Ultraviolet-Based Disinfection Technologies in Water and Wastewater Treatment: Developments and Roadblocks

Abstract

Ultraviolet (UV) disinfection is widely used in municipal wastewater and reuse systems, yet its full-scale outcomes depend strongly on how fluence is delivered under real-water conditions rather than on nominal lamp output. This review consolidates the photochemical basis and plant-relevant limitations of UV disinfection, with emphasis on key factors such as matrix optics (e.g., UVT254, color, and turbidity/suspended solids) and particle shielding. Building from these constraints, UV-enhanced disinfection is examined as an engineering strategy that couples UV with oxidants or catalytic/physical processes to expand inactivation pathways and improve robustness in challenging effluents. Representative configurations have been reported (e.g., UV/free chlorine, UV/monochloramine) and are compared in terms of dominant reactive species and reaction networks, matrix dependence and scavenging effects, disinfection performance trends across microbial targets, and process-specific trade-offs including transformation products/disinfection byproducts, energy and chemical demands, and materials durability. Finally, practical considerations for implementation are summarized, including monitoring and control variables, validation approaches (e.g., biodosimetry, challenge testing), and operating windows that balance inactivation with risk and resource inputs, to support more reliable selection and operation of UV and UV-hybrid disinfection for water reuse.

1. Introduction

Growing pressures on aquatic environments and the rising demand for wastewater reclamation have made disinfection an indispensable safeguard in modern water treatment [1,2]. In municipal wastewater systems, the disinfection step determines whether pathogenic microorganisms and other microbiological hazards are released into receiving waters or recycled for reuse. Beyond conventional fecal indicators, treatment plants must now address a wider range of public health risks, including opportunistic pathogens and resistance-associated determinants such as antibiotic-resistant bacteria (ARB) and antibiotic resistance genes (ARGs), which can survive biological treatment and remain detectable in secondary effluents [3,4]. Meanwhile, effluent characteristics are highly variable, with notable fluctuations in optical properties, particle loading, chemical composition, as well as disturbances from storm events and industrial discharges [5,6]. Such variability requires evaluating disinfection not by nominal capacity alone, but by its reliability under dynamic and often unfavorable field conditions.

Ultraviolet (UV) irradiation has become a widely implemented alternative to chemical disinfectants in both drinking water and wastewater applications because it achieves rapid microbial inactivation without generating a persistent residual [7]. Germicidal UV-C (200–280 nm) primarily acts through direct photochemical damage to nucleic acids, producing pyrimidine dimers and other lesions that block transcription and replication [8], while longer-wavelength or hybrid configurations can contribute indirect oxidative effects through the formation of reactive oxygen species [9]. Despite this well-established mechanism, the actual performance of UV systems at full scale is often less predictable than laboratory results suggest. In practice, disinfection efficacy depends on the fluence field, which is the spatial and temporal distribution of UV dose within the reactor, rather than on the nominal lamp output or a single target dose [10]. Reductions in UV transmittance, absorption by dissolved organic matter (DOM), and scattering by suspended solids (SS) can markedly decrease effective irradiance [7,11,12]. Microorganisms attached to or embedded within particles may be shielded, resulting in tailing behavior even when the average doses appear adequate [13]. Hydraulic non-idealities such as short-circuiting, dead zones, and broad residence-time distributions create under-dosed fractions that dominate the risk of microbial breakthrough, especially under stringent log-reduction requirements [14]. Operational factors (e.g., lamp aging, sleeve fouling, sensor drift, and inconsistent maintenance) introduce additional uncertainty. Moreover, photoreactivation and dark repair may partially offset the achieved inactivation if the treated water is subsequently exposed to light or stored before reuse [15].

To address these limitations and enhance overall robustness, recent research has focused on coupling UV with complementary physical, chemical, or electrochemical processes [16]. Such hybrid or UV-enhanced disinfection systems aim to extend microbial inactivation pathways, reduce shielding effects, and simultaneously remove co-contaminants. For example, UV/free chlorine and UV/monochloramine (NH₂Cl) processes can generate reactive chlorine species; UV/ozone (O₃) intensifies hydroxyl-radical chemistry; UV/peracetic acid (PAA) and UV/persulfate (PS) can produce additional reactive intermediates. Furthermore, combinations with other energy inputs such as ultrasound, electricity, and photons can enhance mass transfer or sustain in situ oxidant generation [17,18]. While these configurations often exhibit improved microbial control, they also introduce new engineering challenges. Notable issues include radical scavenging by natural matrices, oxidant demand and dosing control, higher energy use, potential material degradation, and the formation of transformation products or DBPs that may offset the intended benefits [19,20].

