Efficient Degradation of Acesulfame by Ozone/Peroxymonosulfate Advanced Oxidation Process

Artificial sweeteners (ASWs), a class of emerging contaminants with good water solubility, have attracted much attention recently because of their wide use and negative impact on the aquatic environment and drinking water. Efficient technologies for removing ASWs are in urgent need. This study investigated degradation of typical ASW acesulfame by ozone-activated peroxymonosulfate process (O3/PMS) in prepared and real waters. O3/PMS can degrade >90% acesulfame in prepared water within 15 min at a low dosage of O3 (60 ± 5 µg∙min−1) and PMS (0.4 mM). Ozone, hydroxyl radical (HO•), and sulfate radical (SO4•−) were identified as contributors for ACE degradation and their contribution proportion was 27.1%, 25.4%, and 47.5% respectively. O3/PMS showed the best degradation performance at neutral pH and were sensitive to constituents such as chloride and natural organic matters. The qualitative analysis of degradation products confirmed the involvement of hydroxyl radical and sulfate radical and figured out that the active sites of ACE were the C=C bond, ether bond, and C-N bond. The electrical energy per order ACE degradation were calculated to be 4.6 kWh/m3. Our findings indicate that O3 is an efficient PMS activator and O3/PMS is promising due to its characteristic of tunable O3−HO• SO4•− ternary oxidant involving.


Introduction
As a class of emerging pollutant, artificial sweeteners (ASs), have recently received increasing attention [1][2][3]. ASs are synthetic or semi-synthetic organic compounds that replace sucrose and are widely used in food, beverage, pharmaceutical, and personal care products [4]. Most artificial sweeteners are hardly converted by the human body (called as non-caloric sugars) and are generally highly water-soluble. Thus, the aqueous environment is their main destination. There are more than 20 kinds of ASs currently used, and there are five kinds of sweeteners that are often considered in the water environment, namely saccharin (SAC), cyclamate (CYC), aspartame (ASP), acesulfame (ACE), and sucralose (SUC) [5]. The high water solubility, large amount of use, and anti-biodegradation property (SAC, CYC, ACE, and SUC biodegradation cycle > 15d, [6]) of ACSs make them frequently detected in surface water [2], groundwater [7], drinking water [6,8], and sewage treatment plant effluent [2]. Concentration of ASs in drinking water has reported to be tens of ng·L −1 to several hundred µg·L −1 , which is much higher than that of other emerging pollutants such as drugs and personal care products and endocrine disruptors. The toxicology of ACSs is not clear yet, but its negative effects on the human health have been reported [9]. (1) To fill the abovementioned gaps, the ACE degradation by O 3 /PMS was particularly focused in this study. The contributions of the reactive oxidative species were distinguished. Moreover, the mineralization rate and degradation products were detected. Specifically, the degradation behaviors in real waters (four effluent of filter tank of waterworks) were tentatively studied for the first time. Then, the influence of operational parameters (dosage of O 3 and PMS) and common water quality parameters (solution pH, bicarbonate, chloride, and natural organic materials (NOM)) on the degradation processes was systematically investigated. Finally, the economic cost was evaluated.

