Sublithosphere mantle crystallization and immiscible sulphide melt segregation in continental basal magmatism: evidence from clinopyroxene megacrysts in the Cenozoic basalts of eastern China

This study explores the effects of high-pressure crystallization and immiscible sulphide melt segregation under mantle conditions on the compositional variation of basaltic magmas, using clinopyroxene megacrysts in the Cenozoic basalts of eastern China. These clinopyroxene megacrysts are large (up to > 10 cm in size) and homogeneous at the grain scale. They were crystallized from variably evolved parental magmas and then captured by their host basalts. The large and systematic variations of [Sm/Yb]N, Lu/Hf, Fe/Mn, Sc/La, Ni and Cu with Mg# in the clinopyroxene megacrysts suggest their co-precipitation with garnet and with immiscibility between sulphide and silicate melts. This is consistent with the appearance of garnet megacrysts in the host basalts and abundant sulphide globules in the clinopyroxene megacrysts. The covariation between Ni contents of sulphide globules and Mg# of the clinopyroxene megacrysts suggests a genetic relationship between sulphide globules and clinopyroxene megacrysts. High-pressure crystallization of clinopyroxene and garnet results in decrease of Mg# and concentrations of CaO, MnO and heavy rare earth elements (e.g., Yb) and increase of Fe/Mn and [Sm/Yb]N in the residual melts. Therefore, geochemical characteristics of low Mg#, low CaO and MnO contents and high Fe/Mn and [Sm/Yb]N in basalts do not necessarily indicate a pyroxenite mantle source. In addition, caution is needed when applying the olivine addition method to infer the primary compositions of alkali basalts without considering the effects of highpressure crystallization of clinopyroxene and garnet. The calculated P-T conditions of the clinopyroxene megacrysts are close to those of the lithosphere-asthenosphere boundary (LAB) beneath eastern China, and the low primitive [Sm/Yb]N (~ 4.0) of melts parental to the clinopyroxene megacrysts suggests final equilibration at relatively low pressures most likely beneath the LAB. Hence, a melt-rich layer is expected close beneath the LAB. Melt pools in this melt-rich layer provide a stable and closed environment for the growth of compositionally homogeneous clinopyroxene megacrysts. As a result, melts in these melt pools are compositionally evolved with low and variable Mg#. Subsequent pulses of melt aggregation/supply from depths with primitive compositions and high Mg# will disturb these melt pools, cause magma mixing and trigger the eruption of magmas carrying clinopyroxene and garnet megacrysts