Carbonatite metasomatism, the key to unlocking the carbonatite-phoscorite-ultramafic rock paradox

Abstract

One of the features of carbonatite-phoscorite complexes that has received comparatively little attention is that many of them are surrounded by large bodies of ultramafic rocks. Although it is generally assumed that all the rocks are of igneous origin, there is disagreement over whether the carbonatites and phoscorites are genetically related to the ultramafic rocks. This is despite the fact that in complexes where they have been dated, these three suites of rocks are identical in age. A major challenge for any igneous hypothesis is the paradox posed by the observation that the ultramafic and carbonatitic magmas are the products of very high and very low degrees of partial melting of the mantle, respectively. We resolve this paradox by proposing a hypothesis, in which the ultramafic rocks are metasomatic in origin. According to this hypothesis, a carbonatitic magma, generated by partial melting of carbonated mantle, acts as the agent of metasomatism. The interaction of this magma with quartz-rich rocks, including granites and gneisses, transforms them into ultramafic rocks. This is made possible by the extremely low viscosity of the magma that allows it to pervasively infiltrate the wall rocks. As a result, Mg, Ca and other components are transferred to the host, leading to the crystallisation of minerals such as olivine, clinopyroxene and biotite, and the loss of CO2 from the system. Because of its low viscosity, the carbonatitic magma is emplaced as numerous intrusions, ranging in width from the micron to the metre scale, that progressively transform the host rocks into ultramafic rocks. As this metasomatism proceeds, the Na concentration of the magma gradually increases and eventually becomes high enough to permit large scale assimilation of Si (and Al) and the generation of an alkaline silicate magma that is emplaced at the margins of complexes, forming ijolites. The metasomatism and assimilation are envisaged as a front that moves progressively from the conduit to the limits of the complex, producing an outward zonation from dunite through clinopyroxenite to ijolite or from clinopyroxenite to ijolite directly. Whether the zonation is fully developed or dominated by a single zone, e.g., dunite or clinopyroxenite, depends on the initial composition of the carbonatitic magma and/or the degree of progress of the metasomatic reaction. The processes described above constitute the “silicate stage” of carbonatite-phoscorite-ultramafic rock complex development and are marked by the complete consumption of the carbonatitic magma. During the waning stages of the formation of these complexes, metasomatism by new batches of carbonatitic magma is restricted to remnants of reactive wall rocks adjacent to the magma conduit. Consequently, the magma is only partially consumed. This results in a residual liquid that is enriched in phosphate and iron and crystallises phoscorite. We refer to this as the “phoscorite stage” of carbonatite-phoscorite-ultramafic rock complex development. The final, “carbonatite stage”, is marked by the cessation of metasomatic activity and the crystallisation of carbonatites from batches of magma that are unaffected by interaction with the wall rocks and are compositionally identical to those that initiated the “silicate stage”. This model explains the spatial association of the different rock types, in which the complexes are zoned outwards from a core of carbonatite, through a zone enriched in phoscorite, into a broad halo of ultramafic/ultrabasic rocks. In so doing, the model accounts for the many features of these complexes that have not been adequately explained in previous genetic models.