Aerial transmission of SARS-CoV-2 virus (and pathogens in general) through environmental e-cigarette aerosol

2.1. Background

As opposed to previous studies of “airborne” pathogen transmission through various respiratory activities, vaping (as a respiratory activity) involves transmission of respiratory droplets through a different carrier fluid: the ECA gas and particulate phase diluted in expired air breath. It is thus essential to consider vaping styles and known physical and chemical properties of ECA (Section 3).

Subsection 3.1: Vaping behavior and demographics. To simplify the complexity of vaping styles and puffing topography we classify usage of e-cigarettes in two main categories: the majority (80-90 %) low intensity ‘mouth-to-lung’ (MTL) and the minority (10-20 %) high intensity ‘direct to lung’ (DTL) vaping.

Subsection 3.2: Exhaled ECA. It is a strongly diluted aerosol (mass retention of inhaled ECA by the respiratory system is about 90 %) whose particulate phase is made of rapidly evaporating submicron droplets composed of propylene glycol (PG), glycerol or vegetable glycerine (VG), nicotine and water. Exhaled ECA droplet numbers (about ∼ 10⁸ – 10⁹) are far higher than those of respiratory droplets potentially emitted by vapers.

Subsection 3.3: Exhaled ECA as a visual tracer of respiratory flows. Because of the fluid dynamical properties of ECA and optical properties of its droplets, the later act effectively as visible tracers of the exhaled respiratory flow, a feature not present in other respiratory activities.

2.2. Methods I: Inferences on respiratory droplets spread by ECA

Given the lack of experimental data on respiratory droplets carried by ECA, we need to infer their possible diameter distribution and rate of emission by looking at available evidence on expiratory activities that can serve as proxies for vaping. We proceed along the following steps in section 4

Subsection 4.1: Vaping as a respiratory process. Cigarette smoking is a useful proxy to estimate the breathing mechanics of vaping, a justified assumption since the vast majority of vapers are relatively recent ex-smokers and current smokers, following the MTL vaping style that resembles the dominant puffing topography of smoking.

Subsection 4.2: Mouthpieces, noseclips and suction. Vaping involves oral suction of ECA through mouthpieces, which together with noseclips are standard instruments in breathing experiments. The effects of these instruments, as well as the differences between oral and nasal breathing, are relevant to infer droplet emission in breathing and outcomes of breathing experiments.

Subsection 4.3: Diameter distribution and emission rate of respiratory droplets. Mouth breathing is the best available proxy to infer the characteristics of respiratory droplets carried by exhaled ECA, as their exhalation velocities are comparable. We examine available data on respiratory droplets from breathing experiments at different levels of inspiration, with subjects breathing in all cases through mouthpieces and noseclips (whose effects we examined in section 4.2).

2.3. Methods II: Assessing direct, indirect exposure and chemical interactions

Given the inferred characteristics of respiratory droplets potentially carried by exhaled ECA, we estimate the scope and risks for direct and indirect exposure in the following three sections:

Section 5: Hydrodynamical modeling of direct exposure. We model ECA flow as a puff with a starting turbulent jet with finite fluid injection (finite exhalation time) to estimate how far can exhaled ECA carry respiratory droplets.

Section 6: Risk model of indirect contagion. To asses this risk we consider a simplified adaptation to incorporate vaping in the dose-response exponential risk model of the spread of infective SARS-CoV-2 quanta developed by Buonanno, Morawska and Stabile [26] (BMS). Our aim is to estimate the relative risk for indoor vaping and vocalizing (intermittent expiratory activities), relative to a “control case” scenario consisting in rest breathing, an unavoidable continuous respiratory activity.

Section 7: ECA as a disinfectant. We review the literature on the laboratory tested aerial disinfectant properties of PG and VG, the main chemical compounds of ECA.

2.4. Summary of results

From the available evidence examined in section 4 we infer in section 8.1 the following plausible characteristics of respiratory droplets associated with vaping

• MTL vaping should produce exhalations of tidal volume of 700 – 900 cm³, emitting on average 2–230 droplets per puff overwhelmingly in the submicron range

• DTL vaping involves much larger exhaled volumes (1000 – 3000 cm³) and thus it should emit on average several hundreds and up to 1000 droplets also in the submicron range but with higher mean diameters.

The comparative reference for droplet characteristics of low intensity MTL vaping is respiratory experiments for tidal volumes close to rest breathing, while for DTL vaping it is experiments at more intense level of inspiration.

From the evaluation of the dynamical equations derived in section 5 we report in section 8.2 that droplets (ECA and respiratory) should be transported horizontally 0.5-meters for MTL vaping and over 2 meters for DTL vaping. Once the injection stops the jet evolves into an unstable puff that becomes rapidly disrupted by turbulent mixing with entrained air, with the submicron droplets (ECA and respiratory) transported by the linear momentum of the starting jet subsequently dispersing, being carried by air currents, thus representing a potential risk of indirect contagion.

From our adaptation of the risk model of BMS introduced in section 6, we find in section 8.3 the following added risks from indirect exposure with respect to the control case of continuous breathing (see also figure 7):

• 1 % for low the dominant intensity MTL vaping (160 daily puffs)

• 5-17 % for the minority high intensity DTL vaping (160 daily puffs)

• 44 % for continuous speaking 10 % of time (6 minutes every hour), up to 90 % for speaking 40 % of time (20 minutes)

• over 260 % for coughing 30 times per hour.

These are proportional risks with respect to the control case, hence they hold for both scenarios (home and restaurant) and ventilation types (natural or mechanical), though the absolute exposure values in terms of the count of infective quanta strongly depend on exposure time and choice of ventilation.

We show in section 8.4 that universal wearing of face masks involves negligible alterations of these relative risks. This protective gear significantly blocks quanta emissions from the baseline continuous breathing (pending on their filtration efficiency), but this effect is compensated by the protection that masks afford to exposed bystanders.

After reviewing the literature presented in section 7 on the disinfectant properties of gaseous PG and VG, we estimate in section 8.5 that the highly idealized conditions of conducted experiments and the high PG concentrations could hardly be applicable to the erratic and intermittent conditions of realistic vaping. Nevertheless, the disinfectant properties of PG on the SARS-CoV-2 virus should be tested.

