A Simple Method for Generating Narrowly-dispersed Bioaerosols in Various Sizes

Biological warfare agents (BWAs) cause disease in humans, animals, and plants when purposefully dispersed in an area. To minimize contamination and personnel exposure and initiate early treatment, effective BWA detection or monitoring techniques are needed. Currently, bioaerosol detection or monitoring techniques are used for detecting BWA; however, these techniques have limitations, such as limited sensitivity. To improve the detection performance and develop novel techniques, an additional step in sample preparation, such as obtaining particles of various sizes, is needed. In this study, we investigated the simple and effective generation of bioaerosol particles using a custom-made inkjet aerosol generator (IJAG). Unlike previous inkjet aerosol generators, the operation conditions of the IJAG are fixed at a nozzle heating temperature of 140 ° C, a driver voltage of 150 V, a pulse width of 60 µ s, and a frequency of 250 Hz. The only controlled factor was the concentration of bioaerosol models, including Bacillus globigii spores, ovalbumin, and polystyrene sphere latex. Our system generated bioaerosols with a diameter of 1–8 µ m and a narrow distribution size. These results suggest that our IJAG system can achieve the simple and versatile generation of narrow-dispersed bioaerosols for a wide range of available materials. Our study can help improve the sensitivity of detection and monitoring systems for BWAs and bioaerosols.


INTRODUCTION
Biological warfare agents (BWAs) are microorganisms, such as viruses, bacteria, fungus, and protozoa, or poisons generated by them that cause disease in humans, animals, or plants when purposefully dispersed in an area (Thavaselvam and Vijayaraghavan, 2010). Many BWAs are stable in air, vary in size from nanometers to microns, and can be easily spread in the air by spraying (Clark and Pazdernik, 2016). Over the decade, some serious illnesses leading cause of numerous human and animal suffering and deaths have been reported such as severe acute respiratory syndrome (SARS), avian influenza, African swine fever, and coronavirus disease (Wang et al., 2021), that were available as BWAs. For the early prediction and treatment and real time warning, rapid and accurate BWAs detection or monitoring system was the goal of various national security programs and researcher involved. Most of the current R&D of BWAs detection and monitoring used the bioaerosol detection and monitoring techniques (Huffman et al., 2020;Greenwood et al., 2009;Švábenská, 2012). Among of bioaerosol techniques, one of the most successful detection and monitoring techniques to date uses matrix-assisted laser desorption time-of-flight mass spectrometry (MALDI-TOF MS) (Jeong et al., 2014;Takahashi et al., 2020;Han et al., 2021;Li et al., 2021b). This technique ensures fast and reliable results in analyzing and identifying BWAs simulants and pathogen biomarkers using an interpretive software package. Many researchers have adopted MALDI-TOF MS as a simple, fast, and reliable method for identifying highly pathogenic organisms (Shaw et al., 2004;Pierce et al., 2007).
Even though MALDI-TOF MS techniques has established as a technology platform for bioaerosol detection with minimal workload, compared to manual analysis methods including antigen/antibody assay, or polymerase chain reaction (PCR), it has a low sensitivity caused no standard methods for bioaerosol sample collection as the upstream of bioaerosol detection and monitoring procedures (Li et al., 2021a). To date, the sedimentation, filtration, centrifugation, impaction and impingement were frequently used for collection of bioaerosol (Franchitti et al., 2020). These methods shown higher collection efficiency with micrometer-scale, but can damage the bioactivity of the collected bioaerosols, resulting in underestimating the risk level. Recently, the microfluidic-based collection methods have shown the advantages with automatic operation, but have small sampling volume, the lower concentration of collected bioaerosols (Wang et al., 2021). Consequentially, bioaerosol sample collection and preparation methods mainly influenced the sensitivity (Mainelis, 2020). Particularly, particle size is essential for the verification and capability enhancement of the detection and monitoring system. The diameter and viability of biological aerosols may be greatly influenced by air temperature and relative humidity (Xie et al., 2021). Therefore, more efficient collection methods were investigated and developed for downstream detection procedure with high sensitivity.
