Detection of Suspended SARS-CoV-2 in Indoor Air Using an Electrostatic Sampler


 A prototype virus sampler using electrostatic precipitation has been developed to investigate aerosol infection by SARS-CoV-2. The sampler consists of a discharge electrode placed inside a vial, and a thin layer of viral lysis buffer at the bottom, working as a collection electrode. The sampler was operated with the sampling air flow rate of 40 L/min. Collection efficiency of the sampler is about 80% for 25nm to 5.0µm diameter particles. We sampled the air of a food court of a commercial facility, a connecting corridor of a clouded train station, and two office rooms (A and B) in September 2021, just after the 5th peak of COVID-19 in Japan. The analysis using a RT-qPCR detected the virus RNA in the air of the office A, B and the food court. Estimated concentration of the virus in the air determined by calibration curve was 2.0 x 102, 7.8 x 102, and 0.6 - 2.4 x 102 copies/m3, in the office A, B, and the food court, respectively. These results indicate that the sampler using electrostatic precipitation can detect SARS-CoV-2 in indoor air. It could be developed as a risk assessment method for aerosol infection.


Introduction
Although droplet and contact infections have been thought to be the main routes of SARS-CoV-2 transmission, aerosol infections have also been found to be an important route 1,2,3,4,5 . The Centers for Disease Control and Prevention report in May 2021 also pointed out the aerosol transmission 6 . Aerosols are particles emitted by coughing, sneezing, or breathing, then drying up in the air to become smaller and oating for long time. To prevent aerosol infection, ventilation (complete air exchange) is recommended.
However, in many commercial facilities and restaurants, it is di cult to provide su cient ventilation due to weather conditions and structural problem. Monitoring SARS-CoV-2 oating in indoor air is important to know the degree of contamination for countermeasures against aerosol infection. To date, airborne sampling results have been reported from medical facilities where patients with SARS-CoV-2 are admitted 7,8 . In addition, there are examples of measurements in general environments such as commercial facilities, o ces, and buses 9 , but due to the limitations of the samplers, long sampling times are required. Midget impingers and gelatin lters are commonly used as bioaerosol samplers 10,11 . The gelatin lter requires elution of the collected virus, and the midget impinger has low collection e ciency.
Recently, samplers using electrostatic precipitator 12 have been reported 13,14,15,16 . Electrostatic precipitator has advantages such as low pressure drop, high collection e ciency for ne particles, and collection on liquid, and can be designed to suit the application.
We have developed a prototype sampler that incorporates electrostatic precipitation into a vial and collects the particles in a liquid. Using a viral lysis buffer as the collection electrode, we were able to reduce the risk of infection to workers. Using the prototype electrostatic sampler, sampling was conducted in the third week of September 2021, just after the peak of the fth wave of COVID-19 in Japan, at a food court in a commercial facility, a station passageway, and two o ce locations (A and B).
The virus RNA was detected in the air of the food court and the two o ces.

Results And Discussion
Collection e ciency of the electrostatic samplers Collection e ciency of the electrostatic sampler was measured using a Fast Mobility Particle Sizer for particles less than 0.3 µm in diameter and an optical particle counter for particles larger than 0.3 µm. The sample gas of incense smoke diluted with air was used to contain a wide range of particles from 5.6 nm to 5.0 µm, with the most frequent particle size being 60 nm. The collection e ciency of the sampler was 88% at 25 nm, 81% at 50 nm, 81% at 100 nm, 83% at 200 nm, 80% at 300 nm, 87% at 500 nm, and 100% at 5 µm.
Possible damage to nucleic acids by corona discharge Corona discharges produce reactive species such as NO and O. These reactive species can denature the nucleic acids in the collected microorganisms and viruses 17,18,19,20 and affect the analysis of PCR. To con rm the effect on nucleic acids, bacteriophage MS2 was added to the AVL buffer, then, exposed to the discharge, and analyzed using RT-qPCR. The comparison was made using the same condition of the discharge and the buffer for conducting the sampling. Figure 1 shows the results. Each test was performed twice. Bacteriophage MS2 exposed to the discharge for 90 min. was detected at 21.89 and 22.19 cycles by the PCR, while untreated bacteriophage MS2 was detected at 22.46 and 22.40 cycles.
These results indicate that the effect of the exposure to the discharge on detection of RNA of collected MS2 by running the sampler for 90 minutes is negligible.

