Electrochemical Co-Degradation of Acetaminophen and Bisphenol A in Aqueous Solutions: Degradation Competition and Pathways

Abstract

This study investigated the degradation competition and pathways of electrochemical co-degradation of two emerging environmental contaminants, polar acetaminophen (AP) and (moderately) non-polar bisphenol A (BPA), on a boron-doped diamond (BDD) electrode in aqueous solutions. The results showed that both compounds mainly relied on hydroxyl radicals (•OH) to trigger indirect oxidation for their electrochemical degradation, although AP also underwent direct oxidation during electrolysis. The effect of increasing current density on the increases in degradation performance was almost the same for AP and BPA. However, BPA exhibited a better performance in mono-degradation than AP, while the opposite tendency was observed for their co-degradation. Their degradation efficiencies were better in 1 M Na₂SO₄ solution than in a real water matrix. Both UV-vis and excitation–emission matrix (EEM) fluorescence analyses demonstrated that all the aromatic rings of AP and BPA were opened after 30 min of electrolysis at 0.5 A cm⁻² in 1 M Na₂SO₄ solution. Regardless of the small difference in intermediate species, the pathways of electrochemical AP+BPA co-degradation were similar to those of their mono-degradation combination. A double exponential decay model is proposed to simulate the formation and degradation rate constants of benzoquinone (an intermediate).

1. Introduction

Acetaminophen (AP) and bisphenol A (BPA) are two phenolic compounds regarded as contaminants of emerging concern (CECs); moreover, they and their transformation products can be co-present to cause environmental and ecotoxicological impacts under biotic and/or abiotic conditions [1]. AP, a well-known drug used to treat fever and mild to moderate pain, is excreted through urine to the sewer systems, wastewater treatment plants (WWTPs), and other aquatic/natural environments, resulting in concentrations from 0.1 to 300 mg L⁻¹ in effluents in many countries; however, its overall removal is insufficient (<85%) in conventional treatments including sedimentation, activated sludge with adsorption, and/or filtration with membranes, mainly because of its high water solubility and low biodegradability [2]. BPA, one of the most commonly used plasticizers, has been ubiquitously detected in the environment including water bodies such as groundwater, surface water, and effluents [3,4,5]. Also, BPA is an endocrine-disrupting chemical (EDC) that has been linked to obesity, altered reproductive behaviors, and decreased fertility in humans and wildlife [3,5]. Industrial wastewater, domestic sewage, and agricultural drainage may discharge BPA into surface and groundwater [3]; however, conventional water treatment processes often exhibit unsatisfactory removal of BPA because the toxicity of BPA can significantly inhibit microbial metabolism and growth in biotreatment, which stresses the need for better treatment processes [4,5].

Electrochemical advanced oxidation processes (EAOPs) have received extensive attention in the field of wastewater treatment, with advantages of using electrons instead of chemicals as pollutant scavengers and low requirements for reaction conditions and energy consumption [6]. Anodic oxidation (AO) is the simplest EAOP, which is suitable for the abatement of various aqueous organic pollutants. In AO processes, organic pollutants in aqueous solution can be degraded through direct oxidation on the anode surface and/or indirect oxidation by electrochemically generated oxidants such as •OH (E⁰ = 2.80 V/SHE [2,6]), SO₄^•− (E⁰ = 2.5–3.1 V/SHE) [6], and S₂O₈²⁻ (E⁰ = 2.01 V/SHE) [7]. The electrogeneration of hydroxyl radicals from water decomposition and mineralization of target compounds (TCs) to form CO₂, H₂O, and other inorganic compounds (OICs) on BDD can be described as reactions 1 and 2, respectively [6,8].

BDD + H₂O → BDD(•OH) + H⁺ + e⁻

(1)

•OH + TC → mCO₂ + nH₂O + OICs

(2)

AO has been adopted to degrade AP or BPA in different solutions using various anode materials (e.g., metal, mixed metal oxide, carbon material, composite materials, and BDD electrodes); among these electrodes, BDDs are good candidates and usually show excellent function for AP or BPA degradation and mineralization [2,8,9]. For instance, it was reported that the performance of a Ti/PbO₂ anode was better than those of Ti/Pt and dimensionally stable anodes (DSAs) for AP electrochemical degradation [10]; in addition, the presence of Ce(IV) improved the degradation performance of AP on Ti/PbO₂ or BPA on BDD in Na₂SO₄ solutions [11,12], while a better performance of AP electrochemical degradation on Ti/IrO₂ was observed in NaCl than in Na₂SO₄ solution because electrochemically generated active chlorine species also contributed to target compound degradation [13]. Nevertheless, the removal of total organic carbon (TOC) from BPA electrochemical degradation on Si/BDD was better in Na₂SO₄ than in NaCl or NaNO₃ solution [14], while the degradation efficiency of chemical oxygen demand (COD) from BPA electrochemical degradation on Ti/Co₃O₄/β-PbO₂ could reach 92.2% in 0.02 M NaCl solution [15]; furthermore, Ti/BDD was found to be more promising than Ti/Sb–SnO₂, Ti/RuO₂, and Pt for the effective electrochemical treatment of BPA and similar EDC pollutants [9]. In an electro-Fenton or a photoelectro-Fenton process with Fe²⁺ and H₂O₂ addition, AP was observed to achieve a 98% degradation efficiency when using a Ti/RuO₂/IrO₂ DSA anode and two stainless steel cathodes, but TOC removal was low (≤20%) [16].