To investigate and develop the efficient bioaerosol collectors and test the performance, bioaerosol generators certainly used (Zhen et al., 2014). Pneumatic and collision nebulization is the most commonly used method in a variety of bioaerosol research and applications. Two nebulization methods were able to produce high concentration of bioaerosol with 0.95-15 µm of particle size, but, due to high-speed jet flow and shear forces, microbial viability can be affected (Thomas et al., 2011;Zhen et al., 2014). To minimize the damage to microbial cells, several new generators have been introduced for bioaerosol research. However new generators including a flow-focusing aerosol generator (FFAG), and C-Flow concentric nebulizer were based on pneumatic technique, accordingly, there was a limit to the minimizing stress (Thomas et al., 2009). Recently, inkjet printing has been used in the bioaerosol generation fields, controlled the bioaerosol formations using piezoelectric actuator such as driving voltage, pulse width, and waveform. This technique has great advantages in viability of microbial cells due to thermally consistent process. Utilized piezoelectric inkjet heads, the polystyrene sphere latex (PSL) and bacteria spore aerosol were generated (Lin et al., 2019). These studies have demonstrated the feasibility of piezoelectric inkjet printing techniques to generated viable bioaerosols. However, in initial step, driving voltage, pulse width, and waveform have to be optimized.
Herein, we investigated generation techniques to control the bioaerosol particle size with a narrow dispersion using a custom-made inkjet aerosol generator (IJAG) with the fixed operation conditions and controlled concentration of bioaerosol models. The IJAG system uses a cartridge containing multiple nozzles to create water droplets through pressure applied to the liquid in the microchamber above the nozzle (Kesavan et al., 2014). The liquid jet leaving the nozzle usually breaks up into single droplets from which one or more satellites may form. The concept of using an inkjet cartridge for near monodisperse aerosol generation was developed to enable testing of bioaerosol detection equipment in an aerosol chamber where the desired bioaerosol concentration can be a few particles per liter. Different bio-aerosol cluster sizes can be produced by adjusting the concentration of cells in the liquid suspension. The IJAG system to generate near monodisperse clusters of Bacillus atrophaeus spores (King and McFarland, 2015). We chose BWA simulants, including B. globigii (BG) spores, ovalbumin protein, and polystyrene sphere latex as various bioaerosol models. Our experimental setup achieved a standardized bioaerosol generation with a narrow size-diameter distribution between 1-and 8-µm by simply adjusting the model concentration.

Design of the Particle Generator and Generation of BWA Simulants
For the size-controlled generation of bioaerosol particles, we designed a custom-made IJAG including a micro-droplet component Microdrop Technologies Gmb,Norderstedt,Germany). The IJAG is equipped with an aerosol generator module, including a piezo-type inkjet nozzle and reservoir, a control unit with a function generator and a CCD camera, a flow splitter, a heater, and a drying module (Fig. 1). With Micro-droplet™, you can control the droplet size and frequency. After confirming and storing the generated droplet through the image of the CCD camera, the size of each droplet is roughly measured. In addition, by controlling the temperature of the supplied nitrogen gas with the heater of the drying module, the liquid phase of the droplets is dried to prepare a pure sample. By manually adjusting the amount of nitrogen gas, the speed of droplets and pure samples can be controlled. The gas is provided as clean nitrogen through a HEPA filter. The flow meter and regulator are regulated and control the amount of gas and the velocity of the gas through a cylinder tube of a given diameter. A heater is installed with a thin heating wire and insulation material on the cylinder tube. The heat generated by supplying electric current to the heating wire heats the tube to heat the passing gas. Nitrogen was passed through the reactor as a carrier gas, and the temperature of the nitrogen gas was raised with a heating coil inside the reactor. Nitrogen gas heated to 200°C formed a sheath flow around the inkjet nozzle and carried the droplets downstream in the drying module.
