Comparative Study of UV-Based AOPs for Degradation of Hydrophilic Ribavirin and Hydrophobic Chloroquine Phosphate: Performance, Radical Pathways, E_EO, and Water Matrix Effects

Abstract

Ribavirin (RBV) and chloroquine phosphate (CQP) are two typical pharmaceutical contaminants with distinct hydrophilic (RBV) and hydrophobic (CQP) properties. This polarity contrast led to markedly different degradation behaviors. Interestingly, hydrophobic CQP consistently degraded faster and with lower E_EO than hydrophilic RBV across all combined systems, highlighting pollutant polarity as a key determinant. This study systematically compared their degradation by three UV-based advanced oxidation processes (UV/AOPs): UV/H₂O₂, UV/PMS, and UV/KMnO₄. Degradation kinetics, electrical energy per order (E_EO), radical pathways, and water matrix effects were investigated. Sole UV or sole oxidant achieved negligible removal (<7.2%). All UV/AOPs greatly enhanced degradation in a dose-dependent manner. At equal molar oxidant concentration (0.2 mM), the efficiency order was UV/PMS > UV/H₂O₂ ≫ UV/KMnO₄, with the gap widening at higher dosages. UV/H₂O₂ exhibited the best overall performance, with remarkably low E_EO values (0.59 kWh/m³ for RBV and 0.46 kWh/m³ for CQP at 0.2 mM), whereas UV/PMS showed faster kinetics but much higher energy consumption (e.g., 28.67 kWh/m³ for RBV and 28.60 kWh/m³ for CQP at 0.4 mM) and secondary pollution risks. UV/KMnO₄ had low energy but poor degradation. Radical quenching experiments revealed that in UV/H₂O₂, hydroxyl radicals (•OH) predominantly drove degradation regardless of pollutant polarity. In UV/PMS, •OH primarily drove RBV degradation, while CQP removal involved the combined action of •OH, sulfate radicals (SO₄•⁻), and other reactive species. For the optimal UV/H₂O₂ process, acidic pH (5.0) favored degradation; Cl⁻ slightly promoted CQP removal but inhibited RBV, whereas SO₄²⁻, CO₃²⁻, and HCO₃⁻ suppressed both pollutants. Collectively, UV/H₂O₂ is recommended as the most energy-efficient and robust UV/AOP for treating both hydrophilic and hydrophobic pharmaceuticals, with the additional insight that pollutant polarity governs both degradation kinetics and radical mechanisms.

1. Introduction

Emerging contaminants include pharmaceuticals and personal care products (PPCPs), persistent organic pollutants (POPs), endocrine-disrupting chemicals (EDCs), per- and polyfluoroalkyl substances (PFAS), and others. These compounds generally exhibit high environmental stability and can pose hazards to soil, water bodies, atmosphere, and human health through natural and social cycles [1]. At present, emerging contaminants enter wastewater systems via human metabolism and other pathways, imposing a heavy burden on existing wastewater treatment facilities. Their proportion in organic pollutants from wastewater treatment plants (WWTPs) worldwide has been increasing, and in China, their share can reach up to 21% [2]. Ribavirin is a common antiviral drug with in vitro activity against a wide spectrum of viruses. It is used to treat various viral infections and has also been applied in the therapy of colorectal cancer and hepatocellular carcinoma [3,4,5]. Chloroquine phosphate (CQP) is an antiviral agent used against coronavirus disease and is also a long-established drug for malaria, rheumatoid arthritis, and asthma. Accordingly, ribavirin (RBV) and CQP are extensively consumed in daily life and represent two typical PPCPs [6,7]. Both RBV and CQP possess high persistence and biotoxicity. For instance, RBV has been found to exert significant teratogenic and embryocidal effects in nearly all animal studies and shows high delayed toxicity toward Spodoptera litura, a destructive agricultural pest. Its discharge into aquatic environments can lead to severe water pollution and viral drug resistance. CQP released into wastewater can affect biomass activity and process stability, including anaerobic ammonia oxidation, in biological treatment systems [8,9,10,11,12,13]. Excessive amounts of RBV, CQP, and other pharmaceuticals enter wastewater systems through human excretion. Owing to their high structural stability, the large-scale consumption of these compounds has placed a substantial burden on conventional wastewater treatment processes [14,15,16]. RBV and CQP have been detected in urban wastewater, surface water, groundwater, and sediments at concentrations ranging from ng/L to μg/L [17,18,19]. Uncontrolled accumulation of RBV, CQP, and their metabolites in aquatic environments would exert toxic effects on algae, fish, and humans, posing elevated risks to water safety [4,16,20,21]. Wastewater treatment plants serve as critical hubs for water treatment; however, traditional processes show limited removal efficiency toward RBV and CQP. Moreover, RBV is hydrophilic, while CQP displays hydrophobic characteristics, making the two compounds ideal candidates for comparative investigation into the degradation behavior of pharmaceuticals with distinct physicochemical properties [22,23]. Therefore, it is necessary to develop novel methods or optimize existing technologies to enhance their elimination [14,16,21].

