Microlites of natural and experimental peraluminous pseudotachylytes: a comparison

Pseudotachylyte is a quenched coseismic frictional melt. As such, pseudotachylyte may provide invaluable information on the processes occurring on fault at hypocentre depths. Of particular interest are pseudotachylytes hosted in high-grade rocks, as they may record seismic ruptures propagated in the middle and lower crust. However, pseudotachylyte in high-grade rocks may also result from shallow deformation after uplift, thus constraining ambient conditions of faulting is crucial although not trivial.The mineralogy of pseudotachylyte includes microlites crystallized during melt quenching, glass recrystallization products and, for deep-seated pseudotachylytes, minerals reflecting re-equilibration to the ambient metamorphic conditions. In absence of ductile deformation of pseudotachylyte promoting re-equilibration, the estimate of P–T conditions is typically based on the microlites. For example, the presence of microlitic ‘cauliflower’ garnet has been interpreted to reflect high-grade ambient conditions of faulting. However, Papa et al. (2023), described cauliflower-garnet-bearing pseudotachylytes hosted in granulite facies garnet-sillimanite-rich gneiss from Calabria and proposed shallow faulting conditions based on radiometric dating, suggesting that garnet can be transiently stable during quenching at shallow conditions.Here we quenched at room conditions superheated (>2100 °C) melts produced by instantaneous laser-heating of the same peraluminous gneisses hosting the natural pseudotachylyte and compare the microlite population of the experimental glass with the microlites of the natural pseudotachylyte. Both the natural pseudotachylyte and the experimental glass contain: (i) acicular-shaped corundum microlites; (ii) sillimanite/mullite microlites overgrowing sillimanite clasts; (iii) skeletal-, dendritic-shaped spinel microlites, spatially associated with garnet, epitaxially nucleating on sillimanite/mullite and dispersed in the glass; (iv) microlitic cordierite, present in the natural pseudotachylyte as spherulitic aggregates and in the experimental glass as plumose microlites in melt-filled fractures of the wall-rock garnet; (v) newly formed euhedral rims of garnet epitaxial on garnet clasts and wall-rock garnet. The observed microlites crystallized during melt quenching following the same sequence, with slight differences due to the faster cooling rate of the experiments.By comparing natural pseudotachylytes and experimentally produced analogues, we show that the mineralogy of natural microlites is essentially constituted by high-melting point phases and it is controlled by the local availability of chemical constituents and nucleation seeds (i.e. host-rock clasts). The experiments also prove that garnet can crystallize during quenching even at room conditions if seeds are available and the melt has the right composition. This observation calls for caution when using the mineralogy of pseudotachylytes, and in particular the presence of cauliflower garnet, to infer the depth of faulting. Finally, the melting experiments under static conditions highlight the relevance of thermal fracturing as deformation process aiding pseudotachylyte formation.Papa et al. (2023), Lithos 460, 107375