key: cord-0850084-uh928n3q
authors: Luo, B.; Schaub, A.; Glas, I.; Klein, L. K.; David, S. C.; Bluvshtein, N.; Violaki, K.; Motos, G.; Pohl, M.; Hugentobler, W.; Nenes, A.; Krieger, U. K.; Stertz, S.; Peter, T.; Kohn, T.
title: Acidity of expiratory aerosols controls the infectivity of airborne influenza virus and SARS-CoV-2
date: 2022-03-14
journal: nan
DOI: 10.1101/2022.03.14.22272134
sha: c8f7d1fafe7778844d6ac45afe70658183aa8ee4
doc_id: 850084
cord_uid: uh928n3q

Enveloped viruses are prone to inactivation when exposed to strong acidity levels characteristic of atmospheric aerosol. Yet, the acidity of expiratory aerosol particles and its effect on airborne virus persistence has not been examined. Here, we combine pH-dependent inactivation rates of influenza A virus and SARS-CoV-2 with microphysical properties of respiratory fluids under indoor conditions using a biophysical aerosol model. We find that particles exhaled into indoor air become mildly acidic (pH $approx$ 4), rapidly inactivating influenza A virus within minutes, whereas SARS-CoV-2 requires days. If indoor air is enriched with non-hazardous levels of nitric acid, aerosol pH drops by up to 2 units, decreasing 99%-inactivation times for both viruses in small aerosol particles to below 30 seconds. Conversely, unintentional removal of volatile acids from indoor air by filtration may elevate pH and prolong airborne virus persistence. The overlooked role of aerosol pH has profound implications for virus transmission and mitigation strategies.

impact of air composition beyond RH has been overlooked by scientists to date. To the best of our knowledge, the only attempt to inactivate airborne viruses by -likely inadvertently -modulating aerosol pH is the use of acetic acid from boiling vinegar during the 2002/03 outbreak of SARS-CoV-1 (see (21) and Supplementary Material).

Outdoor airborne particulate matter is often highly acidic, with pH values ranging between -1 and +5 (17, 19) . Contrary to expectations, the strength of the acid or base contained in aerosols (expressed by its dissociation constants) may not be the dominant parameter controlling aerosol pH. Rather, the volatility of species is of importance. For example, strong organic acids like HCOOH and CH 3 COOH partition negligibly to aerosol and bear a minor impact on aerosol pH for most atmospherically relevant conditions (22) . In contrast, HNO 3 and NH 3 partition into aerosol particles and impact pH, albeit buffered by the formation of ammonium nitrate.

Indoor aerosol particles have a variety of sources, including outdoor air, human transpiration and respiration, and building materials. Indoor air tends to have lower levels of gas-phase inorganic acids (e.g., HNO 3 ) than outdoor air, owing to their condensation on aerosol particles as well as their efficient removal via deposition on surfaces. Human activities are a source of organic acids and NH 3 (19, 23, 24) , often elevating their levels compared to outdoors.

The ratio of indoor to outdoor concentrations is typically 0.1-0.5 for HNO 3 and 3-30 for NH 3 , causing the pH of indoor aerosol particles to increase compared to outdoor levels.

Operation of humidification, ventilation, and air conditioning (HVAC) systems also affect air composition (25) and, hence, likely the pH of indoor aerosol particles. While many outdoor and indoor aerosol particles are in equilibrium with their environment, this can only be expected for exhaled aerosol if given enough time. In the interim, freshly exhaled aerosol can change its pH considerably.

Exhaled air, before mixing into the indoor air, contains high concentrations of ammonia and is characterized by very high concentrations of CO 2 and high number densities of expiratory aerosol particles. These particles are emitted by breathing, talking, coughing or sneezing, and 3 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 14, 2022. ;  https://doi.org/10.1101/2022.03.14.22272134 doi: medRxiv preprint contain a complex aqueous mixture of ions, proteins and surfactants. Although the pH of exhaled breath condensate has been investigated (26) , there is no study that quantifies the pH of respiratory aerosol -especially when it equilibrates with the acidic or alkaline gases present in the indoor air within a few seconds to minutes of exhalation.

