Michael Bau, Robert Frei, Dieter Garbe-Schoenberg, Sebastian Viehmann

ABSTRACT
Unique deposits that formed in the Precambrian oceans are the Banded Iron-Formations (BIFs) which are chemical sediments comprised of alternating layers of iron- and silica-dominated precipitates. The origin of this enigmatic banding is still debated, with most favoring a primary sedimentary and a few others arguing for a secondary diagenetic origin. We here used a high-resolution integrated multi-proxy approach and report on the Ge-Si-Fe and Th-U systematics and on the isotopic composition of Cr in exceptionally pure adjacent magnetite and metachert bands from the ~2.7 billion years old Temagami BIF, Ontario, Canada.
While each of the geochemical proxies alone may still be controversially discussed, our multi-proxy approach sheds new light on their applicability, as the two types of adjacent BIF bands are characterised by distinct geochemical signatures. The magnetite bands show super-crustal (i.e. fractionated) Ge/Si ratios approaching those of modern marine high-temperature hydrothermal fluids, unfractionated Th/U ratios similar to those of crustal rocks, and unfractionated Cr isotope ratios similar to "Bulk Silicate Earth". Adjacent metachert bands, however, show lower (i.e. unfractionated) Ge/Si ratios close to those of modern seawater, river water and crustal rocks, sub-crustal (i.e. fractionated) Th/U ratios and fractionated Cr isotope ratios above those of Bulk Silicate Earth. The composition of the magnetite bands suggests that the original iron-oxyhydroxide precipitated during periods when the Ge, Th, U, and Cr inventory of seawater was dominated by input from mafic crust via anoxic high-temperature hydrothermal fluids. In contrast, the composition of the metachert bands indicates that these formed during times when seawater was dominated by low-temperature riverine input from an evolved landmass that despite its Neoarchean age had provided environments for oxidative Cr and U mobilization during terrestrial weathering. This suggests that habitats with oxygenic photosynthesis may have existed almost 400 million years before the GOE. The systematic difference between the geochemical signature shown by the two types of adjacent BIF bands also demonstrates that the prominent banding of BIFs is a primary depositional feature reflecting precipitation from different watermasses and is not the result of post-depositional separation of an initially homogenous chemical sediment.