ALBERT

Description: The Tundulu and Kangankunde carbonatite complexes in the Chilwa Alkaline Province, Malawi, contain late-stage, apatite-rich lithologies termed quartz-apatite rocks. Apatite in these rocks can reach up to 90 modal% and displays a distinctive texture of turbid cores and euhedral rims. Previous studies of the paragenesis and rare earth element (REE) content of the apatite suggest that heavy REE (HREE)-enrichment occurred during the late-stages of crystallization. This is a highly unusual occurrence in intrusions that are otherwise light REE (LREE) enriched. In this contribution, the paragenesis and formation of the quartz-apatite rocks from each intrusion is investigated and re-evaluated, supported by new electron microprobe (EPMA) and laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) data to better understand the mechanism of HREE enrichment. In contrast to the previous work at Tundulu, we recognize three separate stages of apatite formation, comprising an "original" euhedral apatite, "turbid" apatite, and "overgrowths" of euhedral late apatite. The crystallization of synchysite-(Ce) is interpreted to have occurred subsequent to all phases of apatite crystallization. The REE concentrations and distributions in the different minerals vary, but generally higher REE contents are found in later-stage apatite generations. These generations are also more LREE-enriched, relative to apatite that formed earlier. A similar pattern of increasing LREE-enrichment and increased REE concentrations toward later stages of the paragenetic sequence is observed at Kangankunde, where two generations of apatite are observed, the second showing higher REE concentrations, and relatively higher LREE contents. The changing REE distribution in the apatite, from early to late in the paragenetic sequence, is interpreted to be caused by a combination of dissolution-reprecipitation of the original apatite and the preferential transport of the LREE complexes by F- and Cl-bearing hydrothermal fluids. Successive pulses of these fluids transport the LREE out of the original apatite, preferentially re-precipitating it on the rim. Some LREE remained in solution, precipitating later in the paragenetic sequence, as synchysite-(Ce). The presence of F is supported by the F content of the apatites, and presence of REE-fluorcarbonates. Cl is not detected in the apatite structure, but the role of Cl is suggested from comparison with apatite dissolution experiments, where CaCl 2 or NaCl cause the reprecipitation of apatite without associated monazite. This study implies that, despite the typically LREE enriched nature of carbonatites, significant degrees of hydrothermal alteration can lead to certain phases becoming residually enriched in the HREE. Although at Tundulu the LREE-bearing products are re-precipitated relatively close to the REE source, it is possible that extensive hydrothermal activity in other carbonatite complexes could lead to significant, late-stage fractionation of the REE and the formation of HREE minerals.