Here, we investigate the role of aerosol acidity in the inactivation of airborne influenza A virus (IAV) and two coronaviruses, SARS-CoV-2 and HCoV-229E in indoor environments. We accomplish this in three steps by first determining the pH-dependent inactivation kinetics of IAV, SARS-CoV-2 and HCoV-229E in bulk samples of representative respiratory fluids, then measuring the thermodynamic and kinetic properties of microscopic particles of these fluids, and finally jointly applying the inactivation kinetics and aerosol properties in a biophysical model to determine inactivation in the aerosol system. We then use the model to investigate the possibility of using gaseous nitric acid (HNO 3 ) in indoor environments at non-hazardous concentrations to lower the pH of respiratory aerosol for a wide range of sizes, and thus to effectively reduce the risk of transmission.

Inactivation kinetics of IAV (strain A/WSN/33), SARS-CoV-2 (BetaCoV/Germany/BavPat1/2020) and HCoV-229E (strain HCoV-229E-Ren) were determined over a pH range from neutral to strongly acidic, after immersion in bulk solutions of synthetic lung fluid (SLF; see Table S1 for composition), mucus harvested from primary epithelial nasal cultures grown at air-liquid interface (nasal mucus) or aqueous citric acid/Na 2 HPO 4 buffer. Figure 1 summarizes the inactivation times (here expressed as the time to reach a 99% infectivity loss) as a function of pH. All viruses were stable in all matrices at neutral pH, with inactivation times of several days. From pH 6 to 4, IAV inactivation times decreased from days to seconds, or by about five orders of magnitude. This decrease was evident in all matrices studied. It is noteworthy that inactivation in nasal mucus, which is most representative of the matrix comprising expiratory aerosol particles, is well described by SLF. However, inactivation times did depend on the SLF concentration. Specifically, we determined IAV inactivation at three different levels of SLF enrichment (1× and 18× SLF, determined experimentally; 24× SLF, determined by extrapolation), corresponding to water activities a w = 0.994, 0.8 and 0.5. This represents the fluid in equilibrium with a gas phase at 99.4%, 80% and 50% RH, i.e. from physiological equilibrium to common indoor conditions. While inactivation times in aqueous buffer, 1× SLF and nasal mucus were very similar, 18× enrichment of the SLF coincided with an increase in inactivation time by up to a factor 56 (blue triangles in Fig. 1 ). This protective effect of concentrated SLF was most prominent around the optimal pH for A/WSN/33 viral fusion of ∼ 5.1 (27) . Coronaviruses were less affected by acidic pH than IAV. Both, SARS-CoV-2 and HCoV-229E remained largely stable down to pH 3, where their inactivation still required 24 hours. When further decreasing pH down to 2, the inactivation times rapidly reduced to < 10 seconds for SARS-CoV-2, but never dropped below 2 hours for HCoV-229E. Compared to aqueous buffer, SLF provided some protection against inactivation below pH 3, both at 1× and 5× SLF concentrations (while measurements for pH < 3 in 18× SLF were not possible due to precipitation). The measured differences in pH-sensitivities between IAV and the coronaviruses may be explained by their different mechanisms of virus entry into host cells. IAV relies on an acid-induced conformational change in haemagglutinin during endosomal entry. This conformational change is irreversible (28) ; if IAV encounters the fusion pH (typically pH < 5.5) outside the host cell, e.g. whilst within an aerosol particle, the acid-triggered haemagglutinin can no longer bind to host-cell receptors and the virus is inactivated. Conversely, the spike glycoprotein of coronaviruses becomes fusion competent through cleavage by host proteases, instead of relying on acidic pH triggering conformational changes (29) . The different behavior of SARS-CoV-2 and HCoV-229E at pH < 3 remains unclear.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) ; buffer (circles) and nasal mucus (diamonds) correspond to a w ∼ 0.99. Each experimental condition was tested in replicate with error bars indicating 95% confidence intervals. While IAV displays a pronounced reduction in infectivity around pH 5, SARS-CoV-2 develops a similar reduction only close to pH 2, and HCoV-229E is largely pH-insensitive. Solid lines are arctan fits to SLF data with a w = 0.994 (blue: IAV; red: SARS-CoV-2; black: HCoV-229E). The dashed line is an arctan fit to the SLF data with a w = 0.80. The dotted line is an extrapolation to a w = 0.5 (24× SLF). Upward arrows indicate insignificant change in titer over the course of the experiment, and downward arrows indicate inactivation below the level of detection at all measured times. The fitted curves below pH 2 (grey shaded aera) are extrapolated with high uncertainty. Examples of measured inactivation curves are shown in Fig. S1 . The arctan fit equations, which are also used for the model simulations, are given by Eqns. S28, S29, and S30.

