Remediation of Surfactants Used by VUV/O3 Techniques: Degradation Efficiency, Pathway and Toxicological Analysis

Surfactants are increasingly used in systems that come into contact with the human body, such as food, pharmaceuticals, cosmetics and personal hygiene products. Increasing attention is being devoted to the toxic effects of surfactants in various human contact formulations, as well as the removal of residual surfactants. In the presence of ozone (O3), anion surfactants—a characteristic micro-pollutant—such as sodium dodecylbenzene sulfonate (SDBS) in greywater, can be removed using radical advanced oxidation. Herein, we report a systematic study of the SDBS degradation effect of O3 activated by vacuum ultraviolet (VUV) irradiation and the influence of water composition on VUV/O3, and determined the contribution of radical species. We show a synergistic effect of VUV and O3, while VUV/O3 reached a higher mineralization (50.37%) than that of VUV (10.63%) and O3 (29.60%) alone. The main reactive radicals of VUV/O3 were HO•. VUV/O3 had an optimal pH of 9. The addition of SO42− had almost no effect on the degradation of SDBS by VUV/O3, Cl− and HCO3− slightly reduced the reaction rate, and NO3− had a significant inhibition on the degradation. In total, SDBS had three isomers, with which the three degradation pathways were very comparable. Compared with SDBS, the toxicity and harmfulness of the degradation by-products of the VUV/O3 process decreased. Additionally, VUV/O3 could degrade synthetic anion surfactants from laundry greywater effectively. Overall, the results show the potential of VUV/O3 in safeguarding humans from residual surfactant hazards.


Introduction
The biological changes caused by surfactants and their metabolites in the body, i.e., the possible toxic side effects on the body, include acute toxicity, subacute toxicity, chronic toxicity, reproductive effects, embryotoxicity, teratogenicity, mutagenicity, carcinogenicity, allergenicity, hemolysis, etc. [1]. Surfactants are used as various processing aids or synergists in the food and pharmaceutical industries. This increases the exposure of surfactants to the human digestive tract and blood system. Hence, strict requirements are imposed on the oral toxicity, hemolysis, heredity, carcinogenicity and teratogenicity of surfactants [2]. When the concentration of surfactants is relatively high, it can adversely affect aquatic organisms [3]. For example, chronic and sublethal toxicity to aquatic animals occurs when the concentration of anionic surfactants exceeds 0.1 mg/L [4]. Surfactants also inhibit the growth, activity and photosynthetic capacity of aquatic algae [5]. Moreover, a number of synthetic surfactants and their decomposition products do present a potential risk to human health and the environment, and they tend to increase the problem of water treatment [1].
Surfactants are a class of chemicals that reduce the surface tension of liquid or liquid interfaces. They have hydrophobic tails and hydrophilic heads. Surfactants are usually found as monomers, but sometimes in more complex polymeric forms [6]. According to the charge of hydrophilic groups of surfactants, they are divided into anionic, cationic, and nonionic, semi-polar, and amphoteric. SDBS is a typical anionic surfactant, which has been Figure 1 illustrates the procedure of SDBS degraded by O 3 treatment, VUV treatment, and combined VUV/O 3 treatment. During the 30 min of reaction, VUV/O 3 was efficient to degrade SDBS at the 223 nm optical band as shown in Figure 1a. Based on the DFT calculation as shown in Figures S2-S4, 16(C)-21(C) on-ring bonds were the more vulnerable sites to radicals' attack than the (C) on carbon chain. Interestingly, with the progress of VUV/O 3 treatment, the absorbance peak widened, and a visible shoulder peak appeared, indicating the presence of intermediates. According to previous studies, SDBS can be effectively degraded by O 3 itself as well as by generating free radicals, and is more efficient than other commonly used oxidants [11]. At the same time, O 3 can be activated by VUV irradiation to increase the yield of radical species, which can enhance the degradation effect of surfactant [11]. Overall, it can be concluded that VUV/O 3 was an effective method for degrading surfactants.

