Insights into the Kinetics, Theoretical Model and Mechanism of Free Radical Synergistic Degradation of Micropollutants in UV/Peroxydisulfate Process

Abstract

The degradation of acyclovir (ACY) and atenolol (ATL) in the UV/peroxydisulfate (UV/PDS) process has been systematically considered, focusing on the degradation kinetics, theoretical models, and reaction pathways via applying a microfluidic UV reaction system. The removal efficiencies of ACY and ATL were >94.8%, and the apparent degradation rate constants (k_obs) were 0.0931 and 0.1938 min⁻¹ at pH 6.0 in the UV/PDS system. The sulfate radical (SO₄^•−) and hydroxyl radical (^•OH) were identified as the major reactive radicals. The pH-dependent reaction rate constants of ACY and ATL with ^•OH and SO₄^•− were measured via the competing kinetics. Meanwhile, the contributions of ^•OH and SO₄^•− for ACY and ATL degradation were calculated by the radical steady-state hypothesis, and the results revealed that SO₄^•− occupied a decisive position (>84.5%) for the elimination of ACY and ATL. The contribution of ^•OH became more significant with the increasing pH, while SO₄^•− was still dominant. Moreover, ACY and ATL degradation performance were systematically evaluated via the experiments and Kintecus model under different operational parameters (Cl⁻, Br⁻, HCO₃⁻, NOM, etc.) in the UV/PDS process. Furthermore, the plausible reaction pathways of ACY and ATL were elucidated based on the Fukui function theory and ultra-performance liquid chromatography-tandem quadrupole time-of-flight mass spectrometry (UPLC-QTOF-MS) analysis. The UV/PDS process has been demonstrated to be an efficient and potential application for micropollutants mitigation.

1. Introduction

Acyclovir (ACY) and atenolol (ATL), as the typical and most widely used antiviral and β-blocker agents, are frequently adopted against cardiovascular diseases and two common virus infections (herpes simplex and varicella-zoster) [1,2]. ACY and ATL have been frequently detected in different aquatic environments (from ng L⁻¹ to μg L⁻¹) owing to their scarce biodegradability and widespread usage during the last decade [2,3,4]. ACY and ATL could not be effectively eliminated by the conventional treatment processes, including precipitation, biodegradation, and flocculation; hence, wastewater treatment plant effluents are regarded as one of the most important pollution resources. Therefore, in order to effectively respond to Green China and Carbon Neutrality [5,6], it is urgent to explore efficient and eco-friendly technologies for the diminution of ACY and ATL in water treatment.

In this sense, the UV/peroxydisulfate process (UV/PDS) has received considerable attention recently due to the in situ generation of sulfate radicals (SO₄^•−, 2.5–3.1 V) and hydroxyl radicals (^•OH, 1.8–2.7 V), relatively high stability, the wider operational pH range, and its high mineralization efficiency of organic pollutants [7,8,9,10,11]. To date, many studies have been conducted to investigate the removal efficiencies of diverse pollutants, the effects of main influential parameters, and the identification of reactive species, etc., in the UV/PDS process [12,13,14,15]. Liu et al. (2013) compared the degradation performance of ATL in UV/hydrogen peroxide (UV/H₂O₂) and UV/PDS processes and observed that UV/PDS was superior to UV/H₂O₂ for efficient ATL degradation due to the longer half-life time of SO₄^•− and its higher reactivity with ATL than ^•OH [3]. Although the removal efficiencies of diverse micropollutants in the UV/PDS process have been widely investigated, limited reports have taken note of the contributions and reaction mechanisms of SO₄^•− and ^•OH under different operational conditions.

On the basis of previous studies, the degradation of pollutants in the UV/PDS process is mainly attributed to ^•OH, SO₄^•−, and UV direct photolysis [9,12,16]. In order to accurately investigate the contributions and reaction mechanisms of SO₄^•− and ^•OH in the UV/PDS system, the microfluid-based UV reaction system (MVPS) was deployed in this study. The MVPS system, with the advantages of a small total volume, minimum waste liquid generation, and controllable reaction rate via radiation flux, could explore the dynamics process in detail and in-depth [17]. Moreover, both the steady-state assumption and the Kintecus model were applied to compare the radical contribution rates. Furthermore, the elucidation of the degradation pathways of ACY and ATL based on density functional theory (DFT) has rarely been reported in the UV/PDS system.

Therefore, the aims of the present study are (1) to explore the degradation performance of ACY and ATL and the effects of influencing factors in the UV/PDS process; (2) to determine the pH-dependent reaction rate constants of ^•OH and SO₄^•− with ACY and ATL; (3) to calculate the contributions of ^•OH and SO₄^•− at different operational factors via steady-state assumption and the Kintecus software; (4) to elucidate the degradation pathways of ACY and ATL based on DFT and ultra-performance liquid chromatography-tandem quadrupole time-of-flight mass spectrometry (UPLC-QTOF-MS) analysis in the UV/PDS process. It is worth noting that this research would provide some theoretical support and preliminary exploration, especially for the applicability in wastewater for micropollutants abatement replying on the UV/PDS process.

