key: cord-0913838-j5jcn8bx
authors: Lin, Xiao-Ting; Hsu, Nai-Yun; Wang, Jen-Ren; Chen, Nai-Tzu; Su, Huey-Jen; Lin, Ming-Yeng
title: Development of an efficient viral aerosol collector for higher sampling flow rate
date: 2017-11-25
journal: Environ Sci Pollut Res Int
DOI: 10.1007/s11356-017-0754-z
sha: 44031097dd39d7fde7b22c1115fc23ed0ac076dd
doc_id: 913838
cord_uid: j5jcn8bx

Viral aerosol infection through cough generates large amounts of viral aerosol and can result in many adverse health effects such as influenza flu and severe acute respiratory syndrome (SARS). To characterize the coughed viral aerosol, the sampler needs to sample at higher flow rate and possess high physical collection efficiency as well as high viral preservation. However, most current inertia-based high flow bioaerosol samplers are not suited for viral aerosol sampling since the viability will be lost doing the sampling process. Current condensation growth methods only have good physical collection efficiency and viral preservation at low flow rate (< 10 LPM). In this study, we developed a viral aerosol sampling system using a cooler and steam-jet aerosol collector (SJAC) for bioaerosol collection for the first time. The system is based on mixing condensation growth method and has high viral preservation at a higher flow rate (12.5 LPM). We control the inlet aerosol flow temperature and the SJAC mixing reservoir temperature to improve the physical collection efficiency and viability preservation of the viral aerosol. Results indicate that the physical collection efficiency is 70–99% for aerosol 30–100 nm when the aerosol flow and mixing reservoir temperature was 19 and 50 °C, respectively. In addition, the system was 7 and 22 times more efficient for viability preservation of MS2 bacteriophage than the commonly used All Glass Impinger 30 (AGI-30) and BioSampler®, respectively. Finally, the system can be applied to sample at a lower concentration (10(5) PFU/m(3)), and results shows the system was 4.7 times more efficient for viability preservation than using AGI-30 alone. The developed viral collection system will improve our understanding of the characteristics of coughed aerosol and can be used for future evaluation of respiratory protective equipment and environmental sampling.

Viral aerosol spreads infectious diseases such as severe acute respiratory syndrome (SARS) and influenza, which can have significant adverse health effects (Lee et al. 2003; Lipsitch et al. 2003 ; Thompson et al. 2003) . Considerable global economic losses and public fear are also associated with viral aerosol infection. In 2008, 90 million children were infected by influenza which can result in acute lower respiratory infections, birth defects, and even death. Among the recorded number of deaths, 99% occurred in developing countries (Liang et al. 2017; Osterhaus et al. 2015) . Another example is the outbreak of pandemic influenza HIN1 in 2009, resulting in an estimated 201,200 of respiratory deaths and an additional 83,300 cardiovascular deaths worldwide (Dawood et al. 2012 ). Viral aerosol originated from cough has initial sizes of 1-100 μm, but evaporates to droplet nuclei of 20-300 nm in a short amount of time (Kowalski et al. 1998; Morawska 2006; Tseng and Li 2005) . Due to their small size, viral aerosol can penetrate deep into the lungs and are less easily removed by macrophages (Buzea et al. 2007 ). In addition, viral aerosol can remain airborne for hours to Responsible editor: Philippe Garrigues days (Verreault et al. 2008) , adding the chance of infection. The viability of the aerosol is also affected by the content of salt concentration in saliva and relative humidity during droplet evaporation (Yang et al. 2012) . Agencies such as the US and European Centers for Diseases Control (CDC) and World Health Organization (WHO) have expressed urgent need for research in influenza virus transmission (see references within (Bischoff et al. 2013) ). Therefore, detecting the viral aerosol concentration and viability is critical in identifying and preventing disease transmission in public health.