Types and Roles of Radicals Existing in VUV/O3 Process
In order to obtain a quantitative evaluation of the role of various radicals on the SDBS degradation, the relevant reaction rate constants were calculated from the gradient dose of scavengers. The results of the research are displayed in Figure 2. The inhibitory degrees of SDBS degradation of two used scavengers were correlated with the oxidative effect of the relative radicals. As the TBA concentration increased from 1 to 100 mM, the degradation efficiency of SDBS by VUV/O3 decreased by 64.45-78.92%; for FA there was less of a concentration effect but an overall stronger inhibition: 27.58-51.24% for 5 µL/L-50 µL/L FA, respectively. Since the inhibition of the reaction was evident by the injection of low concentrations of TBA, it indicated that HO• was the main active substance in the degradation of SDBS by VUV/O3. Notably, the inhibition of the FA increased markedly with the increase of FA concentration. However, FA can preferentially absorb 185 nm photons over other components (H2O, O3, SDBS), so it can be inferred that O3 could also degrade SDBS,  Figure 1b shows the removal efficiency of SDBS by different treatments. It can be observed that C t /C 0 progressively reduced with time, demonstrating an effective removal of SDBS. The removal efficiency of SDBS by single VUV and single O 3 were only 59.21% and 71.27% when the time 30 min, respectively. The VUV/O 3 had a better effect by 99.35% for SDBS removal. It can be found that VUV/O 3 can promote the removal efficiency of SDBS by O 3 due to its higher radical species yield. Figure 1c depicts the mineralization of SDBS by the above three treatments during a 30 min reaction time. It can be seen that the DOC t /DOC 0 gradually decreased with time, indicating a gradual increase in the mineralization of SDBS. The effect of single VUV on the Molecules 2023, 28, 3312 4 of 18 removal of DOC from SDBS solution was not obvious, and the removal efficiency of DOC by single O 3 was only 29.60%. The coupling of VUV and O 3 had a high mineralization capacity (50.37%) for SDBS and showed a synergistic effect. This is due to the fact that O 3 is able to generate radical species via other pathways as the VUV was activated, at which point O 3 utilization increased significantly [29]. As the removal of SDBS by VUV/O 3 reached 99.35% during 30 min, these results suggest that the generated intermediates make an important contribution to the determination of DOC. It is clear that VUV combining with O 3 produced a synergistic effect and enhanced the mineralization capacity of O 3 . Figure 1d shows the concentration of SO 4 2− for the three treatments. The production of SO 4 2− was significantly accelerated after 30 min of VUV/O 3 treatment, but the effect of O 3 and VUV treatments were not obvious and the final concentrations of SO 4 2− were 1.856 mg/L, 0.788 mg/L and 0.628 mg/L, respectively. According to DFT calculation in Tables S1-S3, 1(S) is most susceptible to attack by radicals to produce SO 4 2− , and their presence implies that advanced oxidation was effective to cleavage the S-O bond. Higher energy photons enhanced the activated efficiency of O 3 and led to the acceleration of the required electron transfer, which may account for the increased rate of SO 4 2− production [32]. Overall, this may imply the oxidation capacities of VUV alone and O 3 alone were enhanced by the coupling.

