Inactivation characteristics of a 280 nm Deep-UV irradiation dose on aerosolized SARS-CoV-2

Abstract

A non-filter virus inactivation unit was developed that can control the irradiation dose of aerosolized viruses by controlling the lighting pattern of a 280 nm deep-UV (DUV)-LED and the air flowrate. In this study, the inactivation properties of aerosolized SARS-CoV-2 were quantitatively evaluated by controlling the irradiation dose to the virus inside the inactivation unit. The RNA concentration of SARS-CoV-2 remained constant when the total irradiation dose of DUV irradiation to the virus exceeded 16.5 mJ/cm². This observation suggests that RNA damage may occur in regions below the detection threshold of RT-qPCR assay. However, when the total irradiation dose was less than 16.5 mJ/cm², the RNA concentration monotonically increased with a decreasing LED irradiation dose. However, the nucleocapsid protein concentration of SARS-CoV-2 was not predominantly dependent on the LED irradiation dose. The plaque assay showed that 99.16% of the virus was inactivated at 8.1 mJ/cm² of irradiation, and no virus was detected at 12.2 mJ/cm² of irradiation, resulting in a 99.89% virus inactivation rate. Thus, an irradiation dose of 23% of the maximal irradiation capacity of the virus inactivation unit can activate more than 99% of SARS-CoV-2. These findings are expected to enhance versatility in various applications. The downsizing achieved in our study renders the technology apt for installation in narrow spaces, while the enhanced flowrates establish its viability for implementation in larger facilities.

1. Introduction

The COVID-19 pandemic has remarkably affected daily life. Several countries have implemented measures to prevent the spread of SARS-CoV-2 (Leung et al., 2023, Akter et al., 2022, Srivastava et al., 2019, Wu et al., 2020, Ye et al., 2020). In particular, restricting human activity is one of the major measures implemented globally (Doung-Ngern et al., 2020, Richter et al., 2021, Guzzetta et al., 2021). The lockdown (Wurtzer, et al., 2020, Redlberger-Fritz et al., 2021) was temporarily effective (Hyafil and Moriña, 2022, Tobías, 2020, Cao et al., 2020, Plümper and Neumayer, 2022). However, when restrictions on human activities are removed, the infection started to spread again. As a result, COVID-19 is yet to be completely eradicated. Furthermore, long-term restrictions on human activity are discouraged as they slow down economic activity. Many people hope to return to their pre-pandemic lives, which requires the establishment of technologies to control infections without restricting actions.

The key to controlling viral infections is to reduce the virus concentration in living spaces (Tang et al., 2020). In crowded spaces (e.g., classrooms [Foster and Kinzel, 2021, Narayanan and Yang, 2021, Zimmerman et al., 2021], offices, hospital waiting rooms [Arjmandi et al., 2022], and trains [Woodward et al., 2021]), the virus present in the human breath fills the spaces and increases the risk of infection. To reduce the infection risk, several reports indicate the importance of outdoor ventilation with open windows or air-conditioned ventilation to let in outdoor air (Morawska and Milton, 2020, Park et al., 2019, Elsaid et al., 2021). However, outdoor ventilation significantly changes the indoor environment in summer and winter owing to differences in indoor and outdoor temperatures. In addition, spaces where outdoor ventilation is difficult, such as hospitals (examination rooms and operating rooms) and commercial facilities (movie theaters and theme parks), rely on internal ventilation systems for indoor circulation (Park et al., 2020). Filters with high collection performance, including MERV, HEPA, and ULPA filters are often used in internal ventilation systems (Yan et al., 2023), but they have various limitations, such as a high-pressure drop, clogging due to long-term use, and limited flowrate depending on the effective area of the filter (Davis and Kim, 1999). In addition, frequent maintenance of the filter depends on the ventilation volume, which places a heavy burden on the operator.

Deep ultraviolet (DUV)-LEDs have attracted attention in recent years as a technology for killing bacteria and viruses (Gerchman et al., 2020, Beck et al., 2017, Kim et al., 2017). Recently, the performance of DUV-LEDs has increased dramatically, achieving a lifetime of more than 10,000 h (Ippommatsu, 2013), which can significantly reduce the maintenance frequency. The UV-C light (100–280 nm) emitted by DUV-LEDs has a shorter wavelength than visible light (Diffey, 1999). Ultraviolet rays in the UV-C wavelength range are effective in sterilizing microorganisms and inactivating viruses by causing DNA and RNA damage (Beck et al., 2017, Rattanakul and Oguma, 2018). In addition, it is known that UV light with wavelengths below 240 nm can induce damage not only to nucleic acids but also to viral components, such as proteins (Beck et al., 2014, Beck et al., 2016). Several studies have reported that ultraviolet light in the wavelength range of 200–280 nm is effective in inactivating SARS-CoV-2 (Negishi et al., 2023, Ma et al., 2021, Muramoto et al., 2021, Biffi et al., 2022, Ruetalo et al., 2021).

