Degradation of UV Filter Ethyl 4-Aminobenzoate (Et-PABA) Using a UV-Activated Persulfate Oxidation Process

In this paper, the ultraviolet/persulfate (UV/PDS) combined oxidation process was used to remove the ethyl 4-aminobenzoate (Et-PABA), one of the typical 4-aminobenzoic acid (PABA)-type UV filters. The effects of various factors on the removal of Et-PABA using the UV/PDS process were investigated, and the degradation mechanisms of Et-PABA were explored. The results showed that the UV/PDS process can effectively remove 98.7% of Et-PABA within 30 min under the conditions: UV intensity of 0.92 mW·cm−2, an initial concentration of Et-PABA of 0.05 mM, and a PDS concentration of 2 mM. The removal rate of Et-PABA increased with the increase in PDS dosage within the experimental range, whereas humic acid (HA) had an inhibitory effect on Et-PABA removal. Six intermediates were identified based on HPLC–MS and degradation pathways were then proposed. It can be foreseen that the UV/PDS oxidation process has broad application prospects in water treatment.


Introduction
In recent years, skin cancer has attracted widespread attention owing to a series of risks related to ultraviolet (UV) radiation, sunburn, and premature skin aging [1]. Thereby UV filters are widely used in personal care products (PCPs) as UV-absorbing (shielding) agents that protect against UVA and UVB radiation and protect human health [2]. Most PCPs such as cosmetics, sunscreens, hair care products, detergents, etc., contain one or more UV filters in everyday products that are closely related to our lives. These organic UV filters are prone to the persistent pollution of the environment due to their large dosage, stable physical and chemical properties, and resistance to degradation. Those UV filters have become a new class of contaminants as they continue to enter the environment daily [3]. Residues of organic UV filters have been detected in various media such as surface water, sediment and aquatic organisms [4][5][6][7]. Several studies have shown that UV filters have potential endocrine disrupting effects that adversely affect organisms such as human tissues, mammals, animals, and fish, and cause serious health threats [8][9][10][11]. UV filters are relatively stable in nature and are resistant to biodegradation, making it difficult for municipal wastewater treatment plants (MWTPs) to remove them thoroughly [12][13][14]. Among them, p-aminobenzoic acid (PABA)-type compounds are the earliest and most widely used sunscreen products, thus their exposure characteristics, ecological effects and value that can effectively remove organic contaminants in water via an oxidation reaction. UV-based AOPs have great potential due to their high treatment efficiency, non-toxicity and low cost. Many researchers [32,33] used the UV-activated persulfate process to study the removal of refractory organic pollutants. Since UV technology has been widely used in MWTPs, it is advantageous to use UV to activate persulfate. In this case, UV cannot only disinfect and sterilize the water [34], but also activate persulfate to produce strong oxidizing SO 4 − to remove organic pollutants, thus it is more practical to study UV-activated persulfate to remove PABA-type UV filters [35]. However, previous studies have been mainly focused on the photocatalytic degradation of PABA [36][37][38][39] or ethyl 4-aminobenzoate (Et-PABA) [40], one of the typical PABA-type UV filters. Information regarding the SO 4 − degradation process of Et-PABA is limited, which hinders the application of this technique in PABA-type UV filter treatments. In this paper, the UV-activated PDS process was used to remove Et-PABA. The effects of PDS dosage, pH value, humic acid (HA) and other factors on the removal of Et-PABA during the UV/PDS process were investigated. The degradation intermediates of Et-PABA were identified and the possible degradation pathways were proposed. This work provides basic data, application references and theoretical guidance of the UV/PDS process for the removal of PABA-type UV filters in water.

Experimental Methods
A schematic diagram of the laboratory-made ultraviolet light irradiation device is shown in Figure 1. The size of the reactor was 30 × 30 × 30 cm. The ultraviolet light irradiator was equipped with two low-pressure mercury lamps (254 nm, G4T5, Philips, Kossaka, Poland; 4W). The UV lamps were arranged in parallel above a crystallizing dish (90 cm in diameter) and were approximately 10 cm from the surface of the solution, where the intensity of UV irradiation was determined by a digital UV-C meter (0.92 mW·cm −2 ). In order to ensure the stability of the light source, the UV lamps were preheated for 30 min. Generally, UV filters have been detected in water with concentrations up to µg/L. In this experiment, 0.05 mM of Et-PABA was chosen for convenience detection by LC and LC-MS [29,30]. 100 mL solution of Et-PABA and 2 mM PDS were added into a crystallizing dish, then magnetically stirred at 800 rpm. A quantity of 800 µL of the solution was sampled at different reaction times, filtered through a 0.45 µm filter, and added into 1.5 mL chromatography vials. The reaction was stopped with 200 µL of MeOH quencher.