Types and Roles of Radicals Existing in VUV/O 3 Process
In order to obtain a quantitative evaluation of the role of various radicals on the SDBS degradation, the relevant reaction rate constants were calculated from the gradient dose of scavengers. The results of the research are displayed in Figure 2. The inhibitory degrees of SDBS degradation of two used scavengers were correlated with the oxidative effect of the relative radicals. As the TBA concentration increased from 1 to 100 mM, the degradation efficiency of SDBS by VUV/O 3 decreased by 64.45-78.92%; for FA there was less of a concentration effect but an overall stronger inhibition: 27.58-51.24% for 5 µL/L-50 µL/L FA, respectively. Since the inhibition of the reaction was evident by the injection of low concentrations of TBA, it indicated that HO• was the main active substance in the degradation of SDBS by VUV/O 3 . Notably, the inhibition of the FA increased markedly with the increase of FA concentration. However, FA can preferentially absorb 185 nm photons over other components (H 2 O, O 3 , SDBS), so it can be inferred that O 3 could also degrade SDBS, but not as effectively as HO• [31]. The results showed that both scavengers inhibited the degradation of SDBS during VUV/O 3 , and the degree of inhibition increased significantly with the increase of scavenger concentrations. It is thus clear that HO• played a more important role than UV photons and O 3 in the degradation of SDBS during the VUV/O 3 process. To further clarify the role of radicals, EPR spectra were performed to record. According to Figure 3, a 1:2:2:1 DMPO-OH characteristic peak spectrum could be recorded in O3 alone, and the intensity of such peaks was relatively diminished in VUV alone. Obviously, significant increase in the signal intensity of HO• by O3 under VUV activation [33]. This  To further clarify the role of radicals, EPR spectra were performed to record. According to Figure 3, a 1:2:2:1 DMPO-OH characteristic peak spectrum could be recorded in O 3 alone, and the intensity of such peaks was relatively diminished in VUV alone. Obviously, significant increase in the signal intensity of HO• by O 3 under VUV activation [33]. This explains why the mineralization and concentrations of SO 4 2− can be considerably increased during SDBS degradation by the VUV/O 3 system. From the analysis of each EPR spectrum, it is clear that VUV led to a considerable activated efficiency of O 3 , producing more HO• as Equations (1)-(3). All these confirm that HO• was the main active substance in the VUV system and its effect was greater than that of O 3 . To further clarify the role of radicals, EPR spectra were performed to record. Accord ing to Figure 3, a 1:2:2:1 DMPO-OH characteristic peak spectrum could be recorded in O alone, and the intensity of such peaks was relatively diminished in VUV alone. Obviously significant increase in the signal intensity of HO• by O3 under VUV activation [33]. Thi explains why the mineralization and concentrations of SO4 2− can be considerably increased during SDBS degradation by the VUV/O3 system. From the analysis of each EPR spectrum it is clear that VUV led to a considerable activated efficiency of O3, producing more HO• as Equations (1)-(3). All these confirm that HO• was the main active substance in the VUV system and its effect was greater than that of O3.   As shown in Figure 4, the removal rate of SDBS gradually increased with increasing ozone concentration. The constant k (SDBS removal) considerably increased from 0.064 to 0.162 min −1 with O 3 dosage ranging from 0.247 to 0.927 mg/L for the VUV/O 3 process. According to the calculations, at ozone doses of 0.927 mg/L, the k was 1.5-and 2.5-fold larger than the k for 0.247 mg/L and 0.462 mg/L, respectively. Increasing the ratio of ozone concentration to surfactant concentration may increase the surfactant removal rate. This was mainly due to an increase in O 3 concentration that increased the absorption efficiency of 185 nm and 254 nm photons, which led to an increase in the yield of radicals. Furthermore, ozone itself can be involved in SDBS degradation. Thus, an increase in ozone concentration can significantly increase the removal rate of SDBS. The result demonstrated that the decomposition of SDBS was highly effective during the VUV/O 3 process, and that the removal of SDBS was facilitated by increasing the concentration of dissolved O 3 .
was mainly due to an increase in O3 concentration that increased the absorption efficiency of 185 nm and 254 nm photons, which led to an increase in the yield of radicals Furthermore, ozone itself can be involved in SDBS degradation. Thus, an increase in ozon concentration can significantly increase the removal rate of SDBS. The resul demonstrated that the decomposition of SDBS was highly effective during the VUV/O process, and that the removal of SDBS was facilitated by increasing the concentration o dissolved O3.  As the concentrations were increased from 5 to 20 mg/L, the percentage absorbance of 185 nm photons by SDBS in competition with O 3 and H 2 O increased from 12.31% at 5 mg/L to 19.58%, 24.27%, and 26.61% as the concentrations increased from 10 to 20 mg/L, as depicted in Figure 6 [34]. In theory, the higher the absorption ratio of O 3 and H 2 O to 185 nm photons, the higher the activation efficiency of 185 nm photons. Since the concentration of O 3 dissolved in solution remained constant, an increase in [SDBS] 0 led to a decrease in the concentration of produced radicals, which would reduce the chance of effective collisions of radicals with contaminant molecules. Similarly, a previous study has also revealed that the removal of SDBS at lower rates by VUV/O 3 for higher initial concentrations of contaminants [35].
As the concentrations were increased from 5 to 20 mg/L, the percentage absorbanc of 185 nm photons by SDBS in competition with O3 and H2O increased from 12.31% at mg/L to 19.58%, 24.27%, and 26.61% as the concentrations increased from 10 to 20 mg/L as depicted in Figure 6 [34]. In theory, the higher the absorption ratio of O3 and H2O t 185 nm photons, the higher the activation efficiency of 185 nm photons. Since th concentration of O3 dissolved in solution remained constant, an increase in [SDBS]0 led t a decrease in the concentration of produced radicals, which would reduce the chance o effective collisions of radicals with contaminant molecules. Similarly, a previous study ha also revealed that the removal of SDBS at lower rates by VUV/O3 for higher initi concentrations of contaminants [35].
The yield of radicals was constant for a fixed reaction time at the same ozon concentration, and the radical/reactant ratio decreased as the initial concentration of SDB increased, leading to a decrease in the SDBS removal rate. Despite this, the degradatio efficiency of SDBS was still very high even at high concentrations. The results show tha VUV/O3 is a promising method for effective removal of anionic surfactants in greywate treatment [36]. The yield of radicals was constant for a fixed reaction time at the same ozone concentration, and the radical/reactant ratio decreased as the initial concentration of SDBS increased, leading to a decrease in the SDBS removal rate. Despite this, the degradation efficiency of SDBS was still very high even at high concentrations. The results show that VUV/O 3 is a promising method for effective removal of anionic surfactants in greywater treatment [36]. Figure 7 depicts the change in the degradation rate of SDBS degraded by VUV/O 3 as the pH increases from 3 to 11. Solution pH is a major environmental variable, among others, that affects the degradation rate of advanced oxidation reactions, as it may determine the applicability of a process under practical treatment. The change in pH value resulted in the highest degradation rate of VUV/O 3 process under weak alkaline conditions: when pH value is 9, the K value of VUV/O 3 is 0.254 min −1 . The rate of VUV/O 3 oxidation reaction was higher in an alkaline environment. SDBS can exist as dodecylbenzene sulfonic acid under acidic conditions, and the undissociated SDBS is not as vulnerable to oxidative HO• as the anionic form [37]. Meanwhile, under the strong acidic condition, O 3 oxidation dominates the degradation of SDBS and excess H 3 O + might become a scavenger for HO•, which increased the consumption of reactive HO• (Equations (4) and (5)), thus inhibiting the degradation of SDBS [38]. In the VUV/O 3 system, the higher the amount of HO − , the faster the decomposition of O 3 to HO•, which is responsible for the increased degradation rate of SDBS in a high pH environment. HO• was a more reactive than O 3 , and alkaline conditions improved the increase of HO• yield, which facilitated the removal of SDBS [19,39]. The optimal pH was 9 rather than 11, indicating that HO• had a stronger oxidizing effect 8 of 18 under weakly alkaline conditions and on ionized form of SDBS (DBS − ) [40]. Previous studies have also observed comparable effects of pH upon the VUV/O 3 process. Jiang et al. found that the optimal treatment condition by VUV/O 3 was at pH of 9.0 [29], and Zhang et al. found that the removal rate of sodium n-butyl xanthate increased with increasing alkalinity of the solution in VUV/O 3 process [30]. Experimental data verify that pH has a significant influence on the degradation of SDBS by VUV/O 3 , which is increased in strongly alkaline environments.

