Laboratory study of bioaerosols: Traditional test systems, modern approaches, and environmental control

Abstract

Biological aerosol particles have been studied in the laboratory for many decades to understand their roles in human health and disease and in environmental processes. These studies have used a variety of instrumentation under varying conditions. This review covers the most common types of bioaerosol chambers (e.g., static, flow-through and rotating drum), the study of biological particles captured on spider's webs, and single particle levitation systems (optical, acoustic and electrodynamic). It also discusses the variety of ways the study environment (temperature, humidity, light and trace gas) may impact biological particles through modification of both biological properties, such as proteins, DNA, and viability, and the physical properties such as size and fluorescence that are relied upon for measurement. The result is a broad look at the many ways biological particles can be studied in the laboratory and the many factors that past studies have determined impact the properties of bioaerosols.

1. Introduction

Laboratory studies are central to understanding aerosol behavior in many fields of aerosol science. Observations in the ambient environment are complicated by the uncontrollable nature of the atmosphere (e.g., Santarpia 2016, Huffman and Santarpia 2017), and aerosol chambers provide a way to study aerosol behavior where it is possible to limit the variables impacting the particles, contain hazardous particles and gas-phase species and control the environment around the particles to study subsets of aerosol processes. In particular, studies of biological particles use chambers and control devices with a variety of forms and utility. Several key aspects of chamber design are unique to the study of bioaerosols. First, excepting fragments, bioaerosol size range spans four orders of magnitude, from 0.01 to 10 s of micrometers (which equates to twelve orders of magnitude of volume or mass) (Jonsson, Olofsson, and Tjärnhage 2014). This vast range has dramatic effects on aerosol dynamics and its subsequent chemical/biological processing, while simultaneously affecting the particle sedimentation rate/fractionation (Frohlich-Nowoisky et al. 2016), all of which need to be considered in the design of a laboratory system. Second, the challenges associated with the complexity of aerosol plumes has limited the amount of information that can be inferred from any one experiment. By definition, any population of bioaerosol will have a broad size distribution (Allegra et al. 2016). This variability alone will cause significant changes in the parameters critical to understanding bioaerosol dynamics, such as microorganisms per droplet, surface to volume ratio, and fractionation. Thus, the study of bioaerosol is often the study of the average behavior of populations, and as a result, it is difficult to accurately quantify stimulus-response relationships to the degree necessary to produce rigorous predictive models. To compensate for this limitation, novel single particle control techniques have been developed (e.g., Hopkins et al. 2004; Redding et al. 2015; Redding and Pan 2015; Pope 2010; Fernandez et al. 2019). Third, experimentalists often need protection from the bioaerosols, as many of interest are pathogenic, or at the very least may cause unwanted immune response in large quantities (e.g., Kim, Kabir, and Jahan 2018). Lastly, the aerosols in the chamber need to be isolated from the ambient environment or portions of it. The effects of the ambient environment on bioaerosols is an active area of research (Santarpia et al. 2013; Pan, Santarpia, et al. 2014; Ratnesar-Shumate et al. 2015; Kinahan et al. 2019) and other areas of bioaerosol research, where these effects are not considered in the design or analysis of data may be adversely affected. The intent of this work is to provide an overview of the variety of chambers and control systems used in bioaerosol studies, the environmental variables that may impact them, and recommendations for the application of various control technologies to current bioaerosol research study issues.

2. Aerosol control and study systems

There are many and varied approaches to delivering, containing, and measuring aerosol for study. Depending on the needs of the experiment, different types of aerosol confinement and methods can be employed. The primary drivers for the design of biological aerosol systems are containing the biological aerosol and minimizing the physical loss of particles, so that the chemical and biological properties of aerosol can be studied over time. An overview of the systems that have been developed to study biological aerosol, are described in this section and range from systems that contain large volumes of high concentration aerosol in stagnant conditions to systems that attempt to counteract gravitational settling to systems designed to study single or small populations of confined particles.

2.1. Static chambers

A basic static chamber (Figure 1a) can be used in bioaerosol studies in which the aerosol is not required to be suspended for long periods of time or do not require agitation or reaerosolization (Dybwab and Skogan 2017; Kesavan and Stuebing 2009). A limitation to the use of these chambers is the loss of aerosol due to gravitational settling. Static chambers have been constructed in varying sizes with large chambers constructed up to 1 million Liters (Ruoff 1998; Christopher et al. 1997; Reed, Nalca, and Roy 2018). Fans or blowers may be utilized to achieve uniform mixing and increased loft time within static chambers if needed. Physical characterization of the residence time in the chamber should be performed for the given bioaerosol and associated size distribution at each humidity level to be studied in order to account for losses in the static chamber and to ensure no bias in the analysis of results.

Figure 1. Representative drawings of bioaerosol control systems and considerations for their use.

