The role of trace element partitioning between garnet, zircon and orthopyroxene on the interpretation of zircon U–Pb ages: an example from high-grade basement in Calabria (Southern Italy)

Abstract

The recognition of the coeval growth of zircon, orthopyroxene and garnet domains formed during the same metamorphic cycle has been attempted with detailed microanalyses coupled with textural analyses. A coronitic garnet-bearing granulite from the lower crust of Calabria has been considered. U–Pb zircon data and zircon, garnet and orthopyroxene chemistries, at different textural sites, on a thin section of the considered granulite have been used to test possible equilibrium and better constrain the geological significance of the U–Pb ages related to zircon separates from other rocks of the same structural level. The garnet is very rich in REE and is characterised by a decrease in HREE from core to outer core and an increase in the margin. Zircons show core–overgrowth structures showing different chemistries, likely reflecting episodic metamorphic new growth. Zircon grains in matrix, corona around garnet and within the inner rim of garnet, are decidedly poorer in HREE up to Ho than garnet interior. Orthopyroxene in matrix and corona is homogeneously poor in REE. Thus, the outer core of garnet and the analysed zircon grains grew or equilibrated in a REE depleted system due to the former growth of garnet core. Zircon ages ranging from 357 to 333 Ma have been determined in the matrix, whereas ages 327–320 Ma and around 300 Ma have been determined, respectively, on cores and overgrowths of zircons from matrix, corona and inner rim of garnet. The calculated DREEzrn/grt and DREEopx/grt are largely different from the equilibrium values of literature due to strong depletion up to Ho in zircon and orthopyroxene with respect to garnet. On the other hand, the literature data show large variability. In the case study, (1) the D zrn/grt values define positive and linear trends from Gd to Lu as many examples from literature do and the values from Er to Lu approach the experimental results at about 900 °C in the combination zircon dated from 339 to 305 Ma with garnet outer core, and (2) D opx/grt values define positive trends reaching values considered as suggestive of equilibrium from Er to Lu only with respect to the outer core of garnet. The presence of a zircon core dated 320 Ma in the inner rim of garnet suggests that it, as well as those dated at 325–320 Ma in the other textural sites and, probably, those dated at 339–336 Ma showing depletion of HREE, grew after the garnet core, which sequestered a lot of HREE and earlier than the HREE rich margin of garnet. The quite uniform REE contents in orthopyroxene from matrix and corona and the low and uniform contents of HREE in the zircon overgrowths dated at about 300 Ma allow to think that homogenisation occurred during or after the corona formation around this age. The domains dated around 325–320 Ma would approximate the stages of decompression, whereas the metamorphic peak probably occurred earlier than 339 Ma.