Efficient Degradation of Bisphenol S by Ultraviolet/Persulfate Oxidation in Ultra-Pure and Saline Waters: Effects of Operating Conditions and Reaction Mechanism

Abstract

As an alternative to bisphenol A, bisphenol S (BPS) is considered an emerging concern. In this study, the degradation of BPS by persulfate (PS), ultraviolet (UV), and UV/PS was comprehensively examined in ultra-pure and saline waters. UV/PS effectively degraded BPS, and the observed first-order rate constant, k_obs, increased from 0.021 to 0.382 min⁻¹ with an increasing PS concentration from 100 to 1000 μΜ. The addition of humic acid (HA) inhibited the degradation of BPS, and 1/k_obs was directly proportional to the concentration of HA. In salty water containing 540 mM Cl⁻ or 0.8 mM Br⁻, UV/PS possessed a higher degradation ability for BPS: the corresponding k_obs values were 1.45 and 1.66 times that of the control sample, respectively. Eighteen degradation products, including β-scission, sulfate addition, quinone type, ring-opening, and cross-coupling, were identified using high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). Two possible pathways were proposed: (i) the initial step was considered to be an electron transfer reaction from BPS to SO₄^•−, leading to the formation of a phenyl radical cation R1, and then phenol radical R4, 4-hydroxybenzenesulfonate cation R5, phenoxyl radical R3, resonant-type carbon-centered radical R2, and their secondary products; (ii) another pathway was the sulfate addition and hydroxylation. These primary reaction sites were further verified by theoretical calculation. This study highlights the effectiveness of UV/PS as a promising strategy for the remediation of BPS and other endocrine-disrupting chemicals in ultra-pure and saline waters (540 mM NaCl or/and 0.8 mM NaBr).

1. Introduction

Bisphenol S (BPS) has gained commercial importance as an alternative to bisphenol A (BPA) for a variety of industrial applications [1]. It is used as a fastening agent in cleaning products, a developer in thermal paper, a monomer in synthetic polymers, including epoxy, an electroplating solvent, and a constituent of phenolic resin [2,3]. Since the application of BPA-containing polymers can lead to negative impacts on reproductive abilities and the induction of precocious puberty in babies, the use of BPA in baby bottles was first restricted in October 2008 by Canada, and banned in March 2011 by the European Union [4,5]. Due to restrictions and regulations on the use of BPA, BPS is increasingly developed in the manufacturing of polycarbonates and epoxy resins to meet the demand of the market as a replacement [6]. Based on the survey data of the European Chemical Agency (ECHA), the European Economic Area manufactured or imported approximately 1000–10,000 million metric tons of BPS [7]. Given the increasing use of BPS, its widespread occurrence has been found in environmental media, animals, and the human body of China, Japan, Korea, America, Romania, Colombia, India, Pakistan, Vietnam, and Saudi Arabia, including in foods, drinks, house dust, surface water, sediments, wastewater treatment plants, and paper, as well as in human urine, sera, and breast milk, at concentration levels within the ng/L–μg/L or ng/g–μg/g range [8,9,10].

