1. Introduction
Conventional disinfection technologies (e.g., chlorine, ultraviolet (UV), ozone) are effective to ensure the water quality standard based on fecal indicator bacteria, with reasonable costs and wide applications
[12]. However, they are challenged with the increasing detection of the emerging biological contaminants (BCs), in view of the developing microbial detection techniques (from culture-based to molecular-based). BCs include not only pathogenic microorganisms, but the endogenous pollutants including living prey (intracellular pathogens), endotoxin, and antimicrobial resistance genes (ARGs). The removal of culturable indicators such as
Escherichia coli (
E. coli) neither indicates the removal of other pathogens, nor even indicates the complete removal of their persistent life forms and endogenous pollutants
[3]. Opportunistic human pathogens that in their stubborn survival states or protected by their predator (typically, the protozoa), have been widely detected in the effluents of drinking water and wastewater treatment, which can cause severe infections if directly ingested or inhaled in aerosol, especially for the individuals with weakened immune systems
[4]. Statistically, waterborne diseases result in approximately annual 1.8 million death worldwide
[5]. Moreover, the retained pathogenic bacteria are prone to harbor ARGs and mobile genetic elements in the disinfected waters, further raising global burden of disease
[6],
[7]. Therefore, it’s highly desirable to develop new powerful technologies against BCs.
The interdisciplinary investigations of BCs control, especially in the field of biomedical and materials, have inspired proof-of-concept trials for water treatment. Emerging antimicrobial technologies (EATs), including nanotechnology, advanced oxidation process (AOP), biological control process, and their integrated techniques, are increasingly reported. Recent advanced branches of the four focused categories are described and introduced
. Though efficient antimicrobial effects of EATs have been highlighted and partially reviewed
[8],
[9], their tailored improvements from engineering perspectives are always circumvented. Scattered information is presented about the how the EATs adapt to the complex water matrices. Comparison between EATs and conventional disinfection technologies in terms of engineering-concerned trade-offs (i.e., reconciling between antimicrobial efficiency, economy, biosafety, and sustainability) is rarely involved. Indeed, development of EATs still prioritizes the antimicrobial efficiency, while inclined to avoid addressing the specific challenges for practice applications.
In this review, we firstly list the hard-to-remove BCs in the disinfected effluents in wastewater and drinking water systems. Then we briefly summarize the performance of conventional treatments. Afterwards, we try to share valuable insights into the EATs, elaborating on the controlling factors, advantages, and challenges. On this basis, we tentatively evaluate the application potential of the representative EATs with the assistance of life cycle assessment (LCA) and analytic hierarchy process (AHP) before recommending their feasible extension in water treatment system. Herein, we provide qualitative and semi-quantitative supports for the scale-up of emerging BCs control technologies.
2. Why do BCs persist in the effluents of current water treatment?
Wastewater contains sufficiently higher microbial loads in the influents than drinking water, as proven by the several orders of magnitude higher diversity and abundance (
Table 1 [1],
[2],
[10],
[11],
[12],
[13],
[14],
[15],
[16],
[17]). Drinking water systems partially remove BCs through physicochemical treatments, primarily ensuring biological safety through endpoint disinfection. Wastewater systems employ a combination of secondary, tertiary, and/or advanced treatments for BCs removal and seem to exclude more indicators than drinking water systems, but still contain more pathogens in the effluents and pose high health risks to the receiving waters
[2]. Despite meeting standards based on fecal indicators, growing BCs have been commonly identified in the final waters of both wastewater and drinking water treatments, threatening biological safety. These BCs are classified as “persistent” and summarized in Table S1 in Appendix A, and their persistent pattern is closely correlated with the self-defense strategies and treatment procedures
[6],
[18].
2.1. Persistent BCs and their survival strategies
2.1.1. Bacteria
Proteobacteria, Bacteroidetes, and Firmicutes were the predominant phyla in the wastewater effluents
[19]. Myriad genera, the most abundant of which are
Proteobacteria, are clinically relevant (Table S1).
Escherichia, Legionella, and
Mycobacterium ranging from 10
0 to 10
2 colony-forming unit (CFU)·mL
−1 were examined using cultivation methods
[1]. Pathogens such as
Legionella,
Salmonella,
Klebsiella, and
Acinetobacter have also been identified in tap water
[20],
[21], although health risks can be overestimated due to indiscriminate metagenomic analysis regardless of bacterial viability.
Currently, the effluent standard for bacterial removal in water treatment is the loss of cultivability
[22],
[23]. However, the biological risks are likely to be underestimated. This is, on the one hand, owing to the trace amount of culturable microbes (below detection limit) that persist after disinfection. It is achieved by the increased gene expression of SOS response (a global regulatory network that aids bacterial propagation by inhibiting cell division), efflux pump, DNA repair, and porin regulation
[24],
[25]. Self-adaptive behaviors, such as cell aggregation, spore production, and excretion of extracellular polymeric substances, which are regulated by quorum sensing
[26], also enable bacteria to survive. On the other hand, viable but nonculturable (VBNC) bacteria have been detected in the effluent. Typical pathogens, including
E. coli,
Staphylococcus aureus,
Pseudomonas aeruginosa,
Salmonella, and so forth, could enter the VBNC state under external stress (Table S2 in Appendix A), forming the majority (> 90%) of total bacterial communities after drinking water disinfection
[27],
[28]. They can also develop tolerance to antibacterial agents by altering their morphology to reduce the surface-to-volume ratio for less contact
[29],
[30], decreasing metabolism and respiratory activity to enter the dormant phase
[31], and retaining high levels of genetic material and adenosine triphosphate (ATP) inside the cells
[32] for recovery and regrowth under favorable conditions.
2.1.2. Viruses
Human enteric viruses (e.g., norovirus, adenovirus, enterovirus, and rotavirus) are regulated for water treatment in the United States
[33] because of their pervasive identification
[34],
[35],
[36]. The concentration of enteric viruses ranges from 10
0 to 10
1 genome equivalents (GE) copies·L
−1 in treated wastewater effluent, and their inherent resistance to conventional disinfection is attributed to their unique three-layer capsid protein structure
[37] and the utilization of host repair systems for some double-stranded DNA viruses
[38]. Hepatitis viruses are detected in the effluents of both drinking water and wastewater systems and elicit liver symptoms through blood or body fluid transmission. Notably, severe acute respiratory syndrome coronavirus 2, the culprit of the coronavirus pandemic, is occasionally detected in wastewater effluents
[39] and may be encapsulated by biofilms and retained in drainage systems
[40], posing escalating risks to public health and safety.
2.1.3. Fungi
Fungi are routinely regarded as the functional communities involved in biological treatment
[41]. However, recent outbreaks of
Candida auris have increased awareness of the pathogenicity of fungi in the aquatic environment.
Fungi exhibit higher resistance to disinfection than bacteria, primarily because of their larger size and more complex cellular components, such as melanin, in thicker cell wall
[42],
[43],
[44]. The ubiquitous occurrence of waterborne fungi in wastewater, drinking water, and portable hospital water has become particularly concerning
[45]. Basidiomycota and Ascomycota are the dominant fungal phyla in urban water systems
[46], among which
Fusarium,
Aspergillus, and
Candida are the most prevalent genus in clinical. Increased resistance to antifungal agents is another challenge
[47], while systematic research on this topic is still rare. Only a few studies have investigated the antifungal resistance levels of
Candida [48] and
Fusarium [49] species in wastewater.
2.1.4. Protozoa
The most extensively studied waterborne protozoa are
Giardia and
Cryptosporidium [50],
[51], which are occasionally detected (generally 0–10
1 (oo)cysts·L
−1) in treated wastewater. Sequencing-based detection results also indicated the widespread incidence of other intestinal parasites (e.g.,
Entamoeba,
Blastocystis,
Naegleria, and
Entamoeba)
[52],
[53],
[54]. There are growing concerns about live protozoa (mainly amoebae
[55]) present in drinking water systems because pathogenic fungi are more resistant to disinfection than bacteria.
2.1.5. ARGs
ARGs are another typical and widely detected type of BCs in the effluents of water treatment systems
[56],
[57],
[58],
[59], in which the most abundant classes are multidrug, beta-lactamase, sulfonamide, aminoglycoside, and macrolide–lincosamide–streptogramin. ARGs concentrations vary from 10
1 to 10
6 copies·mL
−1 and 10
0 to 10
3 copies·mL
−1 in wastewater and drinking water effluents, respectively
[60]. Their health risks include the discussions surrounding the fate of clinically relevant ARGs, especially those conferring resistance to multiple antimicrobial agents
[56], and the horizontal gene transfer (HGT) of ARGs to pathogens via mobile genetic elements .
