The reaction of ferric (hydr)oxides with dissolved sulfide does not lead to the instantaneous production of thermodynamically stable products but can induce a variety of mineral transformations including the formation of metastable intermediates. The importance of the various transformation pathways depends, among other factors, on the characteristics of the ferric (hydr)oxides but a mechanistic model which relates the mineralogy of the ferric (hydr)oxides to the type of reaction products and their evolution over time is still missing. Here, we investigate the kinetics of the reaction between dissolved sulfide (6.7–7.5mmolL−1) with ferrihydrite (Fh, 12mmolL−1), lepidocrocite (Lp, 26.6mmolL−1), and goethite (Gt, 22mmolL−1) in batch experiments at pH7 and room temperature. The time evolution of solution and solid phase composition was monitored over 2weeks while TEM, and Mössbauer spectroscopy were used to characterize the transformations of the solid phases.Dissolved sulfide was consumed within 2 (Fh, Lp) to 8h (Gt) with methanol extractable sulfur and HCl extractable Fe(II) (Fe(II)HCl) being the main products after this time. The mass balances of Fe and S indicated that a large fraction of the Fe(II)HCl in the reactions with Fh (46% of Fe(II)HCl) and Lp (36% of Fe(II)HCl) was solid-phase bound but not associated with sulfur. This excess Fe(II) exceeded the adsorption capacity of the solids and remained associated with the oxides. Over the time scale of days, the concentrations of MES and Fe(II)HCl decreased and this process was accompanied by the formation of secondary iron oxides and pyrite in all experiments. The pyrite yield after two weeks showed the same trend as the amounts of intermediately produced excess Fe(II): Fh (84% of initial S(−II))>Lp (50%)>Gt (13%). Besides the formation of pyrite, Fh transformed completely into thermodynamically more stable iron oxides such as hematite or magnetite. In contrast, formation of other iron oxides was only minor when Lp or Gt reacted with sulfide.We propose that the extent of pyrite and secondary iron mineral precipitation is controlled by the ratio between the competing formation rates of excess Fe(II) and surface bound FeS (FeSs) in the early stage of the reaction. Formation of excess Fe(II) is a prerequisite for rapid pyrite formation and induces secondary formation of iron oxides. The competition between excess Fe(II) and FeSs formation, in turn, is ruled by two factors: 1) the ratio between added sulfide and available surface area, and 2) the capability of the iron(hydr)oxide to conduct electrons from surface bound Fe(II) to bulk Fe(III) and to accommodate structural Fe(II). This capability is largest for Fh and explains the most pronounced excess Fe(II) production and, by this, the greatest pyrite yield in experiments with Fh. During the reaction with Gt, in contrast, formation of FeSs outcompetes the accumulation of excess Fe(II) and consequently the precipitation of pyrite is only minor.This conceptual model constrains conditions at which relatively fast pyrite formation within the time scale of days or weeks might be relevant in natural environments. Suitable conditions are expected in environments with low sulfide levels in which formation of reactive iron (hydr)oxides is stimulated by redox oscillations (e.g., wetlands, riparian soils, tidal flats).

The interaction between S(-II) and ferric oxides exerts a major control for the sulphur and iron cycle and in particular for the carbon and electron flow in many aquatic systems. It is regarded to be a key reaction leading ultimately to pyrite formation, the pathways still remaining unresolved. We have studied the reaction between lepidocrocite (γ-FeOOH, 21–42mmolL−1) and dissolved S(-II) (3–9mmolL−1) in batch experiments at pH 7 in a glove box using TEM, XRD, Mössbauer spectroscopy, and wet chemistry extraction to explore the nanocrystalline products forming at different time steps in close contact to the lepidocrocite surface. S(0) and acid extractable Fe(II) (Fe(II)HCl) were the main products detected by wet chemistry extraction. The reaction could be divided into three steps: a rapid (<15min) consumption of dissolved S(-II), formation of S(0) and the build-up of an Fe(II)HCl pool. Then in the absence of dissolved S(-II) concentrations of S(0) and Fe(II)HCl increased only slightly. TEM measurements revealed the occurrence of a mackinawite rim covering the lepidocrocite crystals and being separated from the lepidocrocite surface by an interfacial magnetite layer that can be regarded as a steady state product of the interaction between lepidocrocite and mackinawite. A significant fraction of Fe(II) was formed in excess to FeS within the first 2h. The amount of this fraction increased with decreasing ratio between dissolved S(-II) concentration and the concentration of surface sites, which we attributed to a kinetic decoupling of S(-II) oxidation and Fe(II) detachment from the lepidocrocite surface. At low ratios, S(-II) seems to transfer electrons to lepidocrocite faster then stoichiometric amounts of FeS could. After 2days Fe(II)HCl and S(0) started to decrease resulting in pyrite formation accompanied by traces of magnetite. TEM measurements indicated that mackinawite completely dissolved and precipitation of pyrite occurred dislocated from the lepidocrocite surface. The absence of dissolved sulphide under these conditions suggest that excess Fe(II) is involved in the formation of polysulphides which are key precursors during pyrite formation. We propose that the occurrence of excess Fe(II) is a common phenomenon particularly in low sulphide – high iron environments attributing significant reactivity to ferric (hydr)oxides.

Reduction of ferric oxides by dissolved sufide species plays an important role in the redox dynamics of anaerobic subsurface environments. For example, S(-II) may be the dominant reductant for ferric oxides in sulfidic sediments and the reaction significantly contributes to the retention of reduced sulfur in these environments. The mechanism and rates of the reductive dissolution of ferric oxides by S(-II) have been frequently investigated and the reaction is generally considered to be a surface controlled process. The proposed mechanism includes the reduction of Fe(III) at the mineral surface upon S(-II) adsorption followed by the release of Fe(II) into solution, which is eventually followed by the precipitation of FeS at suitable conditions. However, several reported experimental observations, such as the remarkable high rate of lepidocrocite (g-FeOOH) reduction and the accumulation of surface bound Fe(II) (Poulton et al., 2004) are difficult to align with the proposed reaction mechanism. Here, we report experimental results which indicate that the prevailing idea about a surface controlled reaction between lepidocrocite and S(-II) has to be revised.