2. Material and Methods

2.1. Chemicals

Sinopharm Chemical Reagent (Shanghai, China) provided the majority of the commonly used experimental drugs. The PDS, humic acid sodium salt, ACY (>99.0%), ATL (>98%), and 5,5-dimethyl-1-pyrrolidine N-oxide (DMPO, >95%) were obtained from Sigma-Aldrich (St. Louis, MO, USA) and Aladdin (Shanghai, China). Methanol (MeOH) and acetonitrile (ACN) were chromatographically pure and supplied by Fisher Scientific (Geel, Belgium). Ultrapure water (>18.2 MΩ cm), obtained from an A10 Milli-Q system (Millipore, Boston, MA, USA), was used in all of the experiments.

2.2. Experimental Procedures

The batch experiments were conducted in the MVPS equipped with a micro-fluorescent silica detector (MFSD) in the present study (Figure S1). The detailed structure and operational parameters were described in the study by Li et al. [17]. The photon fluence rates at 254 nm (${\mathrm{F}}_{\mathrm{p},\mathrm{o},\mathrm{UV}}$) in the MVPS was determined and calculated to be 1.12 × 10^–3 Einstein m^–2 s^–1 in this study, based on the degradation of uridine (0.01 mM) (Text S1, Figure S2) [18].

Before the batch experiments in MVPS, the lamp was warmed for 30 min until the MFSD signal was stabilized. A mixture solution of moderate PDS reagent, ACY and ATL solution, and phosphate buffer (5 mM) was pumped through the UV tube to receive irradiation in MVPS, and the reaction was started. All of the reactions were performed at 20 °C. The samples were withdrawn from the MVPS system at interval times, immediately quenched with Na₂S₂O₃ solution, and stored at 4 °C for subsequent analysis.

2.3. Analytical Methods

The e2695 high-performance liquid chromatography (HPLC) system equipped with a Sunfire C18 column (150 mm × 4.6 mm, 5μm) and a 2489 Vis-UV detector (Waters, Milford, MA, USA) was used to measure the concentrations of ACY and ATL during the photolysis reaction. MeOH and water (v:v = 40:60), as the mobile phase with a flow rate of 0.55 mL min⁻¹, were deployed for ACY detection, while phosphate and ACN (v:v = 95:5) at a flow rate of 1.0 mL min⁻¹ was used for ATL analysis. DMPO, a spin-trapping reagent, was employed for the identification of radicals for the electron paramagnetic resonance test (EPR, Bruker A-300 spectrometer, Karlsruhe, Baden Wuerttemberg, Germany). Moreover, the UPLC-QTOF-MS system (Agilent, Wilmington, DE, USA) was applied to determine the intermediates of ACY and ATL in the UV/PDS process. The detection parameters of the UPLC-QTOF-MS system are Supplemented in Text S2 in detail.

2.4. Determination of the Second-Order Reaction Rate Constants

The second-order reaction rate constants of ACY and ATL with ^•OH ($k_{{}^{•}OH,ACY}$ and $k_{{}^{•}OH,ATL}$) and SO₄^•− ($k_{{SO}_4^{•-},ACY}$ and $k_{{SO}_4^{•-},ATL}$) at different pH values were determined using the competitive kinetics method. The p-chlorobenzoic acid (pCBA) ($k_{{}^{•}OH,pCBA}$ = 5 × 10⁹ M⁻¹ s⁻¹) and benzoic acid (BA) ($k_{{SO}_4^{•-},BA}$ = 1.2 × 10⁹ M⁻¹ s⁻¹) were chosen as the competitive chemicals in the UV/H₂O₂ and UV/PDS processes, respectively [19]. The calculation steps and results are depicted in Text S3 and Figures S3–S7.

2.5. Determination of the Relative Contributions of ^•OH and SO₄^•−

The degradation of ACY and ATL in the UV/PDS system could be represented as Equations (1) and (2).

$$-\ln \frac{{[\mathrm{A}\mathrm{C}\mathrm{Y}]}_T}{{[\mathrm{A}\mathrm{C}\mathrm{Y}]}_0}=k_{{}^{•}OH,ACY}{\displaystyle \int [{}^{•}\mathrm{O}\mathrm{H}]}dt+k_{S{O}_4^{·-},ACY}{\displaystyle \int [\mathrm{S}{\mathrm{O}}_4^{•-}]}dt+k_{PDS,ACY}{\displaystyle \int [\mathrm{P}\mathrm{D}\mathrm{S}]}dt+k_{UV,ACY}t$$