Sampling of indoor air
The sampling was conducted at a food court of a commercial facility, a passage of a clouded train station, and two different o ces (A and B) in the third week of September 2021, just after the peak of the fth wave of COVID-19 in Japan. The sampled specimens were analyzed by RT-qPCR. Figure 2a shows the ampli cation curve of the RT-qPCR targeting the N2 region, and Figure 2b shows the calibration curve using the step dilution series of the positive control of the N2 region. The results are summarized in Table   1. In the o ce A, the virus RNA was detected in all ampli ed regions. In the o ce B, the virus was detected in two different ampli ed regions. In the food court of a commercial facility, the detection was made in one of two conditions. In the connecting passage of the train station, the virus was not detected.
In the station, everyone wore masks and moved in silence, therefore, aerosol was not generated easily. On the other hand, in food courts and o ces, most of the people did not wear masks, and generation of aerosol from human was easily took place.
The concentration of the virus was calculated based on the calibration curve in the N2 region of the realtime PCR. The o ce B had the highest airborne concentration of 7.8 x 10 2 copies/m 3 , followed by O ce A with 2.0 x 10 2 copies/m 3 , and the food court with 0.6 -1.9 x 10 2 copies/m 3 . These concentrations are higher than the results from the hospital shown in the literature 9 . Concentration measurements from the number of PCR cycles are not very accurate, and possibly these values are not reliable. Or, possibly infected persons were present at the time of sampling. The concentrations of suspended viruses were calculated using the copy number obtained based on the calibration curve, the aspiration volume, and the collection e ciency of 0.1µm particles.

Conclusions
Page 5/11 The indoor air was sampled using a prototype electrostatic sampler. As a result, the following conclusions were obtained (1) The collection e ciency of the sampler was about 80% for particles from 25 nm to 5.0 µm under the conditions of -6 kV, -80 µA, and sampling air ow rate of 40 L/min.
(2) Bacteriophage MS2 in Buffer AVL was exposed to the discharge for 1.5 hours under the same conditions as the sampling, and the effect of discharge on the detection by RT-PCR was proved to be negligible.
(3) In a food court, and o ce, SARS-CoV-2 RNA oating in the air was detected using the electrostatic sampler and RT-qPCR.
Air sampling using electrostatic precipitation could be developed as a risk assessment method for aerosol infection.

Structure of the electrostatic samplers
Electrostatic precipitation is a method of charging and collecting suspended particles by corona discharge. The smaller the radius of curvature of an electrode, the lower the voltage required to generate corona discharge. A bundle of 100 metallic bers with 12 µm diameter was used as the discharge electrode 21 . About 2 mL of Buffer AVL (QIAGEN, Germany) supplied in the vial, which was used as the collection electrode. The buffer is a viral lysis solution used for purifying viral nucleic acids. Using the buffer as the collection electrode, we were able to reduce the risk of infection to workers. Figure 3 shows the schematic diagram of the sampler. A metal tube with 20 mm diameter was inserted into the center of a vial with 40 mm diameter. 11 bundles for corona discharge were attached to the tip of the tube. The gap between the discharge electrode and the surface of the buffer AVL was xed to 10 mm. The applied voltage was DC-6 kV and the discharge current was -80 µA. The sampling air ow rate was 40 L/min. The air owing in from the central metal tube was turned back in the vial and exhausted to the surrounding area. Meanwhile, the charged aerosol is collected into the buffer by electrostatic force.

Collection e ciency of the electrostatic samplers
Incense smoke was mixed with air to form sample gas and used in a room with a volume of 100 m 3 . This was conducted to increase the number of airborne particles with 300 nm to 5000 nm of diameter in the indoor air. The sample gas was passed through the electrostatic sampler to measure the collection e ciency. A Fast Mobility Particle Sizer (FMPS; TSI, USA) and an optical particle counter Model 3888 (Kanomax, Japan) were used. FMPS can measure particles from 5.6 nm to 560 nm, and the particle counter can measure particles from 0.3 µm to 5.0 µm. The collection e ciency was calculated using the ratio of particle-count with the high voltage (HV) turned ON and OFF.

Possibility of damage to nucleic acids by corona discharge
To con rm the effect on nucleic acids, bacteriophage MS2 was exposed to the discharge, and analyzed Two types of prototype devices (A and B) were used in the demonstration test. Device A consists of one sampler with the sampling air ow rate of 40 L/min, and device B consists of three samplers in parallel with the sampling air ow rate of 120 L/min. The size of both devices was 300 mm (length), 300 mm (width), and 150 mm (height), and they were powered by a battery with 12 V output. The buffers for each sampler were 2.9 mL (1.12 mL of Buffer AVL and 1.78 mL of DW A calibration curve was prepared using the positive control for the N region. For the ORF region, a calibration curve could not be drawn due to the low concentration of the commercial positive control. The positive control was adjusted by 10-fold dilution from 1.0 x 10 0 to 1.0 x 10 3 copies/µL, and calibration curves were drawn for the N 1 and N 2 regions using the PCR conditions described above. The RT-qPCR data were quantitatively analyzed using the calibration curve and the suspended concentration per 1 m 3 of aspirated air was calculated.