Table 1 shows the inactivation characteristics of SARS-CoV-2 by LED irradiation based on previous studies. Ultraviolet light in the wavelength ranges of 280–320 nm (UV-B) and 320–400 nm (UV-A) requires a high radiation dose to inactivate SARS-CoV-2 (Biasin et al., 2022, Minamikawa et al., 2021, Santis et al., 2021). In contrast, ultraviolet light in the UV-C wavelength range (<280 nm) acts quickly to inactivate SARS-CoV-2 and is highly effective (Inagaki et al., 2020, Kitagawa et al., 2021, Sabino et al., 2020, Ueki et al., 2022, Lee et al., 2022). Furthermore, Inagaki et al., 2020, Inagaki et al., 2021 irradiated several viral mutants with DUV light at a wavelength of 280 nm under identical experimental conditions and found that all mutants exhibited similar inactivation characteristics. In addition to the displacement strains mentioned above, UV-C light has been shown to be effective against various strains of SARS-CoV-2 (Mancini et al., 2022, Shimoda et al., 2021, Trivellin et al., 2021).

Table 1.

Summary of inactivation experiments in SARS-CoV-2.

Virus solution (on the plate or in a Petri dish or a well plate)
Authors (Virus type)	LED wavelength [nm]	Infection titer reduction ratio [%]	Total doses of LED energy [mJ/cm2]	Exposure time [s]
Kitagawa et al. (2021) (SARS-CoV-2 2019-nCoV/Japan/AI/I-004/2020)	222	88.5 99.7	1 3	10 30
Sabino et al. (2020)	254	90 99 99.9	0.016 0.706 6.556	0.32 2.98
Ueki et al. (2022) (SARS-CoV-2 UT-NCGM02/Human/2020/Tokyo)	265	90 99 99.9	2.16 5.08 9.02	0.04 0.094 0.167
Lee et al. (2022) (NCCP43326)	275	≥ 99.99	10	< 10
Inagaki et al. (2020) (SARS-CoV-2/Hu/DP/Kng/19-027、LC528233)	280 ± 5	87.4 99.9 > 99.9	3.75 37.5 75	1 10 20
Inagaki et al. (2021) (UK strain, B.1.1.7)	280 ± 5 (Continuous irradiation)	96.3 > 99.9	3.75 18.75	1 5
Inagaki et al. (2021) (South African strain, B.1.351)	280 ± 5 (Continuous irradiation)	94.6 > 99.9	3.75 18.75	1 5
Inagaki et al. (2021) (Brazilian strain, P.1)	280 ± 5 (Continuous irradiation)	91.9 > 99.8 > 99.8	3.75 18.75 37.5	1 5 10
Minamikawa et al (2021) (SARS-CoV-2/Hu/DP/Kng/19–020, Genbank: LC528232)	265 280 300	99.9 99.9 99.9	1.8 3.0 23	- - -
Biasin et al. (2022) (2019- nCoV/Italy-INMI1, Rome)	278 308 366 405	99 99 99 99	2.07 309.07 9719.65 13224.56	- - - -
De Santis et al. (2021)	413 ± 5	∼ 99	–	3600
Floating aerosol particle
Authors (Virus type)	LED wavelength [nm]	Infection titer reduction ratio [%]	Total doses of LED energy [mJ/cm2]	Exposure time [s]
Ruetalo et al. (2022) (icSARS-CoV-2-mNG)	254	> 99.9	0.42–0.51	–
Ueki et al. (2022) (SARS-CoV-2 UT-NCGM02/Human/2020/Tokyo)	265	90 99 99.9	0.23 0.4 1.04	0.0043 0.0074 0.0193
Takamure et al. (2022) (GISAID ID# EPI_ISL_568558)	280	> 99.38	35.36	7.6

The effects of DUV-LEDs on aerosolized SARS-CoV-2 have rarely been reported (Ruetalo et al., 2022, Ueki et al., 2022, Takamure et al., 2022). Among the few reports, Ruetalo et al., 2022, Ueki et al., 2022 irradiated aerosolized SARS-CoV-2 with 254-nm and 265-nm DUV-LEDs and achieved a virus inactivation rate of 99.9% at 0.42–0.51 mJ/cm² and 1.04 mJ/cm², respectively. This finding indicates that approximately 1/10 of the total irradiation dose can inactivate the virus to the limit of detection, compared to adding the virus solution to the Petri dish. However, because the motion of aerosolized viruses is not controlled, precisely estimating the radiation doses required is difficult. Therefore, the total irradiation dose required for viral inactivation remains unknown.