Performance of VUV/O 3 in Varied Initial pH
become a scavenger for HO•, which increased the consumption of reactive HO• (Equations (4) and (5)), thus inhibiting the degradation of SDBS [38]. In the VUV/O system, the higher the amount of HO − , the faster the decomposition of O3 to HO•, whic is responsible for the increased degradation rate of SDBS in a high pH environment. HO• was a more reactive than O3, and alkaline conditions improved the increase of HO• yield which facilitated the removal of SDBS [19,39]. The optimal pH was 9 rather than 11 indicating that HO• had a stronger oxidizing effect under weakly alkaline conditions and on ionized form of SDBS (DBS − ) [40]. Previous studies have also observed comparabl effects of pH upon the VUV/O3 process. Jiang et al. found that the optimal treatmen condition by VUV/O3 was at pH of 9.0 [29], and Zhang et al. found that the removal rat of sodium n-butyl xanthate increased with increasing alkalinity of the solution in VUV/O process [30]. Experimental data verify that pH has a significant influence on th degradation of SDBS by VUV/O3, which is increased in strongly alkaline environments.

Implications of Typical Anions on SDBS Degradation
The implications of varied typical anions on the SDBS degradation during VUV/O process was examined, as shown in Figure 8. For the VUV/O3 process, the presence o SO4 2− hardly affected SDBS degradation; there was almost no decrease in the degradatio rate when 1 mM and 2 mM SO4 2− were dosed. When 1 mM and 2 mM of Cl − and CO3 − wer

