Jibamitra Ganguly
University of Arizona, USA
Symposium Live
University of Arizona, USA
Aside from being an important subject of inquiry by itself, the thermal history of metamorphic rocks provides quantitative constraints on their tectonic histories. Mineral ages of rocks determined by different decay systems coupled with their respective closure temperatures (Tc-s) have been routinely used to determine the cooling rates of metamorphic rocks. However, the Tc of a mineral with respect to a specific decay system such as, for example, Tc of garnet with respect to the Sm-Nd system, depends on several factors, including the cooling rate itself, that can only be ascertained from appropriate theoretical formulations relating Tc to different factors that affect it; this task, however, requires availability of robust diffusion kinetic data of the decay systems in the minerals of interest. A complication arises in the interpretation of a mineral age, such as a Lu-Hf age of garnet (based on garnet-whole rock (Grt-WR) “isochron”), in which the daughter nuclide has higher Tc than the parent nuclide owing to its relatively slower diffusivity in the mineral. An apparently puzzling observation with respect to a garnet age in a metapelite is that its Lu-Hf age is always older than the Sm-Nd age determined using the same aliquot. There is some indication that in some metabasic rock, the situation may be the reverse.
In this talk, I will review (a) the theoretical advances in the formulation of closure temperature of a mineral as a function of different factors that affect it, based on the diffusion kinetic data collected in our laboratory; (b) discuss some of the remaining problems that need to be addressed; (c) present closure temperature calculations of the Sm-Nd and Lu-Hf systems in garnet; and (d) the results of numerical simulation of the evolution of these two systems in garnet surrounded by biotite (a situation that typically holds for metapelites) or clinopyroxene (that may hold in some metabasic rocks). The simulations show that Lu-Hf Grt-WR ages are sometimes spurious in that they do not relate to any specific marker event during the thermal evolution of a rock such as nucleation of garnet, end of garnet growth, attainment of peak temperature, closure of Hf or Lu during cooling. The Sm-Nd age of garnets, on the other hand, usually records the cooling age (i.e. time lapse since the TC (Nd) that is effectively the same as TC(Sm) owing to the very similar diffusivities of the two nuclides in garnet). Also, as will be shown by specific “real world” examples, numerical modeling of the discrepant Sm-Nd and Lu-Hf ages of garnet, along with some additional constraints, could enable retrieval of the full thermal history of the host rock if it was subjected to a single thermal pulse.
Acknowledgments: This talk is based on many years collaborative research with Max Tirone, Eli Bloch and Rick Hervig.