The authors previously developed a non-filter virus inactivation unit with low-pressure drops (Takamure et al., 2022). This unit has multiple 180° sharp turns, and a DUV-LED with an emission wavelength of 280 nm is mounted on the sharp turns. The design of this unit has been described in detail (Takamure et al., 2022). A significant advantage of using LEDs with an emission wavelength of 280 nm is the lower Al concentration in AlGaN compared to LEDs operating at shorter wavelengths, which facilitates the device manufacturing process and improves quality (Yoshinobu et al., 2012). In addition, 280 nm wavelength LEDs generally have a longer lifetime than shorter wavelengths (Yoshinobu et al., 2012). In a previous study by the authors, the SARS-CoV-2 virus was inactivated to detection limits (>99.38%) when all 12 DUV-LEDs in the inactivation unit were turned on at maximum output, although this may have been an excessive radiation dose (Takamure et al., 2022). Understanding the appropriate radiation dose to inactivate aerosolized viruses is important for optimizing the conditions, and it provides essential information for the realization of energy savings and miniaturization of devices.

In our previous research (Takamure et al., 2022), the intensity of the LED illuminating the interior of the virus inactivation unit remained constant. Therefore, the scope of the findings was considerably limited and could solely be applicable under specific conditions. However, in the present study, we elucidated the effect of varying LED irradiation intensities on SARS-CoV-2. We achieved uniform changes in the irradiation intensity of the viral inactivation unit by controlling the irradiation pattern of the DUV-LEDs. The inner wall was coated with Teflon to increase the reflectivity of DUV rays to achieve uniform LED irradiation. This study aimed to determine the irradiation dose required to inactivate aerosolized SARS-CoV-2. The viral inactivation characteristics of aerosolized SARS-CoV-2 were investigated by varying the total irradiation dose of DUV rays used to irradiate SARS-CoV-2. The findings obtained in this study will enable the design optimization of various devices and appliances, including medical devices and home air purifiers that use deep ultraviolet rays. This ultimately contributes to the improvement of versatility, such as installation in narrow spaces due to notable downsizing and installation in large-space facilities due to higher flowrates.

2. Experimental and numerical simulation setups

2.1. Virus inactivation unit

The virus inactivation unit used in this study was similar to that designed by Takamure et al. (2022). Fig. 1 (a) shows a photograph of the virus inactivation unit, and Fig. 1(b) shows its internal structure and dimensions. The inner diameter of the conduit is a square of 50 mm on each side, and the length of the straight part between the sharp turns is 800 mm. The inlet and outlet of the conduit consist of circular cross-sectional channels with an inner diameter of 48 mm. Three DUV-LEDs (VPS164) from Nikkiso Giken Co., Ltd. were installed equally spaced at 20 mm intervals on the end face of each sharp-turn section, and a total of 12 LEDs were installed. The peak emission wavelength, current, and output of the DUV-LEDs were 280 nm, 350 mA, and 40 mW, respectively. The relative radiant intensity of the LED peaks at an angle of ± 25° and maintains more than 70% in the range from − 50° to 50°. Considering the heat dissipation characteristics of the DUV-LEDs, aluminum with good thermal conductivity was used at the end face of each sharp-turn section in contact with the LED. Except at the end face of each sharp-turn section, the conduit was composed of a stainless-steel prismatic pipe. The inner wall of the prismatic pipe was coated with Teflon to reflect and diffuse the DUV rays uniformly.

Fig. 1.

Virus inactivation unit. (a) Photo, (b) dimension outline.

