TESTING SOLUTION MODELS IN PHASE EQUILIBRIA MODELING

Although phase equilibria modeling is routinely used in metamorphic petrology, few studies have compared the results of using different combinations of various solid solutions for mineral phases and melt solution models. In this contribution, four different sets of solution models are compared through phase equilibria modeling with the software Perple_X while attempting to replicate the results of six partial-melting experiments of metasedimentary rocks. Two of these sets comprise standard packages of solution models that are provided with two internally-consistent thermodynamic datasets, whereas the other two comprise a mix of solution models compatible with one of these thermodynamic datasets. Parameters of comparison include the stability field, modal amount and composition of the main mineral/melt phases, as well as whether commonly used compositional isopleths cross in the calculated stability field that correspond to the mineral paragenesis observed in the experiments. While it was found that none of the tested sets perfectly replicate the experimental results for any of the tested parameters, one set does systematically yield better results. The most significant difference of this set relative to the others is the use of a biotite solution model in which Ti orders preferentially onto the M2 rather than onto the M1 octahedral site. In order to confirm that the preferred set is applicable to natural systems, we present examples of pelitic schists metamorphosed at supra- and subsolidus conditions that have been successfully modeled with it.