Field trial assessing the antimicrobial decontamination efficacy of gaseous ozone in a public bus setting

The widespread COVID-19 pandemic caused by the Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) necessitated measures aimed at preventing the spread of SARS-CoV-2. To mitigate the risk of fomite-mediated transmission, environmental cleaning and disinfection regimes have been widely implemented. However, conventional cleaning approaches such as surface wipe downs can be laborious and more efficient and effective disinfecting technologies are needed. Gaseous ozone disinfection is one technology which has been shown to be effective in laboratory studies. Here, we evaluated its efficacy and feasibility in a public bus setting, using murine hepatitis virus (a related betacoronavirus surrogate) and the bacteria Staphylococcus aureus as test organisms. An optimal gaseous ozone regime resulted in a 3.65-log reduction of murine hepatitis virus and a 4.73-log reduction of S. aureus, and decontamination efficacy correlated with exposure duration and relative humidity in the application space. These findings demonstrated gaseous ozone disinfection in field settings which can be suitably translated to public and private fleets that share analogous characteristics.


H I G H L I G H T S G R A P H I C A L A B S T R A C T
• Disinfection practices are important to contain the spread of SARS-CoV-2. • Gaseous ozone is an effective disinfectant that can be safely applied in confined spaces such as public transport. • Increase in duration of gaseous ozone exposure and in relative humidity correlated with improved disinfection efficacy.

A B S T R A C T A R T I C L E I N F O Editor: Warish Ahmed
Keywords: Gaseous ozone Murine coronavirus Antimicrobial efficacy Staphylococcus aureus Field trial The widespread COVID-19 pandemic caused by the Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) necessitated measures aimed at preventing the spread of SARS-CoV-2. To mitigate the risk of fomite-mediated transmission, environmental cleaning and disinfection regimes have been widely implemented. However, conventional cleaning approaches such as surface wipe downs can be laborious and more efficient and effective disinfecting technologies are needed. Gaseous ozone disinfection is one technology which has been shown to be effective in laboratory studies. Here, we evaluated its efficacy and feasibility in a public bus setting, using murine hepatitis virus (a related betacoronavirus surrogate) and the bacteria Staphylococcus aureus as test organisms. An optimal gaseous ozone regime resulted in a 3.65-log reduction of murine hepatitis virus and a 4.73-log reduction of S. aureus, and decontamination efficacy correlated with exposure duration and relative humidity in the application space. These findings demonstrated gaseous ozone disinfection in field settings which can be suitably translated to public and private fleets that share analogous characteristics.
Science of the Total Environment 876 (2023) 162704