The co-existence of different organic compounds might have a worse performance in electrochemical degradation than single compounds in aqueous solutions. For example, it was reported that the rate constants of electrochemical tetracycline and metronidazole co-degradation were lower than those of corresponding compounds [17]. However, less attention has been paid to the electrochemical AP or BPA degradation using BDD, and the electrochemical AP and BPA co-degradation using BDD has not been studied yet. AP is a polar compound, while BPA has a (moderately) non-polar chemical structure; this difference in polarity between AP and BPA may influence their adsorption on the BDD surface, so they are suitable for being compared as target compounds. Accordingly, this study explored the electrochemical co-degradation of AP and BPA in aqueous solutions and focused on their degradation competition and pathways.

2. Materials and Methods

2.1. Chemicals and Solutions

The target compounds of this study, AP and BPA (≥99%, Sigma, USA), were used as received to prepare stock solutions, while sodium sulfate (≥99%, Showa, Japan) dissolved in distilled deionized water (DDW) (Millipore Milli-Q, Merck KGaA, Darmstadt, Germany) was used to prepare the supporting electrolyte for the experiments. Methanol (99.9%, Echo), acetonitrile (99.9%, Echo), and p-benzoquinone (pBQ) (98%, Alfa, UK) were purchased from Echo Chemical Co., Ltd. (Taiwan). The real river water sample was collected from a river in southern Taiwan. A TS-100 pH meter was used to record pH values of the tested solutions while a TOC-5000 analyzer (Shimadzu, Japan) was adopted to measure the TOC. The river water sample was filtered using a 0.45 μm filter to remove suspend solids prior to testing.

2.2. Chemical Analysis

The excitation–emission matrix (EEM) fluorescence spectra were obtained using a Hitachi F-7000 fluorescence spectrometer (Japan), and the ultraviolet-visible (UV-vis) absorbance was measured with a Hitachi U-2900 UV-Vis spectrophotometer (Japan) [8,18]. The electron spin resonance (ESR) spectroscopy spectra were recorded on a Magnettech model MS5000 spectrometer (Germany).

A high-performance liquid chromatography (HPLC) instrument (Hitachi Chromaster 5420, UV-Vis Detector, Japan) was used to measure the concentrations of AP, pBQ, and BPA. Moreover, a Mightysil RP-18 GP column (5.0 µm, 4.6 × 250 mm) was used for the AP and pBQ analyses at 254 nm (retention times = 2.9 and 4.3 min, respectively), while a Shiseido C18 column (5.0 µm, 4.6 × 250 mm) was used for the BPA analysis at 225 nm. The mobile phases used for the analyses of AP, pBQ, and BPA were acetonitrile/water = 45/55, acetonitrile/water = 45/55, and methanol/water = 70/30, respectively (flow rate = 1.0 mL min⁻¹).

The degradation intermediates of AP, BPA, and AP+BPA were analyzed using a Thermo LCQ Deca XP liquid chromatography–mass spectrometry (LC-MS) system (USA), which performed chromatographic separation on a Phenomenex C18 100A column (5.0 µm, 2.1 × 150 mm) at 0.2 mL min⁻¹ and 50–600 (m/z) (electrospray ionization: + or −) [8]. The mobile phases comprised distilled 5%ACN (mobile phase A) and 95%ACN (mobile phase B). The gradient elution programs were A/B = 98%/2% (0–5 min), (98–25%)/(2–75%) (5–13 min), 25%/75% (13–17 min), (25–98%)/(75–2%) (17–18 min), and 98%/2% (18–30 min). Each sample was diluted in 5% ACN + 0.2% FA (formic acid).

2.3. Electrochemical Analysis

Cyclic voltammetry (CV) analyses in 1.0 M Na₂SO₄ (with/without AP, BPA, and AP+BPA) were performed in the potential scan ranges of −0.2→2.0 V (scan rate = 100 mV s⁻¹) using a CHI 660B electrochemical work station (CHI, USA). The BDD electrode served as the working electrode, while the counter and reference electrodes were a platinum wire and an Ag/AgCl electrode (3 M KCl, 0.207 V vs. SHE (standard hydrogen electrode) at 25 °C), respectively.

2.4. Degradation Experiments

The electrolytic experiments were all conducted in a divided thermostated device with a Nafion separator at 0.25−1.00 A cm⁻² and 25°C (Figure 1). In each test, the catholyte was 1.0 M Na₂SO₄ (100 mL) only, while the anolyte was 1.0 M Na₂SO₄ (100 mL) containing AP, BPA, or AP+BPA (100 mg L⁻¹ each), except those used for degradation intermediate tests. In the GGE electrochemical degradation experiments, an Nb/BDD Diachem^® (CONDIAS GmbH, Germany) (10 cm × 1.0 cm × 2.0 mm) was used as the anode, and a titanium plate (10 cm × 1.0 cm × 2.0 mm) was used as the cathode. Both electrodes had the same geometric working surface areas (2 cm²). A DC power supply (Gwinstek GPR-6060D, Taipei, Taiwan) was employed to perform the experiments. Samples were taken at intervals in the electrolysis experiments.

