Trace elements and isotope geochemistry (C, O, Fe, Cr) of the Cauê iron formation, Quadrilátero Ferrífero, Brazil: Evidence for widespread microbial dissimilatory iron reduction at the Archean/Paleoproterozoic transition

Abstract

The Banded Iron Formation (BIFs) of the Cauê Formation, Quadrilátero Ferrífero, Brazil, are part of the great volume of BIFs deposited worldwide around 2.45Ga. Samples of carbonate (ankerite/dolomite) BIF from this unit were collected from a drill core ca. 600m deep in the Alegria region. Geochemical data suggest low clastic contamination (Y/Ho=37.7) and seawater-like signatures, with positive La and Y anomalies, no true Ce anomalies and mean Eu/Eu∗=1.62. U/Th shows a sharp increasing-downwards pattern (up to ca. 15) below the 535m mark. Values of δ13C and δ18O are all negative, reaching a minimum of −10.0‰ and −24.6‰, respectively, in the 540–560m interval. Values of δ56Fe are mostly positive, between 0.5 and 1.3‰; the higher values are attained in the same interval. Values of authigenic δ53Cr range between 0.01‰ and 0.26‰, with an increasing-downwards pattern which mirrors the U/Th variation. A possible explanation for both the low δ13C and the high δ56Fe is the activity of microorganisms in the dissimilatory iron reduction (DIR) of the precursor ferric oxides. Thus, we put forward a model for deposition of the Cauê BIFs that starts with the generation of Fe(II)aq through hydrothermal input to the deeper basin, which upon reaching a chemocline in the platform border, oxidizes to Fe(III) and settles as iron (oxy)hydroxides to the basin floor. When reaching the sediment layer, Fe(III) is then reduced by microbial activity to generate magnetite and iron-rich carbonates with negative δ13C and positive δ56Fe. An important issue is how U(VI) and Cr(VI) were mobilized to the basin to form the positive U/Th and δ53Cr excursions. These could reflect oxidative continental weathering, but an alternative possibility is that those elements became available through local oxygenation in continental microenvironments, which would better explain the lack of Ce redox cycling. In either case, our results, together with a growing body of data for BIF sequences of similar age around the world (e.g. Kuruman, South Africa and Hamersley, Australia), suggest that microbial-mediated DIR was a common and widespread mechanism for the genesis of BIFs at the Archean/Paleoproterozoic transition.