Diffusion Kinetics of Lutetium and Hafnium in Garnet and Clinopyroxene: Experimental Determination and Consequences for ¹⁷⁶LU-¹⁷⁶HF Geochronometry

Abstract

The ¹⁷⁶Lu-¹⁷⁶Hf and ¹⁴⁷Sm-¹⁴³Nd decay systems have been extensively used as geochronological tools to determine ages from garnet (Grt) - whole rock (WR) parent-daughter isotopic ratios; however, the ¹⁷⁶Lu-¹⁷⁶Hf age of garnet is almost always found to be significantly older than the ¹⁴⁷Sm-¹⁴³Nd age determined from the same aliquots. This dissertation presents new experimental diffusion data for Lu and Hf in garnet, and numerical simulations using these data, which explain these age discrepancies and also show that Grt-WR ¹⁷⁶Lu-¹⁷⁶Hf isochrons do not generally yield ages which correspond to an unambiguous temporal event in the evolutionary history of the host rocks. This is a result of (a) partial or complete retention of very slow-diffusing radiogenic ¹⁷⁶HF produced during prograde heating, and (b) the lower closure temperature of Lu relative to Hf; these complexities do not affect the interpretation of Grt-WR isochrons based on the ¹⁴⁷Sm-¹⁴³Nd system. In addition, the diffusion kinetic properties of Hf in clinopyroxene were experimentally determined in order to address the age controversy of the shergottite suite of Martian meteorites (~200 Ma ages determined by ¹⁷⁶Lu-¹⁷⁶Hf, ¹⁴⁷Sm-¹⁴³Nd, and various other decay systems as opposed to a ~4 Ga whole rock Pb-Pb age). This was achieved by calculating the timescales needed to re-equilibrate Hf isotopes in clinopyroxene (the primary host of rare earth elements amongst the minerals used to compose the ¹⁷⁶Lu-¹⁷⁶Hf isochrons) with the surrounding matrix at the peak- and post-shock P-T conditions likely to have been experienced by the shergottites. It is concluded that, contrary to the earlier suggestions, impact heating is highly unlikely to have significantly reset the ¹⁷⁶Lu-¹⁷⁶Hf ages of these Martian samples. These calculations are bolstered by the nature of measured Cr and Fe-Mg concentration profiles across olivine-melt boundaries, and Ti profiles across clinopyroxene-maskelynite interfaces in the shergottite RBT-04262. The lack of any evidence of diffusion in the measured concentration profiles, coupled with the qualitative incompatibility of the measured crystal-melt fractionation at these interfaces with the nature of fractionation expected from equilibrium partitioning, provides strong evidence that no substantial chemical exchange took place between the solid and melt phases during peak-shock P-T conditions.