Although BPS is an effective alternative to BPA, it can cause oxidative stress, obesity, acute toxicity and neurotoxicity, alter adult reproductive function, and has adverse effects on the endocrine system, with similar results in the same magnitude order of concentration as BPA [11,12]. In addition, previous studies have demonstrated that BPS exhibits greater resistance to environmental degradation, with a longer half-life, and is more permeable to the skin than BPA [13,14,15]. Thus, it is critical to remove this contaminant from water and wastewater. In recent years, advanced oxidation technologies have demonstrated advantages in the field of pollutant removal [16,17]. Sulfate and hydroxyl radicals are generated as a result of several processes including ozone, hydrogen peroxide, peroxymonosulfate (PMS), UV, and persulfate (PS)-based AOPs [18,19]. While conventional advanced oxidation processes (AOPs) such as UV/H₂O₂ and electrochemical oxidation have shown efficacy, UV/H₂O₂ often suffers from rapid H₂O₂ decomposition and limited radical lifetime, and electrochemical oxidation requires a high energy input and faces challenges with electrode stability [20]. In contrast, UV/PS systems generate longer-lived sulfate radicals (SO₄^•−) with a higher redox potential compared to hydroxyl radicals (^•OH) in UV/H₂O₂, while avoiding secondary pollution risks associated with electrochemical by-products [21,22]. Additionally, UV/PS typically presents an environmentally friendly process for contaminant removal, since no secondary pollution needs to be monitored [23]. Several studies demonstrate that UV/PS represents a viable technology for the removal of recalcitrant organic contaminants in water and wastewater, such as endocrine-disrupting chemicals, chlorinated compounds, pharmaceuticals, perfluorinated compounds, and algal toxins [24,25].

Up to now, there have been some studies on the removal and transformation of BPS, mainly by biodegradation, bioelectrochemical degradation, thermal degradation, and chemical oxidation including chlorination, ozonation, UV/O₃(H₂O₂) oxidation, Fe@C carbonized resin (CuCo₂S₄)/peroxymonosulfate, nano-graphite (magnetic spinel CuFe₂O₄/SBC)/peroxydisulfate, and photocatalyst degradation [3,26,27,28,29]. These studies have demonstrated that operating conditions (i.e., pH, temperature, oxidant dose), humic acid, cations, anions, and water matrices, can affect the removal of BPS significantly [3,30]. Nevertheless, there is limited information available regarding the degradation behavior and mechanism of BPS in the UV/PS system, despite its great significance for water and wastewater treatment.

The present study aimed to evaluate the degradation behaviors of BPS by UV/PS, with an emphasis on exploring the effects of PS concentration, pH, and humic acid (HA) on BPS removal efficiency. Combining theoretical calculations and LC-MS analysis, the transformation products and degradation pathways were illustrated. This study can provide fundamental information for the remediation of BPS in ultra-pure and saline waters.

2. Chemicals and Methods

2.1. Chemicals

BPS (99%) and PS (99%) were provided by Sigma-Aldrich (Oakville, ON, Canada). HPLC-grade methanol was purchased from Merck (Darmstadt, Germany). Other chemicals, including NaCl, NaBr, NaClO₄, phosphate, HA, and formic acid, were obtained from Nanjing Chemical Reagent Co., Ltd. (Nanjing, China). Ultra-pure water (≥18.2 MΩ cm) was generated with a Stakpure OmniaTap water purification system (Peculiar Instrument1 Technology, London, UK). Salty waters were prepared in our lab (Nanjing, China), including NaCl (540 mmol/L), NaClO₄ (540 mmol/L), NaBr (0.8 mmol/L), and NaCl (540 mmol/L) samples.

2.2. Experimental Setup

Photodegradation experiments were conducted using a BL-GHX-V photochemical reactor (Bilang, Beijing, China). The design of the rotating bottom unit in the reactor ensured uniform illumination. A 15 W UV-LED lamp (Perfectlight, Beijing, China) was used as the light source, emitting a radiation wavelength of 254 nm. The absolute light density was measured to be 6.67 × 10⁻⁵ mol m⁻³ s⁻¹. Additionally, 50 mL of the reaction solution containing 5 µM BPS and PS with a certain concentration range of 100–1000 µmol/L was transferred into the quartz tubes for irradiation. 10 mM phosphate buffer was used in this study. The reaction solution was mixed using a magnetic stirrer with a rotational speed of 100 r/s. At 0, 5, 10, 15, 20, 25, 30, 40, 50, and 60 min, 1 mL of sample was transferred into 5 mL centrifuge tubes pre-added with 50 µL of 0.1 mol/L Na₂S₂O₃. All experiments were conducted in triplicate.

