Water microdroplet platforms for sustainable, reagent-free viral disinfection

TL;DR

This paper introduces a new disinfection method using water microdroplets that safely and effectively inactivate viruses without chemicals or harmful by-products.

Contribution

The study introduces a reagent-free viral disinfection platform using water microdroplets that generate reactive oxygen species for broad-spectrum inactivation.

Findings

Water microdroplets achieved over 99.999% viral inactivation within 20 minutes using reactive oxygen species.

Microdroplet treatment disrupted viral capsids and degraded nucleic acids, confirmed via microscopy and DNA analysis.

The method effectively disinfected surfaces like fresh produce and textiles with over 98% inactivation and no chemical residues.

Abstract

The repeated emergence of global pandemics has highlighted the urgent need for safe, sustainable, and effective disinfection platforms that eliminate viruses without producing toxic by-products or causing surface damage associated with conventional methods such as ultraviolet irradiation and chemical disinfectants. Here, we present water microdroplet platforms that exploit reactive oxygen species (ROS) spontaneously generated at the air–water interface of micron-sized water droplets, providing a reagent-free and cost-effective approach to viral inactivation. Bacteriophage T7 and lambda (λ), together with MS2 (a non-enveloped RNA bacteriophage) and Phi6 (an enveloped RNA bacteriophage), were selected as model viral systems to evaluate disinfection efficacy across different viral structures.

Water microdroplets with an average diameter of approximately 5 μm, generated by gas nebulization or mesh nebulizers, achieved more than 99.999% viral inactivation within 20 min. Transmission electron microscopy, protein profiling, and DNA analyses revealed that microdroplet-derived ROS disrupted viral capsid integrity and degraded nucleic acids, leading to loss of infectivity. The in situ generation of multiple ROS species was directly confirmed by mass spectrometry using a TEMPO probe and by fluorescence imaging with ROS-sensitive dyes, while scavenger assays verified the ROS-dependent nature of viral inactivation. Practical feasibility was demonstrated by treating fresh produce surfaces such as lettuce and potato, as well as porous and textile materials, resulting in more than 98% viral inactivation without chemical residues.

Together, these results demonstrate that water microdroplets provide an effective, reagent-free, and environmentally benign viral disinfection strategy with broad substrate compatibility for applications in food safety, healthcare, and textile-associated environments.

The online version contains supplementary material available at 10.1186/s13036-026-00626-z.

Figures

Schematic illustration of viral disinfection by reactive oxygen species (ROS) generated at the air-water interface of micron-sized droplets

Experimental setup for microdroplet-based viral disinfection. (A) Schematic of viral disinfection system using a gas nebulizer. (B) Photograph of the microdroplet spraying chamber. (C) Experimental setup of the mesh nebulizer for viral disinfection. Both devices generated microdroplets with an average diameter of approximately 5 μm

Inactivation of non-enveloped DNA bacteriophages (T7 and lambda), non-enveloped RNA bacteriophage (MS2), and enveloped RNA bacteriophage (Phi6) by water microdroplet treatment. (A, B) Representative plaque assay images of T7 (A) and lambda (B) after 10 min of treatment with bulk water, microdroplet spray, or sodium hypochlorite (NaClO, 100 ppm), compared with untreated controls. (C, D) Representative plaque assay images of MS2 (C) and Phi6 (D) under the same treatment conditions. (E) Time-dependent viral inactivation under microdroplet exposure. Lambda (gray) shows a gradual decline, whereas T

Morphological changes in bacteriophages after microdroplet treatment. (A) Untreated T7 showing intact head and tail structures. (B, C) T7 after microdroplet exposure, displaying disrupted and aggregated morphology. (D) Untreated lambda showing intact head and tail. (E, F) Lambda after microdroplet treatment, exhibiting damaged head and tail proteins. Scale bars, 200 nm

Microdroplet-generated reactive oxygen species (ROS) induce capsid degradation, genome damage, and ROS-dependent loss of infectivity in bacteriophages. (A) SDS-PAGE analysis of viral proteins. Blue Box: major lambda capsid protein (gpE) and tail fiber protein (gpJ). Red box: T7 capsid head protein (gp10) and tail proteins (gp8, gp12). Bands are markedly fainter in microdroplet-treated samples, indicating protein degradation and aggregation. (B) Quantification of capsid protein degradation after microdroplet treatment based on densitometric analysis of SDS-PAGE bands. (C) ddPCR analysis of lamb

