LIFETIME OF PHASES VS. LIFETIME OF CRYSTALS – AN UPPER LIMIT TO TIMESCALES ACCESSIBLE BY DIFFUSION CHRONOMETRY

An implicit assumption in diffusion chronometry is that any grain of the studied mineral has been present through the entire history dictated by its chemical stability. However, increasing lines of evidence suggest that this assumption needs to be tested for each specific application. It has been seen, for example, that even if olivine is a stable phase in mafic systems all the way from the mantle to the eruptive products, specific grains of olivine may have had much shorter lifetimes due to processes of recrystallization.

Particularly impressive samples of metapelitic garnets from the Higher Himalayan crystallines in Sikkim provide an opportunity to study this process in detail. Garnet has been stable in these bulk compositions since the crossing of the garnet isograd in the prograde history. They are compositionally homogeneous, excepting at the rims, in terms of major as well as trace elements (e.g. REE). Known durations of metamorphism, as well as the young age of the samples preclude diffusive homogenization as a likely explanation. Polyphase inclusions spread over entire volumes of large garnet grains (e.g. several mm in size) suggest that these were melts that were trapped and hence, that the entire volume of the garnet that is seen today grew in the presence of melts. To test this hypothesis, piston cylinder experiments were carried out in which pieces of these garnets were annealed at pressure and temperature conditions corresponding to the peak metamorphic conditions attained by the rocks (e.g. 10 kbar, 800 °C). All of the polyphase inclusions homogenized to glass, and several homogenized to accessory minerals (e.g. apatite, zircon) + glass. Collectively, these evidence point to massive recrystallization of garnets in the presence of melt, so that no matter how slow the diffusivity of an element (e.g. some REE), records earlier than the partial melting cannot be retained by the garnets.

Thus, it is essential to understand textural evolution resulting from the processes of recrystallization and grain growth in partially molten or deforming systems before diffusion chronometry is carried out. The lifetime of a given grain of a mineral is not necessarily the same as the lifetime of the phase during the evolution of a rock, and the former sets an upper limit to timescales that may be accessible to diffusion chronometry.