Albite-spodumene pegmatite are considered to be the product of extreme fractionation of melts or fluids deriving from large alkaline granite intrusions. Anatectic melts deriving from partially molten metasediments are in contrast not believed to be a possible source for such pegmatite. In the Austroalpine Unit of the Eastern Alps, albite-spodumene pegmatite are associated with simple pegmatite and relatively small inhomogeneous leucogranite bodies, all Permian in age. Parent granites were never observed. Instead, field relations, petrography, geochronology as well as phase and whole rock major- and trace-elements geochemistry suggests that this pegmatite bodies and leucogranite derived from anatexis in upper amphibolite facies of Al-rich metapelite. Bulk rock and LA ICPS-MS mineral geochemistry indicate that before melting, metapelite could have contained significant Li (average 120 ppm) and that the main Li-carrier in the protholith was staurolite (with up to 800 ppm Li). The aim of this study is to test with geochemical modelling if melting of such metapelites could be the origin of the albite-spodumene pegmatite.
Using conservative parameters and realistic hypotheses, these models show that 15 to 25 vol% melt containing more than 200 ppm Li can escape the migmatite. Following fractionation of the melt with 99% in mass within simple pegmatite and leucogranite containing 100 ppm Li yields high-evolved melts with 10,000 ppm Li. Our geochemical model shows that partitioning of Li between restite and anatectic melt coeval with breakdown of staurolite, followed by fractionation is a realistic genetic process for the formation of the albite-spodumene pegmatite of the Austroalpine Unit.