In this thesis, the implementation and results of studies into the effect of pH on the kinetics of various iron transformations in natural waters are described. Specific studies include i) the oxidation of Fe(II) in the absence and presence of both model and natural organic ligands, ii) the complexation of Fe(III) by model organic compounds, and iii) the precipitation of Fe(III) through the use of both laboratory investigations of iron species and kinetic modeling. In the absence of organic ligands, oxidation of nanomolar concentrations of Fe(II) over the pH range 6.0 -- 8.0 is predominantly controlled by the reaction of Fe(II) with oxygen and with superoxide while the disproportionation of superoxide appears to be negligible. Oxidation of Fe(II) by hydrogen peroxide, back reduction of Fe(III) by superoxide and precipitation of Fe(III) have been shown to exert some influences at various stages of the oxidation at different pH and initial Fe(II) concentrations. In the presence of organic ligands, different effects on the Fe(II) oxidation kinetics is shown with different organic ligands, their initial concentrations and with varying pH. A detailed kinetic model is developed and shown to adequately describe the kinetics of Fe(II) oxidation in the absence and presence of various ligands over a range of concentrations and pH. The applicability of the previous oxidation models to describe the experimental data is assessed. Rate constants for formation of Fe(III) by a range of model organic compounds over the pH range 6.0 -- 9.5 are determined. Variation of rate constants for Fe(III) complexation by desferrioxamine B and ethylenediaminetetraacetate with varying pH is explained by an outer-sphere complexation model. The significant variation in rate constants of Fe(III) complexation by salicylate, 5-sulfosalicylate, citrate and 3,4-dihydroxylbenzoate with varying pH is possibly due to the presence of different complexes at different pH. The results of this study demonstrate that organic ligands from different sources may influence the speciation of iron in vastly different ways. The kinetics of Fe(III) precipitation are investigated in bicarbonate solutions over the pH range 6.0 -- 9.5. The rate of precipitation varies by nearly two orders of magnitude with a maximum rate constant at a pH of around 8.0. The results of the study support the existence of the dissolved neutral species Fe(OH)30 and suggests that it is the dominant precursor in Fe(III) polymerization and subsequent precipitation at circumneutral pH. Variation in the precipitation rate constant over the pH range considered is consistent with a mechanism in which the kinetics of iron precipitation are controlled by rates of water exchange in dissolved iron hydrolysis species.

In this thesis, the implementation and results of studies into the effect of pH on the kinetics of various iron transformations in natural waters are described. Specific studies include i) the oxidation of Fe(II) in the absence and presence of both model and natural organic ligands, ii) the complexation of Fe(III) by model organic compounds, and iii) the precipitation of Fe(III) through the use of both laboratory investigations of iron species and kinetic modeling. In the absence of organic ligands, oxidation of nanomolar concentrations of Fe(II) over the pH range 6.0 -- 8.0 is predominantly controlled by the reaction of Fe(II) with oxygen and with superoxide while the disproportionation of superoxide appears to be negligible. Oxidation of Fe(II) by hydrogen peroxide, back reduction of Fe(III) by superoxide and precipitation of Fe(III) have been shown to exert some influences at various stages of the oxidation at different pH and initial Fe(II) concentrations. In the presence of organic ligands, different effects on the Fe(II) oxidation kinetics is shown with different organic ligands, their initial concentrations and with varying pH. A detailed kinetic model is developed and shown to adequately describe the kinetics of Fe(II) oxidation in the absence and presence of various ligands over a range of concentrations and pH. The applicability of the previous oxidation models to describe the experimental data is assessed. Rate constants for formation of Fe(III) by a range of model organic compounds over the pH range 6.0 -- 9.5 are determined. Variation of rate constants for Fe(III) complexation by desferrioxamine B and ethylenediaminetetraacetate with varying pH is explained by an outer-sphere complexation model. The significant variation in rate constants of Fe(III) complexation by salicylate, 5-sulfosalicylate, citrate and 3,4-dihydroxylbenzoate with varying pH is possibly due to the presence of different complexes at different pH. The results of this study demonstrate that organic ligands from different sources may influence the speciation of iron in vastly different ways. The kinetics of Fe(III) precipitation are investigated in bicarbonate solutions over the pH range 6.0 -- 9.5. The rate of precipitation varies by nearly two orders of magnitude with a maximum rate constant at a pH of around 8.0. The results of the study support the existence of the dissolved neutral species Fe(OH)30 and suggests that it is the dominant precursor in Fe(III) polymerization and subsequent precipitation at circumneutral pH. Variation in the precipitation rate constant over the pH range considered is consistent with a mechanism in which the kinetics of iron precipitation are controlled by rates of water exchange in dissolved iron hydrolysis species.