2.5. Final discussion and conclusion

We discuss in section 9 the limitations of the study and provide an assessment of our exposure risk analysis within the context of safety measures to address direct and indirect aerial contagion in the current COVID19 pandemic.

3.1. Vaping styles and demographics

3.1.2. Demographics and markets

It is crucial to examine how representative among vapers are the different puff topographies and levels of intensity, something that has varied with time depending on the popularity and availability of different devices. Currently, low powered devices (mostly closed) are the most representative in the largest and most established markets. As shown in figure 1 (Credit to ECig Intelligence [32]) consumer surveys reveal that the overwhelming majority of vapers (80% in the USA) and 90% in the UK) utilize low powered devices (mostly kits for beginners and closed systems), with advanced open tank systems taking the rest.

• Download figure
• Open in new tab

Figure 1:

Usage of different classes of e-cigarette devices in the US and UK markets. Notice that only 15 % and 20 % of consumers in the USA and the UK use advanced kits that allow for the DTL vaping style (Credit to ECig Intelligence [32]).

The USA and the UK are the biggest and more developed markets, a fact that explains why the closed system category is more prevalent. In a natural evolution of markets the vape category takes off with a more hobbyist segment of users who are more likely to vape with DTL topography in high powered devices that yield large clouds. In nascent markets the ‘easy to use’ open system devices are not of great quality, though recent innovations are likely to improve this. Smokers in large markets are also likely to have higher disposable incomes and a more developed attitude of willing to (and being able to afford to) switch to a less harmful alternative. Such markets also have extensive distribution networks (convenience stores, tobacconists, etc). These factors influence the dominance of the market share of closed system devices and thus to characterize low intensity MTL style as the most prevalent among the vast majority of vapers. Prevalence of mouth puffing (puffing without lung inhalation) is marginal, as an overwhelming majority of vapers inhale to the lung for being ex-smokers or current cigarette smokers.

3.2. Inhaled and exhaled E-cigarette aerosol (ECA)

The ECA is generated by various physicochemical processes: self–nucleated condensation in a super saturated medium initiates immediately once the e-liquid vapor leaves the coil, the nucleated centers generate small nm scale droplets that grow through coagulation and diffusion [33]. The particulate phase is made of liquid droplets whose chemical composition closely matches that of the e-liquid: propylene glycol (PG), vegetable glycerin or glycerol (VG), nicotine, water [34], together with a negligible contribution of nanometer sized metal particles [35]. The gas phase is chemically similar. The aerosol contains nicotine and residues produced from the pyrolysis of the glycols and the flavorings (mainly carbonyls), which can be in either the gas or particulate phase depending on their vapor pressure and volatility [36], with most of the PG evaporating into the gas phase and VG tending to be remain in the droplets [34].

Count mean diameter (CMD) distributions of mainstream ECA droplets vary depending on the device, puffing style of users, flavors and nicotine content [33, 37]. Droplet number count is heavily dominated by submicron droplets with CMD distributions having either single modes below 100 nm or bimodal forms (one mode well below 100 nm and one in the range 100-300 nm) [33, 37, 38, 39, 40, 41]. However, particle size grows with increasing coil power [37] and even in low powered devices the mass distribution is dominated by droplets larger than 600 nm [33]. In fact, [33] found a third mode around 1 µm that becomes more prominent at increasing power of the tested device while the nm sized modes decrease, likely because higher power involves larger vaporized mass that favors coagulation and scavenging of nm sized droplets by larger droplets.

The inhaled aerosol mass yield depends on the topography parameters given in Table 1. At inhalation of mainstream ECA instrument measured droplet density numbers are in the range n = 1 – 5 × 10⁹/cm³ [37, 38, 39, 40, 41]. Total average droplet numbers of N_p = 7.6 × 10¹⁰ were reported in [42] for a tank system using e–liquids with high nicotine content in a 2 second machine puff regime § with V_b = 50 mL puff volume (N_p decreases 25 % with nicotine-free e-liquids). Using the same experimental design [41] reported an increase of up to 30% for 4 second machine puff regime. The estimation N_p ∼ 10¹⁰ – 10¹¹ is reasonable given a particle number concentration of ∼ 10⁹/cm³ and V_b = 20 – 100 mL of low intensity vaping, with N_p ∼ 10¹² for high intensity vaping with V_b = 500 mL.

Data on the gas/particle phase partition of the aerosol mass yield m_b is roughly: 50% Total Particulate Matter (TPM), 40% PG/VG gas phase, 7% water vapor, < 3% nicotine [43], roughly a similar gas/particulate phase partition to that of tobacco smoke [44]. As shown in [34] and [36] the presence of compounds in gas or PM form depends on their vapor pressure, with PG tending to be gaseous, VG in PM, for nicotine it depends on its PH, while some aldehydes (like formaldehyde) are most likely in the gas phase.

Values of particle numbers and densities for the exhaled ECA can be estimated by considering its retention by the respiratory system. Retention of ∼ 90% of total inhaled aerosol mass was reported in [45] for a wide variety of devices and e–liquids, with the following average compound specific retention percentages: 86% VG, 92% PG, 94% nicotine, while [46] reported 97% total aldehyde retention. This high retention percentages are consistent with the mass distribution of inhaled ECA dominated by larger micron sized droplets which tend to be efficiently deposited in the upper respiratory tracts [33]. Assuming equal retention rate for the particulate and gas phases, we take as total mass of exhaled aerosol and total numbers of exhaled ECA droplet to be 10 % of the values of m_b listed in Table 1 and 10 % of the values of N_p = 6.7 × 10¹⁰ reported in [42] for a 2 second machine inhalation puff and 50 mL puff volume. Droplet number density of ECA as it is exhaled can be estimated from these values of N_p bearing in mind that the exhaled ECA is now diluted in tidal volumes V_T listed in Table 1 for the various vaping topographies. This yields number densities in the approximate range n_p = 10⁶ – 10⁷ cm⁻³ (lower to higher vaping intensities).