Although recent reviews have reported encouraging results, they have mainly focused on individual UV-AOP chemistries or contaminant degradation, while a reliable framework for guiding full-scale implementation remains lacking. Reported performances are commonly based on average UV doses, whereas the actual dose distribution (influenced by reactor hydraulics, optical conditions, and fouling) remains insufficiently characterized. Matrix effects are often treated qualitatively, rather than quantified through UV transmittance loss, particle shielding, or radical scavenging capacity, which limits the ability to predict performance across seasons and sites [21,22]. Because of inconsistencies in metrics and experimental conditions, comparative data among different UV-hybrid systems are difficult to reconcile, making the enhanced performance claims context-dependent. Moreover, byproduct formation and toxicity are rarely evaluated alongside disinfection efficiency, and operational monitoring or validation practices that determine long-term reliability remain unevenly documented.

Against this background, the present review does not aim to repeat those summaries; rather, it aims to connect UV-based process chemistry with full-scale disinfection reliability by linking fundamental photochemical mechanisms to the operational realities that ultimately determine UV performance in full-scale water treatment. In this review, UV-enhanced disinfection is used as a broad term for UV-centered processes that improve microbial control; UV-based AOPs refer more specifically to UV-activated oxidant systems that generate reactive species; and hybrid oxidation systems include wider UV-coupled physical, chemical, catalytic, or electrochemical configurations. We summarize the major factors that constrain dose delivery, including matrix-induced optical losses, particle shielding, hydraulic non-idealities, and equipment-related variability, and discuss how these constraints shape inactivation behavior, tailing, and post-UV repair or regrowth. On this basis, we critically evaluate representative UV-enhanced disinfection strategies. Throughout, emphasis is placed on mechanistic pathways, matrix dependence, performance evidence under realistic conditions, and potential trade-offs associated with transformation products and disinfection byproducts. This engineering-oriented perspective is intended to clarify when UV-hybrid processes provide practical value and which constraints should be checked before full-scale implementation, rather than simply re-listing available UV-AOP configurations.

Literature search strategy: A targeted literature search was conducted using Web of Science, Scopus, and Google Scholar, with keywords including UV disinfection, UV-based AOPs, UV/chlorine, UV/NH₂Cl, UV/O₃, UV/PAA, UV/persulfate, UV/ultrasound, UV-electrochemical disinfection, wastewater reuse, dose delivery, DBPs, and full-scale validation. Priority was given to peer-reviewed studies and reviews related to water and wastewater disinfection, real-water matrices, operational conditions, and engineering implementation. Studies focusing only on unrelated photochemical synthesis or non-aqueous systems were excluded.

2. Impactors for UV Disinfection in Full-Scale Water Treatment

UV disinfection inactivates microorganisms mainly through photochemical damage to nucleic acids rather than through a persistent chemical action [23]. Absorption of germicidal UV, particularly within the UVC range, induces lesions in DNA and RNA, including pyrimidine dimers, which inhibit replication and transcription and thereby prevent microorganisms from remaining infectious [24] (Figure 1). In full-scale applications, UV disinfection is valued for rapid, broad-spectrum inactivation with typically fewer regulated DBPs than chlorination-based approaches and without maintaining a chemical residual in the treated stream [25,26]. Its performance, however, is governed less by nominal lamp power than by effective dose delivery, which depends on the fluence field established in the reactor and the extent to which the water matrix attenuates or shields UV irradiation [27]. Conventional low- and medium-pressure mercury lamps remain common in municipal plants due to established reactor designs, while UV-LEDs are increasingly explored for their compact form factor and wavelength tunability [28,29,30]. Regardless of the source, plant-scale reliability is often limited by optical losses, hydraulic under-dosing, and gradual output decline associated with lamp aging, sleeve fouling, and imperfect monitoring/maintenance [15]. Mechanistically, UV-induced damage includes cyclobutane pyrimidine dimers, 6-4 photoproducts, and possible strand lesions, which suppress genome replication and transcription and may be partly repaired through photoreactivation or dark-repair pathways if post-UV conditions are favorable.

