A comprehensive kinetic model for phenol oxidation in seven advanced oxidation processes and considering the effects of halides and carbonate

1. Introduction

Advanced oxidation processes (AOPs) are among the most important water treatment technologies for organic contaminant removal. They often rely on the activation of oxidants, such as peroxymonosulfate (PMS), peroxydisulfate (PDS), and H2O2, by either energy inputs or catalysts to induce the formation of reactive oxygen species (ROSs), including sulfate (SO4•−) and hydroxyl (•OH) radicals (Lee et al., 2020a; Wang and Wang, 2018). Recent studies also reported the involvement of singlet oxygen (1O2) or direct electron transfer processes in some of the AOP systems (Duan et al., 2018b; Huang and Zhang, 2019). Utilizing the unique long-distance electron transfer property of the prepared catalyst, recent studies developed a “Galvanic oxidation process” that achieved contaminant oxidation even when PMS and the contaminant were physically separated from each other (Huang and Zhang, 2019; 2020a). The reaction mechanisms in AOPs are usually complex as the reactions between ROSs and organic contaminants generate many intermediates/products, which can further react with the ROSs. The persistence and risks of these intermediates/products may also vary significantly. Moreover, various impact factors, such as the coexisting anions, exist in real water treatment scenarios, as mentioned later. These anions can sometimes significantly alter the reaction mechanisms and impact the contaminant removal efficiency, oxidant utilization efficiency, and product formation. Therefore, advanced experimental, computational, and modeling approaches are needed to investigate these AOPs.

Có thể bạn quan tâm

Kinetic modeling is a powerful approach to gaining mechanistic understandings of these systems and can help formulate strategies for improving the treatment efficiency. Extensive modeling studies have been reported for many AOPs, such as H2O2/UV, PDS/UV, chlorine/UV, and chloramine/UV (Chuang et al., 2017; Crittenden et al., 1999; Grebel et al., 2010; Yang et al., 2014; 2016). Some best examples include 180 reactions for an H2O2/UV system (Grebel et al., 2010), 188 reactions for H2O2/UV and PDS/UV systems (Yang et al., 2014), 140 reactions for H2O2/UV and PDS/UV systems (Zhang et al., 2015), and 203 reactions for HOCl/UV, H2O2/UV, and O3/UV systems (Bulman et al., 2019). A number of studies have also considered intermediate transformation in their kinetic models (Duesterberg and Waite, 2006; C.K. 2007; Guo et al., 2014b; Kang et al., 2002; Li et al., 2004; K. 2007; Luo et al., 2021; Mora et al., 2011; Qian et al., 2016; Zhao et al., 2021; Zhou et al., 2018). For example, Li et al. performed a kinetic modeling of trichloroethene degradation considering four intermediates and 67 reactions in one system (H2O2/UV) (K. Li et al., 2007). Duesterberg and Waite modeled the oxidation of p-hydroxybenzoic acid in the H2O2/Fe system with 49 reactions and considering one intermediate (3,4-dihydroxybenzoic acid) (C.K. Duesterberg and Waite, 2007). However, most of these models were built on limited numbers of systems under well-defined experimental conditions, so they cannot be readily applied to other systems or conditions. New models therefore need to be developed every time for new systems. This is time consuming and labor intensive. Furthermore, a large portion of the reaction rate constants in some studies were set as the fitting parameters (e.g., 12 out of 67 (K. Li et al., 2007), or 13 out of 101 (Qian et al., 2016)). The number of these fitting parameters is generally much larger than that of the datasets (experimental conditions) used for the model fitting, which may cause overparameterization. This is similar to solving multivariable equations, where a unique solution is only possible when the number of the equations is equal to or larger than that of the variables. If such a condition is not met, the true values of the variables (or, in this case, the unknown rate constants) cannot be obtained. Therefore, perfect fittings under the studied conditions can be easily obtained but poor fittings are usually expected for other conditions because the fitted rate constants are likely not accurate. To overcome both the applicability limitation and overparameterization issues, we need to build a robust and comprehensive model considering multiple systems and experimental conditions.

