UNDERSTANDING PROGRADE METAMORPHIC PATHS BY COMBINING SM-ND AND LU-HF GEOCHRONOLOGY WITH GARNET REE ZONATIONS AND SIZE DISTRIBUTIONS

It has been shown that garnet growth in metamorphic rocks can last for extended periods of time (several m.y.). Hence measured Sm-Nd and Lu-Hf bulk garnet ages may not represent a single point in time, but instead reflect the integrated effects of their growth histories. Distinctly different core-to-rim zoning of these REEs in garnet will lead to Sm-Nd and Lu-Hf isochron ages that record different times along a prograde garnet growth history. Lu-Hf ages are biased towards the onset of garnet growth due to enrichment of Lu in the early grown core (Lapen et al. 2003, EPSL 215). As REE ages are obtained from bulk mineral separates, garnet crystal size distributions (CSD) will also exert an important control on bulk garnet ages. Differently sized garnets are typically interpreted to have grown during a prograde metamorphic event, nucleating at different times. The large garnets nucleated first, the smallest garnets at the end. Small garnets, which only record the late stages of prograde garnet growth, have been found to be volumetrically more abundant than large garnets. Hence bulk garnet ages will always be somewhat biased towards the end of garnet growth, typically the peak metamorphic conditions. By combining the effects of crystal size distributions and the different zoning profiles measured for Lu and Sm in low-grade garnets, that have not experienced significant re-equilibration, estimates of garnet growth time can be obtained. We have combined garnet REE profiles, CSD and possible growth rate laws for an example UHP eclogite of the Zermatt-Saas Fee ophiolite (Lago di Cignana), where Lu-Hf resulted in older ages than Sm-Nd for the same rock. Comparison of our model ages to published Sm-Nd (40.6 ± 2.6 Ma; Amato et al. 1999, EPSL 171) and Lu-Hf (48.8 ± 2.1 Ma; Lapen et al. 2003, EPSL 215) ages have suggested that garnet in this sample grew over 30 ± 10 m.y. (Skora et al. 2009, EPSL 287). Although such a protracted growth interval is surprising, this is supported by plate tectonic reconstructions, suggesting that subduction of the Liguro-Piemont ocean occurred through slow and oblique convergence (Stampfli & Borel 2002, EPSL 196).