While Figure 1 shows the pH that must be attained in the aerosol particles for rapid virus inactivation, it lacks information on aerosol particle pH after exhalation into indoor air. To model the pH in these particles it is essential to know the particle composition in thermodynamic equilibrium (liquid water content), as well as the kinetics that determine how rapidly the equilibrium is approached (water and ion diffusion coefficients). To obtain this information, we measured thermodynamic (equilibrium) and kinetic (diffusion-controlled) properties 6 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 14, 2022. ;  of individual micrometer-sized SLF and nasal mucus particles levitated contact-free in an electrodynamic balance (EDB). Each particle was exposed to prescribed changes in RH (see Fig. 2 ). Figure 2A shows two moistening/drying cycles of an SLF particle obtained over a period of two days. They allow determination of the particle equilibrium composition (water content or mass fraction of solutes, see Fig. S2A ) during time intervals with slowly changing RH.

The particle clearly takes up and loses water when the RH is changed. It has a size growth factor at 90% RH of 1.3 (see also Fig. S3 ) and deliquesces at 75%, indicating that NaCl is the predominant salt in the particle. Nasal mucus shows a similar size growth, but deliquesces over an RH range of 55 to 70%, indicating that it contains significant amounts of other salts ( Fig. S3 ). We have no evidence for liquid-liquid phase separation in any of these particles ( Fig. S4A and S5) but Mie-Resonance spectra indicate inhomogenities in the particles even at high RH.

The kinetics of water uptake/loss as derived from periods with rapid RH change or efflorescence are highlighted in Fig. 2 . Figure 2B zooms in on one efflorescence event, first showing rapid water loss (< 10 s), then switching to a much slower rate of water loss over the next hour.

This two-stage diffusion process was confirmed in measurements of additional SLF and nasal mucus particles (see Fig. S6 ). We attribute the fast process to an initial dendritic growth of an NaCl crystal ( Fig. S4A -C), which ends abruptly when the crystal reaches the droplet surface, followed by a slow crystal growth mode (Fig. S4D) . Initially, crystal growth is limited by the liquid phase diffusivity of water molecules with D ,H 2 O > 10 −7 cm 2 /s (Fig. 2C) , which is expelled from the particle as long as water activity is still high. Subsequently, the slow crystal growth is limited by the diffusivities of Na + and Clions through the progressively viscous liquid to the crystal (Fig. S4D ). From Figs. 2B and S4D we estimate the ion diffusion coefficient to be about D * ,ions ≈ 10 −10 cm 2 /s, which determines the low rate of continued loss of water molecules. The diffusion coefficients determined in this way are "effective" (indicated by a star), as they represent the molecular diffusivities under the specific morphological conditions associated with the dendritic growth of the salt crystals inside the droplets (see next section for details on how these diffusion coefficients were further constrained).