Implications of Typical Anions on SDBS Degradation
The implications of varied typical anions on the SDBS degradation during VUV/O 3 process was examined, as shown in Figure 8. For the VUV/O 3 process, the presence of SO 4 2− hardly affected SDBS degradation; there was almost no decrease in the degradation rate when 1 mM and 2 mM SO 4 2− were dosed. When 1 mM and 2 mM of Cl − and CO 3 − were present in the water, the degradation rate of SDBS decreased only slightly, with Cl − causing a decrease of 2.75% and 7.25%, respectively, and CO 3 − causing a decrease of 4.50% and 7.75%, respectively. Unlike these, there was a substantial decrease in the degradation rate of SDBS when 1 mM and 2 mM NO 3 − were added, reaching 23.09% and 32.72%, respectively. The coexisting anions could trap the HO• radical species and convert them to second-order anionic radicals bearing relatively low redox potentials, according to chemical reaction formulae outlined as Equations (5)-(10) [41,42]. The SO 4 2− could capture HO• and 185 nm photons to produce SO 4 •− (Equations (5) and (6) [41,43], so the combined reaction of multiple radicals does not influence the degradation rate of SDBS. A number of studies have found that sulfate radicals are similar in efficiency to hydroxyl radicals for the removal of certain micropollutants, and even higher for the removal of some pesticides [44,45]. As present Cl − , it can directly react with O 3 to generate Cl 2 or HOCl at low pH conditions [35], and they will undergo oxidative degradation reaction with SDBS. However, studies have reported that Cl• could react with H 2 O to form HO•, which might improve the degradation rate of SDBS to a certain degree [46]. These are the reasons for the diminished inhibitory degree of Cl − on removal efficiency of SDBS during the VUV/O 3 process. The reaction of HCO 3 − /CO 3 2− with HO• to produce CO 3 •− is the main radical in resolution at higher HCO 3 − concentrations [47]. The redox potential of CO 3 •− is significantly lower than that of HO•, which would decrease the effectiveness of SDBS removal. Since SDBS is easily degraded, CO 3 • could also participate in the reaction, and the addition of HCO 3 − might increase the solution pH, which enhanced the generation of HO•. The above reasons could alleviate the decrease of SDBS degradation rate by HCO 3 − . It was obtained that NO 3 − had the greatest inhibition on SDBS degradation in the VUV/O 3 process. NO 3 − could absorb more 185 nm photons than any other because of its higher molar absorbance coefficient at 185 nm wavelength [41,48]. This led to a decrease in HO• yield and the production of lower oxidizing anionic radicals (NO 3 •), which inhibited the degradation of SDBS. Thus, these anions could also compete with H 2 O and O 3 molecules to absorb 185 nm photons.
In conclusion, the presented results showed that HO• can degrade SDBS more effectively than other radical species.  [41,43], so the combined reaction of multiple radicals does not influence the degradation rate of SDBS. A number of studies have found that sulfate radicals are similar in efficiency to hydroxyl radicals for the removal of certain micropollutants, and even higher for the removal of some pesticides [44,45]. As present Cl − , it can directly react with O3 to generate Cl2 or HOCl at low pH conditions [35], and they will undergo oxidative degradation reaction with SDBS. However, studies have reported that Cl• could react with H2O to form HO•, which might improve the degradation rate of SDBS to a certain degree [46]. These are the reasons for the diminished inhibitory degree of Cl − on removal efficiency of SDBS during the VUV/O3 process. The reaction of HCO3 − /CO3 2− with HO• to produce CO3 •− is the main radical in resolution at higher HCO3 − concentrations [47]. The redox potential of CO3 •− is significantly lower than that of HO•, which would decrease the effectiveness of SDBS removal. Since SDBS is easily degraded, CO3 • could also participate in the reaction, and the addition of HCO3 − might increase the solution pH, which enhanced the generation of HO•. The above reasons could alleviate the decrease of SDBS degradation rate by HCO3 − . It was obtained that NO3 − had the greatest inhibition on SDBS degradation in the VUV/O3 process. NO3 − could absorb more 185 nm photons than any other because of its higher molar absorbance coefficient at 185 nm wavelength [41,48]. This led to a decrease in HO• yield and the production of lower oxidizing anionic radicals (NO3•), which inhibited the degradation of SDBS. Thus, these anions could also compete with H2O and O3 molecules to absorb 185 nm photons. In conclusion, the presented results showed that HO• can degrade SDBS more effectively than other radical species.

Proposed Degradation Pathways of SDBS
In order to derive a hypothetical pathway for the SDBS degradation during the VUV/O3 process, the formation of by-products was analyzed by high performance liquid chromatography/mass spectrometry (HPLC/MS) in ESI (−) and ESI (+) modes, and HOMO, LUMO, and Fukui indices of SDBS were calculated using DFT (Figures S2-S4 and Tables S1-S3). The obtained total ion chromatogram (TIC) and mass-to-charge ratio are presented in Figure S5. Ten by-products were deduced, and their structural formulae are