Photochemical oxidation processes feature mild reaction conditions, simple operation, fast pollutant degradation, short reaction time, and high removal efficiency. They have been widely applied in water and wastewater treatment and show promising development prospects. The main mechanism of ultraviolet (UV)-based photochemical oxidation systems involves the addition of oxidants such as H₂O₂, peroxymonosulfate (PMS), peroxydisulfate (PDS), NaClO, and KMnO₄ into solution. Under UV irradiation, these substances are activated to generate strong oxidizing radicals or species, including hydroxyl radicals (•OH), sulfate radicals (SO₄•⁻), chlorine radicals (Cl•), and Mn(V). These reactive species undergo substitution, addition, electron transfer, and other reactions with organic pollutants, leading to their partial or complete mineralization and removal [16,24]. H₂O₂ is a commonly used oxidant in advanced oxidation processes (AOPs) [25]. It is chemically unstable and readily decomposes, and its UV-activated oxidation produces only water and oxygen, avoiding secondary pollution. H₂O₂ is fully miscible with aqueous solutions and exhibits high safety, making it suitable for degrading various organic pollutants [26,27]. PMS is an emerging oxidant in AOPs that decomposes into water and sulfate. Upon activation, it generates reactive species such as •OH and SO₄•⁻, enabling efficient pollutant degradation. PMS is convenient, safe, and stable, with good water miscibility and high security [28,29]. The degradation of pollutants by PMS-activated radicals highly depends on the structural active sites of catalysts [30]. PMS can be effectively activated by UV, Fe²⁺, various catalysts, and their combinations, significantly improving the removal of organic pollutants [28,29,31]. KMnO₄ is widely employed as an oxidant in water treatment due to its easy dosing, convenient storage and transportation, strong oxidation capacity, and absence of toxic disinfection byproducts during oxidation. It is used to remove iron, manganese, and other metals; eliminate algae; control odor; and degrade emerging organic compounds in water [32,33]. Compared with other oxidants, KMnO₄ generates fewer byproducts during pollutant treatment, and its final reduction product in water treatment is usually MnO₂, which exhibits adsorption and coagulation-aid effects, providing synergistic benefits for pollutant removal [32,33]. The UV/KMnO₄ system is a typical UV-driven photochemical oxidation process and an emerging technology. Previous studies have investigated the degradation of nitrobenzene, benzoic acid, terephthalic acid, p-chlorobenzoic acid (as probe compounds), gemfibrozil, and nalidixic acid (as representative micropollutants). Results have verified a distinct synergistic effect in the UV/KMnO₄ system, which exhibited significantly enhanced degradation kinetics compared with UV alone. Such improvement is mainly attributed to the UV-induced activation of KMnO₄, generating Mn(V) and hydroxyl radicals (•OH) that contribute to the efficient oxidation and decomposition of target contaminants [34].

Although UV-based AOPs have been individually applied to various pharmaceuticals, most existing studies treat target pollutants in isolation without systematically comparing how intrinsic molecular properties—particularly hydrophilicity/hydrophobicity—affect degradation behavior across different AOPs. The role of pollutant polarity in determining not only degradation kinetics but also energy efficiency (E_EO) and dominant radical pathways remains largely unexplored. Furthermore, the selection of suitable UV/AOPs for pharmaceutical wastewater still lacks a clear, property-based guiding principle. To fill this gap, RBV (logKow = −1.85) and CQP (logKow = 3.89) serve as two representative antiviral pharmaceuticals with sharply contrasting polarities, making them ideal candidates for elucidating the polarity-dependent mechanisms that govern UV/AOP performance [22,23].

In this work, hydrophilic ribavirin (RBV) and hydrophobic chloroquine phosphate (CQP) were selected as model pollutants. The degradation performance and electrical energy per order (E_EO) of three UV-based AOPs (UV/H₂O₂, UV/PMS, and UV/KMnO₄) were systematically compared at equivalent molar oxidant dosages. Particular attention was paid to how pollutant hydrophilicity/hydrophobicity affects degradation kinetics and energy consumption. Radical quenching experiments were designed to identify the dominant reactive species responsible for pollutant removal. Furthermore, the effects of key water quality parameters (pH, common anions) on the most energy-efficient process were investigated. The findings provide both practical guidance for selecting suitable UV/AOPs in treating antiviral pharmaceutical wastewater and mechanistic insights into polarity-dependent radical pathways.

