Oxidative Treatments of Pesticides in Rainwater Runoff by HOCl, O 3 , and O 3 /H 2 O 2 : Effects of pH, Humic Acids and Inorganic Matters

: This study systematically investigated the oxidative treatment of ﬁve selected pesticides, alachlor (ALA), carbendazim (CAR), diuron (DIU), pyrimethanil (PYR), and tebuconazole (TEB), by comparing their relative reactivities as a function of three different oxidative treatment processes (i.e., chlorine (HOCl), ozone (O 3 ), and ozone/hydrogen peroxide (O 3 /H 2 O 2 )) under various oxidant dosages, reaction times, and pH conditions. For oxidative treatment, pesticide standards were spiked into rainwater. The removal efﬁciency of the selected pesticides varied considerably depending on the oxidative treatment processes. HOCl, O 3 , and O 3 /H 2 O 2 treatments were highly effective at eliminating CAR (>80%) and PYR (>99%), while they were not signiﬁcantly effective in removing TEB (<20%). In the case of DIU, HOCl (81%) was shown to be more effective than O 3 (24%) and O 3 /H 2 O 2 (49%). The removal efﬁciency of ALA was in the order of O 3 /H 2 O 2 (49%) > O 3 (20%) > HOCl (8.5%). The effect of increasing the solution pH from 5.0 to 9.0 on pesticide degradation varied between the oxidative treatment processes. Additionally, NH 4+ , NO 2 − , and humic acid in rainwater signiﬁcantly inhibited pesticide degradation.


Introduction
Pesticides are commonly used to enhance crop production and prevent plant diseases, such as fungal infections [1]. The United States Environmental Protection Agency reported that the global application of pesticides totaled more than 2 million tons and over USD 39 billion in 2007 [2]. Although most pesticide contamination occurs due to pesticide use in agricultural fields, the use of pesticides in urban areas has gained further attention owing to their potential risk to the environment and human health (e.g., mutagenicity, toxicity, and carcinogenicity) [3]. Several studies have demonstrated the presence of pesticides in urban rainwater globally [4,5]. Alachlor (ALA) and diuron (DIU) were detected in rainwater samples in Strasbourg, France, with maximum concentrations of 5590 and 1025 ng/L, respectively [4]. Tebuconazole (TEB) was detected in Brunswick Land (Germany) in concentrations ranging between 12 and 187 ng/L [5] (Supplementary information, Table S1).
Water shortage has become a significant global problem due to climate change, rapid population growth, and urbanization. There is an increasing demand for alternative water sources, such as rainwater, to secure a sustainable water supply [6]. In recent years, rainwater has attracted attention as an urban water resource for plant irrigation, flushing toilets, laundry, and cleaning [7,8]. However, a wide range of micropollutants, including pesticides, pharmaceuticals, and endocrine-disrupting chemicals, have been detected in Rainwater samples (n = 4) were collected within initial precipitation (within 4 h after rainfall) between July and September 2019 from the engineering building rooftop in Kangwon National University (Chunchen-si, Republic of Korea; latitude 37 • 52 05 N, longitude 127 • 44 19 E) using acid-cleaned funnels (45 cm diameter) connected to a polyethylene bottle. Chuncheon-si is located approximately 110 km to the east of Seoul (capital city), Republic of Korea. The rainwater samples were immediately filtered using cellulose nitrate membranes (Sartorius Stedim Biotech, Goettingen, Germany) with a pore size of 0.45 µm and stored in amber glass bottles (2 L) at 4 • C.

Experimental Procedures
Oxidant consumption experiments for HOCl and O 3 were conducted for rainwater, where each oxidant was applied at a dosage of 40 µM at pH 6.6 in 20 mL reaction volume. Samples (5 mL) were collected to measure residual oxidants at several time intervals: 0.5, 1, 2, 5, 10, and 30 min. HOCl, O 3 , and O 3 /H 2 O 2 oxidation experiments were carried out on a bench-scale using rainwater at pH 6.6 and the range of specific O 3

Experimental Procedures
Oxidant consumption experiments for HOCl and O3 were conducted for rainwater, where each oxidant was applied at a dosage of 40 μM at pH 6.6 in 20 mL reaction volume. Samples (5 mL) were collected to measure residual oxidants at several time intervals: 0.5, 1, 2, 5, 10, and 30 min. HOCl, O3, and O3/H2O2 oxidation experiments were carried out on a bench-scale using rainwater at pH 6.6 and the range of specific O3 values (0.17-0.20 mg/s; O3 generator, LAB-II, Ozone Tech, Daejeon, Republic of Korea) and HOCl (oxidant/dissolved organic carbon (DOC) = 10, 20, 30, and 40 μM) and H2O2 (molar ratio = 0.5) dosages at room temperature (25 ± 1 °C) (Figure 1). Each pesticide was spiked into a 25 mL reaction bottle at a concentration of 2 μM. For the O3/H2O2 experiment, a stock solution of H2O2 was added before adding the O3. At the end of the reaction time (30 min), the samples were immediately quenched with 50 μL of thiosulfate solution (100 mM) and analyzed for residual pesticide concentrations using HPLC.