Appl. Sci. 2019, 9, x 3 of 11 magnetically stirred at 800 rpm. A quantity of 800 μL of the solution was sampled at different reaction times, filtered through a 0.45 μm filter, and added into 1.5 mL chromatography vials. The reaction was stopped with 200 μL of MeOH quencher.
The removal efficiencies can be obtained by Equation 1.
where, r is the removal efficiency; C0 is the concentration of the original solution; C is the concentration of the solution after treatment using the oxidation process; and t is the reaction time.
The samples were extracted and enriched using Waters Oasis HLB columns (Waters, USA), and then were analyzed by an Qtrap 5600+ (AB Sciex, Redwood, CA, USA) MS coupled with an ESI source and an ACQUITY UPLC (Waters, Milford, MA,USA) [31].

Degradation of the Et-PABA Using Different Oxidation Processes
The removal efficiencies of Et-PABA using UV alone, the PDS alone and the UV/PDS combined process, respectively, were evaluated. As shown in Figure 2A, the removal efficiency of Et-PABA using a single UV process was only 23.7%, because the main ultraviolet radiation wavelength of 254 nm lies just within the absorption band of Et-PABA [32]. The UV process may cause photolysis of a part of Et-PABA because a small amount of active radicals such as • OH may be generated in the water to indirectly oxidize the contaminants [41]; the degradation efficiency of Et-PABA using the PDS process alone was only 21.2%, because PDS can theoretically oxidize and degrade parts of the organic compounds. By contrast, under the same reaction conditions, the removal efficiency of Et-PABA using the UV/PDS combined process increased rapidly, and the degradation efficiency can reach 98.7% in 30 min. The synergy value was calculated according to Reference [42]. It was a positive value (0.918), which stands for a synergistic effect that may greatly improve the removal efficiency of Et-PABA. SO4 •− generated from the PDS activated by the ultraviolet light has a higher oxidationreduction potential of E0 = 2.60 V, and its strong oxidizing ability can transform organic matter into an excited state and thus, the organic contaminants can be removed from water efficiently via an oxidation reaction [43]. Moreover, Figure 2B shows that the UV/PDS process can also efficiently remove PABA, indicating the universality of a UV/PDS process for the removal of PABA-type UV
The removal efficiencies can be obtained by Equation (1).
where, r is the removal efficiency; C 0 is the concentration of the original solution; C is the concentration of the solution after treatment using the oxidation process; and t is the reaction time.
The samples were extracted and enriched using Waters Oasis HLB columns (Waters, USA), and then were analyzed by an Qtrap 5600+ (AB Sciex, Redwood, CA, USA) MS coupled with an ESI source and an ACQUITY UPLC (Waters, Milford, MA, USA) [31].

Degradation of the Et-PABA Using Different Oxidation Processes
The removal efficiencies of Et-PABA using UV alone, the PDS alone and the UV/PDS combined process, respectively, were evaluated. As shown in Figure 2A, the removal efficiency of Et-PABA using a single UV process was only 23.7%, because the main ultraviolet radiation wavelength of 254 nm lies just within the absorption band of Et-PABA [32]. The UV process may cause photolysis of a part of Et-PABA because a small amount of active radicals such as OH may be generated in the water to indirectly oxidize the contaminants [41]; the degradation efficiency of Et-PABA using the PDS process alone was only 21.2%, because PDS can theoretically oxidize and degrade parts of the organic compounds. By contrast, under the same reaction conditions, the removal efficiency of Et-PABA using the UV/PDS combined process increased rapidly, and the degradation efficiency can reach 98.7% in 30 min. The synergy value was calculated according to Reference [42]. It was a positive value (0.918), which stands for a synergistic effect that may greatly improve the removal efficiency of Et-PABA. SO 4 − generated from the PDS activated by the ultraviolet light has a higher oxidation-reduction potential of E 0 = 2.60 V, and its strong oxidizing ability can transform organic matter into an excited state and thus, the organic contaminants can be removed from water efficiently via an oxidation reaction [43]. Moreover, Figure 2B shows that the UV/PDS process can also efficiently remove PABA, indicating the universality of a UV/PDS process for the removal of PABA-type UV filters in water. Particularly, the degradation efficiencies of Et-PABA and PABA using the UV process were slightly higher than that using the PDS process after a 20 min reaction. This may be due to the fact that if UV radiation is increased or the reaction time is extended, UV alone can also remove contaminants from the water, Appl. Sci. 2019, 9, 2873 4 of 11 while PDS does not. Therefore, as time goes on, the individual UV process exhibits a better removal efficiency than the PDS alone. These results were consistent with the results of the previous study [32].
contaminants from the water, while PDS does not. Therefore, as time goes on, the individual UV process exhibits a better removal efficiency than the PDS alone. These results were consistent with the results of the previous study [32].
