Experimental melting of carbonated K-rich garnet harzburgite and origin of kimberlite melts


Bulatov V.K.a, Girnis A.V.b, Brey G.P.c


a Vernadsky Institute of Geochemistry and Analytical Chemistry, Russian Academy of Sciences, ul. Kosygina 19, Moscow, 119991 Russia

b Institute of Geology of Ore Deposits, Petrography, Mineralogy and Geochemistry, Russian Academy of Sciences, Staromonetny 35, Moscow, 119017 Russia

cInstutut für Geowissenschaften, J.-W. Goethe Universität, Altenhöferallee 1, D-60438 Frankfurt am Main, Germany


Experimental melting of carbonated K-rich garnet harzburgite was carried out at 6 and 10 GPa, 11001600oC in a Walker-type multianvil apparatus. The main goal of the experiments was to determine the composition of melt near the solidus of harzburgite with addition of potassium and magnesium carbonates as a proxy for metasomatized mantle peridotite. The starting material was a mixture of natural olivine, orthopyroxene, and garnet with 5% MgCO3 + K2CO3 (~1 wt % K2O). It was found that the near-solidus mineral assemblages of this material could not be reliably established because of the small grains size and low amounts of K-bearing phases. Therefore, in order to clarify possible phase relations near the solidus of K-bearing carbonated peridotite, an experimental series was performed with the SC1 lherzolite (Brey et al., 2008) blended with 10% MgCO3 and 10% K2CO3.

In the both systems, the beginning of melting was detected at ~1100C at 6 GPa and ~1200C at 10 GPa. Only olivine, orthopyroxene, garnet and magnesite were found in the products of harzburgite-MgCO3-K2CO3 experiments. The amount of K2O in the starting mixture (1.4 wt %) was too high to be accommodated in the silicates and carbonates. Therefore, a K-bearing phase (or phases) must be present near the solidus of this mixture. Experiments with the K richer starting material based on the SC1 peridotite showed that two K-bearing phases are stable near the solidus of the lherzolite (SC1)MgCO3K2CO3 system. Phase X was found in the subsolidus experiments at 8 and 10 GPa, and K-Mg carbonate, K2Mg(CO3)2, crystallized at 610 GPa. In the 10 GPa experiments it coexisted with magnesite over a wide temperature range. The stability field of magnesite expands considerably with increasing pressure in the SC1MgCO3K2CO3 system. In the harzburgite system, magnesite is stable up to about 1500C both at 6 and 10 GPa. The melt from the two experimental series are sharply different in K content. Those from the SC1MgCO3K2CO3 system are dominated by K2CO3 and MgCO3 components with only minor CaO content. In contrast, the low-temperature melts from the SC1MgCO3K2CO3 system are dominantly CaMg-carbonate. The main temperature effect is an increase in SiO2 content and a decrease in the Ca/Mg ratio. Similar to experiments in the lherzoliteCO2 system (Brey et al., 2008), the melt composition changes from carbonatitic (<5 wt % SiO2) near the solidus to carbonated silicate with up to 30 wt % SiO2 at 6 GPa 1600C and 10 GPa 1700C. The contents of alkalis and Ca decrease with increasing temperature. The potassium-rich melts from the harzburgiteMgCO3K2CO3 experiments are significantly poorer in MgO compared with those from the SC1MgCO3K2CO3 system. The content of Al2O3 depends primarily on pressure and is lower than 0.5 wt % in the 10 GPa experiments and up to almost 2 wt % in the 6 GPa experiments. FeO content is controlled by temperature and increases from about 5 wt % at 1200C to 2025 wt % at 1700C and 10 GPa. However, the very high FeO content in the high-temperature experiments is in part due to the high oxygen fugacity in the experiments and considerable amount of Fe3+ in the melt.

A comparison of the experimental carbonated silicate melt with the supposed primary kimberlite magmas (Becker, Le Roex, 2006) has led us to the conclusion that the latter cannot be produced by single-stage melting of the asthenospheric mantle. In order to explain the origin of protokimberlite carbonated silicate melts, we proposed a two-stage scenario, including the formation of a liquid enriched in volatile and incompatible components in the asthenospheric (lherzolitic) mantle and its interaction with depleted harzburgites (possibly, preliminarily metasomatized) in the lower part of the continental lithosphere.



Becker, M., Le Roex, A.P. (2006). Geochemistry of South African on- and off-craton, group I and group II kimberlites: Petrogenesis and source region evolution. Journal of Petrology 47, 673703.

Brey, G.P., Bulatov, V.K., Girnis, A.V., Lahaye, Y. (2008). Experimental melting of carbonated peridotite at 610 GPa. Journal of Petrology 49, 797821.

" "