2. Materials and Methods

2.1. Chemicals

RBV (C₁₀H₁₃N₅O₄), CQP (C₁₈H₃₂ClN₃O₈P₂), disodium hydrogen phosphate dodecahydrate (Na₂HPO₄·12H₂O), sodium dihydrogen phosphate (NaH₂PO₄), potassium permanganate (KMnO₄), ammonium acetate (CH₃COONH₄, HPLC grade), tert-butanol (C₄H₁₀O, chromatographically pure), isopropanol (C₃H₈O, chromatographically pure), and ethanol (CH₃CH₂OH, chromatographically pure) were purchased from Aladdin Reagent (Shanghai, China). Methanol (CH₃OH) and acetonitrile (CH₃CN) were obtained from Merck KGaA (Darmstadt, Germany). Sodium chloride (NaCl), anhydrous sodium sulfate (Na₂SO₄), sodium bicarbonate (NaHCO₃), sodium carbonate (Na₂CO₃), sodium hydroxide (NaOH), hydrochloric acid (HCl), and hydrogen peroxide (H₂O₂) were acquired from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Potassium peroxymonosulfate (PMS, KHSO₅·0.5KHSO₄·0.5K₂SO₄) was purchased from Ron Reagent (Shanghai, China). All solutions were prepared with ultrapure water (18.2 MΩ·cm) produced by a Milli-Q system (MilliporeSigma, Guyancourt, France).

2.2. Experimental Design and Methods

The experimental setup employed in this study was a Trojan parallel beam UV reactor, and the UV fluence rate was accurately determined using a UV radiometer. The typical UV fluence rate applied was 206 μW/cm², as detailed in Figure S1 in the Supporting Information. Stock solutions including KMnO₄, PMS, H₂O₂, RBV, CQP, 100 mM disodium hydrogen phosphate (Solution A), and 100 mM sodium dihydrogen phosphate (Solution B) were prepared separately. Each stock solution was stirred for 24 h to ensure complete dissolution before use. All stock solutions were stored at 4 °C and renewed weekly. All experiments were conducted at room temperature of 25 ± 2 °C. Before each experiment, the UV lamp was preheated for 30 min to ensure stable UV radiation.

A 5 mM PBS buffer solution was prepared by mixing appropriate volumes of Na₂HPO₄ and NaH₂PO₄ stock solutions. The solution pH was adjusted using NaOH or HCl if any fluctuation occurred. RBV, CQP, and corresponding oxidant stock solutions (PMS, H₂O₂, KMnO₄) were added into the buffered solution in sequence. After mixing, the solution was transferred into the reactor and exposed to UV irradiation aligned with the cylindrical chamber of the parallel beam apparatus, initiating the UV/AOPs degradation process. Samples were collected at predetermined time intervals for analysis.

The degradation of RBV and CQP was monitored by high-performance liquid chromatography (HPLC). All degradation experiments were performed in duplicate (n = 2). For quantitative analysis of RBV via high-performance liquid chromatography (HPLC), the mobile phase was a mixture of methanol/water (10:90, v/v). The UV detection wavelength was set at 208 nm, using a C18 column (4.6 mm × 250 mm, 5 μm). The flow rate was 0.8 mL min⁻¹, and the column temperature was maintained at 35 °C. For CQP quantification, the mobile phase consisted of methanol/acetonitrile/0.1 mM ammonium acetate (40:25:35, v/v/v). The detection wavelength was 325 nm, using the same C18 column (4.6 mm × 250 mm, 5 μm). The flow rate was 1.0 mL min⁻¹ with a column temperature of 35 °C. Calibration curves for both RBV and CQP were established with excellent linearity (R² = 0.9999 for both analytes). The LOD and LOQ were approximately 0.1 μM and 0.3 μM for both compounds, well below the experimental concentration of 2.5 μM, ensuring reliable quantification. Recovery rates from spiked samples ranged from 95% to 105%, with RSD < 5%. Prior to HPLC analysis, samples were quenched by adding excess methanol. Control experiments confirmed that none of the oxidants (H₂O₂, PMS, KMnO₄) or radical scavengers (EtOH, IPA, TBA) at the concentrations used produced interfering peaks at the analytical wavelengths for RBV (208 nm) or CQP (325 nm).

2.3. Date Analysis

The UV-based photochemical oxidation process in this study is a homogeneous reaction, in which oxidation occurs in the aqueous phase. The UV light intensity was kept constant during the experiments, so the generated strong oxidizing radicals could be regarded as being produced continuously and uniformly. However, these oxidative radicals exhibit extremely short lifetimes in water and cannot be directly quantified. Therefore, the pseudo-first-order kinetic model was generally adopted to describe the degradation process, as shown in Equation (1):

−In(C/C₀) = kct

(1)

where

• C is the concentration of RBV or CQP, mmol·L⁻¹;
• t is the reaction time, min;
• k_C is the apparent pseudo-first-order rate constant, min⁻¹.

Moreover, the electrical energy per order (E_EO), a standard measure commonly used for operating cost assessment, was introduced to evaluate the unit energy consumption of the various UV/AOPs for RBV and CQP removal [35]. In this study, E_EO-UV was calculated based on the effective UV radiant power reaching the reaction solution (measured by a UV radiometer), rather than the total electrical input power of the UV lamp, as this approach normalizes the UV fluence input for a fair inter-process comparison in a laboratory-scale collimated beam apparatus. E_EO-OX is defined as an auxiliary equivalent energy indicator for cost-weight estimation of oxidant consumption; electrical energy consumption and chemical cost should be evaluated separately in practical applications. The detailed calculations are presented in Supplementary Material Text S1.

