Due to increased population, rapid urbanization, economic growth and intensive agriculture, fresh water resources are reaching their limits . As a consequence of a human progresses, vast array of pollutants including industrial chemicals, pharmaceuticals, personal care products and pesticides has occurred in water in the level of concentration from ng L–1 to µg L–1 [2, 3]. Sources of organic pollutants are various; from industrial and domestic wastewaters, hospital effluents livestock and agriculture , and the concern about their occurrence in different environmental compartment has grown in last decades [5, 6]. Directive 2013/39/EU is promoting preventive actions, identifying pollution causes, dealing with emissions of pollutants at the source, and development of innovative (waste)water treatment technologies. This document includes a list of 45 substances or groups of substances and other pollutants with defined environmental quality standards , requiring attention due to their high frequency of occurrence, the expected risk for human health and aquatic life . Seven pesticides included in the list of priority substances have been chosen as targeted water pollutants in this study: alachlor, atrazine, chlorfenvinphos, cybutryne, diuron, isoproturon and simazine.
Water contaminated by pesticides presents a severe problem due to their generally high solubility in water, low-sorption affinity to soils, toxicity, chemical stability, bioaccumulation and low biodegradability. The major sources of pesticides pollution are related to drainage waters from intensive agriculture, including water from washing pesticides containers and application equipment, and effluents from agricultural industries and pesticide manufacturing plants [5, 7, 8].
Conventional water and wastewater treatments, consisting of aeration followed by filtration in primary and secondary rapid sand filters, are not designed to remove pesticides, thus effectiveness is rather limited . Several studies reported findings on pesticides removal from drinking water using technologies such as adsorption and oxidation processes -. Biological treatment has also been considered as a treatment option. However, biological treatment is limited to low pollutant concentrations due to the fact that higher loads exert inhibitory effects on microorganism activity, while lower temperatures level-off the treatment kinetics [11–13]. As alternative, advanced treatment technologies are considered. One of them that have been recognized as especially efficient compared to conventional technologies, are advanced oxidation processes (AOPs) . AOPs offer an effective barrier for organic pollutants, ensuring the removal of wide range of compounds, and are considered as a clean technology for water treatment . It applies the concept of producing primarily hydroxyl radicals (HO•), which attack organic pollutants. Hydroxyl radicals are highly reactive, unselective and powerful oxidant (E° = 2.80 V), capable to degrade organic pollutants with reaction rate of approximately 109 M–1 s–1, providing complete degradation/mineralization to CO2, H2O and inorganic ions as final products of the majority of organics in water [14–16]. Typical oxidant used for their generation is hydrogen peroxide (H2O2) [15–17]. In recent years, persulfate (S2O82–) is also used as an alternative oxidant in AOPs, resulting in generation of sulfate radical (SO4•–, E° = 2.60 V). Persulfate, commonly available in a salt form (with K+, Na+ and NH4+ cations) is characterized as a stable, high soluble at the ambient temperature and relatively harmless and environmentally friendly. Accordingly, SO4•– is increasingly discussed as an oxidizing agent for degradation of pollutants in water [17–25]. Due to its strong electron acceptor ability, SO4•– enables the degradation of recalcitrant compounds which are refractory towards HO•. The generation of SO4•– resembles an AOP based on H2O2. The peroxide bond of S2O82– and H2O2 can be broken by UV irradiation (eqs (1) and (2)) [19, 23–25]:
Photooxidative processes were applied in this study to degrade 7 selected pesticides included in the list of priority substances in water . The degradation of atrazine was in focus of numerous studies available in literature [18–20], while information on other studied pesticides by HO• and particularly SO4•– driven processes, is still scarce. Accordingly, the goal of the study was to determine degradation kinetics of targeted pesticides in water, evaluate and compare the effectiveness of photooxidative AOPs, UV-C/H2O2 and UV-C/S2O82–. Response surface modeling was used as a tool for screening the influence of key process parameters: pH and pollutant/oxidant ratio, on the process effectiveness, and for determining the optimal conditions within the investigated range.
