Fluids and gold mineralisation in the late Archaean Harare Greenstone Belt, Zimbabwe
Abstract
The study of selected gold deposits (i.e. Venus, Gladstone, Ceylon, Mashona Kop, Graham Nigel, Chigimira and Viking) in the late Archaean Harare Greenstone Belt (IIGB) has revealed the presence of more than one fluid associated with mineralisation. An early fluid was responsible for the first episode of mineralisation that produced an ore mineral assemblage dominated by arsenopyrite. A later cooler, higher salinity fluid remobilised the gold to form higher grade quartz-rich ore bodies often associated with pyrrhotite mineralisation. The early mineralisation occurred in the PT range of 300-520°C at 1-2.4 kbar and the fluid was low salinity (0-11 wt% NaCl eq.). The remobilisation occurred at 200-480°C and 0.6-2.3 kbar and the fluid had a higher salinity in the range of 19-36 wt% NaCl eq. Both fluids were aqueousCO<sub>2</sub>-CH<sub>4</sub>-NaCI-KCI-MgCI<sub>2</sub>-CaCI<sub>2</sub>-bearing.
The presence of C0<sub>2</sub> and chlorides in the mineralising fluids and the close association between gold and sulphide minerals (e.g. arsenopyrite, pyrrhotite, pyrite etc.) makes the anions of carbonate, chloride and sulphur good candidates for forming complexes with the Au" for transportation. Fluid/rock interaction and fluid unmixing were probably the dominant processes resulting in gold deposition. This is implied by the extensive wall-rock alteration, and the occurrence of contemporaneously trapped aqueous and gaseous fluid inclusions in quartz associated with mineralisation. Fluid/rock interaction would have resulted in changes in Eh, pH and oxygen fugacity, resulting in the breakdown of gold complexes, and deposition of the gold. The ubiquitous association of gold and sulphides, suggests that the process of sulphidation was intimately related to gold mineralisation. Fluctuations in fluid pressure during vein propagation by the crack and seal process could also have resulted in the precipitation of gold.
Several common characteristics which include alteration assemblages (biotite-actinolitc-quartz-microcline), metasomatism (gains in K, Rb, Ba, Si, Ca, Fe, Mg, Pb, Zn, Sc, Co, Au, As, Mn and volatiles) and associated volume increase, ore mineral assemblages (arsenopyrite, pyrrhotite, free gold), PT conditions of mineralisation, fluid chemistry (low pH, log fO<sub>2</sub> =-5) and isotope signatures (δ<sup>18</sup>O = 4.4 - 8.0 %), make these deposits a coherent group at greenstone belt scale. The sources of fluids indicated for these deposits are metamorphic devolatilisation and magmatic. However, the consistent chemistry of the fluids and the conditions at which mineralisation occurred suggest that the system that produced these deposits was homogeneous at greenstone belt scale, with all deposits having been formed by similar processes with fluids coming from both possible sources.
The study of the relationship between the Chinyika Tonalite (CT) and the late Harare Greenstone Belt (HGB) and what role it may have played in the deformation, metamorphism and mineralisation highlighted several points. Firstly, it showed the emplacement of the CT into the HGB to be syn-tectonic, as evidenced by the presence of a shear zone at their contact, a metamorphic aureole in the metabasalts around the CT and basaltic xenoliths in the CT near the contact. Secondly, the age of mineralisation of 2687±14.6 Ma as derived from Ar/Ar geochronology on hydrothermal muscovite is slightly earlier than the emplacement o f the Chinyika Tonalite whose inferred age is 2.66 Ga. This is consistent with a phase of deformation and metamorphism, and mineralisation of the HGB predating the emplacement of the Wedza-type TTG’s. If the age of (2832±2 Ma) obtained for zircons from the Venus Shear Zone (VSZ) felsite is taken as its true age then it can not be related to the emplacement of the CT as previously thought, and its juxtaposition with the Arcturus Formation metabasalt is testimony to the complexity of the stratigraphy of the HGB. Detailed structural and age data are required to better understand the timing relations between the CT and, the HGB and its mineralised shear zones.