Proposed mechanism of virus inactivation by water microdroplets. ROS generated at the air-water interface oxidize viral capsid proteins, compromise structural integrity, and damage nucleic acids, leading to complete loss of infectivity

Disinfection of food surfaces and porous materials using water microdroplets. (A) Schematic of the experimental procedure. (B) Plaque assay of bacteriophages on lettuce after microdroplet treatment. (C, D) Inactivation efficacy on lettuce: T7, 98.99%; lambda, 98.01%. (E) Plaque assay of T7 on potato after microdroplet treatment. (F) Inactivation efficacies of T7 on lettuce and potato were 98.99 and 98.9%, respectively. (G) Comparison of droplet generation methods showing similar inactivation efficacy: microdroplet spray, 98.99%; mesh nebulizer, 97.17%. (H) Representative plaque assay images of

Full text

Background

Viruses are a major threat to human health, causing a wide range of infectious diseases. They are responsible for approximately 15 million deaths annually, accounting for 26% of total global mortality [1]. The COVID-19 pandemic further underscored the severity and rapid transmission of viral infections, resulting in more than 5.37 million deaths worldwide [2] and a 4.5% reduction in the global gross domestic product, equivalent to 3.94 trillion US dollars [3]. While vaccines effectively prevent severe disease, the emergence of variants with reduced vaccine responsiveness, together with delayed vaccine distribution in low- and middle-income countries, sustains the global public health risk [4]. Therefore, to effectively control viral transmission, there is an urgent engineering challenge to develop sustainable, scalable, and reagent-free disinfection platforms for rapid virus control.

Viruses such as COVID-19 can remain viable on surfaces for extended periods, making effective disinfection essential to reduce indirect transmission, particularly in hospitals where infections threaten patients, healthcare workers, and visitors by prolonging hospitalization, raising costs, and increasing mortality [5–9]. In addition, foodborne viruses like rotavirus persist in crops, emphasizing the need for effective disinfection methods to mitigate transmission risks and ensure food safety [10]. Notably, rotavirus has been detected in vegetables and crops including sweet potatoes, carrots, radishes, potatoes, and lettuce [11–13].

Current viral disinfection methods usually rely on chemical agents such as sodium hypochlorite, hydrogen peroxide, alcohol-based solutions, and benzalkonium chloride, which are widely applied in both food processing and healthcare settings [14, 15]. However, these disinfectants present major drawbacks, including the formation of toxic by-products, limited efficacy against non-enveloped viruses, adverse health effects, poor biodegradability, and negative environmental impacts [16–20]. Thermal and ultraviolet (UV) irradiation are alternative approaches, but their use is limited to unoccupied spaces because of harmful effects on human health and the degradation of environmental surfaces [21]. These limitations highlight the engineering need for safe, sustainable, and deployable disinfection technologies.

Recent studies have demonstrated that water-air interface of aqueous microdroplets exhibits unique physicochemical properties distinct from bulk water [22–24], including strong concentration enhancement of solutes [25] and spontaneous formation of intrinsic interfacial electric fields at their surfaces [26]. They have been associated with various phenomena, including accelerated reaction rates, spontaneous water oxidation to form reactive oxygen species (ROS), and the reduction of solutes by electrons derived from the oxidation process, all occurring without the need for external catalysts, reductants [27–29]. Previous studies have reported that ROS generated from microdroplets along with the droplet surface charge destroyed bacterial cell walls, resulting in effective bacterial inactivation [30–32]. The spontaneous generation of ROS from microdroplets using ordinary water offers a promising basis for reagent-free disinfection platforms [20].

Virus disinfection is more difficult than bacterial inactivation because viruses are smaller, allowing easy transmission, and non-enveloped viruses exhibit higher resistance to disinfectants [29]. To address this challenge, we sought to translate microdroplet ROS chemistry into a biological engineering platform for virus inactivation. This disruption enables the penetration of ROS into the nucleic acid, resulting in virus inactivation and effective disinfection (Fig. 1). We selected bacteriophage T7 and lambda, both non-enveloped and specific to Escherichia coli, as well as MS2, a non-enveloped RNA bacteriophage that infects Escherichia coli, and Phi6, an enveloped RNA bacteriophage that infects Pseudomonas syringae. Together, these four enable evaluation across RNA and DNA bacteriophages and both non-enveloped and enveloped viral architectures [33–35]. Viral disinfection efficacy of bacteriophage T7 and lambda was evaluated using standard plaque assays. The disinfection mechanism is investigated with TEM, SDS-PAGE, and qPCR.