2.1.6. Bacterial endotoxin
Endotoxin is a lipopolysaccharide mixture in the cell wall of some Gram-negative bacteria and cyanobacteria, and is released with the death of bacteria
[61]. The concentration of bacterial endotoxin is expressed in endotoxin units per milliliter (Eu·mL
−1), approximately ranging 10
2–10
3 and 10
−1–10
2 Eu·mL
−1 in reclaimed and drinking water, respectively
[62]. While there is no sufficient data for health risk evaluation.
2.1.7. Intracellular pathogens
Free-living protozoa can inject and harbor bacteria, fungi, and viruses during water treatment
[63]. Among them, the ingested pathogens can avoid being detected by traditional culture-based methods, thus significantly increasing their resistance to the water disinfection process (trojan horse effect)
[64],
[65]. Therefore, it is challenging to examine the existence of endogenous microorganisms that pose potential public health risks
[19],
[55]. For example, protozoan ingestion can explain the different levels of
Legionella measured by polymerase chain reaction (PCR) and culture-based methods in drinking water
[65]. Opportunistic pathogenic bacteria, such as
Pseudomonas fluorescens and
Pseudomonas putida have been identified inside the protoza in wastewater (although their viability is unknown), with higher relative abundances in the effluent than in the influent
[19]. One possible reason is that external stresses, such as limited nutrients and disinfection, may cause protozoa to form spores and cysts, thereby providing shelter for ingested pathogens
[66]. This weakened the inactivation efficiency of chlorine, ozone, UV, and chlorine dioxide disinfection, where a ten-fold increase in the effective dose occurred achieve a 3 logarithmic (log) removal
[67].
2.2. Inapplicability of the conventional methods for BCs removal
2.2.1. Wastewater treatment
In full-scale wastewater treatment, secondary/tertiary physicochemical treatments and disinfection procedures can effectively remove BCs. Interception, adsorption, and gravity sedimentation have been extensively studied as physicochemical mechanisms for the removal of BCs. The total bacteria are randomly removed for 0.1–5.0 log by solid–liquid separation
[68]. Sand filtration can intercept protozoa (up to three log)
[69]. Coagulation or clarification partially removed smaller bacteria and viruses (< 1 log)
[70]. Microfiltration and ultrafiltration can be used to capture micro-scale bacteria and viruses
[71]. Nanofiltration and reverse osmosis sufficiently remove nanoscale viruses and even ARGs
[70],
[72], but are limited by their high cost and energy requirements. Adsorption techniques using activated carbon, typically employed in tertiary treatments to remove chemical contaminants, can synergistically remove BCs via electrostatic attraction and pore trapping
[73]. However, physiochemical approaches transfer rather than eliminate BCs. Studies have paid little attention to the transferred and concentrated microbial loads during downstream procedures or in receiving environment.
Disinfection provides the endpoint protection of biological safety. The most widely used disinfection method is chlorine (sodium hypochlorite) disinfection, which offers the competitive costs. In China, the application ratio of chlorine reached 87.7% in 56 wastewater treatment plants, according to the recent statistics
[74]. UV disinfection is recognized for its short contact time and high inactivation efficiency and is typically combined with sodium hypochlorite to ensure persistent disinfection. Over 3000 wastewater treatment plants in China have adopted UV disinfection, with the total treatment capacity exceeding 160 million tonnes per day
[75]. Ozone disinfection, which has a broad antimicrobial spectrum, is limited by the on-site generation requirements, and lack of residual disinfection efficacy. However, because of its high energy consumption and operational complexity, it is less commonly used in large-scale wastewater treatment facilities
[76]. Conventional disinfection methods require higher doses to successfully remove persistent BCs, even exceeding the actual doses by 10–100 times for intracellular ARGs and protected cells
[67],
[77].
The effectiveness of conventional disinfection is significantly reduced in the presence of complex water sources. The typical dosage of free chlorine in wastewater disinfection is 5–25 mg·L
−1, decreasing to 1–5 mg·L
−1 in drinking water
[68]. Nitrogen-containing substances in wastewater significantly consume free chlorine, converting it into organic chloramines and reducing the disinfection efficiency by 2–3 orders of magnitude
[78]. Owing to the corrosiveness of chlorine, high dosages increase the maintenance work required for equipment and pipelines. The penetration of UV light is influenced by colored substances and suspended solids, leading to fluctuations in the treatment efficiency. The recommended UV radiation dose in wastewater is not less than 80 mJ·cm
−2 (twice that for drinking water treatment)
[68]. Impurities such as iron and manganese ions accelerate the scaling of UV lamps (especially after the coagulation units), generally decreasing their inactivation efficiency and lifespan. The dosage of ozone applied in wastewater is much higher than in drinking water (5–15 vs 1.5–3 mg·L
−1), mainly limited by the mass transfer process, with the effective dosage sufficiently fluctuating with varied organic load
[79].
Trade-offs must be considered between residual disinfection effects and ecological toxicity. Wastewater quality standard requires the residual chlorine to be lower than 0.5 mg·L
−1 (DB11/307–2013) for direct discharge. Additional dechlorination facilities are necessary, which weakens the cost advantage of chlorine disinfection. Chlorine also reacts with natural organic matters and produces disinfection byproducts (DBPs). Although DBPs are not yet regulated for wastewater discharge, they potentially increase the ecological risk of receiving waters and ultimately affect human health. Take the trichloromethane for an example, a 10 mg·L
−1 dosage of free chlorine in wastewater produces about 60 μg·L
−1 of trichloromethane, already reaching the toxicity threshold for aquatic organisms
[80],
[81]. Elevated ozone dosage also led to the generation of high concentrations of aldehydes and bromate.
Increasing the doses of UV and ozone results in significant cost and maintenance challenges. The annual operating costs of UV radiation are mainly derived from electricity consumption, which is linearly related to the UV dose
[68]. When the production capacity of the ozone generator is increased ten-fold, the fixed equipment investment and maintenance costs increase by a factor of 6.3 times
[79].
2.2.2. Drinking water treatment
Prior treatment with coagulation, sedimentation, and filtration can reduce microbial load, although the removal efficiency is unstable. Physicochemical processes mainly remove larger protozoa and bacteria attached to particles but perform less effectively against free bacteria, and they have no impact on the diversity of microbial communities
[82]. Common disinfection methods for drinking water include the use of chlorine (sodium hypochlorite), chlorine dioxide, UV, and ozone. Chlorine disinfection remains predominant in small- to medium-scale drinking water treatment plants, with over 50% occasionally employing chlorine dioxide as a supplement to reduce formation of DBPs
[83]. On-site preparation of chlorine dioxide faces operational challenges in terms of reaction temperature control, waste liquid separation, and dosing methods, which require higher technical skills
[84]. UV disinfection has been applied in over 60 large- and medium-scale drinking water treatment plants in China, with the total treatment scale of exceeding 10 million tonnes per day
[85]. Although co-control with residual chlorine is still required, UV disinfection significantly reduces the use of chemicals and by-product generation, eliminating transportation and handling costs. UV irradiation has a well removal effect on chlorine-resistant
Cryptosporidium and seems perform better at high-energy wavelengths (e.g., 222 nm)
[86]. Ozone is typically applied in drinking water treatment as a catalytic oxidation procedure before activated carbon filters rather than as a terminal disinfectant
[87].
There is a large discrepancy between the disinfectant dosage required to remove persistent BCs and the actual applied dosage, especially in the case of UV radiation. For instance, the effective UV dose required to destroy intracellular ARGs (> 500 mJ·cm
−2) far exceeds the recommended value of 40 mJ·cm
−2. Although higher standards for biological safety in drinking water treatment are desired, increasing DBPs become unavoidable. High-dose UV radiation does not directly initiate DBPs; however, the low-molecular-weight components generated through photolysis significantly promote the formation of chlorine DBPs in the distribution network
[88]. Another challenge is that residual clorine can exert co-selection pressure on microbial communities, enriching microorganisms with strong resistance (typically human pathogens) and promoting the transfer in distribution systems
[7],
[89].
To satisfy the increasingly stringent demands for BCs control, we propose developing a highly-efficient, persistent, anti-interference, and deep-level disinfection paradigm. Specifically, the new disinfection system should ➀ shift from antibacterial targets to antimicrobial targets, successfully inactivate microbes including indicator bacteria, pathogenic bacteria, viruses, fungi, and protozoa; ➁ maintain effective from inside the treatment plant to the outside distribution network; ➂ regardless of impacts from environmental impurities; and ➃ eradicate all BCs of both culturable and unculturable, including the intracellular biomolecules.