Introduction
The Coronavirus disease 2019  pandemic has caused almost 500 million infections and about 6.1 million deaths as of April 2022 (WHO, n.d.). Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the causative agent of coronavirus disease , is known to be spread through close contact, inhalation of contaminated aerosols in poorly ventilated settings, and through fomite-mediated transmission where persons come in physical contact with infectious droplets deposited on surfaces (CDC, n.d.). Reports of cases linked only through shared spaces in Singapore and in other countries, suggest the likelihood of aerosolmediated or fomite-mediated transmission (Goh, n.d.;Pung et al., 2020). Studies have also shown that individuals can get contaminated through snot-oral transmission from virus droplets found on surfaces (Xie et al., 2020). To reduce disease spread and severity, vaccination programmes have been rolled out across the globe. Various countries have also implemented public and environmental health measures targeted at reducing disease transmission, including case isolation and quarantine, improving ventilation of enclosed premises, and enhancing environmental cleaning and disinfection regimes, among others.
To mitigate risks of fomite-mediated transmission of SARS-CoV-2, several techniques for disinfection and sanitization of surfaces have been proposed. The most common method is surface wipe-downs using disinfectants effective against coronaviruses, such as sodium hypochlorite and hydrogen peroxide solutions (Cheng et al., 2022;NEA, n.d.). However, as such conventional approaches are laborious and time consuming, scalable and automated technologies have also been increasingly introduced. Examples include different types of ultraviolet light (UV) (Kumar et al., 2021) such as ultraviolet germicidal irradiation (UVGI) (Hamzavi et al., 2020), heat inactivation (Kampf et al., 2020), hydrogen peroxide vapor (HPV) (Schwartz et al., 2020), self-disinfecting materials and surfaces (Castaño et al., 2021), and ozone (Zucker et al., 2021). Most of these technologies have been under review and rapidly adapted due to the pandemic as a response for public spaces disinfection (Blanco et al., 2021), yet efficacy data are still lacking, especially of their use in specific settings and conditions.
Ozone has been extensively used in killing bacteria (Aydogan and Gurol, 2006;Knobling et al., 2021;Huang et al., 2012;Megahed et al., 2018;Selma et al., 2008;Sharma and Hudson, 2008), fungi, and molds (Mohamed and Barbara, 2021) and in inactivating viruses (Pekovic and Kacimi, 2015;Lee et al., 2021;Tanaka et al., 2009;Hudson et al., 2007;Brié et al., 2018;Dubuis et al., 2020); its antimicrobial properties and effectiveness have been well documented in literature (Table S1). Ozone technology has been applied in water and wastewater treatment , air disinfection (Grignani et al., 2021a), food sanitation (Brodowska et al., 2018) and even as a novel medical treatment for various illnesses (Manjunath et al., 2021). Ozone can disrupt surface proteins and/ or membrane receptors, compromising the infectivity of the pathogen by reacting with amino acids and functional groups of protein (Tseng and Li, 2008). Furthermore, ozone can damage the viral envelope and genetic material, due to the direct oxidation of several molecules or the peroxidation of phospholipids and consequent production of reactive oxygen species (Cristiano, 2020). The efficacy of gaseous ozone against enveloped viruses (e.g., coronavirus, influenza virus, mumps virus) has been consistently shown to be effective (Table S1). Nevertheless, its efficacy can vary largely depending on environmental factors in the field, and it is important to demonstrate its effectiveness for each specific use case.
Public transport (such as buses and trains) settings require frequent disinfection as they are often high-footfall facilities with multiple common touch points (such as hand grips, grab poles, windows, seats, etc.) that may serve as a medium for fomite-mediated transmission (Moreno et al., 2021). The survivability of SARS-CoV-2 on non-porous surfaces commonly found in transport carriages such as stainless steel and plastic for up to 48 h post-deposition (van Doremalen et al., 2020), further necessitates regular disinfection in these settings. This therefore presents an ideal use-case to trial the efficacy of gaseous ozone surface disinfection, as the technique offers ease of disinfection and scalability.
This study aimed to determine the decontamination efficacy of gaseous ozone generated by a gaseous ozone generator in a public bus setting. Adapting our methods from the standard BS EN 17272:2020, we used murine hepatitis virus (a betacoronavirus surrogate) and the bacteria Staphylococcus aureus as our challenge organisms. Decontamination efficacy correlated with exposure duration and relative humidity in the application space, with an optimal gaseous ozone regime resulting in a 3.65-log reduction of murine hepatitis virus and a 4.73-log reduction of S. aureus. These findings provide real-world estimates of the effectiveness of gaseous ozone as a method for decontamination of surfaces, with the findings translatable to spaces that are analogous to buses.

Study protocol
The trial methodology was adapted from the standard BS EN 17272:2020 -"Chemical disinfectants and antiseptics. Methods of airborne room disinfection by automated process".

Preparation of host cell line and test organisms
Murine hepatitis virus (MHV; a betacoronavirus) and Staphylococcus aureus were used as challenge organisms. MHV, a Risk Group 2 pathogen, has been used as a surrogate for SARS-CoV-2 in many disinfection studies (Boegel et al., 2021), and it is genetically and morphologically related to SARS-CoV-2 (Körner et al., 2020). To prepare the working stock of MHV, virus was subcultured on H2.35 mouse hepatocytes (ATCC CRL-1995™) and maintained in HyClone RPMI 1640 media (Cytiva) with 10 % FBS supplementation and incubated at 33°C with 5 % CO 2 . S. aureus (ATCC 25923) was prepared through two successive subcultures. Briefly, S. aureus was sub-cultured through streak-plating from a stock culture on tryptone soy agar (TSA). The bacterial plate was incubated at 37°C overnight for 18 h to 24 h before a single colony was picked for the next subculture. Approval was sought and obtained from the Biosafety Branch (Public Health Group), Ministry of Health, Singapore, for the use of the test organisms in the field trial.