Overall, this study demonstrates that water microdroplet platform provides a practical, scalable, and sustainable disinfection method with direct implications for food processing and healthcare environments.

Fig. 1. Schematic illustration of viral disinfection by reactive oxygen species (ROS) generated at the air-water interface of micron-sized droplets

Methods

Preparation of sample and virus inoculation

Bacteriophage T7 and lambda were purchased from Lysentech (South Korea). Their host strains, Escherichia coli KCTC 1115 and KCTC 1116 (K-12), were used for propagation. LB broth base and LB agar were obtained from Thermo Fisher Scientific (USA). The MS2 and Phi6 bacteriophages used in this study were kindly provided by Prof. Jae Hee Jung (Sejong University, South Korea). Bacteriophages were loaded on crystal-grade polystyrene Petri dishes (SPL, South Korea). Prior to microdroplet treatment, 5 µL of bacteriophage suspension was deposited and air-dried for 20 min. For field tests, fresh lettuce and potatoes were purchased from a local market in Gwanggyo (South Korea). Porous materials, specifically disposable woven cotton cleaning cloths, were purchased from a local market and used to evaluate viral inactivation on porous substrates. Samples were washed with distilled water, surface moisture was removed, and they were cut into 1 × 1 × 0.1 cm^3^ squares (Fig. S1). Each sample was exposed to ultraviolet light on both sides to eliminate surface contaminants [10]. UV pretreatment was used solely to sterilize background microorganisms on produce surfaces prior to viral inoculation, at an intensity level previously reported to cause no significant surface modification [10]. A 5 µL bacteriophage suspension was then inoculated onto each sample and dried for 20 min to ensure viral attachment. Although our workflow does not replicate the ISO 15216-1:2017 protocol in full detail, the recovery procedure follows the same functional sequence recommended in the standard—virus detachment, elution into buffer, and collection of eluates for quantitative analysis. Surface-associated viruses were released by immersing lettuce and potato samples in buffer with gentle agitation, and the resulting eluate was directly used for plaque assay–based infectivity measurement.

Microdroplet devices

Microdroplets were generated using two different methods: gas nebulization and a mesh nebulizer. For gas nebulization setup, microdroplets were produced using a 100-µm inner-diameter silica capillary, with a gas pressure of 100 psi. The microdroplet diameter at this setup was estimated to be approximately 5 μm [28]. For gas nebulization, deionized water (Sigma-Aldrich, USA) was introduced into silica capillaries (inner diameter 100 μm). Three capillaries were used simultaneously, each delivering DI water at a flow rate of 10 µL/min via a syringe pump. Nitrogen gas (120 psi) served as the nebulizing agent, generating microdroplets at the capillary tips. For the mesh nebulizer, droplets were generated by vibrating a perforated membrane at high frequency, forcing liquid through micron-sized pores to produce uniform droplets. The device (Tekceleo, France) was equipped with a 5 μm mesh, generating droplets of similar size. The nebulizer flow rate was 0.8 mL/min.

Deposition rate calculation

To enable dose-normalized comparison across exposure times, the deposited microdroplet dose (mL cm⁻²) was calculated following the framework described by Wood et al. (2021) as:

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{D}\mathrm{e}\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\:\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\hspace{0.17em}=\hspace{0.17em}Q\:\times\:\:t\:/\:A$$\end{document}

where Q is the spray flow rate (mL min⁻¹), t is the exposure duration (min), and A is the deposition area defined by a 15-mm-diameter region (1.77 cm²) [36].

Double agar layer assay

A double agar layer assay was performed to assess bacteriophage infectivity based on plaque formation. Host strains were prepared in early stationary-phase culture suspensions. After microdroplet treatment, phages were collected in LB broth and 50 µL of suspension was incubated with the host strain for 15 min at room temperature. The mixture was combined with top agar (LB broth containing 0.6% agar, preheated to 50 °C) and poured onto a bottom agar layer (LB broth with 1.6% agar). Plates were incubated at 37 °C overnight, and plaques were counted visually. Viral inactivation efficiency was calculated as:

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{\%}\:\mathrm{v}\mathrm{i}\mathrm{r}\mathrm{u}\mathrm{s}\:\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}\:=\:\frac{{P}_{0}-{P}_{N}}{{P}_{0}}\times\:100$$\end{document}

where P₀ is the plaque number in untreated controls and Pₙ is the plaque number after n minutes of microdroplet exposure.

