Cadmium isotope composition of the Earth's mantle inferred from analysis of oceanic basalts and komatiites

Cadmium stable isotope ratios can serve as valuable tracers for biological and geological processes, including nutrients sampling from the surrounding medium, volatilization events, and terrestrial differentiation. However, studies of the isotope fractionation occurring during geological processes require the characterization of the isotope compositions of the different terrestrial reservoirs. Notably, the cadmium isotopic composition of the Bulk Silicate Earth (BSE) is still a matter of debate. To address this issue, we present high-precision cadmium isotope data from a diverse set of samples, including mid-oceanic ridge basalts (MORB) and komatiites obtained using a 106Cd111Cd double spike system. We processed nine biological and geological reference materials (AGV-2, BCR-2, BHVO-2, BIR-1, COQ 1, GSP-2, NIST 2711, NOD-A-1, ERM BB 186) and two pure Cd solutions (AAS and ChemLab) to assess the accuracy and precision of the method.Our study revealed that twenty-three MORB samples (including ten analyses from the literature) from the Atlantic, Indian, and Pacific oceans have a homogeneous cadmium isotopic composition. The weighted mean isotope composition for these samples, expressed as δ114Cd (the permil deviation of the 114Cd/110Cd ratio from the NIST SRM 3108 Cd standard), is 0.07 ± 0.10 ‰ (2SD, n = 23). We also studied twelve komatiite samples from the Abitibi (Superior Craton, Canada), Barberton (Kaapvaal Craton, South Africa), and Kostomuksha (Baltic Shield, Fennoscandia) greenstone belts to characterize the deep Archean mantle isotope composition. These samples, combined with four analyses from the literature (Pickard et al., 2022) exhibited isotopic compositions ranging from −0.33 ± 0.04 ‰ to 0.26 ± 0.08 ‰. We observed a negative correlation between cadmium abundances and the degree of secondary alteration (measured by loss on ignition-LOI). For samples with low LOI (≤ 6 wt%), the isotope compositions of komatiites are quite consistent. On the other hand, the samples with higher LOI (between 6 and 8.4 wt%) tend to be more scattered and can have lighter isotope compositions. These correlations suggest that, although Cd isotope compositions in komatiites are quite resilient to alteration, heavier Cd isotopes can be mobilized into secondary alteration fluids, driving altered samples toward isotopically lighter compositions. This implies that secondary alteration exerts an important control on the isotope compositions of komatiites. Therefore, only the Cd isotope compositions of the least altered komatiites (i.e., those with LOI ≤ 6 wt%) have been used as proxies for their mantle source isotope composition. These samples (two from the Abitibi greenstone belt, one from the Barberton greenstone belt, one from the Baltic shield and one from the Belingwe greenstone belt) gave a weighted mean isotope composition of 0.14 ± 0.18 ‰ (2SD, n = 5). No significant isotope variation was found between the pristine samples from each region, despite ages ranging from 3549 ± 99 Ma to 2713 ± 29 Ma, indicating a possible isotopic homogeneity (within the currently achievable precision) of the Archean mantle for Cd starting as early as 3.55 billion years ago. We determined a Cd isotope composition for the mantle by combining the data for mean MORB and komatiites. This gave a δ114Cd value of 0.11 ± 0.10 ‰ (2SD, n = 28) for the Bulk Silicate Earth. This value is significantly heavier than previously proposed.