Preparation of contaminated carriers
Stainless steel disks (Fujitson, rated grade 2B, 301), 4 cm in diameter and 1.2 mm thick were used in this trial. The disks were cleaned with 70 % v/v ethanol and then autoclaved (121°C, 20 min) prior to use. A 50 μL volume of MHV or S. aureus inoculum was evenly spread over the surface of the carrier, covering an internal area 2 cm diameter wide. The carriers were then placed onto petri dishes and into a humidified chamber for drying, with lid opened.

Tissue culture infectious dose 50 % (TCID50) assay
Tissue culture flat-bottom 96-well plates were seeded with H2.35 cells at a density of 2 × 10 5 cells per well, one day prior to the tests. Postdecontamination regime, 1 mL of each recovered sample was serially diluted 10-fold, and 50 μL of each dilution was added into 5 replicate wells. The plates were then incubated at 33°C with 5 % CO 2 for three days. Cytopathic effects were determined by visualizing under a light microscope. The cell cultures were subsequently stained with crystal violet, and virus titer was determined by TCID50/mL were quantified based on the Spearman and Kärber method.

Bacterial enumeration assay
Following exposure, the spiked carriers were transferred to a sterile urine container containing 10 mL of tryptone salt broth (TSB). The container was capped and sealed with parafilm, placed in a sonicating waterbath (Elma P 30H) and sonicated at 37 kHz at 100 % power for 5 min. 1 mL of the TSB was recovered for bacterial enumeration using the drop-plate method. Briefly, the TSB was serially diluted 10-fold to 10 −8 E.S. Neves et al. Science of the Total Environment 876 (2023) 162704 with PBS, and 10 μL of each dilution was deposited onto a marked quadrant of a TSA agar plate. A total of 5 drops were deposited for each dilution, and the agar plate was incubated at 37°C for 24 h. The dilution that had 0-30 isolated colonies within each drop was used to determine the CFU recovered from each carrier.

Virestorm ozone generator
The Virestorm V5 ozone generator generates ozone gas from oxygen in ambient air and breaks down the generated ozone gas back into oxygen. This is accomplished through the combination of a first stage high-powered (average working power is approximately 1000 watts) atmospheric-pressure plasma generator in a dense dielectric barrier discharge configuration to generate the gaseous ozone and a second stage miniaturized high-efficiency catalytic converter using a proprietary blend of catalysts designed to accelerate the decomposition of ozone gas into oxygen.

Phase 1 -trial location and set-up
Phase 1 of the trial was conducted at a bus depot, with support from Singapore Bus Service Transit (SBS Transit) and the Land Transport Authority (LTA) of Singapore. Two double-deck buses were used for this trial. Each bus was an air-conditioned 80-seater with internal dimensions of 11 m (L) × 2.5 m (W) × 2.5 m (H), approximate volume of 57.96 m 3 for lower deck, 11.5 m (L) × 2.5 m (W) × 1.8 m (H), approximate volume of 51.89 m 3 for upper deck. One bus was subjected to gaseous ozone while the second served as a control (not exposed to gaseous ozone) (Fig. S1).
Two spiked carriers were placed at each corner, and on all 4 corners of both decks (2 carriers per corner × 4 corners × 2 decks = 16 carriers total). For the corners where carriers were close to the ceiling, 1 carrier with the inoculated side faced the wall (vertically) and 1 carrier faced the ceiling (horizontally); while at the corners where carriers were close to the floor, 1 carrier faced the wall (vertically) and 1 carried faced the floor (horizontally) (Fig. 1A). The same setup was performed in the control bus. The inoculated carriers in the control and test buses were secured with clip stands attached to Styrofoam blocks (Fig. 1B). All carriers were located about 50 cm from the wall and ceiling/floor (Fig. 1B). MHV spiked carriers were tested in the first cycle, and then removed after the decontamination cycle. Following that, the S. aureus inoculated carriers were then tested using the same layout in a second decontamination cycle. The decontamination cycle consisted of 15 min of gaseous ozone generation, aiming to achieve a gaseous ozone concentration of between 13 and 18 ppm as the ozone sensors (EcoSensors, USA) were calibrated to accurately measure up to 20 ppm. This was followed by an approximately 40 min for ozone degeneration (Fig. S2). During the test, gaseous ozone concentration, temperature and humidity were monitored. At the step of ozone degeneration, the ozone concentration inside the bus was continuously monitored until levels were below 0.5 ppm, before the bus was immediately ventilated by opening the bus doors with an outward-facing ventilation fan. Following which, operators were allowed to enter the bus to retrieve the carriers. A cycle was run for each organism separately; the first cycle with MHV, followed by a second cycle with S. aureus.