2.1. Water Matrix and Optical Constraints

In full-scale systems, the water matrix often sets the upper bound for UV performance by controlling photon delivery to microorganisms. UV transmittance at 254 nm (UVT₂₅₄) is widely used as an operational surrogate, but it reflects multiple mechanisms, including absorption by chromophoric dissolved organic matter (CDOM), scattering by SS and colloids, and particle-associated shielding [31]. Elevated SS can cause shadowing and reduce effective exposure; under low UV doses, SS reduced E. coli inactivation by 0.4–0.8 log [7]. Increased turbidity further enhances scattering and decreases penetration depth, which is particularly problematic in reactors with low-irradiance zones [11]. Dissolved organics can impair UV disinfection through both light screening and reactive intermediate consumption in UV-hybrid systems. Humic substances attenuate UV directly and can compete for oxidizing species. For example, Deng et al. reported that humic acids decreased naproxen degradation in a UV/chlorine system by 10–15% and reduced E. coli inactivation by 38.3% through competition for reactive oxygen species, with additional evidence of diminished hydroxyl radical stability [12,32,33]. Similar scavenging effects were reported for dissolved black carbon (DBC); DBC suppressed UV/H₂O₂ and UV/PDS performance during carbamazepine degradation, with inhibition increasing with DBC concentration and reaching 72% and 76% reductions, respectively [34]. These observations indicate that low UVT₂₅₄ may signify not only optical attenuation but also high radical demand when UV is coupled with oxidants. Inorganic constituents can further exacerbate optical losses and performance variability. Iron/manganese-related color and persistent colloids sustain attenuation and scattering, while bicarbonate/carbonate alkalinity scavenges radicals and can weaken oxidation-driven contributions in UV-hybrid processes [35]. Overall, matrix-dependent absorption, scattering, and shielding narrow the effective operating window and help explain why identical nominal doses can yield different log removals across plants and seasons. Routine characterization of UVT₂₅₄, turbidity/SS, and indicators of particle association and oxidant/radical demand is therefore essential for both design and stable operation.

2.2. Hydraulics and Dose Distribution

UV performance at full scale is often constrained by hydraulics because microorganisms experience a dose distribution rather than a single uniform dose [36]. Non-ideal residence time distributions (RTDs), short-circuiting, dead zones, and heterogeneous velocity fields create under-dosed fractions that dominate breakthrough risk [14], so the average or nominal target dose may correlate poorly with observed log-removal under changing flow conditions. Flow effects reflect competing mechanisms. Higher velocities reduce retention time and thus delivered dose per pass, yet enhanced mixing can partially improve exposure uniformity by redistributing particles and weakening local shielding [37]. Shen et al. [14] reported that during UV treatment of milk, the sterilization efficiency first decreased and then increased with increasing flow velocity, but remained below the initial level, indicating that mixing-induced gains did not fully compensate for the loss in residence time. Similar trade-offs are expected in wastewater UV reactors, where hydrodynamics may mitigate low-dose tails, but cannot eliminate them entirely. Practically, performance improvement relies on limiting under-dosed pathways within the specific reactor geometry. Reactor configuration (open-channel vs. closed-vessel), lamp layout, baffles, and inlet/outlet design govern RTD and the fluence-rate field. CFD- and RTD-based validation are therefore valuable for diagnosing “same set dose, different log-removal” behavior and for targeting short-circuiting and dead zones rather than simply increasing lamp power. In this context, CFD coupled with RTD testing or biodosimetry can be used to identify low-fluence zones, optimize baffle and lamp arrangements, and balance turbulence-enhanced mixing against reduced hydraulic residence time.

2.3. Reactor and Operation: Lamp, Sleeve, Monitoring, Validation

Stable UV disinfection depends on maintaining delivered irradiance over time. UV sources (low-pressure, medium-pressure, and UV-LED) differ in emission spectra and energy efficiency, but all exhibit output decline with aging and operational drift [38], which directly affects dose delivery. Quartz sleeve fouling represents another common limiting factor in wastewater treatment. Both inorganic scaling and biofouling reduce sleeve transmittance and lower effective irradiance. Scaling is promoted by thermal gradients near the lamp and by deposition/precipitation of trace metals and inorganic species during operation [38,39,40]. Given the low UV transmittance of most deposits, even thin fouling layers can cause appreciable dose loss and reduce inactivation performance [41]. Reliable operation therefore requires measurement and control strategies that track dose rather than lamp age alone. Sensor drift and fouling can bias intensity readings, and dose pacing without UVT₂₅₄ compensation can misrepresent delivered dose under fluctuating water quality [31]. Routine calibration, sleeve cleaning, and setpoints linked to flow and UVT₂₅₄ could improve operational stability. Performance validation using biodosimetry or challenge testing remains important, as it integrates optical and hydraulic effects into outcome-based metrics [38]. Furthermore, microbial photoreactivation, dark repair, and regrowth may affect apparent efficacy [23], so sampling and assay conditions should be standardized for meaningful evaluation. For UV-LEDs, wavelength selectivity may allow better matching with microbial action spectra or oxidant absorption bands; however, current applications are still constrained by wall-plug efficiency, thermal management, array-scale uniformity, lifetime under continuous operation, and scale-up cost.

3. Development of UV-Based AOPs

The limitations of UV disinfection in municipal effluents have spurred UV-enhanced approaches that couple UV with oxidants or catalytic/physical processes to improve robustness under complex water matrices [3]. In UV-based AOPs, UV activation of free chlorine/NH2Cl [17], O3 [35], PAA [42], or PS generates highly reactive species (e.g., •OH) [18], thereby complementing direct photodamage with radical-mediated oxidation. Numerous studies have reported high disinfection efficiency for these configurations, supporting their continued development for challenging treatment scenarios [3].

