Petrogenesis of Kamthai and Amba Dongar Carbonatite Complexes, India: Constraints from <i>In Situ</i>  87Sr/86Sr Composition and U–Pb Ages of Apatite

Abstract The 87Sr/86Sr composition of apatite from carbonatites and associated silicate rocks can provide important petrogenetic constraints. This study reports for the first time, in situ 87Sr/86Sr composition of fluorapatite from calcite carbonatite, nephelinite, and phonolite from the Kamthai and Amba Dongar complexes, along with their major and trace element compositions. It also reports for the first time, U–Pb ages of apatite in phonolite from the Kamthai complex. Most fluorapatite are oscillatory-zoned, occurring as cumulates or disseminated crystals in the groundmass. Apatite grains from Kamthai carbonatites are distinctly enriched in Sr (2.26–4.08 wt %) and $\Sigma $REE+Y (0.87–1.82 wt %), and depleted in CaO than those in phonolites (0.25–2.28 wt %; 0.17–0.86 wt %), which suggests that Sr and REE substituted for Ca. The similar 87Sr/86Sr of fluorapatite from carbonatite (0.70425–0.70442) and phonolite (0.70381–0.70537) indicates that both were derived from a common parental magma and shared a common source. These results and the association with phonolites strongly suggest that the carbonatites may have formed by liquid immiscibly from a parental phonolitic melt. The limited Sr isotopic variability in fluorapatite of phonolite can be explained by the assimilation of felsic rocks of the Proterozoic Malani Igneous Suite (MIS; 8–15%) or by Archean banded gneisses (BGC; &amp;lt;3%) of the Aravalli Delhi Belt. However, crustal assimilation even up to ~30% of MIS-like or ~ 20% of BGC-like crust, does not affect the 87Sr/86Sr of carbonatites, which reflects the characteristics of its enriched mantle source. Fluorapatite from Amba Dongar rocks has lower concentrations of Sr (0.35–2.06 wt %), $\Sigma $REE+Y (0.37–1.12 wt %), and other incompatible elements compared to those from Kamthai. A positive correlation of $\Sigma $REE with SiO2 in fluorapatite from nephelinite is suggestive of substitution of Si4+ + REE3+ for P5+ + Ca2+, possibly because of elevated silica activity due to crustal assimilation. The 87Sr/86Sr of fluorapatite (0.70532–0.70578) and calcite (0.70558–0.70589) from carbonatite and nephelinite (0.70590–0.70669) are distinct from each other and more radiogenic than those of Kamthai. Approximately 20% assimilation of basement gneisses by a parental magma having 87Sr/86Sr similar to apatite of carbonatites can explain the more radiogenic 87Sr/86Sr ratios of fluorapatite from the nephelinite. The trace element and Sr isotope variations suggest that the mantle source for the Kamthai rocks was more enriched than that for the Amba Dongar rocks. Trace element modeling using apatite–melt partition coefficients suggests that carbonatite and alkaline silicate melts in both complexes are compositionally different and derived from different parental melts. The silicate rocks formed from parental melt, similar to basanite of the Deccan LIP. Uranium–Pb ages of fluorapatite from Kamthai phonolite (64 ± 12 Ma) indicate that it is contemporaneous with the Deccan flood basalts and can, therefore, be linked to the Deccan–Réunion mantle plume.