Rapid inactivation of influenza A virus at low humidity and low HOCl concentration

We evaluated the survival percent (SP) of H1N1 influenza A virus in moist aerosols and aerosols dried using a diffusion dryer under environmental conditions subjected to a HOCl_(g) concentration of 20 ppb at a relative humidity of 50% for a gas contact time of 1.8 s (Fig. 1a). For influenza viruses in moist aerosols, a significant reduction in infectivity of 99.84% was observed compared to that for the 0 ppb blank (Wilcoxon signed-rank test, p = 0.016, n = 6, Z = 2.097, r = 0.856). However, no significant reduction in infectivity was detected for viruses in dry aerosols (Wilcoxon signed-rank test, p = 0.078, n = 6, Z = 1.468, r = 0.599). The 20-ppb concentration is fully applicable in real human environments. These findings suggest that HOCl_(g) effectively inactivates influenza viruses in moist aerosols released by infected individuals. Additionally, the substantial difference between moist and dry aerosols under the same environmental conditions is noteworthy as this implies that the presence of moisture on the virus surface is critical. Wang et al.³¹ indicated that the inactivation effect of ClO₂ is linked to water vapour condensing on the surface, acting as a carrier for ClO₂, which is easily dissolved. A similar inactivation process is believed to have contributed to the observed effect of HOCl in the present study. Miyaoka et al. also demonstrated that spraying 100–500 ppm HOCl_(aq) on aerosolised infectious bronchitis virus containing 0.5% foetal bovine serum (FBS) led to inactivation within a few seconds²⁵. This effect can be ascribed to either coagulation between the droplet-shaped airborne virus and microdroplets containing HOCl_(aq) prior to volatilisation or to dissolution of HOCl_(g) into the virus-laden droplets, both of which reduced infectivity. These findings suggest that the combined effect of the virus, moisture, and HOCl was responsible. In contrast, in the present study, the droplet-form airborne virus came into contact with HOCl_(g), which dissolved into the droplets; thus, the combination of virus, moisture, and HOCl contributed to rapid inactivation.

Survival percent of influenza virus (a) in wet/dry aerosol subjected to HOCl [20 ppb, 50% relative humidity, 1.8 s] and (b) in wet aerosol subjected to 10 and 20 ppb concentrations of HOCl [50% relative humidity, 1.8 s]. The box represents the 25th to 75th percentile range, the whisker plot shows the maximum and minimum values, the thick line within the box denotes the median, and the ■ symbol represents the average value.

Furthermore, Imoto et al. confirmed a 2.56 log inactivation effect against the influenza virus by spraying HOCl_(aq) with a contact time of 5 min under low concentration conditions (HOCl_(g) : 20ppb)²⁶. In contrast to this study, they sprayed microdroplets containing HOCl_(aq) onto aerosolised viruses that had evaporated several minutes after spraying. However, both conditions commonly involve the presence of viruses, moisture, and HOCl, suggesting that significant effects can also be expected in low-concentration HOCl gas due to the presence of moisture. Next, we consider real-world spaces and discuss the results for safer and more comfortable gas concentrations, as well as realistic humidity ranges. The SP of H1N1 influenza A virus in aerosols containing moisture under environmental conditions of a HOCl_(g) concentration of 10 ppb, gas contact time of 1.8 s, and relative humidity of 50% was evaluated (Fig. 1b). Even with a reduced HOCl concentration of 10 ppb, a significant inactivation effect of over 99% was observed (Wilcoxon signed-rank test, p = 0.031, n = 5, Z = 1.923, r = 0.860). Compared to 20 ppb, the inactivation effect was reduced by a factor of 1.33 in terms of logarithmic reduction. This decrease is likely due to the reduction in the equilibrium concentration within the droplet moisture as the air concentration decreases, leading to a lower adsorption rate. The 10-ppb concentration is not only 1/50 of the safety standard concentration but also a level that is detectable by the human sense of smell with minimal discomfort, making it a more realistic concentration for practical use.

