The first step of the project was to develop a continuous model for the density of carbonate melts in the system MgO-CaO-Na2O-K2O-Li2O-H2O-CO2. Such a model was missing, in spite of its unique potential to investigate a diversity of carbon-mediated magmatic processes, from carbonatitic eruptions to carbonate melts migration and crystal fractionation, and to constrain the seismic signature of carbonate melts in the upper mantle/transition zone. The present density model is calibrated to span most of these conditions of interest. After having tested different equations of state (EoS), the pressure-volume relations of each oxide component are ultimately described by employing a simple Murnaghan EOS based on 5 adjustable parameters, and describing the partial molar volume and compressibility of the different oxide components. The calibrated density model reproduces well the calibration dataset and predicts densities with an average relative error lower than ±5%. Combined with reasonable assumptions for the volumetric properties of SiO2 and FeO, the applications of the model to density and seismic P-wave velocities Vp predictions in complex carbonatite systems issued from partial melting carbonated lithologies provide new insights into their migration and geophysical signature. Some of the mean observations issued from our calculations are:
1- Both density and VP contrasts between carbonatitic melts and the host solid mantle diminish with depth, which implies larger effect of carbonate-rich melts on the seismic velocity signature of the Earth’s upper mantle at shallower depths;
2- We pinpoint that carbonate melts are seismically faster than silicate melts, which together with their higher electrical conductivity, are promising diagnostic features to identify molten phases at depth and to discriminate between carbonate and silicate melts in the Earth’s upper mantle using geophysical observations.
In a second step, we developed a continuous model for the viscosity of mantle melts, spanning the whole compositional spectrum of melts prevailing within the Earth’s upper mantle. The model is calibrated to be used in most conditions of interest for the crust and upper mantle. We have developed a new non-Arrhenian three-parameters model to predict the contrasted viscosity properties of mantle melts, from carbonate-rich to silicate-rich compositions, inspired of a reformulation of the MYEGA model. The model is parameterized in a way to include effects of pressure-temperature-composition, and is still under construction while most of the calibration and related results have been obtained so far, and a publication is now under preparation.
Viscosity of mantle melts with various compositions, from carbonatites rich in CO2 to basalts enriched in SiO2, will be calculated and presented in the publication, along with estimations for the mobility of melts within the host rocks.
The database used for the development of the continuous viscosity model has benefited from the viscosity points of mixed alkali and alkaline-earth melts at upper mantle conditions, i.e. up to 13 GPa and 1000 – 2000 K, obtained by MD simulations. The latter were performed by a former PhD student in the host group, Dr. Xenia Ritter, in collaboration with Profs. B. Guillot and N. Sator from Sorbonne University (Paris, France). I have participated to this study which provides first evidence for non-Arrhenian temperature dependent viscosity of carbonate melts at upper mantle pressures, which may be a behaviour extensible to other carbonate melt compositions. To facilitate the integration of the new data into carbonate mobility and geodynamic modelling, I employed the data to parameterize a model describing the pressure-temperature dependence of the viscosity of alkali(ne) carbonate melts at upper mantle conditions.
The density model and applications are reported in a publication currently in revision (moderate revisions) in ‘Chemical Geology’, where I am the first author, as well as at different conferences: GeoMünster – Annual Conference of the German Geological Society and the German Mineralogical Society, Münster, Germany; EMPG 2021, Potsdam, Germany (online); Goldschmidt 2021, Lyon, France (online); and the Japan Geoscience Union 2022. The viscosity model and applications will be submitted to ‘Earth and Planetary Science Letters’.