2.2. Flow-thru or wind-tunnel chambers

Bioaerosol experiments may also require a continuous feed of either air or particulates for a study. Systems such as continuous flow or wind-tunnels (Figure 1b) have been used to assess the performance of bioaerosol sampling devices or optical sensors (Upton et al. 1994; Griffiths et al. 1997; Hairston, Ho, and Quant 1997; Su et al. 2012; Sivaprakasam et al. 2007; Emanuel et al. 2012), as well as test and develop new collection systems designed for maintaining bioaerosol integrity (McFarland et al. 2010; King and McFarland, 2012a, 2012b). Depending on the total flow rate of the system, control of temperature and humidity may be possible. Test systems have been developed that allow for simultaneous introduction of multiple types of aerosols into a wind-tunnel for dynamic sensor challenges (Ratnesar-Shumate et al. 2011; National Research Council 2005). Physical characterization in flow-thru or wind-tunnel chamber is also needed to ensure that size-dependent losses are accounted for in data analysis. Depending on the flow velocity, turbulence and length of test system, aerosol will be lost due to gravitational settling and other turbulent loss mechanisms (e.g., Montgomery and Corn 1970). Depending on the desired outcome of the experiment, this may be problematic and can be assessed by measuring the particle size distribution at the beginning and end of the system using either a real-time optical particle counter, spectrometer or cascade impactor.

2.3. Rotating chambers

Aging of bioaerosols in the natural environment involves many different chemical reactions that occur on multiple time scales (minutes, hours, days) and may involve sequential reactions. To realistically simulate conditions in the natural atmospheric environment for studies on aging of bioaerosols, it may be necessary to maintain aerosol populations for long intervals of time. Rotating drum chambers (Figure 1c) are an alternative to static chambers and provide longer residence times for bioaerosol studies in which observation of changes over the course of time are the intent (Goldberg et al. 1958; Krumins et al. 2008; Thompson, Bennett, and Walker 2011 Verreault et al. 2014,; Piercy et al. 2010; Haddrell and Thomas 2017). Similar to static chambers, physical losses occur in rotating drum chambers due to gravitational settling and diffusion of the aerosol to the walls of the chamber. The rotational speed of the drum can be changed depending on the particle size distribution to optimize residence time (Gruel, Reid, and Allemann 1987; Asgharian and Moss 1992). However, despite optimization, size dependent loss over long experiment times is unavoidable (Gruel, Reid, and Allemann 1987). If the goal of the study is to observe changes in the bioaerosol over time, those changes may be more or less likely to occur on the bioaerosol depending on the particle size. One example is when the fate of a bioaerosol needs to be determined when exposed to a light source as a function of size. Size dependent loss in activity has been observed for single versus clusters of spores (Kesavan et al. 2014). Given results like those of Kesavan et al. (2014) and an understanding of size dependent loss in a rotating drum chamber it is important to consider that observations of aerosol fate over long periods of time in a rotating drum chamber will be biased towards particles of the size optimized for the longest lifetime. This bias may cause over or underestimation of the effect being studied if the bias is not considered when the observations are extended to the entire polydisperse distribution of aerosol in the chamber. Therefore, consideration of both size dependent aging mechanisms and the size dependent physical losses are critical to the interpretation of results in these types of chamber systems, particularly when experiments are done over long periods of time.

2.4. Microthread systems

Microthreads, harvested from spiders or synthetically derived, have been used to capture and suspend bioaerosol particles for studies (Figure 1d; Druett 1971; May and Druett 1968; Smither et al. 2011; Handley and Roe 1994). The use of microthreads avoids the problem of physical losses in the chamber and allows for continued observation of bioaerosol over long periods of time. Physical characterization of the chamber and recovery of the bioaerosol from the microthreads needs to be quantified for a given type of bioaerosol particle, size and exposure condition (e.g., relative humidity). Microthreads have notable advantages for studying larger aerosols that have high physical loss rates in static and rotating drum chambers. One limitation of this technique is that the bioaerosol particles are stationary on the surface of the microthread. This potentially poses two issues. First, the portion of the bioaerosol particle that is in contact with the microthread is not exposed to the surrounding air and the water vapor, trace gas species, etc. contained therein. This is likely not an issue for bioaerosols that are much larger than the thread, which have been reported to range from 0.01 to 15 μm (May and Druett 1968; Druett 1971), but may lead to observed differences due to shielding of the adhered side of the aerosol from the environmental conditions in the chamber for those that are smaller or similar in size of the diameter of the threads. The second potential side effect of having the particle deposited on a microthread surface is that it is no longer suspended in the air and rotating, but rather is stationary. For systems in which changes due to exposure to different types of light are the intended experiment, it might be possible that that active components of the larger bioaerosol are shielded because they are on the opposite side of the incident light, and thus are conserved from the intended exposure. In both cases, the side of the bioaerosol in contact with the thread may or not be experiencing the same type of exposures as the opposite side which may impact the degree to which the effect attempting to be observed can occur.

