EXCESS GARNET GROWTH IN METAMORPHIC ROCKS DRIVEN BY REACTION AFFINITY ASSOCIATED WITH OVERSTEPPED REACTIONS

We investigated the progressive change in reaction affinity during the growth of 12 garnet crystals from 6 samples of medium grade pelitic schist from the Chloride Cliff area in the Funeral Mountains, California, an area of classic Barrovian metamorphism. Isochemical plots show that garnet nucleation occurs at higher temperature than the garnet-in line (11 to 158 °C higher), which we interpret to represent an overstep of the garnet-in reaction. The Gibbs free energy (G) associated with the overstep (reaction affinity) at the conditions of garnet nucleation was calculated using Theriak as the difference in G (system) with and without garnet included as an available phase. For each sample, the difference in G can be attributed entirely to a single garnet growth reaction. Pressure-temperature (PT) paths were calculated using the G-minimization method of Moynihan and Pattison (2013, JMG), which calculates a PT point for each analysis along a garnet traverse and fractionates the predicted garnet grown from the bulk composition before calculating the next point. Each predicted change in garnet volume can be compared to the actual change observed in the crystal. For most of the garnets, the early stage of growth fits well with equilibrium predictions; however, the middle and late stages do not. Throughout the middle stage of growth (~30-80% of the radius), the actual volume grown exceeds the predicted volume, reducing the reaction affinity in the rock. During growth of the last 5-20% of the radius, the actual volume grown is less than the predicted volume, increasing the reaction affinity back to the level calculated for the time of nucleation. We interpret the pattern as (1) garnet nucleation and early growth occurred during a rapid change in PT conditions that drove growth with little change to reaction affinity, (2) growth of the middle portion was partially driven by reaction affinity, possibly due to a decreased rate of change in the PT conditions, and (3) growth during the late stage was slowed by a depletion of available cations surrounding each garnet, possibly due to overlapping depletion halos, resulting in less growth than predicted. Implications include (1) reaction affinity fluctuates during garnet growth, and (2) G-minimization modeling can be used to evaluate changes in reaction affinity during garnet growth.