Some commonly used instruments for viral aerosol collection are the All Glass Impinger 30 (AGI-30) and SKC BioSampler® (Willeke et al. 1998 ). AGI-30 and BioSampler® collect aerosol by liquid impingement and tangential impingement, respectively. An ideal viral aerosol sampler should possess both high physical collection efficiency and good viability preservation. However, studies point out that for aerosol 30-100 nm, AGI-30 and BioSampler® have physical collection efficiency less than 10% (Hogan et al. 2005) . Using filter and electrostatic precipitators can lead to higher physical collection efficiency for viral aerosol, but can damage the viability of viral aerosol (Kettleson et al. 2009; Roux et al. 2016; Yang et al. 2011) . Recently, condensation growth method gained significant attention in viral aerosol collection due to its ability to possess high physical collection efficiency and viability preservation. Condensation growth method helps Bgrows^the aerosol to a larger size that can be easier for collection or analysis. Another advantage of the condensation growth methods is that the bioaerosol is collected in liquid phase which is beneficial for later biochemical analysis (Alvarez et al. 1995) . Several studies have achieved promising results using this principle at lower flow rates (< 10 LPM). When pairing the condensation growth method with the AGI-30 or BioSampler®, the physical collection efficiency and viability preservation of viral aerosol can be significantly enhanced. For instance, when adding condensation growth methods such as the mixing-type bioaerosol amplification unit (mBAU) and growth tube collector (GTC) to the BioSampler®, the collected viable MS2 bacteriophage efficiency increased 2-3 and 10-100 times, respectively (Oh et al. 2010; Pan et al. 2016) . However, at higher flow rates (> 10 LPM) as compared to button samplers or deposition devices, applying condensation growth methods for viral aerosol collection were only similar, not significantly better than using AGI-30 impinger and BioSampler® alone Milton et al. 2013) . For example, Gesundheit II (G-II) is inefficient for collecting aerosol smaller than 50 nm. However, many viral aerosols have sizes smaller than 50 nm. Therefore, when using G-II for sampling coughed influenza virus, the system has a high physical collection efficiency, but a low viability preservation (10%) (Hatagishi et al. 2014) . Adopting higher sampling flow rate can increase the amount of sampled air and improve the limit of detection (LOD). In addition, a higher flow rate viral aerosol sampler is also needed for sampling exhaled aerosol during tidal breathing. Nevertheless, most of the current bioaerosol sampler that uses condensation growth methods cannot increase the viable aerosol collection efficiency and viability preservation of the BioSampler® at flow rates higher than 10 LPM.

One option is to use the steam-jet aerosol collector (SJAC) which is also based on condensation growth methods for viral aerosol collection at higher flow rates. SJAC has been widely used for ambient aerosol collection in different campaigns (Rees et al. 2004; Slanina et al. 2001 ). SJAC has a wide range of sampling flow rate from 10 to 50 LPM, enabling aerosol sampling at higher flow rates (Khlystov et al. 2000) . For aerosol larger than 20 nm, SJAC has a physical collection efficiency over 99% for concentration up to 6 × 10 5 #/cm 3 (Khlystov et al. 1995; van Rens et al. 2007) . In this study, we will collect viral aerosol by pairing the SJAC with the AGI-30 and BioSampler® for the first time. In addition, we will test this system at a lower viral aerosol concentration to simulate ambient conditions. The results will help future viral aerosol collection using condensation growth methods at higher flow rate and for ambient environments.

The steam-jet aerosol collector ( Fig. 1) is made up of two main parts, the mixing reservoir and cyclone. Aerosol flow mixes with the steam vapor flow in the mixing reservoir, in which supersaturation is reached and condensation growth occurs. The enlarged aqueous phase aerosol is collected by the cyclone connected behind the mixing reservoir. In this study, we only use the mixing reservoir of the SJAC which is made of stainless steel. The cyclone is replaced by the AGI-30 or BioSampler® for viral aerosol collection. We mounted a 500-ml three-neck round-bottom flask (Pyrex, Tewksbury, MA, USA) which is half filled with ultrapure water (Reference Water Purification System, Milli-Q®, Kenilworth, NJ, USA) to a heating mantle (NewLab Instrument Co., Ltd., New Taipei City, Taiwan). The SJAC is connected to the center of the flask neck. We also inserted a temperature controller (HT-720, Newlab Co Ltd., New Taipei City, Taiwan) and temperature sensor in the side flask to control and monitor the temperature.