2.5. Degradation Efficiency, Mineralization Current Efficiency, and Pseudo-First-Order Rate Constant

The degradation efficiencies of AP and BPA are defined as Eff_AP and Eff_BPA, respectively.

Eff_AP or Eff_BPA = (1 − C_t/C₀) × 100%

(3)

where C₀ and C_t are the initial and time t concentrations of AP and BPA, respectively. Likewise, the TOC degradation efficiency (Eff_TOC) is expressed as the following equation:

Eff_TOC = (1 − TOC_t/TOC₀) × 100%

(4)

where TOC₀ and TOC_t are the initial and time t concentrations of TOC, respectively.

The mineralization current efficiency (MCE) of AP+BPA in electrolysis solutions at a given time t (h) is calculated by the following equation [19]:

$$MCE(\%)={\displaystyle \frac{nF{V}_s\Delta {(TOC)}_{\exp }}{4.32\times {10}^{7}mIt}}\times 100$$

(5)

where F is the Faraday constant (96,487 C/mol), V_s is the volume of solution (L), Δ(TOC)_exp is the value of TOC_t minus TOC₀ (mg/L), 4.32 × 10⁷ is a conversion factor (3600 s/h × 12,000 mg C/mol), m is the number of carbon atoms of AP+BPA, and I is the applied current (A). n is the number of electrons consumed in the mineralization process. The overall mineralization of the AP+BPA is assumed to proceed via the following reaction:

C₈H₉NO₂ + C₁₅H₁₆O₂ + 42H₂O → 23CO₂ + NH₄⁺ + 105H⁺ + 106e⁻

(6)

3. Results and Discussion

3.1. CV Analyses for Target Compounds

Figure 2a shows the voltammograms (100 mV s⁻¹) in 1.0 M Na₂SO₄ without/with AP on BDD. The BDD electrode exhibited an oxygen evolution potential (O1) of ~1.90 V vs. Ag/AgCl; moreover, an oxidation peak (O2) was observed around 1.57 V vs. Ag/AgCl, which was associated with the oxidation of SO₄²⁻ to S₂O₈²⁻ (persulfate) (2SO₄²⁻ → S₂O₈²⁻ + 2e⁻) or SO₄^•− (sulfate radical) [8]. When 1.0 M Na₂SO₄ contained 100 ppm AP, an additional oxidation peak (O3) was shown at 0.84 V vs. Ag/AgCl, which was attributed to the oxidation of AP, while the corresponding reduction peak (R) appeared at 0.63 V vs. Ag/AgCl. Additionally, the current peak at 0.84 V vs. Ag/AgCl increased with the increase in AP concentration (100–800 ppm) in 1.0 M Na₂SO₄, verifying its correspondence to AP oxidation. This quasi-reversible electrochemical behavior of oxidation/reduction, also observed on a glassy carbon electrode, was reported to follow Reaction 1, which involves in the formation of N-acetyl-p-benzoquinone-imine (NAPQI) (due to AP oxidation) and its reduction with 2-proton and 2-electron transfer [20].

AP = NAPQI +2H⁺ + 2e⁻

(7)

The CV of AP (100 ppm) still exhibited the oxidation/reduction coupled peaking at 0.84 and 0.63 V vs. Ag/AgCl, respectively (Figure 2b). However, no BPA oxidation peak was observed when 100 ppm BPA was present in 1.0 M Na₂SO₄. Interestingly, the voltammogram of AP+BPA (100 ppm each) showed an oxidation peak at 0.78 V vs. Ag/AgCl, significantly smaller than that of AP only, revealing that the BDD electrode was partially inactivated due to possible electrochemical BPA polymerization and its adsorption on the electrode surface. It was reported that the electrochemical polymerization of BPA might occur to hinder the oxidation of BPA on the electrode surface [21]. Accordingly, it is inferred that direct oxidation could occur on BDD during the electrochemical degradation of AP, while it was not the case for that of BPA.

3.2. Effect of Current Density on Electrochemical Degradation of Target Compounds

It is well-known that current density usually influences the degradation efficiency of a target organic compound. According to Figure 3a, all Eff_AP could reach 100% at different current densities (0.25−1.00 A cm⁻²); likewise, this was also true for Eff_BPA (Figure 3b). For the electrochemical co-degradation of AP+BPA in 1 M Na₂SO₄ solution, the time required for 100% (complete) AP degradation (t₁₀₀) at 0.25 A cm⁻² was 60 min, while the t₁₀₀ was only half of that at 0.50 A cm⁻². The t₁₀₀ was further shortened to 20 min at 0.75 A cm⁻², although the values of t₁₀₀ at 0.75 and 1.00 A cm⁻² were the same. This phenomenon was also observed for BPA degradation. Therefore, the degradation of AP or BPA increased with the increase in current density.