Degradation Effeciency of ACE by O 3 /PMS
ACE degradation by O 3 , PMS, and O 3 /PMS were compared. The results are presented in Figure 1a. PMS oxidation alone nearly did not degrade ACE for 15 min reaction time, while O 3 oxidation showed a 52.7% degradation rate of ACE at the same reaction time. The fastest ACE degradation (90.4%) was observed in O 3 /PMS system. O 3 oxidation usually includes direct oxidation (pollutants react with O 3 molecular directly) and indirect oxidation (pollutants react with radicals generated from O 3 decomposition). Given the high oxidation potential of O 3 (2.07 V), direct oxidation of O 3 was believed to play an important role. As O 3 can activate PMS to produce SO •− 4 and HO• (Equations (1)-(6)), the excellent degradation performance of O3/PMS process may be also contributed by these two oxidative radicals. The activation effect of O 3 on PMS was confirmed by the accelerated PMS decomposition rate in the presence of O 3 (Figure 1b). believed to play an important role. As O3 can activate PMS to produce SO • and HO• (Equations 1-6), the excellent degradation performance of O3/PMS process may be also contributed by these two oxidative radicals. The activation effect of O3 on PMS was confirmed by the accelerated PMS decomposition rate in the presence of O3 (Figure 1b).   (Figure 2a). That is, 0.4 mM PMS combining with O3 dosing 60 ± 5 µg·min −1 showed the best degradation of ACE (Figure 2b). Because ozone was dosed in a continuous way and PMS was added in one time, the scavenging effect of HO• by PMS (HSO ) is expected to be more and more significant with the dosage increase of PMS (Equations 7-10, [25][26][27][28]) and consequently leads to a deteriorating degradation performance. Therefore, a PMS dosage of 0.4 mM was used in the following experiments.  is expected to be more and more significant with the dosage increase of PMS (Equations (7)-(10), [25][26][27][28]) and consequently leads to a deteriorating degradation performance. Therefore, a PMS dosage of 0.4 mM was used in the following experiments.

Contributions of Different Reactive Species
Based on the above discussion, we can preliminarily assume that ACE degradation mainly contributed by direct O3 oxidation and SO • / HO• attack. To clarify this issue, TBA (HO• scavenger, [19]) and EtOH (scavenger of both HO• and SO • , [19]) were introduced into the O3/PMS system. As shown in Figure 3, the addition of TBA and MeOH made the ACE degradation decrease by 22 (Figure 4).

Contributions of Different Reactive Species
Based on the above discussion, we can preliminarily assume that ACE degradation mainly contributed by direct O 3 oxidation and SO •− 4 / HO• attack. To clarify this issue, TBA (HO• scavenger, [19]) and EtOH (scavenger of both HO• and SO •− 4 , [19]) were introduced into the O 3 /PMS system. As shown in Figure 3

Degradation Products
Considering that the degradation products of an oxidation system are usually highly associated with the oxidative species, degradation products of ACE by O3/PMS were determined through HPLC-MS to testify the participation of HO• and SO • . The 32.8% total organic carbon (TOC) removal rate ( Figure 5) indicates that many transformation intermediates are generated.

Degradation Products
Considering that the degradation products of an oxidation system are usually highly associated with the oxidative species, degradation products of ACE by O3/PMS were determined through HPLC-MS to testify the participation of HO• and SO • . The 32.8% total organic carbon (TOC) removal rate ( Figure 5) indicates that many transformation intermediates are generated.

