Adakite generation as a result of fluid-fluxed melting at normal lower crustal pressures

Abstract

Adakites with continental origin have K-rich compositions distinctive from the sodic, subduction-related adakites and are generally considered as partial melting products of thickened mafic lower crust at high pressure conditions (above 1.5 GPa; >50 km depth). In order to compare the geochemical differences and to constrain their source compositions and partial melting conditions, we compiled published data for the sodic adakites related to subduction processes in the Central Asian Orogenic Belt (CAOB) and the potassic adakite-like rocks related to continental collision in the Lhasa Terranes (LT) of the Tibetan Plateau. Based on the spatial similarity of Sr-Nd isotopes and nature of inherited zircons, the origin of the CAOB sodic adakites is best explained by fluid-fluxed melting of low-K arc basalts, while the LT potassic adakites are best explained as generated by fluid-fluxed melting of rocks similar to the voluminous high-K Linzizong mafic volcanic successions (LVS). The phase equilibria and trace-element modeling reveal that these low Mg# mafic sources would expand garnet stability to pressures as low as 7-8 kbar and fluid-fluxed melting (X(H2O) = 2-3 wt%) would significantly decrease the modal content of plagioclase, resulting in melts with “high pressure” signature (e.g., high Sr/Y and La/Yb). This paper provides an alternative model for the origin of sodic or potassic continental adakites arguing that they originate from fluid-fluxed melting at garnet amphibolite facies at normal lower crustal pressures (10-11 kbar). Our findings highlight the importance of the source composition and fluid content in generating the “high pressure” signatures of adakites, which may also explain the origin of Archean tonalite-trondhjemite-granodiorites (TTGs).