2.3. Analytic Methods

The concentrations of BPS were determined by a Hitachi L-2000 high-performance liquid chromatography system (LC, Hitachi Technologies, Tokyo, Japan). A photo diode array detector (DAD) was used. The mobile phase consisted of a mixture of 50% methanol containing 0.1% formic acid and 50% water containing 0.1% formic acid. The target compound was eluted through a ZORBAX Eclipse Plus C18 column (Agilent (Beijing, China), 250 mm × 4.6 mm, 5 μm) at the flow rate of 1.0 mL/min. The specific wavelength was set at 257 nm.

For solid extraction (SPE), CNWBOND LC-C₁₈ SPE cartridges (2.CA0955.0001, 1 g, 6 mL) (CNW, ANPEL Laboratory Technologies (Shanghai), Int., Shanghai, China) were used to concentrate sample solutions for the authentication of the reaction products. LC-MS analysis was performed by using the AB Sciex HPLC system, equipped with a high-resolution hybrid quadrupole-time-of-flight mass spectrometer (X500R QTOF, AB Sciex, Framingham, MA, USA). An electrospray ionizer (ESI) in negative mode was acted as an ion source. The analytes were separated on an Agilent ZORBAX Eclipse Plus C18 column (100 mm × 2.1 mm, 2.6 μm). The binary solvent system used was composed of 0.3% formic acid in water (A) and methanol (B) at a flow rate of 0.20 mL/min. The gradient program was 10% B for 4 min, followed by 100% B in 11.5 min and kept for 9.5 min, then returned to 10% B in 0.5 min and equilibrated for 14.5 min. The injection volume was 10 μL. Mass spectra were set as m/z 80–1100 in a full-scan negative electrospray ion (ESI−) ionization mode. Ion source parameters were set as follows: ionspray voltage floating, −4500 V; declustering potential, −80 V; collision energy, −10 V; temperature, 550 °C; ion source gas 1, −55 psi; ion source gas 2, −55 psi; curtain gas, −35 psi. Nitrogen gas was used throughout the overall process. The SCIEX OS program (AB Sciex Version 3.1) was used as the data processing software. For the purposes of identifying unknown products, the mass accuracy in this study was set at <5 ppm.

2.4. Theoretical Gaussian Setup

The structure optimization of BPS was conducted by using the Gaussian 09 program at the M06-2X/6-311G** level based on the least quantifiable point. The influence of the bulk solvent, represented by water, was taken into account using the solvent model of the density continuum solvation model, with keywords of “scrf = (solvent = water, smd)”. The 2FED²_HOMO of each atom and the spin densities of the BPS radical were calculated at the same level to seek the possible reaction sites.

3. Results and Discussion

3.1. Oxidation of BPS by UV/PS

PS alone did not result in the detectable removal of BPS, but it degraded rapidly in the UV/PS oxidation process. As illustrated in Figure 1, approximately 45% removal was achieved with the PS dose of 100 µM. The degradation can be modeled by pseudo-first-order kinetics with the rate constant of 0.026 min⁻¹. Technically, the removal of BPS in the UV/PS oxidation process could be ascribed to both UV photolysis and free radicals (SO₄^•− and ^•OH) oxidation. Having a molar extinction coefficient of 14,339 M⁻¹cm⁻¹ (254 nm), BPS absorbs UV light strongly (Figure S1), which is a prerequisite for photolysis. To distinguish the contribution of each mechanism, 1 mM methanol was chosen as the radical scavenger. The degradation of BPS due to reaction with SO₄^•− and ^•OH was expected to be completely suppressed under such conditions. Figure 1 shows that the k value of BPS decreased to 0.013 min⁻¹ and the data were very close to those of direct photolysis (UV alone), indicating that free radicals and UV photolysis each contributed around equally, each accounting for 50% of the BPS removal.

$${\mathrm{S}}_2{{\mathrm{O}}_8}^{2-}\stackrel{\mathrm{U}\mathrm{V}}{\to }2\mathrm{S}{{\mathrm{O}}_4}^{•-}$$