Degradation Products
Considering that the degradation products of an oxidation system are usually highly associated with the oxidative species, degradation products of ACE by O 3 /PMS were determined through HPLC-MS to testify the participation of HO• and SO •− 4 . The 32.8% total organic carbon (TOC) removal rate ( Figure 5) indicates that many transformation intermediates are generated. As shown in the mass spectra ( Figure 6), several obvious peaks (m/z=117, 164, 178) were observed, indicating that ACE was transformed into several intermediates. The ACE molecular possessed a charge-to-mass ratio (m/z) of 162. Like the previous studies [29], a hydroxylated product of ACE (m/z 179.1, P1) was detected in present work. SO • was ready to undergo reaction with organic pollutants through electron transfer. Sulfate radicals react with the olefinic double bond of ACE and form shortlived sulfate radical adducts [30]. Then nucleophilic attack of water and oxygen on the SO • adducts results in the formation of hydroxylated product. Hydroxylated product can also be generated via electron transfer from the double bond to SO • , causing the formation of the intermediate radical [31]. The latter reacts with water and oxygen to produce the hydroxylated product too. The HO• attack on the organic molecular mainly follows or electrophilic addition or hydrogen abstraction mechanism. The detected hydroxylated product can be formed by HO• addition on double bond and dehydration [32]. In addition, the intermediate with an m/z of 165.1 (P2) was also identified in the oxidation processes. This product can be formed through HO• addition on double bond and demethylation. Besides these two products, an intermediate with m/z of 118.1 (P3) appeared in the mass spectra. Such intermediate can be produced from P2 decomposition through break of C-O and C-N bonds. Based on the information of these identified products, HO• and SO • are believed to involve in the degradation of ACE and the attack sites are C=C, C-O, and C-N bonds. As shown in the mass spectra ( Figure 6), several obvious peaks (m/z = 117, 164, 178) were observed, indicating that ACE was transformed into several intermediates. The ACE molecular possessed a charge-to-mass ratio (m/z) of 162. Like the previous studies [29], a hydroxylated product of ACE (m/z 179.1, P1) was detected in present work. SO •− 4 was ready to undergo reaction with organic pollutants through electron transfer. Sulfate radicals react with the olefinic double bond of ACE and form short-lived sulfate radical adducts [30]. Then nucleophilic attack of water and oxygen on the SO •− 4 adducts results in the formation of hydroxylated product. Hydroxylated product can also be generated via electron transfer from the double bond to SO •− 4 , causing the formation of the intermediate radical [31]. The latter reacts with water and oxygen to produce the hydroxylated product too. The HO• attack on the organic molecular mainly follows or electrophilic addition or hydrogen abstraction mechanism. The detected hydroxylated product can be formed by HO• addition on double bond and dehydration [32]. In addition, the intermediate with an m/z of 165.1 (P2) was also identified in the oxidation processes. This product can be formed through HO• addition on double bond and demethylation. Besides these two products, an intermediate with m/z of 118.1 (P3) appeared in the mass spectra. Such intermediate can be produced from P2 decomposition through break of C-O and C-N bonds. Based on the information of these identified products, HO• and SO •− 4 are believed to involve in the degradation of ACE and the attack sites are C=C, C-O, and C-N bonds.

Effect Water Matrix Components on ACE Degradation
Considering the possible scavenging effects of background water matrices, the degradation performance of ACE by O 3 /PMS in four real waters was also tested. Table 1 summarizes the water quality parameters of these four real waters (RWs). They are significantly different in indexes of dissolved organic matters (DOC), alkalinity, Cl − , NO − 3 , SO 2− 4 , and Ca 2+ . As shown in Figure 7, the degradation rates of ACE in RWs generally suffered some extent of decrease compared to the case of DI water. It may result from scavenging of HO• and SO •− 4 by cosolutes like natural organic matters (NOM) and bicarbonate (HCO − 3 ). Such significant inhibition of ACE degradation by background cosolutes makes screening of main inhibitors in real waters necessary. Thus, we evaluate the effects of possibly relevant water quality parameters one by one.

