2. Materials and Methods

2.1. Media and Sample Preparation
The modified protocol by Bonilla et al. [19] was used for media and sample preparation. The collected PRD1 bacteriophage samples and the Salmonella enterica Serovar Typhimurium LT2 (RD1) host (courtesy of Carlos Gonzalez at the Department of Plant Pathology and Microbiology at Texas A&M University, College Station, TX, USA) were both pipetted into soft agar overlay and poured onto prepared Luria Bertani (LB) agar plates. This process was repeated for all the collected samples as well as stock solution dilutions, nebulizer solution dilution, and the test room exhaust filter sample. The plates were placed in the 37 °C incubator over night to grow and the plaque forming units were counted the next day.
To prepare the Luria Bertani (LB) culture medium, 10 g of peptone, 5 g of yeast extract, and 5 g of NaCl were added per liter of MQ water and autoclaved. All media compounds were from Becton, Dickinson and Company (BD, Franklin Lakes, NJ, USA). The soft agar medium was prepared by adding 25 g of LB broth powder and 7.5 g of agar per liter of MQ water. The soft agar was prepared in 50 mL batches and autoclaved.
The bacteriophage host Salmonella enterica Serovar Typhimurium LT2 (RD1) bacterium was grown in LB culture medium containing 100 mg/L Ampicillin, added after autoclaving, to select for cells with pRD1 as the PRD1 phage infects cells that harbor a conjugative plasmid [20]. The SM (Sodium-Magnesium) phage dilutant buffer for phage isolation and propagation with a pH of 7.4 was prepared by adding 5.8 g of NaCl, 2.0 g of MgSO4·7 H2O, and 50 mL of 1 M Tris-HCl per 1 L of MQ water, autoclaved, 0.2 μm filter sterilized and stored at room temperature. Calcium chloride (CaCl2) solution was prepared as a filter-sterilized 1 M stock solution and added to the LB broth to a final concentration of 0.001 M. The overlay was prepared by adding 0.1 mL of fresh mid-log phase (OD600 = 1.0) S. enterica host culture incubated at 37 °C, and 0.1 mL of phage (102–103 plaque forming units (PFU)/mL) to 4 mL of soft agar containing 100 mg/L Ampicillin at 56 °C in a 5 mL tube, poured over an LB agar plate and incubated overnight at 37 °C.

Phage Isolation and Propagation via Plate Lysate
After incubating the LB agar plate with the bacteria to form plaques at 37 °C, the plate was lysed by pouring 5 mL of SM buffer on top of the plate and shaking, gently, for 15 min. The phage suspension in the buffer was collected and centrifuged at 4000× g for 5 min to remove cell debris. The phage lysate was filter sterilized and stored at 4 °C.

2.2. Plexiglass Model Chamber Test
In a controlled plexiglass model chamber at Biosafety Level 2, a Collison 6-jet nebulizer (BGI, Waltham, MA, USA) was used to disseminate known numbers of bovine coronavirus (BCoV) aerosols. The laboratory plexiglass chamber experiments were used to simulate SARS-CoV-2 in controlled environmental conditions and mitigate uncertainty of testing in the open environment of the hospital model room. BCoV (bovine coronavirus, Mebus, Catalog No. NR-445, https://www.beiresources.org/, accessed on 11 March 2022), an enveloped, positive-sense, single-stranded RNA betacoronavirus, served as the simulant virus for the SARS-CoV-2 coronavirus throughout this study which allowed for the generalization of results to SARS-CoV-2. A series of experiments were performed at different temperatures and relative humidity values to assess the effect of environmental factors on particle deposition. At regular time intervals after aerosolization, the cabinet surfaces were swiped to quantify the number of impacted viral particles using quantitative polymerase chain reaction (qPCR). The total virus concentrations that became resuspended in the air were collected with the wetted wall cyclone (WWC) bioaerosol collectors including stainless steel and 3D printed cyclone units and the MD8 Airport Sampler (Sartorius, Goettingen, Germany) as reference. The WWC collected bioaerosols at 100 L/min which continuously collected particles into MQ water by concentrating virus particles from the air in the cyclone into the exhaust in the MQ water. The virus samples were quantitated with qPCR.