(1)

SO₄^•− + H₂O →HSO₄²⁻ + ^•OH

(2)

3.2. Effects of PS Dose and pH on the Degradation of BPS by UV/PS

The PS dose plays a crucial role in the degradation of pollutants by UV/PS. The degradation of BPS was investigated with PS doses within the range of 0–1000 μM, the typical dosages reported for both laboratory studies and practical applications [31,32]. The results are shown in Figure 2. The degradation of BPS increased with increasing PS dose. When 100 µM PS was added to the reaction solution, 73% of BPS was degraded in 60 min. In addition, 200, 500, and 1000 μM PS resulted in near-complete BPS removal within 60, 30, and 9 min, respectively. The observed k_obs value increased from 0.021 to 0.393 min⁻¹ with increasing PS concentration from 100 to 1000 μM. Overall, the k_obs of BPS increased linearly with the PS dose (R² = 0.988), indicating the production of free radicals increased linearly with the increase in the PS dose. This result was consistent with the decreasing removal of BPS in the existence of 1 mM MeOH, due to its quenching effects on reactive radicals (the k_obs of BPS by UV/PS in the presence of MeOH decreased by half from 0.026 to 0.013 min⁻¹ (Figure 1)). Alternatively, different observations were found in several studies, in which increasing the initial S₂O₈²⁻ dose could not continuously improve the substrate removal efficiency. The photooxidative reaction rate constants increased from 0.0042 to 0.0642 min⁻¹ as S₂O₈²⁻ concentrations rose from 0.1 to 5 mM, with a slight slowdown in C.I. Basic Yellow 2 decay rate increment beyond 5 mM. There was an optimal dose of PS because excess S₂O₈²⁻ could react with generated SO₄^•− following Equation (3). However, this radical quenching phenomenon was not observed in this study. This is similar to observations by Gao et al. [33], who investigated the UV-activated PS oxidation of sulfamethazine in water and observed that an excess of PS was able to improve the reaction by nearly 23 times with the increasing PS dose from 0 to 0.5 mM.

SO₄^•− + S₂O₈²⁻ → SO₄²⁻ + S₂O₈^•−

(3)

The influence of pH on BPS removal by UV/PS was investigated by controlling the initial PS dose 500 μM in the pH range of 4.0–11.0. The results from Figure S3 show that the degradation was slightly higher in acidic and neutral conditions than in the alkaline system. Specifically, 100%, 99%, and 85% degradation within 20 min was observed at pH 4.0, 7.0, and 11.0, respectively. When the reaction time was further extended to 30 min, almost complete degradation (95%) can also be observed at pH 11.0, indicating that BPS has a good acid-base adaptability. This result is similar to that of Shao et al. [29], where the removal efficiency showed negligible difference between acidic and alkaline solutions for degradation of ofloxacin and phenol.

3.3. Effects of Natural Organic Matter on the Degradation of BPS by UV/PS

Natural organic matter (NOM) is commonly found in natural water bodies, and it is mostly composed of dissolved humic substances. Here, humic acid (HA) was used to examine the effects of NOM on the degradation of BPS in the UV/PS process. As depicted in Figure 3A,C, HA suppressed the direct photolysis of BPS, and such suppression increased with an increasing HA concentration. In the presence of 20 mg L⁻¹ HA, the removal of BPS by direct photolysis decreased from 61% to 25% in 30 min. The degradation of BPS in the presence of HA was also consistent with the pseudo-first-order kinetics (R² ≥ 0.99). The corresponding k_obs value decreased from 0.016 to 0.005 min⁻¹. These changes in the degradation efficiencies over time with the addition of HA may result from a combination of factors, including the light-shielding effect and competition for reactive species with BPS [1,34]. The effect of the HA (0–20 mg L⁻¹) on BPS removal by UV/PS is shown in Figure 3B,D. At any given HA concentration, the BPS degradation exhibited pseudo-first-order kinetics. A similar inhibiting effect was also observed in the UV/PS process. The HA concentration increased from 0 to 20 mg L⁻¹, leading to the decrease in BPS removal percentages from 99% to 25% in 30 min, respectively. The observed k_obs decreased from 0.18 to 0.01 min⁻¹. This may be due to the following reasons. On the one hand, the light shielding effect of HA could hinder the activation of PS by UV, resulting in the reduced generation of reactive radicals. On the other hand, because HA contains phenolic groups and carboxyl groups, HA molecules were susceptible to attack by SO₄^•− through electron transfer and decarboxylation mechanisms, which led to the decreasing degradation efficiency of BPS. In total, the addition of HA would inhibit the degradation of BPS, and 1/k_obs was directly proportional to the concentration of HA. Similar inhibiting effects of HA were also reported in SO₄^•−-based advanced oxidation processes for atrazine and sulfamethoxazole removal [35,36].

