REE geochemistry of 3.2 Ga BIF from the Mapepe Formation, Barberton Greenstone Belt, South Africa

Banded iron formations (BIFs) are chemical sediments interbedded with Fe- and Si-rich layers, characteristically present in the early history of the Earth. A popular hypothesis for the formation of BIFs postulates that dissolved oxygen produced by photosynthesizers such as cyanobacteria oxidized dissolved ferrous Fe supplied by submarine hydrothermal activities. During precipitation of Fe-oxide minerals, phosphorus and rare earth elements (REEs) were most likely adsorbed on their surface. Therefore, chemical compositions of REEs that adsorbed onto Fe-oxide have useful information on the seawater chemistry at the time of deposition. Especially, information on the redox state of seawater and the extent of the contribution of hydrothermal activity during BIF deposition are expected to have been recorded. Occurrence of BIF has been traditionally tied to the chemical evolution of the atmosphere. Rise of atmospheric oxygen, or as known as GOE (Great Oxidation Event: e.g., Holland, 1994), has been widely believed to have occurred at around 2.4 Ga ago. Contrary, however, some studies have suggested that such oxygenation could have occurred much earlier (e.g., Hoashi et al., 2009). In this study, we used 3.2 Ga old BIF from the Mapepe Formation at the bottom of the Fig Tree Group of the Swaziland Supergroup in the northeastern part of the Barberton Greenstone Belt, South Africa. We aimed to constrain the marine environment, and by inference atmospheric environment, at the time of BIF deposition from REE geochemistry. Major elements and REE compositions of 37 samples were measured using XRF and ICP-MS, respectively. Samples with less than 1.0 wt% Al2O3 are considered to be "pure" BIFs with minimal amount of continental contamination, and are expected to have inherited marine REE signatures. Abundance of REE normalized by C1 chondrite for the analyzed samples commonly exhibits positive Eu anomaly and LREE<HREE signature, suggesting significant influence of hydrothermal fluid on BIF deposition. Decoupling of Y/Ho, most likely due to the difference in adsorption behavior onto Fe-oxide, suggests particle reactivity or precipitation of Fe-oxide. Y/Ho ratios show a positive correlation with total Fe2O3 contents and degree of positive Eu anomaly, strongly suggesting that Fe emanated by hydrothermal venting was oxidized and precipitated as Fe-oxide that adsorbed Y and Ho. Significant negative Ce anomaly was not observed, possibly suggesting (1) greater mixing ratios (dilution) of seawater (with negative Ce anomaly) / hydrothermal (no Ce anomaly) during the BIF deposition, or (2) the seawater itself did not possess negative Ce anomaly. Oxygen isotope ratios of Fe-oxide and silica may reveal temperature of mixing for those fluids and by inference the mixing ratios.