The origin and evolution of intrusive and extrusive carbonatites

Carbonatites are among the most volumetrically insignificant igneous rocks found at the earth's surface, and over the past century, a general picture has emerged on the composition, origin and evolution of carbonatitic magmas, however much remains to be done to fully understand them. Carbonatites are characterised by high concentrations of several strategic elements, including rare-earth elements, niobium, tantalum, phosphorous, copper, iron and fluorine. Carbonatites are mostly considered to be derived from a common parental magma, yet the different types of carbonatites are genetically distinct as they form at different depths in the upper mantle and are the result of partial melting, fractional crystallization or liquid immiscibility. Carbonatites have also been largely accepted to be associated with suites of alkaline igneous rocks and not as single rock units. Understanding the relationship between carbonatites and associated silicate rocks is an important aspect in interpreting the mantle to crust evolution of carbonatite complexes. Silicate-carbonate immiscibility is currently the leading mechanism for the association of coexisting carbonatites and alkaline silicate rocks. The advances in microanalytical techniques means that sophisticated mineralogy and melt inclusions methods that have been at the centre of many recent petrological and geochemical discoveries can now be applied to these complexes and provide important clues to the nature of the magmas that give rise to the genesis of carbonatites and the associated economic deposits within alkaline-ultramafic and carbonatitic complexes. The thesis involves case studies on the 1993 carbonatitic eruption at Oldoinyo Lengai and perovskite ore from the Afrikanda alkaline-ultramafic complex. The studies build on, and expand, the use of petrography and mineralogy of carbonatites and associated alkaline silicate rocks to further the understanding of the genesis of carbonate-silicate melts and seeks to address petrological, geochemical and economic enigmas of extrusive and intrusive carbonatitic magmas. The goals of this thesis are to (1) understand the differentiation processes involved in the evolution of the carbonatites (2) determine the composition of the melt prior to emplacement and during magma ascent and (3) understand the genesis of the Oldoinyo Lengai lava and the perovskite ore in the Afrikanda alkaline-ultramafic complex. These goals are addressed by investigating the petrology and geochemistry of minerals and melt inclusions in the carbonatites and associated silicate rocks. The rarity and difficulty in capturing definitive evidence of liquid immiscibility within magmas has posed a challenge for researchers, as intrusive carbonatite rocks are commonly affected by crystallisation and post-magmatic alteration. Liquid immiscibility is currently understood to be the underlying process for the generation of carbonatite magmas from a parental alkaline silicate magma at crustal depth, and is possibly the only magmatic process that can explain the association of coexisting carbonatites and alkaline silicate rocks. Therefore, petrological and mineralogical studies on extrusive carbonatites, like Oldoinyo Lengai, are necessary to understand the genetic history of carbonatitic magmatism. The texture and mineralogy of the 1993 lava presents evidence of immiscibility between silicate, carbonate, chloride, and fluoride melt phases through textural features preserved in the silicate spheroids, melt inclusions, and carbonatite groundmass. The melt inclusions and silicate spheroids present evidence of silicate-carbonate and carbonate-carbonate immiscibility, while the groundmass shows evidence of carbonate-carbonate and carbonate-halide immiscibility. The rapid quenching of the lava facilitates the preservation of the end products of these immiscibility processes within the groundmass. The mineralogy and melt inclusion study at Oldoinyo Lengai provides confirmation that liquid immiscibility is responsible for the formation of natrocarbonatites. Textural evidence (at both macro- and micro-scales) also indicates that the formation of this natrocarbonatite lava did not occur as a simple single-stage process, and that multi-stage liquid immiscibility is a major factor in the petrogenesis of the lava. Therefore, although immiscibility is a common phenomenon in silicate magmas, we can identify evidence of unmixing through melt inclusion and groundmass studies of preserved extrusive carbonatites. The identification of carbonate-carbonate and carbonate-halide immiscibility within the natrocarbonatite lava shows that different types of liquid immiscibility, other than silicate-silicate and silicate-carbonate, can occur in natural magmas. The Afrikanda alkaline-ultramafic complex is one of the smallest intrusions in the Devonian Kola Alkaline Province (KAP). The KAP in NW Russia is one of the largest carbonatite provinces in the world