Timescale of pervasive melt migration in the continental crust

Movement of a large volume of granitic melt is an important factor in the compositional differentiation of the continental crust and the presence of melt in rocks profoundly influences their rheology. Different mechanisms controlling melt migration through crust were proposed. We suggest that pervasive melt flow, analogous to reactive porous melt flow in mantle, could be possibly one of them. It is generally accepted that migration of felsic melts in continental crust starts with short distance pervasive microscopic flow into segregation veins which extract melt. However, we show that pervasive melt flow may be a regional mode of melt migration in continental crust. In such scenario, melt driven by deformation passes pervasively along grain boundaries through the whole rock volume. And the term pervasive melt flow is used for grain-scale, diffuse, porous and reactive flow of felsic silicate melt through rocks. This is effectively an open-system process that thoroughly reworks the resident rock mass. Through-flow of melt destroys pre-existing fabrics and the original chemical and isotopic nature of the protolith. Melt segregation is inefficient and protolith become isotropic granite-like, with partly preserved relics of the original, without ever containing more than a few melt percent at any time. The fabric and geochemical nature of these granites encapsulates the complex history of hybridization.

In order to decipher duration of pervasive melt migration we used precise U-Pb monazite ID-TIMS (isotope dilution thermal ionization mass spectrometry) and U-Pb monazite Laser Ablation Split Stream (LASS) geochronology in combination with monazite chemistry as well as U-Pb zircon SHRIMP geochronology. Monazite reveal continuous chemical equilibration with passing melt. They are getting progressively enriched in HREE and depleted in Eu. Monazites in the least affected rock preserve original magmatic zoning in Th and U, in contrast to more with melt equilibrated rock types, where this zoning is lost. Data for each migmatite type reveal  similar date spread for both cores and the Y-rich well defined rims of single monazite grains, indicating a disconnect between U-Pb dates and chemical zoning. There is also no correlation between U-Pb ages and Yb/Gd ratio. This suggest perturbance of the isotopic system. We interpret these random distribution within-grain date variations as a result of dissolution-reprecipitation reactions between monazite grains and melt. During the coupled dissolution-reprecipitation radiogenic Pb was redistributed within the grain. This is supported by dissolution of apatite into silicate melts that stabilizes monazite during migmatitization, preventing their dissolution but not reaction with passing melt. Redistribution of radiogenic Pb resulted in meaningless individual ages from different migmatite types, but gave overall duration of the thermal event – pervasive melt flow. Duration of pervasive melt flow was dated 8-10myr. This suggest that porous flow of silicate melts in continental crust is a process which can operate over a long time and impacts on the rheology of the crust during orogeny.