The solubility of TiO 2 in olivine: implications for the mantle wedge environment

Abstract

One characteristic of many subduction-zone garnet peridotites is that they contain titanium-bearing phases not otherwise found in mantle rocks. In particular, titanoclinohumite and/or its breakdown assemblage consisting of symplectic intergrowths of olivine and ilmenite is common in many of these bodies. The Alpe Arami garnet lherzolite of the Swiss Alps, while lacking titanoclinohumite, displays instead large numbers of FeTiO 3 rod-shaped precipitates in the oldest generation of olivine, amounting to approximately 1% by volume, indicating that at some time in its past, the peridotite experienced conditions under which the solubility of TiO 2 in olivine was >0.6 wt.%. In order to test the hypothesis that the environment of very high solubility of TiO 2 in olivine is to be found at very high pressures, we have conducted experiments on lherzolite compositions with added ilmenite at pressures between 5 and 12 GPa and temperatures of 1350–1700 K. Our results on anhydrous compositions show that whereas solubility of TiO 2 was not detected in olivine at 5 GPa, 1400 K where it coexists with rutile, when rutile disappeared from the paragenesis, the solubility climbed to 0.4 wt.% at 8 GPa, 0.5 wt.% at 10 GPa and to >1.0 wt.% at 12 GPa, 1700 K. These results support our previous interpretations from titanate morphology and abundance that the Alpe Arami massif has surfaced from P=10 GPa but remove the need to suggest a deeper origin and possible precursor phase such as wadsleyite. They also support the hypothesis that garnet peridotites with unusual Ti-bearing phases reflect a unique mantle environment occurring in the mantle wedge overlying subduction zones.