Similarly, the degradation of TOC also approximately increased with an increasing current density (Figure 3c). The t₁₀₀ at 0.25 A cm⁻² was 90 min, triple that at 1.00 A cm⁻². Although the values of t₁₀₀ at 0.50 and 0.75 A cm⁻² were the same, the former had an Eff_TOC value (58%) significantly lower than the latter (91%) at 30 min. On the other hand, the MCE decreased with the increase in current density (Figure 3d), which was associated with the fact that at the greater current density, more side reactions such as oxygen evolution (from direct water oxidation or •OH combination 2•OH → O₂ + 2H⁺ + 2e⁻) and weaker oxidant formation occurred to waste the current [19].

As discussed in Section 3.1, direct oxidation on BDD occurred for AP but did not for BPA during their electrochemical co-degradation. According to Huang et al. [8], oxidants (O_x) such as •OH, SO₄•⁻, and S₂O₈²⁻ can be electrochemically generated on BDD during electrolysis in sodium sulfate solutions and reacted with target compounds via indirect oxidation; moreover, •OH is favored to be formed over SO₄•⁻ at a low pH (≤2). Additionally, the oxidative power for •OH or SO₄•⁻ is stronger than for S₂O₈²⁻. In this study, the anolyte and catholyte were separated by a Nafion-212 cation exchange membrane. After electrolysis operation for 2.5−5.0 min, the pH value of the anolyte was lower than 2 due to the increase in hydrogen ion concentration from water oxidation; on the other hand, that of the catholyte was greater than 12 because of hydrogen ion or water reduction. In this situation, the anolyte was like a sulfuric acid solution, but the BDD electrode should have good stability and corrosion resistivity in the strong acidic medium [22]. In ESR analysis, only the typical DMPO-OH signal (quartet with an intensity ratio = 1:2:2:1 and hyperfine couplings a_N = a_Hβ = 1.45−1.49 mT = 14.5−14.9 G) was detected for the sample taken from a 0.1 M H₂SO₄ solution (pH < 1) electrolyzed using the BDD electrode (Figure 3e), indicating the formation of hydroxyl radicals based on the reaction of DMPO + •OH → DMPO-OH [8]. Huang et al. [8] also reported that the intensity of the DMPO-OH signal was much stronger that of DMPO-SO₄⁻, also in H₂SO₄ solutions (pH = 2) electrolyzed using BDD electrodes. Accordingly, it is inferred that the indirect oxidation of target compounds mainly relied on their reactions with •OH during electrolysis, although SO₄•⁻ and S₂O₈²⁻ also acted as oxidants. However, most of the target compounds and intermediate molecules should be oxidized by •OH formed on the electrode surface rather by SO₄•⁻ and S₂O₈²⁻ [14].

For the sake of simplicity, we assumed that the concentrations of •OH did not change significantly and that the chemical reaction between SO₄•⁻ (or S₂O₈²⁻) and AP (or BPA) (as well as AP direct oxidation) was minor or negligible in comparison to that between •OH and AP (or BPA) during electrolysis. Such a reaction can be regarded as a pseudo-first-order reaction, expressed as Equation (8), which is integrated to obtain Equation (9). Through the use of linear regression based on the Ln(X₀/X_t) data versus the time (t) plot, the pseudo-first-order rate constant (k) can be obtained from Equation (10):

$${\displaystyle \frac{-d[\mathrm{X}]}{dt}}=k^0[\mathrm{X}][•\mathrm{OH}]=k[X]$$

(8)

$$\mathrm{X}=X_0{e}^{-kt}$$

(9)

$${\mathrm{Ln}(\mathrm{X}}_0{/\mathrm{X}}_{\mathrm{t}})=-kt$$

(10)

where k⁰ is the rate constant (k = k⁰[•OH]); X is AP, BPA, or TOC; and subscripts 0 and t denote the time at 0 and t, respectively.

Accordingly, the k values of AP increased from 1.23 × 10⁻³ s⁻¹ to 3.39 × 10⁻³ s⁻¹, while those of BPA were raised from 6.75 × 10⁻⁴ s⁻¹ to 1.81 × 10⁻³ s⁻¹ (

Table 1. Pseudo-first-order rate constants k (×10⁻³, s⁻¹) for AP and BPA mono-degradation at 0.5 A cm⁻² (marked in *) or co-degradation at different current densities (CD) (0.25–1.00 A cm⁻²) in 1 M Na₂SO₄ solution or at 0.5 A cm⁻² in river water (RW); k₁ (×10⁻³, s⁻¹ A⁻¹) and m for AP and BPA co-degradation in 1 M Na₂SO₄ solution.

	k							k1	m
	1.0 M Na2SO4 Solution						RW
CD	0.25	0.50	0.75	1.00	0.50 *		0.50
AP	1.23	1.94	2.96	3.39	3.50		1.34	1.21	0.76
BPA	0.675	0.887	1.63	1.81		6.43	0.700	0.63	0.76
TOC	0.408	0.575	0.623	0.810	1.89	2.96	0.392

), based on the inset data of Figure 3a,b. Therefore, the Eff_AP or Eff_BPA increased with an increase in current density; however, the Eff_AP was greater than the Eff_BPA at the same electrolysis time, possibly because the chemical structure of AP (C₈H₉NO₂) was simpler than that of BPA (C₁₅H₁₆O₂); furthermore, the former was degraded by both direct and indirect oxidation while the latter was decomposed only by indirect oxidation. If referring to the approach used by Zhao et al. [15], k or [•OH] should be proportional to current (I) at the same electrode area, and thus these parameters can be expressed as the following equations:

$$k=k^0[•\mathrm{OH}]=k_1{I}^m$$

(11)

$$\mathrm{l}\mathit{nk}=\ln k_1+m\ln I$$

(12)

According to the slopes (lnk₁) and intercepts (m) of lnk vs. lnI linear plots for AP and BPA (R² = 0.988 and 0.923, respectively) in Figure 3f, the calculated values of m are the same for both AP and BPA (0.76); however, the k₁ of AP is 1.21, roughly double that of BPA (0.63) (Table 1).

