key: cord-0956385-tq2w4foj
authors: Albert, Simon; Amarilla, Alberto A.; Trollope, Ben; Sng, Julian D.J.; Setoh, Yin Xiang; Deering, Nathaniel; Modhiran, Naphak; Weng, Sung-Hsia; Melo, Maria C.; Hutley, Nicholas; Nandy, Avik; Furlong, Michael J.; Young, Paul R.; Watterson, Daniel; Grinham, Alistair R.; Khromykh, Alexander A.
title: Assessing the potential of unmanned aerial vehicle spraying of aqueous ozone as an outdoor disinfectant for SARS-CoV-2
date: 2021-02-26
journal: Environ Res
DOI: 10.1016/j.envres.2021.110944
sha: 3a1abeae1086ad1fb179d81b4d11ce55b48f6000
doc_id: 956385
cord_uid: tq2w4foj

The COVID-19 pandemic has revealed gaps in our understanding of safe, effective and efficient means of disinfecting high use public spaces. Whilst this creates an opportunity for development and application of innovative approaches such as unmanned aerial vehicle (UAV) based disinfection, unregulated outdoor disinfection using chlorine has led to environmental and public health risks. This study has quantified the efficiency, safety and efficacy of UAV-based spraying of aqueous ozone. Optimised UAV flight characteristics of 4.7 km/h at 1.7 m elevation spraying 2.4 L/min were able to provide >97% and >92% coverage of a 1 m and 2 m wide swath respectively. During spraying operations using 1 mg/L aqueous ozone, atmospheric concentrations of ozone remained within background levels (<0.04 ppm). Highly efficient inactivation of two different isolates of SARS-CoV-2 virus was achieved at aqueous ozone concentrations of 0.75 mg/L after an incubation period of only 5-minutes, with 0.375 mg/L achieving 82-91.5% inactivation in this time. Exposure of diamondback moth larvae and parasitic wasps to 1 mg/L aqueous ozone did not significantly affect their survivorship. These results indicate for the first time that aqueous ozone may provide the required balance between human and environmental safety and viral inactivation efficacy for targeted application in high risk outdoor settings.

Since declared by the World Health Organisation (WHO) on 11 March 2020 as a global health crisis, the COVID-19 pandemic has resulted in >31 million infections and >1 million deaths with the daily number of infections and deaths still remaining very high (https://covid19.who.int). The rapid escalation of the COVID-19 pandemic has in many cases led to ad-hoc large-scale dispersion of chlorine or alcohol-based disinfectants across public areas ( Figure 1 ). Despite recommendations from the WHO that disinfection outdoors is not required (WHO 2020), widespread spraying of public areas has occurred in numerous locations eg China, Russia, India, Philippines and Iran (Nabi, Wang et al. 2020 , Service 2020 . Recent assessments indicate SARS-CoV-2 can remain active on surfaces for up to 28 days (Riddell, Goldie et al. 2020) . High touch areas of partially covered facilities such as sports stadiums and park/playground equipment have been an area of particular concern in many jurisdictions.

Outdoor disinfection is largely unregulated and often sits between medical therapeutic regulation of indoor disinfection and agricultural chemical regulation for outdoor settings.

The urgency with which outdoor disinfection is often applied frequently limits the possibility of even the most basic environmental and public health risk assessments being conducted.

Any proposed widespread spraying of outdoor spaces must consider the potential for secondary adverse impacts alongside the efficacy of outdoor spraying (SanJuan-Reyes, Gómez-Oliván et al. 2021) . The direct exposure to spraying of chlorine as disinfectant during Ebola outbreaks in West Africa led to significant respiratory, skin and eye impacts on J o u r n a l P r e -p r o o f patients, residents and health workers (Mehtar, Bulabula et al. 2016) . Whilst there have been no quantitative studies to date on the secondary impacts of outdoor disinfection spraying in response to SARS-CoV-2, potential impacts on water quality of natural systems from increased use of chlorine based disinfectants have been highlighted (Chu, Fang et al. 2020 ).