The performance of AOPs is also known to be affected by halides (including Cl− and Br−) or carbonate (collectively referred to as “the anions” hereafter), often through generating different reactive species. For example, Grebel et al. demonstrated that carbonate significantly decreased the phenol degradation rate in an H2O2/UV system because the reaction of •OH with carbonate forms less reactive carbonate radicals (CO3•−). Such an effect is much stronger than that of Cl− but weaker than that of Br− (Grebel et al., 2010). In PMS-based systems, halides have reportedly accelerated the contaminant removal because the reactions of PMS with halides form non-radical reactive halogen species (RHSs) (e.g., HOX, X2, X = Cl or Br) (Fang et al., 2016; Luo et al., 2019; Wang et al., 2017). These non-radical RHSs can sometimes be very abundant compared to the original ROSs and result in significantly faster contaminant oxidation, although they may also lead to inhibited mineralization and formation of toxic halogenated byproducts (Fang et al., 2016; Luo et al., 2019; Wang et al., 2017). Given the ubiquitous presence of these anions in the aquatic environment, it is important to systematically evaluate their impacts on the performance of different AOPs under various conditions. However, current studies considering anions only focused on very limited numbers of oxidants, activation approaches, and reaction mechanisms in each model. The inclusion of additional oxidants/activation approaches may require many more reactions which have not been previously considered.

To address the above limitations, a comprehensive model needs to be developed for a large number of systems covering multiple known oxidants and activation mechanisms. It can also be easily applied to, for instance, compare the performance of different systems side by side or simulate scaled-up water treatment systems under various conditions without having to conduct experimental tests. When hybrid systems are employed in some water treatment scenarios where multiple reactants and/or activation mechanisms are involved (e.g., PDS and H2O2 coupled with UV, catalyst, and heat (Monteagudo et al., 2015)), this comprehensive model would be very powerful to describe the systems and provide mechanistic insights.

In this work, four PMS, four PDS, and one H2O2 systems were first experimentally evaluated for phenol oxidation in both the absence and presence of the anions. Then a comprehensive kinetic model was developed for 7 AOPs considering different oxidants, activation approaches, reaction mechanisms, and other impact factors. The HOX/UV systems were also included in the model because they existed under some conditions (e.g., the PMS/UV system containing Cl− and Br− resulting in HOCl/UV and HOBr/UV systems). The intermediate transformation was also modeled. Phenol was selected as the model compound because it is widely used in industries and is toxic to aquatic life (Olmez-Hanci and Arslan-Alaton, 2013). Although phenol can be relatively easily oxidized by AOPs, it is the most basic phenolic compound and shares significant similarity with many other contaminants. Using phenol would be, therefore, helpful to develop a model that can be easily adopted for a wide range of contaminants. Further, recent advances in tools for oxidation pathway generation (Guo et al., 2014a, 2014b) and reaction rate constant prediction (e.g., quantitative structure-activity relationship (QSAR) models (Zhong et al., 2020a, 2020b)) can also greatly expand the application of this model to other organics. Extensive model tuning and validation were then carried out by fitting the model against the experimental data from both this study and the literature. Detailed balancing was performed to fix the reaction loop illegality issue, and the model sensitivity analysis was conducted to evaluate the importance of each reaction. With the developed model, the effect of anions on the reaction mechanisms and the importance of different ROSs responsible for phenol degradation and product formation, were finally evaluated.

Rate this post

Notice: ob_end_flush(): failed to send buffer of zlib output compression (1) in /home/tfbjrndghosting/public_html/hotdealhcm.com.vn/wp-includes/functions.php on line 5373

Notice: ob_end_flush(): failed to send buffer of zlib output compression (1) in /home/tfbjrndghosting/public_html/hotdealhcm.com.vn/wp-includes/functions.php on line 5373