Across these UV-enhanced systems, the radical chemistry differs substantially. UV/free chlorine and UV/NH₂Cl generate both hydroxyl radicals and reactive chlorine or nitrogen-centered species, whereas UV/O₃ is mainly associated with hydroxyl-radical chemistry. UV/PAA can produce hydroxyl and organic acyl/peroxy radicals, while UV/PS introduces sulfate radicals with a stronger dependence on pH and halide composition. Therefore, process performance should be interpreted not only by log inactivation, but also by radical selectivity, scavenging by the water matrix, and byproduct-forming potential (

Table 1. Comparison of redox potential (V, vs. normal hydrogen electrode), lifetime (s), pK_a, and diffusion coefficient (cm² s⁻¹) for radical species.

Reactive Species	Redox Potential	Lifetime	pKa	Diffusion Coefficient	Ref.
SO4•−	E0 = 2.5–3.1	(3–4) × 10−5	<0	1.49 × 10−5	[43]
•OH	E0 = 1.8–2.7	1.0 × 10−9	11.9	2.30 × 10−5
Cl•	E0 = 2.2–2.6	N/A	N/A	2.42 × 10−5
Cl2•−	E0 = 2.1–2.3	N/A	N/A	1.70 × 10−5
HOCl•−	E0 = 1.9	N/A	−4.7	1.54 × 10−5
(CH3C(O)O•)	E0 = 2.24	N/A	N/A	N/A	[44]
(CH3C(O)OO•)	E0 = 1.6	N/A	N/A	N/A

).

3.1. UV/Free Chlorine and UV/NH₂Cl

UV/free chlorine and UV/NH₂Cl are attractive for engineering practice because they combine a conventional disinfectant with UV activation and can be implemented with relatively simple retrofits [17] (Figure 2a). In municipal wastewater, chlorine-resistant microorganisms may reduce the apparent efficacy of chlorination, whereas UV can inactivate chlorine-tolerant populations and thereby strengthen the overall disinfection barrier [4]. Mechanistically, chlorination exerts its biocidal effect primarily through oxidative damage to key cellular functions (e.g., enzyme systems) [45]. Under UV irradiation, free chlorine species (HOCl/ClO⁻) undergo photolysis to produce reactive intermediates, including reactive chlorine species and ^•OH, thereby introducing additional oxidative pathways [46,47,48]. UV exposure can also increase cell envelope susceptibility, facilitating subsequent oxidant attack on membrane-associated structures [49,50]. For NH₂Cl, UV photolysis proceeds through nitrogen-containing radical pathways, generating a more complex reactive network than that of free chlorine [51]. Overall, the process is best viewed as a coupled disinfection and oxidation system, in which improved robustness and broader reactivity are gained at the cost of increased reaction complexity and potentially altered byproduct chemistry [52].

Beyond reporting higher log removal, performance should be interpreted in terms of target microorganisms (bacteria/viruses/protozoa), potential effects on ARB/ARGs (if assessed), and whether observed outcomes exceed the additive expectation of UV and oxidant alone (synergy) [52]. A common example is MS2 inactivation; compared with chlorination alone, UV/chlorine achieved 1.3-fold higher disinfection efficiency, consistent with synergistic action between UV damage and chlorine-derived reactivity [52]. Similarly, for P. aeruginosa, the disinfection efficiency of UV/chloramine increased by 11.34 times relative to chlorination alone and 2.28 times relative to UV alone [56]. Other studies have reported a 7.7-fold increase in bactericidal efficacy against E. coli [45] (

Table 2. Summary of UV-based AOPs for microbial inactivation. Representative operating conditions are retained as reported in the cited studies.