(1)

$$-\ln \frac{{[\mathrm{A}\mathrm{T}\mathrm{L}]}_T}{{[\mathrm{A}\mathrm{T}\mathrm{L}]}_0}=k_{{}^{•}OH,ATL}{\displaystyle \int [{}^{•}\mathrm{O}\mathrm{H}]}dt+k_{S{O}_4^{·-},ATL}{\displaystyle \int [\mathrm{S}{\mathrm{O}}_4^{•-}]}dt+k_{PDS,ATL}{\displaystyle \int [\mathrm{P}\mathrm{D}\mathrm{S}]}dt+k_{UV,ATL}t$$

(2)

The PDS oxidation and UV direct photolysis are negligible (Figures S8 and S9). Meanwhile, Equations (1) and (2) can be simplified to Equations (3) and (4) on account of the steady-state assumption of the radicals.

$$\begin{array}{cc}\hfill -\ln \frac{{[\mathrm{A}\mathrm{C}\mathrm{Y}]}_T}{{[\mathrm{A}\mathrm{C}\mathrm{Y}]}_0}& \approx k_{{}^{•}OH,ACY}{\displaystyle \int [{}^{•}\mathrm{O}\mathrm{H}]}dt+k_{S{O}_4^{•-},ACY}{\displaystyle \int [\mathrm{S}{\mathrm{O}}_4^{•-}]}dt\hfill \\ & =(k_{{}^{•}OH,ACY}{[{}^{•}\mathrm{O}\mathrm{H}]}_{ss}+k_{S{O}_4^{•-},ACY}{[\mathrm{S}{\mathrm{O}}_4^{•-}]}_{ss})t\hfill \end{array}$$

(3)

$$\begin{array}{cc}\hfill -\ln \frac{{[\mathrm{A}\mathrm{T}\mathrm{L}]}_T}{{[\mathrm{A}\mathrm{T}\mathrm{L}]}_0}& \approx k_{{}^{•}OH,ATL}{\displaystyle \int [{}^{•}\mathrm{O}\mathrm{H}]}dt+k_{S{O}_4^{•-},ATL}{\displaystyle \int [\mathrm{S}{\mathrm{O}}_4^{•-}]}dt\hfill \\ & =(k_{{}^{•}OH,ATL}{[{}^{•}\mathrm{O}\mathrm{H}]}_{ss}+k_{S{O}_4^{•-},ATL}{[\mathrm{S}{\mathrm{O}}_4^{•-}]}_{ss})t\hfill \end{array}$$

(4)

where [^•OH]_ss and [SO₄^•−]_ss represent the steady-state concentrations of ^•OH and SO₄^•−, respectively. In addition, the apparent degradation rate constants of ACY and ATL (k_obs) would be further simplified to Equations (5)–(7).

$$k_{obs}{}_{}=k_{{}^{•}OH,ACY}{[{}^{•}\mathrm{O}\mathrm{H}]}_{ss}+k_{S{O}_4^{•-},ACY}{[\mathrm{S}{\mathrm{O}}_4^{•-}]}_{ss}$$

(5)

$$k_{obs}{}_{}=k_{{}^{•}OH,ATL}{[{}^{•}\mathrm{O}\mathrm{H}]}_{ss}+k_{S{O}_4^{•-},ATL}{[\mathrm{S}{\mathrm{O}}_4^{•-}]}_{ss}$$

(6)

$$k_{obs}=k_{exp,{}^{•}OH}+k_{exp,S{O}_4^{•-}}$$

(7)

Nitrobenzene (NB), a typical probe compound of ^•OH ($k_{{}^{•}OH,NB}$ = 3.0 × 10⁹ M⁻¹ s⁻¹, $k_{S{O}_4{}^{•-},NB}$ < 10⁶ M⁻¹ s⁻¹), was used to determine the [^•OH]_ss [20,21] in the UV/PDS process. The [^•OH]_ss during the reaction could be further expressed as Equation (8):

$$k_{obs,NB}=k_{{}^{•}OH,NB}{[{}^{•}\mathrm{O}\mathrm{H}]}_{ss}+k_{UV,NB}$$

(8)

where $k_{obs,NB}$ and $k_{UV,NB}$ indicate the pseudo-first-order rate constant and the single UV photolysis rate constant (min⁻¹) of NB (Figure S10). $k_{{}^{•}OH,NB}$ represents the reaction rate constants of NB with ^•OH.

2.6. Kinetic Model

In this study, a reaction kinetic model was performed using Kintecus 6.80 software (James C. Ianni, Albuquerqua, NM, USA) [22] to simulate the concentrations of ACY, ATL, and reactive radicals in the UV/PDS process on the basis of 150 elementary reactions (Table S1). All of the rate constants of the above reactions were acquired from our determinations, literature, or estimations of similar reactions. The Kintecus model, with its high predictability and accuracy, has been widely used for simulating chemical reaction kinetics in the UV/chlorine [23], UV/H₂O₂ [24,25], and UV/PDS processes [12,25].