Anecdotal evidence from Chongqing, in southwest China suggests wildlife mortality occurred as a result of high intensity outdoor disinfectant application (Yingzi 2020) .

Attention has also been recently drawn to the environmental risks of broad scale outdoor disinfection (Nabi, Wang et al. 2020) . Three principal groups of chemical disinfectants are typically used for virus inactivation; alcohol based (>70%), quaternary ammonium and oxidisers (chlorine, hydrogen peroxide or ozone) (Wigginton, Pecson et al. 2012 , Hora, Pati et al. 2020 . Despite safety and health concerns (Gorguner, Aslan et al. 2004 , Medina-Ramón, Zock et al. 2005 ) chlorine bleach at 1000-5000 mg/L continues to be the most commonly used disinfectant product due to its low cost, ease of use, deodorising and efficacy across a broad spectrum of microorganisms, including viruses (Kampf, Todt et al. 2020) . Oxidisers such as ozone and hydrogen peroxide in contrast, offer promise as a safer alternative. Environmental and public health risks are J o u r n a l P r e -p r o o f alleviated as these oxidisers rapidly convert to oxygen and water without leaving residual toxicity on disinfected surfaces.

Ozone gas is routinely used as an effective disinfectant in the medical industry. In uncontrolled outdoor environments, the gaseous phase of ozone is neither practical nor safe due to potential for exposure of operators to harmful concentrations of ozone (>0.1 ppm).

Ozone can be dissolved in water to form ozonated water, or 'aqueous ozone', which can deliver a liquid form of ozone to surfaces as a disinfectant. The aqueous form of ozone has been utilised as a disinfectant in wastewater treatment (Von Sonntag and Von Gunten 2012), food and livestock industries (Aslam, Alam et al. 2019) , and as a commercial disinfectant (Martinelli, Giovannangeli et al. 2017) . Studies have also demonstrated the effectiveness of aqueous ozone as a hand sanitiser in hospital settings, with improved efficacy and lower secondary impacts (eg skin irritation) compared to traditional alcohol-based sanitisers (Breidablik, Lysebo et al. 2019) . Aqueous ozone can effectively inactivate virus through disrupting the proteins and lipids of the virus spikes (Tizaoui 2020) . However, the instability of aqueous forms of ozone (20-minute half-life) has limited the widespread uptake as a commercial disinfectant. Recent advancements in understanding aqueous ozone chemistry (Eriksson 2005) has led to commercially available ozone generators producing a stabilised form of aqueous ozone, increasing its stability to several hours.

The widespread disruption from the COVID-19 pandemic has driven innovation in the utilisation of autonomy, robotics and artificial intelligence across a range of settings (Zeng, Chen et al. 2020) . Numerous jurisdictions globally have re-purposed agricultural spraying unmanned aerial vehicles (UAVs) to support disinfection efforts in response to COVID-19.

To date, this effort has largely been reactionary with little pre-planning, testing or assessment of efficacy. A range of disinfectant solutions (chlorine (1000-5000 ppm), alcohol, hydrogen peroxide) have been utilised alongside early trials of high intensity UV light for indoor J o u r n a l P r e -p r o o f applications. However, the lack of systematic testing coupled with significant concerns around public and environmental safety of widespread chlorine bleach spraying, has limited the effectiveness and continuation of these UAV disinfection programs.

Using conventional disinfection delivery systems in outdoor areas with complex structures is a major challenge due to health and safety risks to applicators, as well as the logistics of moving large volumes of liquid in confined areas. In addition, where rapid application of disinfectant is required (eg between sporting or entertainment events), this presents a challenge to current manual delivery systems. Furthermore, the scaling of adequate and rapid disinfectant coverage using existing manual delivery systems requires greatly increased person hours. The advent of aerial spraying using UAVs represents a feasible approach to overcome these challenges, as a single system is able to provide rapid coverage within complex outdoor structures.