UV-AOPs	Targets	Reaction Conditions	Performance	Ref.
UV/Cl	MS2	[UV] = 20 mJ/cm2; [Cl] = 5~10 mg/L [contact time] ≤ 120 s	1.5-fold higher than Cl alone	[52]
	P. aeruginosa	[UV] = 50 μM/cm2; [react time] = 5 min [NH2Cl] = 5 mg/L; pH = 7.0	2.28-fold/11.34-fold higher than UV alone/NH2Cl alone	[56]
	E. coli	[UV] = 13 mM/cm2; [Cl] = 5 mg/L; pH = 7.0 [react time] = 10 min	7.7-fold/10.8-fold higher than UV alone/Cl alone	[45]
UV/O3	Six PCBs	[O3] = 1 mg/L; pH = 8.0	8.8-fold higher than UV alone	[57]
UV/PAA	Fecal coliform	[PAA] = 70.0 mg/L [UV] = 4.76 × 10−7 Einstein·L−1·s−1; pH = 7.3; [react time] = 10 min	4.17-fold/1.27-fold higher than UV alone/PAA alone	[58]
	E. coli	[UV] = 2.25 × 10−7 Einstein·L−1·s−1 [PAA] = 60–120 mM; pH = 7.5 [react time] = 3–4 min	2.90-log total bacteria inactivation in actual raw CSOs after 4 min	[59]
UV/PS	SAC	[PS] = 0.42–2.52 mM [SAC] = 0.055–0.220 mM; pH = 7.5 [react time] = 60 min	29.6-fold higher than UV alone	[60]
	quinoline	[UV] = 0.15 μE·s−1; [quinoline] = 200 mg/L [PS] = 4.68 mM; pH = 3, 5, 7, 8 [react time] = 60 min	Quinoline near completely degraded within 40 min	[61]
	tetracycline	[tetracycline] = 5 mg/L [PDS] = 30 mg/L [react time] = 30 min	4.23-fold higher than UV alone	[62]
UV/US	E. coli	[US] = 20–30 kHz; [US] = 40 W [Electrical power per LED] = 0.6 W [Optical power per LED] = 12 mW [react time] = 15–120 s	>95% inactivation	[63]
	E. coli	[UV] = 2.1 ± 0.1 mW/cm2 [US] = 40 kHz; [US] = 180 W [react time] = 1–5 min	1.2-fold higher than UV alone	[64]
UV/PEC	Erythrosine B dye	[dye] = 100 mg/L; [Na2SO4] = 0.05 M [current density] = 20 mA/cm2 pH = 3.0; [react time] = 90–120 min	Faster 60.00% than PEC alone	[65]

).

The synergy and altered reactivity of UV/chlorine also influence disinfection byproduct (DBP) profiles, by transforming chlorine-containing precursors and chlorinated intermediates. UV activation of chlorine/chloramine has been reported to generate photolysis radicals and reduce the concentration of organic chloramines (OCs), which are associated with DBP formation risk [66]. For instance, Xu et al.’s [67] study found that UV promoted OC degradation, reaching 96.6% after 60 min. Consistently, compared with chlorination alone, UV/chlorine reduced the formation of several DBPs, including trichloromethane (THMs), haloacetic acids (HAAs), dichloroacetioc acid (DCAA), and trichloroacetioc acid (TCAA). At a UV dose exceeding 300 mJ/cm², THMs, HAAs, DCAA, and TCAA decreased by 42.25%, 13.75%, 70.70%, and 90.40%, respectively [68,69]. At the same time, however, UV activation of chlorine/chloramine can introduce additional transformation pathways, such as the formation of inorganic oxychlorine species or halogenated oxidation byproducts under specific conditions. This dual effect warrants matrix-dependent assessment and control [52]. In addition, UV/chlorine or UV/chloramine processes may promote chlorate/perchlorate formation and nitrogenous DBPs such as haloacetonitriles, halonitromethanes, haloacetamides, and N-nitrosamines, particularly in waters containing ammonia, organic nitrogen, or high halide levels.

3.2. UV/O₃

UV/O₃ is a representative UV-AOP route in which UV photolysis of O₃ and subsequent chain reactions increase ^•OH formation [19,35]. When ^•OH becomes dominant, performance is strongly controlled by radical scavenging (e.g., NOM and carbonate/bicarbonate). Thus, the same UV/O₃ inputs can yield different outcomes across different water matrices [70]. Ozonation may also modify effluent optics, for example, reducing color and some UV-absorbing constituents, which can indirectly improve UV penetration. Ozone-driven improvements related to turbidity have also been reported [71], though the magnitude is site- and matrix-dependent [10]. UV/O₃ can be valuable when high virus inactivation or control of disinfectant-tolerant microorganisms is required [72]. Albert et al. [73] reported that at an ozone concentration of 0.75 mg/L, inactivation of SARS-CoV-2 reached 82–91.5%. Beyond disinfection, UV/O₃ enhances the oxidation of dissolved organic pollutants. For example, during treatment of six polychlorinated biphenyls (PCBs), O₃ alone achieved 95% removal, whereas UV/O₃ reached 97% [57,74], indicating an incremental benefit from UV activation [74] (Table 2).

Ozone-based systems may reduce DBP precursors. For instance, ozone-induced DOC conversion decreased THM formation potential by 47% [75]. Chen et al.’s [76] study showed that, compared with O₃ alone, UV/O₃ reduced the formation of absorbable organic bromine (AOBr) by 81.4%. However, in bromide-containing waters, bromate risk becomes a central constraint, and additional oxidation byproducts (e.g., aldehydes/ketones) may require attention [76]. For practical application, bromate control in bromide-containing waters and ozone mass-transfer efficiency are two key constraints; insufficient gas–liquid transfer can reduce dissolved ozone availability, increase off-gas management requirements, and weaken the expected benefit of UV/O₃ treatment.