Pesticides used in the study: Alachlor (99.8 %), Atrazine (99.1 %), Chlorfenvinphos (97.1 %), Cybutrine (98.4 %), Diuron (99.6 %), Isoproturon (99.8 %) and Simazine (99.9 %), were all purchased from Sigma Aldrich, USA. Their CAS numbers, formulas, chemical structures, used abbreviation in the study, as well as reaction rate constants with hydroxyl (HO•) and sulfate (SO4•–) radicals are summarized in the Table 1 [15, 18, 26]. Chemicals used as constituents of mobile phases for high performance liquid chromatography (HPLC) were methanol (CH3OH, HPLC grade, JT Baker, USA), acetonitrile (CH3CN, HPLC grade, JT Baker, USA), ammonium acetate (CH3COONH4, 98 %, Sigma Aldrich, USA) and ultra-pure water obtained from Millipore Direct-Q UV 3 system, Merck, USA. Oxidant agents used in photooxidative processes for degradation of pesticides were: hydrogen peroxide (H2O2, 30 % v/v, Gram-mol, Croatia) and sodium persulfate (Na2S2O8, p.a., Sigma Aldrich, USA). Auxiliary chemicals used for pH adjustment, sulfuric acid (H2SO4, >96 %) and sodium hydroxide (NaOH, p.a.), were both purchased from Kemika, Croatia. Humic acid (HA) in a form of sodium salt (technical grade), Sigma Aldrich, was used to simulate natural organic matter (NOM) present in water matrix.
The model solutions of the studied pesticides (PS-P) were prepared by dissolving respective quantities in deionized water (conductivity <1 μScm–1). The initial concentrations of studied PS-P were 0.1 mM or lower, regarding their solubility in water: [ALC] = 0.1 mM, [ATZ] = 0.03 mM, [CFP] = 0.1 mM, [CYB] = 0.02 mM, [DIU] = 0.05 mM, [IPT] = 0.1 mM, and [SZM] = 0.02 mM. All experiments were performed in a glass water-jacketed batch photoreactor (total volume, VT = 0.1 L, solution volume, VS = 0.08 L, and T = 25.0 +/- 0.2 °C). The photoreactor was equipped with UV-C lamp (P0 = 1.04×10–6 Einstein s–1), which was placed in the middle (irradiation path L = 1 cm), and magnetic stirrer in order to provide effective mixing of reaction solution (mixing speed was 550 rpms). Initial pH values ranged from 4 to 10. Oxidant (OX) concentration depended on the initial concentration of targeted PS-P, varying [PS-P]: [OX] ratio from 1: 10 to 1: 200 (Table 2). At the beginning of the process pH was adjusted to desired value using 0.1 M NaOH and H2SO4, then required oxidant aliquot was added, which was followed by the insertion of warmed-up UV-C lamp in the quartz tube to initiate the photooxidative or photolytic process. Photooxidative processes (UV-C/H2O2 and UV-C/S2O82–) and direct photolysis experiments (UV-C alone) were conducted for 90 and 480 s, respectively. At least 6 aliquots (500 μL) were taken in each experiment to establish the kinetic regime and determine apparent rate constants. Samples were quenched with CH3OH and submitted to HPLC analysis to monitor PS-P conversion. All experiments were performed in quintuplicates and averages were reported. The reproducibility of the experiments was >97.6 %.
Additional experiments were performed using the mixture of studied pesticides. The sum of their concentrations was 0.1 mM, while the composition of the mixture was defined by the ratio of the concentrations of individual components in previous experiments. Degradation kinetics of individual mixture components in the presence and absence of NOM was observed.
The conversion of PS-Ps was monitored by HPLC, Series 10, Shimadzu, Japan, equipped with UV-DAD, SPD-M10AVP, Shimadzu, Japan, using XBridge BEH C18 column (13 nm, 3.5 µm, 2.1 mm ˟ 150 mm), Waters, USA. The mobile phases, consistent of organic phase (CH3OH or CH3CN; type depended on targeted PS-P) and polar phase (20 mM of CH3COONH4), which ratio depended on the type of targeted PS-P, were operated by isocratic flow at 0.5 mL min–1. Handylab pH/LF portable pH-meter, Schott Instruments GmbH, Mainz, Germany, was used for pH measurements.
The degradation kinetics of studied PS-Ps was tested for the reaction order using integrated equations displayed by different functional dependences of concentration on time (eqs (3)–(5)) , using the conversion data acquired during performed photooxidative and photolytic treatments.