Here we present the first quantitative assessment of the potential of utilising unmanned aerial vehicles to disinfect complex outdoor environments, demonstrating the efficacy of the aqueous form of ozone to inactivate SARS-CoV2 and show its environmental safety in two experimental insect models.

This study assessed three core areas; efficiency of spray coverage, environmental and operator safety, and efficacy in inactivation of SARS-CoV-2 using aqueous ozone.

A critical component to ensure both effective and efficient UAV based spraying of disinfectant is to assess coverage and drift of spray under different flight conditions. Test strips of 6 m x 5 m 160 gsm geofabric (Grunt Non-Woven Geotextile Membrane, Preston, J o u r n a l P r e -p r o o f Australia) were used to quantify both coverage and density of spray from the UAV (DJI Agras MG-1, Shenzhen, China). Two hundred mL of red food dye (Pillar Box Red Food Colour, Queen, Brisbane, Australia) was added to the 10 L spray tank of water and trials were conducted using flight speeds of 4.7 km/h and 5.8 km/h. Spray heights of 1, 1.7 and 2 m were tested as well as a combination of coarse or mist nozzles either mounted below the motors or on a rigid boom. Following each spray trial a high resolution aerial image was taken from 10 m elevation using a UAV (DJI Mavic 2 Pro, Shenzhen, China). These high resolution images with 3 mm pixel size were georectified (ArcGIS 10.5, ESRI) prior to further analysis.

Linear Spectral Unmixing, a sub-pixel classification method was used to calculate the intensity of the spray. The process offers superior results where the feature size is relatively smaller than the pixel size and with limited spectral bands (Kamal and Phinn 2011) . As a result of spectral unmixing, fractional cover maps were generated and a rule-based feature extraction model was applied to identify the concentration of the red dye in each pixel. The fractional images provide endmember-based data with pixel values ranging between 0 -1.

Pixels with complete coverage of red dye had high pixel value, whereas the areas with less or no spray will provide values close to zero. The areas were then divided into the four distinctive spread classes of high, moderate, low and no spread ( Figure 2 ).

The stability of dissolved ozone in four solutions were assessed: Brisbane municipal tap water pH=7.4; Brisbane municipal tap water filtered through a cation exchange resin (Tersano SAO-24, referred here as Municipal-Tersano) pH=3.2; deionised water pH=7.1; and deionised water filtered through a cation exchange resin (Tersano SAO-24, referred here as Deionised-Tersano), pH=3.1. Five liters of each solution was ozonated for 10 min by recirculating through an industrial ozone generator (Grenof, Brisbane, Australia). The ozone J o u r n a l P r e -p r o o f generator uses corona discharge to produce 5 g/h of ozone which is introduced into the water flow via a venturi injector nozzle. Dissolved ozone concentrations were monitored using a Cronos Ozosense Analyser (Process Instruments, Burnley, UK) every 10 min for the first 4 h and every 30 min for the following 4 h.

Cell line and virus isolates: Vero E6, African green monkey kidney cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% foetal bovine serum (FBS) and maintained at 37 ºC with 5% CO 2 . Two different isolates of SARS-CoV-2, QLD02

(GISAID accession EPI_ISL_407896) and QLD935 (GISAID accession EPI_ISL_436097) were obtained from the Queensland Department of Health as passage 2 in Vero E6 cells and passaged once more in Vero E6 cells to generate virus stocks. The virus titres were then determined by immuno-plaque assay (iPA) as described below.

Virucidal activity assay: viral isolates, each with an estimated 1-5x10 5 foci forming units (FFU)/mL, were prepared in 1.8% NaCl solution in deionised and filtered water (pH=6.0). 