There are 7 generation parameters, and the user decides it with an experimental value according to the purpose. Among them, in the case of the pulse width, the piezo actuator contracts and an overpressure impulse is generated. The pressure accelerates the liquid by pushing it into the nozzle. With a short pulse width, if a pressure is generated for a short period, the droplet size also becomes very small, and if a long pulse width that becomes saturation is applied, the droplet becomes large. That is, a short pulse is generated at a high voltage. The pulse width also affects the occurrence of satellite drops, so it is adjusted so that it does not occur. The satellite drop depends on the properties of the liquid, and the lower the viscosity, the easier it occurs. Therefore, the higher the voltage (v), the higher the generated pressure, the pulse width (µs) determines the duration of the piezo contraction, and the frequency (Hz) plays a role in adjusting the droplet generation frequency. Through the droplet camera, it is possible to observe the generated droplets in real time on the computer for analysis, and by storing them, the droplet size can be measured in the analysis software.
For the generation of the bioaerosol particles, the IJAG conditions were fixed at a nozzle heating temperature of 140°C, a driver voltage of 150 V, a pulse width of 60 µs, and a frequency of 250 Hz. The corresponding conditions were experimentally obtained and fixed and used. The particle size was changed by changing the concentration. First, in the case of PSL, 8 µm was generated by diluting 1/1000, and the particles of the following size were prepared by diluting 1/2 thereof. For the generation of a BG cluster, spores at a concentration of 1 × 10 6 -1 × 10 9 colony forming units (CFU) per mL in sterile water were adjusted using a piezo-type inkjet nozzle. The OV sample concentration was 0.04-4 mg mL -1 . For PSL, serially diluted samples were applied to the IJAG for aerosol generation. Fully dried monodispersed particles of the BG spores, OV, and PSL clusters with specific sizes were generated and analyzed using an Aerodynamic Particle Sizer spectrometer 3321 (TSI Inc., USA), a representative standard analyzer for measuring particle size distribution in real-time.

Determination of Particle Morphology Using Scanning Electron Microscopy (SEM)
An SEM image analysis was conducted using the ImageJ software. SEM images were obtained using a JSM-7000F microscope (JEOL Ltd., Tokyo, Japan). Information about the chemical composition of the water residues was obtained using an energy-dispersive X-ray spectroscopic (Oxford Instruments INCA X-sight LN2 EDS) analysis with a 10-nA probe current, 15-kV acceleration voltage, and 10.0-mm working distance.
Since the specimen of SEM must be electrically grounded, particles were collected through carbon tape. There is the four-split flow distributor (TSI., USA, Model 3708) used was connected to the rear end of the inkjet particle generator and a 1/4 inch Teflon tube of the same length was connected. Using one 3/8-inch outer diameter inlet and four outer diameters 1/4-inch outlets, it is a device that evenly branches biological particles in four directions. By forming the area of the outlet relatively wide, the particles spread more lubricatingly throughout the entire area inside the port so that the concentration distribution is uniformly formed. In addition, since the number of ports to be used can be selected by blocking unused ports, there is a useful advantage when necessary. One port was connected with APS and the other port was connected with carbon tape to collect particles of the same size during 20 min.

RESULTS AND DISCUSSION
To compare and analyze the results of currently used systems and methods, we first tested the aforementioned two methods (Fig. S1). BG spores, which were used as a simulant for B. anthracis spores, and OV, which is a simulant for staphylococcal enterotoxin B, T-2 mycotoxin, and botulinum toxin, were used. For BG, the inkjet system generated 2-to 3-µm diameter bioaerosols. The inject system mainly generated 2-to 3.5-µm diameter OV bioaerosols. In contrast, mechanically dispersed aerosol particles of BG spores and OV through the air gun had broad-range particle distributions. BG spores and OV aerosols generated by mechanical spraying had particles with a diameter exceeding 10 µm. These results showed that biological aerosols could be generated with a 2-to 4-µm diameter uniform size using an inkjet system. However, BWAs are engineered to contain particles with an aerodynamic diameter between 1 and 10 µm (Primmerman, 2000). Thus, to improve the performance of the BWA detection system, the need for an additional step in sample preparation, such as the collection by size and the requirement of obtaining particles with various sizes having a narrow distribution, was addressed.