Condensation growth occurs under supersaturated vapor phase. This occurs when the saturation ratio (S) is larger than 1. Saturation ratio is given as follows:

where p i is the vapor partial pressure and p sat,T is the saturation vapor pressure at temperature T. Supersaturation can be reached through mixing condensation (i.e., mixing two flows with different temperatures, relative humidity, and flow rates) (Kousaka et al. 1982; Okuyama et al. 1984) . Assume flow one (aerosol flow) has temperature (T sl ), relative humidity (H sl ), flow rate (Q sl ), and specific heat (C sl ); flow two (vapor flow) has temperature (T sh ), relative humidity (H sh ), flow rate (Q sh ), and specific heat (C sh ). When aerosol flow and vapor flow mix with each other, saturation point i is initially reached. At point i, according to the enthalpy and mass balance equation, the temperature (T i ), relative humidity (H i ), flow rate (Q i ), and specific heat (C i ) become

The saturation point i then moves adiabatically towards point f as condensation progresses; during this process, the amount of vapor (ΔH) that condenses on the aerosol is

where λ is the latent heat. The temperature and relative humidity at point f obtained from the saturation vapor curve are T sf and H sf , respectively. According to Eqs. 2-6, at the same vapor flow condition (i.e., same T sh , H sh , Q sh , and C sh ) and fixed T sf , lower temperature in the aerosol flow (T sl ) will result in a lower T i . Therefore, ΔH will become larger, indicating more vapors condensing to the aerosol and enhancing aerosol growth. (Hogan et al. 2004 ). In addition, the high stock titer of MS2 bacteriophage has the advantage of better virus detection and lower measurement uncertainty (Ge et al. 2014) . Therefore, MS2 bacteriophage is widely used as a viral aerosol surrogate (Balazy et al. 2006; Walker and Ko 2007) . Plaque assay technique was used to characterize the collected MS2 bacteriophage concentration, and each sample was assayed in duplicates. All the plates were incubated at 37°C over 12 h using the spread plate method (Cormier and Janes 2014) . We used ultrapure water to dilute the supernatant to a titer of 10 9 plaque-forming units (PFUs)/ml for viral aerosol generation. Note that 10 9 PFU/ml is much higher than in real-life environments, but this concentration works better for viral culture when testing the SJAC system. In addition, we also used a lower MS2 concentration (i.e., a titer of 10 4 PFU/ml) to simulate real-life situations.

The schematic figure of the experimental setup is shown in Fig. 2 (Olson et al. 2004) . However, our current setup only allows our aerosol flow temperature to reach 19°C. At a lower aerosol flow temperature (19°C), there is a larger temperature difference between the cooling section and the SJAC mixing reservoir, thus facilitating condensation growth on the viral aerosol (Kousaka et al. 1982) . We wrapped the copper tube between the cooling section exist and the SJAC inlet with a thermal insulator made of polyethylene to maintain constant temperature of the aerosol flow. We also monitored the temperature and relative humidity with a sensor (HMT 330 Series, Vaisala, Helsinki, Finland) at the SJAC inlet. The air flow enters the SJAC at flow rate of 12.5 LPM. Note that the residence time of the aerosol flow in the cooling section and SJAC mixing reservoir are 0.14 and 0.19 s, respectively. The AGI-30 or the BioSampler® was connected behind the SJAC to collect the MS2 bacteriophage. Ultrapure water was used as the collection medium. We used a thermometer (PT100, PU-YANG Electronics Co., Taoyuan, Taiwan) to monitor the temperature in the SJAC mixing reservoir. In the following physical collection efficiency, particle growth, and viability preservation experiments, we examined the MS2 bacteriophage viability preservation at different temperatures of aerosol flow (19 and 22°C) and SJAC mixing reservoir (37, 45, and 50°C). Note that throughout each temperature combination, the SMPS concentration and mode was within 10% of 2.5 × 10 6 #/cm 3 and 40 nm, respectively.

Evaluation of physical collection efficiency and particle growth Figure 2 is the experimental setup for measuring physical collection efficiency and particle growth. We measured the physical collection efficiency (E) for different particle sizes before and after the collector (i.e., the cooler, SJAC, and AGI-30 or Biosampler®) using the SMPS (DMA model 3081 and CPC model 3010, TSI Inc., USA). At each temperature combination, we performed three sets of measurement. Each set contains one upstream and one downstream scan; each scan is 90 s (60 s upscan/30 s downscan). The physical collection efficiency is given as follows: 

To evaluate the particle growth, an aerodynamic particle sizer (APS, model 3221, TSI Inc., USA) was connected to the two sampling ports located before and after the cooler and SJAC. The two sampling ports were connected to the APS through a threeway valve, allowing us to switch the sample airflow between the two ports. At each temperature combination, we performed three sets of measurement (APS scan upstream and downstream of the SJAC, each scan is 10 s).