The Log reduction for H1N1 influenza A virus in aerosols containing moisture under environmental conditions of a HOCl_(g) concentration of 10 ppb, gas contact times of 1.8 and 3.9 s, and relative humidity ranging from 30 to 50% were evaluated (Fig. 2). The relative humidity range of 30–50% was selected based on the average relative humidity in homes in Japan during winter³². Under identical relative humidity conditions, the reaction time was directly proportional to the inactivation effect. Additionally, under the same reaction time, a higher relative humidity resulted in a greater inactivation effect. This can be explained using the concept of CT value, where extending the reaction time increases the contact time between the virus and HOCl, leading to a higher inactivation effect.

Survival percent of influenza virus in wet aerosol subjected to a HOCl concentration of 10 ppb at different relative humidities (RH) (30, 40, 50%RH) and gas contact times (1.8, 3.9 s). The number of trials (n) was n = 3 for 1.8 s and n = 5 for 3.9 s at 30% RH; n = 8 for 1.8 s and n = 5 for 3.9 s at 40% RH. The data are presented as the mean ± 1 SD of ‘n’ independent samples.

Conversely, for droplet-like aerosols containing moisture, the effect of relative humidity was expected to be minimal because the droplets inherently retained the moisture acting as a carrier, and the difference in evaporation time was small—only a few tens of milliseconds. However, a sizeable difference of 3.06-fold was observed between 30 and 50% relative humidity³³.

Previous studies have shown that the final equilibrium diameter and surface area of aerosol droplets change with evaporation under varying relative humidity conditions, and the degree of inorganic salt crystallisation within the droplets also differs^34,35. Based on these droplet properties, it is thought that in a higher-humidity environment, the droplets maintain their liquid state. Consequently, this allows more HOCl to readily partition into the aqueous phase, which is considered a major factor in enhancing the inactivation effect. In contrast, in a low-humidity environment (e.g., 30% relative humidity), salts crystallise, thereby promoting the evaporation of moisture from the aerosol. This prevented HOCl_(g) from dissolving effectively, leading to a reduced inactivation effect. Furthermore, an increased surface area is believed to contribute to a higher inactivation effect by enhancing the initial adsorption, even in situations where diffusion within the droplet is the rate-limiting step. Additionally, the humidity response of O₃ gas absorption characteristics, as reported by Shiraiwa et al.³⁶, suggests that the viscosity of the phase decreases with increasing relative humidity, improving the diffusibility of the components. The partitioning of HOCl from the gas phase is expected to undergo significant changes during the evaporation process. Therefore, future studies should investigate the changes in HOCl concentration and distribution inside the liquid, along with the inactivation mechanism.

Here, we confirmed an inactivation effect of over 99% in just 3.9 s, even in a 40% relative humidity environment, which is considered the lower limit of comfortable humidity according to the American Society of Heating, Refrigeration, and Air-Conditioning Engineers (ASHRAE). The results demonstrate the effectiveness of the inactivation method in a comfortable temperature and humidity environment. In real-world environments, humidification is often used to maintain a comfortable humidity level. As HOCl_(aq) is vaporised to supply HOCl to the air, this inactivation method is highly compatible with such humidification practices.

Comparison with ROS gas to verify solubility differences

Next, we examined whether the difference in solubility of the active substances plays a crucial role in the inactivation effect on these influenza viral aerosols containing significant amounts of water. For comparison with HOCl_(g), we used ROS gas (primarily O₃). The Henry constant of HOCl_(g) is much smaller than that of O₃, thereby making it readily soluble in water^37,38. The gas–liquid equilibrium concentration at 10 ppb of HOCl_(g) was estimated to be 3.3 ppm based on actual measurements³⁹. In contrast, at 10 ppb of O_3(g), the concentration is below 1 ppb at all pH levels and temperatures, based on the Roth–Sullivan equation. This suggests that the differences in solubility could be adequately compared.

