CONSTRAINTS on VAPOR SATURATION IN CUMULATES OF THE UPPER CRITICAL ZONE OF THE BUSHVELD COMPLEX

Controversy over the origins of stratiform PGE deposits in layered intrusions has pitted proponents of orthomagmatic and hydrothermal hypotheses against each other for decades. With appropriate choices of geochemical and physical parameters the hydromagmatic and orthomagmatic models both succeed in reproducing observed ore distributions and compositions. As a result the controversy has not been resolved. The hydromagmatic hypothesis requires that a semi-solid pile of crystal cumulates in a layered intrusion reaches vapor saturation during cooling. The vapors are thought to rise through the cumulates, leaching and transporting PGE and S which are redeposited at a site where vapors are undersaturated. The success of the hydromagmatic models depends on the presence of sufficient trapped liquid and sulfide to generate a significant amount of vapor to move the PGEs upward and also requires that the vapor phase be freely mobile in the partially solidified crystal mush. One envisages a mush zone comprising orthocumulate textured crystals which are slowly compacted by the weight of newly added crystals above. The mush zone would need to be many tens of meters or even hundreds of meters thick to provide the volumes of vapor required to transport the PGE up to the site of the PGE reef. Here I present field observations of the upper Critical Zone of the Bushveld Complex of South Africa, including the UG1, UG2, and Merensky Reef PGE deposits. Based on these observations and model calculations, I suggest that the cumulate pile in the upper Critical Zone matured quickly to a highly annealed texture with minimal proportions of trapped liquid under a mush zone probably only several meters thick. Two problems for the hydromagmatic hypothesis arise from these observations and calculations. First, the volumes of vapor available for PGE transport are severely constrained by the small amounts of trapped liquid inferred to have been present. Second, surface tension between vapor and interstitial melt would tend to bar the upward mobility of vapor bubbles through constricted pores between cumulus mineral grains. It therefore appears unlikely that sufficient trapped liquid persisted beneath the PGE deposits in the Bushveld Complex to have been able to supply the vapors required by the hydromagmatic hypothesis for their genesis.