and hosts more than twenty plutonic and subvolcanic bodies, including alkaline, ultramafic, carbonatite, and melilitolite suites. Many of these complexes host ore deposits, with active mines at Kovdor (apatite-magnetite ore), Lovozero (loparite-ore) and Khibiny (nepheline-apatite ore). The concentric internal structure of the Afrikanda complex hosts olivinites and clinopyroxenites, cross-cut by minor intrusions of carbonatitic rocks with the ultramafic rocks hosting large stock-like bodies of perovskite-titanomagnetite ores. Perovskite-rich segregations are only found in alkaline-ultramafic complexes. Understanding the formation of the perovskite-rich segregations at Afrikanda is important for understanding the source and early history of the silicate and carbonatitic magmas. Afrikanda is assumed to be a magmatic complex, with the intrusive ultramafic rocks derived by partial melting of a metasomatised lithospheric source to produce a Ca-rich melanephelinitic magma, and the later carbonatites from an alkaline silica-rich carbonatitic magma that intruded the earlier-formed ultramafic rocks. The study investigates the mechanisms responsible for the development of the perovskite-rich segregations through a detailed textural and chemical analysis of the perovskite and perovskite- and magnetite-hosted inclusions within the ultramafic rocks (olivinites and clinopyroxenites) and carbonatites of the Afrikanda alkaline-ultramafic complex. Additionally, U-Pb geochronology was combined with major and trace element geochemistry to support the contemporaneous emplacement of the various rock types in the Afrikanda complex. Across the perovskite-rich lithologies, we classified perovskite into three types (T1-T3) based on crystal morphology, inclusion abundance, composition, and zonation. Perovskite in olivinites and some clinopyroxenites is represented by clusters and networks of fine-grained, equigranular grains with abundant polymineralic inclusions (T1). In contrast, perovskite in other clinopyroxenites and some silicocarbonatites are coarse-grained with rare polymineralic inclusions with interlocked (T2) and massive (T3) textures. The polymineralic inclusions in perovskite and magnetite contain a variety of silicate, hydrous silicate, carbonate, oxide, sulphide and phosphate minerals. The unexpected mineralogy of the polymineralic inclusions raised several questions about the genesis of the alkaline complex. The difficulty in explaining the presence and composition of the inclusions in perovskite is associated with the occurrence of both magmatic and recrystallization processes in the complex. The study of the perovskite textures and perovskite- and magnetite-hosted inclusions revealed that perovskite has a magmatic origin but the formation of the polymineralic inclusions and the development of massive ore textures is associated with post-magmatic processes. Therefore, a non-magmatic model has been developed to explain the genesis of perovskite-rich segregations in the Afrikanda alkaline-ultramafic complex. In this model, we propose that the polymineralic inclusions in perovskite formed by trapping the surrounding material between perovskite grains during post-magmatic coalescence at subsolidus temperatures. The continuation of the sintering process resulted in the coarsening of inclusion-rich subhedral perovskite into inclusion-poor anhedral and massive perovskite. Initial crystallisation of perovskite from the magma results in disseminated euhedral crystals enclosed by larger silicate, carbonate and oxide minerals. During subsolidus cooling these small, randomly orientated grains accumulate together and develop into loosely packed aggregates of perovskite. As these grains coalesce, varying amounts of interstitial material are trapped between the inclusion-free grains. These denselypacked perovskite grains undergo textural equilibration to form aggregates of perovskite with the external and internal appearance of a single crystal with internal inclusions. Therefore, the development of polymineralic inclusions is associated with the magmatic crystallisation of perovskite, textural reequilibration and sintering. These perovskite grains then link together and form clusters and chains, with the granoblastic-polygonal texture associated with T1 perovskite. The progressive transition of T1 perovskite to T2 and T3 perovskite involves grain rotation and coalescence of these small equilibrated T1 polygonal clusters and results in the formation of larger anhedral polycrystalline mosaics (T2). In some areas, the continued consolidation and coarsening transforms the large polycrystalline perovskite into massive perovskite. Thus, the progressive development of clusters and networks of fine-grained perovskite crystals (T1) to mosaics of coarse-grained and massive perovskite (T2 and T3) in the ultramafic rocks is due to post-magmatic textural re-equilibration and re-crystallisation. A combination of characteristic features identified in the Afrikanda perovskite...