The viability of the collected MS2 bacteriophage in AGI-30 and BioSampler® was determined by using the PFUs. We calculated the PFUs with and without the SJAC. It should be noted that in both scenarios, the SJAC and cooler were both attached to the system as seen in Fig. 2 . The only difference is in the case without SJAC; the heater of the SJAC was turned off. We can then obtain the viral aerosol collection enhancement factor (VCEF) given as follows:

PFU SJAC-on indicates that the heater of the SJAC is turned on and PFU SJAC-off means that the heater of the SJAC is turned off. VCEF values were used to evaluate the viability preservation enhancement when using the SJAC. Higher VCEF Fig. 2 A sketch of the experimental setup. Note that for the control case without the SJAC, the SJAC is still attached to the system. The only difference for with and without SJAC is that the heater of the SJAC is turned on and off, respectively values indicate that the cooler and SJAC improve the viable MS2 bacteriophage viability preservation over using AGI-30 or BioSampler® alone. We also investigated how different sampling time affect the VCEF values. All data points were generated with three replicates.

To simulate the real-life viral aerosol concentration, a lower MS2 concentration was used. In ambient concentration, the viral aerosol concentration is 10 5 -10 7 #/m 3 (Prussin et al. 2015; Whon et al. 2012) . We first dilute the supernatant to a titer of 10 4 PFU/ml with ultrapure water for viral aerosol generation from the nebulizer. This is equivalent to a concentration of 10 5 PFU/m 3 in the mixing chamber. The viability and total virus concentration were determined by using PFUs. We used a longer sampling time (30 min) as recommended by other studies to collect more viral aerosol for later analysis, but not too long as to reduce the physical collection of AGI-30 (Macher et al. 1995; Willeke et al. 1998 ). The aerosol flow and SJAC mixing reservoir temperatures were set to 19 and 45°C, respectively.

SJAC physical collection efficiency and size enlargement Figure 3 shows the physical collection efficiency of various particle sizes under different temperature combinations of aerosol flow and SJAC mixing reservoir. Overall, for aerosol larger than 30 nm, the physical collection efficiency was over 66 and 57% for AGI-30 and Biosampler®, respectively. The highest physical collection efficiency (70-99% for AGI-30 and Biosampler® when the aerosol was larger than 30 nm) was achieved when the aerosol flow and SJAC mixing reservoir temperature was 19 and 50°C, respectively. This is a significant improvement as compared to using the AGI-30 or Biosampler® alone where the physical collection efficiency is less than 10% in the 30-100-nm range (Hogan et al. 2005) . Moreover, using AGI-30 as a viral aerosol collector in the SJAC system has a higher physical collection efficiency over Biosampler®. In addition, at the same SJAC mixing reservoir temperature, the physical collection efficiency was higher in both AGI-30 and Biosampler® when the aerosol flow temperature was 19°C as compared to 22°C, indicating that more condensation growth in viral aerosol occurred. Figure 4 shows the aerosol concentration and size distribution of MS2 bacteriophage in front and behind the SJAC under different temperature combinations. Higher SJAC mixing reservoir temperature increased the aerosol number concentration and mode by enhancing aerosol growth. In addition, lower temperature in the aerosol flow showed more condensation growth also resulted in a higher aerosol number concentration. The largest growth was seen when the aerosol flow and SJAC mixing reservoir temperatures were 19 and 50°C, respectively. During which, the concentration and mode from APS measurements reached 6.3 × 10 4 #/cm 3 and 1.6 μm, respectively. However, the VCEF decrease in this scenario, and this could also be attributed to the high temperature that can damage the viability of MS2 bacteriophage. Figure 5 shows the collected viable MS2 bacteriophage concentration in the water medium of AGI-30 and BioSampler® over the 15 min of sampling duration. During the 15-min sampling interval, a total of 187.5 l of air was sampled. Assuming indoor viral aerosol concentration to be 10 5 #/m 3 and physical collection efficiency of around 66% using SJAC system, a total of 12,375 would be sampled. This amount sampled is higher than the 3375 viral aerosol threshold dose needed to induce infection (Yang et al. 2011 should be noted that when the SJAC mixing reservoir temperature was 50°C, no significant increase in the collected viable MS2 bacteriophage concentrations was found in AGI-30. This could be attributed to the higher SJAC mixing reservoir temperature affecting the viability of the MS2 bacteriophage.