Exhaled ECA dilutes and disperses very fast. Its chemical composition is similar to that of inhaled ECA, both in the gas phase and the droplets [33], with PG and water in the latter evaporating rapidly. Since hyperfine nm sized droplets deposit efficiently by diffusion in the alveolar region and larger micron sized droplets (which tend to grow from hygroscopic coagulation [31, 33]) deposit by impaction in the upper respiratory tracts [31, 42, 37, 39], the CMD distribution of ECA as it is exhaled should be dominated by modes in intermediate ranges 0.1 – 0.5 µm. Since there are no ECA measurements at the exhalation point (the vaper’s mouth), we can estimate the representative droplet diameter by a rough order of magnitude calculation: assuming an aerosol mass yield of 5 mg of inhaled ECA for a low powered device, a retention of aerosol mass with 50 % made of PM, the total droplet mass of exhaled ECA should be around M_p = 0.25 mg. Since 90 % of droplets are retained, the total number of exhaled droplets should be N_p = 7.6 × 10⁹ droplets [42], leading to a median droplet mass of , where ρ_p is the droplets density that we can assume to be close to VG density: ρ_p = 1.3 gm/cm³, leading to d_p = 0.38 µm. Similar order of magnitud values are obtained for the parameters of high intensity vaping.

The fact that CMD chamber measurements are in the range d_p = 0.1 – 0.2 µm can be explained by the fact that detectors are located 1-2 meters from the exhalation source, thus measured ECA droplets have already undergone significant degree of dilution and evaporation (as shown in [34] droplets’ mass can decrease by one third in just 1 second by evaporation of its PG content). This is consistent with droplet number densities dropping at least two orders of magnitud from ∼ 10⁶ – 10⁷ cm⁻³ as they are exhaled to n ∼ 10⁴ – 10⁵ cm⁻³ at one meter distance from the emission and further dropping to near background levels n ∼ 10³ cm⁻³ at two meters [47, 48, 49].

3.3. Exhaled ECA as a visual tracer of respiratory fluid flow

As opposed to other respiratory expirations (breathing, vocalizing, coughing, sneezing), the actual respiratory flow of ECA expiration can be directly visualized. This is a consequence of the optical properties of its particulate phase [19, 50]. Moreover, when visualizing exhaled ECA the viewer is practically seeing nearly the same respiratory air flow that would result from the same respiratory mechanics without involving ECA (i.e. “mock vaping”). This is so because of its physical properties: exhaled ECA is a “single–phase fluid flow” (SFF) system [51, 52] in which the particulate phase (made of overwhelmingly submicron ECA droplets) have negligible influence on the fluid dynamics, acting essentially as visible tracers or (to a good approximation) as molecular contaminants carried by the fluid.

The fact that exhaled ECA can be used as an effective tracer of respiratory flows is not surprising nor unique, given the existence of numerous gas markers and aerosols in a SFF regime that are widely used as proxies for the study and visualization of expired air [53, 54]. This also applies to mainstream exhaled tobacco smoke, whose particulate matter is also made of submicron liquid and solid droplets. In fact, there are studies that have directly used cigarette smoke as a tracer to visualize respiratory airflows [55, 56, 57]. It is worth mentioning that respiratory droplets potentially carried by exhaled ECA would not change its possible role as a tracer of expiratory flows, since as we show further ahead (section 4.3) these droplets are also overwhelmingly in the submicron range and their numbers are much fewer than ECA droplets.

The particulate phase of the ECA at exhalation consists of submicron droplets made of PG, VG, water and nicotine, whose density should be close to VG density ρ_p ∼ 1.3 gm/cm³. Dilution and PG evaporation further reduce droplets diameters and number densities. These droplets are evolving along a carrier fluid made of a gas mixture: the gas phase of the ECA strongly diluted in exhaled air (in practice, we can think of the carrier fluid simply as exhaled air at mouth temperature ∼ 30 – 35° C). As in all SFF systems, In what follows we examine two criteria that determine the SFF character of the exhaled ECA: the particles’ relaxation time and volume fraction.

Submicron particles in a carrier fluid have little inertia, and thus essentially follow the fluid flow. They are well within the Stokes regime with Reynolds numbers Re_p ≪ 1 and negligibly small relaxation times t_rel, the response time of an aerosol particle to adjust to external forces. For d_p = 0.3 µm we get [50] where µ = 1.895 × 10⁻⁵gm/(sec cm) the dynamic viscosity of air at 35 C and C_c = 1 + (λ/d_p)[2.34 + 1.05 exp(−0.39d_p/λ)] ≈ 1.4 is the Cunningham slip factor with λ = 0.066 µm the mean molecular free path of air. The relaxation time provides the time scale for a particle released into a fluid with velocity U along a horizontal stream to settle into the fluid velocity (neglecting gravity). In this case (see Chapter 3 of [50]) the velocity of the particle becomes practically identical to U in about 10⁻⁵ seconds (instantaneously in practical terms), thus justifying the notion of particles simply following the fluid flow with (practically) no influence on its dynamics. This behavior occurs also for the larger ECA droplets of d_p ∼ 1 µm whose relaxation times are . Evidently, these relaxation times are much smaller than macroscopic characteristic times of the carrier fluid (for example a 2 second inhalation time or even the tenths of a second the ECA stays in the mouth cavity [31]). The Stokes number is defined as St = t_rel/t_f, where t_f is a characteristic fluid time, hence for the exhaled ECA we have St ≪ 1, which is another criterion to define SFF systems.

Another criterion for an aerosol to be described as SFF systems is the ratio ϕ of total volume of the particles to the fluid volume satisfying [51, 52] where N_p is the total number of particles, is the particles’ volume (assuming they are spherical) and V_f is the fluid volume. Substituting the qualitative values we obtained for the exhaled ECA: N_p = 7.6 ×10⁹, d_p = 0.3 µm and an exhaled air volume of V_f = V_T = 300 – 1500 mL for a low intensity regime yields ϕ = 1 – 3 × 10⁻⁷, which fulfills (2). This condition holds even if we assume that a large part of the mass distribution is contained in micron sized (d_p ∼ 1 µm) droplets making (say) 10 % of the total number. The value of ϕ is bound to decrease as the exhaled ECA dilutes and the volatile droplet compounds (PG and water) evaporate. As shown in [34] this process decreases the droplets mass (and thus volume) to one third while the fluid volume increases and thus ϕ necessarily decreases.

