Paper No. 13
Presentation Time: 11:20 AM
CRYSTALLIZATION OF MAGMATIC SULFIDES
I present an empirical parameterization of the low-pressure melting relations of monosulfide solid solution (mss) (Mungall 2007; Geochimica et Cosmochimica Acta, Vol 71, 2809-2819). The sum of available data reporting compositions of coexisting mss and sulfide liquid have been used to parameterize distribution coefficients for Ni, Cu and S between melt and mss as functions of temperature and sulfur content of the melt phase. The model uses the Fleet and Pan (1994; Geochimica et Cosmochimica Acta Vol. 58, 3369-3377) liquidus surface for mss. Natural sulfide ores generally show trends from Ni-rich and Cu-poor compositions (e.g., 8% Ni, 1% Cu) toward Ni-poor and Cu-rich compositions (e.g., 1% Ni, 34% Cu). This trend has been interpreted in the past as a liquid line of descent of decreasing Ni content due to the fractional removal of Ni-rich mss from an increasingly Cu-rich residual sulfide melt. The current model can be used to demonstrate that this interpretation is impossible to support in light of available experimental data. Ni does not become compatible in mss except in very Ni-rich systems like some komatiite-hosted ores, and at very low temperatures. I argue instead that the trend represents a mixing line between mss and intermediate solid solution (iss) cumulates; all compositions along this trend are therefore cumulate compositions and the residual liquids were much more Ni-rich than previously supposed. The presence of a secondary trend from the composition of pentlandite to that of iss is also now interpreted to represent a mixing line between cumulus pentlandite and iss. The crystallization sequence is thus interpreted to be first mss, then mss + iss. At some point as Ni concentration continues to increase in the melt, pentlandite is in a peritectic relation to mss, causing the crystallizing assemblage to change to iss + pentlandite. The sulfide melts from which these assemblages crystallized are rarely, if ever, preserved as whole rocks. The only candidates for such frozen fractionated sulfide liquids are rare examples of extremely Ni- and Cu-rich veins that commonly form in the footwall of very large deposits like those at Sudbury.