3. EATs for BCs control
3.1. Methods
3.1.1. Literature collection
Firstly, we used “Web of Science” and “Google Scholar” as the retrieval database, searched the keywords “disinfection or inactivation or antimicrobial or antibacterial” and “drinking water or wastewater,” and defined the time period from 2013 to 2023. We then downloaded the top 10 000 records sorted by relevance to perform cluster analysis using the VOSviewer software (Leiden University, the Netherlands;
Fig. 1(a)). We identified the most frequent and recently occurring keywords and calculated the average publication years for the manually screened categories (Leiden University, the Netherlands;
Fig. 1(b)). Therefore, advanced categories for antimicrobial fields were selected and focused on nanotechnology, AOP, biological control, and their integrated approaches (
Fig. 1(c)).
3.1.2. Qualitative analysis
Herein, only literature meeting the following criteria were selected (Table S3 in Appendix A).
(1)Peer-reviewed publications.
(2)Incorporate data revealing the dynamics of BCs removal in water matrices, including at least one of the following water types: ideal matrices (PBS solution, ultrapure water, saline, culture medium, tris(hydroxymethyl)aminomethane hydrochloride buffer (tris-Cl), or synthetic wastewater), drinking water, and wastewater.
(3)The incorporated experimental setup can be traced back.
On this basis, we qualitatively summarized the feasibility of using EATs for wastewater and drinking water disinfection, focusing on whether these EATs were able to address the challenges faced by conventional disinfection methods in terms of ➀ highly efficient microbial inactivation, ➁ persistent disinfection, ➂ countering environmental interference, and ➃ deep-level removal of intracellular BCs. Microbial inactivation was evaluated by comparing the overall inactivation rate (IR) and electrical energy per order (EEO), calculated using Eqs. (1), (2) (the detailed calculation methods are presented in Text S2 in Appendix A). One-way analysis of variance (ANOVA) was performed using GraphPad Prism 9.0 (GraphPad Software, USA) to compare the inactivation rates of EATs with those of conventional disinfection (*p < 0.05). The qualitative results are discussed in Section 4.1.
Where,
EEO is the electrical energy required to inactivate the pathogens by 1 log (kW·h·m
−3),
P is the power consumption of the technology (kW), and
t is the treatment time (h). As is widely accepted, the cost of disinfectants is also converted to indirect electricity and incorporated into
Pt based on manufacturer data (Text S2) and the same electrical charges (0.04 USD per kilowatt hour)
[90].
V is the treated water volume (m
3),
C0 is the initial pathogen concentration, and
Ct is the pathogen concentration at treatment time
t.
3.1.3. Semi-quantitative framework
We developed a three-tier hierarchical analytic framework (
Fig. 2 and Text S3 in Appendix A) to quantify the application potential of EATs, which allows us to determine the dominant superiority and shortcomings of monotechnology compared to conventional methods. At the target level, we presupposed two application scenarios: wastewater and drinking water disinfection. At the index level, we selected ➀ IR value, ➁ growth inhibition, ➂ ARGs removal rate, ➃ human health, ➄ global warming, ➅ operating cost, and ➆ freshwater ecotoxicity as essential indices, synthetically considering the trade-offs among BCs removal efficiency, techno-economics, and health-related issues. Their assigned coefficients were determined through expert scoring, which evaluated the pairwise importance of different indices in wastewater and drinking water scenarios. Experts engaged in environmental engineering were investigated based on the classic AHP 1–9 scale, and ten results that passed the consistency test were considered to form an average judgment matrix
A (
Table 2). The detailed procedure is provided in Text S3.
At the scheme level, we selected one competitive technology from each of the four emerging categories based on qualitative analysis, with Schemes 1–4 as immobilized Cu nanocomposites, non-thermal plasma, phage-based treatment, and photocatalytic nanocomposites (Ag/TiO
2/graphene oxide), combining conventional Schemes 5–7 as chlorine, UV, and ozone. Point-estimated indices for each scheme were obtained from previous literature. Among these, Index 1 refers to the immediate inactivation efficiency in nutrient-free media, determining whether EAT is “highly efficient.” The “persistent” and “anti-interference” goals are reflected by Index 2 by monitoring the long-term inactivation rates in culture media, whereby sufficient nutrients are provided for microbial growth and oxidants scavenging. Index 3 weighs the goal of “deep-level,” evaluating the removal effects of intracellular BCs. Index of construction cost is not included due to the lack of reliable normalized parameters for the conversion of treatment capacity (USD·m
−3). The specific estimation methods and results for the seven indices are summarized in Text S4 in Appendix A. The corresponding normalized
WB and judgment matrix
B are listed in
Table 3. Finally, the comprehensive weights
W considering the efficacy indices, cost indices, and scenario requirements were calculated using Eq. (3). The detailed procedures can be found in Text S4. The entire semi-quantitative procedure is described in Text S3 and the results are discussed in Section 4.2
3.2. Nanotechnology
3.2.1. Proper niches in wastewater/drinking water treatment
The application of antimicrobial nanomaterials in water treatment involves the direct addition of nanomaterial powders or their immobilization on macroscopic carriers (such as filters, resins, and magnetic minerals). Nanopowders that are considered feasible for direct addition should exhibit negligible ecological effects. However, they remain controversial and rarely progress to large-scale treatments. The typical antimicrobial components include graphene, chitosan, Cu, Zn, and Fe (Table S4 in Appendix A). Metal-based nanomaterials exhibit higher antimicrobial efficiency than carbon-based materials, ranking in the order Cu > Zn > Fe
[91]. Tunable microscale properties shaped by different synthesis methods also affect the antimicrobial efficiency of free nanoparticles (NPs). In general, a smaller particle size, larger surface area, rough morphology (with sharp corners or piercing edges), and positive surface charge are beneficial factors for antimicrobial practice
[92],
[93],
[94],
[95]. Nanopowders exhibit high stability for hours, months, or even years
[96], eliminating the need for on-site preparation and allowing convenient injection into existing pipelines. Moreover, NPs perform anti-interference disinfection through physical interactions with microbial cells. NPs with sharp edges can aggregate and adsorb bacteria through electrostatic forces, causing shading effects and contact injuries (illustrated in Text S1), which are less affected by complex matrices in wastewater. Unlike chlorine disinfection for drinking water treatment, the physical effects of NPs do not induce new inheritable resistance and limit DBPs production
[97]. In wastewater treatment, nanopowders can be added before tertiary coagulation, causing them to settle and separate after disinfection exposure, and the NPs enriched in the sludge require further disposal. Laboratory trials have shown that NPs can be reduced from mg·L
−1 to μg·L
−1 by enhanced coagulation
[98]. For drinking water treatment, the interception of free NPs may rely on advanced membrane filtration
[99]. However, there is currently no available research investigating the potential leakage of NPs. The integration of magnetic cores (e.g., Fe
3O
4) into nanopowders enables magnetic separation and recovery. Magnetic NPs in small-scale batch reactors (several liters) can be separated using permanent magnets and recycling for 3–10 cycles. Researchers have envisaged that magnetic NPs can be separated in full-scale water treatment with the assistance of high-gradient magnetic separators. However, limitations on energy consumption and separation efficiency remain challenging
[100].
Freely suspended NPs can be immobilized on macroscopic carriers through facile physical chemistry reactions, including crosslinking, adsorption, hot pressing, coating, and three-dimensional (3D) printing
[101],
[102],
[103]. Using immobilized nanomaterials is a more conservative approach to avoid leakage, but may partially sacrifice the nano-scale internalization effects and exposed active sites
[104]. From this perspective, Ag components, which exhibit the strongest inactivation efficacy in the free state and have higher release risks
[91], are more suitable for antimicrobial applications involving immobilized forms. Some small-scale or pilot experiments have immobilized Ag on various commercially available substrates such as clay
[105], silica
[106], resin
[107], and activated carbon
[108] to construct fixed-bed reactors that successfully inactivate indicator bacteria under continuous and intermittent operation. An earlier study constructed resin filter columns coated with nanosilver and demonstrated an effective inactivation period of over 30 h in drinking water (flow rate of 2 L·min
−1)
[107]. However, the concentrations of released silver ions or silver NPs were not measured. Under intermittent operating conditions, the inactivation rate is significantly influenced by the loading amounts, hardness (Ca
2+ and Mg
2+), and organic backgrounds, with contact times ranging from 1 to 3 h, and silver release within acceptable limits (< 21 µg·L
−1)
[106]. Fixed nanomaterials can be integrated with existing processes for full-scale drinking water or wastewater treatment, including sand filtration, activated carbon filtration, and adsorption.
3.2.2. Current challenges for wastewater/drinking water treatment and improving strategies
The free-state and fixed forms of nanomaterials may be feasible for both wastewater and drinking water matrices. However, scale-up cannot be implemented since there is a trade-off between nanodoses and antimicrobial efficiency, as well as the release effects.