In summary, independent of the exact thermodynamic equilibrium state of the particles, our results demonstrate that SLF as well as nasal mucus show a clear diffusion limitation for ions. In contrast, water diffusion in SLF and nasal mucus remains fast even when RH is low. This continuous, rapid diffusion of water indicates that SLF and nasal mucus do not form diffusion-inhibiting, semisolid phase states such as those recently reported by others in particlets containing model respiratory compounds (30) . The voltage required to balance the particle in the EDB against gravitational settling and aerodynamic forces is a measure of the particle's mass-to-charge ratio, allowing the particle radius R to be estimated. (A) Two humidification cycles of an SLF particle with a dry radius R 0 ∼ 9.7 µm. The experiment spanned about 2 days with slow humidity changes, allowing the thermodynamic and kinetic properties of SLF to be determined. Deliquescence/efflorescence points are marked by "Deliq/Effl". (B) Zoom on the drying phase (red box in (A)) with salts in the droplet (mainly NaCl) efflorescing around 56% RH (black line): very fast initial crystal growth (< 10 s) with rapid loss of H 2 O from the particle, followed by slow further crystal growth (1 h). The latter is caused by the abrupt switch from H 2 O diffusion to the diffusion of Na + and Clions through the viscous liquid, resulting in an ion diffusion coefficient of D * ,ions ≈ 10 −10 cm 2 /s. The insert (C) highlights the minute before and after efflorescence, which allows a lower bound of the H 2 O diffusivity (namely D ,H 2 O > 10 −7 cm 2 /s) to be determined.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 14, 2022. ; https://doi.org/10.1101/2022.03.14.22272134 doi: medRxiv preprint

Only the combination of the virological bulk phase data ( Fig. 1) with the microphysical aerosol thermodynamics (vapor pressures and activity coefficients) and kinetics (Figs. 2 and S7) allows the pH attained in the aerosol particles and the resulting rates of viral inactivation, to be determined. Thus, the virological and microphysical data were combined as input for a multi-layer Respiratory Aerosol Model (ResAM). ResAM is a biophysical model that simulates the composition and pH changes inside an expiratory particle during exhalation, and determines the impact of these changes on virus infectivity (see section "Biophysical modeling" in the Supplementary Material). The model performs calculations for particles of selectable size (from 20 nm to 1 mm) with a liquid phase composed of organic and inorganic species representative of human respiratory fluids S1 (more detail in the Supplementary Material). It takes account of diffusion in the gaseous and condensed phase, vapor pressures, heat transfer, deliquescence, efflorescence, species dissociation, and activity coefficients due to electrolytic ion interactions (see Tables S2, S3 ). Ultimately, ResAM computes the species distribution and their activity in the liquid, the resulting pH, and the corresponding virus inactivation rates as function of time and of the radial coordinate within the particle.

When RH changes are slow, the measured mass fraction of solutes in SLF as a function of RH allows the model thermodynamics to be constrained (Fig. S2B ). Under thermodynamic equilibrium conditions the model captures the mass fraction of solutes along the deliquesced and effloresced branches of the particle reasonably well. However, only after kinetic effects (ion and water diffusivities) are also taken into account does the model accurately reproduce the solute composition curve along the deliquesced branch. This demonstrates that even when RH changes are slow (raising RH from 50% to 70% in over one hour), kinetics cannot be neglected.

