Selective transfer of Li-Al-rich phyllosilicate to metamorphic veins (Western Alps): Laser Induced Breakdown Spectroscopy (LIBS) compositional profiles and microstructural characterization

[Display omitted] •Cookeite was transferred to veins via a dissolution-transport-precipitation process.•LIBS profiles were performed between cookeite veins to study Li diffusive transfer.•50% initial cookeite remains homogeneously distributed in host-rocks (flat profile).•Efficient Li diffusion through the connected intergranular free fluid phase.•Mass transfer to veins may be interfacial energy driven or stress-induce. In convergent settings, fluid-rock interactions generally result in quartz and calcite preferential transfer to metamorphic veins in classical metamarls, while phyllosilicates tend to remain in the host-rock. However, the mechanisms responsible for such a selective mass transfer are poorly discussed in the literature. Here, we study Alpine metabauxites in which phyllosilicates (Li-Al-rich chlorite called cookeite, followed by pyrophyllite) were preferentially transferred to veins at blueschist peak P-T conditions, by a dissolution-diffusion-precipitation process without any fluid infiltration or associated reaction. Cookeite fibrous en-echelon veins formed by extensional shear, and part of them evolved towards thicker fluid-filled veins with euhedral cookeite crystallization. We performed chemical profiles across host-rocks between successive cookeite veins, using Laser Induced Breakdown Spectroscopy (LIBS), associated to a microstructural study. Flat LIBS Li profiles show that about half of the initial cookeite remains homogeneously distributed in host-rocks, which suggests a minimum diffusion distance of 2–4cm for Li. The availability of an aqueous fluid during most of the metamorphic cycle is demonstrated here. A simple 1D reaction-diffusion model, assuming Li diffusion through a connected fluid-filled porosity network, is able to account for the observed lithium distribution assuming Li diffusion coefficients consistent with literature values in fluid-bearing systems. Chemical potential gradients that drove phyllosilicate transfer to veins can be either interfacial energy driven (i.e., Ostwald ripening), the anhedral phyllosilicate microsheets maintaining high supersaturation levels in the small host-rock pores compared to veins, or stress-induced: phyllosilicates present the highest solubility variations with pressure in the Vanoise bauxites (contrary to quartz-bearing rocks), which may account for their unusual selective transfer to veins. Therefore, mineral solubility variation with pressure seems to be the chief controlling paramete