In practical water treatment, contact between nanomaterials and microorganisms is significantly limited by the mass transfer kinetics and aggregation of NPs. The dispersibility of nanomaterials is positively correlated with their hydrophilicity
[109]. In mechanical/magnetic stirring reactors, previous studies utilized surface modification with substances such as cyclodextrin, chitosan, polylactic acid-
co-ethylene copolymers, and polyvinyl alcohol to enhance the dispersibility
[110]. However, most metallic nano-approaches face the dilemma of pursuing antimicrobial efficiency and controlling the release risks. The leaching of metal ions and small NPs from metallic nanopowders significantly threatens human health and ecological safety and indirectly promotes ARGs transmission
[111],
[112]. For drinking water systems, the concentrations of heavy metal ions such as Ni, Cr, and Ag are strictly regulated (not exceeding 0.5 mg·L
−1), while ions like Cu, Zn, and Al generally do not exceed 1.0 mg·L
−1. Under the same material conditions and identical physicochemical properties, the order of the van der Waals forces between particles is generally Au < Ag < Fe
2O
3 < ZnO < SiO
2 [113]. Given the significant aggregation interference, low-toxicity metal cores, such as Cu and Zn, typically have effective antimicrobial doses of 10
2–10
3 mg·L
−1 (Table S3), which are prone to exceeding the limits for released Cu
2+ and Zn
2+. CuO doses of no more than 40 mg·L
−1 are suitable for drinking water disinfection
[114]; however, the limit can still be overestimated. Compared to free-state nanopowders, immobilizing NPs onto macrocarriers reduces the leaching of harmful ions and controls the aggregation of NPs
[115]. However, released silver ions, even when below standard limits, can trigger health concerns owing to their cumulative effects.
For wastewater treatment, nanocoatings in wastewater are more subject to the disruption of complex pollutants, but the antimicrobial reusability, regeneration methods, and associated techno-economic issues are scarcely discussed.
3.3. AOP
Oxidation-based mechanisms are suitable and unique for the irreversible destruction of BCs (Text S1). Herein, we mainly focus on emerging methods for producing highly reactive oxygen species (ROS), which have been extensively studied for organic removal
[116], while still being explored for their disinfection potential.
3.3.1. Proper niches in wastewater/drinking water treatment
Electrochemical oxidation: Relying on dual-electrode systems with stirring or flow-force homogenization, the current method electrolyzes inherent precursors in wastewater and drinking water to generate highly reactive oxidants. The most frequently utilized oxidants in electrochemical system are reactive chlorine species (RCS), as they have the uniquely present precursors (Cl
−) in water environments with lower redox potentials (−1.40 V)
[117]. Compared to conventional chlorine disinfection, electrogenerated RCS exhibits a higher antimicrobial capacity. Electrolysis of water molecules can also generate free ROS, including ·OH, O
3, and H
2O
2. The active anodes form an adsorbed oxidation surface with ·OH, directly contacting and damaging BCs
[118]. ROS-mediated electrooxidation consumes more energy than RCS owing to confined oxidation and lower charge efficiency
[117].
Electrochemical oxidation requires the establishment of a dedicated contact apparatus that is independent of the existing terminal disinfection, occupies a small space, and features facile operation and maintenance. The main factors controlling the electrooxidation process are the input voltage, current, and electrode materials. As shown in Table S3, the current density used for controlling the BCs can be as low as 2 mA·cm
−2 and does not exceed 100 mA·cm
−2, because a current higher than 50 mA·cm
−2 results in rising solution temperatures
[119]. The voltage provided by the direct current (DC) power supply ranges from to 0–30 V. Emerging anode materials, including boron-doped diamond and titanium-doped materials such as Ti/Sb and Ti/IrO
2, exhibit efficient ROS generation, wide potential windows, low impurity adsorption, and corrosiveness
[118]. The effectiveness of electrooxidation is also affected by various external factors, including pH, temperature, ionic strength, and organic matter. When the conductivity of treated water is low, the required energy increases significantly because of the constricted current
[120]. In a pilot-scale study, electrochemical disinfection was used to treat rural drinking water without the addition of electrolytes. The current density was elevated to an impractical range (250–500 mA·cm
−2); however, ensuring compliance with coliform standards under continuous-flow conditions remains challenging
[121]. Exogenous electrolytes in drinking water are prone to exceed inorganic ion limits and initiate more toxic chlorates and perchlorates
[122]. The only full-scale trial adopted an indirect approach, electrolyzing the ideal NaCl solution for the on-site production of RCS
[123]. The replacement of Cl
2 gas with mixed electrolysis products has operational cost advantages for drinking water treatment
[123]. Electrooxidation is more appropriate for disinfecting wastewater that already contains sufficient electrolytes, whereas the scale-up of electrochemical devices, especially the standardization of parameters, including electrode size and fluid dynamics, remains a subject of debate
[124],
[125].
Non-thermal plasma: Non-thermal plasma dissociates the working gases (air, He, Ar, and H
2) at lower power inputs under atmospheric pressure, generating various reactive species (e.g., ·OH,
1O
2, O, N atoms, H
2O
2, O
3, NO
3−, and NO
2−) and releasing UV radiation at the air–water interface
[126]. Among them, the roles of ROS and reactive nitrogen species (RNS) have been highlighted
[127], whereas the contribution of incidental UV radiation is always negligible. Non-thermal plasma has three applications: direct contact dissociation, gas–water interface dissociation, and indirect use of activated plasma carriers
[128]. In a reactor designed for contact dissociation, dual electrodes are directly inserted into water, and the reactive species preferentially interact with the BCs near the electrodes. The working electrode for the gas–water interface dissociation mode is placed above the water, and the reactive species diffused downward from the water surface. Another variant involves dissociating the working gas in a pre-reactor to generate a plasma airflow, which is then loaded onto the surface of water or macroscopic carriers or directly introduced into the water to be treated.
For drinking water treatment, plasma does not require exogenous chemicals and is applied within very short periods, simultaneously removing trace pollutants and improving sensory indicators such as odor. However, it introduces NO
3− and NO
2− when air is used as the working gas. In an Ar/air plasma system, up to 5 mg·L
−1 nitrate can be produced in 20 min, which further increases to 113 mg·L
−1 with the addition of NO to working gas
[129], markedly exceeding the limit in drinking water (10 mg·L
−1 nitrate, according to GB5749–2022). Although rarely discussed, this may lead to the formation of nitrogenous DBPs, and caution is advised in the future. Common small-scale treatments (< 1 L) result in a significant decrease in pH after treatment
[130]. Overall, non-thermal plasma is not suitable for direct use as an endpoint disinfection method for drinking water. Water indirectly activated by plasma can be temporarily stored for 1–3 h and used for surface disinfection of fruit and vegetables
[131]. When scaling up, the partial return and dilution of indirectly activated drinking water may address the negative impacts related to NO
3−, NO
2−, and acidic pH, providing persistent disinfection efficacy to some extent
[132]. This operating mode can be used as an alternative for endpoint disinfection and provides continuous disinfection in pipeline networks, although there is no available literature.
In wastewater treatment, in addition to being employed for end-point disinfection, plasma can be used for pre-oxidation before biological treatment. The oxidation products of plasma promote the growth of downstream microorganisms (e.g., secondary activated sludge and tertiary biofilters) and plants (e.g., constructed wetlands and irrigation reuse) with appropriate supplementation of nitrogen sources
[133]. However, the energy consumption of plasma is significantly higher than that of conventional disinfection, and the relationship between the discharge power and amplified treatment volume still requires optimization.
Photodynamic reaction: ROS are the core substances for photodynamic therapy, with photosensitizers (Table S5 in Appendix A) serving as inducing agents. The heat dissipated from the photodynamic reaction may synergistically shock the bacteria, increasing their susceptibility to other antimicrobial agents
[134]. Natural photosensitizers are sourced from plant pigments, secretions, or bacterial culture products and have negligible environmental impacts. Crude extracts of natural photosensitizers are affordable (less than 50 USD·kg
−1) and readily available, offering prospective pathways for enhancing solar disinfection within minutes to hours
[135]. Second- and third-generation synthetic photosensitizers have been developed to promote hydrophilicity, ROS production, and near-infrared adsorption (750–1700 nm) for clinical applications. However, their prices far exceed the acceptable ranges for water treatment. Natural photosensitizer extracts are suitable for the disinfection of wastewater and decentralized emergency drinking water. It may lead to color issues and a potential increase in DBPs for large-scale drinking water treatment. The most economical mode is to enhance the removal of BCs in backend open-air treatment facilities such as artificial wetlands, which utilize natural light. Research can also focus on optimizing ROS production through synergism between natural photosensitizers and ultraviolet C (UVC) irradiation.
