Iron-Titanium Oxide Thermometry: Is Chromium a Perturbing Element?

The Fe-Ti-oxide thermometer and oxygen barometer of Buddington & Lindsley (1964, JPET 5, 310-357) has been recently re-examined on the basis of a large experimental data set retrieved by Evans et al. (2006, CMP 152, 149-167) and Lattard et al. (2005, CMP 149, 735-754) in the Fe-Ti-Mg-Mn-Al-O system. Ghiorso & Evans (2008, AJS in press) have presented a new thermodynamic model for the rhombohedral oxide solid solutions and derived a revised version of the thermo-oxybarometer which yields much better T-fO₂ estimates for magmas equilibrated under relatively oxidizing conditions. In contrast, the numerical fits of Sauerzapf et al. (2008, JPET 49, 1161-1185) are directly based on experimental results and designed for use at temperatures above 800°C under reduced or moderatly oxidized conditions. However, tests on chromium-bearing ilmenite-titanomagnetite pairs from gabbroic xenoliths in basalts point to temperature overestimates by as much 100°C.

Because Cr is an important supplementary element in Fe-Ti oxide pairs from a variety of terrestrial and lunar rocks equilibrated at high temperatures and reducing conditions (e.g. xenoliths in kimberlites, lunar basalts) relevant phase equilibrium experiments appear necessary to test the new formulations of the Fe-Ti oxide thermometer.

We have performed synthesis and re-equilibration experiments in the Fe-Ti-Cr-O and Fe-Ti-Cr-Mg-O systems at 1 bar in the sub-solidus temperature range 900-1300°C at low to moderate oxygen fugacity values (ΔNNO between -4 and -1). The run products are polycrystalline assemblages of titanomagnetite and ilmenite solid solutions with Cr₂O₃ contents between 12 and 18 wt.% for the spinels and between 0.7 and 4 wt.% for the ilmenites. Preliminary results show that in most cases both our formulation and that of Ghiorso & Evans (2008) overestimate the temperature, especially for assemblages equilibrated under strongly reducing conditions.