2.5. Fielding aerosol systems in-situ

Understanding aerosol behavior under ambient conditions has numerous technical challenges. Field measurements of existing particles provides only limited data on the composition and history of the primary particle. For biological particles, this is further complicated by the limited ability to cultivate and characterize individual ambient bioaerosols (Santarpia 2016 provides a review). Aerosolizing seed particles of known composition into the open atmosphere is not necessarily a viable option, both due to the difficulty in tracking released particles and due to potential hazards posed by the seed particles themselves. One approach to overcoming these issues is to develop aerosol control systems (that can be fielded in-situ and used to contain known aerosol that is exposed to local environmental conditions (Figure 1e). This approach has been applied in multiple studies. For instance, the quasi-atmospheric aerosol evolution study (QUALITY; Peng et al. 2016) chamber has been used to study changes to black carbon particles in both Houston, TX and Beijing, China. The Captive Aerosol Growth and Evolution (CAGE; Antonetti 2014) chambers were used for the study of biological aerosol evolution in Adelphi, MD in 2012. These chambers allow ambient gas-phase species to penetrate through a permeable, expanded PTFE membrane and allow solar radiation to penetrate through transparent FEP walls to contain particles in an environment that is representative of the local environment. In addition, the CAGE chambers were designed as rotating drums (described above) to allow biological and any super-micron particle to be suspended for extended periods of time (hours).

2.6. Single particle levitation and analysis

Some conditions, such as solute supersaturation and high surface area to volume ratio, are unique to aerosols and cannot be mimicked in the bulk phase. (Svenningsson et al. 2006). As a result, to understand the behavior of microorganisms in the aerosol phase, they must be measured directly in the aerosol phase (Haddrell and Thomas 2017). For this reason, coupled with the challenges associated with polydisperse bioaerosol analysis, there is a strong motivation to develop strategies to study individual bioaerosol droplets/particles (Figure 1f). This approach addresses many of the issues associated with studying polydisperse bioaerosol, but still has some limitations of its own.

A major benefit of single particle analysis is that an individual bioaerosol droplet/particle can be confined in a fixed position in three-dimensional space; consequently, it can be probed through numerous different detection methods, such as light scattering, while levitated (Reid et al. 2007). The ability to collect elastic light scatter or fluorescence spectra from individual bioaerosol droplet/particle in a plume online has been reported and found to be primarily useful for general identification (e.g., fungi versus pollen) (Healy et al. 2012; Toprak and Schnaiter 2013). However, Raman spectra are far more informative, and thus desirable, than elastic scattering or fluorescence as it yields information about composition and structure of the chemicals present in biological samples (Xie, Dinno, and Li 2002; Xie et al. 2003; Li et al. 2019). Given the relative sensitivities between Raman scatter and fluorescence, the droplet must be confined to a much larger degree for a Raman spectrum to be collected. Access to online Raman analysis affords the opportunity to investigate precise characterization of bioaerosol in real-time.

As discussed, a hallmark of single particle analysis is that the individual droplet is trapped in a small, confined and tailorable environment, termed the “trap”. The internal volume of the trap is typically described in milliliters. This small size allows the conditions that the droplet experiences, such as the temperature, pressure or relative humidity, to be both readily controlled and altered rapidly. Additionally, as the position of the droplet in three-dimensional space is focused, the individual droplet can be exposed to any light source; this control allows for the precise luminous flux to be known. Collectively, these features enable one to explore in detail the specific parameters that govern bioaerosol longevity, etc.

The initial solute composition of aerosol produced in a controlled environment, with devices such as nebulizers or droplet on demand dispensers, is entirely dependent on the makeup of the starting formulation. If the starting formulation is homogeneous (lacks any potential sediment, such as crystals or biological components), a single droplet produced will be a good representative sample of the whole; meaning that a low number of droplets will be enough to generate a robust dataset. For this reason, single droplet analysis technologies have been shown to be effective in a range of disciplines from climate science to pharmaceutics (Haddrell et al. 2013). However, when a biological component, such as a cell, is added to the starting formulation, the reproducibility of the solute composition will be reduced. This results from a combination of having both biological variability (e.g., genetic/phenotypic) and solute variability (non-normal distribution in the number of microbes per droplet). This variability means that a higher number of droplets will need to be measured to garner a representative and robust dataset.

Numerous strategies and instrumentation have been developed for decades to trap, suspend and probe individual droplets in a confined space. Of these instruments, only a select few have been further adapted to study bioaerosol . The techniques that have been successfully adapted to trap an individual bioaerosol droplet use either optical (light scattering/absorption) or electrodynamic forces. Note that the force used to confine droplet(s)/particle(s) will have their own intrinsic advantages and/or disadvantages. For example, while both optical and electrodynamic traps can be used to trap both individual and multiple droplets/particles, only electrodynamic levitation can be used to confine and sample a large population (>50 of individual) bioaerosol droplets/particles. A detailed explanation into the utility of various optical and electrodynamic trapping techniques is as follows.

