Calcium isotope fractionation during melt immiscibility and carbonatite petrogenesis

Stable calcium isotopes have been used to suggest that subducted marine carbonates are frequently involved in the formation of carbonatites. Significant Ca isotope fractionations during carbonatite petrogenesis, however, could lead to a dramatically different picture. We present Ca isotope data for (i) coexisting (immiscible) carbonatite and silicate melts from high temperature centrifuging piston cylinder experiments, (ii) primary apatite and calcite/dolomite from natural carbonatites, and (iii) ab initio estimates for equilibrium Ca isotope partitioning in calcite, dolomite, and ankerite. Carbonatitic melts have lower δ⁴⁴Ca than their conjugate silicate melts, with an equilibrium fractionation factor [1000lnα(1000K)] of −0.21 ± 0.06 (tSE). We develop a quantitative four stage model for carbonatite petrogenesis (partial melting followed by fractional crystallisation, silicate-carbonatite melt immiscibility, and calcite/apatite accumulation) that fully explains our natural data (average δ⁴⁴Ca_BSE of −0.30 ± 0.03 ‰) and those from recent studies, without requiring isotopic contributions from recycled marine carbonates. Our results suggest that lighter isotopes of similarly bound cations (e.g., Mg, Fe, Sr, Ba, Zn) should be preferentially incorporated into carbonatitic melts and that calciocarbonatite formation involves melt immiscibility after differentiation of mantle-derived alkaline CO₂-bearing silicate melts.