Operator safety is a key consideration for any disinfection operation, therefore a field test was undertaken to examine atmospheric ozone levels during UAV spraying operations. The Australian occupational limit for atmospheric ozone over an 8 h period is 0.1 ppm and it is important to ensure pilots as well as support crew are not exposed to atmospheric ozone J o u r n a l P r e -p r o o f levels above this limit during aerial spraying operations. An atmospheric ozone logger (Aeroqual Inc., Series 500 -Portable Ozone Monitor) was placed 20 cm above ground level directly in the spray zone to capture the likely maximum atmospheric ozone levels during spray operations. Spray tests using 1.5 mg/L aqueous ozone were undertaken on 20th May 2020 at a spray height 2 m above the ground with a spray duration of 2 min to maximise local atmospheric ozone levels. 

Ozone is highly unstable in the dissolved phase in water, which is dependent on pH, temperature and presence of metal ions and other impurities (Eriksson 2005) . We assessed Municipal water (dark blue, circle), deionised water (red, square), municipal water prefiltered through Tersano cation exchange resin (purple, circle), deionised water pre-filtered through Tersano cation exchange resin (green, triangle).

The current pandemic of SARS-CoV-2 provides an ideal example to use aqueous ozone as a disinfectant. Two clinical isolates of SARS-CoV2; an early Australian isolate QLD02 sampled from a patient on 30/01/2020, and the more recent isolate QLD935 sampled from a patient on 25/03/2020 were tested here for inactivation by aqueous ozone. QLD935 isolate represents the currently dominating virus isolate with a characteristic D614G mutation in a Spike protein which potentially increases virus infectivity (Korber, Fischer et al. 2020) . Both viral isolates were efficiently inactivated by as little as 5 minutes incubation with 1.5 mg/L and 0.75 mg/L of aqueous ozone with more than 1.7 log 10 reduction ( Figure 4A ). Thirty minutes incubation produced similar results for these ozone concentrations with complete inactivation of both virus isolates ( Figure 4B ). In fact, no virus was detected at all in treatments with these ozone concentrations with the limit of detection being 2.9 log 10 FFU/ml (400 FFU/ml). Further reduction in ozone concentration to 0.375 mg/L decreased efficiency of inactivation (Figure 4) , however, we still observed inactivation of 91.5% (1.07 log 10 ) for QLD02 isolate 82% (0.74 log 10 ) for QLD935 virus isolate after 5 min incubation ( Figure 4A ), or 92% (1.09 log 10 ) for QLD02 virus isolate and 84.5% (0.81 log 10 ) of QLD935 virus isolate after 30 min incubation ( Figure 4B ). based on the mean log 10 reduction (L) was calculated using the formula: P= (1-10 -L ) x 100.

The safety of UAV operators was assessed by comparing background ozone levels to insitu ozone sensors within spraying operation. Ambient ozone levels at the nearby (less than 7 km) Southport air quality monitoring station (https://apps.des.qld.gov.au/air-quality/) ranged between 0.002 to 0.04 ppm over the duration of the sampling month ( Figure 5A ). Local measurements from the spray area showed background levels to be 0.015 ppm immediately prior to testing ( Figure 5B ). The highest reading recorded during spraying of stabilised aqueous ozone was 0.026 ppm within the spray zone ( Figure 5B ). Monitoring of atmospheric ozone levels 5 m downwind from the spray zone showed concentrations similar to background levels of 0.015 ppm. Based on these initial findings, the atmospheric exposure to ozone during spraying is unlikely to be a health concern for operators as the observed levels all lay within the range for background ambient levels.

J o u r n a l P r e -p r o o f 

Environmental safety is an important consideration for disinfection operations. We evaluated the potential effect of aqueous ozone and other disinfectants on the survivorship of larvae of the worldwide crucifer crop pest, the diamondback moth, and adults of its parasitoid, D. semiclausum, which has been widely introduced for biological control of the pest (Furlong, Wright et al. 2013) . Parasitoids are typically far more sensitive to insecticides and other xenobiotic chemicals than their insect hosts (Devine and Furlong 2007, Kim, Lee et al. 2019 ) and as such their responses can serve as important bio-indicators of xenobiotic compounds in the environment (Furlong, Zu-Hua et al. 2004 ) Bleach and hydrogen peroxide significantly affected the survival of diamondback moth larvae ( Figure unintended consequences on the environment in some instances (Nabi, Wang et al. 2020) , and a placebo effect or "disinfection theatre" in others (Lowe 2020).

