Extreme iron enrichment and liquid immiscibility in mafic intrusions: Experimental evidence revisited

This paper examines phase equilibria and mass balance constraints on Fe enrichment in tholeiitic liquids in plutonic environments. The peak of Fe enrichment is thought to roughly coincide with magma saturation in Fe–Ti oxides; and olivine starts to react with the liquid at about the same time. This crucial stage of crystallization is examined in detail using a compilation of chemical analyses of 64 experimental charges that comprise liquids (quenched glasses) equilibrated with the liquidus assemblage of olivine, plagioclase, high-Ca pyroxene, and low-Ca pyroxene. Some samples also contain Fe–Ti oxides. It is shown that the 4-phase liquidus assemblage does not constrain a narrow range of liquid compositions. The concentrations of SiO2 in the selection of experimental glasses vary broadly from 42 to 66 wt.%. Silica content shows strong negative correlations with FeO and CaO/Al2O3, and strong positive correlation with alkalis. Extreme Fe enrichment above 22 wt.% FeO is observed only in alkali-free or alkali-poor liquids. Broad compositional variations for the multiply-saturated liquids are attributed to strong non-ideality and complex concentration-activity relationships in ferrobasaltic melts. Liquid immiscibility characteristic of Fe-rich silicate liquids is the ultimate consequence of non-ideality. Petrogenetic implications of phase equilibria and mass balance constraints are discussed for a classical example of the Skaergaard intrusion in East Greenland, where the trend of extreme Fe enrichment has been in contention. It is proposed that seemingly conflicting results of experiments on Skaergaard natural cumulate rocks and model melt compositions can be reconciled if it is assumed that silicate liquid immiscibility in Skaergaard started not at the very end of crystallization but earlier, soon after the start of ilmenite and magnetite crystallization.