Effect Water Matrix Components on ACE Degradation
Considering the possible scavenging effects of background water matrices, the degradation performance of ACE by O3/PMS in four real waters was also tested. Table 1 summarizes the water quality parameters of these four real waters (RWs). They are significantly different in indexes of dissolved organic matters (DOC), alkalinity, Cl , NO , SO , and Ca . As shown in Figure 7, the degradation rates of ACE in RWs generally suffered some extent of decrease compared to the case of DI water. It may result from scavenging of HO• and SO • by cosolutes like natural organic matters (NOM) and bicarbonate (HCO ). Such significant inhibition of ACE degradation by background cosolutes makes screening of main inhibitors in real waters necessary. Thus, we evaluate the effects of possibly relevant water quality parameters one by one.    ACE degradation efficiency increased with pH elevation in the pH range 5.0-7.4 and the removal rate dropped from 89.3% to 77.9% as the pH further increased from 7.4 to 8.0 (Figure 8a). The increase of pH from 8.0 to 9.0 made the degradation rate decrease to 39.8%. These results indicated that the most efficient degradation of ACE by O3/PMS is under neutral condition (insert in Figure 8a). Notably, Yang et al. [23] found that degradation of nitrobenzene and atrazine were promoted with increasing pH, which is different from what we observed here. Based on Equations 1-6, the primary precursors of SO • /HO• are SO • /O • and the increasing pH will inhibit the formation of HO•, which cannot explain the phenomena observed in present work. According to the mechanism proposed by Tomiyasu, Fukutomi, and Gordon (TFG mechanism) [33], O3 can react with OH to produce hydroperoxide (HO ) under neutral or alkaline condition (Equation 13, [26] [20]). In summary, when the solution pH shifted from neuter to alkaline region, the proportion of O3 directly reacting with ACE dropped, leading to enhanced formation of SO • and suppressed formation of HO•. Given that SO • degraded ACE more slowly than HO• did (Equations 20-21, [35]), the oxidation capacity of the system was weakened due to the decrease of HO•. Thus, we can reasonably explain the inhibition effect caused by pH increase from 7.4 to 9.0. ACE degradation efficiency increased with pH elevation in the pH range 5.0-7.4 and the removal rate dropped from 89.3% to 77.9% as the pH further increased from 7.4 to 8.0 (Figure 8a). The increase of pH from 8.0 to 9.0 made the degradation rate decrease to 39.8%. These results indicated that the most efficient degradation of ACE by O 3 /PMS is under neutral condition (insert in Figure 8a). Notably, Yang et al. [23] found that degradation of nitrobenzene and atrazine were promoted with increasing pH, which is different from what we observed here. Based on Equations (1)-(6), the primary precursors of SO •− 4 /HO• are SO •− 5 /O •− 3 and the increasing pH will inhibit the formation of HO•, which cannot explain the phenomena observed in present work. According to the mechanism proposed by Tomiyasu, Fukutomi, and Gordon (TFG mechanism) [33], O 3 can react with OH − to produce hydroperoxide (HO − 2 ) under neutral or alkaline condition (Equation (13), [26]). Similarly, PMS can also react with OH − to form HO − 2 (Equation (14), [34]). Hydroperoxide reacts with O 3 and PMS to generate HO• (Equation (15), [26]) and SO •− 4 (Equation (16), [13], respectively. Because PMS was in excess over O 3 , the formed HO − 2 was believed to mainly react with PMS. In addition, conversion of SO •− 4 to HO• and HO• to O •− is weak under conditions of pH<9.0 according to previous work (Equation (17)- (19), [20]). In summary, when the solution pH shifted from neuter to alkaline region, the proportion of O 3 directly reacting with ACE dropped, leading to enhanced formation of SO •− 4 and suppressed formation of HO•. Given that SO •− 4 degraded ACE more slowly than HO• did (Equations (20)-(21), [35]), the oxidation capacity of the system was weakened due to the decrease of HO•. Thus, we can reasonably explain the inhibition effect caused by pH increase from 7.4 to 9.0. As can be seen from Figure 6 (e), temperature was not a factor which significantly affected the ACE degradation. ACE removal rate increased slightly when temperature rose from 5 to 40 ℃. These results indicate that O3/PMS process is not thermodynamically controlled in the investigated temperature range. This is quite similar to O3/H2O2, which was almost not influenced by reaction temperature [41].

EE/O Analysis
In order to determine whether O3/PMS is cost-effective for a given situation, EE/O concept was applied [42]. The electrical energy related to O3 and PMS consumption (EE/O and EE/O ) which is required for an order of ACE removal (i.e., 90% destruction of ACE) were calculated using Equations 31-33: where [O3] and [PMS] are the amounts of O3 and PMS consumption with the unit of g·L −1 , k and Natural organic matter (NOM) is also a common radical scavenger in real waters. Here, HA was selected as representative of NOM to research the effect of NOM. It can be seen from Figure 8d  As can be seen from Figure 6e, temperature was not a factor which significantly affected the ACE degradation. ACE removal rate increased slightly when temperature rose from 5 to 40°C. These results indicate that O 3 /PMS process is not thermodynamically controlled in the investigated temperature range. This is quite similar to O 3 /H 2 O 2 , which was almost not influenced by reaction temperature [41].