3.3.2. Current challenges for wastewater/drinking water treatment and improving strategies
AOP are applicable in both full-scale wastewater and drinking water treatments as enhanced disinfection methods, serving at pre-oxidation or advanced treatment stages. They exhibit highly efficient and broad-spectrum disinfection performance, successfully eliminating Cl-resistant protozoa
[136], UV-resistant spores
[137], and viruses
[137], including their intracellular ARGs and endotoxins, as well as their VBNC states.
However, most AOP-based disinfection methods are non-targeting and confined, indicating that other chemical pollutants pre-scavenge the reactive species and attenuate disinfection efficiency. Increasing oxidant doses or energy input is necessary to guarantee the BCs removal capacity, albeit against cost and sustainability constraints. Additionally, AOP can hardly provide residual disinfection in the pipeline network because ROS represented by ·OH and 1O2 have short lifespans. Although long-lived species, such as H2O2, HNO3, and HNO2 provide persistent disinfection efficacy, they cause severe pipe corrosion. Therefore, residual chlorine needs to be added when AOP are used for endpoint disinfection in drinking water systems. For wastewater treatment, it is essential to investigate the threshold doses of reactive species for the control of persistent BCs and restriction of pathogen regrowth. Given the co-occurrence of complex backgrounds, we outline additional mechanistic studies to explore the reaction path, intermediate byproducts, and holistic ecological effects of emerging AOP.
3.4. Biological control process
3.4.1. Phage-based antimicrobial process
In clinical trials, bacteriophages (abbreviated as phages) are revived as the “last line of defense” against multidrug-resistant bacteria
[138], which has also inspired environmental researches. In contrast to the aforementioned nanotechnology and AOP, phages present a highly selective antimicrobial method at the strain-level
[139]. They are capable of precisely splitting target host cells via natural parasitic processes without significantly affecting other bacteria, viruses, fungi, and protozoa, even amplifying ARGs
[140]. As bio-agents, high concentrations of phages are obtained, separated, and purified from co-cultures with host bacteria. In laboratory-scale treatments, phage storage is directly added to artificially contaminated water, where phages take only 15–20 min to complete one lytic cycle and a few hours to inactivate higher than 3 log bacteria
[141],
[142]. The plaque counts of phages decrease with a decline in hosts, and no significant health risk from residual phages has been validated to date.
In actual wastewater and drinking water treatment, the addition of free phages is challenged by low lysis efficiency. Environmental factors, such as pH, suspended impurities, inorganic ions, and nutritional conditions, affect the adsorption and invasion of phages. To overcome the inherent bacterial resistance and adaptive costs (Fig. S2 in Appendix A), application scenarios with higher host loads and phage doses (generally, phage:host > 10:1 and initial phage > 10
6 CFU·mL
−1) are necessitated
[142]. In microenvironments containing high bacterial concentrations, such as filter biofilms, membrane fouling, and activated sludge, free phages have been successfully applied to control the excessive growth of foam-forming bacteria
[143],
[144],
[145]. As shown in
Table 1, the concentrations of pathogenic bacteria in the secondary effluent were approximately 10
4 CFU·mL
−1, and the total bacterial load in the drinking water was less than 10
3 CFU·mL
−1. Phage storage can slowly lyse bacteria in wastewater but fails for endpoint disinfection in drinking water. Phage cocktails and polyvalent phages have been isolated from natural environments as potential allies. They can simultaneously target multiple host bacteria, thereby reducing survival costs. For instance, polyvalent phage cocktails preferentially utilize their production host to accumulate titers, thereby rapidly capturing more elusive pathogenic hosts
[146]. Emerging drug delivery methods in the medical field that encapsulate phages in liposomes and alginate polymers may further preserve their lytic capabilities
[147].
Another important challenge is the remained dormant hosts. Phages cannot effectively lyse all host bacteria and randomly induce a portion of the bacterial population to enter the dormant state, potentially leading to regrowth and the evolution of more resistant communities
[148]. Even when used in wastewater, a combination with conventional disinfection is required to meet water quality standards (mainly discussed later as an integrated technique). Except for a newly identified phage of
Pseudomonas aeruginosa, it is capable of lysing dormant cells and exhibits the potential for deep-level inactivation of VBNC bacteria in wastewater
[149].
Phage-based treatment is applicable as an endpoint of wastewater treatment and specifically targets human pathogenic bacteria. Strategies, such as the use of phage cocktails, polyvalent phages, and dormant-overcoming phages, can be employed to enhance antimicrobial efficiency.
3.4.2. Genetic engineering based antimicrobial process
Toxin–antitoxin (TA) systems: The free-living bacterium inherently possesses “suicidal” toxins that can be neutralized in an unstable way by the expression of antitoxin
[150]. TA sequences are commonly identified in the chromosomes and plasmids of bacteria but do not overlap with the human genome
[151]. Their regulation is closely associated with bacterial tolerance, making the targeting of TA complexes an attractive, tailored antimicrobial approach. This “suicidal” effect has been artificially activated by disrupting antitoxin synthesis and the ectopic expression of toxin genes, mainly to combat persistent bacteria
[150],
[152],
[153].
Using plasmids as carriers to deliver toxin genes requires a high bacterial density and contact time for plasmid transformation and expression in environmental microbial communities. Only one study tentatively introduced plasmids containing TA loci (
pNJR6 plasmid carrying the
susB gene) to inactivate
Elizabethkingia meningosepticum in wastewater
[154]. The process involves no addition of chemicals or harmful byproducts and, to a certain extent, achieves the targeted inactivation of pathogenic bacteria. Nevertheless, the inactivation efficiency was dramatically low (< 100%, even after 72 h), presumably because of the low transformation efficiency of the plasmids. Moreover, the application of TA systems as antibacterial agents is still premature because their long-term roles in bacterial dormancy, biofilm formation, quorum sensing, and other physiological processes are still debatable
[155].
Clustered regularly interspaced short palindromic repeats-associated systems (CRISPR/Cas systems): CRISPR systems serve as natural defense barriers developed by bacteria during their evolutionary competition with phages, specifically recognizing and breaking foreign genetic material
[156],
[157]. Artificial replacement of target sequences in CRISPR spacers facilitates the gene-level control of pathogens and ARGs
[158]. During water treatment, CRISPR genes must be delivered through biological carriers such as phages, plasmids, and live bacteria. Similarly, high concentrations of CRISPR carrier complexes are necessary for large-scale applications. A recent groundbreaking study achieved the efficient removal of target ARGs (100% elimination within 3 h) from wastewater using the donor
E. coli carrying CRISPR plasmids, with minimal energy input required for aeration
[159]. Bacterial conjugative transfer provides a higher transformation efficiency than phage transfer in wastewater, although it requires extra separation of the host bacteria. However, this gene-level disinfection can be a laborious task for various ARGs, pathogen subtypes, and changeable sequences unless there are shared antimicrobial resistance or virulence sequences among different BCs.
Genetic engineering-based treatments are still at the conceptual stage, with only a few laboratory experiments proving their feasibility in wastewater matrices. It targets pathogenic microorganisms and ARGs with minimal effort. However, the potential ecological effects still require long-term monitoring. Trials in drinking water are limited because of the lower microbial loads and, thereby, lower expression efficiency.
3.4.3. Biological intervention based on macroscopical ecology
Biological interventions based on macroorganisms have been implemented in full-scale wastewater treatment with virtually no application in drinking water treatment. The initial design objectives of these technologies were not focused on the removal of BCs, but on incidentally reducing some pathogens. The constructed ecosystem is environmentally and economically friendly with no requirement for additional chemicals.
Microalgae: Microalgal culture and the symbiotic system of fungi and algae inactivate pathogenic bacteria, viruses, and protozoa to varying degrees (< 1–5 log) after an operational cycle of 3–7 d
[160]. Apart from competition, poisoning effects, and occasional predation by microalgae, the optimized light irradiation, pH, temperature, dissolved oxygen, and hydraulic retention time required for algal production are detrimental to the survival of pathogens
[161]. Algal-bacterial consortia significantly reduce exogenous ARGs, which is attributed to ARGs captured by extracellular polymeric substances and intracellular transcription interference via certain DNA self-protection mechanisms
[162]. However, there is limited research on the association between removal efficiency of BCs and regulation of the culture parameters, including light intensity, pH, and nutrient supply. It is recommended that operational parameters be optimized with regard to the potential synergism between BCs removal and microalgal culture, potentially assisted by advanced forecasting models
[163].
