CONCENTRATION OF PLATINUM-GROUP ELEMENTS BY MAGMATIC FLUIDS IN LAYERED INTRUSIONS

Evidence leading to a model which supports the formation of Pt-Pd reefs in a layered intrusion by the action of magmatic fluids is presented, as well as some of the constraints imposed on such a model. The Stillwater Complex is used as an example of this model. Within the Stillwater Complex, no laterally continuous sulfide-enriched cumulates occur above the lowermost parts of the Ultramafic series until one reaches the platiniferous J-M reef. This sulfide-poor stratigraphic interval beneath the J-M reef is characterized by unusually Cl-rich apatite. Although reported sulfide modes and whole-rock sulfur contents in the cumulates below the J-M reef are approximately an order of magnitude lower than that expected for a sulfide-saturated magma; sulfide saturation prior to the formation of the J-M reef is suggested by relatively high Pd and Pt concentrations reported in the cumulates below the J-M reef as well as by the presence of sulfide as inclusions in silicate and oxide cumulus minerals. Hence it is suggested that the footwall cumulates to the J-M reef were originally sulfide saturated but later lost sulfur to a fluid phase exsolving from inter-cumulus liquid. Sporadic sulfide-bearing pegmatoids in some of these rocks are additional evidence for late loss of sulfur. These observations, as well as the association of the J-M reef at the first significant stratigraphic break in the Cl/(Cl + F) ratio of apatite and the relatively high solubility of Pt and Pd as aqueous Cl complexes, are consistent with the formation of the J-M reef by exsolution of Cl- and volatile-rich fluids from footwall intercumulus liquids which leached and transported ore components (both the PGE and S) upward to the level of the reef. Using published whole-rock platinum-group element (PGE) concentrations for the Ultramafic and Lower Banded series, the cumulates below the J-M reef must have been enriched by a factor of, at most, 1.24 for Pt and 1.75 for Pd if these rocks were to be the source for the metals in the J-M reef. Model Raleigh fractionation calculations suggest that the cumulates beneath the J-M reef could have had the required metal concentrations if, during the crystallization of a parent magma containing no more than 15 ppb of either metal, the average bulk solid/liquid distribution coefficients for Pt and Pd were no more than 5; these values are consistent with assumption of sulfide saturation prior to the formation of the J-M reef. During solidification of intercumulus liquid, the separation and loss of a Cl-rich fluid will deplete the residual assemblage of S and PGE. If 0.1 to 0.5 wt percent fluid containing 5 percent S evolved during solidification of the footwall cumulates (the amount of fluid expected to be exsolved by the crystallization of approximately 2 to 10 percent fluid-saturated intercumulus liquid), the fluid/liquid distribution coefficients for Pt and Pd would need to be almost as large as the (sulfide liquid)/(silicate liquid) distribution coefficients to remove significant amounts of Pt and Pd. Much smaller distribution coefficients are needed if sulfide is completely lost to fluid during degassing. The latter possibility is consistent with the low S contents of the footwall cumulates beneath the J-M reef. Numerical models of the degassing of intercumulus liquids suggest that thick cumulus sequences can act as chromatographic columns to separate the PGE and S during the degassing of intercumulus silicate liquids, the PGE being enriched at a sulfide-dissolution front as upward migrating sulfide-undersaturated fluids resorb cumulus sulfides. Zones of PGE and S enrichment can occur where degassing fronts encounter stratigraphic physical-chemical discontinuities such as changes in the fluid saturation temperature of the intercumulus silicate liquids.