The full factorial design (FFD) and response surface modeling (RSM) were employed in order to estimate the influence of pH and oxidant concentration on the photooxidative treatments (UV-C/H2O2 and UV-C/S2O82–) efficiency based on the apparent degradation rates. In such manner, the treatment time was indirectly introduced in the used design of experiment as a third variable. Process parameters are represented by independent variables: X1 (pH) and X2 (oxidant concentration), and their combined influence on degradation kinetics (i. e. dependent variable; Y) was described by polynomial equation:
where β0 is intercept; β1, β2, are linear; β11, β22 quadratic; and β12, interaction coefficients . The values of independent variables were transformed in dimensionless coded values; three levels for X1 and X2. The Table 2 summarizes the full experimental matrix and the apparent rate constants for each studied PS-P. The fitting of models was evaluated using the coefficient of determination (R2) and the analysis of variance (ANOVA). Calculations and analyses were performed using STATISTICA 12.7, Dell-Inc., USA, Design-Expert 10.0, StatEase, USA, and Mathematica 9.0, Wolfram Research, USA.
3 Results and discussion
Collected data on PS-P conversion during applied UV-C/H2O2 and UV-C/S2O82– treatments were tested for reaction order using above given eqs (3)–(5). It was establish that degradation kinetics obey first-order in all cases, thus such calculated apparent rate constants can be sued as appropriate response in RSM modeling. The RSM approach was applied to study the influence of pH and [PS-P]:[OX] on degradation kinetics of PS-Ps within the studied range of process parameters. Hence, taking into account the DoE matrix given in Table 2 and obtained apparent kinetic rates, RSM models in a form of quadratic polynomials were derived for each studied PS-P and used oxidant type (Table 3). In the spirit of good practice in such statistical/empirical modeling [22, 25, 28], we evaluated the derived RSM models for their significance and accuracy using standard statistical tools such as analysis of variance (ANOVA; including F, t, p, R2, Radj2, Rpre2 values) and residual diagnostic (RD; including normal probability test, Levene’s test, and constant variance test). According to the results (Table 3), 0.942 < R2 < 0.999 and 0.0002 < p < 0.0456, it can be concluded that all derived RSM models are accurate and significant, thus can be used hereinafter as a tool to enlighten the influence of key process parameters of applied UV-C/H2O2 and UV-C/S2O82– treatments on degradation kinetics of on studied PS-Ps. ANOVA results (Table 3) revealed statistical significance of each model term within RSM models derived. Based on that, it was found that in most cases both pH and [PS-P]:[OX] contribute to degradation kinetics of studied PS-Ps. The exceptions can be found in the cases of ALC and CFP treated by UV-C/S2O82– and UV-C/H2O2, respectively, whereas model terms corresponding to pH were found to be statistically insignificant within the studied range (Table 3).
The mutual interactions of key process parameters of photooxidative treatments on degradation kinetics of PS-Ps were presented through 3D surface and contour diagrams. Corresponding 3D plots for ALC degradation kinetics by UV-C/H2O2 and UV-C/S2O82– processes in dependence on pH and [ALC]:[OX] are given in Figure 1 and 1(B), respectively. As can be seen, in the case of UV-C/H2O2 process ALC degradation kinetics strongly depends on both studied parameters, which is not the case when UV-C/S2O82– process was applied, whereas pH was found as statistically insignificant (Table 3). Such results indicate that ALC can be effectively treated by sulfate radical driven process through wide pH range. As can be clearly seen (Table 2, Figure 1(A)), kapp(ALC) in the case of UV-C/H2O2 process varies a lot through studied range of process parameters. For instance, at the lowest [ALC]:[H2O2] ratio of kapp(ALC) ranges from 8.3 to 10.9˟10–3 s–1 (Figure 1(A), Table 2), exhibiting the highest rate at neutral pH. By an increase in oxidant concentration, pH influence became more evident. Hence, at the highest [ALC]:[H2O2] ratio kapp(ALC) in acidic medium was found to be 3 times higher than that at basic conditions (Figure 1(A)., Table 2). The possible explanation for such effect can be found in the fact that at basic conditions H2O2 dissociate to less photo active HO2– , resulting in lower generation of HO• and