3.3. UV/PAA

UV/PAA combines the intrinsic disinfecting capability of peracetic acid with UV irradiation and is often discussed as a low-halogen, DBP-conscious option for complex waters [42]. PAA has been reported to be effective against bacteria [77], viruses [78], parasites [79], and fungi [80]. Under UV irradiation, cleavage of the O–O bond in PAA yields radicals such as ^•OH and acetoxy radicals (CH₃C(O)O^•), creating an oxidative system that can complement UV photodamage and strengthen overall inactivation [81,82] (Figure 2d). Conceptually, UV/PAA provides a dual pathway of chemical inactivation by PAA combined with photochemical/radical-driven damage under UV [42]. UV/PAA has demonstrated strong performance across multiple wastewater types, particularly when short contact time and broad-spectrum inactivation are required [83]. In toilet blackwater treatment, the UV/PAA system (70 mg/L PAA) achieved 5.95-log reduction within 10 min, outperforming UV alone (1.36 log) and PAA alone (4.59 log) [58]. In combined sewer overflows, UV/PAA achieved 4.58-log inactivation of E. coli, significantly higher than that of UV or PAA alone [59] (Table 2). Similar enhancement was observed in agricultural wastewater [82]. PAA is often described as producing fewer chlorinated DBPs because it does not rely on free chlorine chemistry [84]. However, low DBP should be interpreted as a different byproduct spectrum rather than an absence of transformation products [85]. In practice, residual PAA management and matrix-dependent side reactions, especially in mixed-halide waters, should be taken into consideration [83].

3.4. UV/PS

To maintain rigor, persulfate should be distinguished as peroxydisulfate (PDS, S₂O₈²⁻) and peroxymonosulfate (PMS, HSO₅⁻) [86]. Under UV irradiation, persulfates are activated to produce sulfate radicals (SO₄^•−) and, via interconversion reactions, ^•OH; the SO₄^•−/^•OH balance depends on pH and matrix chemistry [87] (Figure 2b). Although persulfates can be activated thermally [88], by metal catalysts [89], or via ultrasonication [90], UV activation is often considered operationally straightforward, requiring mild conditions and simple equipment [3]. In waters containing chloride/bromide, secondary halogen-radical chemistry can emerge, affecting both performance and byproduct formation [91]. UV/PS has been explored for waters containing recalcitrant organics and resistant microorganisms [87]. In UV/PMS treatment, simultaneous enhancement of microbial inactivation and oxidation of cyanotoxins/microcystins has been reported [92]. For saccharin (SAC) removal, UV or persulfate alone showed limited effect, whereas UV/PS increased SAC removal to 85.39% [60]. For quinoline, PS alone provided negligible removal, but UV/PS achieved 77.2% removal [61]. UV/PS also achieved up to 95.73% removal of tetracycline in wastewater [62] (Table 2). More specifically, PDS mainly yields sulfate radicals through S₂O₈²⁻ bond cleavage, whereas PMS can generate both sulfate and hydroxyl radicals through HSO⁵⁻ activation; the relative contribution of these pathways depends on UV wavelength, pH, alkalinity, halides, and background organic matter.

Compared with chlorine-based disinfection, UV/PS avoids formation of traditional chlorinated DBPs in many cases. However, sulfate-based oxidation can generate distinct transformation products, and halide-containing matrices may shift pathways toward halogenated byproducts [87]. For NOM removal, UV or persulfate alone showed limited effect, whereas UV/PS increased NOM removal to 19.29%, attributed to ROS formation upon UV activation of persulfate [60]. Residual oxidant control and matrix-dependent scavenging (by NOM and carbonate) represent practical constraints [70]. Although UV/PS is effective for disinfection, it can also generate a greater amount of DBPs under certain conditions. For instance, when treating solutions containing SAC, the production of DBPs such as TCAA and DCAA increases significantly, an issue that should be taken into account in practical applications [88].

3.5. UV/Other Energy

UV/US leverages ultrasound cavitation to assist UV disinfection. At 20 kHz, ultrasound has been reported to inactivate 93% of mycobacteria [93]. The collapse of cavitation bubbles generates strong shear forces and localized microenvironments with high temperature and pressure, which can disrupt cell envelopes and produce oxidative radicals that accelerate the leakage of intracellular components [94]. From a UV perspective, ultrasound can reduce particle/floc shielding by fragmentation and improve mass transfer near irradiated zones, thereby indirectly narrowing dose tails [64]. Because ultrasound is energy-intensive, it is typically treated as an auxiliary step rather than a stand-alone disinfection method [64]. When integrated with UV, the mechanical disruption from ultrasound can increase UV accessibility to cellular targets and shorten the required exposure time. UV/US inactivation of E. coli was reported to be 1.2-fold more effective than UV alone, achieving in 3 min what required 5 min under UV-only conditions [64]. Another study by Li et al. [63] reported 95% inactivation within a short period using UV/US (Table 2). A practical advantage of UV/US is the absence of intentional chemical oxidant addition, which generally limits classical DBP concerns [63]. However, treatment outcomes remain matrix-dependent, and ultrasonic operation may promote the release of intracellular organic matter, which can alter downstream oxidation demand [93].