Researches have indicated that high temperature can result in lost of infectivity of the MS2 bacteriophage (Anders and Chrysikopoulos 2006; Brie et al. 2016; Verreault et al. 2015) . Another reason may be attributed to the relatively large error bar at 50°C from AGI-30. Also, note that the collected viable MS2 bacteriophage concentration in liquid media from AGI-30 was higher than the BioSampler®, and this is similar to other studies that show AGI-30 has a higher viable collection efficiency at higher flow rates (> 9 LPM) Hogan et al. 2005) . respectively. The larger temperature difference between the aerosol flow temperature and SJAC mixing reservoir resulted in a higher viability preservation and viral aerosol condensation growth. This is consistent with the mixing condensation theory mentioned in BMaterials and methods^ (Kousaka et al. 1982) . In addition, the VCEF was highest when the SJAC mixing reservoir temperature was 45°C instead of 50°C. At 50°C, the hot steam vapor from the SJAC may damage the MS2 bacteriophage viability during the condensation growth process.

The VCEF values from this study are mostly higher than the results from the mixing-type bioaerosol amplification unit (mBAU) which has VCEF around 2-3 (Oh et al. 2010) , and both studies have a sampling flow rate of 12.5 LPM and use the BioSampler® as the reference sampler. However, the VCEF values herein are mostly lower than the growth tube collector (GTC) (Pan et al. 2016) . The GTC has a lower sampling flow rate (7 LPM) compared to this study (12.5 LPM). At higher flow rate, according to Eqs. 2-6, the amount of vapor ΔH condensing on the viral aerosol is smaller, leading to smaller VCEF values. Furthermore, at higher flow rate, the residence time in the SJAC mixing reservoir was also smaller, leading to less time for aerosol condensation growth. 

Here, we compared the MS2 bacteriophage viability over different sampling periods (5, 10, and 15 min). Note that 15 min of sampling time is often used for both AGI-30 and BioSampler® (Riemenschneider et al. 2010) . The temperature in the aerosol flow and SJAC mixing reservoir was set to 19 and 45°C, respectively, in this experiment. From Fig. 6 , VCEF values increased with increment sampling time for both AGI-30 and BioSampler®. The largest increase was seen between 10 and 15 min. The VCEF values while sampling for 15 min were significantly higher (P < 0.01) than sampling for 5 min from the Kruskal-Wallis test. The VCEF result for 15 min sampling duration is similar to another study using the BioSampler®, but with a lower flow rate (Pan et al. 2016 ). The AGI-30 and BioSampler® can be used for long sampling duration up to 30 min and 8 h, respectively. As the sampling time increases, the collection efficiencies of the AGI-30 and BioSampler® decrease due to aerosol reaerosolization (Hogan et al. 2005) . At longer sampling periods for viral aerosol, adding the cooler and SJAC to the sampling apparatus could provide better results. This shows the potential of applying the cooler and SJAC to viral aerosol collection. This further indicates the feasibility of using condensation growth method for viral aerosol collection at higher flowrate (> 10 LPM).

When sampling at 10 5 PFU/m 3 in the mixing chamber for 30 min with AGI-30 sampler, the aerosol flow and SJAC mixing reservoir temperatures were set to 19 and 45°C, respectively. The collected MS2 concentrations in the AGI-30 were 527 and 113 PFU/ml with and without using the SJAC, respectively. This is equivalent to a VCEF value of 4.7. The system shows the ability to enhance the viability preservation of AGI-30 sampler during low viral aerosol concentration. Therefore, the SJAC system can be applied to sampling real-life conditions where the viral aerosol concentration is lower. However, more field studies is needed to verify the SJAC system.

Significant improvements in viral aerosol physical collection efficiency and viability preservation were seen when adding the cooler and SJAC to the AGI-30 and BioSampler® during aerosol collection. The SJAC system's physical collection efficiency is 57-99% for viral aerosol larger than 30 nm. When the temperature of the aerosol flow was at 19°C and the SJAC mixing reservoir was 45°C, the VCEF are 7 and 22 for AGI-30 and BioSampler®, respectively. Furthermore, adding the cooler and SJAC to AGI-30 and BioSampler® can significantly increase the VCEF values when the sampling time increased from 5 to 15 min. This shows the potential to use this sampling system for viral aerosol sampling for a longer time period. Finally, the SJAC system has VCEF of 4.7 when the MS2 concentration was as low as 10 5 PFU/m 3 , indicating that the SJAC system can be applied to future ambient environment sampling.