Evidently, larger droplets (diameters larger than a few µm) are present in ECA particle diameter distributions and such particles should contain a significant portion of the aerosol mass [33], but they are too few in numbers and deviate from the flow following ballistic trajectories, thus do not affect the dynamics of the carrier fluid to consider ECA as a biphasic fluid flow system.

4.1. Vaping as a respiratory process

Since most vapers are either cigarette smokers or ex–smokers of cigarettes, it makes sense to examine the respiratory parameters of smoking as reported in the literature (see reviews in [58, 59], see also Table 3). Tobacco smoke is a valid reference for ECA, as it is also an aerosol in a SFF regime characterized by a particulate phase made of predominantly submicron particles with similar particle numbers and diameter distributions [33, 39, 60] (though the particulate and the gas phases of each aerosol have very different chemical properties).

4.2. Mouthpieces, noseclips and the breathing route

Mouthpieces (MP) and nose-clips (NC) (to block nasal inspiration) are standard instruments in observational studies, not only those aimed at studying droplet emission, but of respiratory patterns and flows in human subjects. Since the results of these studies can serve as appropriate proxy values to infer droplet emission in vaping, it is important to assess the effects of these instruments in respiratory mechanics. For the purpose of the present article, this issue is interesting because ECA is inhaled in e–cigarettes through mouthpieces (though without obstruction of nasal breathing).

4.3. Diameter distribution and emission rate of respiratory droplets

The discussion in the previous sections has allowed us to infer the characteristics and parameters of the respiratory mechanics of vaping. We need now to identify among respiratory processes the ones that most closely fit these parameters in order to use their available experimental data to infer the capacity of vaping for respiratory droplets emission.

5. Hydrodynamical modeling of direct exposure

In the previous sections we have inferred the submicron characteristics and rate of emission of respiratory droplets expected to be carried by exhaled ECA. We need to estimate now how far can these respiratory droplets be carried to evaluate the distance for direct exposure of bystanders to pathogens potentially carried by these droplets Exhaled ECA is injected into surrounding air a given horizontal distance roughly in the direction of the exhaled flow. Since it involes a finite fluid mass of a SFF aerosol during a finite injection time (exhalation time), the appropriate dynamical model for it is a turbulent puff with a starting momentum dominated jet that lasts while the fluid injection is on [104, 105, 106, 107, 108, 109, 110, 111]. A schematic description of this system is furnished by Figure 3. We will not be concerned with the few larger particles (diameters d ∼ 1 – 5µm and over) that initially follow the fluid stream but (depending on their size) exit the main flow to follow ballistic trajectories until they either deposit on surfaces, settle on the ground or evaporate [87, 112].

Given the distance and time dispersion scales (< 3 meters and < 2-3 minutes) we can approximate the ECA as an airflow at constant atmospheric pressure, air density and dynamical viscosity ρ_a and µ. For a jet source (vaper’s mouth) approximated as an orifice of 1.5 – 3 cm² area [56] (diameter d₀ =1.25-1.75 cm) and initial velocities U₀ given by (5), exhalation Reynolds numbers Re = (ρ/µ)U₀d₀ = 600 – 4400 are in the transition between laminar and turbulent, values well below the high Reynolds numbers expected near a jet source [104, 105], but we are mostly concerned with the jet evolution and displacement (penetration) along horizontal distances z ≫ d₀. Other parameters to consider are the injection time t_exh = 2 – 5 seconds and a temperature gradient from exhalation (initial) T = 30° −35° C (mouth temperature) into an assumed T = 20° C for the surrounding air. For such values and scales the starting jet can be regarded as isothermal with thermal buoyancy becoming relevant only in the puff stage [108, 109]. It is well known that steady and unsteady jet/puff systems can be well approximated by analytic models that assume axial symmetry and a self similar profile for the average centerline and radial components of the velocity field in cylindrical coordinates [104, 105, 113, 114] (see figure 3) where f, g are empiric Gaussian or polynomial functions of the self similar variable η = r/z and the centerline velocity is U_c = U_z for r = 0 along the z axis, hence f (η), g(η) must satisfy U_z = U_c and U_r = 0 at r = 0 (see examples in [87, 104, 105, 106, 107, 108, 109, 110, 111]). An axially symmetric self similar jet/puff system fulfills the conservation of linear specific momentum 𝒬 = V U_c (puff) and force (jet) where V is the penetration volume [104, 110, 111], hence for an initial time t = t₀. The stream wise centerline penetration distance and velocity for the jet and puff stages can be given by [110, 111]: where the constants C_jz, C_jr, C_pz, C_pr are empirically determined, and z_j0 is the z coordinate value of the ejection orifice and z_pd is the virtual origins of the puff (see Figure 2), which is an appropriate parameter to separate the starting jet and puff stages though it lies within the starting jet region (see detailed explanation in [110]). For the axial geometry of the jet/puff system under consideration we have and Following [110, 111], we will choose the following numerical values for the constants in (7)–(10): where the time t_pd and position of the puff virtual origin z_pd is determined numerically from (7) by the condition z_j(t_pd) = z_pd [110, 111]. Many vapers exhale at a downward angle typically γ ∼ 30 degrees, thus reducing the horizontal penetration of the starting jet given by (7) roughly as z_j cos γ. From the dynamical equations (7)–(10) with the parameter values in (11)–(12) we display in section 8.2 the horizontal distance and jet/puff velocities characterizing direct exposure.

• Download figure
• Open in new tab

Figure 2:

Puff and initial Jet with axial symmetry. The stating jet is propelled by linear momentum parallel to the centerline velocity U_c, the arrows above and below represent the entrainment velocity mixing surrounding air with the carrier fluid. As the fluid injection terminates (end of exhalation), the entrained air makes about 40 % of the fluid mass making the the transition into a ellipsoidal puff through highly turbulent vortex structures. At this point the puff is likely to disperse rapidly as horizontal displacement velocities are comparable to velocity fluctuations characterizing high turbulence and thermal buoyancy.

