Evolution and origin of the Miocene intraplate basalts on the Aleppo Plateau, NW Syria

Abstract

The source of intraplate basalts has long been a controversial topic, particularly in continental settings where ambiguity increases because both crustal contamination and crystal fractionation may mask important source characteristics. We present geochemical data to constrain the source and the chemical evolution of the continental intraplate magmas from the Aleppo Plateau and vicinity, NW Syria. New 40Ar/39Ar ages, coupled with published 40Ar/39Ar and K–Ar ages, reveal two discrete Miocene volcanic phases, ~19–18Ma (Phase 1) and ~13.5–12Ma (Phase 2), in the studied area. New chemical and isotopic compositions [87Sr/86Sr=0.7036–0.7051, εNd=+4.5 to +1.1 and (187Os/188Os)t=0.151–0.453] of the lavas reflect the unequivocal influence of crustal assimilation and fractional crystallisation (AFC). Despite the effects of the AFC processes, there still appear to be some differences between the most-primitive, least contaminated magmas of the two volcanic phases, interpreted as a result of source heterogeneity. Whereas the Phase 1 lavas, with relatively high Si, low Ti and trace-element contents, are consistent with partial melting of a largely peridotitic mantle source, the origin of the Phase 2 lavas is more complicated. The latter are characterised by a source component depleted in Si and enriched in Ti, Fe, Ca, P, alkalis, light and middle rare earth elements (REEs) relative to heavy REEs and with sub-chondritic Th–(U)/Nb, Pb/Ce and Zr/Sm. They approach compositions of experimental melts of amphibole-rich metasomatic veins. The compositional variations among the most primitive Phase 2 lavas are difficult to reconcile with varying degrees of partial melting of either the metasomatic veins or peridotite, but could be explained if partial melts of both lithologies were variably mixed, a scenario that could be sensibly envisioned as ascending (peridotitic) plume/asthenosphere derived melts assimilating highly fusible metasomatic veins during their traverse through the lithosphere. This process can be loosely quantified by trace-element forward partial melting modelling that suggests mixing of up to 80% metasomatic melts derived from ~40% melting of amphibole-rich metasomatic veins (which themselves were inevitably compositionally and mineralogically heterogeneous) with 20% plume/asthenospheric melts derived from ~7% melting of a garnet peridotite. Within the compressional framework of northern Arabia, invocation of diapiric material reasonably accounts for the generation of the intraplate basalts in Syria. Derivation of the Phase 2 hybrid melts was probably triggered by lateral flow of this diapiric material beneath the lithosphere subsequent to its arrival, with the migrating flow-front controlling the locus of volcanism. The increase in degree of Si-undersaturation with time for the Phase 1 and Phase 2 lavas is best explained by decreasing temperatures of this flow-front that resulted in less melt contribution from the diapiric mantle while the amphibole-rich veins within the lithosphere continued to be easily fusible, although we cannot totally exclude the possibility that the Phase 2 volcanism tapped a vein-richer domain which formed subsequent to the Phase 1 volcanism.