Constraining the timing of metamorphism: U-Pb and Sm-Nd ages from a transect across the Northern Torngat Orogen, Labrador, Canada

The Paleoproterozoic Torngat Orogen of northern Labrador, a segment of the Trans-Hudson Orogen, records the consequences of the convergence and collision of the Rae and Nain cratons and the intervening Burwell Domain. To better characterize the ages of the main lithologies and the timing of metamorphism, multiple samples were collected at sites along a transect across northern Torngat Orogen and analyzed for U-Pb and Sm-Nd. The results support previous models in which the Burwell Domain is an eastern extension of the Nain Province proper and experienced protracted exhumation between 1877 and 1800 Ma above an east-dipping Abloviak Shear Zone. The steep Kormaktorvik Shear Zone to the east, which separates the Nain Province proper and the Burwell Domain, facilitated late, rapid exhumation of the Nain Province. Ages from multiple samples, phases, and localities indicate that metamorphism traversed similar PT paths diachronously from west to east and at different rates. The data also demonstrate that metamorphic zircon forms by composition-dependent reactions over a range of temperature conditions, including below the closure temperature of zircon. This affirms that in deeply exposed orogenic belts, metamorphic zircon ages do not inherently represent the time of peak or near-peak temperatures, and that a range of ages across a region cannot be assumed to signal diachronous peak conditions. Furthermore, rocks may experience granulite-facies metamorphism without forming metamorphic zircon. Once formed, zircon may survive subsequent granulite-grade metamorphism and preclude formation of zircon during later thermal events. Sm-Nd analyses of minerals from units precisely dated by U-Pb geochronology show that Nd closes in different minerals over a range of temperatures. In particular, Nd in garnet, monazite, and apatite closes at lower temperatures than in pyroxenes (and perhaps titanite). The use of garnet to derive Sm-Nd isochrons will, therefore, generally derive an age younger than the true age of mineral formation and peak conditions. Exceptions are predicted in rapidly cooled domains (where all phases close within the time resolution of the chronometer) and in rocks without a second, rare earth element-bearing phase that closes at low temperature.