The iron oxide-catalyzed production of hydroxyl radical (*OH) from hydrogen peroxide (H2O2) has been used to oxidize organic contaminants in soils and groundwater. The goals of this study are to determine which factors control the generation rate of *OH (VOH) and to show that if VOH and the rate constants of the reactions of *OH with the system's constituents are known, the oxidation rate of a dissolved organic compound can be predicted. Using 14C-labeled formic acid as a probe, we measured VOH in pH 4 slurries of H2O2 and either synthesized ferrihydrite, goethite, or hematite or a natural iron oxide-coated quartzitic aquifer sand. In all of our experiments, VOH was proportional to the product of the concentrations of surface area of the iron oxide and H2O2, although different solids produced *OH at different rates. We used these results to develop a model of the decomposition rate of formic acid as a function of the initial formic acid and hydrogen peroxide concentrations and of the type and quantity of iron oxide. Our model successfully predicted the VOH and organic compound oxidation rates observed in our aquifer sand experiment and in a number of other studies but overpredicted VOH and oxidation rates in other cases, possibly indicating that unknown reactants are either interfering with *OH production or consuming *OH in these systems.

This work examines the contribution of solution phase reactions, especially those involving a chain reaction mechanism, to the decomposition of hydrogen peroxide (H2O2) and organic compounds in the presence of dissolved iron and ferrihydrite. In solutions at pH 4, where Fe was introduced as dissolved Fe(III), both H2O2 and 14C-labeled formic acid decomposed at measurable rates that agreed reasonably well with those predicted by a kinetic model of the chain reaction mechanism, using published rate constants extrapolated to pH 4. The ratio of the formic acid and H2O2 decomposition rates, as well as the dramatic effect of tert-butyl alcohol on these rates, confirmed that a solution chain reaction mechanism involving *OH controlled the decomposition kinetics of both compounds. In the presence of ferrihydrite as the iron source, the ratio of the rate of formic acid decomposition to that of H2O2 decomposition was significantly lower than that observed in the presence of only dissolved Fe. Moreover, neither rate diminished drastically upon addition of tert-butyl alcohol, indicating that the solution phase chain reaction is not a dominant decomposition pathway of H2O2 and formic acid. Relative decomposition rates of formic acid and a second *OH probe, benzoic acid, were consistent with oxidation of these compounds by *OH. These observations can be reproduced by a kinetic model including (a) decomposition of H2O2 at the iron oxide surface, producing *OH with lower yield than the reaction sequence with dissolved Fe, and (b) low concentrations of dissolved Fe in the presence of ferrihydrite, preventing propagation of the solution phase chain reaction.

Electrostatic effects influence the oxidation rates of charged dissolved organic compounds in systems where the hydroxyl radical (*OH) is produced by the iron oxide-catalyzed decomposition of hydrogen peroxide (H2O2). Experiments were performed using goethite and the *OH probes 14C-labeled formic acid, 2-chlorophenol (2-CP), and nitrobenzene. At pH 4 and an ionic strength of 0.01 M, formic acid (pKa = 3.745) detected a steady-state concentration of *OH ([*OH]ss, calculated as a solution average) approximately 50 times higher than the two neutral probes did in the same systems, indicating significant enrichment of formate at the surface of the positively charged iron oxide where the *OH is being produced. Increasing the pH and ionic strength decreased formic acid oxidation rates by factors consistent with predicted decreases in electrostatic effects. In the presence of high 2-CP concentrations, the [*OH]ss measured by formic acid decreased with time, and goethite coagulation increased, due to loss of positive charge on the oxide surface as the oxidation products of 2-CP complexed surface Fe species. The [*OH]ss detected by 2-CP did not change significantly, indicating that neither goethite coagulation nor surface complexation of 2-CP oxidation products interfered with the rate of *OH generation; however, such an effect could have occurred in experiments using dissolved Fe instead of goethite. Model predictions of organic compound oxidation rates in mineral-catalyzed Fenton-like systems were improved by taking electrostatic effects into account.