6.1. Infective quanta

To evaluate indirect exposure risks from vaping we simplify and adapt the analytic risk model of Buonanno, Morawska and Stabile (hereafter BMS) [26] who have examined the potential SARS-CoV-2 virus transmission in various indoor micro-environments (see also their previous paper [116]). BMS develop this model by means of Montecarlo simulations in which variability of droplet emission rates and exposure parameters is described by suitable probability distributions. Our approach is to assume median values for these variables (50 percentiles) of these distributions, similar to their approach in their previous paper [116]. This is justified because our aim is to evaluate the risks from indoor COVID-19 transmission from vaping, speaking and coughing (all episodic or intermittent expirations) in comparison with what can be denoted as a “control case” scenario of risks in a space were the infectious vaper is only rest breathing (a continuous expiration). We are not aiming at providing a full comprehensive risk analysis for each respiratory activity separately under more realistic conditions (something that would justify a full separate study in itself).

BSM consider the notion of an infective “quantum”: the dose of airborne respiratory droplet nuclei necessary to infect 63 % of exposed susceptible individuals. They introduce the “quantum emission rate” ER_q (emitted quanta per hour) for various respiratory expirations where c_v is the viral load (RNA copies/mL) in the sputum of a SARS-CoV-2 infected person (symptomatic or not), 𝒞_RNA is the number of RNA copies per PFU (plaque forming unit) needed to generate infection and 𝒞_PFU is quanta-to-PFU conversion parameter, f_br is the number of breaths per hour and V_T the tidal exhaled volume, 𝒞_d is the droplet volume concentration (in mL/m³, hence 𝒞_d V_T is the total volume of exhaled droplets in mL). BMS define the product “IR = V_T × f_br” as an “inhalation rate”, but it is really an exhalation rate expressible in units m³/h.

For the infection parameters BMS consider values that have emerged from recent data: c_v = 10⁷ RNA copies/mL (average in the range 10³ – 10¹¹), c_RNA = 1.3 × 10² RNA copies/PFU and c_PFU = 2.1 × 10² PFU/quanta. For the droplet volume concentration they take as reference an experimental value that incorporated dehydration effects in droplets associated with loud speech [117], then using experimental data from Morawska et al [94] to scale this reference to other respiratory expirations, leading to the following values (in mL/m³)

In order to fit vaping expirations into these values we need to make some assumptions on the involved parameters, besides considering the effects on exposure from the time duration of expiratory activities. In particular, we need to evaluate their mean quanta emission rate only in the times when they occur and compare with the rates of normal rest breathing (which takes place all the time). To simplify matters, we assume that c_v, c_I and f_(br) are largely unaffected by the timing of these expiratory activities. We have then

• Low intensity MTL Vaping. A vaper breathes N_(tot) times in (say) one hour and of these breaths N_(vp) coincide with vaping expirations (puffs), the expression for ER_q in (13) must be modified as where N_(vp) N_(tot) are the number of vaping puffs and total number of breaths per hour, V_{T (br)} V_{T (vp)} and 𝒞_d(vp), 𝒞_i(br) are the tidal volumes and droplet volume concentration for vaping and rest breathing. For low intensity MTL vaping we assume a tidal volume of V_T = 750 cm³ supported by inference from data discussed in previous sections, while for droplet volume concentration we assume 𝒞_d = 3 × 10⁻³ mL/m³, a plausible value denoting emissions slightly above rest breathing but below normal speech in (14), fitting the ‘whispered counting’ data of [94]. For the number of breaths we can take the average values of 160 daily puffs in a 16 hour journey [27, 72] and breathing frequency of f_(br) = 16/min (in the range 12-20), so that N_(tot) = 960 breaths/h and N_(vp) = 10 breaths/h.

• High intensity DTL vaping. We assume V_T = 2000 cm³ as an average tidal volume. However, there is ambiguity in inferring a value for droplet volume concentration because of insufficient data on how much the larger tidal volume and deeper inhalation of DTL vaping can modify respiratory droplet numbers and diameters. As mentioned in section 3.2, higher powered devices associated with DTL vaping tend to increase ECA droplet sizes and diameters [37, 33] but it is not certain if this applies to respiratory droplets. However, as mentioned in section 4.3, speech involves droplet generating mechanisms that are distinct from those of breathing [92, 94, 101], resulting in higher rate of droplet emission even with a tidal volume only slightly larger than the breathing rest value of 400 – 600 cm³ [118, 119]. Thus, we have two plausible options to account for a higher total volume of exhaled droplets V_d = V_T 𝒞_d: it may follow simply from a larger V_T with the same value 𝒞_d = 3 × 10⁻³ mL/m³ of low intensity vaping, or we might assume the larger value of 𝒞_d for normal speech in (14). Instead of choosing one option, we will keep the continuous range of 𝒞_d = 3 – 6 × 10⁻³ mL/m³. Regarding the number of breaths we can assume the same values as low intensity vaping: N_(tot) = 960 breaths/h and N_(vp) = 10 breaths/h.

• Normal speech. The equation for ER_q in (13) needs to be modified in a similar way as (15), replacing the droplet volume concentration 𝒞_d with the value for normal speech in (14) and we take as tidal volume the value V_T = 600 cm³, roughly 10 % larger than the average rest value [118, 119]. To incorporate the timing we replace N_(vp) with a number count of breaths coinciding with a given percentage of an hour interval spent on continuously speaking in a given indoor environmant. For 5, 10, 20, 30, 40 % of the hour (960 total breaths) we have N_(sp) = 48, 96, 192, 288, 384 breaths/h.

• Coughing. The emission data from coughing in [94] is comparable to that of ‘unmodulated vocalization’ (repeating the vowel “aahh”). Hence, we can use (15) with the value for droplet concentration volume of loud speaking in (14) as a proxy for coughing, while for coughing tidal volume we have V_T = 1400 cm³ [55]. Assuming a cough every 2 and 3 minutes, N_(vp) is replaced by N_(cf) = 20, 30.