This study evaluated alternatives to continuously regenerated greensand for iron and manganese removal. Specific objectives were (1) to investigate the applicability for removing soluble manganese by adsorption and oxidation onto the surface of manganese oxide-coated media, and (2) to develop mathematical models to predict the removal of soluble manganese both in the presence and absence of free chlorine. Results from a pilot-scale experiment in Columbus, Indiana, showed that when the filters were operated in a conventional oxidant addition mode (i.e., with the addition of HOCI and KMnO₄) the anthracite-sand and anthracite-sand-garnet configurations both provided efficient treatment because of the reduced rate of head loss. Further experiments, without the use of KMnO₄ or greensand, equally removed manganese by adsorption and oxidation onto oxide-coated media; however, the treatment process resulted in reduced head loss and oxidant costs. Results from this study and other previous research demonstrated that manganese removal by oxide coatings is an efficient and functional treatment mechanism. However, little quantitative information was available to predict these processes. Therefore, mathematical models were developed to help predict design and operational conditions needed to implement oxide-coated media as a treatment process. Two separate models were developed to predict (1) the continuous removal of soluble manganese in the presence of free chlorine (continuous regeneration model), and (2) the eventual breakthrough of soluble manganese without oxidant addition (intermittent regeneration model). Each model was derived from chemical reaction, mass balance, or isotherm equations and was further developed by a sensitivity analysis and parameter estimation. The two models were then verified by predicting manganese removal from independent research. The continuous regeneration model can help predict the removal of soluble manganese by adsorption and oxidation on the surface of oxide-coated media and is useful in the design of filters for continuous Mn(Il) removal. The intermittent regeneration model effectively predicts the performance of filters without the addition of an oxidant and is useful for treatment plants which cannot apply chlorine continuously to their filter applied water. / Master of Science

The fate of iron complexed by various organic compounds was investigated as a function of both chemical oxidative and coagulation removal methods. Dissolved organic carbon (DOC) utilized in the studies was obtained from a variety of sources and included humic and fulvic acids, tannic acid and oxalic acid. Oxidants evaluated were potassium permanganate, free chlorine, and chlorine dioxide. Both laboratory-scale and field monitoring studies were performed. The relative molecular weight distribution (MWD) of the DOC present was analyzed to evaluate how changes in this parameter affected the efficiency of soluble iron removal by oxidation. In addition, the MWD of selected coagulated samples was evaluated to determine how this parameter affected the fate of complexed iron during the coagulation of dissolved organic matter with alum. A high degree of ferrous iron complexation occurred with the DOC dominated by higher molecular weight organics. This complexation rendered the iron stable against the addition of each of the oxidants evaluated. However, soluble Fe(II) complexed by low molecular weight organics was successfully removed by chemical oxidation. Potassium permanganate was found to be the most effective oxidant of the three oxidants utilized in the study. The results indicated that soluble Fe(II) complexed by high molecular weight DOC can be efficiently removed by alum coagulation. The pH and alum dose utilized to produce effective DOC removal was also found to promote efficient complexed Fe(II) removal. / Master of Science