Prediction of Oxidation Efficiency and Byproduct Formation during Ultraviolet-based Advanced Oxidation Processes

Abstract

Over the last few decades, the presence of micropollutants in various water environments has become an important issue. One of the main concerns is their effect on the quality of drinking water sources. Because micropollutants cannot be removed effectively during conventional municipal wastewater treatment plants (WWTPs), the quality of the drinking water sources can be deteriorated. Advanced oxidation processes (AOPs) have gained attention from the perspective of treating micropollutants. The aim of this doctoral dissertation was to predict the oxidation efficiency and formation of byproducts during ultraviolet (UV)-based oxidative water treatment processes. This study was focused on the following four sub-topics: i) understanding of the photochemistry of chlorine and bromine species for intrinsic quantum yield, oxidation efficiency, and oxyhalide formation, ii) investigation of bromate (BrO3−) formation during photolysis of chlorine species during the treatment of bromide (Br−)-containing water, iii) development of a computer-based prediction platform for •OH reactivity with organic compounds, and iv) assessment of the elimination efficiency of halogenated disinfection byproducts (DBPs) in a UV/H2O2 treatment process. In the first part, UV254 photolysis of free available chlorine and bromine species was investigated. The intrinsic quantum yields for •OH and X• (X = Cl or Br) generation were determined by model fitting of formaldehyde formation using a tert-butanol assay, and found to be 0.61/0.45 for HOCl/OCl− and 0.32/0.43 for HOBr/OBr−, respectively. The steady-state •OH concentration in UV/HOX was higher than that in UV/OX− by a factor of 23.3 and 7.8 for Cl and Br, respectively. This was attributed to the different •OH consumption rate by HOCl versus OCl−, while for HOBr/OBr− both the •OH formation and consumption rates were implied. This was supported by a k of 1.4×108 M−1 s−1 for the •OH reaction with HOCl, which was >14 times lower than the k for •OH reactions with OCl−, HOBr, and OBr−. Formation of chlorate (ClO3−) and BrO3− was found to be significant with apparent quantum yields of 0.12 − 0.23. A detailed mechanistic study on the formation of XO3− (including a new pathway involving XO•) is presented, which has important implications because the level of XO3− can exceed the regulation (BrO3−) or guideline (ClO3−) values during UV/halogen oxidant water treatment. The new kinetic models well simulate the experimental results for halogen oxidant decomposition, probe compound degradation, and formation of ClO3− and BrO3−. In the second part, the formation of BrO3− was investigated during UV254 photolysis of chlorine in the presence of Br− (UV/chlorine/Br−). In the UV/chlorine/Br− system, the molar and quantum yields for the formation of BrO3− were linearly correlated with the initial concentration of Br− because the formation of BrO•, precursor of BrO3−, is proportional to the [bromine]/[chlorine] ratio. In particular, the in situ generation rate of bromine species can be a rate-limiting step for BrO3− formation when the pH is basic (9 – 10). In this case, the apparent k values for the reaction of chlorine with Br− were relatively low (48 and 5 M−1 s−1 at pH 9 and 10, respectively). In the presence of dissolved organic matter (DOM), the formation of BrO3− can be reduced below the regulation value (10 ppb) with careful adjustment of the UV fluence, even if the concentration of Br− is high (1 ppm). A new formation pathway for BrO3− is proposed, including interhalogen reactions for XOx species. In the third part, a report is given of the development of a computer-based prediction platform for the kinetics of the •OH reaction with organic compounds. Second-order rate constants of the •OH reaction with various types of organic compounds (k•OH ) can be predicted using the newly developed prediction platform. A combined group contribution method (GCM) model was proposed and achieved by the integration of two GCM models to predict "k" _•"OH" for a wider range of organic compounds and to improve prediction accuracy. A database of k for 161 micropollutants was compiled, predicted, and compared (experimental k•OH vs. predicted k•OH ): 72% of the data were found to be within a factor of 0.5 – 2 in relation to the experimentally measured data. The new platform could also predict the degree of micropollutants elimination using a kinetic model based on the predicted k•OH and water quality parameters. The degree of elimination of 50 micropollutants during UV/H2O2 treatment could be successfully predicted using this platform (r2 = 0.94). In the last part, the elimination efficiency of methylparaben (MeP) and halogenated methylparabens (halo-MePs) during UV/H2O2 treatment was investigated. Second-order rate constants for the reaction of •OH (k•OH) were in the range 2.3 – 4.3×109 M−1 s−1 for the halo-MePs. During the UV/H2O2 (10 mg L−1) treatment, the elimination rates of MeP and its halo-MePs were more than 90% at a UV fluence of 1500 mJ cm−2. In particular, Br-containing halo-MePs, which have high UV reactivity, showed more than 90% elimination by UV alone treatment (1500 mJ cm−2). A kinetic model based on the photochemical and kinetic parameters obtained from this study could successfully predict the degree of elimination of MeP and its halo-MePs during UV/H2O2 treatment. This dissertation provides a kinetic and mechanistic model framework developed by understanding of basic and/or intrinsic (photo)chemistry of UV-based AOPs and related radical reactivity with organic compounds. This information can be useful for prediction of the elimination degree of micropollutants and the formation of byproducts during treatment.