For rapidly evaporating expiratory particles, kinetics effects are even more critical. By matching the model to the fast changes during the efflorescence and deliquescence processes, ion diffusion coefficients can be derived for different water activities. Interpolation together 9 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. As an example, Fig. 3 shows the evolution of the physicochemical conditions within an expiratory particle with 1 µm initial radius during transition from nasal to typical indoor air conditions with 50% RH (Table S1) , and the concomitant inactivation of IAV and SARS-CoV-2 contained within the particle. The rapid loss of water leads to concentration of the organics and salts, to the point when NaCl effloresces. Nitric acid from the indoor air enters the particle readily, lowering its pH from an initial value of 6.6 (resulting from the high concentrations of CO 2 and NH 3 in the exhaled air) to pH 5 within ∼ 10 s. This, in turn, pulls NH 3 into the particle, partly compensating the acidification. The pH further decreases to ∼ 4 within 2 minutes, then slowly approaches pH 3.7 due to further uptake of HNO 3 from the room air. This result confirms the importance of trace gases in determining the pH of indoor aerosol particles (24) . If only CO 2 is considered, its volatilization from the particle would lead to an expected increase in pH after exhalation (31) . Owing to aerosol acidification, rapid influenza virus inactivation occurs at ∼ 2 minutes, whereas SARS-CoV-2 (and the even more pH-tolerant HCoV-229E) remain infectious.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . Evolution of physicochemical conditions within a respiratory particle leading to inactivation of trapped viruses during the transition from nasal to typical indoor air conditions, modeled with ResAM. The initial radius of the particle is 1 µm. Thermodynamic and kinetic properties are those of synthetic lung fluid (SLF, see Fig. 2 and Table S1 ). The indoor air conditions are set at 20°C and 50% RH (see Fig. S8 for the corresponding depiction of physicochemical conditions at 80% RH). The exhaled air is assumed to mix into the indoor air using a turbulent eddy diffusion coefficient of 50 cm 2 /s (32) . The temporal evolution of gas phase mixing ratios is shown in Fig. S20 . The gas phase compositions of exhaled and typical indoor air are given in Table S4 . Within 0.3 s, the particle shrinks to 0.7 µm due to rapid H 2 O loss, causing NaCl to effloresce (grey core). The particle then reaches 0.6 µm within 2 minutes due to further crystal growth, after which it slowly grows again due to coupled HNO 3 and NH 3 uptake and HCl loss. ResAM models the physicochemical changes in particles including (A) water activity, (B) molality of organics, (C) NO -3 (resulting from the deprotonation of HNO 3 ), (D) molality of total ammonium, (E) molality of Cl -, (F) pH, as well as inactivation of (G) IAV and (H) SARS-CoV-2 (decadal logarithm of virus titer C at time t relative to initial virus titer C 0 ). Inactivation times vary with particle size: larger droplets take longer to reach low pH than smaller ones as they are impeded by longer diffusion paths of the relevant molecules (mainly HNO 3 and NH 3 ) or ions through both the air and liquid phases. The black line in Fig. 4C illustrates this relationship for IAV, showing 99% inactivation after about 2 minutes in particles with radii < 1 µm, but longer than 5 days for millimeter-sized particles. As a rule of thumb, a 10-fold increase in particle size leads to roughly a 10-fold increase in IAV inactivation time under typical indoor conditions. Conversely, the black line in Fig. 4D for SARS-CoV-2 shows 11 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 14, 2022. ; https://doi.org/10.1101/2022.03.14.22272134 doi: medRxiv preprint that inactivation is inefficient for SARS-CoV-2, irrespective of particle size.

is slightly acidified, provided that the gaseous acid molecules meet two conditions: their volatility must be sufficiently low, such that they readily partition from the gas phase to the condensed phase, and, once dissolved, they must be sufficiently strong acids to overcome any pH buffering by the particle matrix. Figure 4 compares the aerosol pH in typical indoor air (panel A) with that in air enriched with 50 ppb HNO 3 (panel B). This concentration of HNO 3 is well below legal 8-h exposure thresholds (0.5 ppm (33) or 2 ppm (34)). Notably, 50 ppb HNO 3 reduces the time to reach an aerosol pH of 4 from minutes to seconds. More importantly, 50 ppb also allowed the pH value to drop below 2, which is required for efficient SARS-CoV-2 inactivation (Fig. 1 ). For comparison, enriching the air with the more volatile and weaker acetic acid at concentrations below exposure threshold values could not achieve this, see Fig. S9 . The dark blue lines in Fig. 4C -D show the resulting inactivation times for IAV and SARS-CoV-2 (and Fig. S10 for HCoV-229E) as a function of particle radius. Remarkably, inactivation times of SARS-CoV-2 diminished by 4-5 orders of magnitude compared to typical indoor air (black lines). For particles with radii < 1 µm, which constitutes the majority of expiratory particles (see panel E), inactivation is expected to occur within 30 s.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. While an enrichment of acidic gases in air leads to an acceleration of IAV and SARS-CoV-2

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 14, 2022. ;  https://doi.org/10.1101/2022.03.14.22272134 doi: medRxiv preprint inactivation, the depletion of these gases, for instance by air filtration, has the opposite effect.