2.6.1. Optical trapping (general)

The ability to trap, manipulate and probe an individual airborne droplet in or around the focal point of a laser beam passed through an objective lens has been used for decades (Hopkins et al. 2004). Aerosol optical tweezers (AOT) can hold and manipulate single, or multiple, nano- and micro- sized droplets for prolonged periods of time, ranging from seconds to hours. Once trapped, the droplet can be subsequently probed, typically by Raman spectroscopy. Analysis of the images of the droplet coupled with their Raman spectra can afford a wealth of information about the droplet including diameter, refractive index, chemical composition, viscosity, and surface tension. Many of these measurements are not possible via any other techniques. In recognition of the many unique features of AOT, despite the potential challenges of using laser light scatter to trap a bioaerosol droplet, in the last few years there has been an influx of optical traps developed to specifically probe bioaerosol, described in the following sections.

There are primarily two distinctly different forces that the various optical trap designs use to trap the droplet: radiative pressure and photophoresis.

2.6.1.1. Radiative pressure trapping

Standard AOT trap individual nebulized droplets with a Gaussian laser beam focused by a microscope objective (Reid and Mitchem 2006). Thus, the position in three-dimensional space that the droplet occupies is inundated with a high luminous flux. This feature, when coupled with the same objective being used to collect the light scattered from the droplet results in AOT being able to collect high quality Raman spectra of the levitated droplet (e.g., an absolute detection limit of 4 pg of substance has been reported; Haddrell et al. 2017).

The AOT is an excellent instrument to probe optically transparent, non-absorbing droplets, such as those commonly studied in atmospheric chemistry. However, many components of bioaerosol are not optically transparent, and as such are not able to be probed with AOT. The gradient forces used in AOT require that the droplet is near or completely spherical to be trapped. This is another byproduct of the high luminous flux being passed through the core of the trap. Consequently, the inability to study non-spherical particles severely limits the utility of the AOT to study bioaerosol. Additionally, the weak forces in AOT limits the size of droplet that can be suspended, ranging between ∼4 µm to ∼14 µm (Hopkins et al. 2004). This covers only a portion of the size regime commonly associated with bioaerosol.

While optical traps have been used for decades to study and manipulate microbes in the suspensions, the limitations (particle size/shape, etc.) of radiative pressure trapping of aerosols have severely inhibited their use studying microbes in the aerosol phase. To address these shortcomings, other optical trap approaches have been taken.

2.6.1.2. Photophoretic trapping

Although they both use a Gaussian laser to build a trap, the forces used to confine a droplet in a photophoretic trap (PT) differs entirely from a conventional AOT. Unlike in the AOT, a photophoretic trap requires the droplet to absorb some of the energy from the laser. The photophoretic force results indirectly from the particle absorbing photons in a spatially non-uniform manner, causing a non-uniform temperature distribution at its surface. The temperature gradients created cause gas molecules to collide with the particle and effectively push it into the center of the trap. The photophoretic force can be up to four or five times stronger than the AOT, meaning much larger droplets can be trapped. Note that absorption of the laser light is critical to trap; to trap larger bioaerosol a wavelength range in the visible spectrum is sufficient while a deep UV wavelength is needed to trap smaller species such as biological molecules (β-NADH) (Redding et al. 2015).

Many different methods to use photophoretic forces to trap an absorbing droplet have been shown (Gahagan and Swartzlander 1996; Liu, Zhang, Fu, et al. 2014; Liu, Zhang, Wei, et al. 2014; Zhang, Prakash, et al. 2011). Experimentally, a photophoretic trap is typically differentiated from an AOT by presence of a pair of axion lens that the Gaussian laser beam through prior to the objective(s). This forms a hollow beam that is used to create the trap; the particle is held in a region of low light intensity near the focal point. The ability to readily trap non-spherical/absorbing particles means that photophoretic traps can effectively deal with the major problems associate with the conventional AOT’s ability to study bioaerosols.

PT was initially developed to study light absorbing species in air (Pan, Hill, and Coleman 2012). PT can capture an absorbing droplet with a diameter anywhere between 6.2 to 41.8 µm, which encompasses the size regime of most bioaerosols. It has been successfully used to levitate various bioaerosol types, including many pollen grains and spores (Redding et al. 2015; Wang et al. 2015). Given that PT requires heating of the droplet/particle, studies using PT to probe thermally sensitive species, such as bacteria, have yet to be reported.

Efforts have been made to establish a “pollen Raman spectra database”(Guedes et al. 2014). The PT has been successful measuring distinguishable Raman scattering signals consistent with previous reports of Raman peaks in the 1600–3400 cm⁻¹range for pollen grains and spores. Additionally, characteristic peaks associated with specific chemical structures, such as CH₃, CH₂ and CH, were found, further demonstrating the utility of using Raman spectroscopy to probe bioaerosols (Wang et al. 2015).

2.6.1.3. Combination (radiative pressure and photophoretic) trapping

In order to be able to trap both transparent and absorbing particles in the same system, a trap that utilizes either radiative pressure or photophoretic forces in the same system has been developed. Termed a universal optical trap (UOT), they have a similar experimental setup as the PT where axicon lenses are used to produce hollow beams. However, in the UOT the beams are not counter propagating, rather a single beam is pointed vertically, forming an optical “funnel” shape where the particle is trapped at the bottom (Redding and Pan 2015). Absorbing droplets/particles are photophoreticly trapped slightly above the focal point of the beam while transparent droplets are radiatively trapped at the focal point.

