ILMENITE LAMELLAE IN {111} OF HOST MAGNETITE: EXSOLUTION OR OXIDATION ORIGIN?

Formulations of the Fe-Ti oxide thermometer/oxybarometer assume that the original spinel (Mt-Usp_ss) was stoichiometric, and that ilmenite (Ilm_ss) lamellae in {111} formed by oxidation of Fe₂TiO₄ component in the spinel. But experiments at high T and atmospheric pressure show limited but real cation deficiency in spinel coexisting with ilm_ss, which is chemically equivalent to limited but real solid solution of ilm_ss in the spinel [e.g., 1300ºC (Taylor, Am. Min. 49, 1964); (Senderov et al., Am. Min. 78; 1200ºC (Webster and Bright, J. Am. Ceram. Soc., 44, 1961); 1100ºC (Katsura et al., Bull.. Volc. Soc. Japan, 21]. Lattard (Am. Min., 80, 1995) showed that such spinels can exsolve ilm_ss on {111} when heated at lower T in a closed system.

Lattard’s work has been taken by some authors as grounds for rejecting the oxidation origin of {111} ilmenite in spinel, despite her statement “In most terrestrial rocks, however, spinel + ilmenite intergrowths are certainly best interpreted as resulting from oxidation...” Here we point out severe limits on the amount of ilm_ss that can form by direct exsolution.

At 1300ºC, the cation deficiency at 1 atm. corresponds to ~15 mol % equivalent ilm_ss. The amounts at 1200ºC and 1100ºC are even lower, ~ 4-12 mol % and ~3-8 mol %. Yet commonly magmas do not become saturated with two Fe-Ti oxides until they cool to 1100ºC and below, so the possible extent of cation deficiency in the spinel must be in the latter range or lower. Furthermore, limited data show that the molar volumes of cation-deficient Fe-Ti spinels are larger than the equivalent stoichiometric spinel plus ilm_ss, suggesting that the extent of deficiency will be lower in plutonic rocks than in volcanics at the same T. We are testing this idea by synthesizing Fe-Ti spinel (ilm_ss) at 1300ºC and 1 atm., then reheating it at the same T but at 5 Kbar (0.5 GPa).

Bottom line: The amount of ilm_ss that can form by direct exsolution from Fe-Ti spinel is probably 3-8% or less for volcanic rocks, and even lower for plutonic rocks. Greater amounts of {111} ilm_ss must have formed by oxidation of Fe₂TiO₄ component in the spinel.