2.7. Quantum Chemistry Calculation

The molecular regioselectivity of ACY and ATL was investigated by the quantum chemical calculation based on DFT. All of the geometry optimizations were implemented at the B3LYP/6–31 G (d) level using the Gaussian 09 program [26]. The wavefunction analysis was carried out at the B3LYP/6–31G (d) level using Multiwfn 3.8 (T. Lu, Beijing, China) [27], based on the structure configuration from the Gaussian optimization. The condensed Fukui function (f⁰) was employed as a popular and powerful tool in this study to predict the regioselectivity of ACY and ATL to radical attack at the atomic level. All of the calculation details of f⁰ are Supplemented in Text S4.

3. Results and Discussion

3.1. Degradation Efficiencies of ACY and ATL in Different Processes

Figure 1 reveals that the removal of ACY and ATL was significantly promoted in the UV/PDS process with the k_obs values of 0.0931 and 0.1938 min⁻¹ (R² ≥ 0.99), respectively. However, the ACY and ATL degradation via direct PDS oxidation and UV photolysis can be neglected (Figure 1). The UV/PDS process can produce both SO₄^•− and ^•OH due to UV activation of peroxide bond and sulfate conversion [19] (Figure S11) (Equations (9) and (10)) [12,28], resulting in distinct improvement for the elimination of quinolone drugs [29], dyestuff [30], nonsteroidal anti-inflammatory drugs [10], etc.

$${\mathrm{S}}_2{\mathrm{O}}_8{}^{2-}+{\mathrm{hv}}_{254}\to 2\mathrm{S}{\mathrm{O}}_4^{•-}$$

(9)

$$\mathrm{S}{\mathrm{O}}_4{}^{•-}+{\mathrm{OH}}^{-}\to {}^{•}\mathrm{O}\mathrm{H}+\mathrm{S}{\mathrm{O}}_4{}^{2-}$$

(10)

To verify the roles of SO₄^•− and ^•OH for the enhanced ACY and ATL removal, the scavenger experiments were performed in the UV/PDS process (Figure 1). Tertiary butanol (TBA) was used to quench ^•OH ($k_{{}^{•}OH,TBA}$= (3.8−7.6) × 10⁸ M⁻¹ s⁻¹ and $k_{{SO}_4^{•-},TBA}$= (4.0−9.1) × 10⁵ M⁻¹ s⁻¹) [31,32,33], and MeOH can quench both SO₄^•− and ^•OH ($k_{{}^{•}OH,M\mathrm{e}OH}$ = 9.7 × 10⁸ M⁻¹ s⁻¹ and $k_{{SO}_4^{•-},M\mathrm{e}OH}$ = 2.5 × 10⁷ M⁻¹ s⁻¹) [34,35,36]. It is notable from Figure 1 that in the presence of TBA and MeOH, the k_obs values of ACY degradation reduced by 7.53% and 57.0%, while they declined by 6.73% and 42.0% for ATL, respectively. It is noteworthy from Figure 1 that SO₄^•− instead of ^•OH played a vital role in both ATL and ACY degradation in the UV/PDS process. Therefore, it is essential to evaluate the specific contribution rates of SO₄^•− and ^•OH for pollutant removal under different parameters in the UV/PDS process.

3.2. Effects of PDS Dosage

The effects of PDS dosage on the degradation of ACY and ATL were explored and modeled by the Kintecus 6.80 software(James C. Ianni, Albuquerqua, NM, USA). As depicted in Figure 2, with the increment in the PDS dosage from 0.1 to 1.0 mM, the degradation of both ACY and ATL was significantly strengthened, and the k_obs values of ACY and ATL degradation were promoted from 0.0104 to 0.1213 min⁻¹ and 0.0432 to 0.3031 min⁻¹ (R² > 0.99), respectively. The specific rate constants of ^•OH and SO₄^•− for ACY and ATL degradation (k_exp,_•OH and $k_{{exp,SO}_4^{•-}}$) calculated by the steady-state assumption are also presented in Figure 2. The modeled degradation rates of ACY and ATL (k_obs,mod) highly agreed with the experimental values (k_obs) in the UV/PDS process within a credible range (<25%).

In addition, it can be noted from

Table 1. [^•OH]_ss and [SO₄^•−]_ss in UV/PDS process at different PDS dosages and pHs.