Bioaerosols are known to contain many organic compounds with strong fluorescence emissions (Pan, Hill, et al. 2014); this large signal will typically overwhelm the weak Raman signal. The stability of the UOT allows for an individual bioaerosol to be held long enough to adequately photo-bleach the particle such that high-quality Raman spectra are able to be collected (Gong et al. 2018).

Using optical trapping as a test system for bioaerosols is still in its early stages. Currently, optical trapping has been adapted quite well to identify and differentiate bioaerosol based on their Raman scattering and fluorescence profile; this, despite many of the potential challenges of using optical trapping to probe bioaerosol: droplet heating affecting microorganisms, particle shape, etc. However, the successful use of optical traps to test microbial health has not yet been demonstrated. Many of the tools developed thus far to probe and classify pollen and fungi may be used in the future to probe pathogenic bacterial and viral activity and monitor microbe health in real time.

2.6.2. Electrodynamic levitation (general)

The ability to trap a particle with net charge, in this case an ion under vacuum, in an electrodynamic field was first reported in 1953 (Paul and Steinwedel 1953). The electrodynamic field operates in the radio frequency range while the trajectory of charged particles is described by the Matheiu equations. Through lowering the frequency that the device traps, it became possible to suspend micron sized particles with net charge at atmospheric pressure (Wuerker, Shelton, and Langmuir 1959). In the ensuing decades, a series of different electrodynamic devices have been designed (Bogan and Agnes 2004; Davies, Haddrell, and Reid 2012; Davis, Buehler, and Ward 1990).

Electrodynamic levitation devices are capable of trapping supermicron sized droplets and particles across a broad size range. Additionally, through rapidly altering the frequency of the electrodynamic field, the electrodynamic balance (EDB) is capable of accounting for massive changes in the mass of an evaporating droplet, such as a 50 µm saline droplet being dried to <3 µm in a second (Haddrell et al. 2014). The benefit of this is that the aerosol source for an electrodynamic levitation device can be a droplet on demand dispenser, where the complete chemical and biological composition of the starting formulation is tailorable. This ability to create and trap individual droplets of known size and composition is critical to controlling many of the variables associated with bioaerosol analysis, such as microbes per droplet.

2.6.2.1. Double ring electrodynamic balance

The double ring EDB uses a pair of ring electrodes to capture a single droplet with net charge injected from a droplet on demand dispenser (Davis, Buehler, and Ward 1990). Once trapped, the bioaerosol can be probed in many ways: the absolute radius can be measured through laser light scatter, the relative mass of the particle can be inferred from the change in the electrodynamic field needed to hold the droplet in the center of the trap, or the chemical composition can be probed through collection of the Raman or fluorescence microscopy (Laucks et al. 2000). Variations of the EDB have been used for over three decades to study a multitude of bioaerosol properties, ranging from mass, vapor pressure, hygroscopicity, and chemical processing.

The EDB has been found to be a robust technique to study bioaerosol, capable of probing properties of bioaerosol. For example, studies using the EDB have shown that pollen grains are an efficient cloud condensation nuclei (Pope 2010). The largest uncertainty in the aggregate of the radiative forcing of the main drivers of climate change is the role of clouds and aerosol (IPCC. 2013). Although the number density of pollen is likely too low to be a major source of cloud condensation nuclei, the role of bioaerosol in climate change should not be overlooked (Frohlich-Nowoisky et al. 2016).

Much like the series of optical traps discussed, the EDB has largely been used to examine the physical properties of bioaerosol (e.g., Pope 2010). While this information is critical to understanding how and why pathogens in aerosol remain infectious in the aerosol phase, the viability of the bioaerosol cannot be measured without first extracting from the EDB.

2.6.2.2. Controlled electrodynamic levitation and extraction of bioaerosol onto a substrate (CELEBS)

The ability to extract bioaerosol injected into an electrodynamic trap, in a controlled and reproducible fashion, for subsequent analysis of microbe viability has been recently reported (Fernandez et al. 2019). CELEBS uses a droplet on demand dispenser to produce a small (quantifiable) population of droplets with known and designed chemical and biological composition. The absolute number of microbes per droplet can be described by a Poisson distribution, wherein if the starting formulation has an initial concentration of 8 x 10⁶ CFU/mL, the majority of the droplets will have between 0 and 2 CFU/droplet (λ = 0.795); typically a starting formulation of ∼10⁸ CFU/mL is used resulting in each droplet/particle having ∼20 CFU/droplet (λ = 20.6). These bioaerosols are then injected into an electrodynamic trap where they can be held at a controlled relative humidity and temperature for time periods as low as 5 seconds to days. Each droplet has <5 fC of net charge in the form of an ion imbalance of sodium to chloride ions; the net charge (e.g., an excess of <32,000 sodium ions in a droplet containing ∼10¹³ sodium and chloride ion pairs) are located exclusively at the surface of the droplet. No adverse effects as a result of this net charge has been observed. After a set period, the electrodynamic field is then manipulated to eject the bioaerosol onto a substrate, typically an agar plate coated with a thin layer of broth. After a few minutes, the droplets/particles are assumed to be fully dissolved in the broth, thus freeing the individual microbes from the droplets/particles. The broth is then spread across the agar evenly until the broth is absorbed by the agar. After an incubation period, the number of viable microbes is quantified. The viability of the microbes as a function of environmental conditions and particle composition is inferred from the reduction of total microbes resulting from the aerosolization process (the number of microbes per droplet is known, as is the number of droplets per levitation).

