Use and abuse of REE spectra to constrain the eclogite-facies event
Résumé
In the absence of other indicators, the link between the growth of a specific zircon crystal (or part of it) and the eclogite-facies part of the P–T evolution of a rock is commonly validated using the analysis of its REE signature (e.g. Rubatto, 2002). The absence of an Eu anomaly and a flat HREE pattern are considered to reflect a growth in a plagioclase-free garnet-bearing assemblage, typical for the eclogite facies. The ages obtained using this approach can reveal, however, to be in conflict with other geological evidence. Using two eclogite samples from the south of the French Massif Central, we argue that this approach should be used with caution. (1) Zircon grains from an eclogite embedded in massive amphibolite in the Najac-Carmaux Klippe yield a well-constrained U/Pb age of 386.0 ± 1.2 Ma (MSWD = 1.5) consistent with the Lu/Hf age of 382.8±1.0 Ma (MSWD = 0.61) and the Sm/Nd age of 377.8 ± 1.1 Ma (MSWD = 1.4) obtained on garnet. Yet, chrondrite-normalized REE spectra of the dated zircon crystals display variably negative, positive or no Eu anomaly, and their HREE patterns are either enriched or flat. (2) U/Pb zircon data from an eclogite wrapped in a cordierite-bearing migmatite in the Montagne Noire Dome, yield two ages at 339.9 ± 3.4 Ma (MSWD = 0.093) and 315.3 ± 1.4 Ma (MSWD =0.13). With the exception of one analysis, all the REE spectra display neither Eu anomaly nor HREE enrichment, which should suggest that the zircon grains crystallized during the HP metamorphism. The younger age is identical, however, to the ages found in the LP-HT cordierite-bearing migmatite host (320-300 Ma, Poujol et al., 2017). We question therefore the link between the U/Pb age and REE patterns in zircon and suggest that they can be disconnected: “magmatic” spectra (strong negative Eu anomaly and HREE enrichment) can correspond to eclogitic zircon and conversely “eclogitic” spectra can be found in zircon that (re)crystallized after the HP event. Resetting of the U/Pb system does not necessarily go hand in hand with the resetting of the REE system. We propose that REE spectra in zircon may fossilize the REE budget of the rock at the time of the zircon growth: REE patterns in zircon reflect the successive crystallization of the different mineral assemblages in the rock. A U/Pb date, even if obtained on zircon grains yielding an “eclogite-like” REE spectrum cannot be automatically interpreted as the age of the eclogite-facies metamorphic stage, unless its significance is supported by other methods. Poujol, M., Pitra, P., Van den Driessche, J., Tartèse, R., Ruffet, G., Paquette, J.L., Poilvet, J.C. (2017) Two-stage partial melting during the Variscan extensional tectonics (Montagne Noire, France). IJES 106: 477-500. Rubatto, D. (2002) Zircon trace element geochemistry: partitioning with garnet and the link between U–Pb ages and metamorphism. Chemical geology 184:123-38.