Rare-metal enrichment in carbonatite-bearing magmatic systems: Part B. Understanding the role of fractional crystallization and liquid immiscibility by experimental simulations of silicate-carbonatite systems

Carbonatites constitute the main source of REE and Nb. Most ore-bearing carbonatitic complexes are closely related to silica-undersaturated alkaline igneous rocks and are considered to differentiate from a parental carbonated basanitic-nephelinitic magma of mantle origin by crystal fractionation. This fractionation process, and in particular the role of immiscibility between a silicate and a carbonate melt at some stage of the differentiation as well as the partitioning of trace elements between co-existing melts is not understood. The few available experimental data indicate that trace element should partition to the silicate melt, which is contrasting with the data obtained from the investigation of natural case studies. The proposed work program is designed to fill these gaps in our knowledge. Experiments will be conducted at high temperatures and pressures to understand at which conditions crystallization of carbonated nephelinitic magma leads to the formation of immiscible melts and if silicate-carbonate immiscibility can bypass immiscibility, ending up with the direct crystallization of calcite and other carbonate minerals. Pressure conditions will vary from 1,5 to 0.1 GPa, allowing us to simulate processes in magmas ascending from mantle to near-surface condition. The products from crystallization experiments will be used to constrain the trace element partitioning between silicate melt, carbonate melt and mineral phases (e.g., clinopyroxene, perovskite, apatite). Pilot experiments will also be conducted to understand the effects of additional ligands (F-, Cl-, PO43- and SO42-) on the immiscibility gap between silicate and carbonate liquids and to identify and characterize possible metasomatic reactions between residual natrocarbonatitic melt and minerals crystallized from nephelinitic magmas in early differentiation stages . The starting material for the experiments will be selected so that the results can be applied to natural case studies, especially the Kola Alkaline Province, representing a classical and well-exposed case of carbonatites and related