Bacteriophage DNA analysis

Digital droplet Polymerase Chain Reaction (ddPCR) (QX200 Droplet Digital PCR, Bio-rad) was conducted to quantitatively analyze the changes in the DNA concentration of bacteriophages before and after microdroplet treatment. DNA of the bacteriophages were extracted by Viral Gene-spin Viral DNA/RNA Extraction kit from iNtRON Biotechnology (South Korea). EvaGreen Supermix (Bio-rad) was used in the ddPCR reactions to enable fluorescence-based quantification of DNA. PCR analysis was performed using specific primers targeting the lambda phage genome: Lambda For (5′-GCG TTA CCG TTT CGC GGT GC-3′) and Lambda Rev (5′-TCG CAG CAT TGC CCG TCA GG-3′) [37, 38].

Transmission electron microscopy (TEM)

Microdroplet treated bacteriophage samples and control samples were dropped onto glow charged carbon grids and kept for 10 min at room temperature. After the liquid was blotted by filter paper, the grids were covered with 1% uranyl acetate. The transmission electron microscope (TEM) instrument (JEOL, JEM 1010) was utilized with 200 kV at ×25,000 magnification.

SDS-PAGE

T7 and lambda bacteriophages that were untreated or microdroplet treated were mixed with Laemmli buffer (Sigma-Aldrich) and heated for 2 min at 85 °C. Electrophoresis was performed for 90 min at 100 V. After electrophoresis, gels were stained with Coomassie blue R-250 (Thermo Fisher Scientfic) and scanned using Chemiluminescence Imaging System (iBright CL1500, Thermo Fisher Scientific).

Mass spectrometric analysis

Reactive oxygen species (ROS) generated in water microdroplets were detected using the stable radical probe 2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO). A 500 µM aqueous TEMPO solution was nebulized into a mass spectrometer inlet using a nitrogen-assisted sprayer (120 psi) at a flow rate of 10 µL/min and a spray distance of 9 cm, matching the conditions used in the viral inactivation assays. ROS-oxidized TEMPO products were analyzed using a Q Exactive Quadrupole–Orbitrap mass spectrometer (Thermo Fisher Scientific). The spray voltage was set to 3 kV, and the inlet temperature was maintained at 250 °C.

Results

Inactivation of viruses by ROS generated from microdroplets

According to the U.S. Centers for Disease Control and Prevention (CDC), a 10-minute contact time is recommended for effective disinfection, and most EPA-registered hospital disinfectants specify this duration on their labels [39]. To allow comparison, bacteriophage suspensions were also treated with 100 ppm sodium hypochlorite under the same conditions used for microdroplet exposure.

Viral inactivation by microdroplets was tested using two microdroplet-generation methods: gas nebulization and a mesh nebulizer (Fig. 2). Both approaches produced microdroplets with an average diameter of approximately 5 μm [28]. Reactive oxygen species (ROS), including hydrogen peroxide, hydroxyl radicals, and superoxide radicals, were spontaneously generated at the air–water interface of these droplets [20, 28, 40]. Deposition rate onto a target surface was calculated using equation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{De}\mathrm{position\:rate}=Q\times\:t/A$$\end{document} . Table S1 summarizes the deposition rate for each device used for microdroplet-based viral inactivation. For the gas nebulizer spray (30 µL/min), 10-, 15-, and 20-minute exposures resulted in areal doses of 170, 255, and 340 µL/cm^2^, respectively. In contrast, the mesh nebulizer (800 µL/min for 10 min) generated a substantially higher deposited dose of 4520 µL/cm^2^ due to its greater liquid output.

Fig. 2. Experimental setup for microdroplet-based viral disinfection. (A) Schematic of viral disinfection system using a gas nebulizer. (B) Photograph of the microdroplet spraying chamber. (C) Experimental setup of the mesh nebulizer for viral disinfection. Both devices generated microdroplets with an average diameter of approximately 5 μm

To distinguish the effects of bulk water from those of water microdroplets, an equivalent volume of bulk water was applied to bacteriophage suspensions under the same experimental conditions. To evaluate inactivation efficacy, suspensions were exposed to microdroplet spray for different durations. A pronounced reduction in plaque-forming units (PFUs) was observed in the microdroplet-treated samples compared with both bulk water and untreated controls, as evidenced by visual inspection of plaque assays (Fig. 3A-B). To evaluate whether the antiviral efficacy of microdroplet treatment extends beyond DNA bacteriophages, we tested two RNA phages with distinct structural properties: MS2, a non-enveloped RNA virus, and Phi6, an enveloped RNA virus. As shown in Fig. 3C–D, MS2 produced dense plaques in untreated controls and exhibited only limited reduction with bulk water, whereas microdroplet treatment resulted in near-complete plaque disappearance. Phi6 showed similarly high susceptibility to microdroplet treatment, with substantial loss of plaques relative to both untreated and bulk-water conditions.