EE/O Analysis
In order to determine whether O 3 /PMS is cost-effective for a given situation, EE/O concept was applied [42]. The electrical energy related to O 3

and PMS consumption (EE/O O 3 and EE/O PMS )
which is required for an order of ACE removal (i.e., 90% destruction of ACE) were calculated using Equations (31)- (33):
HA stock solution was prepared in a procedure similar to that described in our previous work [45]. The accurate concentration of the HA stock solution was calibrated using a total organic carbon (TOC)-VCPH analyzer (Shimadzu, Japan).

Experimental Procedures
A Guolin CF-G-3-10g ozone generator (Qingdao, China) was used to produce O 3 . Then O 3 stock solution was prepared by bubbling O 3 into 1500 ml DI water of pH = 4.0 (adjusted with HClO 4 ) which was cooled by ice bath. The experiment was conducted in a 500-mL glass reactor. ACE (8.0 mg·L −1 ) was initially prepared with ultrapure water. HClO 4 /NaOH (0.1 M) was used to adjust pH value from 5.0-6.0 and 2 mM borate buffer was used to adjust pH value from 7.4 to 9.0. A magnetic stirrer was used to mix the reaction solution evenly throughout the whole process. Using a water bath to maintain the temperature of the reaction solution at 15 • C so as to slow down the decomposition of O 3 itself. PMS solution (100 mM) was then added to generate an initial concentration of 0.4 mM. At the same time, O 3 solution was added to the reaction system by a peristaltic pump (Longer, Baoding, China) at a dosing rate of 60 ± 5 µg·min −1 . Timing was started simultaneously. The O 3 concentration was determined immediately before and after the reaction, and the average of the two concentrations was taken to calculate the dosage of O 3 . The residual oxidant in each sample was removed by NaNO 2 before HPLC analysis.

Analysis Methods
The concentration of O 3 solution was determined by indigo method [46]. The absorbance at the wavelength of 612 nm was detected by a Hach DR6000 ultraviolet-visible spectrophotometer (Hach, Loveland, CO, USA). ACE was quantified by an Agilent 1200 HPLC (Agilent, Palo Alto, CA, USA). Separation was performed with an Agilent Eclipse XDB-C18 column (5 µm, 4.6 × 150 mm) at 30 • C. The mobile phase consisted of 90% ammonium acetate (0.02 mol·L −1 ) and 10% methanol and had a flow rate of 1mL·min −1 . Detection wavelength was set at 230 nm. 20 µL sample injection was employed. Products analysis was performed by Agilent 6460 triple-quad HPLC-MS (Agilent, Palo Alto, CA, USA). The samples were concentrated by solid phase extraction 50 times before product analysis.
Typical water quality indexes were measured for the four collected effluent samples of waterworks filter tank. Alkalinity (as CO 2− 3 ) was quantified according to the Standard Methods for the Examination of Water and Wastewater [47]. DOC (sample were filtrated with 0.45 µm membrane) and TOC was determined via a Shimadzu TOC analyzer (Shimadzu, Kyoto, Japan). The concentrations of cations (Ca 2+ , Mn 2+ , Cu 2+ , and total Fe) were determined using a PerkinElmer NexION 350Q ICP-MS Spectrometer (PerkinElmer, Shelton, CT, USA). The Cl − and NO − 3 measurements were carried out via a Dionex ICS-2000 ion chromatograph (Chameleon 6.8, Sunnyvale, CA, USA). UV absorbance at 254 nm (UV 254 ) was determined with a Shimadzu UV-250 spectrophotometer (Shimadzu, Kyoto, Japan). The pH was determined using an Orion 3-Star pH meter (Thermo Fisher, Shanghai, China).

Conclusions
In this study, the ACE degradation by the system of O 3 /PMS was studied in detail. It was demonstrated that efficient degradation of ACE was achieved due to the coaction of O 3

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