Abstract
Serpentinization has been widely documented and investigated at mid-ocean ridges (MOR) and subduction zones. In contrast, at magma-poor rifted margins serpentinization has received much less attention, despite its importance in controlling rheology and mass fluxes during breakup and establishing of a steady-state MOR. In this study, we present new petrological and geochemical data on subcontinental exhumed serpentinized peridotites from the spectacularly exposed Platta, Tasna and Totalp nappes in the Eastern Central Alps in SE Switzerland, belonging to the Alpine Tethys Ocean Continent Transition (OCT). The results testify of a complex history of fluid–rock interactions recorded by several serpentinization events starting with lizardite mesh and bastite textures (S1), subsequently followed by a succession of serpentine-filling veins with distinct textures and serpentine polysomes that include spherical polyhedral serpentine (S2); chrysotile ± polygonal ± lizardite banded veins (S3); lamellar antigorite veins and patches (S4) and chrysotile crack-seal (S5). The serpentinization sequence differs at proximal (i.e. continentwards) and distal (i.e. oceanwards) domains of the OCT. At proximal domains of the OCT (Upper Platta, Tasna) serpentinites record the complete serpentinization sequence (S1 to S5), whereas at distal domains (Lower Platta) serpentinization is restricted to pseudomorphic mesh and bastite (S1) and chrysotile crack-seal (S5). We attribute this discrepancy to contrasted mechanisms of mantle exhumation along the OCT. While at proximal domains mantle is unroofed along continuous and single large offset detachment faults allowing for the formation of all serpentine generations, mantle exhumation at distal domains is a more discontinuous process, controlled by sequential out-of-sequence detachment and flip-flop faults preventing the full development of all serpentine generations. In this frame, the nature and order of formation of the serpentine polysomes are directly controlled by the conditions of serpentinization (i.e. temperature, mantle composition and fluid/rock ratio). We propose that this new conceptual model can be extrapolated to serpentinization at slow to ultra-slow MORs, where close similarities in the serpentinization sequences have been recently reported.