Experiments and Models Bearing on the Role of Magma Mixing and Contamination on Chromite Crystallization in Ultramafic Magmas

To better understand the origin of stratiform chromitites, we experimentally evaluate the role of bulk compositional shifts arising from mixing or contamination of a mafic magma on both the liquid line of descent and the chromite saturation state. Experiments equilibrated synthetic Cr-bearing komatiite containing 0-50 wt.% Cr-free contaminants (granodiorite - GD, Fe-rich shale – Fe-shale, metasediment – MS, natural magnetite - NM) on Fe-pre-saturated Pt or Ir loops at 1192-1462°C and 0.1 MPa at the fayalite-magnetite-quartz (FMQ) oxygen buffer. Experimental run products involving mixtures of komatiite with silicate contaminants produced some combination of euhedral olivine, chromite and glass, whereas initial cubes of NM nearly completely dissolved to produce Fe-enriched melt plus Cr-rich magnetite. The addition of contaminants results in the chromium content of the melt at chromite saturation (CCCS) to decrease with increasing melt FeO or decreasing melt SiO2 abundance. Assessment of the temperature dependence reveals that the log of the CCCS decreases linearly with inverse temperature, allowing for the chromite crystallization temperature to be predicted for a given melt Cr concentration. The addition of any of the studied contaminants decreases the modal abundance of olivine. The origin of the melt composition dependence of the CCCS is traced to variation in the activity coefficient of Cr3+ (γCr3+) and molar Cr3+/[total Cr], both in the melt phase, with log(γCr3+) increasing and Cr3+/[total Cr] decreasing with the wt% ratio of SiO2/FeO in the melt. The high solubility of magnetite in komatiite at experimental conditions would preclude the survival of entrained Fe-Ti-oxide xenocrysts and their transformation to chromite, as proposed by Lesher et al. (2019). We apply the chromite solubility data to models of a) mixing between differentiated komatiite recharged by primitive komatiite and b) assimilation by stoping of pre-heated blocks of Fe-shale, GD and model banded iron formation (BIF). Mixing between differentiated and recharging primitive komatiite results in olivine-out and chromite as the sole crystallizing phase only if the recharging komatiite is contaminated, requiring >10% Fe-Shale or BIF; GD-contaminated mixtures yield Cr-concentrations too low for chromite saturation. Assimilation by stoping suppresses olivine crystallization and induces chromite-only crystallization if >10% Fe-Shale or BIF, or >20% GD, comprise the contaminated magma mass. In all models, contamination by Fe-shale and BIF yield a larger olivine-out temperature interval than for GD, but in general the chromite crystallization interval is short (5-50 degrees). As a measure of the relative efficacy of these processes, for the contaminants considered, we estimate that meter-scale thicknesses of chromitite produced by 30% contamination of the recharging primitive komatiite or by stoping require magma columns of ~0.5 to ~6 km. These are similar to the column of B1 magma (2 km) required by the pressure-reduction hypothesis of Latypov et al. (2018) to form chromitite of the Bushveld complex. For several chromitites world-wide, the geochemical evidence for chromite-only crystallization is lacking, instead consistent with chromite co-precipitation with olivine or orthopyroxene. Chromitite formation may therefore reflect intervals of chromite-only precipitation but also the possible role of fluid dynamic-related sorting of predominantly cotectic chromite-silicate mixtures.