3. Results

3.1. Comparative Analysis of RBV and CQP Degradation by UV/H₂O₂, UV/PMS and UV/KMnO₄

H₂O₂, PMS, and KMnO₄ are common oxidants. This study compared the degradation of RBV and CQP in sole UV, sole oxidant, and UV/AOP systems at equivalent molar concentrations (Figure 1). Pseudo-first-order rate constants (kc) are given in Tables S2 and S3. Under sole UV, RBV removal was only 7.2% (45 min), and CQP removal was 6.8% (30 min). Sole oxidants (0.2 mM H₂O₂, 0.2 mM PMS, or 0.02 mM KMnO₄) gave negligible removal (<5.5%), because unactivated oxidants cannot generate reactive oxygen species without light.

UV/AOPs dramatically enhanced degradation in a dose-dependent manner. For RBV in UV/H₂O₂ (0.02 → 0.4 mM H₂O₂), removal increased from 9.5% to 79.5% (kc 0.0021 → 0.0347 min⁻¹). In UV/PMS (0.02 → 0.4 mM PMS), RBV removal rose from 10.2% to 86.8% (kc 0.0022 → 0.0444 min⁻¹). In UV/KMnO₄ (0.02 → 0.2 mM), RBV removal increased from 9.4% to 27.8% (kc 0.0020 → 0.0070 min⁻¹). For CQP, UV/H₂O₂ gave removal from 13.0% to 76.7% (kc 0.0044 → 0.0482 min⁻¹); UV/PMS from 14.7% to nearly 100% (kc 0.0056 → 0.0659 min⁻¹); UV/KMnO₄ from 10.5% to 40.3% (kc 0.0032 → 0.0136 min⁻¹). No obvious quenching was observed, indicating oxidant dosages were below the radical-scavenging threshold [32,33].

Pollutant polarity significantly influenced degradation. Hydrophobic CQP consistently showed higher kc than hydrophilic RBV. At 0.4 mM H₂O₂, kc(CQP)/kc(RBV) = 1.39; at 0.4 mM PMS, the ratio was 1.48. The absolute kc difference between CQP and RBV widened with oxidant dosage: in UV/PMS at 0.02 mM, the difference was 0.0034 min⁻¹; at 0.4 mM, it reached 0.0215 min⁻¹ (six-fold increase). Thus, the hydrophobic pollutant benefits disproportionately from increased radical flux. The relative advantage also varied among systems: in UV/KMnO₄ at 0.2 mM, the ratio reached 1.94, suggesting that Mn(V) intermediates [34] may favor hydrophobic targets more strongly than •OH or SO₄•⁻.

A horizontal comparison (Figure 1g,h) confirmed the degradation order: UV/PMS > UV/H₂O₂ > UV/KMnO₄ at all tested concentrations (0.02–0.2 mM). At 0.02 mM, removal differences were small (e.g., RBV: 9.4–10.2%; CQP: 10.5–14.7%). As oxidant concentration increased, the gap widened. At 0.2 mM, RBV removals were 27.8% (KMnO₄), 60.8% (H₂O₂), and 63.0% (PMS); CQP removals were 40.3%, 55.3%, and 62.6%, respectively. UV/PMS produced the most reactive species (SO₄•⁻ + •OH), while UV/KMnO₄ suffered from poor activation efficiency. In summary, both oxidant choice and contaminant polarity are key determinants of degradation efficiency in UV/AOPs. To identify the dominant radical species in UV/H₂O₂ and UV/PMS and to interpret their different degradation behaviors, quenching experiments were performed.

3.2. Mechanistic Insights

In this section, the radical quenching method was employed to investigate the main radical species responsible for the degradation of RBV and CQP in UV/H₂O₂ and UV/PMS. UV/KMnO₄ was not further studied because its intense purple color (even at 0.2 mM) makes the solution deeply colored, hindering light penetration and complicating radical identification; moreover, its poor degradation performance already limited practical value. Therefore, mechanistic studies focused only on UV/H₂O₂ and UV/PMS. In UV/H₂O₂ and UV/PMS, possible reactive species include •OH, SO₄•⁻, ¹O₂, and O₂•⁻. To distinguish them, tert-butanol (TBA, quenches •OH but not SO₄•⁻), ethanol (EtOH, quenches both •OH and SO₄•⁻), and isopropanol (IPA, also quenches both •OH and SO₄•⁻) were selected based on their well-known rate constants. All three react very fast with •OH (k ≈ 10⁸–10⁹ M⁻¹·s⁻¹) [36,37]. TBA reacts slowly with SO₄•⁻ (<10⁶ M⁻¹·s⁻¹), while EtOH and IPA show moderate reactivity (k ≈ 7.8 × 10⁷ and 8.5 × 10⁷ M⁻¹·s⁻¹) [38,39]. Their reactions with ¹O₂ are extremely slow (<10³–10⁵ M⁻¹·s⁻¹) and negligible with O₂•⁻ [36,40]. Strong inhibition by TBA, EtOH, or IPA indicates involvement of •OH and/or SO₄•⁻; no inhibition suggests ¹O₂ or O₂•⁻. IPA rapidly quenches both •OH and SO₄•⁻ but not ¹O₂, so strong IPA inhibition excludes ¹O₂ [41,42].

