CO2-rich melts (carbonatite) imprint a HIMU isotopic signature in the shallow mantle lithosphere

Monday, 15 December 2014
James Scott1, Tod E Waight2, Michael Palin1, Quinten Van der Meer2 and Carsten Munker3, (1)University of Otago, Dunedin, New Zealand, (2)University of Copenhagen, Copenhagen K, Denmark, (3)Department of Geology and Mineralogy, University of Köln, Germany, Köln, Germany
Evidence for metasomatic enrichment of the shallow lithospheric mantle by CO2-rich melts (carbonatites) has been documented in peridotite xenoliths entrained in alkali basalts all over the world. The strongest chemical and isotopic imprint of a carbonatite melt is preserved in peridotites that were formerly very depleted (i.e., harzburgite or dunite) because there is little or no pre-existing clinopyroxene (and its trace element budget) to dilute the carbonatite metasomatic signature. Clinopyroxene in many spinel peridotite xenoliths from New Zealand are LREE-enriched, HREE-depleted, have very low K and Ti/Eu, and high Th/U. Several of these properties are indicative of reaction between depleted peridotite and carbonatite during devolatisation reaction: carbonatite + orthopyroxene = clinopyroxene + olivine + CO2. Isotopic analysis of clinopyroxene grains in affected peridotites, particularly of those that were already extremely depleted, indicates that the carbonatite melts have generated isotopic compositions in lithospheric mantle peridotite of 87Sr/86Sr ~ 0.703, eNd ~ +5, 206Pb/204Pb ~ 20 and eHf ~ +5. The compositions generated by these CO2-rich melts is very close to the values commonly quoted for the HIMU mantle component inferred to exist in the source of many ocean island and continental intraplate basalts. Subsequent low-degree partial melting of the carbonatite-refertilised spinel facies peridotitic lithosphere could therefore generate a basaltic magma low in HREE and K, high in LREE, and with an apparent HIMU-like source composition. The lost CO2 should migrate upwards towards the crust. CO2-rich melts in the mantle may therefore play a role in the long-term geological carbon cycle. Scott et al. (2014, Geochemistry, Geophysics, Geosystems).