Earthworm filter: Earthworms exhibit cross-linked ecological relationships with protozoa and bacteria (natural predation, parasitism, and symbiosis), partially eliminating pathogenic bacteria, fungi, protozoa, and ARGs within one operational cycle lasting several hours to days
[164]. In addition to the toxic effects of earthworm coelomic fluid, diverse communities of competitors and predators of pathogenic bacteria have been screened from earthworm biofilms without specific identification at the genus or species levels
[165]. Earthworm load is a crucial performance parameter affecting BCs removal and is closely linked to water quality parameters (i.e., chemical oxygen demand, biochemical oxygen demand, and seasonal temperature)
[166]. Operationally, it is influenced by factors such as the type of filtration medium and the thickness of the filter bed. Pilot studies indicated that riverbed gravel is more favorable for the removal of pathogenic bacteria than mud balls, coal, and glass balls, and the removal rates surpass the reported general ranges
[167].
Earthworm biofilms degrade a large portion of organic pollutants, but increasing co-exposure to chemical pollutants (e.g., antibiotics and heavy metals) enhances the prevalence and transferability of ARGs in the earthworm gut
[168]. Earthworm biofilms, which serve as a long-term domesticated microenvironment, should be prevented from becoming a new ARGs reservoir.
3.5. Integrated techniques
The above EATs can address some (mostly not all) of the unresolved challenges during conventional disinfection but often require higher energy or chemical inputs. To fulfill the four aspects of the disinfection diagram proposed in Section 2, the integration of multiple barriers with different antimicrobial mechanisms represents a mainstream strategy (as demonstrated by the interconnected network of keywords in
Fig. 1(b)). In particular, the synergy between EATs and conventional treatment enables deep-level elimination of BCs without massively upgrading current facilities (
Fig. 3).
3.5.1. Photo-, sono-, and electro-catalysis nanomaterials
Existing wastewater treatments inherently provide energy imports, including hydraulic energy, aeration, homogenization mechanical energy, and natural light irradiation in follow-up semi-open facilities or ponds, as well as energy-concentrative UV light in endpoint disinfection. Drinking water treatment also offers hydraulic energy, mechanical stirring, and UV irradiation energy, but natural light is less available. The tunable characteristics of nanomaterials (mostly semiconductors, as shown in
Fig. 4) take advantage of these energy sources to enhance external ROS production, further increasing their applicability in wastewater and drinking water treatment.
Light-driven nanotechnology provides green and energy-efficient options that can be combined with flexible settings. Electrons are excited by light absorption from the valence band to the conduction band, leaving holes as reactive centers for rapid ROS initiation. TiO
2 is the most extensively studied heterogeneous photocatalyst, with the advantages of low toxicity and strong commercial availability. However, pure TiO
2 has a wide bandgap (3.9 eV), leading to a limited absorption range (only in the UV region) and rapid electron-hole recombination. Improvements have been proposed, including doping with metallic or semiconductor materials and cross-linking with other photosensitizers, to develop the catalytic capacity of TiO
2 under visible light
[169]. Metal-based nanomaterials have abundant charge carriers on their surfaces, which synergistically enhance the reactivity of photogenerated electrons and holes. In addition to photocatalysis, inorganic nanomaterials such as Au, MoS
2, CuS, Pd, transition metal carbides/nitrides (MXene), and carbon-based nanomaterials possess photothermal properties. An increase in the local temperature promotes photocatalytic reactions and provides heat shock to BCs. Photocatalytic nanomaterials are typically used in immobilized forms, rather than as directly injected nanopowders, for end-point disinfection in the treatment of wastewater and drinking water, either in continuous flow or intermittent operation. The color and turbidity of wastewater obstruct the transmission of natural light. To enhance antimicrobial efficiency, tailored reactors with enhanced photon flux have been designed, with the widely used representative of compound parabolic collector reactors
[170].
Under ultrasound (US) treatment (20 kHz–3 MHz), sonosensitive nanomaterials can be well pre-dispersed in an aqueous matrix, primarily absorbing the vibrational energy and dissipated energy from cavitation to induce ROS (assisted by piezoelectric effects)
[171]. Generally, semiconductors are also endowed with sonosensitive properties (e.g., ZnO, TiO
2, MoS
2, and ferroelectric ceramics)
[172]. Acoustic cavitation is an emerging water disinfection technique that has been applied to drinking water and wastewater
[173]. Mono-US treatment leads to the rapid growth and violent collapse of bubbles, disrupting pathogens through mechanical, thermal, and chemical effects
[174]. However, the disinfection efficiency of mono-US seems insufficient, requiring longer treatment times (> 60 min) and higher energy inputs, hindering its large-scale application
[174]. Therefore, the incorporation of sonosensitive NPs enhances cavitation disinfection. For instance, an optimized TiO
2 nanocomposite-US system achieved over 90% inactivation of
Staphylococcus aureus after only 1 min treatment
[175]. This integration takes advantage of the low sensitivity to interference from complex backgrounds; however, high-frequency US requires additional fixed equipment to locally apply high-energy input to the existing pipeline
[176]. Low-frequency energy inputs (such as mechanical aeration, stirring, and water flow) can also induce nanocatalytic activity. Successful antimicrobial practices have been implemented by the direct addition of nanopowders to wastewater, increasing hydraulic and residence times
[177]. To optimize the utilization of mechanical water flow, innovative application forms of NPs involve embedding them into pipe materials or coating them as inner linings
[178]. For drinking water treatment, confined zone-separated semi-permeable membranes have been proposed at the end of the pipeline to avoid leakage of nanopowders while transmitting the reactive species
[177]. However, these catalytic nanomaterials can lose their catalytic activity within several hours
[179], which poses a vital challenge that requires urgent resolution for future applications.
Semiconductor-like nanomaterials have also been used as electrodes in nanocoatings, nanosheets, and nanowires. Dense networks composed of metal-based nanowires (e.g., Ag, Co
3O
4, CuO, and ZnO) and carbon-based nanomaterials (e.g., graphene) have highly efficient electrode morphologies and are harnessed in antimicrobial processes dominated by electrocatalytic trapping, electrooxidation, and electroporation
[180]. Nanoelectrodes offer the main advantages of improved effective contact area, discharge efficiency, and anti-interference capacity. Specifically, nanowires concentrate the charge density at their scattered tips and create localized electric fields, performing ultrafast antibacterial electroporation (nanosecond to millisecond) at low voltages (alternating current < 10 V)
[181].
3.5.2. Hybrid membrane process
The hybrid membrane process applies a pressure-driven membrane as a base material, immobilizes functionalized materials (e.g., nanocatalysts, chemical oxidants, and phages), and thereby
in-situ eliminates BCs at the end-point treatment for both wastewater and drinking water systems. The most studied nanocomposition is Ag, followed by graphene. They are easy to coat or integrate onto membranes and their immobilization methods are relatively mature
[115]. Efficient BCs removal from the embedded Ag membrane is attributed to the release of Ag
+ [182], and thus strictly depends on the loading amount of Ag, requiring careful assessment in drinking water treatment. Furthermore, an increase in Ag loading limits the membrane retention capacity
[183]. By enhancing the surface localization of Ag rather than embedding it within membrane pores, the hindrance to water flux was partially addressed, and synergistically optimized antimicrobial performance was observed
[184],
[185]. Graphene-based nanocomposite membranes have the advantages of maintaining water flux and prolonging the membrane lifespan
[186], generally with concessions to antimicrobial effects compared to Ag membranes.
Nanocatalysts
[187], photosensitizers
[188], and strong oxidants, such as persulfate and peracetic acid
[189], are immobilized on membranes to enhance ROS production, indiscriminately removing BCs and other membrane-fouling substances. Typically, TiO
2-modified membranes display better BCs control and upgraded anti-fouling capacity under UV irradiation and are stable for multiple operation cycles
[187],
[190]. Photosensitive metal-coordinated porphyrin coatings also successfully degrade various hormones (up to 78% for 17β-estradiol) during ultrafiltration, and maintain persistent catalytic efficiency for nearly a month under outdoor sunlight conditions
[188].
The condensed biofilms on the membrane create a favorable microenvironment for phages in wastewater treatment, omitting some survival costs because of easier adsorption
[191]. Highly concentrated phage solutions can efficiently mitigate biofouling in ultrafiltration
[192] and membrane bioreactors
[193]. Compared to other functionalized materials, phages exert minimal detrimental effects on the membrane structure and more rapid inhibition of biofouling (3–6 h)
[194]. However, most trials are limited to conventional indicator
E. coli strains, and future culture-based experiments are outlined for differentially dominant bacterial species in the biofouling flora.