Considering the plausible assumptions stated above, we display in figure 3 the logarithmic plots of quanta emission rate ER_q from an infectious individual as a function of viral load c_v, for rest breathing, low and high intensity vaping, speaking for 10 %, 20 %, 30 % and 100 % of the time, as well as coughing every 2 and 3 minutes. The numerical values of ER_q in quanta per hour for c_v = 10⁷ RNA copies/mL are where the symbols br, vpL, vpH, sp10, sp20, sp30, sp40, sp100 and cf respectively denote breathing, vaping low and high intensity, speaking 10, 20, 30, 40, 100% of the hour and coughing 30 times. Notice that for low and high intensity vaping ER_q is very close to the control case of rest breathing (almost indistinguishable for low intensity vaping), while even speaking 10 % of the hour (6 minutes) yields a larger ER_q value than the upper end of high intensity vaping. Also, normal speech for a full hour (not uncommon) produces a higher quanta emission than coughing 30 times

• Download figure
• Open in new tab

Figure 3: Quanta emission rates.

The curves display ER_q (quanta/hour) as a function of viral load c_v (RNA copies/mL) for various expiratory activities: rest breathing (br), low and high intensity vaping (vp), speaking (bottom to top) 10, 20, 30, 40 % of the hour (sp), coughing (cf) and speaking 100 % of the time (sp100). Numerical values of ER_q for c_v = 10⁷ RNA copies/mL (vertical line) are listed and discussed in the text. The ratios between these activities and rest breathing (taken as the case control scenario) is displayed in figure 7.

6.2. Exponential dose-response risk model

In order to evaluate a time dependent risk for expiratory activities that incorporates quanta emission rates and indoor environment variables, BSM consider the “dose response exponential model” given in terms of the the density of the quanta n(t) in units quanta/m³ under the assumption that n(0) = 0 (no exposure at initial time t = 0) where V is the volume (m³) of the indoor micro-environment, N is the number of exposed susceptible individuals, IR is the inhalation rate (m³/h) of these individuals and IVVR is the infectious virus removal rate, which which BMS take as the sum of three factors: IVVR = AER + κ + λ, where AER is the ventilation air exchange rate, κ is the particle deposition on surfaces and λ is the virus inactivation (all of these quantities given as h⁻¹).

We evaluate in section 8.3 the risk R for vaping exhalations and other respiratory activities, aiming at the evaluation of their relative risk with respect to the control state of continuous breathing, assuming a home and restaurant scenarios with natural and mechanical ventilation.

8.1. Respiratory droplets emission

From the evidence and data examined in section 4 it is plausible to assume that droplet emission in smoking and vaping (at least MTL style) can be reasonably inferred from outcomes of studies in Table 3 with comparable exhaled tidal volumes (see Table 2), including outcomes of studies mentioned previously that examined breath holds.

The studies we have summarized and listed in Table 3 only involve mouth breathing, but share some common respiratory features with vaping and smoking: oral inspiration with usage of MP’s (in vaping), as well as qualitatively similar exhalation velocities and respiratory parameters: inhalation/exhalation times and tidal volumes. However, there are also differences: smoking and vaping do not involve the nose occlusion of these experiments, but involve suction which the subjects of the latter experiments did not experience. While absence of NC’s would imply a tidal volume very close to rest values in MTL smoking and vaping, this absence is compensated by the increase due to the need to overcome airflow resistance through suction. The decrease of droplet emission from the mouth/oropharynx hold in MTL topography (absent in normal breathing) was a detected outcome in two of the studies listed in Table 3. We have then the following inferences regarding emission of respiratory droplets

• MTL vaping and smoking (and even DTL vaping not involving deep inspiration). The outcomes displayed in Tables 2 and 3 suggest that exhaled droplets in average tidal volumes V_T = 700 – 900 cm³, overwhelmingly in the submicron range (typically peaking at d_p = 0.3 – 0.8 µm), as well as a small rate of droplet emission: roughly N_p = 2 – 230 per exhalation (per litter), with droplet number densities well below n_p = 1 cm⁻³, though the wide individual variation reported in these studies should also apply to vaping, including the existence of a small minority of individuals that can be thought of as “super emitters” reaching over N_p ∼ 1000 per exhalation.

• DTL vaping. It involves a spectrum of deeper respiratory intensity than MTL vaping and thus should involve a higher rate of droplet emission. Perhaps the closest analogue in the studies listed in Table 3 to infer droplet emission for intense DTL (2-3 LT exhalation) breathing at fractional residual capacity in [95] that reported emission rates of around 1000/LT. However, this style of vaping is practiced by a small non-representative minority of vapers (roughly 10-20 %, see figure 1), while extreme vaping with big clouds (the so called “cloud chasers”) is even less frequently practiced in competitions or exhibitions. Evidently, this type of vaping cannot be sustained for long periods.

While the inferred droplet numbers in the upper end of high intensity DTL vaping can be comparable with low end numbers for vocalizing, the latter involves modes with larger mean diameters because of distinct droplet generation processes [92, 94, 101].

8.2. Distance for direct exposure

Direct exposure to respiratory droplets carried by exhaled ECA can be inferred from the horizontal displacement or penetration distance of the jet/puff system whose dynamics follows from equations (7)–(10) with the parameter values in (11)–(12). We display in figure 4 displacement distances and centerline velocities for assorted values of initial exhalation velocities U₀ corresponding to the vaping intensities we have considered.

• Download figure
• Open in new tab

Figure 4:

Jet/Puff horizontal displacement and centerline velocity. Panel (a) displays the displacement z_jc of the staring jet (blue) and z_pc of the puff (red) as functions of time from (equations (7)–(10)), for the three vaping topographies described in section 3.1: DTL (Direct to Lung), MTL (Mouth to Lung) and MP (Mouth Puffing). We assumed as injection (exhalation times) 3, 4 and 5 seconds. The initial velocities from top to bottom are U₀ = 50, 100, 150, 200, 250, 300, 400 cm/s. Panel (b) depicts centerline velocities U_c for the starting jet (equation (8)), as functions of the horizontal displacement z_jc during the injection times and initial velocities of panel (a) (green for MP, blue for MTL and red for DTL). Notice that once injection stops the jet has reached velocities comparable to those of indoor air currents.