3.3. Mono- vs. Co-Degradation, Na₂SO₄ vs. River Water, and pBQ Concentration Variation

Interestingly, the performance of AP single (mono) degradation was lower than that of BPA, although the opposite trend was shown for their co-degradation at 0.5 A cm⁻² in 1 M Na₂SO₄ solution (Figure 4a). As a result, the k value of mono-degradation was smaller for AP (3.50 × 10⁻³ s⁻¹) than for BPA (6.43 × 10⁻³ s⁻¹); moreover, these values were greater than their corresponding data of co-degradation. This result is probably related to the hydrophilic/phobic property of the electrode surface. It is well-known that the surface of the BDD layer is generally hydrophobic with H-terminated functional groups (usually =CH₂ and ≡C–H groups), although hydroxyl, carbonyl, carboxylic, or other functional groups may be formed after use or anodic polarization [23,24]. The hydrophobic surface of the BDD layer, not favoring adsorption of polar molecules [25], favored BPA more than AP adsorption to undergo electrochemical reactions, because BPA has a K_ow (octanol/water partition coefficient) (10^3.4) much greater than AP (2.88 on average) [26,27]. Moreover, BPA has more unsaturated bonds (e.g., C=C) than AP, which are preferred by •OH attack; also, BPA has two aromatic rings, while AP only has one, revealing that the former has more sites to form hydroxylated derivates via •OH oxidation than the latter. Therefore, the mono-degradation performance was better for BPA than for AP. Nevertheless, the k value of AP mono-degradation (3.50 × 10⁻³ s⁻¹) was higher than that of our earlier study using a Ti/PbO₂ anode (1.55 × 10⁻³ s⁻¹) [11], because the oxygen evolution potential of BDD may be ~0.4−0.6 V higher than that of PbO₂ [22].

On the other hand, the electrochemical degradation of target compounds is usually limited by mass transfer (i.e., the diffusion of AP or BPA in the diffusion layer near the electrode surface) [9,14,22]. At 25 °C, the diffusion coefficient of AP (6.90 × 10⁻¹⁰ cm² s⁻¹ [28]) is greater than that of BPA (5.08 × 10⁻¹⁰ cm² s⁻¹ [29]). Therefore, AP was more competitive at electrode surface active sites for direct and indirect oxidation (•OH consumption) than BPA (indirect oxidation only), and thus the co-degradation performance was better for AP than for BPA, although the •OH from water electrolysis was largely responsible for the destruction of most organic chemicals during the electrochemical process [9]. The better performance for mono- than for co-degradation of the same target compound can also be explained by more active sites available for mono- than for co-degradation, because the electrochemically active surface area (ECSA) is proportional to active sites associated with reaction activity [30]. The TOC removal performance of AP or BPA mono-degradation was also better than that of their co-degradation, and the TOC degradation was better for BPA mono-degradation than for AP mono-degradation (Figure 4b), so the magnitude of k values took the order BPA mono-degradation (2.96 × 10⁻³ s⁻¹) > AP mono-degradation (1.89 × 10⁻³ s⁻¹) > AP+BPA co-degradation (5.75 × 10⁻⁴ s⁻¹) (Table 1).

Although the t₁₀₀ values at 0.50 A cm⁻² for AP and BPA co-degradation were both 30 min in the river water (RW) matrix (spiked with 1 M Na₂SO₄) (Figure 5a), the same as those in 1 M Na₂SO₄ (Figure 4a), their corresponding k values were 1.34 × 10⁻³ and 7.00 × 10⁻⁴ s⁻¹, respectively, smaller than those in 1 M Na₂SO₄ (Table 1). Also, the k value of TOC removal was greater in 1 M Na₂SO₄ (5.75 × 10⁻⁴ s⁻¹) than in RW (3.92 × 10⁻⁴ s⁻¹). This result is chiefly related to the inherent TOC (10 mg L⁻¹) in RW, which led to a slightly higher initial TOC concentration in RW (TOC: 10 + AP + BPA = 10 + ~64 + ~79 = ~153 mg L⁻¹) than in 1 M Na₂SO₄ (TOC: AP + BPA = ~143 mg L⁻¹) (Table 1). This phenomenon was also observed for BPA (75 °C) or acesulfame (25 °C) electrochemical degradation on BDD in a RW matrix [12,18]. Additionally, the possible presence of dissolved organic matter in RW might lower the TOC degradation by •OH oxidation.