Radical Species (M)	PDS (mM)				pH
	0.1	0.3	0.6	1.0	6.0	9.0	10.0	11.0
[•OH]ss × 10–14	1.00	3.00	9.83	28.72	9.44	12.27	17.72	17.86
[SO4•–]ss × 10–13	1.46	3.39	6.03	8.38	6.13	3.83	2.66	1.79

that the [^•OH]_ss and [SO₄^•−]_ss in the UV/PDS process gradually increased from 1.00 × 10⁻¹⁴ to 28.72 × 10⁻¹⁴ M and from 1.46 to 8.38 × 10⁻¹³ M as the PDS dosage rose from 0.1 to 1.0 mM. Due to the rapid photodecomposition rate of PDS (Figure S12), the yields of both [^•OH]_ss and [SO₄^•−]_ss in the UV/PDS system were evidently expedited (Table 1) [37], resulting in the gradual increment of the k_exp,_•OH and $k_{{exp,SO}_4^{•-}}$ values of ACY and ATL (Figure 2). The higher [SO₄^•−]_ss led to the higher contribution of SO₄^•− to the degradation of ACY and ATL, manifesting that SO₄^•− rather than ^•OH played a major role in pollutant removal. It can be obviously seen from Figure 2 that the modeled results ($k_{obs,mod}$) derived from the kinetic model were in good agreement with the experimental values ($k_{obs}$) in most cases.

3.3. Effects of Solution pH

The $k_{{}^{•}OH,ACY}$,$k_{{SO}_4^{•-},ACY}$, $k_{{}^{•}OH,ATL}$, and $k_{{SO}_4^{•-},ACY}$ values at pH 6.0−11.0 were determined based on the competition kinetics, and the results are depicted in Figure 3. The $k_{{}^{•}OH,ACY}$ and $k_{{SO}_4^{•-},ACY}$ values declined from 1.50 × 10⁹ to 1.30 × 10⁹ M⁻¹ s ⁻¹ and from 1.58 × 10⁹ to 1.00 × 10⁹ M⁻¹ s⁻¹, respectively, with the pH value rising from 6.0 to 11.0 (Figure 3a,b). Russo et al. reported the $k_{{}^{•}OH,ACY}$ to be 2.3 × 10⁹ M⁻¹ s⁻¹ at pH 6.0 [2], and the $k_{{}^{•}OH,ACY}$ and $k_{{SO}_4^{•-},ACY}$ values at different pHs have been seldom investigated. For ATL (Figure 3c,d), with the increment of pH (6.0−11.0), the $k_{{}^{•}OH,ATL}$ decreased from 4.6 × 10⁹ to 4.11 × 10⁹ M⁻¹ s ⁻¹, while the $k_{{SO}_4^{•-},ATL}$ increased from 4.56 × 10⁹ to 5.64 × 10⁹ M⁻¹ s ⁻¹. The pH dependency of the second-order reaction rate constant could be ascribed to a combination of various effects due to the reaction of free radicals (^•OH and SO₄^•−) with dissimilar pollutant species, which are the significant parameters of the following kinetic study.

As shown in Figure 4, with the increase in the solution pH from 6.0 to 11.0, the k_obs of ACY and ATL distinctly declined from 0.0931 to 0.071 min⁻¹ and 0.1938 to 0.101 min⁻¹, respectively, and the $k_{obs,}{}_{mod}$ were consistent with the experimental results. Additionally, as the pH increased from 6.0 to 11.0, the k_exp,_^•OH slightly enhanced from 0.00878 to 0.01378 min⁻¹ for ACY degradation and from 0.02607 to 0.04357 min⁻¹ for ATL degradation, respectively. However, the overt reduction in $k_{{exp,SO}_4^{•-}}$ for ACY and ATL degradation was observed. The $k_{{exp,SO}_4^{•-}}$ of ACY and ATL decreased by 30.59% and 63.79%, respectively (Figure 4).

Moreover, the [^•OH]_ss and [SO₄^•−]_ss exhibited similar change patterns with the $k_{{exp,}_{•OH}}$ and $k_{{exp,SO}_4^{•-}}$ (Table 1). It is notable that the [^•OH]_ss obviously increased from 9.44 × 10⁻¹⁴ to 17.86 × 10⁻¹⁴ M, while the [SO₄^•−]_ss reduced from 6.13 to 1.79 × 10⁻¹³ M with the increasing pH from 6.0 to 11.0 in the UV/PDS processes. Guan et al. (2011) [38] discovered that when the solution pH > 9.3, the distinct fast conversion of SO₄^•− to ^•OH caused the declining formation rate of SO₄^•− from photolysis of peroxymonosulfate (PMS) and the reduction in the oxidative power, which were in agreement with the changes of [^•OH]_ss and [SO₄^•−]_ss in the current study. Liu et al. 2013 [3] investigated the elimination of ATL in the UV/PDS process with a pH range of 3.0−11.0 and discovered the declined degradation rate of ATL with the pH decreasing from 7.0 to 9.0. Furthermore, the reaction of OH⁻ with SO₄^•− by Equation (10) at high pH, could also result in the decreasing trend of SO₄^•− [29].

