Hydrous melting of garnet peridotite under reducing conditions

• Applicant: Professor Dr. Alan Butler Woodland
• Subject Area: Mineralogy, Petrology and Geochemistry
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 566456556

Partial melting is fundamental to mass transfer and differentiation in the Earth at a variety of scales. The solidus temperature and melt composition are essential parameters for understanding this process. This project will experimentally investigate the process of redox melting under reducing conditions, which are increasingly relevant for the deeper portions of the mantle lithosphere and asthenosphere. In contrast with previous studies of hydrous melting, the low fO2 imposed means that the fluid phase present under sub-solidus conditions will have a distinctly different composition in terms of dissolved species compared with more oxidizing conditions. This should cause a shift in the solidus temperature as well as modify the incipient melt composition. In addition, the roll of elevated H2 concentrations and the possible influence of a coexisting Fe-hydride melt on the partial melting of mantle silicates will be investigated. The redox melting process will be simulated at 4-6 GPa using the experimental three-capsule technique recently developed in Frankfurt for imposing reducing conditions. Carbon will be omitted so that the fluid phase will be a H2O-H2 mixture, which is justified on the basis that it is actually the availability of H2O under reducing conditions that drives the redox melting reaction rather than C. Natural fertile lherzolite and a depleted harzburgitic compositions will be investigated in the presence of a small range in H2O content (0.5 and 3 wt %) using olivine capsules. The resulting samples will be studied by microprobe imaging and chemical analysis to identify incipient melting. As a further aid, photo-induced force microscopy (PiFM) will also be employed. This innovative approach available at ANU allows chemical imaging at the nanometer scale, allowing better characterization of grain boundaries. Since PiFM yields information analogous to FTIR absorption, but at a much higher resolution, if melt is quenchable, then at least a estimate of H2O concentration can be made. If the melt is unquenchable, then PiFM will permit identification of the quench phases present and their relative abundances by 2-D mapping. The olivine sample container serves as a trap for fluids, which can subsequently be analyzed by Raman spectroscopy to determine whether H2O and H2 were immiscible or not during the experiment. In addition, Raman will be used to scan for H2 incorporation in olivine itself, as well as for the presence of a metallic hydride phase. Quantitative measurement of OH concentrations in the olivine sample container will be made by FTIR (Clermont-Ferrand). These data, combined with SIMS measurements (Heidelberg) of total H provide a way to quantify both OH and H2 incorporated in the olivine at both sub-solidus and super-solidus conditions.

• DFG Programme: Research Grants
• International Connection: Australia, France
• Cooperation Partners: Dr. Nathalie Bolfan-Casanova; Professor Gregory Yaxley, Ph.D.