Formation and degradation tests of para-benzoquinone (pBQ) during AP and BPA mono- and co-degradation were conducted in this study because some researchers indicated that in single AP or BPA electrolysis, pBQ was an intermediate [9,11,12,31], which is a recognized emerging micro-contaminant with chronic toxicity to plants and animals [32]. Figure 6a shows the plots of normalized pBQ (an intermediate of AP or BPA electrochemical degradation) with time for electrochemical co-degradation of target compounds at different current densities. Clearly, the time (t_max) to reach the pBQ maximum concentration (C_max) decreased but the pBQ formation/degradation rate (R_f/R_d) increased with the increase in current density, similar to the trend of AP or BPA elimination in their electrochemical co-degradation (Figure 3a). Likewise, the t_max was shorter and R_f/R_d was greater for AP mono-degradation than for BPA mono-degradation and for mono- vs. co-degradation (Figure 6b), similar to the tendency shown in Figure 4a,b. The degradation of pBQ intermediate during electrochemical AP and BPA mono- and co-degradation suggests the subsequent formation of diacid intermediates from a pBQ ring-opening reaction (see more discussion in Section 3.5).

The precursor for forming pBQ is usually para-hydroquinone (pHQ) [9,32], while the products of pBQ degradation can be diacids [33] (see more discussion in Section 3.5). For simplification, the pBQ formation/degradation mechanism along with equations is assumed as follows:

$$\mathrm{Precursor}\stackrel{{\mathrm{k}}_2}{\to }\mathrm{pBQ}\stackrel{{\mathrm{k}}_3}{\to }\mathrm{Product}$$

(13)

$${\displaystyle \frac{\mathrm{d}\left[\mathrm{pBQ}\right]}{\mathrm{dt}}}={\mathrm{k}}_2\left[\mathrm{Precursor}\right]–{\mathrm{k}}_3\left[\mathrm{pBQ}\right]$$

(14)

where k₂ and k₃ are the rate constants of pBQ formation and degradation, respectively. The pBQ concentration ([pBQ]) (varying with time) is assumed to be described as the following equation based on a simplified double exponential decay model [34]:

$$\left[\mathrm{pBQ}\right]=\mathrm{Z}\left({\mathrm{e}}^{-{\mathrm{k}}_2\mathrm{t}}–{\mathrm{e}}^{-{\mathrm{k}}_3\mathrm{t}}\right)$$

(15)

where Z is a constant related to k₂ and k₃. This model was examined by fitting the data of 0.50 A cm⁻² in 1 M Na₂SO₄ and RW from Figure 6a and Figure 6b, respectively. The data-fitting curves in these figures are shown in pink, which approximately match these data with R² = 0.8876 and 0.8952, respectively (

Table 2. Formation and degradation constants of pBQ (k₂ and k₃, respectively, s⁻¹) for AP+BPA co-degradation at 0.50 A cm⁻² in 1 M Na₂SO₄ (SS) and RW and those of data fitting (*: BPA degradation at 0.05 A cm⁻² on Ti/BDD) taken from [9].

K2			k3
1 M SS	RW	0.1 M SS *	1 M SS	RW	0.1 M SS *
1.38 × 10−3	1.38 × 10−3	2.87 × 10−4	1.42 × 10−3	1.40 × 10−3	5.42 × 10−4
This study	This study	[9]	This study	This study	[9]

). The obtained k₂ and k₃ were 1.38 × 10⁻³ and 1.42 × 10⁻³ s⁻¹, respectively, in 1 M Na₂SO₄, while those in RW were 1.38 × 10⁻³ and 1.40 × 10⁻³ s⁻¹, respectively (Table 2). A better data-fitting result (R² = 0.9534) was observed when using this model to fit the data from BPA electrochemical degradation on Ti/BDD at 0.05 A cm⁻² in 0.1 M Na₂SO₄ [9]; the obtained k₂ and k₃ were 2.87 × 10⁻⁴ and 5.42 × 10⁻⁴ s⁻¹, respectively (Table 2). In fact, the precursor or product also underwent consecutive reaction, which can partially account for the discrepancy between the fitting curve and data.

3.4. UV-Vis and EEM Analyses of Target Compounds in Electrolysis

Before electrolysis, the UV-vis spectrum of AP+BPA in 1.0 M Na₂SO₄ showed three absorbance peaks at 227, 245, and 280 nm (Figure 7). Among these bands, the peak at 245 nm was attributed to the adsorption of UV light by AP, close to the typical peak at 243 nm due to a n → π* electronic transition of the C=O group [35]. The two absorbance peaks at 227 and 280 nm were associated with BPA, similar to those documented in the literature [9,31]. Agrahari et al. [32] indicated that in the BPA π-conjugated system, the delocalization of O atoms lone pairs resulted in the π–π* peak at ~228 nm, while the HOMO and LUMO molecular orbitals extending over the phenyl moieties led to the characteristic UV absorption at λ_max = 280 nm.

During the co-electrolysis of AP+BPA (100 mg/L each) at 0.50 A cm⁻² in 1.0 M Na₂SO₄, the intensities of the aforementioned three absorbance peaks decreased with the increase in electrolysis time, and these peaks disappeared at 30 min, reflecting the disappearance of the AP structure and the opening of aromatic rings of BPA, in accordance with the results shown in Figure 3a,b.