Subsequently, we also studied the removal efficiency of Et-PABA using PDS activated by heat energy. As shown in Figure 2C, the degradation rate did not increase significantly with the increase in the reaction temperature, mainly due to the relatively stable properties of PDS, thus PDS could not be activated by heat energy at low temperatures (25~45 °C) [44]. Dhaka et al. [32] also found the UV/PDS process was a time efficient process, compared to the heat-activated PDS oxidation of methyl paraben.

Effects of Various Factors on the Removal of Et-PABA
The concentration of oxidant dosage directly affects the amounts of free radicals produced; thus, it is necessary to study the effect of PDS dosage on the degradation of Et-PABA. As shown in Figure  3A, the degradation rate of Et-PABA increased as the concentrations of PDS increased from 1.0 to 4.0 mM within 15 min, which indicated that the removal of pollutants increased rapidly with the increase of oxidant dosage within a certain range. The main reason may be that as the concentration of PDS increases, the production of SO4 •− and • OH free radicals also increases, which promotes the degradation efficiency of Et-PABA in a short time [45,46]. The remaining concentration of PDS (when adding 2.0 mM PDS) was determined using the iodometric titration method. The remaining rate was Subsequently, we also studied the removal efficiency of Et-PABA using PDS activated by heat energy. As shown in Figure 2C, the degradation rate did not increase significantly with the increase in the reaction temperature, mainly due to the relatively stable properties of PDS, thus PDS could not be activated by heat energy at low temperatures (25~45 • C) [44]. Dhaka et al. [32] also found the UV/PDS process was a time efficient process, compared to the heat-activated PDS oxidation of methyl paraben.

Effects of Various Factors on the Removal of Et-PABA
The concentration of oxidant dosage directly affects the amounts of free radicals produced; thus, it is necessary to study the effect of PDS dosage on the degradation of Et-PABA. As shown in Figure 3A, the degradation rate of Et-PABA increased as the concentrations of PDS increased from 1.0 to 4.0 mM within 15 min, which indicated that the removal of pollutants increased rapidly with the increase of oxidant dosage within a certain range. The main reason may be that as the concentration of PDS increases, the production of SO 4 − and OH free radicals also increases, which promotes the degradation efficiency of Et-PABA in a short time [45,46]. The remaining concentration of PDS (when adding 2.0 mM PDS) was determined using the iodometric titration method. The remaining rate was about 27%, indicating that increasing the oxidant dosage is effective but not economical.  The acidity may affect the existence of free radicals in the UV/PDS process. Thus, the effect of pH on the Et-PABA removal was examined in this experiment. The initial pH values of the solution were adjusted from 3 to 11 with 0.1 M H2SO4 or 0.1 M NaOH, respectively, to investigate the impact of pH on the removal efficiency. As Figure 3B shows, the removal efficiency of Et-PABA was slightly better under acidic conditions, while it was slightly worse under the alkaline conditions in a short time of less than 20 min. It may be due to the existence of a large amount of H + in the reaction system which reacts with PDS to produce more SO4 •− , the predominate radicals in the UV/PDS system. SO4 •− has a higher redox potential under acidic conditions, thereby accelerating the reaction rate [47]. Liang and coworkers found that acid-catalyzed decomposition of persulfate anion under acidic pH gives a higher concentration of SO4 •− [48]. The slight decrease in the removal efficiency at pH 11 may be due to the fewer radicals generated as a result of the alkaline condition. However, as the reaction progressed, the final removal efficiency was greater than 97%, indicating that a wide range of pH value was applicable for the Et-PABA degradation using the UV/PDS process. From these results, the UV/PDS process is expected to be a prospective treatment in the remediation of emerging contaminants in wastewater.
