CONTROLS ON ZIRCON GROWTH IN MIGMATITES: A PHASE DIAGRAM MODELING APPROACH

With an ability to retain valuable trace element and isotopic information through protracted events at elevated temperatures, zircon is a valuable chronometer for placing constraints on deep crustal processes. Zircon growth has been described in a wide range of metamorphic settings, and is interpreted to date high temperature segments of P-T-t paths. Traditionally, most if not all zircon in partially melted rocks is considered to grow when individual rock compositions cross the solidus. Scatter in zircon ages within and between samples could be considered an effect of local differences in rock composition, and therefore differences in the P-T at which the solidus is crossed. However, detailed studies of migmatites where zircon growth has been tied to major mineral assemblages by trace element partitioning, suggests that zircon may grow during melting and melt extraction.

New phase equilibria modeling work using THERMOCALC (Powell & Holland, 1998) is aiming to understand better the controls on zircon growth from anatectic melts. With the recent addition of zirconium to thermodynamic datasets (Kelsey et al., 2011), it is now possible to integrate zircon saturation models within the framework of an internally consistent dataset to fully evaluate the roles of melt and mineral composition on zircon growth in a more holistic way. Phase diagrams have been calculated to directly assess the role of water on zircon growth under conditions equivalent to those experienced in metasedimentary migmatites. Using the composition of melt generated at granulite facies conditions (P~7kb, T~860°C), T-X_H2Opseudosections were constructed to simulate the effects of reactions between hydrous melt and comparatively anhydrous wall rocks during melt ponding in leucosomes. Results show that loss of water from the melt leads to immediate saturation of zircon, and continued zircon growth is possible without cooling. Multiple “ponding” events lead to additional zircon growth from melt significantly reducing melt-zirconium content. Results are consistent with observations on deep crustal migmatites, and provide a mechanism for trace element depletion in S-type granites, and relative enrichment of the granulite residue.

Kelsey et al. (2011), J. Met. Geol., 29, 151-166. Powell et al. (1998), J. Met. Geol., 16, 577–588.