The effect of network-modifiers (Mg, Ca, Na and K) on ferric/ferrous ratio in silicate melts: experimental study and improvement of existing models.

Geochemical processes occurring in magmatic systems depend strongly on oxygen fugacity and on the Fe3+/Fe2+ ratio in silicate melts. This ratio can be theoretically predicted using available thermodynamic models. These models take compositional parameters for basaltic systems into account, but have not been calibrated for a wide compositional range. Previous studies of the proponents on the role network formers (Si, Al, Ti, P) on the Fe3+/Fe2+ ratio in silicate melts also show that the formalism adopted in the thermodynamic models needs to be improved by adding interaction terms for some key elements in the equations (a simple formalism adding the individual effect of each oxide, ΣdiXi, is not sufficient). In this project, we explore experimentally the effect of network-modifiers (Mg, Ca, Na, K) on the Fe3+/Fe2+ ratio in silicate melts. The experimental results obtained so far were used to develop an equation predicting the Fe3+/Fe2+ ratio in multicomponent aluminosilicate melts (published in 2017). However, this equation is only calibrated at oxidizing conditions. In an extension of this project, we plan to incorporate various datasets obtained in previous studies at reducing conditions. It is also planned to make some additional experiments to test the formulation developed so far with natural multicomponent systems (oxidizing and reducing conditions). Using this extended dataset, a consistent thermodynamic model predicting accurately the effect of oxygen fugacity and melt composition on the Fe3+/Fe2+ ratio of natural mafic to silicic melts will be proposed.