Experimental determination of ferric iron partitioning between pyroxene and melt at 100 kPa

Pyroxene is the principal host of Fe3+ in basalt source regions, hosting 79 and 81% of the Fe(3+ )in spinel and garnet lherzolite, respectively. In spinel peridotite, orthopymxene (opx) and clinopymxene (cpx) host 48% and 31%, respectively, of the total Fe3+. Yet the relationship between mantle mineralogy, pyroxene chemistry, and the oxygen fugacity (f(o2)) recorded by mantle-derived basalts remains unclear. To better understand partitioning of Fe3+ between pymxene and melt we conducted experiments at 100 kPa with f(o2) controlled by CO-CO2 gas mixes between Delta QFM -1.19 to +2.06 in a system containing andesitic melt saturated with opx or cpx only. To produce large (100-150 mu m), homogeneous pyroxenes, we employed a dynamic cooling technique with a 5-10 degrees C/h cooling rate, and initial and final dwell temperatures 5-10 degrees C and 20-30 degrees C super and sub-liquidus, respectively. Resulting pyroxene crystals have absolute variation in Al2O3 and TiO2 < 0.05 wt% and < 0.02 wt%, respectively. Fe3+/Fe-T in pymxenes and quenched glass were measured by Fe K-edge XANES. We used a newly developed XANES calibration for cpx and opx by selecting spectra with X-rays vibrating on the optic axial plane at 45 +/- 5 degrees to the crystallographic c axis. Values of D-Fe3+(cpx/melt) increase from 0.03 to 0.53 as f(o2) increases from AQFM -0.44 to +2.06, while DFe3+ (cpx/melt) remains unchanged at 0.26 between Delta QFM -1.19 to +1.37. In comparison to natural peridotitic pymxenes, Fe3+/Fe-T in pyroxenes crystallized in this study are lower at similar f(o2), presumably owing to lower Al-3(+) contents. Comparison to thermodynamic models implemented in pMELTS and Perple_X suggest that these over-predict the stability of Fe3+ in pymxenes, causing these models to underpredict the f(o2) of spinel peridotites under conditions of basalt genesis.