It is well-known that concentrations of strong inorganic acids, such as HNO 3 , are lower indoors than outdoors by at least a factor 2, and in buildings with special air purification, such as museums and libraries, by factors 10-80 (24) . If air is purified to contain only a fraction of the initial trace gas concentrations (see Table S4 ), the aerosol pH increases compared to typical indoor air and intermittently reaches neutral or even slightly alkaline values (up to pH 8.4 in particles with 5 µm radius in air purified to 1%). As a result, air purification is expected to enhance virus persistence, especially for IAV, as indicated by the red curves in Fig. 4 . 

Given the high pH sensitivity of many viruses (18, 36, 37, 38) and the readiness of expiratory aerosol particles for acidification, we next investigated the extent to which the modification of indoor air composition could mitigate the risk of virus transmission. To this end, we consider a ventilated room with occupants who exhale aerosol containing infectious viruses. We further make the assumption that, given the low concentration of airborne viruses, the transmission risk is directly proportional to the infectious virus concentration, respectively inhalation dose.

We use the term "relative risk of transmission" to express how the risk changes from standard 14 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 14, 2022. ; https://doi.org/10.1101/2022.03.14.22272134 doi: medRxiv preprint conditions (here typical indoor air according to Table S4 ) compared to air slightly enriched by HNO 3 or air that has been purified.

For the ventilated room we assume steady-state conditions where the exhalation defines the source of virus, which is balanced by three sinks, namely air exchange through ventilation, aerosol deposition, and pH-moderated virus inactivation within the aerosol particles (see Supplementary Material). We describe the virus source by the mean size distributions of number emission rates of expiratory aerosol particles (Fig. 4E) and assume each particle with radius > 50 nm to carry one virus irrespective of size. We describe the virus sinks by expressing ventilation by Air Change per Hour (ACH, mixing ventilation), applying mean aerosol deposition rates (39), and computing the inactivation rates similar to Fig. 4C ,D. This allows the airborne viral load and, thus, the relative risk of transmission, to be calculated as displayed in Fig. 5 for IAV and SARS-CoV-2 (and Fig. S15 for HCoV-229E). Black bars show the results for typical indoor conditions, blue bars indicate an enrichment of HNO 3 to 10 or 50 ppb, and red bars indicate purification of air to 20% or 1% of trace gases (see Table S4 ).

The results are unambiguous: while adding 50 ppb HNO 3 only has a moderate impact on HCoV-229E (Figs. S10 and S15), it promises to diminish the relative risk of transmission of IAV by a factor of ∼ 20 and of SARS-CoV-2 by a factor of 800 in rooms with ACH 2. Using HNO 3 is a more effective measure than increasing ventilation from ACH2 to ACH10, which for SARS-CoV-2 leads to a mere dilution by a factor 5 and for IAV does not help at all as the IAV 99%-inactivation time in typical indoor air is already short (2 minutes).

The ResAM estimates for purified air with significant reduction of trace gases (red bars) are also striking. While even normal air conditioning systems with air filters can lead to a reduction in "sticky" molecules such as HNO 3 (40), acid removal is likely even more pronounced in museums, libraries or hospitals with activated carbon filters (24) . In such public buildings, the relative risk of IAV transmission can increase significantly compared to buildings supplied with unfiltered outside air.

