University of Leicester
Smith_etal_2014_MinDep.pdf (5.04 MB)

Precious and base metal geochemistry and mineralogy of the Grasvally Norite-Pyroxenite-Anorthosite (GNPA) member, northern Bushveld Complex, South Africa : implications for a multistage emplacement

Download (5.04 MB)
journal contribution
posted on 2014-07-31, 13:51 authored by J. W. Smith, David A. Holwell, I. McDonald
The Grasvally Norite-Pyroxenite-Anorthosite (GNPA) member within the northern limb of the Bushveld Complex is a mineralized, layered package of mafic cumulates developed to the south of the town of Mokopane, at a similar stratigraphic position to the Platreef. The concentration of platinum-group elements (PGE) in base metal sulfides (BMS) has been determined by laser ablation inductively coupled plasma-mass spectrometry. These data, coupled with whole-rock PGE concentrations and a detailed account of the platinum-group mineralogy (PGM), provide an insight into the distribution of PGE and chalcophile elements within the GNPA member, during both primary magmatic and secondary hydrothermal alteration processes. Within the most unaltered sulfides (containing pyrrhotite, pentlandite, and chalcopyrite only), the majority of IPGE, Rh, and some Pd occur in solid solution within pyrrhotite and pentlandite, with an associated Pt-As and Pd-Bi-Te dominated PGM assemblage. These observations in conjunction with the presence of good correlations between all bulk PGE and base metals throughout the GNPA member indicate the presence and subsequent fractionation of a single PGE-rich sulfide liquid, which has not been significantly altered. In places, the primary sulfides have been replaced to varying degrees by a low-temperature assemblage of pyrite, millerite, and chalcopyrite. These sulfides are associated with a PGM assemblage characterized by the presence of Pd antimonides and Pd arsenides, which are indicative of hydrothermal assemblages. The presence of appreciable quantities of IPGE, Pd and Rh within pyrite, and, to a lesser, extent millerite suggests these phases directly inherited PGE contents from the pyrrhotite and pentlandite that they replaced. The replacement of both the sulfides and PGM occurred in situ, thus preserving the originally strong spatial association between PGM and BMS, but altering the mineralogy. Precious metal geochemistry indicates that fluid redistribution of PGE is minimal with only Pd, Au, and Cu being partially remobilized and decoupled from BMS. This is also indicated by the lower concentrations of Pd evident in both pyrite and millerite compared with the pentlandite being replaced. The observations that the GNPA member was mineralized prior to intrusion of the Main Zone and that there was no local footwall control over the development of sulfide mineralization are inconsistent with genetic models involving the in situ development of a sulfide liquid through either depletion of an overlying magma column or in situ contamination of crustal S. We therefore believe that our observations are more compatible with a multistage emplacement model, where preformed PGE-rich sulfides were emplaced into the GNPA member. Such a model explains the development and distribution of a single sulfide liquid throughout the entire 400-800 m thick succession. It is therefore envisaged that the GNPA member formed in a similar manner to its nearest analogue the Platreef. Notable differences however in PGE tenors indicate that the ore-forming process may have differed slightly within the staging chambers that supplied the Platreef and GNPA member.



Mineralium Deposita, 2014, 49, pp. 667–692

Author affiliation

/Organisation/COLLEGE OF SCIENCE AND ENGINEERING/Department of Geology


  • VoR (Version of Record)

Published in

Mineralium Deposita


Springer Link





Copyright date


Available date


Publisher version



Usage metrics

    University of Leicester Publications