To elucidate the radical pathways, quenching experiments were performed using the UV/H₂O₂ (0.2 mM) and UV/PMS (0.2 mM) processes. The degradation efficiencies of RBV (45 min) and CQP (30 min) were measured upon addition of various quenchers, and the results are summarized in Figure 2 and Table S4 (Supporting Information).

UV/H₂O₂: Without quencher, RBV removal = 60.9% (kc 0.020 min⁻¹), CQP = 55.3% (kc 0.0274 min⁻¹). At 10 mM quencher ([quencher]:[H₂O₂] = 50:1), RBV removal dropped to 5.7% (EtOH), 2.9% (IPA), 5.6% (TBA); CQP to 6.3% (EtOH), 11.1% (IPA), 5.4% (TBA). TBA suppressed both pollutants as strongly as EtOH (e.g., CQP: 5.4% vs. 6.3%), proving that •OH is the predominant radical; other ROS contributions are negligible.

UV/PMS: Without quencher, RBV removal = 63.0% (kc 0.023 min⁻¹), CQP = 62.6% (kc 0.0319 min⁻¹). For RBV, 10 mM TBA reduced removal to 5.4%, and 10 mM EtOH to 4.2%, indicating •OH dominates with minor SO₄•⁻ contribution. For CQP, TBA (10 mM) left 44.8% removal, whereas EtOH (10 mM) left only 22.2%. The much weaker TBA inhibition proves SO₄•⁻ plays a major role. After 10 mM EtOH, residual CQP removal (22.2%) exceeded the sum of direct photolysis (6.8%) and direct PMS oxidation (5.2%) = 12.0%. Given the low reactivity of EtOH/IPA with ¹O₂, non-radical pathways likely contribute to CQP degradation (but the specific involvement of ¹O₂ remains a reasonable inference based on indirect evidence). Thus, in UV/PMS, radical pathways depend on pollutant polarity: hydrophilic RBV is primarily oxidized by •OH in the aqueous phase; hydrophobic CQP, due to its low aqueous solubility and tendency to associate with radical species in solution, undergoes degradation involving the combined action of •OH, SO₄•⁻, other possible radicals, direct photolysis, and direct PMS oxidation.

Mechanistic explanation: •OH acts as “short-range, fast oxidation” (H-abstraction/addition [43]), with near diffusion-controlled rates (10⁹–10¹⁰ M⁻¹·s⁻¹) and minimal steric hindrance, rapidly oxidizing hydrophilic RBV [44,45,46]. However, its half-life is ~20 ns, limiting it to attack only pollutants near its generation site in the aqueous phase [47,48]. SO₄•⁻ acts as “long-range, migratory oxidation” via outer-sphere single electron transfer (SET) [43], favoring electron-rich aromatics (e.g., CQP quinoline ring) [46,49]. Its reaction with hydrophilic substrates is ~10× slower (10⁸–10⁹ M⁻¹·s⁻¹) than •OH, and it suffers more steric hindrance [44,45,46,49]. However, SO₄•⁻ has a much longer half-life (30–40 μs, ~2000× that of •OH), enabling greater diffusion distance [47,48]. These differences in reactivity and diffusion explain why •OH dominates RBV degradation, whereas the degradation of CQP in UV/PMS involves the combined action of •OH, SO₄•⁻, and other reactive species.

3.3. Analysis of Electrical Energy per Order (E_EO)

As shown in Figure 3a,b and Table S5, the electrical energy per order (E_EO) for RBV and CQP degradation by UV/H₂O₂, UV/PMS, and UV/KMnO₄ was analyzed. E_EO consists of UV contribution (E_EO-UV) and oxidant contribution (E_EO-OX).

For UV/H₂O₂, energy was dominated by UV. Increasing H₂O₂ from 0.02 to 0.4 mM reduced E_EO for RBV from 4.28 to 0.51 kWh/m³ and for CQP from 2.05 to 0.44 kWh/m³. CQP (hydrophobic) consistently required lower E_EO than RBV. In contrast, UV/PMS showed an opposite trend: raising PMS from 0.02 to 0.4 mM sharply increased E_EO to 28.67 kWh/m³ for RBV and 28.60 kWh/m³ for CQP, with similar values regardless of polarity. E_EO-OX became dominant, worsening energy efficiency. For UV/KMnO₄, energy was also UV-dominated. Increasing KMnO₄ from 0.02 to 0.2 mM decreased E_EO for RBV from 4.57 to 2.16 kWh/m³, and for CQP from 2.89 to 1.54 kWh/m³ (minimum 1.30 kWh/m³ at 0.05 mM). CQP again showed lower E_EO than RBV.