3.4. The Removal of BPS in Salty Water

Research on the removal of BPS has mostly focused on its degradation in water and wastewater matrices. Few studies have examined the transformation and fate of BPS in saline waters. The removal of BPS by UV/PS in the presence of different halogen ions (A–E groups) at seawater levels was investigated (Figure 4). As seen, the degradation in 15 min reaction was 84.25% in the absence of halides (E group), and a slight inhibition was observed in the presence of 540 mM NaClO₄ (B group, 81.80%). However, under the A (540 mM NaCl), C (0.8 mM NaBr), and D (540 mM NaCl + 0.8 mM NaBr) groups, the degradation efficiencies were increased to 98.88, 97.99, and 100%, respectively. In addition, the measured degradation data showed that the oxidation reaction of BPS in these five groups (A to E) also followed pseudo-first-order kinetics, and their corresponding rate constants (k_obs) were 0.168, 0.105, 0.193, 0.225, and 0.116 min⁻¹, respectively. In salty water containing 540 mM Cl⁻ or 0.8 mM Br⁻, UV/PS possessed higher degradation ability of BPS, the corresponding k_obs values are 1.45- and 1.66-times that of the control sample, respectively. This result is attributed to the potential of SO₄^•− to react with Cl⁻ (E⁰_Cl^•_/Cl⁻ = 2.41 V) and Br⁻ (E⁰_Br^•_/Br⁻ = 1.62 V), generating active chlorine/bromine species (Cl^•, Cl₂^•−, HOCl, Br^•, Br₂^•−, HOBr) through Equations (4)–(13) [37,38,39,40,41]. The presence of 540 mM Cl⁻ and 0.8 mM Br⁻ together increased the value of k to 1.94 times that in the control sample. However, the degradation of BPS in the D group was slower than the sum of BPS removal in the A and C groups (Figure 4). This is reasonable because Cl^• could react with Br⁻ and Br^• to generate BrCl^•− (11) and BrCl (12), respectively [42,43]. Overall, these results highlight the feasibility of applying UV/PS for the practical treatment of BPS in saline water with sea-level halides.

SO₄^•− + Cl⁻ → Cl^• + SO₄²⁻ k = 3.0 × 10⁸ M⁻¹s⁻¹

(4)

Cl^• + Cl⁻ → Cl₂^•− k = (7.8 ± 0.8) × 10⁹ M⁻¹s⁻¹

(5)

Cl^• + Cl^• → Cl₂ k = 1 × 10⁸ M⁻¹s⁻¹

(6)

Cl₂^•− + Cl₂^•− → Cl₂ + 2Cl⁻ k = (9 ± 1) × 10⁸ M⁻¹s⁻¹

(7)

Cl₂ + H₂O → HOCl + H⁺ + Cl⁻

(8)

SO₄^•− + Br^– → Br^• + SO₄²⁻ k = 3.5 × 10⁹ M⁻¹s⁻¹

(9)

Br^• + Br⁻ → Br₂^•− k = 1.2 × 10¹⁰ M⁻¹s⁻¹

(10)