As the pandemic transitions to local community transmission, developing effective yet safe disinfection protocols will be critical to enable large scale outdoor events to proceed safely.

Examples from Italy have demonstrated that high density community events, such as sporting matches, can lead to rapid escalation of community transmission and are a critical gap in knowledge (Sassano, McKee et al. 2020) . The complexity and scale of surfaces in sporting stadiums and the short turn around between events limits the potential of targeted manual disinfection of high touch surfaces. Recent work demonstrating SARS-CoV-2 may remain active on surfaces for up to 28 days highlights the need to investigate new approaches for disinfection (Riddell, Goldie et al. 2020) . Aerial spraying of traditional disinfectants (eg chlorine bleach) introduces too many environmental and human health risks to make it a viable option in reducing COVID-19 transmission risk. The potential of ultraviolet (UV) radiation has recently been demonstrated as a cost-effective sterilizing method for SARS-CoV-2 (Bianco, Biasin et al. 2020) , however the integration of high energy UV systems on aerial platforms is less developed than spraying systems. Aqueous ozone may provide the ideal balance between efficacy, cost effectiveness and safety to be utilised at scale for disinfection control of large community gatherings. Recent studies showed effectiveness of J o u r n a l P r e -p r o o f ozone spraying in the inactivation of aerosolized airborne viruses (Dubuis, Dumont-Leblond et al. 2020) indicating that in addition to intended inactivation of SARS-CoV-2-contaminated surfaces aqueous ozone spraying may also aid in inactivating SARS-CoV-2 potentially remaining in the air after large community gatherings such as sporting events.

Although aqueous ozone has been utilised in limited health care settings, uptake across the cleaning and disinfection industry has been limited by a number of factors including stability of ozone in solution, presence of ozone-consuming compounds such as organic substances, and of by-products from ozone action on organic compounds (Khadre, Yousef et al. 2001 ). In our experiments, deionised and filtered water showed the best performance in providing extended ozone solubility and stability. The highest antiviral activity also required virus dilution in NaCl solution prepared in deionised and filtered water. Virus preparations in other solutions, such as complete cell culture media with or without foetal bovine serum were inactivated less efficiently (data not shown). Whether these factors could potentially reduce the efficacy of SARS-CoV-2 inactivation by aqueous ozone spraying in the settings of large community gatherings remains to be determined. Field trials of aqueous ozone spraying in large stadiums using model microbial organisms as the readout will provide highly valuable data to further assess the potential of aqueous ozone as an effective disinfectant.

Background levels of exposure to high atmospheric ozone and particulate matter concentrations have been linked to increased transmission and severity of COVID-19 (Fattorini and Regoli 2020 , Zhu, Xie et al. 2020 , Bontempi 2020a , Bontempi 2020b . The low atmospheric concentrations of ozone released during outdoor spraying indicates these broader relationships between ozone pollution and transmission and severity of COVID-19 will not be influenced by disinfection spraying. Despite the human health risks associated with ozone exposure (Ito, Inoue et al. 2005 , Wang, Wild et al. 2020 , our field trials suggest aqueous ozone sprayed in outdoor settings maintains atmospheric concentrations below regulatory levels, thus ensuring operator safety. 

