Key Factors Controlling Biotite–Silicate Melt Nb and Ta Partitioning: Implications for Nb−Ta Enrichment and Fractionation in Granites

Abstract

AbstractBiotite–melt Nb and Ta partition coefficients (DNb and DTa) are crucial for understanding Nb−Ta enrichment and fractionation in rare metal granites (RMGs). However, the key factor(s) affecting biotite–melt DNb, DTa, and DNb/DTa values remain unclear. To elucidate the physicochemical factors that control the partition coefficients, we performed piston−cylinder experiments at 0.5−1.0 GPa and 850−1000°C with H2O‐added (4−10 wt.%) mixtures of granitic and biotitic glasses as starting materials. Two series of experiments (graphite‐buffered and unbuffered fO2 conditions) were conducted with calculated fO2 values ranging from ∼FMQ–1.5 to ∼FMQ+4. Under these experimental conditions, biotite–melt DNb, DTa, and DNb/DTa values are 0.30–2.63, 0.24−1.02, and 1.01−2.15, respectively. Biotite–melt DNb, DTa, and DNb/DTa values increase with decreasing melt NBO/T value (non‐bridging oxygens per tetrahedron), melt H2O content, and biotite Mg#T value [molar 100 × MgO/(MgO + FeOT)]. In addition, DNb and DTa exhibit good correlations with DTi, suggesting that DNb and DTa are predictable via DTi values. With our and literature data, we used multiple linear regressions to obtain empirical expressions of DNb and DTa as functions of the three parameters. By applying the empirical models to granite differentiation process, we found that ∼99% crystallization of biotite ± muscovite‐bearing assemblages results in an enrichment in magma Ta contents by >10 times with a decrease in Nb/Ta values from 10−13 to ∼1, which reproduces the Ta–Nb/Ta features of most RMGs. However, additional processes, such as columbite‐group mineral precipitation, may be required to account for the extremely low Nb/Ta values (<1) of some RMGs.