Origin of chromitites in the Esker Intrusive Complex, Ring of Fire Intrusive Suite, as revealed by chromite trace element chemistry and simple crystallization models

Abstract

To better constrain the origin of the chromitites associated with the Esker Intrusive Complex (EIC) of the Ring of Fire Intrusive Suite (RoFIS), a total of 50 chromite-bearing samples from the Black Thor, Big Daddy, Blackbird, and Black Label chromite deposits have been analysed for major and trace elements. The samples represent three textural groups, as defined by the relative abundance of cumulate silicate phases and chromite. To provide deposit-specific partition coefficients for modeling, we also report on the results of laboratory experiments to measure olivine- and chromite-melt partitioning of V and Ga, which are two elements readily detectable in the chromites analysed. Comparison of the Cr/Cr+Al and Fe/Fe+Mg of the EIC chromites and compositions from previous experimental studies indicates overlap in Cr/Cr+Al between the natural samples and experiments done at >1400oC, but significant offset of the natural samples to higher Fe/Fe+Mg. This is interpreted to be the result of subsolidus Fe-Mg exchange between chromite and the silicate matrix. However, little change in Cr/Cr+Al from magmatic values, owing to the lack of an exchangeable reservoir for these elements. A comparison of the composition of the EIC chromites and a subset of samples from other tectonic settings reveals a strong similarity to chromites from the similarly-aged Munro Township komatiites. Partition coefficients for V and Ga are consistent with past results in that both elements are compatible in chromite (DV = 2-4; DGa ~ 3), and incompatible in olivine (DV = 0.01-0.14; DGa ~ 0.02), with values for V increasing with decreasing fO2. Simple fractional crystallization models that use these partition coefficients are developed that monitor the change in element behaviour based on the relative proportions of olivine to chromite in the crystallizing assemblage; from 'normal' cotectic proportions involving predominantly olivine, to chromite-only crystallization. Comparison of models to the natural chromite V-Ga array suggests that the overall positive correlation between these two elements is consistent with chromite formed from a Munro Township-like komatiitic magma crystallizing olivine and chromite in 'normal' cotectic proportions, with no evidence of the strong depletion in these elements expected for chromite-only crystallization. The V-Ga array can be explained if the initial magma responsible for chromite formation is slightly reduced with respect to the FMQ oxygen buffer (~FMQ- 0.5), and has assimilated up to ~20% of wall-rock banded iron formation or granodiorite. Despite the evidence for contamination, results indicate that the EIC chromitites crystallized from 'normal' cotectic proportions of olivine to chromite, and therefore no specific causative link is made between contamination and chromitite formation. Instead, the development of near- monomineralic chromite layers likely involves the preferential removal of olivine relative to chromite by physical segregation during magma flow. As suggested for some other chromitite-forming systems, the specific fluid dynamic regime during magma emplacement may therefore be responsible for crystal sorting and chromite accumulation.