Notice that the maximal penetration goes beyond that afforded by the momentum trust of the starting jet, with the puff further evolving at lesser speeds. Horizontal penetration varies from 0.5 meters for Mouth Puffing (U₀ = 0.5 m/s) through the range between 0.6 and 2.0 meters the MTL regime (U₀ = 0.3 – 3 m/s) and beyond 2 meters for the higher intensity DTL regime (U₀ = 1.5 – 4 m/s). Centerline velocity drops to about 0.2 m/s at different times and distances when fluid injection stops in all cases.

Given its short time duration and close distance scope of the momentum trusted staring jet, the analytic model (7)–(8) remains a reasonably good approximation to infer the necessary distance to minimize the risk of direct exposure of bystanders to respiratory droplets. As the jet evolves while fluid is injected there is increasing entrainment from the surrounding air at velocity U_e ∝ U_r, with entrained air reaching about 40 % of the jet mass at the end of injection in the transition towards the puff (around its virtual origin) [108, 109]. Since there are airflow currents of ∼ 10 cm/s (and up to 25 cm/s) even in still air in home environments with natural ventilation [135, 136], at this stage the puff formation can be easily destabilized by vortex motion generated through turbulent mixing from the large velocity fluctuations produced by the entrainment [137, 115].

Turbulence and thermal buoyancy become important factors when there is human motion or walking [138], or in micro-environments with mechanical ventilation (mixed or displaced) [139, 140, 141], resulting in a faster disruption and dispersion of the slow moving puff, carrying the submicron ECA and respiratory droplets along the air flow. In general, submicron droplets exhaled at the velocities under consideration can remain buoyant for several hours, with mixing ventilation tending to uniformly spread them, whereas directed ventilation tends to stratify them along different temperature layers. In all cases there is a risk of indirect contagion by exposure to these droplets. The detailed description of droplet dispersion after the puff is disrupted is a complicated process that requires computational techniques that are beyond the scope of this paper (see comprehensive analysis in [115]).

8.3. Risks of indirect exposure

To evaluate the risk of indirect exposure to respiratory droplets carried by vaping exhalations in indoor environments we have adapted in section 6 the “dose response exponential model” developed by Bounnano, Morawska and Stabile (BMS) [26]. Specifically, we evaluate equation (17) that defines the risk R (as a fraction < 1) for the value IR = 0.96m³/h taken from the previous paper of BMS [116] and justified as a level of physical activity half way between standing and light activity. For the remaining parameters BSM assume the range AER = 0.2 – 0.5/h for natural ventilation and AER = 9.6/h for a restaurant scenario with mixed ventilation. BMS compute the deposition rate by dividing typical gravitational settling velocity for supermicron particles (10⁻⁴ m/s) by the height of emission (1.5 m), leading to κ = 0.24/h, while for the viral inactivation they take the measured aerosolized SARS-CoV-2 virus mean life of 1.1 hours [142] and even longer periods [143], leading to λ = 0.63/h. We consider the following home and restaurant indoor scenarios:

• Home scenario. We assume one infectious vaper and three exposed susceptible family members (N = 3). Total exposure time T = 12 h. Indoor volume 125 m³ (small 50 m² apartment with roof height of 2.5 m). For natural ventilation: AER = 0.2/h we have IVVR = 1.07/h.

• Restaurant, natural ventilation with open door. Thirty costumers (N = 30), total exposure time T = 3 h. Air exchange rate AER = 0.5/h, indoor volume 300 m³ (100 m² area with roof height of 3 m), results in IVVR = 1.37/h

• Same restaurant endowed with mechanical ventilation: AER = 9.6/h (taken from [116]), results in IVVR = 10.47/h

The infection risk R for home and restaurant scenarios is plotted in figures 5 and 6 as a function of time for breathing, low and high intensity vaping, various percentages of time spent speaking and coughing every 2 minutes, considering natural and mechanical ventilation. As expected from the quanta emission rates displayed in figure 3, the exposure time of different expirations is a crucial factor in computing R. Exposure to vaping expiration (vaper doing 10 puffs per hour) poses an infection risk to bystanders that is very close to that from the control case scenario: exclusive normal rest breathing (for low intensity vaping the infection risk is practically indistinguishable). The infection risk from a person vaping is well below that from the same person speaking and coughing: speaking only for 10 % of the time (6 minutes per hour) already yields a higher infection risk than high intensity vaping, while speaking 30 – 40 % yields up tp 4 times the infection risk, which is roughly the values plotted in figure 7.

• Download figure
• Open in new tab

Figure 5: Infection risk in a home environment.

The curves display R as a function of exposure time T from (17). The abbreviations br, vpL, vpH, sp10, sp20, sp30, sp40 and cf stand for rest breathing, vaping low intensity, vaping high intensity (upper end option), speaking for 10, 20, 30, 40, % of time and coughing. Notice the dramatic reduction of R achieved by mechanical ventilation (air exchange rate of 3/h). Also: the curves for the risks from vaping (all intensities) are practically indistinguishable from that of the case control scenario of rest breathing (red circles).

• Download figure
• Open in new tab

Figure 6: Infection risk in a restaurant.

The same abbreviations as in figure 5 plus sp100 (speaking 100 % of the time, a possible outcome when spending 3 hours in a restaurant). As in figure 5, mechanical ventilation (air exchange rate 9.6/h) achieves a dramatic reduction of R and the curves for the risks from vaping are practically indistinguishable from the curve of the control case scenario of rest breating (red circles).

• Download figure
• Open in new tab

Figure 7: Added percentage risks of expiratory activities with respect to the control case scenario of rest breathing.

The percentage values with respect to the control case are: low intensity vaping 1.3 % (vpL), high intensity vaping 5.2-17.7 % (vpH), speaking 44% (sp10), 88% (sp20), 132% (sp30), 176% (sp40) for 10%, 20%, 30%, 40% of time, coughing 259% 30 times per hour (cf). These values were obtained from (ε − 1) × 100 for ε defined for these expiratory activities by (19)–(21).