An optimised high-throughput immuno-plaque assay for SARS-CoV-2

Sanitization Potential of Ozone and Its Role in Postharvest Quality Management of Fruits and Vegetables

UV-C irradiation is highly effective in inactivating and inhibiting SARS-CoV-2 replication

The europe second wave of COVID-19 infection and the Italy "strange" situation

First data analysis about possible COVID-19 virus airborne diffusion due to air particulate matter (PM): The case of Lombardy (Italy)

Ozonized water as an alternative to alcohol-based hand disinfection

Intensified Disinfection Amid COVID-19 Pandemic Poses Potential Risks to Water Quality and Safety

Insecticide use: Contexts and ecological consequences

Ozone efficacy for the control of airborne viruses: Bacteriophage and norovirus models

Ozone chemistry in aqueous solution

Role of the chronic air pollution levels in the Covid-19 outbreak risk in Italy

Diamondback Moth Ecology and Management: Problems, Progress, and Prospects

Experimental Analysis of the Influence of Pest Management Practice on the Efficacy of an Endemic Arthropod Natural Enemy Complex of the Diamondback Moth

Reactive Airways Dysfunction Syndrome in Housewives Due to a Bleach-Hydrochloric Acid Mixture

Increased Use of Quaternary Ammonium Compounds during the SARS-CoV-2 Pandemic and Beyond: Consideration of Environmental Implications

Mechanism of site-specific DNA damage induced by ozone

Hyperspectral Data for Mangrove Species Mapping: A Comparison of Pixel-Based and Object-Based Approach

Persistence of coronaviruses on inanimate surfaces and their inactivation with biocidal agents

Microbiological Aspects of Ozone Applications in Food: A Review

Transfer and biological effects of arsenate from soil through a plant-aphid system to the parasitoid wasp, Aphidius colemani

Tracking changes in SARS-CoV-2 Spike: evidence that D614G increases infectivity of the COVID-19 virus

Toward COVID-19 secure events: considerations for organizing the safe resumption of major sporting events

Water and air ozone treatment as an alternative sanitizing technology

Asthma, chronic bronchitis, and exposure to irritant agents in occupational domestic cleaning: a nested case-control study

Deliberate exposure of humans to chlorine-the aftermath of Ebola in West Africa

Massive use of disinfectants against COVID-19 poses potential risks to urban wildlife

The effect of temperature on persistence of SARS-CoV-2 on common surfaces

COVID-19 in the environment

Transmission of SARS-CoV-2 and Other Infections at Large Sports Gatherings: A Surprising Gap in Our Knowledge

Does disinfecting surfaces really prevent the spread of coronavirus?

Human Monoclonal Antibody Combination against SARS Coronavirus: Synergy and Coverage of Escape Mutants

Ozone: A Potential Oxidant for COVID-19 Virus (SARS-CoV-2)

Chemistry of ozone in water and wastewater treatment

Health impacts of long-term ozone exposure in China over 2013-2017

Cleaning and disinfection of environmental surfaces in the context of COVID-19 Interim guidance

Virus Inactivation Mechanisms: Impact of Disinfectants on Virus Function and Structural Integrity

Wildlife unexpected outbreak casualty

From high-touch to high-tech: COVID-19 drives robotics adoption

Association between short-term exposure to air pollution and COVID-19 infection: Evidence from China

We thank Mark Phillips and Nige Austin of The Ripper Group for facilitating field trials, Simon Bayley and David Redfern of Grenof for supply of ozone generator, Ecofy and Tersano Inc. for technical advice on ozone stability, and Queensland Health Forensic & Scientific Services, Queensland Department of Health for providing SARS-CoV-2 isolates.Graphical abstract was produced by graphic artist Ryo Tsukui.

J o u r n a l P r e -p r o o f Highlights • Aqueous ozone achieves highly efficient inactivation of SARS-CoV-2 • Spraying aqueous ozone from unmanned aerial vehicles (UAV) has potential for broad scale disinfection • UAV-based spraying can achieve >97% coverage with optimised flight characteristics • Aqueous ozone represents lower environmental and human health risks than other disinfectants J o u r n a l P r e -p r o o f

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: J o u r n a l P r e -p r o o f