『Abstract
　Fe released into solution is isotopically lighter (enriched in the lighter isotope) than hornblende starting material when dissolution occurs in the presence of the siderophore desferrioxamine mesylate (DFAM). In contrast, Fe released from goethite dissolving in the presence of DFAM is isotopically unchanged. Furthermore, Δ⁵⁶Fesolution-hornblende for Fe released to solution in the presence of ligands varies with the affinity of the ligand for Fe. The extent of isotopic fractionation of Fe released from hornblende also increases when experiments are agitated continuously. The Fe isotope fractionation observed during hornblende dissolution with organic ligands is attributed predominantly to retention of ⁵⁶Fe in an altered surface layer, while the lack of isotopic fractionation during goethite dissolution in DFAM is consistent with the lack of an altered layer. When a siderophore-producing soil bacterium is added to the system (without added organic ligands), Fe released to solution from both hornblende and goethite differs isotopically from Fe in the bulk mineral: Δ⁵⁶Fesolution-starting material ＝−0.56±0.19 (hornblende) and −1.44±0.16 (goethite). Increased isotopic fractionation is attributed in this case to the fact that as bacterial respiration depletes the system in oxygen and aqueous Fe is reduced, equilibration between aqueous ferrous and ferric iron creates a pool of isotopically heavy ferric iron that is assimilated by bacterial cells. Adsorption of isotopically heavy ferrous iron (Fe(II) enriched in the heavier isotope) or precipitation of isotopically heavy Fe minerals may also contribute to observed fractionations.
　To test whether these Fe isotope signatures are recorded in natural systems, we also investigated extractions of samples of soils from which the bacteria were isolated. These extractions show variability in the isotopic signatures of exchangeable Fe and Fe oxyhydroxide fractions from one soil sample to another, but exchangeable Fe is observed to be lighter than Fe in soil Fe oxyhydroxides and hornblende. This observation is consistent with isotopically light Fe-organic complexes in soil pore water derived from the Fe-silicate starting materials in the presence of growing microorganisms, as documented in experiments reported here. The contributions from phenomena including organic ligand-promoted nonstoichiometric dissolution of Fe silicates, uptake of ferric iron by organisms, adsorption of isotopically heavy ferrous iron, and precipitation of iron minerals should create complex isotopic signatures in soils. Better understanding of these processes and the timescales over which they contribute to fractionation is needed.』

1. Introduction
2. Methods
　2.1. Experimental setup
　　2.1.1. Medium and bacteria
　　2.1.2. Hornblende experiments
　　2.1.3. Goethite experiments
　2.3. Analysis
　　2.3.1. Concentrations
　　2.3.2. Fe isotopes
　2.4. Soil extractions
3. Results
　3.1. Aqueous chemistry vs. Time
　3.2. Isotopic signature of starting materials
　3.3. Isotopic measurement of aqueous Fe
　3.4. Soil extractions
4. Discussion
　4.1. Siderophores and dissolution
　4.2. Fe redox chemistry
　4.3. Fe isotope fractionation
　　4.3.1. Abiotic hornblende experiments
　　　4.3.1.1. Incongruent dissolution of bulk mineral
　　　4.3.1.2. Precipitation/adsorption
　　　4.3.1.3. Preferential dissolution from a leached layer
　　4.3.2. Goethite experiments
　　4.3.3. Weathering
5. Conclusions
Acknowledgments
References