In the case of scavenger reactions between the radicals generated during the UV/PDS process and the large number of anions in the natural water body such as Cl − , NO3 − , SO4 2− , HCO3 − , H2PO4 − , it is of great significance to investigate the effects of anions in water on the degradation of Et-PABA in the UV/PDS process [49]. To compare the effects of different anions on the degradation of Et-PABA, sodium chloride, sodium nitrate, sodium sulfate, sodium hydrogen carbonate, and sodium dihydrogen phosphate were added to the reaction system at a concentration of 5 mM [31], respectively. As shown in Figure 4A, Cl − showed no inhibition effect on the decomposition, which may be due to the fact that Clcan react with sulfate radicals and produce relatively weak free radicals such as Cl • , Cl2 −• and ClHO • [32]. HCO3 − anions were thought to be scavengers of hydroxyl radicals that can react with sulfate radicals to produce bicarbonate radicals in the solution. Previous studies have shown that Cl − and HCO3 − at small concentrations can promote the degradation of organic compounds by the UV/PDS process [50]. The other ions, such as NO3 − and SO4 2− , had no significant effect on the removal efficiency of Et-PABA. Although most of the common ions in the water matrix may affect the degradation rate of Et-PABA, the removal efficiency of Et-PABA using the UV/PDS process in the presence of anions was still impressive when the reaction time increased to more than 20 min. The acidity may affect the existence of free radicals in the UV/PDS process. Thus, the effect of pH on the Et-PABA removal was examined in this experiment. The initial pH values of the solution were adjusted from 3 to 11 with 0.1 M H 2 SO 4 or 0.1 M NaOH, respectively, to investigate the impact of pH on the removal efficiency. As Figure 3B shows, the removal efficiency of Et-PABA was slightly better under acidic conditions, while it was slightly worse under the alkaline conditions in a short time of less than 20 min. It may be due to the existence of a large amount of H + in the reaction system which reacts with PDS to produce more SO 4 − , the predominate radicals in the UV/PDS system. SO 4 − has a higher redox potential under acidic conditions, thereby accelerating the reaction rate [47].
Liang and coworkers found that acid-catalyzed decomposition of persulfate anion under acidic pH gives a higher concentration of SO 4 − [48]. The slight decrease in the removal efficiency at pH 11 may be due to the fewer radicals generated as a result of the alkaline condition. However, as the reaction progressed, the final removal efficiency was greater than 97%, indicating that a wide range of pH value was applicable for the Et-PABA degradation using the UV/PDS process. From these results, the UV/PDS process is expected to be a prospective treatment in the remediation of emerging contaminants in wastewater.
In the case of scavenger reactions between the radicals generated during the UV/PDS process and the large number of anions in the natural water body such as Cl − , NO 3 − , SO 4 2− , HCO 3 − , H 2 PO 4 − , it is of great significance to investigate the effects of anions in water on the degradation of Et-PABA in the UV/PDS process [49]. To compare the effects of different anions on the degradation of Et-PABA, sodium chloride, sodium nitrate, sodium sulfate, sodium hydrogen carbonate, and sodium dihydrogen phosphate were added to the reaction system at a concentration of 5 mM [31], respectively. As shown in Figure 4A, Cl − showed no inhibition effect on the decomposition, which may be due to the fact that Cl − can react with sulfate radicals and produce relatively weak free radicals such as Cl , Cl 2 − and ClHO [32]. HCO 3 − anions were thought to be scavengers of hydroxyl radicals that can react with sulfate radicals to produce bicarbonate radicals in the solution. Previous studies have shown that Cl − and HCO 3 − at small concentrations can promote the degradation of organic compounds by the UV/PDS process [50]. Humic acid (HA) is one of the main natural organic matter (NOM) in natural water. Thus, it's typically used as a representative of NOM to investigate the NOM effect on the UV/PDS oxidation process. Thus, the effects of HA at different concentrations on the removal of Et-PABA were investigated as shown in Figure 4B. The degradation rate of Et-PABA decreased with the increase in the HA concentration. The removal efficiencies at 30 min were 97.4%, 92.3%, and 66.8%, respectively, suggesting that HA had a significant inhibitory effect on the removal of Et-PABA. HA molecules contain a large amount of photosensitive substances such as aromatic carboxyl groups, hydroxyl groups, amine groups, which can directly absorb ultraviolet light, competitively inhibit the absorption of light energy by the PDS [51], and reduce the yield of free radicals in the solution. At the same time, the active group of HA will also compete with Et-PABA for active radicals such as SO4 •− , thus reducing the removal efficiency of Et-PABA [52,53].
