Mineralisation and fluid processes in the alternation zone around the Chilwa Island and Kangankunde carbonatite complexes, Malawi

Carbonatite magmas characteristically expel fluids into the surrounding country rock, creating metasomatised alkali-rich fenite aureoles. This thesis examines evidence for the mobility ofhigh field strength elements (HFSE), including the rare earth elements (REE), niobium and zirconium into these fenite aureoles in order to provide a better understanding of element mobilisation, transportation and precipitation. The study is timely because the REE and Nb are recognised as critical metals. The REE, in particular, are used in green technologies, and have grown in importance in recent years. The combination of the concentration of REE mines in China, associated export restrictions and a severely limited scope for substitution of these elements, has highlighted the challenges of securing adequate supplies and of a better management of resources, together with the need for an improved knowledge of the processes by which mineral deposits are enriched to economic grade. The approach used here was to study fenite aureoles around two Cretaceous carbonatite complexes at Chilwa Island and Kangankunde in the Chilwa Alkaline Province of southern Malawi by re-examining samples held in the collections of the Natural History Museum, London (BM1968 P37 and BM1957 1056). The geology of the complexes had already been well described by M. Garson and A. Woolley but the results of recent studies elsewhere suggested that it should now be possible to explore the fenitising process in more detail. Kangankunde is a rare earth-rich ferroan dolomite carbonatite, subject to active exploration for the REE. Chilwa Island contains a wider range of carbonatites, including apatite-, magnetite-bearing calcitic carbonatite, and REE mineral-bearing ankeritic and sideritic carbonatites. The presence ofassociated silicate rocks is minor at both these complexes. Samples were selected from across the fenite aureoles to represent areas of differing intensity and style of metasomatism. Whole-rock analyses were made to assess compositional changes across the aureole. Mineralogical studies by SEM-EDS, EPMA, LA-ICP-MS and SEM cathodoluminescence permitted the identification of mineral assemblages, the establishment of paragenetic sequences and also the characterisation of the REE compositions in apatite, zircon and the RE-bearing minerals from zones of varying degrees of alteration within the aureole. Fluid properties were further explored by fluid inclusion studies of secondary inclusions in the country rock quartz. Apatite and zircon in the fenite were dated by fission track analysis and a U-Pb LA-lCP-MS method, respectively, to determine their relationships with the carbonatite. At both complexes, fenite rocks contain micro-assemblages of minerals visible by backscattered electron imaging but not reported in previous studies. Examples include the assemblages of zircon, ilmenite, rutile, apatite and monazite found in veins of up to 500 um width in the sodi-potassic outer fenite at Chilwa Island, and the associations of strontianite, RE-minerals and carbonates in mm-sized veins in the more highly metasomatised parts of the aureole at Kangankunde. Alteration and dissolution-precipitation reactions seen in these micro-assemblages suggest that multiple fluid events occurred in the fenite aureole. The presence of alkaline minerals (aegirine, albite, K-feldspar, arfvedsonite) indicates pervasive fenitisation by fluids expelled from the carbonatite over distances of more than a kilometre. The micro-assemblages provide evidence that fluids were able to mobilise, transport and precipitate the HFSE and REE outwards over a similar distance into the country rock but that for the REE this event followed an earlier fenitisation by alkaline fluids. It is likely that fluids from more than one carbonatite have contributed to the variation in mineralogy and mineral composition. Fenite rocks and minerals have lower light REE:heavy REE ratios than those of the carbonatites in the core of the complexes. This cannot be explained by simple mixing between carbonatite and country rock end members, and fractionation of the REE during fluid transport is therefore invoked with relative preferential mobilisation and deposition of the mid to heavy REE. Zircon and apatite in the mineral assemblages at Chilwa Island appear petrographically to be co-eval. However, although fission track dating confirmed a carbonatitic age of c.130 Ma for apatite, the U-Pb dating of associated zircon produced ages of between 520 and 770 Ma and so zircon must be part of the original country rock rather than having co-precipitated with apatite. In the less altered fenites, zircon is stable, and at Chilwa Island, appears to encourage the nucleation of carbonatite-derived mineral assemblages. In contrast, zircon is unstable in higher- grade fenite at both complexes where metasomatising fluids were most intense. Zirconium is thus released from highly altered rocks and, rather than forming alkali zirconosilicates as happens in alkaline complexes, it appears to be transported outwards and is re-precipitated as a secondary, sub-micron zircon population in low- grade fenite. Interpretative models require the ingress of multiple fluids at each complex. Early alkaline fluids were more extensive and predominantly sodic. Later potassic fluids produced more intensive alteration in the inner aureole. The mineral assemblages were precipitated in veins in the aureole by fluids carrying the REE, which were expelled in multiple episodes from the different carbonatites at each complex, both before and after alteration by potassic fluids. At Chilwa Island, modelling suggests that early REE mineralisation in the fenite rocks from the outer carbonatites of the complex occurred after pervasive initial alkaline alteration by sodic fluids. This was overprinted by intense potassic alteration in the innermost fenites. The subsequent relative MREE and HREE enrichment in the aureole is attributed to fluids from the late-stage carbonatites. At Kangankunde, fluid events produced the same pattern of alkaline alteration followed by HFSE mineralisation, but here enrichment in the MREE and HREE occurred early, after the sodic alteration, with later fluids precipitating minerals with higher LREE:HREE ratios and substantial carbonate. As at Chilwa Island, the distinct characteristics of each mineralising fluid show that it is likely they were expelled from different carbonatites in the complex.