Challenges in applying REE-based thermobarometers: Insights from mafic-ultramafic granulites in Indian southern granulite terrain

High-temperature diffusion of major elements may obscure the records of early and peak metamorphic stages in granulites, while trace elements are more likely to preserve these records due to their lower diffusion rates. Thus, using calibrated REE-based thermobarometers has proved essential for reconstructing these key stages, drawing considerable attention and application from scholars. However, the precision of these thermobarometers depends on including both major and trace elements from coexisting minerals to define the correlation coefficients (A, B, and D) among mineral pairs, indicating that the elemental composition of these pairs can affect the results. Our study examines the mafic-ultramafic granulites in the southern granulite terrain, India, employing integrated methods such as petrography, mineral chemistry, phase equilibrium modeling, and REE-based thermobarometers. We aim to determine their metamorphic conditions and evolutionary history and to identify potential challenges in using REE-based thermobarometers. The garnet, clinopyroxene, and orthopyroxene in the mafic-ultramafic granulite samples display homogeneous compositional profiles, with pronounced Fe-Mg diffusion zones at the interfaces between garnet and clinopyroxene. Conversely, the profiles of trace elements within garnet and clinopyroxene are better preserved. Investigations into Fe-Mg exchange and randomly selected mineral pairs significantly influence the accuracy of REE-based thermobarometers. Fe-Mg exchange can increase in coefficient A, while decreasing coefficient B for light rare earth elements (LREEs) and increasing it for heavy rare earth elements (HREEs), ultimately resulting in over-estimations when calculating REE-based thermobarometers. For example, selecting major compositions with an Ex ( \(=\left(X_{\text{Mg}}^{C}-X_{\text{Mg}}^{0}\right)/X_{\text{Mg}}^{0}\times100\%\) $$=(X_{\text{Mg}}^C-X_{\text{Mg}}^0)/X_{\text{Mg}}^0\times 100\mathrm{\%}$$ ; where \(X_{\text{Mg}}^{C}\) $$X_{\text{Mg}}^C$$ is the value after Fe-Mg exchange, \(X_{\text{Mg}}^{0}\) $$X_{\text{Mg}}^0$$ is the value before Fe-Mg exchange; \(X_{\text{Mg}}=\text{Mg}/\left(\text{Fe}^{2+}+\text{Mg}\right)\) $$X_{\text{Mg}}=\text{Mg}/({\text{Fe}}^{2+}+\text{Mg})$$ ) value of ∼10 for calculation using the REE-in-Grt-Cpx thermobarometer will result in pressures and temperatures being ∼10 kbar and 30–40°C higher than the true values. Random pairing, such as selections based solely on the core or rim of minerals with changes in trace elements, can severely impact the distribution coefficient D, resulting in substantial discrepancies in thermobarometric calculations and potentially producing anomalous results. Thus, to minimize the impact of these factors, it is necessary to first analyze the profiles of major and trace elements in coexisting minerals before applying the REE-based thermobarometers to evaluate the P-T conditions of granulites. Based on this analysis, major element compositions less affected by Fe-Mg exchange (such as avoiding the selection of major compositions at the boundaries of minerals) and in relative equilibrium in trace element compositions among coexisting minerals (same growth periods) should be selected for pairing. Furthermore, integrating additional methods should also be considered when applying the REE-based thermobarometers, to prevent the misinterpretation of the P-T conditions obtained.