Here we demonstrate that a significant reduction in transmission risk can be achieved by 15 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 14, 2022. ;  https://doi.org/10.1101/2022.03.14.22272134 doi: medRxiv preprint air enrichment with HNO 3 levels that correspond to less than 10% of the legal exposure thresholds (33,34). We therefore expect that the resulting acid exposure will not cause harmful effects on human health. Nevertheless, future studies should investigate the effects of acid accumulation in indoor air on the microbiome and immune response in the respiratory tract.

In addition, ResAM should be further refined to include a greater diversity of respiratory matrices. Aerosol particles emitted during different human activities (e.g., coughing, singing) differ in their production mechanisms and site of origin in the respiratory tract, and hence in their matrix composition. The two matrices considered in this work -SLF and nasal mucushave comparable thermodynamic and kinetic properties as well as a similar pH-dependence of viral inactivation. However, we cannot exclude that additional respiratory matrices found in expiratory aerosol plumes (e.g., saliva) exhibit divergent properties (30, 41) . Despite these current unknowns, targeted regulation of aerosol pH promises profound positive effects on virus transmission and disease mitigation strategies.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 14, 2022 Airborne viral load (cm ) Table S4 ) and ACH 2 (thin horizontal line). Typical indoor air is shown by black bars, filtered air with removal of trace gases to 20% and 1% by red bars, and air enriched with 10 or 50 ppb HNO 3 by blue bars. Thick grey horizontal lines indicate the viral load and relative transmission risk in the absence of any inactivation. Results for 2 and 5 ppb HNO 3 are shown in Fig. S17 . Results for HCoV-229E, along with analogous analyses for coughing and speaking/singing are shown in Fig. S15 .

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 14, 2022. ; https://doi.org/10.1101/2022.03.14.22272134 doi: medRxiv preprint

Global mortality associated with seasonal influenza epidemics: New burden estimates and predictors from the GLaMOR Project

Clarification of terminology: In physical chemistry, an "aerosol" is a system of colloidal particles dispersed in a fluid, such as air. An "aerosol particle" refers to one single condensed-phase element in such an ensemble

regardless of particle size. In contrast, in epidemiological or virological parlance "aerosol" or "aerosol particle" usually means a very small (d 1µm) airborne particle, whereas "droplet" is used as its larger counterpart (d 1µm)

Airborne transmission of respiratory viruses

Studies on the experimental epidemiology of respiratory infections

Certain aspects of the behavior of type A influenza virus as an air-borne cloud

Virus survival as a seasonal factor in influenza and poliomyelitis

Airborne micro-organisms: survival tests with four viruses

Influenza virus infectivity is retained in aerosols and droplets independent of relative humidity

The influence of simulated sunlight on the inactivation of influenza virus in aerosols

Airborne SARS-CoV-2 Is Rapidly Inactivated by Simulated Sunlight

The influence of temperature, humidity, and simulated sunlight on the infectivity of SARS-CoV-2 in aerosols

Aerosol and surface stability of SARS-CoV-2 as compared with SARS-CoV-1

Survival characteristics of airborne human coronavirus 229E

Mechanistic insights into the effect of humidity on airborne influenza virus survival, transmission and incidence

Humidity-dependent decay of viruses, but not bacteria, in aerosols and droplets follows disinfection kinetics. Environmental Science & Technology 54, on inactivation of SARS-CoV-2 and other enveloped viruses

High aerosol acidity despite declining atmospheric sulfate concentrations over the past 15 years

Stability of infectious influenza A viruses to treatment at low pH and heating

The acidity of atmospheric particles and clouds

Relationship between Humidity and Influenza A Viability in Droplets and Implications for Influenza's Seasonality

The SARS epidemic and its aftermath in China: a political perspective

Characterization of aerosol composition, aerosol acidity, and organic acid partitioning at an agriculturally intensive rural southeastern US site

Indoor and outdoor concentrations of inorganic acidic aerosols and gases

Indoor acids and bases

Exhaled breath condensate pH is a robust and reproducible assay of airway acidity

Influenza HA subtypes demonstrate divergent phenotypes for cleavage activation and pH of fusion: implications for host range and adaptation

