Combined Fenton's oxidation and biodegradation for the treatment of pentachlorophenol-contaminated water

Abstract

Combinations of chemical and biological degradation processes can be more effective for the treatment of recalcitrant pollutants than either chemical or biological treatment alone. Many chemical oxidation technologies rely on the generation of hydroxyl free radicals (•OH), and can partially oxidize recalcitrant compounds and generate more biodegradable intermediates. In a laboratory-scale project, pentachlorophenol (PCP) was studied as a model recalcitrant compound for a sequential combined system of chemical oxidation and biodegradation. For this combination, Fenton's reagent was selected as the oxidation method. The combined treatment was achieved in a continuous-flow stirred-tank Fenton's reactor and a packed-bed bioreactor. The goal of this research was to rationally optimize the combined system. Kinetic and stoichiometric data for chemical oxidation and biodegradation were measured in separate studies for both the chemical and the biological processes, and then integrated into a combined reactor model. The mathematical model was used to provide an understanding of the combined system and to test two project hypotheses: (i) the combined system can achieve higher mineralization than a single step process, and (ii) recycle of the effluent from the bioreactor back to the chemical reactor can improve the efficiency of the mineralization. To characterize the •OH-mediated degradation of PCP, the second-order rate constant for PCP was measured by the method of competitive kinetics. The value obtained for PCP (4.4x1009 L/mols), together with experimental and previously reported values for other chlorophenols (which were all higher than the value for PCP) were correlated by means of a group contribution method (Hammett's equation), the number of chlorine atoms in the aromatic ring, and also with estimated diffusion coefficients. Since correlation of kinetic data might be indicative of common reaction mechanisms, this analysis provided insight in generalizations about the mechanism of the reaction between •OH and chlorophenols. It was found that all of these correlations were equally valid, which indicates that the very fast rate of reaction between •OH and chlorophenols might be diffusion-limited, instead of limited by the reaction mechanism. Thus, generalizations on the reaction mechanism of these reactions based exclusively on correlations are limited. Data from batch experiments revealed strong dependence of PCP degradation on the hydrogen peroxide dose, and chloride production concomitant to the PCP degradation. Hydrogen peroxide doses of 2.5x10-04 mol/L achieved almost complete degradation of PCP, and lower hydrogen peroxide doses achieved partial degradation of the contaminant. However, the total carbon concentration (TOC) of the solution was not decreased at any of the hydrogen peroxide concentrations tested; i.e. mineralization did not occur. Different PCP intermediates were observed, but only tetrachlorohydroquinone could be identified. A kinetic model for the degradation of PCP under Fenton' s reaction was developed, based on previously reported reactions and rates among all the relevant chemical species (ferrous and ferric iron, hydrogen peroxide, and different free radicals) as well as the reaction rate of the reaction of PCP and hydroxyl free radicals, accounting for scavenging effects of the PCP by-products on the degradation of PCP. The model was validated and applied to predict the experimental time-course concentrations of hydrogen peroxide and ferrous iron, in the absence and the presence of PCP as a hydroxyl free radical scavenger, and also the experimental time-course and equilibrium concentrations of PCP oxidized with hydrogen peroxide and ferrous iron in batch reactors. The model was in good agreement with data obtained during time-course reactions (which in most cases reached completion in the first few minutes) and also with equilibrium concentrations of PCP. Treatment of PCP-contaminated solution in a continuous Fenton's reactor (operating at a retention time of 1.5 h) showed similar behavior to the batch experiments: PCP degradation and dechlorination were strongly dependent on the hydrogen peroxide dose, although reaction extents were lower than in batch experiments. No reduction of TOC was observed. Previously observed intermediates during batch experiments were produced in the continuous Fenton's reactor. Further treatment of the effluent from Fenton's with a continuous bioreactor that included a packed bed column (operating at a retention time of 5.5 h) showed that TOC was biodegraded, although no further reduction of PCP or dechlorination in this reactor were achieved. These biodegradation extents were consistent with the measured production of biomass by this reactor. A small part of the biodegradation observed in the bioreactor was attributable to the biodegradation of the observed intermediates. Experiments in which the retention time in the bioreactor was doubled did not yield significantly different biodegradation results. Recycle of the effluent from the bioreactor to the chemical reactor improved the TOC degradation, but not the extent of the PCP degradation or dechlorination. The Fenton's kinetic model for the degradation of PCP was applied to the continuous Fenton's reactor. Additionally, kinetic information about the biodegradation process was obtained from the continuous operation of the bioreactor and incorporated into a biodegradation model. The two models were integrated into a combined model, which also accounted for recycle effects. The model over-predicted the degradation of PCP and TOC in the combined system, but adequately predicted the sensitivity of PCP and TOC degradation achieved by the combined system at different hydrogen peroxide doses and recycle rates. Given these considerations, the model provided information supporting the two project hypotheses. A suggested reactor operating strategy to improve the system performance was to increase the recycle rate to improve the mineralization of the waste. The model also indicated that the effect of recycle rates was highly dependent on the microbial degradation kinetic parameters. As opposed to many of the previous empirical or semiempirical studies on combinations of chemical oxidation and biodegradation, the approach followed in this work provides basic understanding of the combined system, and provides an alternative tool for process design and optimization.