A time-dependent study was performed by exposing bacteriophage suspensions to microdroplet treatment for different durations, which revealed progressively higher inactivation with longer exposure (Fig. 3E, Fig. S2). Viral infectivity was quantified by counting plaque-forming units (PFUs) after incubation. Raw plaque counts, PFU/mL calculations, and assay limits of detection corresponding to Figs. 3 and 6 are provided in Tables S2-S5. In Fig. 3E, lambda phage displayed a gradual decline in infectivity, reaching more than six log reductions after 15 min of microdroplet exposure. In contrast, T7 remained relatively resistant during the first 10 min, but then underwent a sharp decline between 15 and 20 min, consistent with cumulative ROS-induced capsid damage. These differences may be attributed to structural and physicochemical distinctions between the two phages, with lambda being more vulnerable to early ROS-mediated capsid oxidation, whereas T7 resists initial oxidative stress but undergoes accelerated inactivation after prolonged exposure due to accumulated structural damage. This observation is consistent with previous reports that bacteriophage T7 exhibits higher oxidative stress resistance compared with bacteriophage MS2 and E. coli [34].

We compared the antiviral efficacy of microdroplets with sodium hypochlorite (NaClO) under identical conditions (Fig. 3F-G). After 10 min of treatment, microdroplets achieved 99.16% inactivation for T7 and 99.56% for lambda. Sodium hypochlorite was less effective against T7 but showed strong activity against lambda, whereas microdroplets provided the highest level of inactivation for both bacteriophages. Quantitative analysis (Fig. 3H–I) further confirmed that microdroplets achieved greater than 99% inactivation for both RNA phages (MS2 and Phi6), outperforming bulk water and showing efficacy comparable to NaClO treatment.

These findings demonstrate that microdroplet treatment is more effective than chlorine-based disinfectants within the same exposure time and volume. Moreover, the marked difference between bulk water and microdroplets highlights that the antiviral activity originates from the unique physicochemical properties of microdroplets, particularly their ability to generate reactive oxygen species at the air–water interface.

To benchmark the microdroplet platform against commonly used chemical disinfectants, we compared its antiviral efficacy with ethanol, hydrogen peroxide, and sodium hypochlorite. As shown in Supplementary Fig. S3A–B, microdroplet spraying achieved the highest level of inactivation for MS2 (99.71%), whereas ethanol, hydrogen peroxide, and sodium hypochlorite achieved inactivation efficiencies of 97.45, 95.75, and 97.91%, respectively. For Phi6, microdroplet treatment reached 99.25% inactivation, compared with 97.56–98.20% for the chemical disinfectants. These results demonstrate that the microdroplet platform performs comparably to, or better than, widely used chemical disinfectants while requiring only water and leaving no chemical residues. Environmental factors known to influence ROS generation, including nebulization gas composition (ambient air, O₂, and N₂) and relative humidity (40% and 60%), did not significantly affect antiviral outcomes. As shown in Fig. S4A–D, both MS2 (97.65–98.93%) and Phi6 (98.74–99.51%) remained highly susceptible under all tested conditions, indicating robust microdroplet-mediated disinfection performance across diverse environmental settings.

Fig. 3. Inactivation of non-enveloped DNA bacteriophages (T7 and lambda), non-enveloped RNA bacteriophage (MS2), and enveloped RNA bacteriophage (Phi6) by water microdroplet treatment. (A, B) Representative plaque assay images of T7 (A) and lambda (B) after 10 min of treatment with bulk water, microdroplet spray, or sodium hypochlorite (NaClO, 100 ppm), compared with untreated controls. (C, D) Representative plaque assay images of MS2 (C) and Phi6 (D) under the same treatment conditions. (E) Time-dependent viral inactivation under microdroplet exposure. Lambda (gray) shows a gradual decline, whereas T7 (red) remains stable until 10 min and then decreases sharply between 15 and 20 min. (F) Percent inactivation of T7 after 10 min for bulk water, microdroplets, and NaClO. (G) Percent inactivation of lambda after 10 min for bulk water, microdroplets, and NaClO. (H, I) Quantification of MS2 (H) and Phi6 (I) inactivation efficiencies after 10 min of treatment. Error bars indicate standard deviations from triplicate experiments. Statistical significance was assessed using one-way ANOVA with post-hoc multiple comparison tests (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001)