UV–electrochemical systems integrate photochemistry with electrochemical oxidant generation or photoelectron-catalytic ROS production [95] (Figure 2c). In electrocatalytic disinfection, reactive species (e.g., active chlorine and ^•OH) can be generated at electrodes and oxidize bacteria, while direct electrical effects (e.g., electroporation) can also disrupt cell integrity [96]. In electro-Fenton systems, H₂O₂ generated at the cathode reacts with Fe²⁺ to produce ^•OH, enabling oxidative inactivation [97]. Conventional Fenton-type systems face constraints including high cost, dependence on electrode/catalyst selection, and sensitivity to operating conditions [98]. Consequently, UV/electro-Fenton has been proposed to improve radical generation and process efficiency [95]. UVA (315–400 nm) can be absorbed by iron complexes and facilitate Fe²⁺ regeneration, reducing reliance on continuous catalyst addition [99,100]. Regarding the reported inactivation mechanisms for E. coli, UVA alone provides limited direct lethality, whereas electro-peroxidation and ROS generation compromise cell membranes, enabling deeper penetration of UVA and oxidants. Moreover, UVA can promote ROS formation [101,102].

UV/electrocatalytic disinfection has been reported with broad applicability and high efficiency [65], but energy demand and electrode cost remain major barriers. In Chen et al.’s [99] study for E. coli, UVA alone required 140 min and electro-peroxidation required 160 min, whereas the combined UVA/electro-peroxidation achieved maximal inactivation in 50 min, indicating their synergetic effects. A similar study by Clematis et al. [65] showed that for the dissolution of Erythrosine B dye, electro-peroxidation alone took 50 min to achieve the desired result, whereas UVA/electro-peroxidation took only 20 min (Table 2). Cathode material and long-term stability are critical cost drivers [103]. For example, incorporating Cu into cathodes via sol–gel methods was reported to reduce excessive PTFE accumulation and lower energy consumption, consuming 21.34 kWh/kg COD to treat one ton of wastewater [104]. Byproduct profiles depend strongly on electrolyte composition (especially chloride) and electrode reactions. Previous research has reported that active chlorine formation could introduce chlorinated transformation products under certain conditions [103]. Additionally, electrode fouling, catalyst loss, and performance decay are constraints for long-term operation [98].

4. Implications and Challenges

4.1. Engineering Windows for UV-Based AOPs

Across UV-based disinfection trains, the main engineering variables can be grouped into matrix conditions (e.g., UVT, turbidity/SS, NOM, and pH), process inputs (UV dose, oxidant dose, and contact time), and control requirements (mixing, residual control, and validation). These grouped factors jointly determine photon delivery, radical production, scavenging intensity, and byproduct risk. Beyond these shared requirements, the practical distinctions among processes are better understood from the standpoint of implementation context. To make the comparison more quantitative while keeping the review concise,

Table 3. Summary of reported DBPs in UV-based AOPs.

UV-AOPs	DBPs	Reaction Conditions	Change Amount	Ref.
UV/Cl	THMs	[UV] = 40–300 mJ/cm2	42.25% lower than chlorination alone	[69]
	HAAs	[UV] = 40–300 mJ/cm2	13.75% lower than chlorination alone	[69]
	TCAA	[UV] = 40 mJ/cm2; [Cl] = 3 mg/L	79.40% lower than chlorination alone	[68]
	DCAA	[UV] = 40 mJ/cm2; [Cl] = 3 mg/L	70.70% lower than chlorination alone	[68]
UV/O3	THMFP	[DOC] = 13.8 mg/L; [O3] = 0–0.8 mg O3/mg DOC	47.0% lower than O3 alone	[75]
	AOBr	[UV] = 100 mJ/cm2; [O3] = 1 mg-O3/mg-DOC; pH = 7.0	79.4% lower than O3 alone	[76]
UV/PAA	Almost none	/	/	[84]
UV/PS	TCAA	[SAC]0 = 0.11 mM; [PS]0 = 1.05 mM; UV = 11 W (254 nm); pH = 7.0	91.48% higher than the blank experiment	[88]
	DCAA	[SAC]0 = 0.11 mM; [PS]0 = 1.05 mM; UV = 11 W (254 nm); pH = 7.0	1234.96% higher than the blank experiment	[88]
UV/US	Almost none	/	/	[63]
UV/PEC	Depends on electrolyte composition			[103]

summarizes representative reported reaction conditions, including oxidant dose, UV fluence or intensity where available, pH, reaction/contact time, and observed inactivation or removal performance.