Statistical Analysis
The statistical analysis was carried out using a SigmaPlot (Version 12.5, Systat Software, Inc, CA, USA). All experiments were performed in triplicate. Furthermore, the results (i.e., mean ± standard deviation) are presented in the manuscript.

Analytical Methods
The concentrations of the selected pesticides (i.e., ALA, CAR, DIU, PYR, and TEB) in the rainwater samples were quantified using high-performance liquid chromatography

Statistical Analysis
The statistical analysis was carried out using a SigmaPlot (Version 12.5, Systat Software, Inc, CA, USA). All experiments were performed in triplicate. Furthermore, the results (i.e., mean ± standard deviation) are presented in the manuscript. Table 2 shows the physicochemical characteristics of the rainwater samples at the study site. The conductivity and pH values were not significantly different during collecting periods (conductivity: 1.9-2.0 µS/cm; pH: 6.5-6.6). The average conductivity value (2.0 µS/cm) of the rainwater was considerably lower than that of the Korean urban precipitation (Goyang = 13.9 mS/cm; Gangneung = 30 µS/cm). In contrast, the mean pH value of 6.6 was slightly higher than that of other urban areas (Goyang = 5.6; Gangneung = 5.3) [21,22]. Moreover, the rainwater has fewer inhibition factors for oxidation treatments due to low DOC (0.4 mgC/L) and inorganic nitrogen species (TN = 70.2 µgN/L, NO 3 − = 51.3 µgN/L). Na + and Ca 2+ were the most abundant cations, and Cl − and SO 4 2− were the most abundant anions. Other rainwater contents, such as NO 3 − , NH 4 + , and Mg 2+ , were found in µg/L levels.   32 µM)). HOCl showed rapid consumption (29%) within 2 min from the initial oxidant dose (in terms of the initial phase), followed by a slow decrease over the next 60 min of reaction time (in terms of the second phase, the consumption rate = 70%). In contrast, O 3 depleted in less than 30 min (the consumption rate during the initial phases = 49%). Similar decay behaviors have been observed in the removal of effluent organic matter in secondary wastewater effluent by HOCl and O 3 [23].

Effects of Oxidant Dosages
The removal efficiencies of the selected pesticides were compared in the rainwater samples using selective (HOCl and O3) and non-selective (hydroxyl radicals: O3/H2O2) oxidants. With increasing oxidant dosage, the degradation of the selected pesticide's removal efficiency increased. Figure 3 shows that the removal efficiencies of CAR, PYR, and DIU by HOCl and O3 were between 15% and 50% for 10 μM oxidant concentration after 30 min, while the removal efficiency of ALA and TEB was only 9.5% under the same experimental conditions. After increasing the oxidant concentration to 40 μM, the removal