3.5.3. Phage-based combined process
Synergism exists between the sequential combinations of phages and conventional disinfection methods (solar, chlorine, and UV)
[193],
[195],
[196]. Monophage treatment downregulates the bacterial defense system (including cell wall protection, ROS scavenging, and DNA repair genes), thereby decreasing bacterial resistance to conventional disinfectants
[195]. Solar irradiation also activates functional genes related to phagocytosis, and the phage–solar integrated system significantly shortens the inactivation lag phase by 2 h
[195]. On this basis, we deduced that phages potentially enhance the pathogen removal of conventional wastewater/drinking water treatments as self-limiting supplementary disinfectants, and their plaque counts appear to decrease with the reduction of the host population in terminal waters
[197]. Moreover, phages can intrinsically invade the “active” host including VBNC
[198]. However, there is no direct evidence that phages can move this further when dealing with disinfectant-resistant bacteria and their persistent life forms in distribution networks.
To tackle the difficulty of phage survival in drinking water with low bacterial loads, study have immobilized phages onto carriers to create confined higher phage-host ratios and facilitate the phage-lysis process
[199]. Nanomaterials present an intriguing carrier option because their adsorption and self-propelling characteristics assist in spreading the infectivity of phages
[199]. Moreover, other functional carriers endow the integration of their “confined bursting” with the phages “precise identification.” Successfully combined weaponry includes genetically engineered products
[200], Au NPs
[201], commercial photosensitizers (e.g., Nile blue)
[202], and aggregation-induced emission agents
[203]. For this instance, a low host concentration is no longer the limiting factor because phages only need to locate and adsorb on their hosts instead of undergoing the complete lysis. Among them, ROS-induced weaponry inactivates both phages and hosts within a few minutes
[201],
[202], ultimately eliminating health concerns associated with residual phages in drinking water.
4. Toward application: Where we stand and the road ahead
4.1. Efforts toward more BCs removal
4.1.1. Promoting the microbial inactivation
EEO is an important parameter that incorporates the inactivation rate and cost and is widely accepted for comparing the disinfection efficiency of different technologies. For EATs, the ranking of median
EEO is as follows: biological control process < AOP ≪ integrated techniques < nanotechnology as demonstrated in
Fig. 5 and Fig. S4 in Appendix A. Miklos et al.
[204] examined the
EEO of AOP for removing chemical pollutants. They defined the processes with
EEO < 1 kW·h·m
−3 as feasible for full-scale application, consistently, the conventional disinfection processes show reasonable
EEO values lower than 1 kW·h·m
−3 (
Fig. 5). In this regard, some biological control processes and AOP also seem applicable with
EEO values comparable to or lower than those of conventional UV and ozone processes. Technologies with an
EEO range of 1–100 kW·h·m
−3 have a promising potential for future applications. The four EAT categories have branch solutions within this range.
The calculation of
EEO partly averages the impacts of inactivation rate and cost factor, where the solutions with extremely low costs but also exceptionally low inactivation rates may be considered “competitive;” however, this is not the case in practice. Therefore, we introduce another criterion to screen the “competitive” schemes: the inactivation rate should be greater than 0.1 min
−1 (
Fig. 5). In nanotechnology, metallic NPs exhibit applicable
EEO values at dosages of no more than 100 mg·L
−1 (Table S3). Among these, readily available Cu and Zn nanoparticles, can efficiently inactivate bacteria within minutes to hours. Although Ag NPs cause rapid bacterial inactivation, their high cost contributes to higher
EEO values. Biological control processes endow the lowest
EEO mainly because of their lower energy input. Their inactivation rates were also the lowest, and only the phage-based treatment exhibited a disinfection time (minute to hour) comparable to that of conventional disinfection. Competitive options for AOP include UV-based AOP, nonthermal plasma, and electrochemical oxidation. The electroporation process involves ultrafast disinfection, surpassing conventional disinfection techniques to inactivate more than 95%
E. coli within 20 ns
[181]. Non-thermal plasma in this assembly achieves promising inactivation rates of 10
1–10
2 min
−1, and thoroughly inactivates robust bacteria, regardless of the protection in biofilms or the VBNC state
[205]. Integrated techniques ensure inactivation rates under the reduced chemical or energy input, thereby lowering the
EEO to some extent. The optimized photocatalytic nanocomposites stand out with the treatment time not exceeding 30 min and the rate constants greater than 1 min
−1.
The
EEO primarily provides a method to compare different disinfection technologies from an economic perspective. Essentially, it considers energy consumption and inactivation rate, giving them equal weights. However, the relative importance of these two parameters is partially offset because they are directly divided. Additionally,
EEO is significantly influenced by water quality variations in different studies
[204]; therefore, a lower
EEO does not necessarily equate to a higher application potential.
EEO also does not encompass the subsequent environmental impacts of disinfectants and their byproducts, which is a critical aspect for the application of EATs.
4.1.2. Providing persistent and anti-interference disinfection
An institutional concept for next-generation disinfection is the development of a disinfectant that achieves rapid inactivation of BCs and simultaneously provides residual protection from inside the treatment plant to the outside distribution network without the addition of residual chlorine. This necessitates that emerging disinfectants persist in water for an extended period, neglecting complex background effects while having minimal environmental impact. Although ROS contribute to rapid microbial inactivation, they do not have persistent effects and pose a risk for VBNC regrowth. One alternative residual disinfectant is H
2O
2. Some nano-dominated catalytic processes generate highly reactive H
2O
2 as the main antimicrobial agent, exhibiting moderate but persistent antimicrobial effects (for hours or days) via self-decomposition
[206]. However, H
2O
2 corrodes steel pipes, which severely limits its application. Few studies indicate that phage-based and genetic engineering-based methods provide persistent disinfection effects owing to their consistent variability with the bacterial population. Phage-based treatment achieves complete inactivation of the free host within 14 h without any regrowth in the wastewater matrix
[207]. This persistent disinfection can be species- and scenario-specific and requires careful evaluation in terms of phage resistance and complex microbial interactions. Nanomaterials based on physical destruction and relatively reusable oxidants/catalysts can also ensure residual disinfection owing to their persistence. A case in point is the sharp Cu(OH)
2 nanocoating that tears up bacteria, which remains efficient in tap water and reclaimed water for up to 30 d
[208]. Additionally, the combination of multiple antimicrobial mechanisms effectively countered fluctuations in water quality. For instance, the hybrid membrane process provides highly concentrated BCs and partially excludes organics from membrane surfaces, endowing disinfectants with better efficacy.
4.1.3. Increasing ARGs removal
The absolute abundance of total ARGs in the influents of municipal wastewater and drinking water disinfection processes is generally not less than 10
5 and 10
3 copies·mL
−1, respectively
[209],
[59],
[210],
[211]. The abundance of ARGs in natural waters ranges from 10
2 to 10
6 copies·mL
−1 [212]. Although there are no recognized threshold concentrations of ARGs for environmental regulation, the reduction rates should be more than 1 log for disinfection to intercept ARGs enrichment. Conventional disinfection methods rarely reached ARGs removal rates of 1 log (Table S6 in Appendix A) and retained higher than 10
3 copies·mL
−1 in the final waters. Moreover, the relative abundance of ARGs and HGT potential increased after conventional disinfection processes
[18], increasing the occurrence of multi-drug resistant pathogens.
Oxidation-based mechanisms, especially the induction of ROS, are mainly effective for ARGs degradation. AOP and integrated techniques exhibited improved degradation capacities for both intracellular ARGs and free-state ARGs compared with conventional disinfection methods (Table S6). Plasma rapidly removes ARGs, achieving a reduction of more than 5 log of ARGs within the practical disinfection contact time. An integrated nanocomposite system appears to degrade ARGs more slowly than chemical oxidants but is still capable of removing three log ARGs within 60 min. Thus far, studies have been conducted to evaluate the potential enrichment of ARGs and delay HGT in environmental microbial communities after EATs. We believe that the transient and rapid antimicrobial effects of plasma do not promote undesirable ARGs enrichment. Under conditions without external energy input, sublethal NPs can act as carriers for ARGs dissemination. Nevertheless, this challenge can be overcome by establishing feasible recovery and utilization cycles for NPs.
The genetic targeting-cutting system based on CRISPR/Cas is promising for the “minimally invasive” removal of clinically relevant and mobile ARGs. The bacterial conjunction that delivers artificially designed CRISPR/Cas plasmids displayed a 100% ARGs removal rate in the wastewater matrix
[159].