Proposed Degradation Pathways of SDBS
In order to derive a hypothetical pathway for the SDBS degradation during the VUV/O 3 process, the formation of by-products was analyzed by high performance liquid chromatography/mass spectrometry (HPLC/MS) in ESI (−) and ESI (+) modes, and HOMO, LUMO, and Fukui indices of SDBS were calculated using DFT ( Figures S2-S4 and Tables S1-S3). The obtained total ion chromatogram (TIC) and mass-to-charge ratio are presented in Figure S5. Ten by-products were deduced, and their structural formulae are listed in Table S4. According to the three SDBS isomers and intermediates detected, three possible degradation paths were given in Figure 9. Some studies have extensively inferred that HOMO and LUMO could theoretically determine the sites where organic pollutant molecules tend to lose or gain electrons during redox reactions [49,50]. Considering that the mass charge between the products is relatively large, and that all of them differ by 14 m/z, they must be the products after the attack of branched chain carbon. The above speculation could be confirmed based on the detected MS spectra at RT = 5.376-5.538 min, 3.641 min, and 3.255 min corresponding to molecular ion peaks of m/z = 325.2 (SDBS), 311.1, and 297.2, respectively. Of these, m/z = 325.2 corresponded to the anionic form of SDBS after hydrolysis in solution to release Na + , while m/z = 311.1 and 297.2 corresponded to substances where the alkane chain was decreased by one methyl and two methyl groups, respectively. According to the product types, there were three main branched forms of SDBS, including methyl, ethyl, and propyl branched chains.  Theoretically, the site of HO•-attack could presumably appear at all C sites at the SDBS alkyl chain based on the DFT calculations, though it was more likely to occur at alpha-and beta-carbons. Nonetheless, according to the HPLC/MS, the oxidized byproducts of SDBS branched carbon was detected even more frequently, and the molecular According to the DFT calculation that f − and f + of Fukui index, the most easily oxidized sites of reactive oxygen radicals are the 1(S) bond and 16(C), 17(C), 18(C), 19(C), 20(C), and 21(C). Therefore, the sulfonyl group and branched chain carbon were not included in the final measured by-products. However, almost all of the products obtained during the reaction contain sulfonyl groups and branched carbon chains. It is possible that many intermediates were oxidized and degraded in a short time without their presence being detected. The following conclusions were obtained by deriving the degradation pathways based on the detected substances. In the presence of reactive oxygen species, SDBS was attacked in two main ways: (a) attacking the alpha-carbon, beta-carbon, or branched chaincontaining carbon on the alkane chain; (b) attacking the adjacent carbon of the alkane chain and eventually forming phenolic hydroxyl groups. During the course of the reaction, signal peaks were also detected at RT = 0.9 min and 2.4 min with molecular ion peaks corresponding to m/z = 329 and 343 in the MS spectra. It is the further conversion product of the product after the attack of the above three branched structures, and since 329 m/z differs from 297 m/z and 343 m/z from 311 m/z by 32 m/z, this was the result of the obvious oxygen intervention, which is exactly in line with the (b) mode of attack of SDBS [51,52]. Among the by-products, m/z = 315.2 was the product of further attack on the carbon of the benzene ring of the attacked by-product of the SDBS branched chain; m/z = 185.0 was the product of further attack on the α carbon of the alkane chain. The degradation pathways of three SDBS isomers in VUV/O 3 were derived from the above analysis as shown in Figure 9.
Theoretically, the site of HO•-attack could presumably appear at all C sites at the SDBS alkyl chain based on the DFT calculations, though it was more likely to occur at alpha-and beta-carbons. Nonetheless, according to the HPLC/MS, the oxidized by-products of SDBS branched carbon was detected even more frequently, and the molecular weights of the alpha-and beta-carbon cleavage products were considerably smaller than those of the basal peak cleavage by-products. At first it was the alkyl chain of SDBS that the free electrons formed by HO•-action excited the carbon atoms chemical reactivity. They are induced to react with dissolved oxygen (O 2 ) and O 3 . HO• cleaved the alkyl chains at the β-position to form carbonyl groups to produced by-products (A), (B), (D), and (G). Ultimately, HO• as well as dissolved O 2 and O 3 transform the alkyl chains of SDBS to carboxyl groups, producing by-products (C), (E), and (H). According to the above summary, the degradation modes of the three SDBS isomers were very comparable.

Toxicological Analysis of Intermediates
Toxicological evaluations of SDBS with its degradation intermediates were performed with a toxicity assessment software tool (TEST) [53]. The structural formulae of intermediates A-J are shown in Table S4. As shown in Figure 10a, the LD50 of SDBS was estimated as 3057.19 mg/kg, which was classified to be "harmful" [6]. All by-products remained "harmful" in the process of SDBS degradation by VUV/O 3 , except for by-product J, the toxicity of other products increased. As can be seen from Figure 10b, SDBS and its intermediates/products were "harmful" in terms of bioaccumulation factors, and the bioaccumulation factors values of these compounds were reduced to different degrees compared to SDBS [54]. This indicated that the VUV/O 3 treatment had a positive effect on the protection of the higher consumers in the organism. Figure 10c shows the developmental toxicity, which was maintained for all by-products. The developmental toxicity of all the by-products, except by-product C, were decreased. As illustrated in Figure 10d, there was an increase in the mutagenicity of 6 SDBS degradation by-products relative to the parent contaminant, but only by-product J reached a positive mutagenicity level [55]. The analysis of TEST indicated that the toxicity of most SDBS degradation by-products decreased or remained at the original level. Therefore, VUV/O 3 treatment of SDBS has a broad application prospect on reduction of surfactants toxicity. relative to the parent contaminant, but only by-product J reached a positive mutagenicity level [55]. The analysis of TEST indicated that the toxicity of most SDBS degradation byproducts decreased or remained at the original level. Therefore, VUV/O3 treatment of SDBS has a broad application prospect on reduction of surfactants toxicity.

