Hydrogen isotope composition of the Merensky Reef and related rocks, Atok section, Bushveld Complex

Abstract

Hydrogen isotope compositions of rocks from the Bushveld Complex at the Atok mine have been determined to evaluate the relative importance of hydromagmatic and subsolidus hydrothermal processes for defining the geochemical characteristics of the Merensky reef and adjacent rocks. The rocks include mineralized pyroxenite and pegmatoid of the Merensky reef as well as underlying and overlying pyroxenite, norite, and anorthosite cumulates from the Merensky and Bastard cyclic units of the Upper Critical zone. The samples were chosen specifically because of the absence of deuteric alteration phases. The water contents of all rocks, including the pegmatoid, are less than 0.26 wt percent and typically in the range 0.05 to 0.15 wt percent. The major hydrous phase observable optically is phlogopite. However, there is not enough phlogopite in most samples to account for even these low water contents. Additional water is believed to be present either in solution in pyroxene or in submicroscopic phlogopite lamellae within pyroxene.The Upper Critical zone rocks have relatively homogeneous hydrogen isotope compositions, having delta D smow values in the range of -93 to -55 per mil. There is a small but systematic difference in composition between pegmatoid and pyroxenite of the Merensky reef (avg delta D = -65ppm) and of the generally more leucocratic rocks elsewhere in the section (avg delta D = -79ppm). Within each lithology there is no relationship between isotopic composition and water content, and the data cannot be rationalized in terms of simple mixing of light magmatic water with heavier meteoric-crustal water. The narrow range of the D/H ratio of all the rocks suggests that their isotopic signatures were established under similar conditions. This and the fact that different lithologies exhibit different isotopic compositions but possess the same mineralogy (the rocks differ from each other mainly in modal abundances of the major phases) are used to argue that their isotopic compositions could not have been imposed by subsolidus processes. This contention is supported by the composition of Merensky reef phlogopite. The halogen contents of phlogopite do not vary with the crystal-chemical factors known to influence their substitutional behavior. Phlogopite composition was in part controlled by its equilibration with an intergranular fluid of variable composition, implying that there was no significant flow to homoginize pore fluid during subsolidus cooling in the massive rocks. Subsolidus circulation was therefore limited to large, through-going fractures.The geochemical features of the Merensky reef and adjacent rocks at Atok were established during late-stage solidification. The hydrogen isotope data are consistent with a model involving addition of hydrous vapor to the Merensky protoreef horizon. This vapor originated from crystallization of the intercumulus melt below the reef and may also have included a small portion of water from the country rocks.