Phase 2 -trial location and set-up
Phase 2 of the trial was conducted in a well-ventilated private parking area. Two single-decked buses (one control bus and one test bus) were used for the second trial. Each bus was an air-conditioned 53-seater with internal dimensions of 10.5 m (L) × 2.5 m (W) × 2 m (H), with an approximate internal volume of 52.5 m 3 , similar to one deck of the double-decker buses used in Phase 1 of the trial (Fig. S3).
In this phase, carriers spiked with either MHV or S. aureus were placed in the buses together and run in the same decontamination cycle. A total of 16 spiked carriers (8 × MHV, 8 × S. aureus) were used in each bus ( Fig. 2A). Twenty ultrasonic humidifiers (Xiaomi Mijia Humidifier 4 L, CAT# MJJSQ02LX) were also placed along the aisle in each bus for humidification. ( Fig. 2A and C).

Phase 2 -efficacy test conditions
Two experimental conditions were tested in Phase 2: the first considered the antimicrobial effect of ozone with increased humidity, and the second considered the antimicrobial effect of prolonged exposure to ozone. The decontamination cycle for the first experiment consisted of 20 min of ozone generation (to achieve a concentration of between 13 and 18 ppm), 15 min of humidification to increase the relative humidity to above 90 % concentration, 10 min of incubation, and approximately 40 min for ozone degeneration (Fig. S4.A). The cycle for the second trial comprised 20 min of ozone generation, 150 min (2.5 h) of incubation, and approximately 40 min for ozone degeneration (Fig. S4·B).

Results
Prior to conducting the efficacy trial, we conducted gaseous ozone leak tests to assess the safety of the ozone disinfection trial (S1.1). Additionally, drying tests (S1.2) and residual efficacy tests (S1.3) were performed to confirm the validity of the laboratory test method. Details about these prequalification tests can be found in the Appendix S1. Pre-Qualification Tests.  Due to the operational requirement of a quick turn-around time for decontamination, a short decontamination cycle was initially tested. Gaseous ozone concentration was raised to 13-18 ppm (approximately 15 min) with the degeneration cycle initiated once the concentration was reached ( Fig. 3A and B). The complete decontamination cycle lasted approximately 1 h. One cycle was performed for MHV contaminated carriers, and a second cycle was performed for the S. aureus contaminated carriers.
Thirty-two carriers were spiked with 6.99-log TCID50/carrier of MHV, 16 were placed in the control bus, and 16 were placed in the ozone-exposed bus. The mean recovery of MHV from control carriers was 5.822-log TCID50/carrier. The mean recovery of MHV from ozone exposed carriers was 4.865-log TCID50/carrier (Fig. 3C). Compared to the mean recovery of control carriers, the ozone exposed carriers showed a log reduction of 0.95-log AE 0.16-log SEM (95 % CI -0.62-to 1.28-log reduction) (Fig. 3C).
Similarly, 32 carriers were spiked with 6.74-log CFU/carrier S. aureus, 16 placed in the control bus, and the remaining 16 placed in the ozone exposed bus. The mean recovery of S. aureus from control carriers was 6.59log CFU/carrier, and 5.98-log CFU/carrier from ozone exposed carriers (Fig. 3D). Compared to the mean recovery of control carriers, the ozone exposed carriers showed a log reduction of 0.61-log AE 0.09-log SEM (95 % CI -0.41-to 0.80-log reduction) (Fig. 3D).

