sup 176^Lu-^sup 176^Hf geochronology of garnet I: experimental determination of the diffusion kinetics of Lu^sup 3+^ and Hf^sup 4+^ in garnet, closure temperatures and geochronological implications

The ^sup 176^Lu-^sup 176^Hf and ^sup 147^Sm-^sup 143^Nd decay systems are routinely used to determine garnet (Grt)-whole-rock (WR) ages; however, the ^sup 176^Lu-^sup 176^Hf age of garnet is typically older than the ^sup 147^Sm-^sup 143^Nd age determined from the same aliquots. Here we present experimental data for Lu^sup 3+^ and Hf^sup 4+^ diffusion in garnet as functions of temperature, pressure and oxygen fugacity and show that the diffusivity of Hf^sup 4+^ in almandine/spessartine garnet is significantly slower than that of Lu^sup 3+^. The diffusive closure temperature (T ^sub C^) of Hf^sup 4+^ is significantly higher than that of Nd^sup 3+^, and although this property is partly responsible for the observed ^sup 176^Lu-^sup 176^Hf and ^sup 147^Sm-^sup 143^Nd Grt-WR age discrepancies, the difference between the T ^sub C^-s of Lu^sup 3+^ and Hf^sup 4+^ could lead to apparent Grt-WR ^sup 176^Lu-^sup 176^Hf ages that are skewed from the age of Hf^sup 4+^ closure in garnet. In addition, the slow diffusivity of Hf^sup 4+^ indicates that the bulk of metamorphic garnets retain a substantial fraction of prograde radiogenic ^sup 176^Hf throughout peak metamorphic conditions, a phenomenon that further complicates the interpretation of ^sup 176^Lu-^sup 176^Hf garnet ages and invalidates the use of analytical T ^sub C^ expressions. We argue that the diffusion of trivalent rare earth elements in garnet becomes much faster when their concentration level falls below a few hundred ppm, as in the experiments of Tirone et al. (Geochim Cosmochim Acta 69: 2385-2398, 2005 ), and further argue that this low-concentration mechanism is appropriate for modeling the susceptibility of ^sup 147^Sm-^sup 143^Nd garnet ages to diffusive resetting.