Overall energy efficiency order: UV/H₂O₂ > UV/KMnO₄ > UV/PMS. UV/PMS had high energy cost, especially at high oxidant concentrations, while UV/H₂O₂ and UV/KMnO₄ were more practical. Hydrophobic CQP always needed less electrical energy than hydrophilic RBV in UV/H₂O₂ and UV/KMnO₄, confirming that intrinsic hydrophobicity favors lower energy input when the oxidant does not dominate the budget.

In summary, hydrophobic CQP consistently exhibited faster degradation and lower E_EO than hydrophilic RBV across all combined systems. This polarity-dependent advantage widened with oxidant dosage, though the extent varied. Based on treatment efficiency per mole, the ranking was UV/PMS > UV/H₂O₂ ≫ UV/KMnO₄; based on E_EO, the order was UV/PMS ≫ UV/KMnO₄ > UV/H₂O₂. UV/PMS achieved fast degradation but suffered from high energy consumption and secondary pollution risks (excess PMS and sulfate can form sulfides under reducing conditions, causing blackening and malodor [50]). UV/KMnO₄ had low operating cost but limited oxidation capacity; excessive KMnO₄ causes water coloration and MnO₂ precipitation, leading to filter clogging and turbidity [33,51,52]. In addition, the deep coloration of KMnO₄ solutions substantially attenuates 254 nm UV radiation, reducing the effective photon fluence available for •OH generation and contributing to the lower degradation efficiency of the UV/KMnO₄ process. In comparison, UV/H₂O₂ offered balanced performance: efficiency slightly lower than UV/PMS, energy efficiency improved with oxidant dosage, and H₂O₂ decomposes into H₂O and O₂ without toxic residues [53,54]. Considering degradation efficiency, energy consumption, environmental risks, and detoxification performance, UV/H₂O₂ shows superior application potential for pharmaceutical pollutant control. The following sections analyze water matrix effects on RBV and CQP degradation by UV/H₂O₂.

3.4. Effects of Different Water Matrix Components

3.4.1. Influences of pH

The influence of pH (5.0–9.0) on RBV and CQP degradation by UV/H₂O₂ was investigated (Figure 4a,b; Table S6). For both pollutants, degradation efficiency decreased with increasing pH. The highest removal was achieved at pH 5.0: RBV 60.7% (kc = 0.0211 min⁻¹, 45 min) and CQP 63.9% (kc = 0.0326 min⁻¹, 30 min). At pH 9.0, removal dropped sharply to 27.0% for RBV (kc = 0.0071 min⁻¹) and 44.5% for CQP (kc = 0.0197 min⁻¹). Thus, acidic conditions favor the UV/H₂O₂ process for both hydrophilic RBV and hydrophobic CQP, though the extent of pH-induced attenuation differs between the two pollutants.

Based on advanced oxidation principles and previous studies, the high degradation efficiency of the UV/H₂O₂ system under acidic conditions can be attributed to two main factors. First, an acidic environment effectively stabilizes the molecular structure of H₂O₂, reducing its unproductive decomposition under UV irradiation and thus ensuring the sustained generation of •OH [55]. Second, the quenching of •OH is notably inhibited under acidic conditions, and both RBV and CQP exist mainly in molecular form, making them more susceptible to oxidation by •OH [56]. As pH increases to neutral and alkaline ranges, H₂O₂ tends to undergo self-dismutation or react inefficiently with OH⁻, leading to a substantial reduction in •OH production [57]. Meanwhile, the molecular forms of RBV and CQP change (e.g., dissociate into ionic species), resulting in lower reactivity toward •OH and ultimately a marked decrease in degradation efficiency [55,58].

3.4.2. Influences of Anions

The effects of common anions (Cl⁻, SO₄²⁻, CO₃²⁻, HCO₃⁻, each at 1 mM) on RBV and CQP degradation by UV/H₂O₂ were investigated (Figure 5, Table S7). For hydrophilic RBV, all anions inhibited degradation: the kc decreased from 0.0195 min⁻¹ to 0.013 (Cl⁻), 0.0095 (SO₄²⁻), 0.0082 (CO₃²⁻), and 0.0068 min⁻¹ (HCO₃⁻). For hydrophobic CQP, Cl⁻ slightly promoted degradation (kc increased from 0.0274 to 0.0279 min⁻¹), whereas the other anions suppressed it (kc: 0.0211 for SO₄²⁻, 0.0234 for CO₃²⁻, 0.0238 for HCO₃⁻). These distinct responses are attributed to the different hydrophilic/hydrophobic properties of the two pollutants.