However, the study is limited to flow rates of 12.5 LPM. Future studies on even higher inlet flow rates (preferably over 100 LPM) using condensation methods are also needed for sampling exhaled aerosol during tidal breathing. In addition, application of the SJAC to other coughed viral aerosol collections such as influenza virus is also needed. Furthermore, additional measures such as staining methods for viability or qPCR/ddPCR tools for assessment of total viral particles are also needed to strengthen the study. The work shows the potential of SJAC system in the viral aerosol collection at higher flow rates which can facilitate our understanding towards environmental viral aerosol. This viral collection system can potentially benefit future infectious disease modeling and prevention work.

PCR for bioaerosol monitoring-sensitivity and environmental interference

Evaluation of the factors controlling the time-dependent inactivation rate coefficients of bacteriophage MS2 and PRD1

Do N95 respirators provide 95% protection level against airborne viruses, and how adequate are surgical masks?

Exposure to influenza virus aerosols during routine patient care

The effect of heat on the physicochemical properties of bacteriophage MS2

Nanomaterials and nanoparticles: sources and toxicity

A double layer plaque assay using spread plate technique for enumeration of bacteriophage MS2

Estimated global mortality associated with the first 12 months of 2009 pandemic influenza A H1N1 virus circulation: a modelling study

Airborne virus survivability during long-term sampling using a non-viable Andersen cascade impactor in an environmental chamber

Establishment and clinical applications of a portable system for capturing influenza viruses released through coughing

Capture of viral particles in soft Xray-enhanced corona systems: charge distribution and transport characteristics

Sampling methodologies and dosage assessment techniques for submicrometre and ultrafine virus aerosol particles

Airborne virus capture and inactivation by an electrostatic particle collector

The steam-jet aerosol collector

Characterization of particulate matter in urban air: instrumental development and experimental

Bactericidal effects of high airborne ozone concentrations on Escherichia coli and Staphylococcus aureus

A major outbreak of severe acute respiratory syndrome in Hong Kong

The influence of maternal exposure history to virus and medicine during pregnancy on congenital heart defects of fetus

Transmission dynamics and control of severe acute respiratory syndrome

Sampling airborne microorganisms and aeroallergens. In: Air sampling instruments for evaluation of atmospheric contaminants

Development and performance evaluation of an exhaled-breath bioaerosol collector for influenza virus

Influenza virus aerosols in human exhaled breath: particle size, culturability, and effect of surgical masks

Droplet fate in indoor environments, or can we prevent the spread of infection?

Development and evaluation of a novel bioaerosol amplification unit (BAU) for improved viral aerosol collection

Condensational growth of ultrafine aerosol-particles in a new particle-size magnifier

Effects of freezing and storage temperature on MS2 viability

Why should influenza be a public health priority?

Efficient collection of viable virus aerosol through laminar-flow, water-based condensational particle growth

Total concentrations of virus and bacteria in indoor and outdoor air

Mass balance closure and the federal reference method for PM2

Characterization of reaerosolization from impingers in an effort to improve airborne virus sampling

Development of a new portable air sampler based on electrostatic precipitation

The continuous analysis of nitrate and ammonium in aerosols by the steam jet aerosol collector (SJAC): extension and validation of the methodology

Mortality associated with influenza and respiratory syncytial virus in the United States

Collection efficiencies of aerosol samplers for virus-containing aerosols

Proving the possibility of physical removal of particulate matter with the SJAC-technology

Methods for sampling of airborne viruses

Resistance of aerosolized bacterial viruses to relative humidity and temperature

Effect of ultraviolet germicidal irradiation on viral aerosols

Metagenomic characterization of airborne viral DNA diversity in the near-surface atmosphere

Improved aerosol collection by combined impaction and centrifugal motion

Concentrations and size distributions of airborne influenza A viruses measured indoors at a health centre, a day-care centre and on aeroplanes

Relationship between humidity and influenza A viability in droplets and implications for influenza's seasonality

Acknowledgements We greatly appreciate the help from Prof. Hsion-Wen Kuo, Mrs. Ching-Hui Shih, and Mrs. Wen-Yun Chen with the experiment. Funding information The authors greatly acknowledge the support from Taiwan's National Science Council (NSC 102-2218-E-006-002-MY2) and Ministry of Science and Technology (MOST 104-2221-E-006-186-).

Conflict of interest The authors declare that they have no conflict of interest.