Structure of influenza haemagglutinin at the pH of membrane fusion

Mechanisms of SARS-CoV-2 entry into cells

Evidence for a semisolid phase state of aerosols and droplets relevant to the airborne and surface survival of pathogens

Measuring stability of virus in aerosols under varying environmental conditions

Turbulent eddy diffusion models in exposure assessment -Determination of the eddy diffusion coefficient

Reverse genetics of SARS-related coronavirus using vaccinia virus-based recombination

Virological assessment of hospitalized patients with COVID-2019

Application of a biologically contained reporter system to study gain-of-function H5N1 influenza A viruses with pandemic potential

Compositional characterisation of human respiratory tract lining fluids for the design of disease specific simulants

Design and development of a biorelevant simulated human lung fluid

Antiviral activity of type I, II, and III interferons counterbalances ACE2 inducibility and restricts SARS-CoV-2

The double-ring electrodynamic balance for microparticle characterization

Morphological investigations of single levitated H 2 SO 4 /NH 3 /H 2 O aerosol particles during deliquescence/efflorescence experiments

Photochemical degradation of iron(iii) citrate/citric acid aerosol quantified with the combination of three complementary experimental techniques and a kinetic process model

A Thermodynamic Model of the System HCl-HNO 3 -H 2 SO 4 -H 2 O, Including Solubilities of HBr, from <200 to 328 K

Vapour pressures of

H 2 SO 4 /HNO 3 /HCl/HBr/H 2 O solutions to low stratospheric temperatures

Endobronchial pH: relevance to aminoglycoside activity in gram-negative bacillary pneumonia

Endoscopic sensing of alveolar pH

Effects of inhaled acids on airway mucus and its consequences for health

The buffer capacity of airway epithelial secretions

Thermodynamic model of the system H + -NH + 4 -SO 2− 4 -NO 3 -H 2 O at tropospheric temperatures

Ions' motion in water

Diffusion processes

Low mach number fluctuating hydrodynamics for electrolytes

Coupling Poisson-Nernst-Planck and density functional theory to calculate ion flux

Microstructure of Atmospheric Clouds and Precipitation

Relative humidity of expired air

Random Walks in Biology

Review of the vapour pressures of ice and supercooled water for atmospheric applications

Indoor air quality of 5,000 households and its determinants. Part A: Particulate matter (PM2.5 and PM10-2.5) concentrations in the Japan Environment and Children's Study

Atmospheric chemistry and physics: from air pollution to climate change

The history of vinegar and of its acetification systems

Effectiveness of common household cleaning agents in reducing the viability of human influenza A/H1N1

abstract: Vinegar and Its Active Component Acetic Acid Inhibit SARS-CoV-2 Infection In Vitro and Ex Vivo. bioRxiv : the preprint server for biology

Accurate representations of the microphysical processes occurring during the transport of exhaled aerosols and droplets

Liquid-liquid Phase separation in mixed organic/inorganic aerosol particles

A biocompatible synthetic lung fluid based on human respiratory tract lining fluid composition

Two-dimensional angular light-scattering in aqueous NaCl single aerosol particles during deliquescence and efflorescence

The ionization constant of water over wide ranges of temperature and density

Fate of ammonia in the atmosphere -A review for applicability to hazardous releases

Compilation of Henry's law constants (version 4.0) for water as solvent

The dissociation constant of acetic acid from 0 to 60 • centigrade

Geochemistry of Sedimentary Carbonates

Evaluation of the temperature dependence of dissociation constants for the marine carbon system using pH and certified reference materials

Solubility of NaCl, NaBr, and KCl in water, methanol, ethanol, and their mixed solvents

Corrosion Engineering -Principles and Practice

Exhaled breath concentrations of acetic acid vapour in gastroesophageal reflux disease

Breath concentration of acetic acid vapour is elevated in patients with cystic fibrosis

An introduction to airway assessment and management (concise airway anatomy and pathophysiology)

The authors thank Chuck Haas and Mutian Niu for valuable discussions.