Application of VUV/O3 for Real Laundry Wastewater Treatment
To investigate the anionic surfactant degradation effect of VUV/O3 in a real water matrix, we applied it to treat laundry wastewater. VUV/O3 could effectively remove anionic surfactants from laundry wastewater ( Figure 11). The removal of surfactant was followed by the reduction of CODCr and TOC. The CODCr of laundry wastewater Figure 10. Theoretically calculated analysis of various toxicity of SDBS and its by-products: (a) Oral rat LD50, (b) bioaccumulation factor, (c) developmental toxicity, (d) mutagenicity.

Application of VUV/O 3 for Real Laundry Wastewater Treatment
To investigate the anionic surfactant degradation effect of VUV/O 3 in a real water matrix, we applied it to treat laundry wastewater. VUV/O 3 could effectively remove anionic surfactants from laundry wastewater ( Figure 11). The removal of surfactant was followed by the reduction of COD Cr and TOC. The COD Cr of laundry wastewater decreased from 239.9 to 140.6 mg/L within 60 min of VUV/O 3 treatment. Meanwhile, the DOC was reduced from 67.22 to 34.91 mg/L under VUV/O 3 treatment. The removal efficiency of anionic surfactants in laundry wastewater was identified to be less than that depicted above in pure water. This was probably caused by capturing of radicals and (V)UV rays by inorganic anions, protein-like and greasy contaminants existing from the treated greywater [18]. In addition, the high turbidity of the effluent may hinder the effective propagation distance of VUV and UV photons. It can be found that VUV/O 3 was more suitable for treating low turbidity greywater after pretreatment. efficiency of anionic surfactants in laundry wastewater was identified to be less than tha depicted above in pure water. This was probably caused by capturing of radicals and (V)UV rays by inorganic anions, protein-like and greasy contaminants existing from the treated greywater [18]. In addition, the high turbidity of the effluent may hinder the effective propagation distance of VUV and UV photons. It can be found that VUV/O3 was more suitable for treating low turbidity greywater after pretreatment.

Experimental Procedures
A glass reactor with a capacity of 700 mL was used for the VUV oxidation experiments. The setup used for the tests was placed in 6 mm thick polyvinyl chloride sealed chamber, which was connected to the thermostatic water cycle (kept stable in 298 K) and also defended against UV leakage. A 6 W low-pressure mercury VUV lamp was fixed in a quartz tube to prevent water from coming into direct contact with the lamp. The low-pressure mercury lamp used (GPH150T5VH/4, Heraeus, Hanau, Germany) was mainly available in two wavelengths, 254 nm (~90%) and 185 nm (~10%), and its ligh intensity was tested by the degradation kinetics of photosensitizers, exhibited in Figure  S1 according to existing methods [56,57]. To ensure stable output of optical power, the VUV lamps should be preheated for at least 20 min. In standard experimental procedure a fixed amount of O3 (about 0.927 mg/L O3 dissolved in solution) was introduced into a 10 mg/L SDBS solution and simultaneously exposed to VUV for 30 min, while 10 mL samples were taken every 5 min. These samples were injected into 10 mL sampling tubes after direct filtration with a 0.45 µm microfiltration membrane.

Experimental Procedures
A glass reactor with a capacity of 700 mL was used for the VUV oxidation experiments. The setup used for the tests was placed in 6 mm thick polyvinyl chloride sealed chamber, which was connected to the thermostatic water cycle (kept stable in 298 K) and also defended against UV leakage. A 6 W low-pressure mercury VUV lamp was fixed in a quartz tube to prevent water from coming into direct contact with the lamp. The lowpressure mercury lamp used (GPH150T5VH/4, Heraeus, Hanau, Germany) was mainly available in two wavelengths, 254 nm (~90%) and 185 nm (~10%), and its light intensity was tested by the degradation kinetics of photosensitizers, exhibited in Figure S1 according to existing methods [56,57]. To ensure stable output of optical power, the VUV lamps should be preheated for at least 20 min. In standard experimental procedure, a fixed amount of O 3 (about 0.927 mg/L O 3 dissolved in solution) was introduced into a 10 mg/L SDBS solution and simultaneously exposed to VUV for 30 min, while 10 mL samples were taken every 5 min. These samples were injected into 10 mL sampling tubes after direct filtration with a 0.45 µm microfiltration membrane.
To determine the performance of the radicals and VUV photons for the degradation of SDBS, the scavengers tert-butanol (TBA) and formic acid (FA) were added. TBA is a well-known HO• scavenger [58]; FA can absorb VUV photons in preference to H 2 O molecules (Φ FA,185 > Φ H2O,185 ) [59].
Several sets of univariable experiments were conducted to determine the influences of varied water constituents. The initial SDBS concentrations ([SDBS] 0 ) varied from 5 to 20 mg/L. The initial solution pH was regulated from 3 to 5, 7, 9, and 11 with 0.1 M H 2 SO 4 or 0.1 M NaOH. The influence of the presence of inorganic anions on the degradation of SDBS was measured by adding 1-2 mM Na 2 SO 4 , NaCl, NaHCO 3 , and NaNO 3 . All of the test results were the average of the three replicates.