Phase 2 -efficacy test
Following lower than expected efficacy data in the first phase, the second phase of trials expanded on investigating additional experimental parameters that could influence disinfection efficacy of gaseous ozone.
The first trial investigated the role of a 10 min incubation with increased humidity in disinfection efficacy of gaseous ozone. After the gaseous ozone concentration of 13-18 ppm was achieved (in approximately 25 min) (Fig. 4A), humidifiers were activated for approximately 16 min (Fig. 4B), during this time the ozone concentration was fluctuating likely due to the fine water droplets interfering with the ozone sensor (sensor was rated for 25-75 % relative humidity). After the relative humidity crossed >90 % both the ozone generator and the humidifiers were switched off (Fig. 4B). A 10 min incubation time was allowed before the ozone degeneration cycle was switched on. For the duration of the experiment, the temperature within both control and test buses were between 27 and 29°C.
The test and control carriers were spiked with a 7.8-log TCID50/carrier of MHV or 6.3-log CFU/carrier of S. aureus.
The mean recovery of MHV from control carriers was 4.17-log TCID50/ carrier. The mean recovery of MHV from ozone exposed carriers was 1.46log TCID50/carrier (Fig. 4C), and when compared to the mean recovery of control carriers, the ozone exposed carriers showed a MHV log reduction of 2.71-log AE 0.19-log SEM (95 % CI -2.29-to 3.13-log reduction) (Fig. 4C).
The mean recovery of S. aureus from the control carriers was 5.82-log CFU/carrier while the recovery from ozone-exposed carriers was 2.52-log CFU/carrier, demonstrating a log reduction of 3.29-log AE 0.51-log SEM (95 % CI -2.18-to 4.40-log reduction) (Fig. 4D). These results suggested that increased humidity improved antimicrobial efficacy of gaseous ozone, as compared to the results obtained from Phase 1 of the study.
A second experimental condition in phase 2 investigated the effect of increased exposure duration (2.5 h) on disinfection efficacy. The gaseous ozone concentration was maintained at 13-18 ppm (Fig. 5A). For the duration of the decontamination cycle, the temperature in both control and test buses were between 29 and 31°C, with relative humidity (ambient) maintained at approximately 80 %. The mean recovery of MHV from control carriers was 4.95-log TCID50/ carrier, while the mean recovery of MHV from ozone exposed carriers was 1.3-log TCID50/carrier, accounting for a log reduction of 3.65-log AE 0.06log SEM (95 % CI -3.51-to 3.78-log reduction) (Fig. 5B).
The mean recovery of S. aureus from the control carriers was 6.03-log CFU/carrier while the recovery from ozone-exposed carriers was 1.3-log CFU/carrier, demonstrating a log reduction of 4.73-log AE 0.06-log SEM (95 % CI -4.58-to 4.88-log reduction) (Fig. 5C). These results demonstrated that prolonged exposure to gaseous ozone resulted in improved disinfection efficacy.