Morphological changes in microdroplet-treated bacteriophages

To determine whether ROS generated from microdroplets disrupt viral capsid structures, transmission electron microscopy (TEM) was used to compare the morphology of untreated and microdroplet-treated bacteriophages (Fig. 4, Fig. S5). A total of six virus particles were analyzed in the TEM images for morphological assessments. Viral particles were analyzed per condition in the TEM micrographs. Damage categories were defined based on (i) capsid integrity—loss of icosahedral geometry, surface disruption, or structural collapse—and (ii) tail morphology, including detachment or deformation. Untreated controls showed intact viral particles with well-defined capsids. For T7, the average head diameter of five representative particles was approximately 62 nm, consistent with reported dimensions (Fig. 4A). Lambda phage exhibited a defined head structure with an average diameter of 66 nm and a tail length of 137 nm. (Fig. 4B)

In contrast, microdroplet-treated samples (Fig. 4B, C, E, F) displayed disrupted and aggregated particles, with evident loss of capsid integrity, indicating significant structural damage caused by ROS. These alterations are consistent with previous observations of ROS-mediated disruption in bacteriophages [41, 42].

Fig. 4. Morphological changes in bacteriophages after microdroplet treatment. (A) Untreated T7 showing intact head and tail structures. (B, C) T7 after microdroplet exposure, displaying disrupted and aggregated morphology. (D) Untreated lambda showing intact head and tail. (E, F) Lambda after microdroplet treatment, exhibiting damaged head and tail proteins. Scale bars, 200 nm

Mechanisms of microdroplet-induced viral disinfection

Next, we investigated the mechanism of bacteriophage inactivation by microdroplet treatment. SDS-PAGE analysis was performed to examine whether the structural degradation observed after microdroplet exposure was associated with degradation of viral proteins (Fig. 5A). Protein bands of bacteriophages treated with microdroplets for 10 min appeared markedly fainter than those of untreated controls, indicating substantial protein loss or degradation. The reduced band intensity is likely due to ROS-induced denaturation and aggregation of capsid proteins, which hindered migration through the gel or caused protein loss during sample preparation. Densitometric quantification of the major capsid protein bands revealed a greater than 90% reduction in relative protein intensity following microdroplet treatment (Fig. 5B), indicating extensive ROS-induced oxidative damage and structural disruption at the protein level. This interpretation is supported by TEM observations (Fig. 4), which revealed extensive capsid disruption in microdroplet-treated samples.

The loss of capsid integrity suggests that viral nucleic acids became exposed and were subsequently damaged. To test this, ddPCR was conducted to quantify DNA concentration before and after microdroplet treatment (Fig. 5C). Lambda DNA concentration decreased from 7.73 to 5.53 copies/µL, consistent with partial genome degradation. This reduction is attributed to oxidative damage by ROS, which can cause DNA fragmentation and loss of genetic material.

To verify that these structural and genomic damages originate from microdroplet-generated ROS, we directly characterized the ROS generated from microdroplets. Mass spectrometric analysis using the ROS-sensitive probe TEMPO, performed under the same spray conditions as the viral inactivation assays (9 cm, 120 psi), revealed clear formation of TEMPO–OH and TEMPO–OOH (Fig. S6A), confirming in situ generation of hydroxyl and hydroperoxyl/superoxide species. To further visualize ROS formation within individual microdroplets, we employed fluorescence microscopy using three ROS-responsive dyes: PF1 for H₂O₂, HPF for •OH, and DHE for superoxide. As shown in Fig. S6B, microdroplets produced strong punctate fluorescence across all dye conditions, indicating that each ROS species is generated directly within the droplets. These mass spectrometric and fluorescence measurements together establish that microdroplets generate a spectrum of ROS species capable of inducing the observed capsid and genome damages in bacteriophages.