In order to study the degradation of Et-PABA in real water bodies, two water samples from different sources (river water and tap water from Jiaxing College) were collected and filtered through a 0.45 μm filter. In order to compare with the Et-PABA degradation in pure water, the same concentration of Et-PABA (0.05 mM) was added to the two real water samples, respectively. The effect of real water on the degradation of Et-PABA using the UV/PS process is shown in Figure 5. The degradation rate of Et-PABA in the tap water reached as high as 96.5% within 10 min, which was essentially the same as that in pure water. The degradation rate of Et-PABA in surface water decreased, but the removal efficiency also reached 94.3% after a reaction time of 20 min. This result may be attributed to the inhibition of NOM in the real water body and the interaction of other water components, as discussed above. These results suggest that the UV/PDS process has good adaptability regarding the Et-PABA removal of real water samples. Chen et al. [54] also studied the degradation of 2,4-dichlorophenol in the real water using the UV/PDS process. They determined the water quality of several real water bodies. The concentrations of Cl − , HCO3 − , and NO3 − in different water samples were 42.5-63.0 mg/L, 128.0-136.6 mg/L, and 0.005-0.070 mg/L, respectively. The results also suggested the UV/PDS process can be further extended to practical applications. Humic acid (HA) is one of the main natural organic matter (NOM) in natural water. Thus, it's typically used as a representative of NOM to investigate the NOM effect on the UV/PDS oxidation process. Thus, the effects of HA at different concentrations on the removal of Et-PABA were investigated as shown in Figure 4B. The degradation rate of Et-PABA decreased with the increase in the HA concentration. The removal efficiencies at 30 min were 97.4%, 92.3%, and 66.8%, respectively, suggesting that HA had a significant inhibitory effect on the removal of Et-PABA. HA molecules contain a large amount of photosensitive substances such as aromatic carboxyl groups, hydroxyl groups, amine groups, which can directly absorb ultraviolet light, competitively inhibit the absorption of light energy by the PDS [51], and reduce the yield of free radicals in the solution. At the same time, the active group of HA will also compete with Et-PABA for active radicals such as SO 4 − , thus reducing the removal efficiency of Et-PABA [52,53]. In order to study the degradation of Et-PABA in real water bodies, two water samples from different sources (river water and tap water from Jiaxing College) were collected and filtered through a 0.45 µm filter. In order to compare with the Et-PABA degradation in pure water, the same concentration of Et-PABA (0.05 mM) was added to the two real water samples, respectively. The effect of real water on the degradation of Et-PABA using the UV/PS process is shown in Figure 5. The degradation rate of Et-PABA in the tap water reached as high as 96.5% within 10 min, which was essentially the same as that in pure water. The degradation rate of Et-PABA in surface water decreased, but the removal efficiency also reached 94.3% after a reaction time of 20 min. This result may be attributed to the inhibition of NOM in the real water body and the interaction of other water components, as discussed above. These results suggest that the UV/PDS process has good adaptability regarding the Et-PABA removal of real water samples. Chen et al. [54] also studied the degradation of 2,4-dichlorophenol in the real water using the UV/PDS process. They determined the water quality of several real water bodies. The concentrations of Cl − , HCO 3 − , and NO 3 − in different water samples were 42.5-63.0 mg/L, 128.0-136.6 mg/L, and 0.005-0.070 mg/L, respectively. The results also suggested the UV/PDS process can be further extended to practical applications.    Figure 6 presents the possible transformation pathways of Et-PABA based on the six observed TPs. On one hand, SO4 •− and • OH can directly attack the carbon on the benzene ring, resulting in hydroxylation of Et-PABA. Through this degradation pathway, TP1 was produced. This product was further hydroxylated to produce a product of TP2. On the other hand, TP3 was a product of Et-PABA formed after attacks by ROS caused the C-O bond between the carbonyl group and the ethoxy group to be broken. Under further actions of ROS, due to the oxidation or removal of the amino group, TP3 was converted into TP4 and TP6, respectively. Similarly, hydroxylation of TP4 produced TP5. Finally, these intermediates of Et-PABA can be further degraded to lower organic acids.     Figure 6 presents the possible transformation pathways of Et-PABA based on the six observed TPs. On one hand, SO4 •− and • OH can directly attack the carbon on the benzene ring, resulting in hydroxylation of Et-PABA. Through this degradation pathway, TP1 was produced. This product was further hydroxylated to produce a product of TP2. On the other hand, TP3 was a product of Et-PABA formed after attacks by ROS caused the C-O bond between the carbonyl group and the ethoxy group to be broken. Under further actions of ROS, due to the oxidation or removal of the amino group, TP3 was converted into TP4 and TP6, respectively. Similarly, hydroxylation of TP4 produced TP5. Finally, these intermediates of Et-PABA can be further degraded to lower organic acids.

Conclusions
This study explored the removal efficiency, mechanism, and a novel advanced oxidation route for Et-PABA in a water environment using the UV/PDS combined process. The results showed that the combined treatment could significantly improve the degradation of Et-PABA compared to the single processes. The removal efficiency of Et-PABA after the UV/PDS combined process can reach 98.7% in 30 min. The removal rate of Et-PABA increased rapidly upon increasing the dosage of oxidant.