Melting properties and melting phase relations of the Martian mantle from in-situ measurements on iron-rich mineralogical assemblages

Mars’ mantle dynamical history has certainly been dominated by a stagnant-lid regime, with limited mixing and homogenization. Accordingly, the chemical and mineralogical signatures of early processes, including the crystallization of a primitive magma ocean, are overall well preserved on Mars. The major geological structures visible at its surface are the remains of an intense ancient volcanism, not so dissimilar from the large igneous provinces found on Earth at very old ages (several million/billion years).

Current models used to determine the mantle thermal evolution and the crustal extraction heavily relies on melting properties of materials expected to form the Martian mantle, which, however are poorly known. In particular, the fact that the Martian mantle is probably richer in iron than the terrestrial mantle has a direct impact on the solidus and liquidus and on the chemistry of the magmas that can be produced at different pressures. Thus, the study of Martian volcanism and thermal history requires a precise understanding of the melting properties of the mantle (solidus, liquidus and extent of melting) as a function of pressure and temperature. Studies in literature are scant, mainly address the solidus, and are limited to analysis of recovered samples, missing in situ diagnostics.

To address this problem, we studied the solid-liquid melting relations and, more generally, the melting diagram for a mineralogical assemblage model of mantle composition, by high-pressure and high-temperature experiments in multi anvil press performed at the PSICHE beamline of the SOLEIL synchrotron. We determined the solidus and the liquidus of the investigated rock at pressures up to 12 GPa by complementary in-situ diagnostics (X-ray diffraction and falling sphere technic). The obtained solidus and liquidus are well lower (difference >200K), especially at the highest investigated pressures, compared to previous studies, with strong implications for the origin of volcanism and notably the crystallization of the magma ocean. Furthermore, our experiments provide important data to refine the extent of melting (Φ), modal proportion and the chemistry of all the different phases present between the solidus and the liquidus at different conditions (P, T, Φ).

Altogether, these new results are critical to constrain models of thermal evolution and crust extraction and formation, as well as to address the evolution of the magmatism and volcanism at the Mars surface since 3.5 Ga. Finally, depending on different parameters, such as the thickness of the crust or the concentration of radioactive elements, the estimated areotherm could cross the solidus and lead to partial melting of the mantle, especially close to the core-mantle boundary, where a high extent of melting could be reached.