Brittle initiation of dissolution–precipitation creep in plagioclase-rich rocks: insights from the Bergen arcs, Norway

Abstract

The initiation of ductile shear zones commonly occurs spatially associated with fluid-rock reactions along brittle precursors. In many cases the relative timing of fracturing, fluid infiltration, reaction, and recrystallisation is unclear, making it difficult to disentangle mechanisms of shear zone initiation from subsequent deformation and recrystallisation. Here we present the study of the transition from a dry plagioclase-diopside-garnet-scapolite host granulite-facies lithology to (1) a low strain amphibolite-facies rock, and (2) a transition from low strain to high strain amphibolite-facies lithologies. Hydration of the granulite-facies precursor at amphibolite-facies conditions produces an assemblage comprised dominantly of plagioclase-amphibole-zoisite-clinozoisite-kyanite-scapolite-quartz. Detailed study of plagioclase chemistry and microstructures across these two transitions using Electron Backscatter Diffraction (EBSD) and Wavelength Dispersive Spectrometry (WDS) allows us to assess the degree of coupling between deformation and fluid-rock reaction across the outcrop. Plagioclase behaves dominantly in a brittle manner at the hydration interface and so the initial weakening of the rock is attributed to grain size reduction caused by fracture damage and fluid infiltration at amphibolite-facies conditions. Extensive fracturing-induced grain size reduction locally increases permeability and allows for continuing plagioclase and secondary mineral growth during shear. Based on plagioclase microstructures, such as, an inherited but dispersed crystallographic preferred orientation (CPO), truncation of chemical zoning, and the dominance of fine (5–150 µm), slightly elongate, polygonal grains we conclude that deformation is dominantly facilitated by dissolution–precipitation creep assisted by grain boundary sliding in the shear zone.