Origin of platinum deposits in layered intrusions by in situ crystallization: Evidence from undercutting Merensky Reef of the Bushveld Complex

Abstract

A central issue for understanding platinum group element (PGE) deposits in layered intrusions is whether they formed through gravitational settling or in situ deposition of sulphide-bearing cumulates on the chamber floor. A unique feature pertinent to this issue is the undercutting Merensky Reef (MR) of the Bushveld Complex, which is commonly associated with potholes, roughly circular depressions in which footwall rocks are removed by magmatic erosion. The undercutting MR forms sill-like apophyses of medium- to coarse-grained harzburgite and orthopyroxenite enriched in sulphide and chromite, which extend laterally from pothole margins into the footwall. They vary in thickness from 5 cm to 1-2m and can be traced away from pothole margins for distances from a few centimetres to dozens of metres (in one case up to 300 m). Locally there may be several (six or more) apophyses of undercutting MR that are vertically stacked within the footwall cumulates in the vicinity of a single pothole. The most telling field, textural and geochemical features of the undercutting MR are as follows: (1) there is no evidence of deformation of igneous layering or rotation of xenoliths associated with these reefs; (2) thin reaction-type selvages of almost pure anorthosite after host leuconorite or mottled anorthosite are commonly developed along both margins of undercutting MR; (3) undercutting MR are mineralogically zoned; thin seams of chromitite are commonly developed along both margins and pass inwards into rocks with decreasing amounts of disseminated chromite; sulphides can be most abundant at the base, along both margins or in the centre of the apophyses; (4) undercutting MR are also compositionally zoned, with rocks displaying either an increase or decrease in whole-rock MgO content from the margins inwards; (5) undercutting MR are highly enriched in PGE, which are mostly controlled by sulphides and chromite, with the grade and tenor being comparable with those of the normal MR. A key inference is that the development of the undercutting MR within footwall rocks, in some cases many metres below a temporary chamber floor, completely eliminates their origin by gravitational settling from the overlying magma. In situ crystallization appears to be the only possible mechanism through which these zoned reefs could be produced. The undercutting MR is interpreted to develop along particularly amenable horizons at pothole margins through thermal/chemical erosion of the footwall rocks by superheated magma, followed by in situ crystallization within the resulting cavities. The cavities behaved as open systems where convection continuously supplied fresh magma, which crystallized from the margins inwards, producing mineralized cumulate rocks. The process may locally be accompanied by redeposition of crystals and sulphide droplets that originally grew in situ, but were torn loose and transported by flowing magma. The highly mineralized undercutting MR of the Bushveld Complex is dramatic confirmation of the hypothesis that high concentrations of PGE in stratiform deposits within layered intrusions can be obtained through in situ crystallization, directly on crystal-liquid interfaces, without processes of gravitational settling.