A good inference of the risk from intermittent and episodic expiratory activities (vaping, speaking, coughing) relative to the control case scenario of exclusive rest breathing (a continuous expiration) is furnished by the ratio R_(A)/R_(br), where A = vp, sp, cf (see (16)). Plotting this ratio from (17)–(18) for every expiratory activity yields near constant curves around the values of the quotients ER_q(A)/ER_q(br) (see numerical values in (16)). This is not surprising since ER_q is the only variable in R that characterizes the infectious person (the other variables characterize the indoor micro-environment and the exposed susceptible persons). Hence, given the same indoor micro-environment and same number of susceptible individuals, we consider risks relative to the control case scenario of rest breathing in terms of the ratio of quanta emission. Using (15) we have where is the total exhaled droplet volume (in mL) for each expiratory activity. Since N_(br) = N_(tot), then for a heavy breathing activity in intense aerobic exercise ε might grow only because of the much larger tidal volume. However, for a truly intermittent expiration like vaping we have N_(vp)/N_(br) ≪ 1 and thus ε ≈ 1 holds even if we have (large exhaled amount of droplets as with the large tidal volumes in extremely intense vaping). For the values of tidal volume and droplet volume concentration we have used the numerical values in (16), we have the following relative risks which provides an intuitive indication of the added exposure risks relative to the control case from the different expiratory activities.

We display in figure 7 the numerical values of ε, as an added risk with respect to the control case for various expiratory activities with respect to the continuous presence of risk from rest breathing and under the assumptions we have used. These numbers clearly reflect the effects of the intermittence or duration time of each activity. Under normal vaping conditions (10-15 puffs per hour) the added risk of low intensity vaping respect to the control scenario of exclusive rest breathing is of the order of ∼ 1 % (since ε−1 ∼ 10⁻²). For high intensity vaping it is ∼ 5−17 %, given the ambiguity in the range of V_d = V_T C_d, still it is of the order of ε − 1 ∼ 5 × 10⁻² – 10⁻¹, also a low added risk since the low value of N_(vp)/N_(br) compensates for the large exhaled tidal volume. Notice that the added risk respect to the control case grows to ∼ 40 % just for talking for 10 % of the time and easily reaches 90 % if talking 40 % of the time. Coughing is also intermittent, possibly even more intermittent than vaping, but its large amount of exhaled droplets (large factor of 28 in (21)) can offset this effect. For speaking ε can be large even if normal speech involves a tidal volume close to rest breathing, but it also involves a much larger amount of time (larger number of breaths in typical conversation).

8.4. Effects of universal face mask wearing

In the previous risk evaluations we assumed no face mask wearing. However, these evaluations remain basically unchanged when assuming universal face mask wearing, as it is well known [144] that these protective devices block respiratory droplets both ways: from emissions by an infected person and also for bystanders wearing them exposed to these emissions. The efficiency of face masks for droplet filtering is sensitive to the involved pathogens [145], the mask and/or fiber type and droplet diameters [146, 147] characterizing respiratory activities [147, 148, 149]. In general, surgical masks and N95 respirators efficiently block droplets in the hyperfine (< 200 nm) and supermicron ranges, less so droplets in the fine submicron range (0.2 – 1.0 µm).

Assuming universal wearing of surgical face masks, with filtration efficiency expressed as a fraction k < 1 of blocked droplets, those exhaled by an infected person and those potentially inhaled by exposed bystanders wearing them. Evidently, the control case baseline quanta emission rate sharply decreases, becoming from (16) ER_q(br) × 1 − k = (1 − k) × 0.3416/h, while intermittent respiratory activities (like vaping) carried without mask wearing keep their quanta emission rates. For low intensity vaping involving N_(vpL) breaths (puffs) per hour with mask removed (but worn for the remaining breaths N_(T) − N_(vpL) with N_(T) the total number of breaths) we have an excess amount of quanta with respect to the decreased control case level where we assumed 80 % filtration efficiency, N_(vp) = 10, N_(T) = 960 and (from (16)) ER_q(vpL) = 0.3562 quanta/h emitted in the 10 breaths without face mask. This increase of 4.39 % of emitted quanta is almost 4 times as much as in the case without face masks (1.25 % in (20)) because the baseline emission dropped by 80 %, but the face masks worn by exposed bystanders will block 80 % of the exhaled droplets from the vaper emitted without wearing a face mask, thus fully compensating for the lowered baseline level. In fact, by being protected with face masks bystanders are exposed to an actual (but negligible) decrease of quanta: 0.2×4.39 % = 0.878 %. It is straightforward to show that universal face mask wearing also involves negligible modifications of risks with respect to baseline state of continuous breathing for other intermittent respiratory activities.

8.5. Disinfectant properties under realistic vaping conditions

It is difficult to relate the highly controlled and idealized experiments described in section 7 to the erratic and highly variable conditions in vaping. First, in these experiments pure PG (as aerosol or as vapor) was supplied continuously and spread evenly under carefully controlled conditions, whereas particulate and gas phase concentrations of ECA rapidly vary with time and position. The gas phase of ECA is a mixture of PG and other compounds (VG, nicotine, with residual concentrations of mostly aldehydes) and is supplied into the surrounding air (when inhaled or exhaled) intermittently during puffs and spreads rapidly and unevenly. Second, bactericidal effects in these experiments were registered with PG concentrations of: 50 – 500 mg/m³ which are 2-3 orders of magnitude higher than maximal gaseous PG concentrations of exhaled ECA registered in experiments involving several users vaping in relatively small chambers during hours: 0.3 – 0.4 mg/m³ [150, 151]. Since about 92 % of inhaled PG is retained [45], PG concentrations inside the respiratory tracts could approach the lower end of concentrations in the experiments. However, the disinfectant effect is unlikely to occur, as the transit and absorption time of ECA in the respiratory system is too short (around 5-6 seconds) and this effect is much less efficient in the prevailing relative humidity close to 100 %.

As we have argued throughout this article, exhaled ECA (as an expiratory activity) should spread respiratory droplets in the environment. However, it is very unlikely that its chemical medium could inhibit COVID-19 contagion by disabling or destroying the SARS-CoV-2 virus. Conversely, it is equally unlikely that this chemical medium could (somehow) enhance the probability of contagion in comparison with “normal” airborne transmission. Nevertheless, the chemical interaction between the SARS-CoV-2 virus and PG and/or other glycols and compounds of ECA needs to be probed and tested in well designed experiments, even outside the context of vaping.

9.1. Limitations

9.2. Safety considerations