2.2.1. Plexiglass Model Chamber Set-Up
A controlled temperature–humidity BSL-2 chamber (0.91 m × 0.91 m × 0.61 m) was built of clear Plexi-glass Polypropylene with 3 mm thickness. A Collison 6-jet nebulizer (BGI, Waltham, MA, USA) was used to disseminate known numbers of BCoV aerosols at different relative humidity (30, 60 and 80%) and temperature (10, 15, 25 °C) values monitored by probes. The range of temperature and relative humidity values was based on the measurements the authors conducted in meat processing facilities in preliminary studies. A fan with a HEPA filter (high-efficiency particulate air (filter); to provide particle free background air) and a flow straightener were connected to the inlet to maintain the airflow at approximately 0.1 m/s in the chamber, corresponding to the average indoor air velocity [21]. An additional HEPA filter was inserted at the chamber exit. HOBO® dataloggers (Onset, Bourne, MA, USA) and hot wire anemometer (TSI Inc., Shoreview, MN, USA) were used for temperature, relative humidity, and air velocity measurements. HOBO dataloggers are wireless loggers with built-in sensors that record temperature and relative humidity of the surrounding environment during its operation period, in which data can be downloaded through HOBOware program for further analysis. Hot wire anemometers measure air velocity through heat transfer from an electrically heated wire and display the reading on its integrated screen in real time.
The number of viral particles impacted on the walls and metal, plastic, and wood surface samples (2 in. × 1 in.) at the bottom of the chamber were quantitated to assess the effect of temperature and humidity on particle deposition. The virus aerosols that remained resuspended in the chamber were collected using the WWC, the 3D printed WWC, both operated at 100 L/min, and the MD8 Sartorius Airport Sampler, operated at 50 L/min, using 80 mm gelatin filters as reference. Each test was performed three times.

2.2.2. Virus Quantitation by qPCR
For total viral counts for all samples, whole cell qPCR (quantitative Polymerase Chain Reaction) was performed using BCoV specific oligonucleotides and probes [22,23]. The BCoV sample was added to the qPCR SYBR Green reaction mixture (Applied Biosystems, Warrington, UK), and amplified in an automated thermocycler/analyzer (AB StepOne RT-PCR System, AB, Foster City, CA, USA) as described previously [24]. Broadly reactive primer pairs IN-2 (+) 5′ GGGTTGGGACTATCCTAAGTGTGA 3′ and IN-4 (–) 5′ TAACACACAACICCATCATCA 3′ were used to identify the bovine coronavirus (GenBank Access NC_003045). Dilutions of a plasmid containing the BCoV target sequence were used to create a standard calibration curve for each analysis and allow for the calculation of total viral gene copy number [25].

2.2.3. Surface Virus Assessment
Three different surfaces were placed inside the chamber: wood, metal, and plastic that are commonly found, high-touch surfaces in hospital rooms. Figure 1 shows the chemical compound of the plexiglass chamber, polymethyl methacrylate (PMMA) [26]. The wood surface has a varnish finish made of polyurethane (PU), Figure 2 [27]. The metal surface used was aluminum. Polypropylene, (PP), its chemical compound shown in Figure 3 [28], was the plastic surface used in the chamber.
Bio-layer interferometry (BLI) was performed on the surface swab samples to determine the kinetics of the samples collected for each surface under the different environmental conditions. BLI is a technique used to measure the micromolar interactions through white light interference caused by the sample attaching to the biosensor surface. The BLItz system (ForteBio, Fremont, CA, USA) was used for the testing with Aminopropylsilane (APS) biosensors (ForteBio, Fremont, CA, USA). It is important to understand the hydrophobic interaction between BCoV and the hydrophobic APS biosensor because the S protein of SARS-CoV-2 binds to the angiotensin-converting enzyme 2 (ACE2) receptors of its host cells. This binding interaction occurs in a hydrophobic region. The BLI results will show how the S protein would bind to human cell receptors after coming into contact with these 3 surfaces in a hospital [29,30]. Each biosensor was hydrated in Phosphate Buffer Saline (PBS) buffer (pH 7.4) for 10 min before use. The tests had 3 steps, baseline, association, and dissociation. The baseline step was performed with 4 μL of PBS buffer for 30 s. The association step used 4 μL of sample for 300 s to allow the sample to associate to the surface and measure the association of the sample to the biosensor surface. Dissociation took 300 s with 4 μL of PBS buffer to allow the sample to dissociate from the surface and measure the sample dissociation from the biosensor surface. The BLItz system was used to generate binding curves with local modeling using a designated reference. Based on the curve, the association and dissociation rates and kinetics constants were determined for each surface swab sample.