eventually lower kapp. This negative effect of pH was not so evident at lower [ALC]:[H2O2] ratio due to the fact that degradation occurred simultaneously by HO• driven and direct photolysis mechanisms . Accordingly, at lower oxidant concentration, direct photolysis mechanism came to forth, presumably almost equally contributing to the overall degradation rate along with HO• driven mechanism. As reported above, pH was found as insignificant in the case of UV-C/S2O82– process (Figure 1(B)). It is known from the literature that decomposition of persulfate either by UV-C, heat or transition metals results in the generation of highly active SO4•–, while in side reactions HSO4– is formed, strongly contributing to the lowering of pH [23, 25]. Since HSO4– would be formed, consequently lowering pH, it can be considered that degradation occurred mainly at acidic conditions regardless initial pH. This is supported by ANOVA results (Table 3) and Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7(B) revealing that the contribution of pH is less pronounced in comparison to [PS-P]:[OX] ratio for all PS-Ps treated by UV-C/S2O82– process. Furthermore, kapp(ALC), obtained at the same process conditions (Table 2), by UV-C/S2O82– are 2–4 times higher than in the case of UV-C/H2O2 process. The observed differences can be attributed to the different optical properties of applied oxidants (i. e. quantum yields and molar absorption coefficients) and different reaction rate constants of ALC and HO• and SO4•–. The quantum yields of studied oxidants under UV-C irradiation applied (λ = 254 nm) differ significantly: ΦH2O2 = 0.5 < ΦNa2S2O8 = 0.9 with the regard to consumed oxidants. It should be noted that some authors use values of quantum yields in respect toward formed radicals: 1.0 and 1.8 . The molar absorption coefficients of used oxidants at applied wavelength are also different: ε254(H2O2) = 18.6 M–1 cm–1 < ε254(Na2S2O8) = 47.45 M–1 cm–1. Accordingly, at the same concentration of oxidants, incident photon flux and radiation path, the concentration of HO• would be lower for almost 50 % in comparison to SO4•–. Although literature does offer reaction rate constants between ACL (and other studied PS-Ps) and HO• [15, 18, 26], the corresponding data for SO4•– are still scarce to the best of our knowledge. Accordingly, the comparison regarding reaction rate constants can not be made. However, the differences in mechanisms of reactions between HO• and SO4•– and organic pollutants (taking into account their structural features) can reveal potential differences observed through kapp in our case. The main reaction pathways of HO• in the reaction with organic compounds possessing C-C, C-N and C-S double bonds is addition and H-abstraction. Since the most organic compounds (such as studied PS-Ps) reveal structural moieties which allow addition and H-abstraction, HO• react fast with a large variety of compounds (i. e., > 109 M–1 s–1, as shown for studied PS-Ps in Table 1). On the other hand, SO4•– reacts more selectively by electron transfer enabling reaction pathways which are not possible in case of HO•, while the addition and H abstraction is mostly slower compared to corresponding reactions with HO•. Generally, the reaction rate constants with SO4•– are around one order of magnitude below corresponding rate constants with HO• [31, 32].
In Figure 2 degradation kinetics of ATZ by UV-C/H2O2 and UV-C/S2O82– processes in dependence on pH and [ATZ]:[OX] are compared. The layout of 3D surface presenting the influence of key parameters of UV-C/H2O2 process to kapp(ATZ) is very similar to that obtained in the case of ALC (Figure 1 and Figure 2(A)). The main differences can be found at lower [PS-P]:[OX] ratio, where kapp(ATZ) are approximately two times higher than corresponding kapp(ALC). It should be noted that initial [ATZ] was more than 3 times lower than initial [ALC] and that ALC has twice as higher reaction rate with HO• than ATZ (Table 1). By increasing [PS-P]:[OX] ratio through investigated range, kapp(ATZ) and kapp(ALC) almost equalized. Such behavior can be attributed to structural features of ATZ, i. e. triazine ring, which is highly susceptible to direct photolysis , and in such case direct photolysis mechanism is promoted over radical driven mechanism in overall degradation rate in the case of ATZ. On the other hand, the layout of 3D surface presenting the influence of key parameters of UV-C/S2O82– process to kapp(ATZ) is completely