Iron-cyanide complexes in soil under varying redox conditions: speciation, solubility and modelling

The distribution of iron-cyanide complexes between ferrocyanide, [Fe-II(CN)(6)](4-), and ferricyanide, [Fe-III(CN)(6)](3-), in soils on contaminated sites depends on the redox potential, E-H. We carried out microcosm experiments in which ferrocyanide (20 mg l(-1)) was added to an uncontaminated moderately acidic subsoil (pH 5.2), and varied the E-H of the soil suspension between 200 and 700 mV over up to 109 days. Ferrocyanide and ferricyanide were analysed by capillary isotachophoresis. At redox potentials ranging from 400 to 700 mV, small amounts of iron-cyanide complexes were adsorbed, and ferrocyanide was almost completely oxidized to ferricyanide. Decreasing E-H to 200 mV led to nearly complete removal of iron-cyanide complexes from solution, and the complexes were not mobilized after subsequent aeration (E-H > 350 mV). Under weakly to moderately reducing conditions (E-H approximate to 200 mV), iron-cyanide complexes were removed from solution by precipitation, which occurred, presumably in the form of e.g. Fe-2[Fe-II(CN)(6)], Fe-4[Fe-II(CN)(6)](3) or Mn-2[Fe-II(CN)(6)], after reductive dissolution of Mn and Fe oxides. Four different sets of geochemical model calculations were carried out. The species distribution between ferrocyanide and ferricyanide in solution was predicted reliably under varying pH and redox conditions when iron-cyanide complex concentrations and Fe concentrations, excluding Fe bound in iron-cyanide complexes, were used in model calculations. In model calculations on the fate of iron-cyanide complexes in soil, adsorption reactions must be considered, especially under oxidizing conditions. Otherwise, the calculated iron-cyanide complex concentrations are larger than those actually measured.