For this purpose, we developed a custom-made IJAG, including a piezo-type inkjet nozzle, as shown in Fig. 1. To generate narrowly dispersed bioaerosols, serially diluted samples of BG spores, OV, and PSL samples were applied to the sample reservoir in the IJAG. We successfully controlled the various specific sizes of bioaerosol particles generated with our experimental setup using the IJAG (Fig. 2). Polydisperse aerosols have wider size distributions and higher geometric standard deviation (GSD) values. The geometric standard deviation describes how spread out the values are in the distribution. If the GSD is less than or equal to 1.25 µm, it means generally monodisperse. Since biological particles are very difficult to generate monodisperse, the term narrow distribution has been substituted. Most of the particles were generated below GSD 1.25, but particles of GSD 1.25 or higher were generated in the large range of particles. Corresponds to narrow distribution. The GSD of the PSL 1, 3, 5, and 8 µm were 1.01, 1.07, 1.31, and 2.00 µm, respectively. The GSD of the BG spore bioaerosols consisted of 2, 3, 4, and 5 µm diameter particles were 1.17, 1.36, 1.10, and 1.57 µm, respectively. Finally, the GSD of the OV bioaerosols consisted of 3, 5, and 7 µm diameter particles were 1.17, 1.03, and 1.12 µm, respectively. It can be confirmed that most of the particles generated through our system are generated monodisperse, and the particles are generated narrowly. Volume 23 | Issue 4 | 220218 The PSL aerosols as the control chemical consisted of particles with a diameter of 1, 3, 5, and 8 µm; BG spore bioaerosols consisted of 2, 3, 4, and 5 µm diameter particles; and OV bioaerosols consisted of 3, 5, and 7-µm diameter particles. The 5-µm-diameter PSL particles were generated from a sample concentrated to one drop of PSL suspended in 1 mL of sterile water. PSL particles with 4, 3, and 2 µm diameters were generated from serially diluted PSL samples. For BG spores, 2, 3, 4, and 5 µm diameter clusters were generated from 1 × 10 9 , 1 × 10 8 , 1 × 10 7 , and 1 × 10 6 CFU mL −1 samples, respectively. Moreover, OV particles of 3, 4, and 5 µm diameter were generated from 0.04, 0.4, and 4.0 mg mL −1 , respectively. In addition, we confirmed the different sizes and shapes of PSL, BG, and OV clusters using SEM images (Fig. 3).
Methods for generating general particles are largely divided according to the generation of monodisperse/polydisperse, non-spherical/spherical, and solid/liquid particles. The particle size can be generated in the range of 0.01-100 µm, and the range varies depending on the generation technique. First, there is spraying a suspension solution such as PSL to generate monodisperse particles, which may form agglomerates or cause residues of solvents that are not particles to be generated as particles. Therefore, vibrating orifice related aerosol generators (VOAG) are also used to obtain stable monodisperse particles. This device is advantageous for generating large particles of 0.5-50 µm, and as the diameter increases, particle loss tends to be large. In addition, more than four formulas are used to control the particle size, and the constant value varies depending on the feed material used, making it difficult to calculate easily (Chen et al., 2011). There is a disadvantage that the orifice must be kept clean for continuous occurrence. Spinning top/disc aerosol generators that can compensate for the shortcomings of the device are also used. It has the advantage that the generator is not blocked by the feed solution and the disc is not heavily coated. Electrospray Technique is also used to generate particles as small as 0.01-1 µm (Simpkins, 1997). As the aerosol passes through the bipolar charger, it becomes neutral, and along the electric field of the negatively charged rod inside the DMA, the positive charge goes to the negatively charged outlet, and is generated according to the electric mobility and size. The above-described particle generation techniques have a disadvantage that the generation conditions must be optimized initially, and a number of variables must be considered in order to generate particles of a desired size (Mitchell, 2020).