CELEBS can be considered the next generation rotating drum. The ability to study a population of droplets with near identical size and composition is a major advantage, as is the highly resolved levitation time. Given the small size of the CELEBS (internal volume under 1 L), there is a high degree of control over the environmental conditions the bioaerosol experiences. Thus far, the CELEBS has only been used to probe the viability of Escherichia coli. The viability curves were found to be similar to those collected with the rotating drum.

3. Environmental factors impacting bioaerosol studies

Whether intentionally or inadvertently, the surrounding environment plays a role in the results of experiments in bioaerosol research studies. Historically, humidity, temperature, light, and the chemical composition of the surrounding air have all been demonstrated to have significant effects on bioaerosols (Cox 1987; Mohr 2007; Lighthart and Mohr 2012; Haddrell and Thomas 2017). Therefore regardless of the objective of the study, these parameters should be either controlled or measured during experiments with bioaerosols. Besides a scientific desire to understand these interactions for ambient biological aerosol, it is also important for controlling and subsequently interpreting laboratory results. Laboratory relative humidity can be variable both throughout the year and in different geographic locations. Electrical discharges, such as those from older vacuum pump motors or an overloaded circuit, can generate ozone, and organic vapors or even oil from vacuum pumps may find their way into aerosol experiments where their presence may have unintended or uncalculated effects on the experiment. Ultraviolet and infrared radiation from laboratory sources may also impact a study. The following sections overview known interactions between biological particles and environmental conditions.

3.1. Water vapor and humidity

Water vapor in the air may contribute to differences in the physical (size, shape, spectroscopic), chemical (composition, pH), and biological (activity, viability, culturability, infectivity or ability to assay) properties of bioaerosols (Reponen et al. 1996; Park et al. 2011; Otero-Fernandez et al. 2019; Yang and Marr 2012; Pan, Santarpia, et al. 2014; Ratnesar-Shumate et al. 2015; Mikhailov et al. 2004; Shaman and Kohn 2009; Zhao et al. 2012; Ijaz et al. 1985). The chemical composition of a bioaerosol particles recreated for laboratory studies may differ in composition from those that are naturally occurring. A bioaerosol can consist of a single biological component or can be a mixture of other organic or inorganic materials and the biological component. The presence of these components, and their relative concentrations in the particle will affect the hygroscopic properties of the bioaerosol. The hygroscopic properties, in turn, based on the surrounding humidity, may affect the biological component inside the bioaerosol. For example, consider a microorganism, aerosolized in a buffer solution which contains a significant amount of sodium chloride (Johnson 1999; Lee, Kim, and Kim 2002). This particle, which when primarily atomized, consists of the microorganisms, salts, residual proteins and other components from growth media, and water. Depending on the RH into which the droplet containing the bioaerosol is released, the water component may evaporate either completely or partially, leaving behind a more concentrated solution/suspension of the mixture. The degree to which this will occur will be dependent on whether the humidity is below or above the efflorescence or crystallization point of the multiple component particle. Or conversely, if the local humidity is very high, above the deliquescence point of the mixture, the particle may uptake water changing the local chemistry. If a dry particle is released into the environment or is moving from a dry to a more humid environment it may also uptake water and similar chemical changes may occur (Mikhailov et al. 2004; Seinfeld and Pandis 2016; Verjano and Marr 2018; Marr et al. 2019; Marsh et al. 2019).

The water vapor content of the air may be expressed relative to the saturation vapor pressure, as the relative humidity (RH) saturation ratio (Sr) or water activity, or as the mass of water vapor in a volume of air, or absolute humidity (AH). Humidity has been reported to be correlated with the biological activity of different types of bioaerosols including bacteria, and viruses (Peccia et al. 2001; Hatch and Dimmick 1966; Cox and Goldberg 1972; Hambleton et al. 1983; Yang and Marr 2012; Marr et al. 2019; Donaldson 1972; Noti et al. 2013; Herman et al. 2007; Songer 1967). It is likely that the effect of the amount of water content in the air results in changes in the local chemistry of the particle and therefore affecting the biological component entrained within. The chemical nature of how the bioaerosol is produced for laboratory studies in addition to the moisture content of the air in which the study is being performed needs to be considered because it may play a role in the response of the bioaerosol and the result of the study. For example, the humidity during an experiment may change the susceptibility of bioaerosols to other inactivating factors such as UV light or ozone (Ko, First, and Burge 2000; Tang 2009; Whitney et al. 2003). In studies in which the survival of a bioaerosol is important for understanding transport in the natural environment, for example, transmission of infectious pathogens in health care settings (e.g., Tellier et al. 2019 and Fernstrom and Goldblatt 2013), or atmospheric bioaerosol studies (e.g., Després et al. 2012 and Santarpia et al. 2013) it is important that the RH or AH of the target environment are controlled for or at a minimum measured and reported.

