ESTIMATING ABUNDANCES OF VOLATILE AND OTHER MOBILE COMPONENTS IN EVOLVED SILICIC MELTS THROUGH MINERAL-MELT EQUILIBRIA

Abstract

Silicic igneous rocks (granites, pegmatites, and rhyolites) usually ascribed to A- or S-type sources commonly manifest enrichment in some combination of the rare alkalis and alkaline earths (Li, Rb, Cs, Be, Sr, Ba) and the fluxing components P, F, and B. Because most of these components are incompatible in rock-forming minerals, they remain mobile up to the subsolidus transition and tend to be dispersed into host rocks rather than conserved within the igneous body; hence, the igneous whole rocks do not record the original magmatic abundances of these components. Through mineral–melt equilibria, the abundances of these components can be constrained to variable degrees of accuracy from source (partitioning between residual minerals and anatectic melts), through melt fractionation (partitioning between igneous minerals and residual melts), to the end stages of magma solidification, where higher concentrations of normally trace elements may promote saturation in their crystalline phases. The magmatic abundance of rare alkalis and alkaline earths is controlled largely by reactions among feldspars, micas, and melt. Fluorine in anatectic melts may be buffered at the source by micas (and amphiboles), but F is not usually controlled by other silicate–melt equilibria throughout the remainder of magmatic fractionation. At present, the compositions of micas and of apatite can constrain F contents of melts, though only with some important assumptions. The abundance of B in melt is dictated by the stability of tourmaline with respect to other femic aluminosilicates. Equilibria among tourmaline, biotite, and cordierite (or garnet) in granitic magmas operate at 1–4 wt % B 2 O 3 in melt depending principally on temperature and the activity of Al in melt. The low femic content of evolved B-rich magmas limits the amount of tourmaline that can crystallize; thus, the buffering reactions are readily exhausted and B increases unbuffered in melt. Phosphorous is the best constrained of these fluxing components through (1) equilibria among biotite, garnet, LiAl-silicates and their corresponding phosphate analogs, (2) solubility models for apatite, and (3) the calibration of P distribution between the alkali feldspars and melt. In combination with trapped melt inclusions, the mineral equilibria described here provide useful measures of these petrologically and economically important components in high-silica melts.