Rare peak and ubiquitous post-peak zircon in eclogite: Constraints for the timing of UHP and HP metamorphism in Erzgebirge, Germany

Abstract

Erzgebirge eclogite occurs in three high-pressure (HP) units, including one UHP unit (Unit 1). Each sample investigated contains either magmatic or metamorphic zircon but never both. This is in contrast to many literature examples of eclogite with more than one generation of zircon. We infer that metamorphic zircon formed at the expense of a magmatic precursor, and that the process either went to completion (consuming all magmatic zircon) or did not occur at all. Metamorphic zircon is only present in UHP eclogite and one HP eclogite from Unit 2. Another sample from Unit 2 and one from Unit 3, which experienced the lowest metamorphic PT conditions of all samples, contain only magmatic zircon. The U-Pb SIMS age of c. 540 Ma of igneous zircon, being somewhat younger than average Hf model ages (i.e., 567 and 591 Ma), is attributed to crystallization in the basaltic or gabbroic precursor that formed from a juvenile mantle source.The time of peak metamorphism is recorded by only two zircons from UHP eclogite giving ages of 360.5 ± 5.4 Ma and 359.7 ± 5.3 Ma that well agree with a previous three-point Sm-Nd isochron of 360 ± 7 Ma. All other grains gave younger ages, mainly between 340 and 330 Ma but covering a range of c. 350–315 Ma, which is attributed to retrogression at amphibolite-facies conditions. We infer (i) that this post-eclogitic zircon replaced peak-metamorphic grains (only in UHP eclogite) as well as magmatic zircon and (ii) that peak-metamorphic zircon did not form in non-UHP eclogite. Growth of peak and post-peak zircon was governed by fluid and/or melt. The latter is related to decompressional melting in felsic host rocks (and possibly also to late-orogenic granitic magmatism). Fluid is either of external origin (melt-derived or from dehydration reactions in subducting footwall units) or due to decompressional release of internal fluid (molecular water from fluid inclusions or structural water in omphacite and garnet). The new and former age data can best be reconciled with a two-stage exhumation process. Initial, buoyancy-driven exhumation was fast and terminated as the HP/UHP slices reached lower crustal levels where the rocks presumably rested for 10–20 million years, followed by slow exhumation associated with gravitational collapse of the Variscan orogen. The majority of metamorphic zircon (with ages of 340–330 Ma) grew at lower crustal levels, where the rocks presumably rested for 10–20 million years, and other grains (younger than 330 Ma) during the final, slow exhumation that partly overlapped with the time of granitic magmatism.