3.4. Effects of Operation Parameters in UV/PDS Process

3.4.1. Chloride

The existence of Cl⁻ significantly affected the removal efficiencies of ACY and ATL in the UV/PDS process (Figure 5). As shown in Figure 5a,b, the k_obs of ACY and ATL decreased by 36.56% and 53.36%, with augmenting Cl⁻ concentration from 0 to 3 mM. Additionally, the $k_{\mathrm{mod},{}^{•}OH}$ increased by 62.0−344.4% and $k_{{mod,SO}_4^{•-}}$ declined by 82.7−85.1% in the presence of 3 mM Cl⁻, respectively. With the augmentation of the Cl⁻ concentration from 0.5 to 3 mM, the [^•OH]_ss gradually increased from 6.62 × 10⁻¹⁴ to 32.24 × 10⁻¹⁴ M, while the [SO₄^•−]_ss decreased from 4.52 × 10⁻¹³ to 1.38 × 10⁻¹³ M for ACY and ATL degradation (Table S2). Lutze et al. [23] reported that SO₄^•− would be converted to ^•OH in the presence of Cl⁻ at pH ≥ 5.0 (Equations (11)−(15)). Cl⁻ can react quickly with ^•OH to form ClOH^•− with a rate constant of 4.3 × 10⁹ M⁻¹ s⁻¹(Equation (13), reverse). In addition, the extremely rapid decomposition rate constant of ClOH^•− (6.1 × 10⁹ M⁻¹ s⁻¹, Equation (13)) can obviously suppress the reaction between Cl⁻ and ^•OH. However, only at pH < 3.0 did the yield of Cl^• from the reaction of ClOH^•− with H⁺ become important (Equation (14)) [39,40]. Noteworthily, the reactions between Cl⁻ and ^•OH/SO₄^•− could also produce several reactive chlorine species (Cl^•, Cl₂^•−, ClO^•, etc.).

$${\mathrm{SO}}_4{}^{•-}+{\mathrm{Cl}}^{-}\to {\mathrm{SO}}_4{}^{2-}+{\mathrm{Cl}\hspace{1em}}^{•}k=3.0\times {10}^8{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(11)

$${\mathrm{Cl}}^{•}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{ClOH}}^{•-}+{\mathrm{H}}^{+}\hspace{1em}k=2.5\times {10}^5{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(12)

$${\mathrm{ClOH}}^{•-}\leftrightarrow {}^{•}\mathrm{O}\mathrm{H}+{\mathrm{Cl}}^{-}{k}_{for}=6.1\times {10}^9{\mathrm{M}}^{-1}{\mathrm{s}}^{-1},k_{rev}=4.3\times {10}^9{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(13)

$${\mathrm{ClOH}}^{•-}+{\mathrm{H}}^{+}\to {\mathrm{Cl}}^{•}+{\mathrm{H}}_2\mathrm{O}\hspace{1em}k=2.1\times {10}^{10}{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(14)

$${\mathrm{Cl}}^{•}+{\mathrm{Cl}}^{-}\to {\mathrm{Cl}}_2{}^{•-}\hspace{1em}k=8.5\times {10}^9{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(15)

3.4.2. Bicarbonate

It can be seen from Figure 5c,d that the k_obs for the abatement of ACY and ATL declined by 8.8−11.82% with the addition of 3 mM HCO₃⁻ in the UV/PDS system. HCO₃⁻ can react quickly with ^•OH (8.5 × 10⁶ M⁻¹ s⁻¹) and SO₄^•− (1.6 × 10⁶ M⁻¹ s⁻¹) to form secondary free radical CO₃^•− [6,41], leading to the scavenging effects of ^•OH and SO₄^•−. Furthermore, the presence of 3 mM HCO₃⁻ caused $k_{mod,{}^{•}\mathrm{O}\mathrm{H}}$, and $k_{{mod,SO}_4^{•-}}$ to decrease by 32.6−36.6% and 25.5−32.6%, respectively. Moreover, at pH 6.0, both H₂CO₃ and HCO₃⁻ (pK_a1 = 6.3, pK_a2 = 10.3) reacted with SO₄^•− in an analogous rate constant [42,43,44]. According to the simulation results (Table S2), [^•OH]_ss and [SO₄^•−]_ss distinctly decreased by 6.67%−5.69% and 7.08%−3.77% as the HCO₃⁻ concentration rose from 0.5 to 3 mM, respectively, resulting in the obvious reduction in the degradation rates of ACY and ATL.

