Silicate diffusion in alkali-carbonatite and hydrous melts at 16.5 and 24 GPa: Implication for the melt transport by dissolution–precipitation in the transition zone and uppermost lower mantle

The diffusivity of dissolved Mg2SiO4 in wadsleyite saturated KMC melt (K2Mg(CO3)2 + 25.7 wt.% MgSiO3) at 16.5 GPa and 1700 °C, MgSiO3 diffusivity in perovskite saturated KMCH (K2Mg(CO3)2 × 2H2O + 31.7 wt.% MgSiO3) and HM (H2O + 75.7 wt.% MgSiO3) melts at 24 GPa and 1500 °C were determined experimentally using a scaled-up version of a Kawai-type multi-anvil apparatus. During a diffusion experiment, silicate saturation was maintained at different levels in the two temperature regions by placing the diffusion cell in the thermal gradient of 20 °C/mm. The diffusivity was computed from the total mass of silicate transported from “hot” to the “cold” region during the course of an experiment. At given conditions silicate diffusivities were estimated to be D KMC Mg 2 SiO 4 = 2 × 10 - 9 m 2 / s , D KMCH MgSiO 3 = 4 × 10 - 9 m 2 / s , and D HM MgSiO 3 = 5 × 10 - 8 m 2 / s . obtained diffusivities we estimated possible migration rates of dispersed melt inclusion in the deep mantle by means of dissolution–precipitation considering different driving forces. The rates of melt migration driven by the lateral thermal gradient of 1 °C/km in the mantle plume range from 4 × 10−8 to 8 × 10−7 m/year. This means that during plume ascent time of about 50 Ma, the melt can be moved by 2–40 m. These values clearly demonstrate that the thermal gradient is very weak driving force in terms of melt segregation in the deep mantle. On the other hand, at typical mantle stress of 1 MPa and droplet size of 100 μm the migration rates of the HM, KMCH and KMC melts are estimated to be 22.5, 0.9 and 0.2 m/year, respectively, which are 2–3 orders of magnitude faster than ascent rate of the mantle plume. This implies that all melt droplets on the way of ascending plume would be entrapped by the stressed zone in front of plume and accumulated in the plume head. This mechanism may explain segregation of mantle magmas with the source regions deeper than 150–250 km, such as kimberlites.