different than that obtained in the case of ALC (Figure 1 and Figure 2(B)). As discussed above, pH value in UV-C/S2O82– process lowered rapidly due to the formation of HSO4– in side reactions, resulting in lowering the pH to acidic conditions through the treatment time course. Accordingly, initial pH has been found as insignificant parameter in the case of ALC (Figure 1(B)). The same effect did not come to forth in the case of ATZ. As can be seen from 3D surface (Figure 2(B)), initial pH has almost the same influence as [ATZ]:[OX] ratio. This is in accordance with the findings of Khan et al. , who determined the dependence of ATZ degradation by UV-C/S2O82– process on the initial pH. However, in their case, neutral conditions were found to be most suitable for ATZ degradation, while our experiments revealed that ATZ degradation is promoted at acidic conditions (Figure 2(B)). It should be noted that at acidic conditions and highest [ATZ]:[OX] ratio the highest kapp(ATZ) where obtained by both processes; 50.5 and 67.1 s–1 by UV-C/H2O2 and UV-C/S2O82– processes, respectively (Figure 2, Table 2). Taking into account above discussed higher generation of SO4•– than HO• at same conditions, as well as the fact that reaction rate constants are rather similar (Table 1), one may expect faster degradation of ATZ by UV-C/S2O82– process, which was not the case. The plausible explanation can be found in the fact that kapp summarizes both direct photolysis and radical driven mechanisms, but also the consumption of generated radicals in side reactions with formed by-products. Hence, it can be expected that ATZ by-products could react with SO4•– at higher rates than with parent pollutant, and consequently kapp is lower than expected.
In Figure 3 degradation kinetics of CFP by UV-C/H2O2 and UV-C/S2O82– processes in dependence on pH and [CFP]:[OX] are compared. The influence of [PS-P]:[OX] ratio is similar to that observed in the cases of ALC and ATZ. As can be seen from Table 2 and Figure 3(A), kapp(CFP) at basic conditions are slightly lower than obtained at neutral and acidic, indicating negative influence of H2O2 dissociation. ANOVA analysis determined the influence of pH as statistically insignificant in the case of CFP degradation by UV-C/H2O2 (Table 3). Similarly to UV-C/H2O2, the influence of [CFP]:[OX] ratio is dominant in UV-C/S2O82– process (Figure 3(B)) As can be seen, kapp(CFP) by UV-C/S2O82– process are much higher than that obtained by that UV-C/H2O2, particularly at higher [CFP]:[OX] ratios (Table 2, Figure 3.). The plausible explanation can be found in the aforementioned higher generation of SO4•– than HO• in corresponding processes at the same conditions, but also in the fact that electron transfer, which can be effectively accomplished by SO4•–, is more favorable pathway than the addition and H abstraction in the case of CFP. At lower [CFP]:[OX] ratios these differences are not so pronounced due to the contribution of direct photolysis in overall degradation mechanism in both processes.
In Figure 4 degradation kinetics of CYB by UV-C/H2O2 and UV-C/S2O82– processes in dependence on pH and [CYB]:[OX] are presented. Again, the layout of 3D surface plot for UV-C/H2O2 (Figure 4(A)) is rather similar as obtained in the cases of ALC, ATZ, and CFP reflecting the general chemistry of UV-C/H2O2 process. An increase in [PS-P]:[OX] ratio led to the higher generation of HO•, and consequently higher degradation rate, while an increase in pH toward basic conditions increases dissociation of H2O2, resulting in lower generation of HO•, and consequently degradation rate was lower as well. As can be seen from Figure 4(B), similar 3D surface layout can be observed in the case of CYB degradation by UV-C/S2O82– process, but negative influence of basic pH is less pronounced, presumably due to aforementioned fact that initial pH has been lowered as a consequence of HSO4– formation. It should be noted that kapp(CYB) obtained by UV-C/S2O82– process are significantly lower than those obtained by UV-C/H2O2 (Table 2, Figure 4.). Since higher generation of SO4•– can be expected in comparison to HO• at analogues process conditions, it is more likely that reaction rate constants of these radical species with CYB differ significantly. In Table 1 kHO• for CYB is given , while the corresponding data SO4•– were not found. However, the obtained results indicate that reaction rate constant might be lower for one order of magnitude in comparison to kHO•.
