WATER CONTENT OF BIOTITE AS MONITOR OF CHANGE IN WATER ACTIVITY

Water activity is an important quantity during metamorphic processes. It can vary due to the absence or presence of a fluid phase, or the composition of the fluid phase though (e.g., additional CO₂, NaCl, etc.). Water activity variations have been invoked to explain many processes such as melting in the crust, granulite formation, equilibrium domains and the kinetics of mineral reactions. It is inferred from low-variant mineral assemblages. Hydrous minerals can, in principle, monitor a_H2O through the exchange reaction 2 OH^-_bt -> O^2-_bt + H₂O_fluid. Here we report SIMS analyses of water content in biotites from two different contact aureoles to test the feasibility of this hypothesis.

In the mid-crustal contact aureole of the western Adamello¹, the OH^- content of biotites in pelitic schists decreases with increasing temperature. Ti content increases, while no correlation with Cl or F was observed. Ti-content varies within individual samples, but no correlation with OH^- was observed. Hence the Ti-oxygen substitution² is not solely responsible for water decrease in biotites. We rather interpret the decrease of OH^- content from 3.41 ± 0.09 p.f.u. for sub-solidus samples to 3.27 ±0.08 p.f.u. for partially molten rocks as an indication of a decrease in water activity due to fluid absent partial melting.

In the Ubehebe Peak contact aureole, OH^- content of phlogopites were measured in two different metacarbonate samples with different fluid composition³, as determined by phase petrology. Water content decreases in biotite with increasing CO₂ in the fluid phase from 2.8-3.8 p.f.u. to 1.5-2.2 p.f.u. Since there is virtually no Fe or Ti in these phlogopites we interpret this clear trend of OH^- content in biotite to monitor water activity during crystallisation.

In conclusion, water content in biotite is a powerful tool to constrain the relative evolution of water activity of a system, but it still need a good experimental calibration of the relation between these two variables to have an absolute value.

¹Floess and Baumgartner (2012) Terra Nova 25, 144-150.

²Cesare et al. (2008) American Mineralogist, 93, 327–338.

³Roselle et al. (1999) American Journal of Science, 299, 93–138.