3.4.3. Bromide

It can be observed from Figure 6a,b that the addition of Br⁻ inhibited the removal efficiencies of ACY and ATL in the UV/PDS process. The k_obs for ACY and ATL degradation decreased by 37.7 and 61.14%, respectively, as the Br⁻ concentration increased from 0 to 1.0 mM. Furthermore, $k_{mod,{}^{•}\mathrm{O}\mathrm{H}}$ and $k_{{mod,SO}_4^{•-}}$ decreased by 2.6−66.1% and 51.5−75.1% with the addition of 1.0 mM Br⁻. Because Br⁻ exhibits higher rate constants reacting with SO₄^•− and ^•OH (Equations (16) and (17)) than ACY and ATL do, it can be considered a scavenger of SO₄^•− and ^•OH, which may be the main reason for the decreased removal efficiencies of ACY and ATL. In addition, Lu et al. 2016 [45] also observed that the presence of Br⁻ notably suppressed the removal of o-phthalic acid (PA) in SO₄^•−-based advanced oxidation systems, and PA probably did not react with the generated reactive bromide species directly.

$${\mathrm{SO}}_4{}^{•-}+{\mathrm{Br}}^{-}\to {\mathrm{SO}}_4{}^{2-}+{\mathrm{Br}}^{•}\hspace{1em}k=3.5\times {10}^9{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(16)

$${}^{•}\mathrm{O}\mathrm{H}+{\mathrm{Br}}^{-}\to {\mathrm{BrOH}}^{•-}\hspace{1em}k=1.0\times {10}^{10}{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$

(17)

3.4.4. NOM

It can be observed from Figure 6c,d that the natural organic matter (NOM) markedly inhibited the removal efficiencies of ACY and ATL in the UV/PDS process. It is worth noting that the k_obs for ACY and ATL significantly dropped by 36.6% and 50.9% with the addition of 3.0 mgC L⁻¹ NOM. Meanwhile, the $k_{mod,{}^{•}\mathrm{O}\mathrm{H}}$, and $k_{{mod,SO}_4^{•-}}$ decreased by 35.3−57.8% and 25.7−56.3%, respectively, under the same conditions (3 mgC L⁻¹ NOM). It is noted from Figure 6 that the experimental k_obs for the abatement of ACY and ATL were highly compliant with the simulated results. It is well known that NOM can react with ^•OH (1.4 × 10⁴ L mgC⁻¹ s⁻¹) and SO₄^•− (6.8 × 10³ L mgC⁻¹ s⁻¹), resulting in their reduction [23]. Furthermore, NOM could act as an inner filter of UV light affecting the yields of ^•OH and SO₄^•− derived from direct photolysis of PDS [44,46].

3.4.5. Sulfate and Nitrate

It can be seen in Figure S13 that the abatement of ACY and ATL in the UV/PDS process was markedly restrained by the addition of SO₄²⁻. The k_obs for ACY and ATL decreased by 15.1% and 13.8%, respectively, when the concentration of SO₄²⁻ increased to 3 mM. In addition, the $k_{mod,{}^{•}\mathrm{O}\mathrm{H}}$ and $k_{{mod,SO}_4^{•-}}$ decreased by 18.6−35.3% and 19.2−33.1% in the presence of 3 mM SO₄²⁻. Furthermore, we can intuitively find that the experimental values are consistent with the simulated values. The higher concentration of SO₄²⁻ could induce the reductions of the redox potential of SO₄^•−/SO₄²⁻ and the oxidation capacity of the UV/PDS process on the basis of the Nernst equation (Equations (18) and (19)) [47,48]. Whereas SO₄^•− could also be formed through the reaction of ^•OH with SO₄²⁻(Equation (20)). However, the lower reaction rate constants of SO₄^•− with ACY/ATL than ^•OH would lead to suppressive effects on the k_obs of ACY and ATL degradation.

$${\mathrm{SO}}_4{{}^{•}}^{-}+{\mathrm{e}}^{-}\to {\mathrm{SO}}_4{{}^2}^{-}$$

(18)

$${\mathrm{E}}_{{(\mathrm{SO}}_4{}^{·-}/{\mathrm{SO}}_4{{}^2}^{-})}={\mathrm{E}}^{\mathsf{\theta }}{}_{{(\mathrm{SO}}_4{}^{·-}/{\mathrm{SO}}_4{{}^2}^{-})}+\frac{\mathrm{RT}}{\mathrm{zF}}\ln \frac{\left[{\mathrm{SO}}_4{}^{·-}\right]}{\left[{\mathrm{SO}}_4{{}^2}^{-}\right]}$$

(19)

$${\mathrm{SO}}_4{{}^2}^{-}+{}^{·}\mathrm{O}\mathrm{H}\to {\mathrm{SO}}_4{}^{·-}+{\mathrm{OH}}^{-}$$

(20)

NO₃⁻ (0.5−3 mM) visibly inhibited the removal efficiencies of ACY and ATL in the UV/PDS system (Figure S14). The k_obs for ACY and ATL decreased by 26.0% and 27.7%, as the NO₃⁻ concentration rose to 3 mM. The $k_{mod,{}^{•}\mathrm{O}\mathrm{H}}$ and $k_{{mod,SO}_4^{•-}}$ decreased by 18.6−37.1% and 19.2−36.3% in the presence of 3 mM NO₃⁻, respectively. NO₃⁻ can be used as a photosensitizer, consuming UV₂₅₄ during the reaction. Wang et al. demonstrated the strong inhibition of NO₃⁻ on the degradation of TAP in the UV/PDS system [49]. Lin et al. revealed that NO₃⁻ can significantly reduce the radiation intensity received by PMS, thereby slowing down the degradation of BPA [21].