The values summarized in Table 3 should therefore be interpreted as evidence of process-specific operating windows rather than as a direct ranking of UV-AOPs, because the reported studies differ in water matrix, target organism, reaction time, UV/oxidant dose, and energy-related input.

UV/chlorine and UV/NH₂Cl are often the most immediately deployable options under variable-transmittance or otherwise challenging matrices where UV-alone performance is unstable [66]. Their engineering appeal lies in the relative accessibility of chemical addition and residual-based process control. When treatment goals extend beyond microbial inactivation to concurrent micropollutant abatement, UV/O₃ becomes particularly attractive because it couples disinfection with substantial oxidation capacity. However, its broader application is constrained by ozone mass-transfer efficiency, dissolved ozone control, safety requirements, and energy demand [72]. Therefore, it is most defensible in contexts where ozone generation and control are already feasible at the plant scale. Where such ozone infrastructure is difficult to justify [75], UV/PS provides a more practical route to combine disinfection with oxidation of persistent organics, although its use still must be evaluated against scavenger sensitivity and dose–cost–byproduct trade-offs. If a simpler oxidant-handling framework is preferred [86], UV/PAA offers a more operationally straightforward alternative, especially where stable oxidant feeding, appropriate dosing location and residual PAA management can be maintained within the intended operating window [31], making online residual control especially important.

Taken together, Table 3 shows that adding UV does not uniformly reduce byproduct risk; instead, DBP control depends on the oxidant used, halide content, precursor composition, and residual management.

A different scenario arises when the dominant limitation is not oxidant selection but particle-associated shielding, particularly in high-turbidity waters or systems with strong floc association. Under those conditions, UV/US is more defensible because ultrasound can be applied (including intermittently) to disrupt shielding and recover UV effectiveness [11], albeit with penalties in energy use, equipment complexity, and maintenance [105]. Finally, UV-assisted electrochemical and photo-electrocatalytic routes are most relevant in treatment settings that can support tighter process control and greater operational sophistication. Their practical value depends less on basic UV exposure alone than on managing current density, electrode material, electrolyte composition, UV irradiation conditions [99], and hydraulic configuration, while balancing disinfection benefit against energy and cost, electrode durability, and monitoring complexity (Figure 3) [3].

4.2. Key Constraints of UV-Based Hybrid Disinfection

For scale-up, the central question is not whether UV-based systems can inactivate microorganisms, but whether the delivered dose and process control remain reliable under changing matrix, hydraulic, and maintenance conditions. Therefore, the following discussion summarizes these issues from the perspective of implementation barriers, rather than repeating the mechanism-level constraints described above.

Although UV-based hybrid disinfection and UV-AOPs can improve microbial control in complex municipal effluents, their wider use is still constrained by energy demand, matrix-specific robustness, long-term operational stability, and byproduct/risk management. These barriers require comparable reporting of energy per treated volume or per log inactivation, validation under representative real waters, and monitoring strategies that link UVT-compensated dose control with residual and byproduct management. In addition, future scale-up studies should therefore report comparable techno-economic indicators, including kWh m⁻³, energy per log inactivation, electrical energy per order (EEO), oxidant cost, maintenance frequency, and residual/byproduct monitoring costs, rather than relying only on laboratory inactivation efficiency.

5. Conclusions

In summary, UV-based hybrid disinfection represents a viable and increasingly mature approach to enhance microbial control in complex municipal effluents, provided that its advantages are assessed against process-specific trade-offs in a context-aware manner. These trade-offs include: (i) matrix-dependent performance due to radical scavenging by NOM and carbonate/bicarbonate; (ii) distinct byproduct profiles, ranging from halogenated DBPs and bromate in chlorine-/ozone-based systems to sulfate-driven transformation products in persulfate processes; (iii) elevated energy demand, which must be justified using metrics such as kWh·m⁻³ or EEO; and (iv) operational reliability challenges, including lamp aging, sleeve fouling, sensor drift, and the need for regular calibration and cleaning. This review aims to provide a structured framework for process selection, operation, and validation, encompassing UVT-compensated control, biodosimetry or challenge testing, and the establishment of operating windows that balance log-removal targets against byproduct risk and resource inputs, thereby supporting more reliable and sustainable UV disinfection for water reuse. Thus, no single UV-hybrid process can be regarded as universally superior: chlorine- or chloramine-assisted systems are more readily integrated into existing disinfection trains, UV/O₃ and UV/PS provide stronger oxidation capacity but require stricter control of energy input and byproducts, whereas UV/PAA, UV/US, and UV–electrochemical routes are more application-dependent.

The contribution of this review lies in linking UV-based process chemistry with full-scale disinfection reliability, providing a practical complement to previous process-specific UV-AOP or UV-LED reviews. Future studies should standardize dose, matrix, energy, and DBP reporting and should give greater attention to pilot- and full-scale validation, long-term fouling control, and operation under variable real-water conditions.