Cu(II) removal by Anoxybacillus flavithermus–iron oxide composites during the addition of Fe(II)aq

Abstract

There is currently poor understanding of metal removal by composites of bacteria and iron oxide minerals, even though they commonly co-occur and are among the most important sorbents in near-surface fluid–rock environments. This study evaluated Cu removal by composites of Anoxybacillus flavithermus and iron oxide over time during the addition, oxidation, and hydrolysis of Fe(II)aq and precipitation of the mineral, in comparison to Cu removal in the two single-sorbent end-member systems. In the absence of iron oxide, Cu removal by A. flavithermus was well described by a previously published surface complexation model, after inclusion of additional reactions describing aqueous complexation by exudate ligands released by the bacteria. In the absence of bacterial cells, Cu removal by iron oxide synthesized in the presence of the bacterial exudate ligands demonstrated the formation of ternary surface complexes. Removal of Cu by the A. flavithermus–iron oxide composites was ca. 20% greater than the prediction based on assumption of additivity in the two end-member systems. This non-additive behavior was attributed to (1) progressive physical blockage of bacterial surface sites by the iron oxide particles, (2) physical blockage of adsorption sites as a result of self-aggregation of the iron oxide particles, and (3) the reduction of Cu(II) to Cu(I) at the bacterial cell surface, as demonstrated by X-ray absorption spectroscopy. The extent of reduction of Cu(II) to Cu(I) was proportional to the concentration of solid phase Fe(II), suggesting that iron oxidation and copper reduction are linked. This study has shown that Cu removal by bacteria–iron oxide composites is greatly affected by redox processes such as Cu(II) reduction on the cell surface both by other bacterial surface ligands and the oxidation of sorbed Fe(II), as well as Fe(II) redox interactions, and aging effects of the mineral (i.e. surface site masking).