3.2. Temperature

Similar to humidity, temperature has been demonstrated to affect bioaerosols (Theunissen et al. 1993; Erhlich, Miller, and Walker 1970; Cox 1987; Wathes, Howard, and Webster 1986; Brown 1954; Fernstrom and Goldblatt 2013). Temperature also plays a role in the RH, as temperature increases, the saturation vapor pressure increases, thus decreasing RH (Hinds 1999; Seinfeld and Pandis 2016). Temperature and RH have been linked to increasing inactivation of aerosolized bacteria and viruses (Herman et al. 2007; Tang 2009; Theunissen et al. 1993; Wathes, Howard, and Webster 1986). Controlling temperature in aerosol test systems is generally more difficult than controlling humidity. Therefore it is often a parameter that is reported as “room or lab temperature” in indoor studies. However, indoor temperatures can vary greatly by location, type of building, and season (Nguyen, Schwartz, and Dockery 2014; ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers, and Inc.) 2013; Humphreys 1981; Walikewitz et al. 2015). Similar to the case for RH, temperatures should be measured and reported during studies with bioaerosols.

3.3. Electromagnatic raditaion

The amount and type of incident light on bioaerosols may contribute to inactivation of bioaerosols while suspended. Multiple studies have demonstrated inactivation of bioaerosols for both ultra-violet and visible components of the spectrum (Beebe and Pirsch 1958; Webb 1961; Tong and Lighthart 1998; Paez-Rubio and Peccia 2005; Wright and Batley 1969; Walker and Ko 2007; McDevitt et al. 2007; McDevitt, Rudnick, and Radonovich 2012). Test structures for bioaerosols studies may be designed to intentionally recreate a light spectrum of interest, for example, outdoor or indoor conditions, or may be designed to exclude all light for replicating dark conditions. Regardless of the intent, it is important to account for light as a variable in a study if not controlled for specifically. For example, if studies are done in transparent test chambers, indoor light sources may contain portions of the UV spectrum that could potentially be a source of inactivation of the bioaerosols being studied (Kowal, Allen, and Kahan 2017; Blocquet et al. 2018). If studies are to be conducted outdoors, depending on location, time of day and year, the solar spectrum and intensity may vary significantly (Seinfeld and Pandis 2016). Therefore careful measurement of the spectrum should be made to ensure that any differences in the spectrum from one experiment to the next can either be attributed to or dismissed based the type of light present. Recreating a realistic solar spectrum in a laboratory can be both challenging and expensive. However, efforts should be made to replicate the solar spectrum in a manner that can be understood and reproducible to other researchers. The ASTM 173 and 177 solar standards can be used to guide the design for laboratory solar simulators against a standard value (ASTM 2012a, 2012b). Solar simulation systems are commercially available and can be used to recreate solar intensities indoors for controlled experiments with bioaerosols. When attempting to recreate tropospheric levels of solar intensity, light sources that produce significant amounts of UVC light should be avoided, as UVC light does not penetrate the earth’s atmosphere and are damaging to bioaerosols. Different types of light sources can produce very different spectral distributions, and optical filtering can be used to achieve the desired spectral outcome (Codd et al. 2010; Diffey 2002; Berger 1969; Kolberg et al. 2011; Esen, Sağlam, and Oral 2017).

3.4. Chemistry involving trace gas species

Tropospheric trace gas species, both anthropogenic and biogenic, are known to play a significant role in atmospheric aerosol properties, including biological aerosol. Although some gas-phase species may interact directly with biological particles, it is often the products of gas-phase reactions and specifically reactions driven by sunlight that have the most significant affects. These chemical interactions can change not only the viability of the organisms in the biological particle, but other biological properties that can impact measurement of the particle by other modalities including PCR, protein assays and UV-excited auto fluorescence (e.g., Pan, Santarpia, et al. 2014; Ratnesar-Shumate et al. 2015; Kinahan et al. 2019).

