COPPER IN MAGMATIC EFFLUVIA

The processes that concentrate Cu in arc-related ores extend from the sub-crustal realm to Earth's epizone. The interplay between the chalcophilic and halophilic tendencies of Cu is complex, and both melt/sulfide and melt/volatile phase interactions are important in the magmatic geology of Cu.

Sulfides left behind in magma source regions, or removed from magmas by crystallization, can adversely affect ore formation. H-O-C-Cl-S volatile phases concentrate copper toward ore grade in the crust. Wallace (2005) inferred the following volatile concentrations in primitive arc magmas: S and CO2 up to thousands of ppm, Cl ~ 500 - 2000 ppm, H2O from < 1 wt.% to > 8wt%. Studies continue to show that Cl-bearing phases are critical to porphyry Cu genesis (e.g., Tattitch et al., this meeting; Zajacz et al., 2010; Simon et al., 2006; Williams et al., 1995; Candela and Holland, 1984), and S is clearly important in the genesis of these deposits. Vapor (v) phases exsolved from magmas can have mole fractions of CO2 on the order of 0.5 or greater when CO2 (melt) is >800 ppm. The quantity of high-CO2 v exsolved, however, will be limited by the CO2 concentration in the melt, and the exsolution of high-CO2 v will be biased toward the earlier stages of magmatism. Loss of Cl to the CO2-rich v will generally not be large, due to its low v/melt partition coefficient (~1) under these conditions.

Cl/ H2O wt. ratios in the melt can vary from 0.01-0.05; this upper value is the same as the estimate of Candela and Piccoli (1998) for vapor + brine (b) saturation at 800 C and 100 MPa, consistent with Audetat et al. (2008), who infer that, although most porphyry magmas were saturated with a vapor phase only, a minority were saturated with v+b. Data suggest that Cu in the b is chloride-complexed, and hydrosulfo-alkali chloro-complexed in the vapor (Zajacz et al., 2010). Note that for most melt Cl/ H2O ratios, the mass v > mass b (Candela and Piccoli, 1995, 1998). Further, as T and P approach the critical point, the properties of the v and b approach each other, and D's and K's for v/b exchange of any component approach unity (cf., Simon et al., 2006). Hence, the characteristics of v and b are “moving targets” in terms of their composition, and magmatic b is almost always subordinate to v on a mass basis.