Ferric/Ferrous Iron in Plagioclase of the Skaergaard Intrusion

The oxidation state of magma bears on the liquid line of descent and the crystallization products of magmatic differentiation. For example, the differentiation of basalt towards either silica enrichment (rhyolite) or iron enrichment (ferrobasalt) hinges on the amount and composition of FeTi oxides crystallizing, that in turn hinges on magma composition and the redox state of iron. To evaluate the redox state of iron during crystallization of the Skaergaard intrusion we have carried out synchrotron micro beam Fe-XANES (X-ray Absorption Near-Edge Structure) analyses and determined the proportion of ferric iron to total iron (Fe3+/Fe) in plagioclase of the Layered Series. The advantage of using iron in plagioclase over FeTi oxides is that iron is a trace element in plagioclase and that plagioclase is the last phase affected by subsolidus equilibration. In cumulus plagioclase cores Fe3+/Fe range from 74 to 15 (based on calibration against oxides) and correlate positively with the anorthite content. The highest Fe3+/Fe (74 to 47) are in the early fractionation products formed prior to the onset of magnetite crystallization, that is in Lower Zone a and b. In the middle portion of the fractionation sequence above the appearance of FeTi oxides, that is from Lower Zone c through to Upper Zone b, Fe3+/Fe displays large within sample variation and measured values in this portion of the stratigraphy range between 62 and 34 without showing a systematic stratigraphic trend. Finally, in the extremely iron-rich cumulates with fayalitic olivine and sodic plagioclase of Upper Zone c, Fe3+/Fe ranges from 26 to 15. These Fe-XANES results strongly indicate the Skaergaard magma was relatively oxidised prior to magnetite crystallization but thereafter became more reduced with fractionation. This result corroborates to inference from QUILF equilibrium. Hence the Skaergaard magma chamber must have been closed to oxygen exchange in order to reduce ferric iron in the magma in consequence of magnetite crystallization. While Fe3+/Fe decreases with fractionation, the total amount of FeO in the plagioclase cores (0.24 to 0.46 weight percent) increases with fractionation after magnetite-in. This cannot be explained by crystal chemistry or changing partitioning coefficients. We therefore conclude that the iron content of the magma increased with fractionation even after magnetite-in, and that ferrous iron dominated this iron-rich ferrobasaltic end product.