Abstract
The Pt-graphite double-capsule technique is a very commonly used method in high-temperature, high-pressure experimental petrology, particularly for anhydrous experiments relevant to primitive basaltic magmas and mantle melting. We have performed a series of experiments that place better constraints on the range of oxygen fugacity imposed by this capsule material, on the Fe3+/Fe2+ ratios in experimentally produced melts and minerals, and on the temperature reproducibility in Pt-graphite capsules. Oxygen fugacity in our piston-cylinder experiments using Pt-graphite capsules is CCO-0.7 (IW+1.5, QFM-2.2) at 1.5 GPa and 1360 °C. Comparison with other estimates and thermodynamic calculations indicate that a value of CCO-0.8 ± 0.3 can be used as a first approximation at least over the P-T range relevant for MORB and OIB magma generation (0.5–3.0 GPa, 1100–1500 °C). Under those conditions, the amount of Fe3+ in silicate phases (pyroxenes, olivine, glass) and spinel is negligible (Fe3+/∑Fe < 0.05) and would not significantly affect thermodynamic properties. Significantly higher values of fO2 cannot be achieved using Pt-graphite or graphite only capsules, but fO2 can be tuned to lower values by using small pieces of PtFe alloys. The potential range of fO2 that can be reached in graphite or Pt-graphite capsules is CCO to CCO-4. Temperature reproducibility in piston-cylinder experiments has been examined and can be as low as ±10 °C. Finally, unless capsules are dried overnight at 400 °C before the experiment, small amounts of H2O are always present in nominally dry experiments. These small amounts of H2O should not, however, significantly change phase relations.