Effects of Oxidant Dosages
The removal efficiencies of the selected pesticides were compared in the rainwater samples using selective (HOCl and O 3  oxidants. With increasing oxidant dosage, the degradation of the selected pesticide's removal efficiency increased. Figure 3 shows that the removal efficiencies of CAR, PYR, and DIU by HOCl and O 3 were between 15% and 50% for 10 µM oxidant concentration after 30 min, while the removal efficiency of ALA and TEB was only 9.5% under the same experimental conditions. After increasing the oxidant concentration to 40 µM, the removal efficiency of CAR and DIU reached between 40% and 90% after 30 min, while the efficiency of ALA and TEB was enhanced to 21%. PYR was fully degraded with all oxidants. For HOCl and O 3 (the selective oxidants), there was a lag phase at lower oxidant dosages, where the pesticide concentration decreased slightly with increasing oxidant dosage (Figure 2a,b). The low removal efficiency in the lag phase was attributed to the high competition for the selective oxidants between the CAR and PYR and the DOC [23]. Additionally, with an increase in the selective oxidant dosages higher than the initial consumption by DOC, the residual concentrations of CAR and PYR started to decrease significantly. However, the removal rates of ALA and TEB were less than 21% for HOCl and O 3 processes. In contrast, in the case of O 3 /H 2 O 2 , the removal efficiencies of the selected pesticides were linearly proportional to the applied hydroxyl radical dosage (Figure 3c). These observations indicate that the magnitude of the competition for hydroxyl radicals between pesticides and rainwater matrix remained constant during the entire oxidation process. Based on these experiments on the removal efficiency of the selected pesticides according to the HOCl, O 3 , and O 3 /H 2 O 2 dosages, 40 µM was selected as the optimal dosage and used in subsequent experiments.  Figure 4 presents the removal efficiencies of the selected pesticides in the rainwater sample during the HOCl, O3, and O3/H2O2 processes at pH 6.6 for 0.5 h (each oxidant dose of 40 μM). For ALA and TEB, the removal efficiencies of HOCl, O3, and O3/H2O2 were in the order of O3/H2O2 (ALA = 49% and TEB = 23%) > O3 (ALA = 20% and TEB = 17%) > HOCl (ALA = 9% and TEB = 13%) (Figure 4a,e). ALA was reported to be lowly reactive with HOCl due to the inhibitive effect of the acetanilide functional group in the ALA. In addition, the ethyl groups of the ortho positions in the aromatic ring on the ALA reduced the reactivity toward the electrophilic attack of O3 to the aromatic moieties [24]. The relatively higher reactivity of TEB with O3/H2O2 compared with HOCl and O3 might be caused by the reaction between the aromatic C6-cycle or the C5 aromatic N-heterocycle and hydroxyl radicals [25]. The removal efficiencies of CAR, DIU, and PYR by HOCl (CAR = 98%, DIU = 92% and PYR = 100%) were more efficient than those of O3 (CAR = 81%, DIU = 49%, and PYR = 99%) and O3/H2O2 (CAR = 92%, DIU = 24%, and PYR = 98%) (Figure 4bd). CAR and PYR were highly reactive to HOCl and O3, and O3/H2O2 due to the reaction sites of the CAR (C=O double bond) and PYR (the nitrogen bridge between the two rings, the phenyl ring, and the pyrimidyl ring) [26,27]. DIU was lowly reactive with O3 due to the nitrogens at the urea function groups and the aromatic ring with the two chlorine atoms [28]. These findings suggest that choosing an appropriate oxidation process is important as the reactivity of the oxidant varies with the type of pesticide.  (Figure 4a,e). ALA was reported to be lowly reactive with HOCl due to the inhibitive effect of the acetanilide functional group in the ALA. In addition, the ethyl groups of the ortho positions in the aromatic ring on the ALA reduced the reactivity toward the electrophilic attack of O 3 to the aromatic moieties [24]. The relatively higher reactivity of TEB with O 3 /H 2 O 2 compared with HOCl and O 3 might be caused by the reaction between the aromatic C6-cycle or the C5 aromatic N-heterocycle and hydroxyl radicals [25]. The removal efficiencies of CAR, DIU, and PYR by HOCl (CAR = 98%, DIU = 92% and PYR = 100%) were more efficient than those of O 3 (CAR = 81%, DIU = 49%, and PYR = 99%) and O 3 /H 2 O 2 (CAR = 92%, DIU = 24%, and PYR = 98%) (Figure 4b-d). CAR and PYR were highly reactive to HOCl and O 3 , and O 3 /H 2 O 2 due to the reaction sites of the CAR (C=O double bond) and PYR (the nitrogen bridge between the two rings, the phenyl ring, and the pyrimidyl ring) [26,27]. DIU was lowly reactive with O3 due to the nitrogens at the urea function groups and the aromatic ring with the two chlorine atoms [28]. These findings suggest that choosing an appropriate oxidation process is important as the reactivity of the oxidant varies with the type of pesticide.

Effects of pH on the Removal of the Selected Pesticides
The changes in removal efficiencies of the selected pesticides during the HOCl, O and O3/H2O2 processes as a function of the pH of the rainwater samples are presented Table 3. In general, the reactivities of the selected pesticides with HOCl, O3, and O3/H2 increased as the pH increased. However, in CAR, DIU, and PYR, their removal efficienc in the chlorinated rainwater samples were strongly dependent on the pH condition. The results could be attributed to the accelerated reaction of HOCl in the chlorinated rainwa at pH 7, which was closely related to the removal of CAR, DIU and PYR [29].

Effects of pH on the Removal of the Selected Pesticides
The changes in removal efficiencies of the selected pesticides during the HOCl, O 3 , and O 3 /H 2 O 2 processes as a function of the pH of the rainwater samples are presented in Table 3. In general, the reactivities of the selected pesticides with HOCl, O 3 , and O 3 /H 2 O 2 increased as the pH increased. However, in CAR, DIU, and PYR, their removal efficiencies in the chlorinated rainwater samples were strongly dependent on the pH condition. These results could be attributed to the accelerated reaction of HOCl in the chlorinated rainwater at pH 7, which was closely related to the removal of CAR, DIU and PYR [29].

