Significance of "stretched" mineral inclusions for reconstructing P-T exhumation history

Analysis of mineral inclusions in chemically and physically resistant hosts has proven to be valuable for reconstructing the P-T exhumation history of high-grade metamorphic rocks. The occurrence of cristobalite-bearing inclusions in garnets from Gore Mountain, New York, is unexpected because the peak metamorphic conditions reached are well removed (>600 degrees C too cold) from the stability field of this low-density silica polymorph that typically forms in high temperature volcanic environments. A previous study of samples from this area interpreted polymineralic inclusions consisting of cristobalite, albite and ilmenite as representing crystallized droplets of melt generated during a garnet-in reaction, followed by water loss from the inclusion to explain the reduction in inclusion pressure that drove the transformation of quartz to cristobalite. However, the recent discovery of monomineralic inclusions of cristobalite from the nearby Hooper Mine cannot be explained by this process. For these inclusions, we propose that the volume response to pressure and temperature changes during exhumation to Earth's surface resulted in large tensile stresses within the silica phase that would be sufficient to cause transformation to the low-density (low-pressure) form. Elastic modeling of other common inclusion-host systems suggests that this quartz-to-cristobalite example may not be a unique case. The aluminosilicate polymorph kyanite also has the capacity to retain tensile stresses if exhumed to Earth's surface after being trapped as an inclusion in plagioclase at P-T conditions within the kyanite stability field, with the stresses developed during exhumation sufficient to produce a transformation to andalusite. These results highlight the elastic environment that may arise during exhumation and provide a potential explanation of observed inclusions whose stability fields are well removed from P-T paths followed during exhumation.