Analytical Methods
The SDBS concentration of the taken samples was measured by liquid chromatography (LC, Agilent 1260 LC, Santa Clara, CA, USA) with a C18 column. Dissolved organic carbon (DOC) was analyzed by a German Elementar TOC analyzer. The concentrations of O 3 were determined by indigo spectrophotometry [60]. Inorganic anions were determined with a chromatograph (Metrohm, 883 Basic IC plus, Herisau, Switzerland) on a Metrosep a Supp (250.0 mm × 4.0 mm) column.
The degradation kinetics of SDBS were consistent with a pseudo-first-order kinetic formulation (Equation (11)), following our previous research [37]. C 0 and C t are initial and residual SDBS concentrations (mg/L); k is the degradation rate of the presented formulation (min −1 ).
−ln (C t /C 0 ) = kt The spectra of radicals were recorded by electron paramagnetic resonance (EPR, JEOL JES-FA200, Akishima City, Tokyo, Japan), while all spectra were detected three times separately. O 3 and 5,5-Dimethyl-1-pyrroline N-oxide (DMPO) were dissolved into ultrapure water and the mixture drawn to the capillary for HO• detection in EPR.
Density functional theory (DFT) calculations were all calculated using B3LYP/6-31+G* predictive models on Gaussian 09 software, which helped to further elucidate the radical attack on SDBS molecule in the VUV/O 3 reaction system [61]. The molecular structural formula of generated SDBS oxidation by-products were deduced by an Agilent 1290 HPLC-MS (Santa Clara, CA, USA). Toxicological analysis of SDBS degraded intermediates were evaluated by using the Toxicity Estimation Software Tool (TEST). Use of the TEST (version 4.2.1) software is to estimate toxicity values for chemicals from their molecular structure using a variety of quantitative structure-activity relationship (QSAR) methodologies. (2) VUV/O 3 promoted the production of HO• compared to VUV alone and O 3 alone, which was the major reactive species attacking SDBS molecules. (3) The performance of VUV/O 3 was optimal at pH 9, with lower oxidation efficiency at more acidic levels. The addition of SO 4 2− hardly affected the degradation of SDBS and that of Cl − and HCO 3 − slightly reduced the reaction rate, except that adding NO 3 − had a remarkable inhibitory effect on the process. (4) There were three isomers in SDBS, and the degradation modes of the three SDBS isomers as three pathways were very comparable. The degradation by-products of the VUV/O 3 process decrease in harmfulness and toxicity compared to the SDBS parent. (5) VUV/O 3 could effectively remove anionic surfactants from laundry wastewater. The removal efficiency of anionic surfactants in laundry wastewater was identified to be less than that depicted above in pure water.

Conclusions
The above results suggest that increasing the yield of hydroxyl radicals is the main goal that different processes should pursue when removing surfactants. When using the VUV/O 3 process, a moderate increase in reaction time is recommended for deep oxidation or complete minimization of residual intermediates. This ensures that the toxic intermediates in the treated effluent are mineralized or reduced to a lower concentration.
This study contributes to the development of optimal operating conditions for the VUV/O 3 technology and explores novel methods to remediation of surfactants.
Supplementary Materials: The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/molecules28083312/s1, Figure S1: Photolysis kinetics of 0.12 mM uridine (black, 254 nm) and 100 mM MeOH (purple, 185 nm) under UV or VUV irradiation; Figure S2 Table S1: The methyl isomers of SDBS molecule: natural population analysis (NPA) charge populations and condensed Fukui index distribution for electrophilic attack (f − and f + ); Table S2: The ethyl isomer of SDBS molecule: natural population analysis (NPA) charge populations and condensed Fukui index distribution for electrophilic attack (f − and f + ); Table S3: The propyl isomer of SDBS molecule: natural population analysis (NPA) charge populations and condensed Fukui index distribution for electrophilic attack (f − and f + ); Table S4: Chemical formulas and main fragments (m/z) of intermediate products; Table S5: Organic characteristics and anions concentration of the laundry greywater.

Conflicts of Interest:
The authors declare no conflict of interest.
Sample Availability: Not applicable.