Can pyroxenes be liquidus minerals in the kimberlite magma?

Clinopyroxene and orthopyroxene are generally rare in kimberlites, and believed to originate from disintegrated mantle and crustal xenoliths. We report occurrence of inclusions of low-Ca and high-Ca pyroxenes in the olivine phenocrysts in the Udachnaya–East hypabyssal kimberlite (Yakutia, Russia), and make inferences on their relationships to the kimberlite magma. Pyroxenes are only found as either resorbed macrocrysts or small inclusions in olivine; both types are being very rare and volumetrically insignificant. All clinopyroxene and majority of orthopyroxene inclusions are hosted in the olivine cores (Fo86–92) that are compositionally similar to olivine macrocrysts. Major and trace element compositions of clinopyroxene inclusions in olivine phenocrysts and macrocrysts are similar, and resemble compositions of clinopyroxene from lherzolite xenoliths hosted in the Udachnaya–East kimberlite. Their high Na2O and Cr2O3 abundances (0.65–2.55 and 0.6–2.8 wt.%, respectively) suggest deep mantle origin (> 4.5 GPa). The majority of clinopyroxene inclusions have convex-upward trace element patterns that imply kimberlite-like compositions for the hypothetical equilibrium melts. Re-equilibration of clinopyroxene inclusions with the host kimberlite liquid through micro-cracks in olivine is our preferred explanation in this case, however, we cannot exclude their high-pressure crystallisation from the protokimberlite melt. Orthopyroxene inclusions show significant compositional overlap with the low-Al orthopyroxene from the Udachnaya peridotite nodules. Where such inclusions occur in olivine fragments and contact with the kimberlite groundmass, they are strongly resorbed and partially replaced by monticellite. The mantle origin of orthopyroxene and its host olivine and disequilibrium relationships with the kimberlite melt is most likely. Both clinopyroxene and orthopyroxene are completely absent in the kimberlite groundmass. We conclude that the parental melt of the Udachnaya–East kimberlite was not saturated in either clinopyroxene or orthopyroxene at low pressure. This argues against a perceived mafic-ultramafic lineage of the kimberlite primary melt, and provides further support for its essentially carbonate-chloride composition.