REE and Nd isotope data indicate that most of the iron in banded iron formations is derived from hydrothermal sources but do not exclude a significant contribution from terrestrial sources, such as diagenetic recycling. A diagenetic model has been used to estimate the recycling of iron into overlying seawater, due to microbial reduction and dissolution at depth in anoxic sediment pore waters, followed by diffusion upward through a surface layer of sediment that contains oxygenated pore waters. Rates of iron recycling increase with higher pore-water dissolved iron concentrations, decreasing pH and temperature, and smaller thicknesses of the surface oxygenated layer. Iron can be recycled at rates of 1000–5000 µg cm −2 yr −1 from Proterozoic (pO 2 = PAL) pore waters with dissolved Fe 2+ = 1–5 µg cm −3 , pH 6.5 (and T < 65 °C), or pH 7.0 (and T < 40 °C), or pH 7.5 (and T < 20 °C), provided the thickness of the surface oxygenated layer is less than 0.1 cm. Lower pO 2 levels and more weakly oxygenated surface layers do not significantly increase the maximum recycling rates but enable these to be achieved at larger thicknesses of the surface layer, for all pH 6.5–7.5 and temperatures from 10 to 65 °C. Rates of iron supplied by diagenetic recycling can be substantially modified by the export efficiency (ϵ) from the source area and by the ratio (Source Area)/(Sink Area), which can either disperse or concentrate the recycling flux that is delivered to a sink area of banded iron formation. Banded iron formations that require maximum iron delivery rates of 22500 µg cm −2 yr −1 can be produced only by recycling rates of 5000 µg cm −2 yr −1 (and ϵ = 1) from a source area that is at least four times larger than the area of banded iron formation. Modern basins have ratios of shelf area (<200 m water depth) to deep basin area that commonly range from 0.2 to 4. Basins at either extreme have ratios of (Deep Basin Area)/(Shelf Area) or (Shelf Area)/(Deep Basin Area) that exceed 4 and are potentially able to concentrate iron either from a deep basin source area to banded iron formation on the shelf, or from a shelf source area to a banded iron formation depositing in the deep basin. However, these mass balance constraints require the existence of substantial areas of contemporaneous source sediments (or smaller areas of iron-enriched sediments) located either on the shelf or in the deep basin.

Abstract Modern and ancient euxinic sediments are often enriched in iron that is highly reactive towards dissolved sulphide, compared to continental margin and deep-sea sediments. It is proposed that iron enrichment results from the mobilization of dissolved iron from anoxic porewaters into overlying seawater, followed by transport into deep-basin environments, precipitation as iron sulphides, and deposition into sediments. A diagenetic model shows that diffusive iron fluxes are controlled mainly by porewater dissolved iron concentrations, the thickness of the surface oxygenated layer of sediment and to a lesser extent by pH and temperature. Under typical diagenetic conditions (pH < 8, porewater Fe 2+ = 10 −6 g cm −3 ) iron can diffuse from the porewaters in continental margin sediments to the oxygenated overlying seawater at fluxes of 100–1000 μg cm −2 a −1 . The addition of reactive iron to deep-basin sediments is determined by the magnitude of this diffusive flux, the export efficiency (ɛ) of recycled iron from the shelf, the ratio of source area ( S ) to basin sink area ( B ) and the trapping of reactive iron in the deep basin. Values of ɛ are poorly constrained but modern enclosed or semi-enclosed sedimentary basins show a wide variation in S/B ratios (0.25–13) where the shelf source area is defined as sediments at less than 200 m water depth. Diffusive fluxes in the range 100–1000 μg cm −2 a −1 are able to produce the observed reactive iron enrichments in the Black Sea, the Cariaco Basin and the Gotland Deep for values of ɛ × S/B from 0.1–5. Transported reactive iron can be trapped physically and/or chemically in deep basins. Physical trapping is controlled by basin geometry and chemical capture by the presence of euxinic bottom water. The S/B ratios in modern basins may not be representative of those in ancient euxinic/semi-euxinic sediments but preliminary data suggest that ɛ × S/B in ancient euxinic sediments has a similar range as in modern euxinic sediments.