In particular, ozone can impact viability, fluorescence signature and PCR signatures of biological organisms (Ratnesar-Shumate et al. 2015; Kinahan et al. 2019). In particular, ozonolysis/hydrolysis reactions have been shown to occur in high humidity environments (Pan, Santarpia, et al. 2014). Ozone has been studied as a disinfection agent for bacteria on surfaces (Fan et al. 2002; Kim and Yousef 2000; Selma et al. 2008), and ozone is also known to impact the fluorescence properties of amino acids used in ultraviolet light induced fluorescence (UV-LIF) measurements in animal tissues (Fujimori 1985; Ignatenko et al. 1982; Ignatenko 1988) and aqueous solutions (Mudd et al. 1969; Fujimori 1985; Ignatenko et al. 1982; Kotiaho et al. 2000). The production of N-formyl kynurinene (NFK), a product of ozonolysis-hydrolysis of tryptophan (Mudd et al. 1969; Ignatenko et al. 1982; Fujimori 1985; Kotiaho et al. 2000) fluoresces at a longer excitation wavelength than tryptophan (Fujimori 1985; Ignatenko et al. 1982). Fluorescence spectral shifts in biological particles due to ozone and humidity exposure have also been observed. There have been numerous recent studies of the impact of ozone and water vapor on biological particles that demonstrate the impact of this ozonolysis-hydrolysis mechanism on the collectively change the spectral properties and viability of airborne bacteria and viruses (Santarpia et al. 2012; Pan, Santarpia, et al. 2014; Ratnesar-Shumate et al. 2015). Differences in UV inactivation of bioaerosols as a function of RH have been reported (Peccia et al. 2001; Ko, First, and Burge 2000; Beebe and Pirsch 1958). Given that these parameters do occur in concert and are present in all studies in which bioaerosols are the subject of experiments, they should be considered in the design of test chambers and systems.

There are limited studies of the impacts of other trace gas species on biological particles (e.g., O₃, radical species and VOCs, and mixtures). Open Air Factor (OAF; e.g., Sawyer et al. 1966 and Dark and Nash 1970) was the name of the process used to describe the observed loss of viability of vegetative bacteria when exposed to ambient air. Research into the phenomena indicated that reactions of olefins and ozone was producing lower vapor pressure species that was believed to be toxic to the bacteria in the same way that ozone and many trace organic compounds react to produce secondary organic aerosol (Seinfeld and Pandis 2016). The photochemistry of urban air pollution has been studied for many decades using smog chambers (e.g., Wang et al. 2014; Paulsen et al. 2005; Cocker, Flagan, and Seinfeld 2001; Carter et al. 2005; Martin-Reviejo and Wirtz 2005; Rollins et al. 2009; Donahue et al. 2012). Pollutants and trace gasses in the urban atmosphere are the results of both complex sources (e.g., industrial, automobile and biogenic) and complex chemical interactions. Ozone can be produced through a variety of mechanisms in the troposphere involving both oxides of nitrogen and organic species. Radical species can be produced through photo-degradation of hydrocarbons and water vapor. Much less work has been done in understanding how this chemistry may impact biological particles. Initial work, examining the changes to E. coli aerosol when exposed to combinations of simulated solar radiation from a mercury halide lamp, relative humidity, laboratory generated ozone, toluene and α-pinene (Kinahan et al. 2019) is the first study that examines the impact of complex organic photochemistry on biological particles. Understanding the role that radical species and other short-lifetime atmospheric species play in these processes may also be important. Other recent research has also indicated that some airborne bacteria may be metabolically active in atmospheric cloud water (Sattler, Puxbaum, and Psenner 2001) and possibly metabolize trace gas species (Smets et al. 2016).

4. Discussion and recommendations

Bioaerosol studies in the laboratory can be accomplished using a variety of techniques. Depending on the type of study and the type of data required for the study, many techniques are available to both contain the biological particles and control the environment to which those particles are exposed. Given the variety of techniques available to perform these studies, standardization of other aspects of each study is critical for producing reproducible data. An often overlooked aspect of standardization, in this field, is the control of the experimental environment. Beside a description of the various systems available for the laboratory study of biological aerosol, this review describes the various ways in which these particles may interact with the experimental environment. Since these interactions may change the outcome of the study by modifying various properties of biological particles, characterization and control of environmental variables during any study of biological aerosol is critical for correct interpretation of the study and for placing it in context with other, similar studies.

Undertaking any controlled study of confined biological aerosol or particles requires an understanding and intentional control of many variables. Any bioaerosol study should adhere to several fundamental principles to insure the most defensible outcome:

1. The physical behavior of the aerosol particles themselves must be separated from the chemical or biological processes that are being studied. Fortunately, the physical behavior of bioaerosols is driven largely by their physical characteristics, rather than their biological characteristics, so, for instance, gravitational settling can be considered separately from biological decay.

2. The environment that the bioaerosol is exposed to must be carefully controlled, monitored and considered in the analysis. The impact of the environment on the biological properties of bioaerosols is still a subject of investigation; however, the studies referenced in this article indicate that many environmental factors have been shown to affect the viability, protein integrity, nucleic acid integrity and physical properties of biological particles. Therefore, every bioaerosol study should attempt to control or quantify the study environment, and those control and measurement points should be described.

3. In addition to the recommendations provided above regarding the test systems and environmental controls, an extensive review of other factors that should be considered in bioaerosol research, including generation, sampling, and measurement are reviewed in the following articles included in this special issue. Many of the themes and concepts already described will appear repeatedly in these discussion pieces and highlight how important these factors are for consideration in future bioaerosol research studies.