Metasomatism of the crust-mantle boundary by melts derived from subducted sedimentary carbonates and silicates

Abstract

Substantial quantities of sediments are known to enter the deep lithosphere at subduction zones, but the extent to which sediments melt and the process involved in sediment contribution to the deep lithosphere are inadequately understood. Vigorous debate continues about whether the subducted sediment component is terrigenous or pelagic and transported as a hydrous melt, an aqueous fluid, or bulk sediment. In this contribution, we conduct an integrated study on a variety of deep-seated xenoliths in the Neogene Hannuoba basalts from the northern margin of the North China Craton. Among these xenoliths, clinopyroxenite xenoliths are compositionally and isotopically distinct. Mineral chemistry shows that the clinopyroxenite xenoliths come from a depth near the MOHO, rather than from the mantle as suggested previously. The clinopyroxenite xenoliths have extremely evolved Sr-Nd-Hf isotopic compositions and are interpreted to have a late Archean protolith age. The extremely low contents of Cr and Ni for the clinopyroxenite xenoliths preclude a magmatic origin. Instead, a metasomatic origin is suggested, which is strongly supported for the clinopyroxenites by the occurrence of hydrous minerals and high contents of large-ion-lithophile elements (K, Rb, Ba, Th and Sr) and light rare earth elements, as well as elevated δ18O (9.9–11.3‰) and light δ26Mg (−1.04‰ to −1.42‰) isotopic compositions. Furthermore, their high high-field-strength element (Nb, Ta, Zr and Hf) contents indicate that the metasomatic agent is a hydrous melt, rather than an aqueous fluid. The metasomatic melts are considered to be derived from a mixed source of sedimentary carbonates and ancient, felsic continental materials. A combination of zircon ages and oxygen isotope data for the clinopyroxenite xenoliths further restricts the timing of metasomatism to the late Paleozoic. Considering the regional tectonic setting, the sediments most likely came from the subducted Paleo-Asian oceanic slab. Thus the Hannuoba clinopyroxenite xenoliths provide direct evidence for melting of the subducted Paleo-Asian oceanic slab sediment and its interaction with the deep lithosphere. The data show that melting of subducted sediments can take place at a much shallower depth than commonly thought and place an independent constraint on future models of slab geotherms.