Zircon megacrysts from Devonian kimberlites of the Azov Domain, Eastern part of the Ukrainian Shield: Implications for the origin and evolution of kimberlite melts

Abstract

Zircon megacrysts are commonly found in kimberlites and, together with olivine, low-Cr garnet, pyroxene, phlogopite, and ilmenite megacrysts, they constitute a mineral assemblage known as the “low-Cr suite”. The generally close similarity of ages and similar isotope geochemical characteristics of megacrysts and matrix minerals in the host kimberlites support a cognate origin. However, alteration rims commonly develop on zircon and ilmenite megacrysts, providing evidence for a lack of chemical equilibrium between the megacrysts and kimberlitic melts. Here, we report results of a detailed geochronological and geochemical study of zircon megacrysts found in the Middle Devonian Novolaspa kimberlite pipe and dyke located in the Azov Domain of the Ukrainian Shield. The concordia age of zircons is 397.0 ± 2.0 Ma, and it is 14 m.y. older than the age of kimberlite emplacement as defined by a Rb-Sr isochron on phlogopite. The average εHf(397) value for unaltered zircon megacrysts is 6.8 ± 0.14, with the alteration rims having similar Hf isotope systematics. These hafnium isotope data indicate a moderately depleted mantle source for zircon. Unaltered megacrystic zircons have low abundances of trace elements and fractionated REE, with pronounced positive Ce/Ce* anomalies and almost no Eu/Eu* anomalies. In contrast, alteration rims have very high and variable concentrations of trace elements, indicating a reaction between zircon and kimberlite melt. The melt or fluid responsible for zircon and ilmenite megacryst formation, in contrast to kimberlitic melt, was poor in incompatible trace elements, except for the HFSE (Zr, Hf, Nb, Ta, and Ti). The oxygen fugacity during crystallization of the megacryst suite was close to the FMQ buffer.Azov zircon megacrysts do not demonstrate close geochronological and isotope-geochemical similarities with their host kimberlites. They are cognate in the broad sense of being related to the same plume event, but their direct affinity is not clearly defined. The megacryst suite may have crystallized from the earliest melts/fluids that separated from the ascending mantle plume, whereas kimberlite magmas were emplaced 14 m.y. after this event.