Figure 5 presents the influence of pH and [DIU]:[OX] on degradation kinetics of DIU by UV-C/H2O2 and UV-C/S2O82– processes. In the case of UV-C/H2O2 process similar layout of 3D surface plot is again obtained as in the cases of ALC, ATZ, CFB and CYB, indicating that structural features are not depended on changes of process parameters in HO• driven process. In the case of UV-C/S2O82– process (Figure 5(B)), initial pH was found to be significantly less influential process parameter than [DIU]:[S2O82–] ratio. As in cases of ALC and CFP discussed above, kapp(DIU) obtained by UV-C/S2O82– process are significantly higher than those obtained by UV-C/H2O2 process. Similar observation can be made on the basis of the obtained 3D surface plots for IPT, presented in Figure 6. Such results might be a consequence of similar structures of DIU and IPT (Table 1); both are made of dimethylurea bonded to benzene ring, which is substituted by isopropyl group in the case of IPT, while in the case of DIU by two chlorides. Accordingly, the similar degradation mechanisms can be considered in both cases.
3D surface plot in the case of SMZ degradation by UV-C /H2O2 process exhibited similar layout as observed in the previous cases, regardless the structural features of studied PS-Ps (Figure 7(A)). The layout of 3D surface obtained for UV-C /S2O82– process was also similar to those obtained for the most of studied PS-Ps regarding the dominant influence of [PS-P]:[S2O82–] ratio on degradation rate, but the difference can be seen in the influence of initial pH. As can be observed from Figure 7, the maximal degradation rate would occur at neutral pH throughout entire range of studied [PS-P]:[S2O82–], which was not the case with other studied PS-Ps, where acidic pH was found as optimal for maximal degradation rate. Similar findings, neutral pH favoring maximal degradation, were reported by Khan et al.  who studied degradation of ATZ, which has similar structure to SMZ (Table 1). They differ only in one moiety – propyl in the case of ATZ and ethyl in the case of SMZ, bonded to amino moiety at triazine ring. However, as aforementioned, we did not observe such effect in studying ATZ. The difference observed between those two similarly structured PS-Ps is related to the maximal kapp by UV-C/S2O82– process. Although initial [SMZ] is lower than that of [ATZ] for only one-third, kapp(SMZ) were two or three times higher (Figure 2 and Figure 7(B), Table 2), indicating on the higher reaction rate constant of SMZ and SO4•– (unknown data) than that reported for ATZ in Table 1.
In conclusion to above discussed results it should be pointed out that in both studied processes the contribution of [PS-P]:[OX] ratio on kapp is more pronounced than that of initial pH within the studied range. In spite of the fact that SO4•– tends to react at similar or slower rates with the most of the organic pollutants in comparison to HO• [31, 32], kapp for PS-Ps treated by UV-C/S2O82– process varied within the significantly wider range and were in some cases higher than corresponding kapp obtained in UV-C/H2O2 process, presumably due to the differences in quantum yields of used oxidants and in reaction mechanisms of HO• and SO4•– influenced by PS-P structure.
The developed RSM models were tested for maxima to determine values of process parameters yielding the highest PS-P degradation rate constant as chosen response. The results of experiments performed at such determined process conditions are presented in Table 4. The predicted and experimentally obtained values of apparent rate constant (kpre and kapp, respectively) are close to each other (Table 4), suggesting the accuracy of developed RSM models summarized in Table 3. In all cases the top boundary of [PS-P]:[OX] ratio was determined as optimal, indicating that scavenging effect of excess oxidants did not come to forth within the studied range (Table 4). Acidic conditions are found to be beneficial for both processes, i. e. bottom boundary was found to be optimal in most cases. The exception was SMZ, where optimal values were at initial weak acidic and neutral pH for UV-C/H2O2 and UV-C/S2O82– processes, respectively. In Figure 8 PS-Ps degradation kinetics treated by UV-C alone at natural pH and UV-C/H2O2 and UV-C/S2O82– processes at optimal conditions determined (Table 4) are presented. It can be seen that all studied PS-Ps are susceptible to degradation by direct photolysis. The conversion of PS-Ps within 90 s varied from 17 to 80 % (Figure 8), which can be attributed to differences in their molecular structure, determining their optical properties, as well as initial concentrations. In all cases addition of oxidant significantly influenced observed degradation kinetics, suggesting that the radical driven mechanism play important role in overall process effectiveness. The complete conversion is obtained within 30–90 s in most cases, depending on the type of PS-Ps and oxidant used. With the exception of CYB, S2O82– was found to be more suitable oxidant for the treatment of studied PS-Ps under UV-C irradiation. The observed differences in degradation kinetics can be contributed to the different reaction mechanisms of HO• and SO4•– above discussed (via addition and via electron transfer, respectively). Accordingly, it can be assumed that methylthio and cycloproylamino moieties in CYB structure disfavor electron transfer as a main mechanism of initial degradation step by SO4•–.
