EXPERIMENTAL EXAMINATION OF IRON PHOTOCHEMISTRY IN ANCIENT OCEANS

The extensive deposition of Banded Iron Formations (BIF) between 3.5 and 1.8 Ga is widely interpreted as evidence that Fe2+ was oxidized to Fe3+ in the shallow waters of Archean and Paleoproterozoic oceans. However, there is disagreement about the cause(s) of oxidation during this period of Earth history, when O₂ was rare in the atmosphere and bulk oceans. A key question is whether biology played a role in the oxidation. One possibility is oxidation by photosynthetically derived O₂ in localized “oxygen oases” (Cloud, 1973). Another is anaerobic oxidation by photoferrotrophy (Konhauser et al., 2002). A third possibility is photooxidation of Fe2+ (Cairns-Smith, 1978) by UV light, a “photogeochemical” process that does not involve biology at all. We aim to determine the viability of this non-biological hypothesis through a combination of photogeochemical experiments and modeling.

To assess the viability of the photooxidation hypothesis it is necessary to determine the wavelengths of light driving the reaction. Photooxidation of Fe2+ by light 300 - 420 nm was reported in early experiments (Braterman et al., 1983), and predicts a rate of oxidation that can account for BIF deposits. However, subsequent experimental work found no photooxidation above 254 nm (Konhauser et al., 2007), in which case oxidation rates are much lower. Preliminary prior work by our group similarly found no oxidation above 295 nm (Castleberry, 2007). The validity of these studies is unclear because they used Hg lamps which do not realistically simulate the solar spectrum. They are also difficult to compare because they all used different experimental designs and different (and in some cases unrealistic) solution compositions.

Our new work uses a Xe lamp that accurately simulates the solar spectrum, and solution compositions that approximate possible Archean open ocean seawater. We directly measure change in [Fe2+] and [Fe3+] and total [Fe] to completely characterize the fate of dissolved Fe. Preliminary results with full-spectrum light reveal a photooxidation rate of 5.7x10^-3 mM Fe/day, which extrapolates to global rates insufficient to account for BIF deposition. The next step is to incorporate spectral cut-off filters to pin-point the range of light driving the reaction, and better quantify its potential importance in different environments.