2.3. Model Hospital Room Test

2.3.1. Model Hospital Room and Equipment Setup
The Texas A&M University Biosafety Level 2, air-conditioned 16.14 m3, ¾ scale physical particle test room operated at an air exchange rate of 6 air changes per hour (ACH) was utilized to test PRD1 phages as SARS-CoV-2 simulants. Twelve PM2.5 filter samplers at 15 L/min air flow were used to collect the aerosolized virus (Figure 4 and Figure 5) [31]. Tests were conducted using three different ventilation configurations: (a) with the air inlet in the ceiling on the left, exhaust on the right; (b) with the air inlet in the ceiling above the bed, exhaust at the bottom left on the wall; and (c) with the air inlet in the ceiling above the bed, exhaust at the bottom left on the wall, air curtain entry/door (Figure 5).
During sampling, the ventilation system was operated constantly for each configuration and the environmental conditions (temperature, relative humidity, air velocity) were monitored. The virus particles were injected into the chamber with a Collison 24-jet atomizer at a location where the head of the patient lying on the bed in the room would be typically located. A 50 mL aliquot of the phage sample was added to 300 mL of SM buffer and nebulized as a SARS-CoV-2 simulant for 5 min aerosolization periods, resulting in a concentration of approximately 3.16 × 109 PFU/m3 in the hospital room. The particle size distribution and mass concentration of the virus aerosols generated by the Collison nebulizer at 30.48 cm (head of the bed), 60.96 cm (middle of the bed), and 91.44 cm (foot of the bed) distance from the nebulizer were measured with the Aerodynamic Particle Sizer (APS, Model 3321, TSI Inc., Burnsville, MN, USA) at a sampling flow rate of 1 L/min. Twelve PM2.5 samplers with HTTP 0.4 µm membrane filters (Whatman, Nucleopore, GE Healthcare, Buckinghamshire, Amersham, UK) were placed in the model room to sample air at 15 L/min inflow rate. The HEPA filter at the air exhaust was also collected and analyzed after each test. Sampling was conducted at room temperature during the entire period of 5 min nebulization. Each test was repeated 3 times.

2.3.2. Sample Assessment
For preliminary analysis of viable virus collection, each of the 12 filters collected in each test was placed in 2 mL of PBS. Aliquots of 4 mL of soft agar medium were added to glass tubes with 100 µg/mL Ampicillin. Aliquots of 100 μL of each collected sample in 10x dilution and 100 μL of the fresh mid-log phase S. enterica LT2 RD1 host were added to the medium and poured over a plate. A 10−6 dilution of either the nebulized liquid or the stock solution 100 μL was added to the soft agar medium with 100 μL of the Salmonella host and poured over an LB plate. The same process was performed with the resuspended exhaust filter. The plates were placed in the incubator overnight at 37 °C and the plaque forming units (PFUs) were counted the next morning. Total PRD1 phage numbers were quantitated by qPCR using species specific primers and ABI qPCR chemistry as described in Section 2.2.

2.3.3. Computational Fluid Dynamics (CFD) Modeling and Validation
HOBOware dataloggers were used for the continuous recording of temperature and relative humidity data at the bioaerosol collection sites. The airflow was measured within every cubic meter of the chamber by anemometry to validate the flow model [32,33]. The HOBO units were connected to the 12 PM 2.5 samplers and started when the testing began. The data from the HOBOware Pro 3.7.22 were collected, exported, and analyzed in Excel worksheets. The SolidWorks® 2020 SP5 program was used to create a model of the hospital model room, bed, visitor chair, and monitor based on the dimensions provided by the mechanical blueprint of the chamber. A patient was modeled as a 51.75 inch-long and 9.67 inch-wide rectangular prism based off the average height and waist circumference of a male in the U.S. [34]. The model was imported into ANSYS® Fluent 2020 R2 software to create a detailed mesh. Each model and mesh were adjusted based on the ventilation configuration being modeled. The mesh sized used for each simulation was 700,000 elements. The standard k-ω turbulence model based on the Navier–Stokes equations was used to model unsteady room ventilation in steady flow conditions. Airflow through the space was simulated to determine virus movement in the room. For CFD simulation, in all ventilation configurations, the vertical walls, floor, monitor surface, hospital bed, chair, and air deflecting cones of the air diffuser were considered adiabatic. The transport equations, Equations (1) and (2), used by ANSYS are shown below [35].
Equation (1). Turbulence kinetic energy (k) (1) ∂∂tρk+∂∂xiρkui=∂∂xjΓk∂k∂xi+Gk−Yk+Sk
Equation (2). Specific dissipation rate (ω) (2) ∂∂tρω+∂∂xiρωui=∂∂xjΓω∂ω∂xj+Gω−Yω+Sω
The initial conditions used for the model are shown in Table 1. Velocity and temperature conditions were based on measurements taken during experimental testing. The air exchange rate for all models was 6 air changes per hour (ACH) and the door was closed in all configurations.

2.4. Statistical Analysis
Statistical analysis was performed with Microsoft Excel (version 6.54) on all experimental data using ANOVA with a 95% confidence interval to determine if there was significant difference between replicates and experimental conditions for the plexiglass chamber and between replicates and ventilation configurations for the hospital model room.