Appraising the late-stage magmatic fO2 of diabasic angrites with a novel phase equilibrium approach. Implications for new and existing models of angrite petrogenesis

Abstract

One of the enduring problems in angrite petrogenesis is how to reconcile the extraordinary Fe-rich compositions of many angritic liquids – which seemingly require modestly oxidized fO2 conditions that fall outside the stability field of Fe-Ni alloys during primordial melting – with the geochemical and paleomagnetic evidence that the angrite parent body contains a small metallic Fe-Ni-S-C core. One of the major impediments to resolving these contradictions stems from a distinct lack of rigorous fO2 data for the angrite meteorite clan. To begin addressing this issue, we have developed a new approach for calculating magmatic fO2 values directly from the late-stage olivine-ulvöspinel- silicate liquid assemblages that are common in the volcanic or hypabyssal angrites. The adaptation of the olivine-spinel-aSiO2liq oxybarometer for use in angrites requires a careful assessment of aSiO2liq of angritic liquids. We apply the results of metal saturated angrite crystallization experiments and an analysis of the Ca-Tschermak’s-anorthite-aSiO2liq equilibrium to obtain estimates for the aSiO2liq values of angritic magmas. We then use the aSiO2liq estimates derived from the experiments and thermodynamic analysis in conjunction with the olivine-spinel-aSiO2liq oxybarometer to calculate the magmatic fO2 of Sahara 99555, D’Orbigny, and Northwest Africa 12004. With this approach we estimate oxygen fugacity values that range from ΔIW + 0 to ΔIW + 0.25. The relatively reducing fO2 values obtained from our analysis are inconsistent with redox conditions required by the oxidized melting hypothesis. Some caution is warranted in extrapolating these late-stage magmatic fO2 values to higher temperatures; however, we stress that fractionated silicate liquids typically become more oxidized with increasing crystallization via auto-oxidation (i.e., the accumulation of incompatible ferric iron in the liquid). Therefore, we suggest that late stage magmatic fO2 values from our calculations may represent the most oxidized conditions along the angrite liquid line of descent. If this interpretation is correct, a large-scale oxidation event on the APB need not be invoked to reconcile mildly reducing conditions (ΔIW−1.35) thought to have prevailed during core formation with the magmatic fO2 record preserved in angritic meteorites. Future redox studies of compositionally primitive angrites (e.g., Northwest Africa 12774) may shed new light on the redox relationships among primitive angrite magmas, evolved angrite magmas, and core forming alloys.