Then, we employed scavenger assays targeting O₂•⁻ (SOD), •OH (salicylic acid), and solvated electrons (sodium nitrate). Figure 5D shows inactivation efficacy of microdroplets containing each ROS scavenger, which shows significantly reduced inactivation efficiency in the presence of ROS scavengers. For comparison, bulk water treatments showed minimal changes in viral infectivity (Fig. S7), indicating that ROS-driven inactivation is specific to microdroplet rather than bulk water. In the case of SOD, the higher inactivation efficacy may be explained by the conversion of superoxide radicals into hydrogen peroxide, rather than by complete suppression of ROS. Previous studies have reported that elevated SOD activity can paradoxically increase oxidant toxicity by accelerating H₂O₂ formation and enhancing susceptibility to oxidative damage, supporting the possibility that SOD alters rather than eliminates ROS burden [43, 44]. These results together confirm that ROS spontaneously generated from microdroplets acts as a virucide.

Collectively, these results demonstrate that reactive oxygen species generated from water microdroplets are both necessary and sufficient to induce capsid disruption and genome damage, leading to irreversible viral inactivation. Based on these findings, we propose an integrated mechanism for microdroplet-mediated viral disinfection, as summarized in Fig. 6.

Fig. 5. Microdroplet-generated reactive oxygen species (ROS) induce capsid degradation, genome damage, and ROS-dependent loss of infectivity in bacteriophages. (A) SDS-PAGE analysis of viral proteins. Blue Box: major lambda capsid protein (gpE) and tail fiber protein (gpJ). Red box: T7 capsid head protein (gp10) and tail proteins (gp8, gp12). Bands are markedly fainter in microdroplet-treated samples, indicating protein degradation and aggregation. (B) Quantification of capsid protein degradation after microdroplet treatment based on densitometric analysis of SDS-PAGE bands. (C) ddPCR analysis of lambda genomic DNA concentration, showing a decrease from 7.73 to 5.53 copies per µL after microdroplet treatment, consistent with oxidative genome damage. Error bars represent standard deviations from three independent replicates. (D) Effects of ROS scavengers on microdroplet-mediated viral inactivation. Superoxide dismutase (SOD; superoxide scavenger), salicylic acid (OH scavenger), and sodium nitrate (electron scavenger) reduced antiviral efficacy to varying extents, demonstrating the involvement of multiple ROS pathways during microdroplet treatment. Error bars denote standard deviations (n = 3). Statistical significance was evaluated using one-way ANOVA with post-hoc multiple comparison tests (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001); “ns” indicates non-significant differences relative to the bulk water treatment control

Fig. 6. Proposed mechanism of virus inactivation by water microdroplets. ROS generated at the air-water interface oxidize viral capsid proteins, compromise structural integrity, and damage nucleic acids, leading to complete loss of infectivity

Viral disinfection on food surfaces and porous materials using water microdroplets

To assess the practical applicability of microdroplet-mediated viral disinfection across materials relevant to real-world transmission, we evaluated antiviral efficacy on both food surfaces and porous substrates, including fresh produce and woven textile materials. Lettuce is a major vehicle for the transmission of enteric viruses such as rotavirus in the United States, because it is often consumed raw and its broad, uneven surface facilitates viral adherence and limits the effectiveness of washing [10]. In addition, previous studies have shown that rotavirus can be transmitted to humans through irrigated crops including carrots, potatoes, and sweet potatoes [12]. These findings highlight the need for effective disinfection strategies that can be applied to a variety of fresh foods.

To evaluate the applicability of microdroplet treatment in food systems, viruses were directly inoculated onto the surfaces of lettuce and potato samples. Figure 7A shows a schematic of the experimental procedure. Lettuce leaves and potato slices were cut into 1 × 1 × 0.1 cm³ squares, exposed to ultraviolet light on both sides to remove pre-existing contaminants [10], and then inoculated with 5 µL of bacteriophage suspension. After drying for 20 min, the inoculated samples were treated with microdroplet spray for 10 min.

Following treatment, viruses were recovered from the food surfaces and assessed by plaque assay (Fig. 7B). Inactivation efficacy on lettuce was similar to that observed for Petri dish surfaces (Fig. 7C, D), indicating that the antiviral effect of microdroplets is maintained across different substrates. On potato skin, despite its irregular and curved surface, microdroplet treatment also resulted in effective inactivation (Fig. 7E). Overall, microdroplet treatment achieved more than 98.9% inactivation of T7 on both lettuce and potato surfaces (Fig. 7F), demonstrating robust antiviral efficacy regardless of surface characteristics. Finally, we compared spray-based and mesh-nebulizer systems (Fig. 7G). Both methods achieved comparable inactivation efficiencies, 98.99% and 97.17%, respectively, indicating that antiviral efficacy depends primarily on droplet size rather than generation method.