3.5. Oxidation Mechanisms and Degradation Pathways Speculation

The f⁰ of ACY and ATL based on the Hirshfeld charge was calculated using Multiwfn software [46,50,51], and the results are represented in Tables S3 and S4 and Figure 7 and Figure 8. Moreover, Tables S5 and S6 tabulated the reasonable transformation products (TPs) of ACY and ATL [52,53], which were detected by UPLC-QTOF-MS.

According to the results in Table S3 and Figure 7a, the highest values of f⁰ (blue area in Figure 7a) were achieved in the C1, C2, C5, C8, N6, N16, and O15 positions of ACY (Figure S15a), which were much easier to be attacked by ^•OH and/or SO₄^•−. An et al. [54] also found similar degradation patterns for ACY, and the dominating reaction pathway of ACY with ^•OH initiated by the addition onto the ortho site of methoxy substituent on the benzene ring (i.e., C1 and C2 positions in this study). As shown in Figure 7b, combining the six detected transformation intermediates (Table S5) and the results of the Fukui function calculations, three degradation pathways for ACY in the UV/PDS process were proposed. Pathway I underwent the electrophilic addition reaction through ^•OH attacking the ACY N-heterocycle, resulting in the formation of the hydroxylated products TP1 and TP2 through dihydroxylation. Pathway II: ACY may also transform to TP3 through a de-alcoholic reaction and subsequent hydroxylation. Pathway III: The C5 and N6 could be further attacked by ^•OH and SO₄^•−, further causing ring opening and the production of TP4 through further hydroxylation. Then, TP4 could generate TP5 through H addition or TP6 by de-alcoholic reaction and hydroxylation.

It can be observed from Table S4 and Figure 8a that the ^•OH/SO₄^•− radicals most favorably attack the C5, C6, C7, C18, C19, O1, and O9 positions of ATL (Figure S15b). In addition, Miao et al. [55] also reported that ^•OH would most likely attack the C3 position of ATL (i.e., the C6 position in this study) to produce hydroxylation products. In addition, Miao et al. [55] also reported that the atom with a higher 2FED²_HOMO value in the ATL structure was the O11 position (i.e., O9 position in this study), indicating that the breaking of the ether bond is reasonable. Herein, three plausible transformation pathways of ATL (Figure 8b) on the basis of the 10 transformation products (Table S6) were as follows. Pathway I: the oxidation of the primary amide group in the main side chain would lead to the formation of TP1 with m/z 268. Subsequently, the addition of oxygen to the alkyl group and hydrogen abstraction by ^•OH contributed to the generation of TP2 and was further hydroxylated to produce TP3. ^•OH/SO₄^•− can attack ether linkage and undergo cleavage of ether linkage to form TP4 with m/z 134. Pathway II: The secondary amine structure of ATL could be attacked by ^•OH/SO₄^•−, causing the cleavage of the C−N bond to produce TP5. Afterward, TP4 and TP6 could also be attributed to the split of the ether linkage caused by the attack of ^•OH/SO₄^•−. After that, the reaction initiated at the secondary amine moieties (–NH–) and dimerization would lead to the yield of TP7. Pathway III revealed the electrophilic addition reaction through ^•OH attacking the ATL aromatic ring, producing the hydroxylated product TP8, which was then carbonylated to generate TP9 and hydroxylated to TP10.

4. Conclusions

In summary, the degradation of ACY and ATL derived from SO₄^•− and ^•OH in the UV/PDS process was comprehensively investigated concerning kinetic simulations and degradation mechanisms. The second-order rate constants at 6.0−11.0 for ACY and ATL reacting with ^•OH were estimated to be 1.30−1.50 and 3.45−4.60 M⁻¹s⁻¹, and those for ACY and ATL with SO₄^•− were 1.00−1.58 and 4.56−5.64 M⁻¹s⁻¹, respectively. In addition, the radical quenching experiments and the steady-state assumption calculation demonstrated that SO₄^•− instead of ^•OH was the predominant radical for pollutant elimination in the UV/PDS process. Additionally, the kinetic model deployed in the current study could well predict the degradation rates of ACY and ATL under different operational parameters. The degradation rates of ^•OH and SO₄^•− for ACY and ATL degradation were highly pH-dependent. Based on the prediction of the active sites via the Fukui function and the intermediates of ACY and ATL identified by UPLC-QTOF-MS, the plausible transformation pathways of ACY and ATL in the UV/PDS process have been proposed. This study could provide some theoretical support and preliminary exploration for pollutant removal in the UV/PDS system. Furthermore, the toxicity of the target pollutants and their intermediates in the UV/PDS system could be considered in future studies.