Effects of Humic Acids on the Removal of the Selected Pesticides
The consumption of oxidants by HA is a significant factor in determining the removal efficiency of the selected pesticides. In the presence of HA (Table 4), a significant abatement in the removal efficiency (reaction time = 0.5 h) of the selected pesticides was found for the HOCl processes. Despite the higher removal efficiency of CAR, DIU, and PYR during the HOCl, O 3 , and O 3 /H 2 O 2 processes compared to other pesticides, the influence of HA on the decrease in the selected pesticides was more pronounced in the O 3 /H 2 O 2 treated rainwater samples (the difference between the relative residual concentrations (∆C/C 0 ) of CAR, DIU, and PYR without and with 4 mg/L HA = 0.96, 0.75 and 0.41, respectively) than that for the chlorinated rainwater samples (∆C/C 0 of CAR, DIU, and PYR without and with 4 mg/L HA = 0.31, 0.42 and 0.018, respectively) due to the phenolic group in HA, which could easily react with dissolved O 3 and hydroxyl radicals during the O 3 /H 2 O 2 processes [23]. Similar behaviors were observed for removal efficiencies of micropollutants by microbubble ozone in the HA presence condition [19].

Effects of Inorganic Matters on the Removal of the Selected Pesticides
Inorganic matters, including NO 2 − and NH 4 + , generally exist in surface water and can rapidly consume the selective and non-selective oxidants and affect the removal efficiency of the selected pesticides. The rainwater samples spiked (4 mg/L) with additional NO 2 − and NH 4 + was treated with the predetermined oxidant dosage. Figure 5 shows that, in the case of HOCl and O 3 /H 2 O 2 , NH 4 + considerably decreased the removal efficiencies of ALA, CAR, DIU, PYR, and TEB. Remarkably, the DIU removal efficiency decreased by more than 55% for the HOCl-treated rainwater samples in the presence of NH 4 + as HOCl rapidly reacted with NH 4 + [30]. NH 4 + significantly reduced the removal efficiencies of CAR and PYR during the O 3 /H 2 O 2 treatment process (C/C 0 of CAR without NH 4 + : 0.19; C/C 0 of CAR with NH 4 + : 0.81; C/C 0 of PYR without NH 4 + : 0.004; C/C 0 of PYR with NH 4 + : 0.46). In the case of O 3 , the interference effects on selected pesticides were affected by NO 2 − , which was consistent with the relatively high k-value (3.7 × these observations suggest that suitable pretreatment processes are needed to enhance the removal efficiency of the selected pesticides in rainwater during the oxidation process. suggest that suitable pretreatment processes are needed to enhance the removal efficien of the selected pesticides in rainwater during the oxidation process.

Conclusions
In this study, the potential of the various oxidant processes for the abatement selected pesticides was evaluated and compared to the HOCl, O3 (i.e., the select oxidants), and O3/H2O2 (i.e., hydroxyl radical as the non-selective oxidant) processes provide deeper insights into the removal behavior of the selected pesticides. The prima outcomes were as follows: • DOC is a major rainwater component and has a more significant influence on t consumption kinetics of the oxidants in rainwater than inorganic nitrogen species

Conclusions
In this study, the potential of the various oxidant processes for the abatement of selected pesticides was evaluated and compared to the HOCl, O 3 (i.e., the selective oxidants), and O 3 /H 2 O 2 (i.e., hydroxyl radical as the non-selective oxidant) processes to provide deeper insights into the removal behavior of the selected pesticides. The primary outcomes were as follows: • DOC is a major rainwater component and has a more significant influence on the consumption kinetics of the oxidants in rainwater than inorganic nitrogen species; In general, the reactivities of the selected pesticides toward the HOCl, O 3 , and O 3 /H 2 O 2 increased due to deprotonation when the pH of the rainwater sample was higher than the pKa values of the selected pesticides; • The interference effects of HA and inorganic matter in the rainwater on removing the selected pesticides were more significant during the O 3 /H 2 O 2 process than those of the other oxidation processes; • These findings suggest that the oxidation processes (i.e., HOCl, O 3 , and O 3 /H 2 O 2 ) might be a promising method to enhance the removal efficiencies of organic pollutants, including pesticides, practically applicable for the wastewater treatment process.

Conflicts of Interest:
The authors declare no conflict of interest.