Because porous materials such as textiles can act as reservoirs for viral particles and are challenging to disinfect using conventional surface treatments, we further evaluated the microdroplet platform on porous substrates using woven cotton fabric. Both MS2 and Phi6 exhibited greater than 99% inactivation on fabric surfaces despite their high absorbency and structural heterogeneity (Fig. 7H–J). These results indicate that microdroplet-generated ROS can effectively penetrate and disinfect porous materials, extending the practical applicability of the platform beyond non-porous and food-contact surfaces to textiles and other high-porosity environments.

Fig. 7. Disinfection of food surfaces and porous materials using water microdroplets. (A) Schematic of the experimental procedure. (B) Plaque assay of bacteriophages on lettuce after microdroplet treatment. (C, D) Inactivation efficacy on lettuce: T7, 98.99%; lambda, 98.01%. (E) Plaque assay of T7 on potato after microdroplet treatment. (F) Inactivation efficacies of T7 on lettuce and potato were 98.99 and 98.9%, respectively. (G) Comparison of droplet generation methods showing similar inactivation efficacy: microdroplet spray, 98.99%; mesh nebulizer, 97.17%. (H) Representative plaque assay images of MS2 and Phi6 deposited on woven textile surfaces following treatment with bulk water, microdroplet spray, or NaClO, compared with untreated controls. (I,J) Quantification of inactivation efficacy for MS2 (I) and Phi6 (J) on textile surfaces under the same treatment conditions. Error bars represent standard deviations from three independent replicates. Statistical significance was assessed using one-way ANOVA with post-hoc multiple comparison tests (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001)

Discussion

Virus inactivation using ROS generated from water microdroplets represents a novel, cost-effective, and non-hazardous disinfection strategy. In contrast, conventional chemical disinfectants are limited by their selective reactivity and the formation of harmful by-products. For example, chlorine-based disinfectants react with organic compounds in a highly selective manner, targeting specific sites on aromatic rings depending on molecular structure [45] and produce toxic disinfection by-products such as trihalomethanes (THMs) [46]. These issues raise concerns over chemical safety and sustainability.

Unlike chemical agents, microdroplet-generated ROS interact broadly with biomolecules, including lipids and nucleic acids [47]. Moreover, this method requires only ordinary water, eliminating the risk of hazardous residues. Because ROS decompose into water and oxygen after reaction, microdroplet disinfection is inherently safe for both humans and the environment [20]. Together, these properties overcome the drawbacks of current disinfectants, including chemical toxicity, low biodegradability, and environmental burden.

The present study demonstrated that microdroplet treatment achieved more than 99% inactivation of bacteriophages across different surfaces, including plastics and fresh produce, underscoring its applicability in real-world settings such as food processing and healthcare environments. This high efficacy, combined with the absence of chemical additives, positions microdroplet technology as a safe and eco-friendly alternative to chlorine-based disinfectants and UV irradiation.

The operational simplicity and low cost of microdroplet generation make this platform highly scalable and accessible. Unlike UV systems that require high capital investment or chlorine-based methods with ongoing maintenance, microdroplet systems rely only on water and minimal energy input. A preliminary cost analysis indicates that the microdroplet-ROS process operates at approximately 0.0007–0.0009 US$/min, consuming only distilled water and compressed gas [48, 49]. A detailed comparison with UV and chlorine disinfection costs is provided in Table S6. Future development could focus on continuous-flow chambers, integration into HVAC or conveyor systems, and coupling with biosensor for real-time monitoring. Taken together, our findings establish microdroplet ROS disinfection as not only an eco-friendly alternative but also a scalable platform for virus control in food safety and healthcare environments.

Conclusions

This study demonstrates that ROS generated from water microdroplets effectively inactivate bacteriophages, including T7 and lambda, through capsid protein and nucleic acid damage. The antiviral efficiency was confirmed across various surfaces, with virus inactivation efficacies exceeding 99%, even on irregular or curved food surfaces. In time-dependent experiments, more than 5-log reductions in viral infectivity were achieved with prolonged microdroplet treatment, indicating more than 99.999% bacteriophage inactivation. Because microdroplets are produced from ordinary water and yield only non-toxic by-products, this approach offers a safe, sustainable, and cost-effective alternative to chemical disinfectants. These findings establish water microdroplet platforms as a practical, scalable, and sustainable disinfection technology with direct applications in food safety and healthcare environments.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1