Relative deformation behavior of the three Al₂SiO₅ polymorphs

The three polymorphs of Al₂SiO₅ -- andalusite, kyanite, and sillimanite -- are of great importance in metamorphic geology and tectonics because they are indicators of pressure-temperature conditions. When more than one polymorphis in a rock is present, the sequence of crystallization can be used to infer a pressure-temperature path: the trajectory that the rock followed during burial and heating. Textural indicators of sequence of crystallization can be ambiguous, however, because the polymorphs may form my reactions other than polymorphic transformation, and, in cases where one polymorphs is included within another, it can be unclear whether the inclusion nucleated within the host, or whether the host grew around the inclusion. In rocks containing more than one polymorph, the polymorphs commonly display different deformation textures, leading to the possibility that these textures could be used to make inferences about crystallization sequence.

The motivation of the study originated in an investigation of metamorphosed quartzites containing andalusite, kyanite, and sillimanite from a metamorphic terrane in Turkey. The polymorphs display different deformation fabrics, and we wanted to know if we could use these fabrics to infer crystallization sequence. The ultimate goal of the study of this metamorphic sequence is to understand the behavior of subducted continental material, but a first step in that goal is to work out the metamorphic pressure-temperature path in relation to deformation fabrics.

We started with the observation that in a quartzite containing all three polymorphs, the andalusite was most deformed (folded, boudinaged), kyanite was slightly deformed, and sillimanite was slightly to undeformed. Can these textures be interpreted to indicate that andalusite formed first? To investigate this question and the relative mechanical behavior of the Al₂SiO₅ polymorphs during shearing, Andalusite, kyanite, and sillimanite crystals were separated by hand-picking, then powdered and sieved the samples to a uniform grain size of < 15 micrometers. The powders were cold-pressed and then hot-pressed in a stack to produce an Al₂SiO₅ 'rock', which was then sliced into 1 mm thick discs. An earlier hot press result was X-rayed to check that the polymorphs had not transformed into sillimanite at the high temperature of the hot press (1000 C) (they hadn't). A stack of one disc each of andalusite, kyanite, and sillimanite was sheared in a Patterson apparatus with the assistance of Mark Zimmerman and David Kohlstedt's research group, and then sectioned and polished for examination of the textures.