Discussion
Although antimicrobial properties of gaseous ozone have been wellreported in literature in laboratory settings (Table S1), there have been limited studies in field settings. To our knowledge, this is the first field trial done in a bus setting testing the disinfection efficacy of gaseous ozone against a virus and a bacterium in parallel. Furthermore, we have adapted the standard BS EN 17272:2020 to a public bus setting, and this protocol can be replicated and used in future studies. This study reports the first field trial assessing the disinfection efficacy of gaseous ozone in a realworld public bus setting. At an optimal gaseous ozone regime of 18 ppm with high humidity (>90 %), decontamination efficacies of >3-log reductions were achieved for both murine hepatitis virus and S. aureus. Our findings highlight the utility of gaseous ozone disinfection of environmental surfaces in enclosed spaces, though parameters had to be adjusted to achieve disinfection efficacy.
To fit the operational requirements of a public transport operator, the initial phase of our study investigated the disinfection efficacy of gaseous ozone (15 ppm concentration) with a disinfection cycle of a shorter duration (approximately 15 min ozone generation time). The resulting log reduction observed in this setup was poor, with 0.96-log for MHV and 0.61-log for S. aureus.
To increase disinfection efficacy, we next explored the relationship between ozone exposure time and disinfection efficacy. This was supported by findings by Yano et al. reporting efficacious results using very low ozone concentrations of 1 ppm and 6 ppm, for 60 and 55 min respectively; revealing the strong correlation between ozone disinfection efficacy and incubation time (Yano et al., 2020). Grignani et al. (Grignani et al., 2021b) did a comprehensive literature review on the effectiveness of ozone as disinfectant and concluded that low concentrations of ozone for longer exposure times can have similar results to shorter exposures at higher concentrations. In one study using influenza virus that used ozone concentrations similar to what was used in our bus trial, H1N1 virus was exposed to gaseous ozone at a concentration of 10 ppm for 3.5 h or 20 ppm for 2.5 h; and reported a > 4-log reduction in H1N1 titers at 10 ppm, and > 5-log reduction at 20 ppm (Tanaka et al., 2009). Therefore, in our follow-up trial in phase 2 of our study, we prolonged the exposure of the challenge organisms to gaseous ozone to 2.5 h, and showed a disinfection efficacy of 3.65-to 4.73-log; a substantial improvement from the phase 1 result.
Varying durations for ozone disinfection cycles have been previously reported. Hudson et al. (Hudson et al., 2009) obtained the maximum antiviral efficacy against 12 different viruses at a concentration of 25 ppm of ozone for 15 min. For human coronavirus 229E, a 3-log reduction in virus titers on contaminated face masks was reported at a gaseous ozone concentration of 120 ppm, for as short as 1 min post exposure (Lee et al., 2021). However, it must be noted that unlike our study which explored the environmental decontamination of enclosed spaces, the test setting was for decontamination of specific items (i.e. masks) which were placed in close proximity (1 cm) from the ozone source. Hence, the high ppm and close distance of the generator would not be feasible for the purposes of environmental decontamination of enclosed spaces.
In contrast, others had also demonstrated successful disinfection efficacy of mumps virus (>5-log reduction) using much lower ozone concentrations of 20-25 ppm, albeit with a longer contact time of 20 min (Pekovic and Kacimi, 2015). Of note, the researchers also intentionally increased the relative humidity of the test chamber to >90 %. Mazur-Panasiuk et al. (Mazur-Panasiuk et al., 2021) also demonstrated that increased humidity favored MHV disinfection efficacy. Similarly in the second phase of our trial, we explored using increased humidity, while maintaining a short duration of ozone exposure. In agreement with literature, we showed that a low concentration gaseous ozone used in conjunction with an increased relative humidity resulted in improved disinfection efficacy of 2.71-log against MHV and 3.29-log against S. aureus.

Conclusion
In summary, we demonstrated that different durations of gaseous ozone exposure (both short and long) have shown to be efficacious. Of note, shorter durations of gaseous ozone required close proximity to ozone source and relied on achieving higher ppm which are not feasible in large spaces. We have demonstrated that for shorter durations of gaseous ozone exposure, a large space could be adequately decontaminated at lower gaseous ozone concentrations if relative humidity could be increased. Although technically a higher concentration could reduce disinfection time, our study protocol had a maximum ozone concentration of~20 ppm, due to the use of in situ ozone generators. These generators produce ozone from oxygen in the air that becomes progressively concentrated with ozone, making difficult to increase even more the concentration. Furthermore, high ozone concentrations present a hazard in an operational setting by rising the safety risk to operators.
The occurrence, stability, and infectivity of SARS-CoV-2 varies depending on the surface materials where the droplets land. Although we only tested here one type of surface (stainless steel), the high virus survivability on stainless steel makes this the most stringent condition (van Doremalen et al., 2020). Although the interior of a bus is made of different types of material including porous surfaces, recent studies demonstrated that gaseous ozone is able to penetrate these surfaces effectively (Uppal et al., 2021;Wolfgruber et al., 2022).
As with most global cities, metropolitan buses are an important part of the Singapore public transport system. Approximately half the population of Singapore commute on them every day, and a safe, effective, and quality-controlled method for sanitizing the fleet of vehicles is called as we continue to live with COVID-19. Although the initial disinfection duration could not be achieved, using a longer disinfection cycle or higher humidity, our field trial demonstrated the utility and efficacy of gaseous ozone disinfection in public transport buses. Where bus operators could afford the time, gaseous ozone decontamination could be a less laborious, safe and efficacious solution for disinfection.
Our results agree with the literature on the importance of relative humidity, duration of exposure and concentration as important parameters for optimal efficacy of gaseous ozone as a disinfectant. Due to its wide antimicrobial range, ozone is considered one of the best biocides against microorganisms, and it can provide a valuable decontamination tool for the removal of pathogens in public transport and other larger indoor venues.