Search for: All records

Abstract Drilling related to development of the platinum-group element deposit of the J-M Reef of the Stillwater Complex returned samples of a rare chromitite seam between anorthosite and norite in a discordant anorthositic body. Plagioclase core An concentrations are marginally higher and modestly reversely zoned on the norite side (average Ancore = 83·8; average Ancore – Anrim = –1·1) as compared with the anorthosite side (average Ancore 82·5; average Ancore – Anrim = +1·0). The anorthosites are also characterized by a slightly smaller average plagioclase grain size than plagioclase in the norite (1·41 mm and 1·54 mm, respectively). The chromite can contain single and polyphase inclusions of orthopyroxene, plagioclase, amphibole, biotite and Cl-rich apatite. These and other compositional and textural features, as well as inference from other discordant anorthositic bodies in the Banded series, are all consistent with a chromatographic model of chromite precipitation at a reaction front as a norite protolith reacts with a Cl-rich aqueous fluid saturated in plagioclase alone. Chromitite seam formation is modeled using an infiltration metasomatic model, in which a fluid becomes progressively undersaturated in pyroxene as it rises into the hotter part of the crystal pile. As this pyroxene-undersaturated fluid moves through a noritic protolith, it dissolves the Cr-bearing orthopyroxene to produce an anorthosite. Chromite precipitates at the reaction front between the anorthosite and the norite owing to liberation of Mg and Cr from pyroxene. Continuous redissolution and reprecipitation of chromite occurs as the pyroxene dissolution front moves in the direction of fluid flow, collecting the Cr lost from the anorthosite. Owing to Cr dissolved mainly as a neutral divalent cation complex, CrCl(OH)0, in the solution, but incorporated as a trivalent cation in chromite, the required redox reaction can involve concurrent precipitation of sulfide with chromite. This mechanism differs from some recent models in that the anorthosites are themselves replacement bodies and are not original precipitates from a magma nor formed by loss of mafic material by partial melting. The results show the need for experimental mineral solubility data at T and P conditions appropriate to upper crustal mafic–ultramafic intrusions. 

In their discussion of our recent publication, Scoon and Mitchell (2020) put forward a number of arguments against the hydromagmatic model of Bushveld Complex formation that we present. Their criticisms of our model focus primarily on the formation mechanisms of the discordant bodies present at Bushveld, namely the iron-rich ultramafic pegmatites and the dunite pipes. While this was a minor portion of our paper, we here review evidence in favor of a fluid-related origin for these discordant bodies, in contrast to the primarily magmatic origin that Scoon and Mitchell present. 

Crystallization of the 2.06 Ga Bushveld magma formed a 9 km (maximum) sequence of ultramafic and mafic rocks that generated a large volume of country fluid as it thermally metamorphosed a 3+ km section of previously unaltered underlying sedimentary rocks of the Transvaal sequence – a geometry similar to that seen as subducting lithospheric slabs are heated by overlying mantle rocks. The presence of a diatreme (breccia pipe) and other large, pipe-like features in the Bushveld Complex located proximal to diapiric upwelling of the basement rocks suggest that overpressured fluids generated during dehydration of the footwall sediments are focused by the diapiric structures such that the country fluids rapidly penetrate the Bushveld rock. A re-examination of existing stable and radiogenic isotopic evidence is consistent with contamination of Main Zone magmas by 1–2% country fluid. Numeric modelling of the footwall dehydration similarly shows that most of the country fluids will be confined to pipe-like channels as it percolates into the Bushveld sill. Modelling also suggests that the maximum extent of the metamorphic aureole was reached at about the same time that the Main Zone began to crystallize. It is proposed that rapid inflation of the Bushveld sill induced the sudden and catastrophic expulsion of overpressured country fluids to both generate the diatreme and contaminate the Main Zone magma, resulting in the Main Zone enrichment in crustal stable and radiogenic isotopic signatures (Sr, Nd, O and others). By analogy, it is also suggested that hydration melting in the mantle wedge is episodically driven by similar sudden influxes of slab fluids that are able to retain their geochemical and isotopic character by rapid channelled influx. This can be aided by flow focusing at diapirs structures at the upper slab-mantle contact.