Sm-Nd chronology of porphyroblastic garnets from granulite facies metabasic rocks in Calabria (Southern Italy): inferences for preserved isotopic memory and resetting

Abstract

Metabasic rocks related to pre-Cambrian protoliths from the lower portion of the deep crust of the Serre (Calabria, southern Italy) contain porphyroblastic garnet up to 5–6 cm in diameter. Garnet forms coronas around the inclusions of clinopyroxene and is in contact with various matrix minerals. Both inner and outer coronas formed under granulite facies conditions after the thermal peak during the Hercynian reworking. Six porphyroblastic garnets (≥1 cm in diameter) from four samples have been dated with the Sm-Nd method to constrain the distinct metamorphic stages and, possibly, to investigate the diffusion of Sm and Nd in garnet. They show in the core major element flat profiles whereas one of these, analyzed for REEs, preserves only a feeble zoning. This suggests that the diffusion rates of REEs are effective at the crystal scale. The apparent Nd ages range from 354 to 88 Ma, without any reproducibility in each and in all rock samples. The oldest age of 354 Ma is interpreted as the primary isotopic signatures linked to prograde metamorphism. The interpretation of younger ages (309, 272, 215, 143 and 88 Ma) requires a detailed discussion about: (i) possible modification of chemical and isotopic composition of the rocks during and after garnet growth, (ii) possible contamination by inclusions in garnet, (iii) inherited isotopic disequilibrium, (iv) new growth or recrystallization of garnet and (v) possible isotopic resetting of large crystals which, in principle, is hampered by the slow diffusion of REE’s in garnet. Some of the Nd ages are similar to U-Pb ages of zircon from the metabasic rocks of deep crustal rocks of the Serre (350, 300 and 280 Ma). This convergence of apparent ages can hardly be considered as simply fortuitous. Thus, since: (i) corona formation was fluid-assisted and (ii) all porphyroblasts were broken up into several fragmented subgrains by sets of fractures resulting in smaller volumes, the volume diffusion and the possible role of high-T fluids on the resetting of Sm-Nd ages are discussed. The calculated ages of 354, 309 and 272 Ma are considered as geologically meaningful and related to the thermal peak and subsequent decompression and cooling stage of the Variscan metamorphism.