The influence of Cl⁻ on the UV/H₂O₂ system is particularly complex, as illustrated in Equations (2)–(10) [59,60,61,62,63,64], and can result in either promotion or inhibition of the overall reaction. In this work, Cl⁻ mainly exhibited an inhibitory effect on RBV degradation but a promoting effect on CQP degradation by UV/H₂O₂.

$$•\mathrm{O}\mathrm{H}+\mathrm{C}{\mathrm{l}}^{-}\to \mathrm{C}\mathrm{l}•+\mathrm{O}{\mathrm{H}}^{-}\begin{array}{cc},& \mathrm{k}=3.0\times 10^9 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(2)

$$•\mathrm{O}\mathrm{H}+{\mathrm{Cl}}^{-}\to {\mathrm{ClOH}}^{-}•\begin{array}{cc},& \mathrm{k}=4.3\times 10^9 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(3)

$$\mathrm{C}\mathrm{l}•+\mathrm{O}{\mathrm{H}}^{-}\to \mathrm{C}\mathrm{l}{\mathrm{H}\mathrm{O}}^{-}•\begin{array}{cc},& \mathrm{k}=1.80\times 10^{10} {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(4)

$$\mathrm{C}\mathrm{l}{\mathrm{H}\mathrm{O}}^{-}•\to •\mathrm{O}\mathrm{H}+\mathrm{C}{\mathrm{l}}^{-}\begin{array}{cc},& \mathrm{k}=6.1\times 10^9 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(5)

$${\mathrm{ClOH}}^{-}•\to •\mathrm{O}\mathrm{H}+\mathrm{C}{\mathrm{l}}^{-}\begin{array}{cc},& \mathrm{k}=(6.1\pm 0.8)\times 10^9 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(6)

$${\mathrm{ClOH}}^{-}•+{\mathrm{H}}^{+}\to \mathrm{C}\mathrm{l}•+{\mathrm{H}}_2\mathrm{O}\begin{array}{cc},& \mathrm{k}=(2.1\pm 0.7)\times 10^{10} {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(7)

$$\mathrm{C}\mathrm{l}\mathrm{O}{\mathrm{H}}^{-}•+\mathrm{C}{\mathrm{l}}^{-}\to \mathrm{C}{\mathrm{l}}_2^{-}•+\mathrm{O}{\mathrm{H}}^{-}\begin{array}{cc},& \mathrm{k}=1.0\times 10^5 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(8)

$$\mathrm{C}\mathrm{l}•+\mathrm{C}{\mathrm{l}}^{-}\to \mathrm{C}{\mathrm{l}}_2^{-}•\begin{array}{cc},& \mathrm{k}=(6.5~8.5)\times 10^9 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(9)

$$\mathrm{C}{\mathrm{l}}_2^{-}•+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{H}}^{+}+\mathrm{C}{\mathrm{l}}^{-}+\mathrm{HClO}\begin{array}{cc},& \mathrm{k}=2.1\times 10^9 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(10)

In this study, SO₄²⁻ exerted a partial inhibitory effect on the reaction, which may be attributed to the reactions of SO₄²⁻ as shown in Equations (11) and (12) [65,66]. Sulfate radicals (SO₄$•$⁻) generated during the reaction possess a redox potential (2.5–3.1 V) comparable to that of hydroxyl radicals (•OH, 2.77 V). However, SO₄•⁻ exhibits certain selectivity toward the degradation of organic pollutants, thereby leading to an overall inhibitory effect on the reaction rate [67].

$$•\mathrm{O}\mathrm{H}+{\mathrm{H}}^{+}+{\mathrm{S}\mathrm{O}}_4^{2-}\to {\mathrm{S}\mathrm{O}}_4^{-}•+{\mathrm{H}}_2\mathrm{O}\begin{array}{cc},& \mathrm{k}=(3.5\pm 0.5)\times 10^5 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(11)

$${\mathrm{S}\mathrm{O}}_4^{-}•+\mathrm{O}{\mathrm{H}}^{-}\to {\mathrm{S}\mathrm{O}}_4^{2-}+•\mathrm{O}\mathrm{H}\begin{array}{cc},& \mathrm{k}=6.5\times 10^7 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(12)

In this study, the inhibitory effects of CO₃²⁻ and HCO₃⁻ were particularly pronounced. This phenomenon could be attributed to the formation of CO₃•⁻ and HCO₃• radicals via the reactions depicted in Equations (13) and (14) [61,68,69]. The corresponding redox potentials are E° (CO₃•⁻/CO₃²⁻) = 1.78 V and E° (HCO₃•/HCO₃⁻) = 1.50 V [70], both of which are significantly lower than that of the hydroxyl radical (•OH). Consequently, these radicals exerted inhibitory effects on the degradation of both RBV and CQP by the UV/H₂O₂ system.

$$•\mathrm{O}\mathrm{H}+\mathrm{H}{\mathrm{C}\mathrm{O}}_3^{-}\to \mathrm{O}{\mathrm{H}}^{-}+{\mathrm{H}\mathrm{C}\mathrm{O}}_3•\begin{array}{cc},& \mathrm{k}=(0.85~1.00)\times 10^7 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(13)

$$•\mathrm{O}\mathrm{H}+{\mathrm{C}\mathrm{O}}_3^{2-}\to \mathrm{O}{\mathrm{H}}^{-}+{\mathrm{C}\mathrm{O}}_3^{-}•\begin{array}{cc},& \mathrm{k}=(3.90~4.20)\times 10^8 {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}\end{array}$$

(14)

The distinct response patterns of RBV and CQP to common anions are mainly attributed to their differences in hydrophilicity and hydrophobicity, which lead to variations in their interaction with anions, existing forms in aqueous solution, and reactivity with •OH generated by the UV/H₂O₂ system.

