital porous and post-sedimentary/hydrothermal pyrite grains from different source areas. High Co/Ni ratios and low Au concentrations are typical of post-sedimentary pyrite, which is hydrothermal in origin. In contrast, relatively low Co/Ni ratios and high Au contents characterise detrital porous banded and concentric pyrite grains (Au > 1 ppm), which are syn-sedimentary in origin. In the Elliot Lake area and the Witwatersrand Basin, detrital compact rounded pyrite is characterised by high Co/Ni ratios, which is in agreement with derivation from a hydrothermal source. Low Au concentrations in this pyrite type support the contention of the gold and the pyrite in these deposits coming from different source rocks. In contrast, derivation from an originally diagenetic pyrite is suggested for the detrital compact pyrite in the Ouro Fino Syncline because of low to intermediate Co/Ni ratios. High Au contents may indicate a genetic relationship between pyrite and gold there. Systematic differences exist between the three areas with respect to Au, Ni, Co, Mo and Cu distributions in detrital pyrite, which reflects differences in the provenance. A predominantly mafic/ultramafic source is indicated for the Ouro Fino, a felsic source for the Elliot Lake, and a mixed felsic–mafic provenance for the Witwatersrand pyrite populations. Independently of pyrite type, the higher Au endowment of the studied Witwatersrand and Ouro Fino conglomerates are also reflected by an overall higher Au concentration in the respective pyrite grains compared to the relatively Au-poor samples from Elliot Lake. In general, a strong positive correlation between Au and Pb levels in the various pyrite grains is noted. Analogous to Pb, which is well known for not being easily accommodated in the pyrite crystal lattice but occurring as discrete PbS phases, Au is considered to be present mainly in the form of discrete Au phases in minute pores and interstices of the pyrite grains rather than within the pyrite lattice.

Trace-element characteristics of different pyrite types in Mesoarchaean to Palaeoproterozoic placer deposits

The origin of the gold mineralization in the Witwaterstrand basin of South Africa—the largest known gold province—has been controversial for decades, with arguments favouring detrital1,2 (placer), modified placer3,4 and hydrothermal5,6 origins. Here we present the results of an extensive geological study of Witwatersrand rocks which show that the gold (and associated uranium) mineralization is hydrothermal in origin and postdates a regional high-temperature alteration event. Alteration processes identified on a small scale can be mapped out regionally as roughly strata-bound zones of acid metasomatism extending far into the basin: the fluid flow responsible for this alteration was concentrated in small-scale structures localized along lithological boundaries. We find that the gold precipitated as a consequence of interactions of the fluid with shale-derived hydrocarbons present within the basin.

Hydrothermal gold mineralization in the Witwatersrand basin

Archaean atmospheric evolution: evidence from the Witwatersrand gold fields, South Africa

— Historically, there have been a range of diameter estimates for the large, deeply eroded Vredefort impact structure within the Witwatersrand Basin, South Africa. Here, we estimate the diameter of the transient cavity at the present level of erosion as ∼124–140 km, based on the spatial distribution of shock metamorphic features in the floor of the structure and downfaulted Transvaal outliers. Taking erosion into account (<6 km) and scaling to original final rim diameter, an estimate of close to 300 km for the rim diameter is obtained. Independent estimates of the final rim diameter, based on an empirical relation of central uplift diameter to rim diameter, spatial distribution of pseudotachylites, and concentric large scale structural patterns, give a similar estimate of close to 300 km for the original final rim diameter. An impact structure of this size is expected to have had an original multi-ring form. At this size, the Vredefort impact structure encompasses the bulk of the Witwatersrand Basin, which appears to owe its preservation to the Vredefort impact. In addition, the Vredefort impact event may have been the thermal driver for some of the widespread hydrothermal activity in the area, which, in recent interpretations, is believed to be a component in the creation of the world-class gold deposits of the Witwatersrand Basin.

https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1945-5100.1997.tb01242.x

Neil G. Phillips

Papers

Metamorphism of the Witwatersrand gold fields: A review