• journal_title：Journal of the Geological Society
• Contributor：G. J. GRAY ; S. R. LAWRENCE ; K. KENYON ; C. CORNFORD
• Publisher：Geological Society of London
• Date：1998-
• Format：text/html
• Language：en
• Identifier：10.1144/gsjgs.155.1.0039
• journal_abbrev：Journal of the Geological Society
• issn：0016-7649
• volume：155
• issue：1
• firstpage：39
• section：Article

The nature and origin of the carbonaceous matter (‘carbon’) in the gold-bearing reefs of the Archaean Witwatersrand basin of South Africa has long been a subject of controversy and debate. The importance of the carbon results from its direct association with the gold grade. The studies reported here were directed at resolving the nature and origin of carbon. Techniques used include stope-scale (macroscopic) mapping, mesoscopic (hand specimen) examination, and microscopic studies incorporating the use of optical reflectance measurement. Macroscopic mapping has demonstrated the braided habit of carbon bodies over tens of square metres, with a close parallel to sedimentology. Mesoscopic observations, however, show the vein-like carbon seams are often cross-cutting sedimentary features, with the carbon occupying an anastomosing set of bed-parallel fractures. Microscopic observations show a globular form for the carbon and inclusions of uraninite particles. In seam carbon the globules are generally modified into elongated ‘spindles’ oriented perpendicular to bedding. Optical properties are dominated by generally high reflectance and anisotropy with a ‘swirling’ liquid crystal structure, classifying seam carbon as ‘carbonaceous mesophase’. It is this fine mesophase texture that has previously been wrongly attributed to plant (algal) morphology. Highly significantly, the optical anisotropy is bed perpendicular (rather than bed parallel as in normal sedimentary burial) indicating mesophase formation during fracture opening.

Both reflectance and bireflectance of the seam mesophase show a range of values at all scales. No stratigraphic or depth trends were apparent, though geographic variation points to localization of the maximum temperature event(s). Given high anisotropies, maximum reflectance values in the 4–8% R_o range and minima in the 1–2% R_o range give arithmetic mean reflectance values of about 4.0% R_o. Kinetic modelling equates these values with short term (50 000 years) temperature events of c. 400°C or burial temperatures of c. 280°C. Lower reflectance values of c. 2.0% R_o obtained from the shales suggests thermal detachment of permeable and impermeable beds, a characteristic of pulsed hydrothermal flux. The included uraninite grains influence the orientation of the surrounding anisotropic mesophase demonstrating that uraninite predated the mesophase formation event. Regional palaeo-stress ‘frozen’ into the mesophase anisotropy indicates a NW–SE deviatoric stress.

The combination of hydrocarbon, radiation and hydrothermal fluid flow marks Witwatersrand Basin pyrobitumens out as being different from conventional oil. Putative source rocks for the oil are identified in the intra-basinal Witwatersrand shale units. Lithological control in carbon seams is explained by the poro-perm and mechanical properties of the unconformity sequence of footwall, reef conglomerates and hanging wall quartzites favouring fracture along the unconformity surfaces. An overpressured oxidizing fluid created the fracture network that hosted the oil and also introduced uranium as uranyl ion. On meeting a reducing agent (oil), uraninite precipitated stabilizing the oil by polymerization. Further hotter hydrothermal flux then ‘coked’ the bitumen to often fibrous (spindle-shaped) mesophase. Continued flow deposited gold in the available pore spaces, most notably those formed by lateral contraction between mesophase spindles. The speculative role of methane in the reductive precipitation of gold is indicated by the presence of gold in any co-eval porosity largely independent of the mineral matrix.