Excitation–emission matrix (EEM) fluorescence spectroscopy was also adopted to analyze the electrolysis process in this study. Prior to electrolysis, no peak was found for the 1.0 M Na₂SO₄ solution (Figure 8a), while RW displayed two fluorescence peaks at regions V (E_x = 300–350 nm and E_m = 400–425 nm) and III (E_x = 230–250 nm and E_m = 400–450 nm) in EEM spectra (Figure 8b), which were also associated with the humic-acid-like and fulvic-acid-like geochemical analogs [23,36]. The EEM spectra of RW spiked with AP+BPPA and 1.0 M Na₂SO₄ exhibited two additional peaks at E_x 270–285 nm/E_m 300–325 nm and E_x 220–235 nm/E_m = 300–325 nm in domains IV and I, respectively (Figure 8c, 2.5–30 min), which resulted from the aromatic structures of the two target compounds [25,31,36]. Note that the intensities of these fluorescence peaks lowered gradually with an increasing electrolysis time, and these peaks disappeared after 30 min electrolysis, revealing a breakage of all the aromatic rings, consistent with the results of the UV-vis analysis.

3.5. Pathways of Electrochemical Degradation of Target Compounds

The hydroxyl radicals in aqueous solutions may attack organic molecules via three mechanisms: (i) the dehydrogenation or abstraction of a hydrogen atom to form water, (ii) the hydroxylation or electrophilic addition to a nonsaturated bond, and (iii) electron transfer or redox reactions [37]. Accordingly, Figure 9 shows the possible pathways of AP mono-degradation (see LC-MS spectra in Figure S1, Supplementary Materials), and

Table 3. Mass spectra data and structures of intermediates identified by LC–MS ([M+H]⁺) for electrochemical degradation of acetaminophen (AP) (#: analysis using HPLC; –: not available).

Compound	Formula	Structure	m/z	RT (min)
Acetaminophen (AP)	C8H9NO2		152.15	4.71
N-acetyl-p-benzoquinone imine (NAPQI)	C8H7NO2		150.13	4.75
Hydroxyacetaminophen	C8H9O3		168.18	4.68
Aminohydroquinone	C6H7NO2		126.31	4.68
Benzoquinone	C6H4O2		109.29	14.13
Hydroquinone	C6H6O2		111.04	10.52
Acetamide	C2H5NO		60.10	10.52
p-Aminophenol	C6H7NO		109.94	21.46
4-aminobutanoic acid	C4H9NO2		104.23	10.85
Malonic acid	C3H4O4		104.76	1.46
Oxalic acid	C2H2O4		90.93	4.75
Citric acid	C6H8O7		192.70	4.70
Oxamic acid	C2H3NO3		90.04	21.34
Formic acid #	HCOOH		–	–

lists the mass spectra data of these intermediates. In brief, AP could be transformed to N-acetyl-p-benzoquinone imine (NAPQI) by hydrogen abstraction or hydroxyacetaminophen by hydroxylation. Serna-Galvis et al. [13] also detected NAPQI in AP electrochemical degradation on Ti/IrO₂ in 0.05 M NaCl solution, which electrochemically generated reactive chlorine species. In addition to AP, NAPQI and hydroxyacetaminophen could be degraded into acetamide, hydroquinone, or p-aminophenol. Acetamide was mineralized into CO₂, H₂O, NH₄⁺→NO₂⁻→NO₃⁻. Through dehydrogenation, hydroquinone was oxidized into benzoquinone (e.g., pBQ [11,12,13]) and vice versa (hydrogenation) [9,14,15], which was further degraded with ring cleavage to produce carboxylic acids: formic, citric, malonic, and oxalic acid. On the other hand, the hydroxylation of p-aminophenol generated aminohydroquinone, and then its benzene ring was also broken down to be further oxidized into 4-aminobutanoic acid or oxamic acid. Lastly, these diacids could be degraded into mono-acids and/or mineralization proceeded. Pacheco-Alvarez et al. [2] indicated that the initial degradation of AP by •OH in an advanced oxidation process (AOP) probably involved in three pathways: (i) the release of the acetamide group with hydroxylation of the remaining benzene moiety to form hydroquinone followed by pBQ; (ii) the direct hydroxylation of the benzene ring followed by consecutive conversion into dicarboxylic acids; and (iii) the release of the acetyl group to form p-aminophenol.

Table 4. Mass spectra data and structures of intermediates identified by LC–MS (([M+H]⁺) or [M-H]⁻ marked by *) for electrochemical degradation of bisphenol A (BPA) (#: analysis using HPLC; –: not available).

