No abstract

and Yb between olivine and coexisting basaltic and komatiitic liquids was measured at known fugacity of oxygen, f(O 2). Olivine and glass were analyzed for trace elements using laser-ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS). The purpose of the study was to further refine an empirical oxygen barometer for picritic magmas (>10% MgO) based on the comparison of the f(O 2)-dependent behavior of V with that of other incompatible and immobile trace elements (Ti, Sc, Ga, Zr and Yb) along the liquid line of descent. Results of this study present an improvement on the earlier estimates of redox conditions that utilized this approach because the partition coefficients for V and other incompatible trace elements (Ti, Sc, Ga, Zr and Yb) are measured in the same bulk-compositions under the same experimental conditions, eliminating any previous bias or correlations. The advantages and limits of the approach are demonstrated in applications to modern Hawaiian picrites, Archean komatiitic lavas and lunar rocks. Uncertainties in the method are significant [about ± 0.5 log f(O 2) units], but it is applicable to mafic suites having an igneous phase-assemblage altered by metamorphism, or without a phase assemblage to which more conventional estimates of redox conditions can be applied. In many cases, crucial estimates of the minimum f(O 2) can be made for many terrestrial and extraterrestrial rocks in the geological record.

Diamond oxidation experiments were undertaken in a piston-cylinder apparatus at 1150 to 1500 °C and 1 GPa to understand the mechanism of diamond oxidation in kimberlite melts and to determine the main rate-controlling parameters for this process. Only surface graphitization, and no diamond resorption, occurs in melts that are fl uid undersaturated (synthetic kimberlite, carbonate melt, alkaline basalt, CaO-MgO-SiO 2 -H 2 O-CO 2 melts). In contrast, fl uid oversaturated conditions (as evidenced by the presence of bubbles) produce resorption features commonly seen in natural diamonds recovered from kimberlites. The diamond oxidation rate is the same in the melts with a free fl uid phase, in a pure H 2 O or CO 2 fl uid, suggesting that the process of diamond oxidation is its reaction with the fl uid and not with the melt. Both CO 2 and H 2 O oxidize diamonds at a similar rate, but produce very different surface features. Therefore, the surface features of natural kimberlite-hosted diamonds may provide information on the relative proportion of H 2 O and CO 2 in the kimberlitic fl uid. The common diamond morphologies imply signifi cant amount of H 2 O. The absence of diamonds with surface graphitization and the abundance of resorbed diamonds in kimberlites suggest the presence of a free fl uid phase in kimberlite magmas for hours or days. We found no correlation between the rate and character of diamond oxidation and the physical properties of diamonds (nitrogen content, color).

No abstract

No abstract