4.2. Efforts toward more systematic engineering requirements
Different from the
EEO method, we provided a point estimation of the relative application potential, preliminary comparing the EATs and conventional disinfection by comprehensive weights from systemic engineering aspects (
Table 4). For the scenario of wastewater treatment, experts sequentially assigned the higher importance levels to “inactivation rate,” “operating cost,” and “global warming/carbon emission.” These indices have been incorporated into practical engineering design and environmental regulation. In drinking water treatment, “human health” was also given a higher weight, potentially interpreted by the recent inclusion of DBPs in the water quality standards. The judgment matrix
A (
Table 2) and
B (
Table 3) structured accordingly were multiplied to obtain the comprehensive weight
W.
4.2.1. Wastewater treatment
The point-collected data in
Fig. 6 and comprehensive weights in
Table 4 indicate that conventional methods are challenged by interference from complex wastewater backgrounds. Oxidants such as chlorine and ozone are rapidly consumed by organic substances in culture media, also resulting in the generation of numerous ecotoxic byproducts. As shown in
Fig. 6, the health effects and ecotoxicity of chlorine were more than twice those of UV during wastewater disinfection. The high residual chlorine concentrations necessitate additional dechlorination, thereby diminishing its cost advantage over EATs and outperforming UV.
Plasma more effectively counteracts the consumption of background impurities by performing thorough bacterial suppression in the culture medium within hours of short processing. Moreover, a recent pilot-scale report verified that non-thermal plasma achieved sufficient disinfection with low power inputs during the extended treatment, enhancing its application potential in wastewater compared with conventional disinfection. The emissions of nitrates were not accounted for in freshwater ecotoxicity but significantly contributed to eutrophication in the LCA models. Thus, the endpoint use of plasma is a better fit for agricultural reclaimed scenarios such as irrigation. Another potential solution for mitigating wastewater matrix interference and ecological impact trade-offs is phage treatment. High concentrations of virulent phages were added to allow for self-propagation with host fluctuation. A few studies have revealed that phages lyse their hosts more rapidly in wastewater because of the appropriate bacterial loads with nutrients
[143]. There have been no reports on the residual effects of exogenous phages; thus, eliminating the need for post-treatment and resulting in acceptable operating costs and ecological effects. Metallic nanomaterials lag behind because of their significant ecological risks and insufficient growth inhibition. The integrated photocatalytic nanocomposite showed promising performance in terms of bacterial growth inhibition and ARGs removal. The bottleneck is cost issues, where we assume the lifespan of nanomaterials to be 1200 cycles (i.e., continuous working for 30 d), which is close to the replacement cycle of filter materials in water treatment plants and exceeds the reported ranges
[213] but still results in unrealistic operating costs. Therefore, we highlight further optimization of the recovery and reusability of the nanocomposites, as the current validations for reusability in approximately 10 operating cycles seem to be inadequate.
4.2.2. Drinking water treatment
Consistent with the qualitative discussion in Section 2.1, chlorine remains the most common option for drinking water treatment. Chlorine has high disinfection rates and cost efficiency, despite concerns regarding the health effects and long-term control of BCs. Non-thermal plasma efficiently removed bacteria and ARGs, showing rates comparable to or higher than those of chlorine. Plasma initiates a large quantity of reactive species instantaneously, thereby exerting instantaneous and permanent inactivation and solving the problem of BCs regrowth. However, its scaling-up is mainly restricted by the high energy input, which significantly increases health impacts and carbon emissions by 1–2 orders of magnitude. Similarly, because of the relatively prioritized health impact of drinking water, the weights of metallic and photocatalytic nanocomposites are reduced, and the dominant cause is the higher energy input during the synthesis of nanomaterials. Pilot-scale trials of metallic nanomaterials for drinking water treatment have been limited to immobilized forms and typically require high dosages, thereby elevating their operational and waste disposal costs. Phage-based treatments exhibit unacceptably slow bacterial removal (> 12 h) in nutrient-poor drinking water and are almost ineffective against ARGs. Host concentrations lower than 10
6 CFU·mL
−1 challenge the lytic effects of exogenous phages
[142], indicating the inaccessibility of phages to a few hosts in drinking water. Considering their growth inhibition ability, phages can be used to control biofilms in distribution networks. Overall, EATs, even the scheme with the highest application potential (photocatalytic nanocomposites), demonstrated lower applicability in drinking water treatment systems than the conventional methods. The four exemplified technologies do not address the trade-offs between persistent antimicrobial efficiency and residual health effects unless combined with cost-effective and reliable post-treatments.
Taking the application potential, current status and proper niches in water treatment together (
Table 5), we can deduce whether and how the selected EATs can be brought to real-world in the foreseeable future, and reasonably propose the developing directions.
Immobilized Cu nanocomposites are incompetent for practical disinfection. Nano-immobilization sacrifices the antimicrobial efficiency to some extent, necessitating higher doses for effective disinfection compared to other EATs and conventional methods. The operating cost and health effect are elevated, causing the inferior application potential at this stage.
Non-thermal plasma is promising as a complementary or enhancing disinfectant for real-world application. Plasma can rapidly inactivate broad-spectrum BCs, as confirmed by the highest efficacy indices among the emerging schemes. It already has commercial applications for pilot-scale and decentralized water treatment. Nevertheless, its scaling-up is challenged by ➀ knowledge gap in the byproducts and ➁ significant increase of cost indices (including operating cost, carbon emission, and environmental load) attributed by the high energy consumption. Current strategies tend to collaborate the plasma with existing treatment to synergistically remove the chemical and biological contaminants.
Phage-based treatment is a potential complementary disinfectant. Despite the inefficient and narrow-spectrum inactivation, phages are recommended as green and low-cost auxiliary disinfectants in wastewater. Phages are advantageous for biofilm ablation and thus useful in the control of membrane fouling, sludge bulking and pathogens regrowth in the distribution system. Studies have unraveled the synergism between phages and conventional disinfection. We devise that phages can specifically target, identify and inactivate the host bacteria (especially for the robust pathogens in VBNC state) by the sequentially collaborating with UV or chlorine, while further explorations are required.
Immobilized photocatalytic nanocomposites can serve as complementary or even alternative disinfectants. Scheme 4 demonstrates higher application potential among the EATs. Unlike Scheme 1, the immobilized nanocomposites harness their catalytic properties to generate ROS, herein, with promising disinfection efficiency and decreasing residual ecotoxicity independent on the leaching of metal ions or nanoparticles. However, high costs resulted from the noble nanomaterials and relatively short lifespan significantly limit the scaling-up applications. We believe that the burgeoning advances in nanoscience will shed light on this issue and future discussions on the green, low-cost and regenerated catalytic materials are outlined.
5. Conclusions and prospectives
(1) Constructing a more systematic framework for evaluating the disinfection. In our preliminary framework, normalizing and integrating the efficacy and cost indices can enhance the interpretability of head-to-head evaluation between different treatment strategies. It allows for flexible adjustments such as adding/deleting indices and changing the weights based on actual requirements, and the literature-derived point estimates can be substituted by in-situ monitoring data. But the currently listed efficacy evaluation is not yet exhaustive. More inactivation data, especially for the rarely involved pathogenic organisms, is required to further establish a broad-spectrum control list of BCs.
(2) Developing green, renewable, and low-carbon techniques. We highlight the innovation and interdisciplinarity of green methodology based on biological intervention, solar-powered oxidation, new functional material synthesis and electroporation processes to perform future disinfection in water treatment process.
(3) Monitoring long-term effects of the emerging techniques. For the emerging technologies discussed in this review, there is a great necessity to unravel the intermediate products and the downstream ecotoxicity in AOP and to investigate the potential long-term effects of phages (especially the genetically engineered ones) on indigenous microbial communities.
CRediT authorship contribution statement
Rui Gao: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Shu-Hong Gao: Writing – review & editing, Resources, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization. Jun Li: Formal analysis, Data curation. Yiyi Su: Methodology, Investigation. Fang Huang: Writing – review & editing, Methodology. Bin Liang: Writing – review & editing, Visualization, Methodology. Lu Fan: Writing – review & editing, Visualization, Validation, Methodology. Jianhua Guo: Writing – review & editing, Validation, Methodology. Aijie Wang: Writing – review & editing, Validation, Supervision, Project administration, Methodology, Funding acquisition.
Declaration of competing interest
Rui Gao, Shu-Hong Gao, Jun Li, Yiyi Su, Fang Huang, Bin Liang, Lu Fan, Jianhua Guo, and Aijie Wang declare that have no conflict of interest or financial conflicts to disclose.
Acknowledgments
The study was finacially supported by the National Natural Science Foundation of China (52293443, 52321005, and 52230004), the Natural Science Foundation of Guangdong Basic and Applied Basic Research Foundation (2024A1515010085), Shenzhen Science and Technology Program (GXWD20231127195344001 and JCYJ20241202123735045), and Shenzhen Overseas High-level Talents Research Startup Program (20200518750C).
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2024.08.022.
PDF
(3185KB)
