Carbonatite Isotope Dataset (CID): A critical reappraisal of Sr, Nd, C, O isotopes and trends

Abstract

Here we present a compilation of published SrNd and CO isotope data on carbonatites and phoscorites of the World integrated in the Carbonatite Isotope Dataset, covering the known interval of the carbonatite formation from recent up to ∼3 Ga ago. The dataset brings together 2545 ⁸⁷Sr/⁸⁶Sr, 2692 ¹⁴³Nd/¹⁴⁴Nd and 2528 ¹³C/¹²C-¹⁸O/¹⁶O isotope ratios from a total of 319 deposits and occurrences with known ages. Measured SrNd isotope ratios have been revised, recalculated and errors fixed using a set of unified equations and normalization factors. The final data can be sorted according to chemical composition, formation age, model ages, geotectonic setting, associated silicate magma and geographic affiliations (region, state, and continent). Analysis of SrNd isotope evolution of carbonatites and phoscorites revealed gradual change from primordial depleted mantle character of parental magmas to recent HIMU-like, plume-related character modified to various extents by the assimilation of continental crust. Large perturbations of SrNd isotope compositions are coincidental with the break-up and fragmentation of Columbia and Gondwana. Theoretical examples of isotope fractionation associated with magma mixing, phenocryst accumulation, degassing, and alteration have been calculated and compared with those observed in carbonatites affiliated with mantle plumes. Most variations in the recorded CO isotope composition can be attributed to subsolidus alteration of primary igneous carbonatite with low-temperature aqueous fluids enriched in CO₂. The low-temperature alteration combined with the igneous source heterogeneity and high-temperature magma-crust interactions obscure the stable isotope fractionation coincidental with phenocryst accumulation and magma degassing. Inverse modelling of Sr-Nd-C-O isotope trends showed that the interaction of mantle plume-derived magma with the crust may result in mixing trends with curvatures and slopes dominantly controlled by time-integrated changes in the ¹⁴³Nd and ⁸⁷Sr proportions, thereby modifying Sr/Nd ratios in the mixing phases. Models of mixing mantle components with similar Sr/Nd ratios, such as HIMU and EM1, also showed conspicuous, age-dependent modifications of the mixing line slope. Hence, the HIMU-EM1 trend revealed in modern carbonatites cannot be used as a template for interpreting isotopic signatures of pre-Cretaceous carbonatites. It is therefore likely that the role of enriched mantle component EM1 has been overestimated in Precambrian carbonatites at the expense of a continental crust assimilant. The same conjecture is valid for the enriched mantle component EM2, because sub-horizontal mixing trajectories with a wide range of εSr_CHUR combined with limited range of εNd_CHUR can also be reproduced by the contamination of mantle plume-derived magma with crustal assimilants. Except for specific carbonatites, e.g. crustal-anatectic, ocean island-related or kimberlite-associated, the SrNd and CO isotope ratios themselves cannot unambiguously discriminate individual genetic and geotectonic groups without additional supporting geochemical data and/or radiogenic isotope systems.