FATE OF RECYCLED MATERIAL IN THE MANTLE DURING DECOMPRESSION MELTING

This research project assesses the decompression melting of a heterogeneous mantle to determine the conditions for melts derived from recycled material (pyroxenite) to reach the surface without being in full re-equilibration with the surrounding mantle (peridotite). As 80% of magmatic productivity is produced at mid-ocean ridges, understanding the behavior of recycled material during adiabatic decompression is important for our comprehension of the Earth’s recycling cycle. Thanks to many experimental investigations, the melting behaviors of peridotite and pyroxenite are now well characterized. It has been demonstrated that most pyroxenite begins to melt at a lower temperature/higher pressure than peridotite [1]. Yet, few information relates to the mantle conditions and compositions required for melt derived from recycled material to reach the surface. Lambart et al. [2] proposed the hypothesis that for pyroxenite melt to reach the surface, the surrounding peridotite mantle needs to be above its solidus. This project tests this hypothesis by running piston-cylinder experiments along an adiabatic decompression path that simulates a heterogeneous mantle composed of peridotite and pyroxenite. We use the three-layer setup [3] in which a ‘pyroxenite layer’ is consecutively overlain by a ‘peridotite layer’ and a layer of vitreous carbon spheres (‘melt trap’). We are also testing the effect of the composition of the pyroxenite by conducting two series of experiments: one with the silica-deficient pyroxenite, M7-16 [4], and one with the mid-ocean ridge basalt (MORB)- like eclogite, G2 [5]. Both pyroxenites have lower solidus temperatures than peridotite but melt derived from these two lithologies are expected to react very differently with the peridotitic mantle [2]. The analysis of the composition of the melt through the layer of peridotite and the melt trap will provide information on the degree of re-equilibration of the experimental charge and the conditions for melt extraction. I will discuss the results of these experiments in my presentation.

[1] Hirshmann & Stolper, 1996, doi: 10.1007/s004100050184; [2] Lambart et al., 2013, doi: doi:10.1093/petrology/egr068; [5] Van den Bleeken et al., 2010, doi: 10.1093/petrology/egp066; [4] Lambart et al., 2009, doi: 10.1016/j.epsl.2009.09.038; [5] Pertermann & Hirschmann, 2003, doi: 10.1029/2000JB000118