4. Discussion

Hydrophobic CQP consistently outperformed hydrophilic RBV in all systems regarding degradation kinetics and energy consumption. As shown in Table S8, RBV and CQP differ substantially in multiple physicochemical properties beyond polarity, including molecular weight, LogP, aqueous solubility, pKa, and molecular structure. Therefore, the observed kinetic differences should be interpreted as a compound-specific comparison reflecting the combined effect of these factors, with polarity identified as a key contributing factor rather than the sole determinant. Each process exhibited distinct characteristics: UV/PMS prioritized reaction rate but at the cost of high energy input; UV/KMnO₄ featured low energy consumption yet poor degradation capacity, partly due to the strong absorption of KMnO₄ at 254 nm, which reduces effective photon fluence for •OH generation; and UV/H₂O₂ stood out as the most balanced option with both high efficiency and low E_EO.

The consistently faster degradation of hydrophobic CQP compared with hydrophilic RBV can be rationalized by considering the differential accessibility of the two pollutants to reactive radicals in aqueous solution. Hydrophobic organic compounds tend to exhibit stronger affinity toward radical species in aqueous solution. In contrast, highly water-soluble compounds such as RBV remain fully hydrated and dispersed in the bulk solution, reducing their probability of encountering short-lived radicals such as •OH and SO₄•⁻. This polarity-dependent accessibility, which has been discussed in the literature for structurally diverse pharmaceuticals [71], provides a plausible mechanistic explanation for the observed kinetic differences.

Increasing oxidant dosage further amplified the performance gap among the three systems. Radical quenching experiments clarified the mechanistic differences between the two core processes: UV/H₂O₂ relied predominantly on •OH for degradation, unaffected by pollutant polarity, while UV/PMS adopted a pollutant-specific pathway—•OH dominated RBV removal, and CQP degradation depended on the synergy of •OH, SO₄•⁻, and other active species. It should be noted that quenching experiments provide indirect and semi-quantitative evidence for radical contributions, and the results should be interpreted as indicative of dominant radical pathways rather than rigorous quantitative proof. Direct radical quantification methods (e.g., EPR spin trapping) would be valuable for future confirmation. For the optimal UV/H₂O₂ process, solution pH and coexisting anions exerted significant effects: pH 5.0 (acidic condition) was most favorable, and elevated pH inhibited degradation. Cl⁻ slightly promoted CQP but inhibited RBV, whereas SO₄²⁻, CO₃²⁻, and HCO₃⁻ showed universal inhibition. A limitation of this study is that all degradation experiments were conducted in phosphate-buffered solutions. Although the effects of major coexisting anions and pH were systematically examined, the overall performance in real water matrices (e.g., municipal secondary effluent or pharmaceutical wastewater) may deviate due to the complex interplay of natural organic matter and other constituents. Validation with authentic water samples is therefore necessary to confirm the practical applicability of the UV/H₂O₂ process.

As homogeneous processes, UV/H₂O₂ and UV/PMS do not suffer from catalyst deactivation, and their batch-to-batch reproducibility has been widely reported [72,73]. Nevertheless, the long-term operational stability under continuous-flow conditions warrants dedicated investigation before practical deployment.

It should be noted that the removal efficiency of target pollutants does not directly equate to the safety of the treated water. The potential formation of toxic intermediates during UV/AOP degradation has been reported for structurally similar pharmaceuticals [71]. Although the •OH-dominated UV/H₂O₂ process generally exhibits strong mineralization potential, the ecotoxicity evolution along the degradation pathway of RBV and CQP in UV/AOPs remains to be systematically evaluated in future studies. Furthermore, the findings obtained at μM concentrations should not be directly extrapolated to environmentally relevant ng/L levels, as the pollutant-to-scavenger ratio differs substantially between these concentration regimes.

5. Conclusions

This study establishes pollutant polarity as a key factor governing both degradation kinetics and radical mechanisms in UV/AOPs. Hydrophobic pharmaceuticals are inherently more amenable to treatment than their hydrophilic counterparts, and the choice of AOP determines whether the radical pathway is universal (•OH-dominated in UV/H₂O₂) or polarity-dependent (•OH/SO₄•⁻ synergy in UV/PMS). Among the three processes compared, UV/H₂O₂ offers the most practical balance of efficiency, energy consumption, and operational simplicity, with its performance tunable through pH and anion control. These findings provide a polarity-informed framework for selecting UV/AOPs in pharmaceutical wastewater treatment.

Future work should validate the UV/H₂O₂ process in real water matrices, optimize reactor configuration and operational parameters, assess the ecological toxicity of transformation products, and explore catalyst- or coupling-enhanced UV/KMnO₄ systems to broaden the application of UV-based AOPs in pharmaceutical wastewater treatment. Additionally, experiments with a broader set of probe compounds are needed to decouple the individual contributions of polarity, aromaticity, and electron-transfer reactivity.