In order to investigate the influence of natural organic matter on degradation kinetics of studied PSs in the mixture, the additional set of experiments was performed. Figure 9 summarizes degradation kinetics of PSs treated by UV-C /H2O2 and UV-C/S2O82– processes at pH 4 and [PSs]:[OX] = 1:200, as shown to be optimal in most cases when individual mixture components were treated (Table 4). The values of apparent degradation rates are summarized in Table 5. According to the literature , NOM may provide both inhibitory and synergistic effects on the removal rate of targeted compounds by photooxidative processes. However, as it can be seen from Figure 9 and Table 5, inhibitory effect of NOM was more pronounced than synergistic effect. In most cases, degradation rates of studied PSs by UV-C/H2O2 were lower when NOM was present in mixture. The observed degradation rates of ALC and DIU were not altered (Figure 9 and 9(E), Table 5). Influence of NOM toward degradation rates of PSs is even less pronounced in the case of UV-C/S2O82– process. However, from the calculated degradation rates (Table 5), slight inhibitory effect can be observed in all cases. The decreased removal rates in the presence of NOM can be attributed to the competitive absorption of emitted irradiation influencing the direct photolysis of PSs and photo-induced generation of reactive species. NOM can act as a scavenger, competing with PSs for radical species in the system. However, above presented experimental results suggest that aforementioned inhibitory effect of NOM is quite limited in the case of studied PSs degradation by UV-C /H2O2 and UV-C/S2O82– processes.
Seven pesticides listed as priority substances (PS-Ps) within the EU Water Framework directive; alachlor, atrazine, chlorfenvinphos, cybutryne, diuron, isoproturon and simazine, were effectively degraded by UV-C/H2O2 and UV-C/S2O82– processes. Based on experimental results it was established that degradation kinetics obey first-order, and accordingly apparent rate constants (kapp) were calculated.
The influence of pH and [PS-P]:[OX] on the degradation kinetics, represented through calculated kapp, was successfully evaluated using full factorial plan and response surface modeling. ANOVA analysis revealed that all developed RSM models are significant and accurate, and were used to study process chemistry related to structural features of studied PS-Ps. It was established that degradation kinetics of PS-Ps is highly depended on [PS-P]:[OX], while pH has minor significance, particularly in UV-C/S2O82– process. With the exception of cybutrine, S2O82– was found to be more suitable oxidant for the treatment of studied PS-Ps under UV-C irradiation. The observed differences in degradation kinetics can be attributed to the different reaction mechanisms of HO• and SO4•–, via addition and via electron transfer, respectively. Addition of NOM showed an inhibitory effect toward degradation rates of some of the studied PSs in the mixture.