IJAG is frequently used as a monodisperse particle generation technique and is effective for repetitive generation. Bottiger and his colleagues firstly developed the IJAG system for creating the monodisperse aerosol particle, which applied electrical pulse. This IJAG concept break down the sample to single droplet with satellites, and primary droplet and satellites were separated to generate the monodisperse aerosol particles by aerodynamic fractionation using counter flow and installed a camera for real time. The particles from about 5 µm diameter can be produced (Kesavan et al., 2014). Using the same Bottiger's IJAG concept, King and McFarland (2015) generated the near-monodispersed particles of B. atrophaeus spore with 1.8-8.7 µm by adjusting the concentration of cells in hydrosol. However, related studies have shown that mechanic installations are needed to control bioaerosol particles. Our custom-made IJAG concept doesn't need the separation step, and fixed by a mechanical installation can simply and effectively generate bioaerosol particles of BWA surrogates.
BG spores were used to simulate B. anthracis and other bacterial species, and OV protein was used to simulate protein toxins such as staphylococcal enterotoxin B, T-2 mycotoxin, or botulinum toxin. Powdered anthrax spores were purposefully inserted into letters mailed through the US postal system in 2001, which infected 22 people who subsequently died. Staphylococcal enterotoxin B, T-2 mycotoxin, and botulinum toxin have been used as biological weapons. These toxins act by stimulating cytokine release and inflammation or by inhibiting protein and DNA synthesis. In many cases, there are difficulties, as robust protective equipment and facilities are required to handle BWA, and the generation of aerosols using various biological components has not been investigated. Therefore, experimental results using simulants like those in our study can be directly used to analyze BWAs.
Within the aerodynamics in the human respiratory tract, aerosols with particle sizes < 10 µm differentially deposit in the upper respiratory tract, and those with sites < 3 to 5 µm penetrate deep into the lung. In 2017, about 545 million people worldwide have a chronic respiratory condition, a 39.8% rise since 1990. Recently, the coronavirus disease pandemic has also been shown to widely spread via bioaerosols. Accordingly, the researches of bioaerosol generation, collection, and monitoring were focused in clinical microbiological area. However, the universal system for generation of bioaerosol was difficult to be suggested because original characters such as size, shape, and arrangement of microorganism, virus or biological toxin were diverse. In our system, we intended to measure BWA simulants, including BG spores, ovalbumin protein, and polystyrene sphere latex as various bioaerosol models through IJAG and use it universally. We achieved a narrow dispersion simply by changing the concentration of the material. Thus, bioaerosol generation with narrow sizes and various materials will enable us to analyze the relationship between the size of BWAs and the mechanism, or monitoring system.

CONCLUSIONS
In this study, we investigated an experimental setup to effectively control the size of bioaerosol particles using a microorganism and a biological toxin. This IJAG setup can be used for biological materials, including simulants of BWAs or viruses. Therefore, it will be possible to use it extensively for various biomaterials. Bioaerosol particle size was controlled within a diameter of 1-8 µm, and our method was superior in uniformly manufacturing various aerosol sizes as compared to conventional methods. In the evaluation of BWAs, the method using an aerosol chamber has difficulties in manufacturing and classifying a certain range of biological particle samples, and the method using an inkjet particle generator has a long culprit in terms of size control and reproducibility of biological particles. We secured convenience and reproducibility by minimizing the cumbersome work of changing the voltage (v), the higher the generated pressure, the pulse width (µs), and the frequency (Hz), which are physical conditions, and controlling the concentration of the solution. Thus, our results may help improve detection and monitoring systems for analyzing the causes and BWAs.