Petrologists are generally faced with the dilemma of estimating Fe2+/Fe3+ ratios for minerals on the basis of microprobe analyses that only identify total iron, commonly given as FeO. Common solutions to this problem include the assumption that all iron is ferrous, or simply calculating Fe3+ on the basis of stoichiometry from the microprobe analyses. In order to test these assumptions, ferric iron content has been directly measured on Al-augite, kaersutite, and spinel high-pressure megacrysts from alkali basalts by using the technique of Mössbauer spectroscopy. Measured Fe3+ content does not agree with Fe3+ calculated from microprobe analyses. Calculated Fe3+ content varies widely between three augite megacrysts that have nearly identical microprobe analyses and consistent Mössbauer Fe3+ values. Ferric iron calculations are found to be highly sensitive to very slight variations in Si, 1+, 3+, and 4+ cations, and should be considered generally unreliable.
Mantle petrologists generally assume low or neglible ferric iron content in mantle minerals. This study found that 28%-40% of iron was ferric in these eight megacrysts. Similarity of Al-augite and kaersutite megacryst compositions to minerals in mantle xenoliths suggests that minerals in mantle xenoliths may have higher Fe3+ content than generally believed.