We studied the interdependent effects of phosphate, silicate and Ca on the formation of Fe(III)-precipitates by oxidation of 0.5 mM Fe(II) in near-neutral bicarbonate-buffered aqueous solutions at concentrations relevant for natural water resources. Complementary results obtained by a suite of analytical techniques including X-ray absorption spectroscopy and transmission electron microscopy showed that the ratio of initially dissolved phosphate over Fe(II) ((P/Fe)_init) had a major impact on precipitate formation. At (P/Fe)_init above a critical ratio ((P/Fe)_crit) of ∼0.5 in 8 mM NaHCO₃ and ∼0.8 in 4 mM Ca(HCO₃)₂ electrolyte, Fe(II) oxidation led to exclusive formation of amorphous basic Fe(III)-phosphate or Ca–Fe(III)-phosphate ((Ca-)Fe(III)-phosphate) with maximum precipitate P/Fe ratios ((P/Fe)_ppt) of ∼0.7 in Na and ∼1.1 in Ca electrolyte. Enhanced phosphate uptake in the presence of Ca was due to phosphate-Ca interactions coupled to Fe precipitation, mainly formation of mitridatite-like Ca–Fe(III)-phosphate polymers and Ca-phosphate polymers. At (P/Fe)_init < (P/Fe)_crit, in the absence of silicate, (Ca-)Fe(III)-phosphate precipitation was followed by the formation of poorly crystalline lepidocrocite and concomitant transformation of the (Ca-)Fe(III)-phosphate into a phosphate-rich ferrihydrite-type precipitate with a (P/Fe)_ppt of ∼0.25. In the presence of 0.5 mM silicate, initially formed (Ca-)Fe(III)-phosphate nanoparticles became coated with silicate-rich ferrihydrite during continuing Fe(II) oxidation and only limited transformation of the (Ca-)Fe(III)-phosphate occurred. The results from this study indicate the complexity of Fe(III)-precipitate formation in the presence of interfering solutes and its consequences for precipitate structure and phosphate sequestration. The findings provide a solid basis for further studies of the reactivity of different Fe(III)-precipitate types and for the systematic assessment of their impact on Fe, phosphate and trace elements dynamics in natural and engineered systems.

The oxidation and removal of As(III) by commercially available micro-scale zero-valent iron (mZVI) was studied in aerated synthetic groundwater with initially 6.7 μM As(III) at neutral pH values. Batch experiments were performed to investigate the influence of ZVI and H₂O₂ concentrations on As(III) oxidation and removal. Oxidation and removal kinetics was significantly increased by increasing ZVI concentration or by adding H₂O₂ in micromolar concentrations slightly higher than that of initial As(III). Observed half-lifes for arsenic removal without added H₂O₂ were 81–17 min at ZVI concentrations of 0.15–2.5 g/L, respectively. X-ray absorption spectroscopy (XAS) confirmed that almost all As(III) was converted to As(V) after 2 h of reaction in the pH range 5–9. Addition of 9.6 μM H₂O₂ to 0.15 g/L ZVI suspensions diminished half-lifes for arsenic removal from 81 to 32 min and for As(III) oxidation from 77 to 8 min, i.e., by approximately a factor of 10. The increased rate of As(III) oxidation is attributable to enhanced formation of oxidants by the Fenton reaction with higher initial concentrations of H₂O₂. In practice, results of this study suggest that addition of small amounts (<1 mg/L) of H₂O₂ in various forms (e.g. stable and widely available Na-percarbonate) to water prior to treatment could significantly enhance As(III) oxidation and removal with ZVI.

The main arsenic mitigation measures in Bangladesh, well-switching and deep tube wells, have reduced As exposure, but water treatment is important where As-free water is not available. Zero-valent iron (ZVI) based SONO household filters, developed in Bangladesh, remove As by corrosion of locally available inexpensive surplus iron and sand filtration in two buckets. We investigated As removal in SONO filters in the field and laboratory, covering a range of typical groundwater concentrations (in mg/L) of As (0.14–0.96), Fe (0–17), P (0–4.4), Ca (45–162), and Mn (0–2.8). Depending on influent Fe(II) concentrations, 20–80% As was removed in the top sand layer, but As removal to safe levels occurred in the ZVI-layer of the first bucket. Residual As, Fe, and Mn were removed after re-aeration in the sand of the second bucket. New and over 8-year-old filters removed As to <50 μg/L and mostly to <10 μg/L and Mn to <0.2 mg/L. Vertical concentration profiles revealed formation of Fe(II) by corrosion of Fe(0) with O₂ and incorporation of As into forming amorphous Fe phases in the composite iron matrix (CIM) of newer filters and predominantly magnetite in older filters. As mass balances indicated that users filtered less than reported volumes of water, pointing to the need for more educational efforts. All tested SONO filters provided safe drinking water without replacement for up to over 8 years of use.

Bangladesh relies heavily on groundwater for the irrigation of dry-season rice. However, the groundwater used for irrigation often contains high concentrations of arsenic, potentially jeopardizing the future of rice production in the country. In seasonally flooded fields, topsoil arsenic concentrations decrease during the monsoon season, suggesting that flooding attenuates arsenic accumulation in the soils. Here we examine the chemistry of soil porewater and floodwater during the monsoon season in rice paddies in Munshiganj, Bangladesh, to assess whether flooding releases significant quantities of arsenic from the soils. We estimate that between 51 and 250 mg m⁻² of soil arsenic is released into floodwater during the monsoon season. This corresponds to a loss of 13–62% of the arsenic added to soils through irrigation each year. The arsenic was distributed throughout the entire floodwater column by vertical mixing and was laterally removed when the floodwater receded. We conclude that monsoon floodwater removes a large amount of the arsenic added to paddy soils through irrigation, and suggest that non-flooded soils are particularly at risk of arsenic accumulation.

Corrosion of zerovalent iron (ZVI) in oxygen-containing water produces reactive intermediates that can oxidize various organic and inorganic compounds. We investigated the kinetics and mechanism of Fenton reagent generation and As(III) oxidation and removal by ZVI (0.1m²/g) from pH 3-11 in aerated water. Observed half-lives for the oxidation of initially 500 μg/L As(III) by 150 mg Fe(0)/L were 26-80 min at pH 3-9. At pH 11, no As(III) oxidation was observed during the first two hours. Dissolved Fe(II) reached 325, 140, and 6 μM at pH 3, 5, and 7. H₂O₂ concentrations peaked within 10 min at 1.2, 0.4, and <0.1 μM at pH 3, 5, and 7, and then decreased to undetectable levels. Addition of 2,2′-bipyridine (1-3 mM), prevented Fe(II) oxidation by O₂ and H₂O₂ and inhibited As(III) oxidation. 2-propanol (14 mM), scavenging OH-radicals, quenched the As(III) oxidation at pH 3, but had almost no effect at pH 5 and 7. Experimental data and kinetic modeling suggest that As(III) was oxidized mainly in solution by the Fenton reaction and removed by sorption on newly formed hydrous ferric oxides. OH-radials are the main oxidant for As(III) at low pH, whereas a more selective oxidant oxidizes As(III) at circumneutral pH.