Br^• + Br^• → Br₂ k = 1.0 × 10⁹ M⁻¹s⁻¹

(11)

Br₂^•− + Br₂^•− → Br₂ + 2Br^– k = 1.9 × 10⁹ M⁻¹s⁻¹

(12)

Br₂ + H₂O → HOBr + H⁺ + Br⁻

(13)

Cl^• + Br⁻ → BrCl^•− k = 1.2 × 10¹⁰ M⁻¹s⁻¹

(14)

Cl^• + Br^• → BrCl

(15)

3.5. Product Identification and Reaction Mechanisms

Some researchers have proposed different degradation pathways for BPS based on various water treatment processes. Gao et al. [3] found that the chlorine substitution and electron transfer occurred, resulting in the formation of mono-/di-/tri-/tetrachloro-BPS, dimers, and benzenesulfonic acid products when BPS was subjected to chlorination treatment. Hydroxylation was identified as the main pathway during the photocatalyst process, which is supported by the detection of hydroxy-benzenesulfonic acid and dihydroxy-benzenesulfonic acid [29]. During the heat-activated persulfate oxidation process, BPS was subjected to electron transfer to the sulfate radicals, leading to the generation of BPS radicals. These BPS radicals could subsequently react with each other, forming some dimers via C–C and C–O coupling ways. The ring-opening reaction can also occur during this process to produce a series of low-molecular-weight products containing more than two sulfur atoms [44].

To investigate the degradation mechanism of BPS in the UV/PS process, LC- MS was applied to identify transformation products. We combined the identified degradation products with those previously reported, and proposed two possible pathways for BPS oxidation in the UV/PS process (Figure 5). As seen, the oxidation of BPS involved β-scission, hydroxylation, sulfate addition, ring-opening, and cross-coupling reactions. The initial step was considered to be an electron transfer reaction from BPS to SO₄^•−, resulting in the formation of a phenyl radical cation, denoted as R1. R1 was unstable and would undergo two different processes to form various intermediates. For example, R1 could be converted to phenol radical R4 and 4-hydroxybenzenesulfonate cation R5 via β-scission. SO₄^•−/HO^• could attack R4, giving rise to unstable hydroquinone P1. A sequential oxidation process occurred to produce the corresponding quinone type product P2 and ring-opening product (maleic acid, P3). R4 could undergo a sulfate addition reaction and hydroxylation to produce P4 and P5 sequentially. Sulfate addition reactions were also observed in the degradation of nonylphenol, triclosan, ranitidine, and nonsteroidal anti-inflammatory drugs based on SR-AOPs [45,46,47]. This result was consistent with kinetic data, highlighting the significant role of SO₄^•− in BPS removal in the UV/PS process. The further oxidation of P5 could also be converted to quinone-type product P6. Another intermediate radical (R5) could undergo hydrolysis, leading to the generation of 4-hydroxybenzenesulfonic acid (P7). Afterward, SO₄^•−/HO^• could attack P7 to yield 3,4-dihydroxybenzenesulfonic acid, which would undergo a further oxidation process to form an O-quinone product (3,4-dioxocyclohexa-1,5-diene-1-sulfonic acid, P8) and 4-hydroxycyclohexa-3,5-diene-1,2-dione (P9), accompanied by a desulfonation reaction. Furthermore, a cross-coupling reaction between R5 and P1 could result in the formation of P15. It is worth noting that P12 and P15 had the same experimental m/z value in the mass spectrum (Figure S2), suggesting that two chemical isomers were formed during the UV/PS oxidation process. In addition, R1 could be converted to phenoxyl radical R3 via deprotonation, which could further generate its resonance-type carbon-centered radical R2. Subsequently, these radicals could undergo cross-coupling reactions to form different products. Specifically, the cross-coupling between R2 and P1 could generate P16; R3 could couple with R4 to give P17. The addition of the SO₄H group on P16 and P17 could also occur to generate P14 and P18, respectively. Pathway II corresponded to the substitution of the H atom by an SO₄H and OH group to give P11 and P12, respectively, and then, the hydroxylation of P12 could give rise to P13. The toxicities of all the products detected in this study were predicted by the ECOSAR program. The predicted acute toxicity (LC₅₀ or EC₅₀) and ChV of the degradation products were 0.009–1.68 × 10⁷ and 0.011–1.01 × 10⁶ mg/L, respectively, indicating that most of the intermediates were not toxic. It is worth noting that P1, P2, P6, P10, P12, P13, P14, P15, P16, and P17 exhibit a certain level of toxicity and require careful handling (Text S1, Tables S1 and S2). Specifically, quinone-type products (P2, P6, and P10) were evaluated as “very toxic”, their formation may introduce new hazards, particularly if these intermediates persist in the environment. Thus, combining other oxidation processes may help further degrade these toxic products.