The inactivation effects of HOCl_(g) and ROS gas (O₃ concentration: 10 ppb) generated by corona discharge against H1N1 influenza A virus in aerosols containing moisture were evaluated (Fig. 3a). We also compared the results for viruses in aerosols dried using a diffusion dryer. The conditions were a gas contact time of 1.8 s and a relative humidity of 50%. In the case of ROS gas, no significant reduction in virus infectivity was observed in either the water-containing (Wilcoxon signed-rank test, p = 0.563, n = 5, Z = 0, r = 0) or dry aerosols (Wilcoxon signed-rank test, p = 0.656, n = 5, Z = − 0.271, r = − 0.121). In dry aerosols, the lack of water likely explains why no effect was observed with ROS. Based on the CT values reported in previous studies, achieving 99% virus inactivation in dry aerosols with 10 ppb O₃ is estimated to require 42 min of contact, even in a high-humidity environment, which is consistent with the lack of effect observed in our short-term experiment^29,40. Conversely, in wet aerosol, the low solubility of O₃ gas—which accounts for a majority of ROS—may have prevented it from achieving the same rapid inactivation effect as HOCl_(g). Based on these results, it is suggested that the high solubility of HOCl_(g) is a key factor enabling the inactivation of H1N1 influenza A viruses in wet aerosols during short-term spatial exposure.

Survival percent of influenza virus (a) subjected to different chemical disinfectants [10 ppb, 50% relative humidity, 1.8 s]. (b) Subjected to HOCl when 0.3% mucin was added [20 ppb, 50% relative humidity, 1.8 s]. PBS, Phosphate-buffered saline. The box represents the 25th to 75th percentile range, the whisker plot shows the maximum and minimum values, the thick line within the box denotes the median, and the ■ symbol represents the average value.

Klug et al.⁴¹ evaluated the influence of humidity on O₃ and ClO₂ against MS2 in aerosols supplied at a constant humidity. ClO₂ showed a higher inactivation rate than O₃, even under conditions of lower humidity and concentration than O₃. As ClO₂ is a substance that readily partitions into the aqueous phase, similar to HOCl, it possibly exhibits a higher degradation rate than O₃ under higher humidity conditions, which is consistent with the results of this study.

However, factors other than solubility are also thought to influence the effect, as HOCl and O₃ also differ in chemical properties such as oxidation potential and reaction rate constants. It has been reported that the inactivation by O₃ under high-humidity conditions may be assisted by hydrolysis, suggesting that its effective environment could differ from the conditions in our experiment^42,43. Furthermore, Ratnesar-Shumate et al.⁴⁴ reported that differences in the ease of water uptake between Bacillus spores and MS2, as well as the hygroscopicity of components contained in the aerosols, might explain differences in the speed of the oxidation process⁴⁴. Therefore, further investigation is necessary, as the conditions under which inactivation is effective may differ not only with the type of gas but also with the physical properties of the target aerosol.

Effect of the presence of mucin

Thus far, our results have shown that HOCl_(g) exhibits a strong inactivation effect on droplet-containing airborne viruses. However, droplets from actual infected individuals contain proteins that are believed to consume HOCl_(g). Makimura et al.⁴⁵ reported that 0.2% mucin reacts with HOCl_(aq), thereby reducing the free chlorine concentration to 1/2600 or less within 1 min⁴⁵. Therefore, we investigated whether HOCl_(g) is similarly consumed in an aerosol state and whether its inactivation effect on viruses is inhibited.

We assessed the inactivation effect of HOCl_(g) on H1N1 using a nebuliser containing 0.3% mucin to simulate saliva. The inactivation effects of HOCl_(g) with and without 0.3% mucin under conditions of 20 ppb, 50% relative humidity, and 1.8 s were evaluated (Fig. 3b). The reduction in inactivation due to the presence of 0.3% mucin was only a 0.45 log reduction compared to that using phosphate-buffered saline (PBS) alone, which confirmed a significant inactivation effect (Wilcoxon signed-rank test, p = 0.031, n = 5, Z = 1.888, r = 0.844). Pan et al.³⁰ reported that when PBS-mucin droplets containing viruses evaporate, a coffee-ring effect is formed, with viruses and mucin distributed on the droplet surface via capillary flow. Yang & Marr³⁴ also noted that enveloped viruses tend to accumulate on the droplet surface. During evaporation, the presence of the virus on the droplet surface, which retains moisture, is believed to bring the virus closer to the HOCl_(g) adsorbed on the droplet surface and enhanced its reactivity. Despite this, the observed 0.45 log reduction is likely due to mucin consumption of some HOCl_(g). Continued evaporation leads to the loss of moisture causing HOCl_(g) to no longer dissolve, thereby significantly reducing the inactivation effect.