1. Holden J. Water resources: An integrated approach. New York, NY: Routledge, Taylor and Francis Group; 2014. Google Scholar
2. Jurado A, Vazquez-Sune E, Carrera J, Lopez De Alda M, Pujades E, Barcelo D. Sci Total Environ. 2012;440:82–94. Google Scholar
3. EC. Off J Eur Commun. 2013;226:1–17. Google Scholar
4. Luo Y, Guo W, Ngo HH, Nghiem LD, Hai FI, Zhang J,et al. Sci Total Environ. 2014;473–474:619–641. Google Scholar
5. Kuster M, Lopez De Alda MJ, Hernando MD, Petrovic M, Martin-Alonso J, Barcelo D. J Hydrol. 2008;358:112–123. Google Scholar
6. Ribeiro AR, Nunes OC, Pereira MF, Silva AM. Environ Int. 2015;75:33–51. Google Scholar
7. Geissen V, Mol H, Klumpp E, Umlauf G, Nadal M, Van Der Ploeg M,et al. Inter Soil Water Conserv Res. 2015;3:57–65. Google Scholar
8. Moreira FC, Vilar VJ, Ferreira AC, Dos Santos FR, Dezotti M, Sousa MA,et al. Chem Eng J. 2012;209:429–441. Google Scholar
9. Al Hattab MT, Ghaly AE. J Environ Prot. 2012;3:431–453. Google Scholar
10. Rodriguez A, Garcia J, Sotelo JL, Ovejero G, Mestanza M. Fresen Environ Bull. 2009;18:2093–2101.Google Scholar
11. Chelme-Ayala P, El-Din MG, Smith DW, Adams CD. Wat Res. 2011;45:2517–2526. Google Scholar
12. Gisi D, Stucki G, Hanselmann KW. Appl Microbiol Biotechnol. 1997;48:441–448.Google Scholar
13. Budaev SL, Batoeva AA, Tsybikova BA. Miner Eng. 2015;81:88–95. Google Scholar
14. Oller I, Malato S, Sanchez-Perez JA, Maldonado MI, Gasso R. Catal. Today. 2007;129:69–78. Google Scholar
15. Wols BA, Hofman-Caris CH. Wat Res. 2012;46:2815–2827. Google Scholar
16. Litter MI, Candal RJ, Meichtry JM. Advanced oxidation technologies: Sustainable solutions for environmental treatments. London, UK: CRC Press, Taylor and Francis; 2014. Google Scholar
17. Vilhunen S, Sillanpää M. Rev Environ Sci Biotechnol. 2010;9:323–330. Google Scholar
18. Luo C, Ma J, Jiang J, Liu Y, Song Y, Yang Y,et al. Wat Res. 2015;80:99–108. Google Scholar
19. Khan JA, He X, Shah NS, Khan HM, Hapeshi E, Fatta-Kassinos D,et al. Chem Eng J. 2014;252:393–403. Google Scholar
20. Chu W, Wang YR, Leung HF. Chem Eng J. 2011;178:154–160. Google Scholar
21. Tan C, Gao N, Deng Y, Ana N, Deng J. Chem Eng J. 2012;203:294–300.Google Scholar
22. Kusic H, Jovic M, Kos N, Koprivanac N, Marin V. J Hazard Mater. 2010;183:189–202. Google Scholar
23. Kusic H, Peternel I, Ukic S, Koprivanac N, Bolanca T, Papic S,et al. Chem Eng J. 2011;172:109–121. Google Scholar
24. He X, Mezyk SP, Michael I, Fatta-Kassinos D, Dionysiou DD. J Hazard Mater. 2014;279:375–383. Google Scholar
25. Brlenic V, Kusic H, Juretic D, Loncaric Bozic A. J Water Proc Eng. 2016;10:78–88. Google Scholar
26. Olasehinde EF, Ogunsuyi HO, Sakugawa H. IOSR J App Chem. 2012;1:7–14. Google Scholar
27. Connors KA. Chemical kinetics: The study of reaction rates in solution. New York, USA: Wiley-VCH; 1990. Google Scholar
28. Myers RH, Montgomery DC, Anderson-Cook CM. Response surface methodology: Process and product optimization using designed experiments;, 3rd ed.. Hoboken, NJ: John Wiley & Sons; 2009. Google Scholar
29. Beltran FJ, Tarr MA. Chemical degradation methods for wastes and pollutants, environmental and industrial applications. New York, NY: Marcel Dekker, Inc; 2003:1–77. Google Scholar
30. Juretic D, Kusic H, Dionysiou DD, Rasulev B, Loncaric Bozic A. Chem Eng J. 2014;257:229–241. Google Scholar
31. Buxton GV, Greenstock CL, Helman WP, Ross AB. Phys Chem Ref Data. 1988;17:513–586. Google Scholar
32. Neta P, Huie RE, Ross AB. J Phys Chem Ref Data. 1988;17:1027–1284. Google Scholar
33. Klementova S, Zlamalb M. Photochem Photobiol Sci. 2013;12(4):660–663. Google Scholar
34. Kwon M, Kim S, Yoon Y, Jung Y, Hwang T-M, Lee J,et al. Chem Eng J. 2015;269:379–390. Google Scholar
About the article
Published Online: 2017-02-10
Published in Print: 2017-01-01
We gratefully acknowledge on the financial support from Croatian Science Foundation through project entitled Modeling of Environmental Aspects of Advanced Water Treatment for Degradation of Priority Pollutants (MEAoWT) (IP-09-2014-7992).