3.6. Theoretical Calculation

In recent years, many studies have highlighted that theoretical calculations are valuable tools for exploring the reaction pathways and behaviors of environmental organic pollutants [48]. The spin densities of the BPS radical were obtained to predict the probability of the presence of the unpaired electron, as shown in Figure 6. The spin densities of 15O (0.4153), 8C (0.3803), 10C (0.3196), and 12C (0.3156) are relatively high in the BPS radical, indicating the spin-unpaired electron tends to be distributed in these locations. This distribution suggests that these sites are highly susceptible to SO₄^•−, leading to the formation of key intermediates. Specifically, a phenol radical (R4) was formed through the abstraction of a hydrogen atom from the hydroxyl group; 4-hydroxybenzenesulfonate cation (R5) was generated via the oxidation of the sulfonate group; and phenoxyl radical (R3) was resulted from the loss of a hydrogen atom from the aromatic ring. A carbon-centered radical (R2) was formed due to the delocalization of the unpaired electron across the aromatic ring. This prediction was confirmed through the detection of P1, P4, P7, P16, and P17. The electrophilic reaction can easily occur at positions with higher 2FED²_HOMO values [49,50,51]. As shown in Figure 6B, the 14O (0.1742), 15O (0.1742), 5C (0.1660), and 8C (0.1661) atoms of the BPS molecule were found to exhibit a higher 2FED²_HOMO value. This suggests that these sites are most likely for SO₄^•− attack and the extraction of an electron. These results were consistent with the prediction of spin densities and the formation of P16, P17, P1, and P7.

4. Conclusions

This study demonstrated the effectiveness of UV/PS for degrading BPS in ultra-pure and saline waters (540 mmol/L NaCl or/and 0.8 mmol/L NaBr). Quenching experiments suggest that BPS removal in the UV/PS process was largely due to direct photolysis and reactive radicals (SO₄^•− and/or ^•OH). The reaction rate constants of BPS degradation under UV, UV/PS, and UV/PS (with 1 mM MeOH) were 0.014, 0.026, and 0.013 min⁻¹, respectively. The presence of HA inhibited the degradation of BPS, and the corresponding k_obs value decreased from 0.016 to 0.005 min⁻¹ in the UV oxidation process and from 0.180 to 0.010 min⁻¹ in the UV/PS oxidation process, with increasing HA concentration from 0 to 20 mg L⁻¹. In addition, SO₄^•− can react with Cl⁻ and Br⁻ to generate active chlorine/bromine species, improving the degradation efficiency of BPS. Eighteen intermediates of BPS were identified based on accurate mass measurement by TOF-MS analysis. The oxidation of BPS by UV/PS involved β-scission, hydroxylation, sulfate addition, phenol ring opening, and cross-coupling reactions. To further confirm the primary reactive sites, calculations were performed for 2FED²_HOMO values of each atom and spin densities of BPS radical. This study provided insights into the effective degradation and sulfate addition and cross-coupling reactions of BPS by UV/PS treatment.