Compound	Formula	Structure	m/z	RT (min)
Bisphenol A (BPA)	C15H16O2		229.14	14.97
BPA tricatechol	C15H16O5		277.04	15.64
BPA catechol	C15H16O3		245.10	14.08
p-hydroxybenzophenone	C13H10O2		198.98	12.11
Hydroquinone *	C6H6O2		108.89	26.47
Benzoquinone *	C6H4O2		107.25	9.66
Isopropylphenol	C9H12O		137.09	16.17
Isopropenylphenol	C9H10O		135.20	16.17
Tricarballylic acid	C6H8O6		176.93	15.96
Acetic acid	CH3COOH		61.19	1.62
Tartaric acid *	C4H6O6		148.73	5.69
Oxalic acid	C2H2O4		91.17	12.07
Citric acid	C6H6O7		193.01	1.29
Formic acid #	HCOOH		–	–
Malic acid #	C4H6O5		–	–

provides the mass spectra data of BPA mono-degradation intermediates, which are used to propose the probable degradation pathways (Figure 10, with LC-MS spectra in Figure S2). During electrolysis, BPA was initially degraded into BPA catechol by hydroxylation and p-hydroxybenzophenone via demethylation and hydroxylation. Further hydroxylation of the BPA catechol led to the formation of BPA tricatechol. BPA or BPA tricatechol underwent isopropylidene bridge cleavage to form unstable isopropylphenol and isopropenylphenol, which might be broken down along with hydroxylation into hydroquinone. Likewise, p-hydroxybenzophenone might also be degraded into hydroquinone, which then could be converted into benzoquinone via dehydrogenation and vice versa (hydrogenation) [9,14,15]. Benzoquinone was further degraded with aromatic ring opening to produce dicarboxylic acids (citric, malic, oxalic, tartaric, and tricarballylic acid) and mono-acids (formic and acetic acid). Finally, these diacids could be degraded into mono-acids, which subsequently underwent mineralization through decarboxylation into carbon dioxide and water. Parts of the electrochemical BPA mono-degradation routes are similar to those operated at 0.05 A cm⁻² on Ti/BDD in 0.1 M Na₂SO₄ solution [9]; furthermore, the BPA catechol intermediate was also documented for its electrochemical degradation at 0.05 A cm⁻² on NeoCoat Nb/BDD in 0.05 M Na₂SO₄ solution [38].

The intermediates and pathways of AP+BPA electrochemical co-degradation are given in

Table 5. Mass spectra data and structures of intermediates identified by LC–MS ([M+H]⁺) for electrochemical co-degradation of AP+BPA (#: analysis using HPLC; –: not available).

Compound	Formula	Structure	m/z	RT (min)
Acetaminophen	C8H9NO2		152.15	4.83
Bisphenol A	C15H16O2		229.08	4.97
BPA tetracatechol	C15H16O6		293.01	11.98
BPA tricatechol	C15H16O5		277.02	13.01
BPA catechol	C15H16O3		245.14	14.02
Isopropylphenol	C9H12O		137.19	13.03
Isopropenylphenol	C9H10O		135.40	13.03
4-Hydroxycatechol	C6H6O3		127.25	10.56
Aminohydroquinone	C6H7NO2		126.19	8.91
Hydroquinone	C6H6O2		110.97	10.56
p-Aminophenol	C6H7NO		109.93	13.00
Benzoquinone	C6H4O2		109.09	15.96
Phenol	C6H6O		95.20	16.08
Oxalic acid	C2H2O4		91.13	12.10
Citric acid	C6H6O7		192.95	11.42
Oxamic acid	C2H3NO3		90.12	14.77
Formic acid #	HCOOH		–	–

and Figure 11 (LC-MS spectra in Figure S3), respectively, similar to the combination of pathways from AP and BPA mono-degradation. However, additional intermediate species such as phenol, 4-hydroxycatechol, and BPA tetracatechol were detected, while acetamide and a few aliphatic acids did not appear. Therefore, in addition to the transformation routes of AP mono-degradation, it is also proposed that AP could initially have been broken down to form phenol, which was then oxidized to hydroquinone by hydroxylation. It is likely that hydroquinone and benzoquinone entered into mutual conversion and subsequently underwent ring cleavage with further oxidation by •OH to produce aliphatic acids. On the other hand, the BPA derivatives (BPA tetra-, tri-, and catechol) might have been degraded to generate hydroxylated phenol derivatives (e.g., 4-hydroxycatechol), which were then converted into aliphatic acids, followed by mineralization [9].

4. Conclusions

In this study, direct and indirect oxidation occurred for AP electrochemical degradation during electrolysis, while that for BPA only proceeded via indirect oxidation. The indirect oxidation of AP and BPA mainly relied on hydroxyl radicals electrochemically generated in acidic anolytes during electrolysis. Moreover, competition between AP and BPA for active sites was observed with their co-presence in the electrolyte.

The k of electrochemical AP, BPA, or TOC degradation increased but MCE decreased with an increasing current density. Additionally, the effect of an increasing current density on the increase in k was the same for AP and BPA. However, the performance of AP mono-degradation was lower than that of BPA in 1 M Na₂SO₄ solution, while the opposite trend was observed for their co-degradation. TOC removal was also better for BPA mono-degradation than for AP mono-degradation, while the performance of TOC degradation was better for mono-degradation than for co-degradation. The k value of electrochemical AP, BPA, or TOC degradation was also smaller in a river water matrix than in 1 M Na₂SO₄ solution.

In EEM examination, the fluorescence peak intensity gradually lowered with the increase in electrolysis time, and these peaks disappeared after 30 min of electrolysis, revealing the breakage of all the aromatic rings, consistent with the results of the UV-vis analysis. The pathways of electrochemical AP+BPA co-degradation were similar to the combination of pathways from AP and BPA mono-degradation, although a few intermediates were different. The formation and degradation rate constants of pBQ were obtained by fitting data using the simplified double exponential decay model.