Coseismic formation of eclogite facies cataclasite dykes at Yangkou in the Chinese Su‐Lu UHP metamorphic belt

Eclogite facies cataclasite is recognized at Yangkou in the Chinese Su‐Lu ultrahigh‐P metamorphic belt. The cataclasite dykes (5?15 cm wide) are bounded by mylonite/ultramylonite zones, cutting through unfoliated metagabbro and/or eclogite. The cataclasite veins (generally 2–4 cm wide) are free of mylonite boundary zones, cutting through the foliation of the high‐P host rock. The dykes and veins are dominated by eclogite fragments consisting of debris of omphacite, garnet, quartz, phengite and kyanite, in a matrix of variable amounts of a schist rich in quartz, phengite and kyanite. Garnet clasts in the fragments are welded and overgrown by more Ca‐rich garnet containing mineral inclusions different from those in the garnet cores. The micropoikilitic texture of garnet is typical of eclogitic pseudotachylytes. Crack‐sealing K‐feldspar veinlets in the cataclasite dykes also imply frictional or shock‐induced melting of K‐mica. The modal abundances in the cataclasite and the schist imply that the dykes formed by flow of the omphacite and garnet‐dominated cataclasites into the fractures during seismic faulting, while the lower density minerals (quartz, phengite and kyanite) were largely left in the ultramylonite boundary zones. The dykes have the same composition as their host rocks, except for slightly lower Si and large ion lithophile elements and higher Mg, Ca, Cr, Co and Ni. Chromite, probably spurted from the nearby ultramafic rock, is found as rare particles in the cataclasite fragments. This indicates that material exchange occurred by mechanical mixing between the dykes and the ultramafic rock during seismic faulting. The Cr‐rich eclogite minerals grown on the chromite are evidence for coseismic high‐P crystallization. Short‐lived crystal growth is implied by the fine grain sizes of the eclogite minerals and very limited element diffusion between the garnet clasts and their overgrowths. The fact that the host rocks are more hydrated implies that the dyke formation was not related to fluid infiltration. It appears, therefore, that stress was the key factor inducing the high‐P phase transformation in the dykes. Both stress and temperature were only transiently high in the dykes, which have been metastable since they were formed.