2.2. Setup for inactivation experiments with SARS-CoV-2

The SARS-CoV-2 inactivation experiments using the virus inactivation unit were conducted at a Biosafety Level 3 (BSL-3) facility owned by the National Hospital Organization Nagoya Medical Center. The B1.1 lineage isolate of SARS-CoV-2 (GISAID ID# EPI_ISL_568558) was used in the virus inactivation experiments as previously described (Takamure et al., 2022). Briefly, the virus was propagated using Vero E6 cells (American Type Culture Collection) in Dulbecco’s modified Eagle’s medium (Merck) supplemented with 10% fetal bovine serum and penicillin (100 units/mL) and streptomycin (100 μg/mL) (Thermo Fisher Scientific, Waltham, MA, USA). The harvested virus was once stored at − 80 °C before use. For nebulizing experiments, the virus stock was diluted in a 1:10 ratio with phosphate-buffered saline (Thermo Fisher Scientific). The viral infection titers and RNA concentrations of the diluted SARS-CoV-2 used in the experiments were 5.2 × 10³ plaque-forming units (PFU)/mL and 8 × 10⁷ copies/mL, respectively.

The experimental system was based on the previous report by Takamure et al. (2022). A schematic diagram of the experimental setup is shown in Fig. 2 (a). The air flowrate through the virus inactivation unit was feedback-controlled using an air sampler (MD8 Airscan, Sartorius). A gelatin filter (17528-80BZD, Sartorius) with a diameter of 80 mm and a pore size of 3 μm was installed above the air sampler to collect the aerosolized SARS-CoV-2 passing through the conduit. A SARS-CoV-2 solution of 7 mL was set in the nebulizer kit of the compressor nebulizer (NE-C 28, OMRON) in advance, and when the compressor was operated, particles of nebulized virus solution with an average diameter of 5 μm were generated from the tip of the nebulizer kit at the flowrate of 0.4 mL/min. A total of 0.4 mL of virus solution was nebulized for 1 min. Air containing nebulized virus particles was transported to a cylindrical acrylic virus dissemination channel with an inner diameter of 100 mm, where it merged with clean air. The DC fan directly connected to the virus dissemination channel was rotated at a very low speed to prevent the backflow of nebulized virus particles and condensation on the wall. Air-containing viral particles that passed through the virus dissemination channel flowed into the virus inactivation unit. Virus-inactivated particles were eventually collected using a gelatin filter, and only air was released via an air sampler. The gelatin filter guarantees a virus collection rate of 99.76% at temperatures below 30 °C and humidity below 80–85% (Jaschhof, 1992). Fig. 2(b) shows the experimental system was placed in an acrylic box 900 mm wide, 350 mm high, and 400 mm deep, and sealed to prevent virus leakage. In this study, the temperature and humidity for each experiment were monitored using an external sensor-equipped thermometer and hygrometer (AD-5682, A&D) and a placed-type thermometer and hygrometer (TT557, TANITA). The temperature and humidity inside the acrylic box were maintained at 15–16 °C and 30–45% humidity, respectively. An acrylic box containing the experimental system was installed in the safety cabinet of the BSL −3 facility.

Fig. 2.

Experimental system. (a) Configuration diagram and (b) photograph of the experiment in an acrylic box located in the safety cabinet of the biosafety lab (BSL)-3 facility.

In this study, the irradiation intensity inside the conduit was controlled by changing the irradiation pattern of the 12 DUV-LEDs installed in the virus inactivation unit. Fig. 3 shows five irradiation patterns (Pattern A–E). Each pattern corresponds to the following case:

Pattern A: All 12 LEDs are turned on, as shown in Fig. 3(a).

Pattern B: A total of eight LEDs are turned on; both ends of three LEDs mounted on the edge of each sharp turn part are turned on, as shown in Fig. 3(b).

Pattern C: A total of four LEDs are turned on; the centers of the three LEDs mounted on the edge of each sharp turn part are turned on, as shown in Fig. 3(c).

Pattern D: A total of two LEDs are turned on; the centers of the two LEDs mounted on the edge of the sharp turn located at the upper right and lower left of the virus inactivation unit are turned on, as shown in Fig. 3(d).

Pattern E: All LEDs are turned off, as shown in Fig. 3(e).

Fig. 3.

Deep UV-LED lighting pattern.

Table 2 lists the irradiation characteristics of the LEDs for each pattern. The irradiation intensity weakens from patterns A to E. Patterns A, B, and C uniformly irradiate the inside of the conduit, whereas Pattern D results in non-uniform irradiation where the irradiation intensity is stronger in the central straight part compared to that in other areas.

Table 2.

Characteristics in each irradiation pattern.

LED irradiation pattern	Number of illuminated LEDs	Irradiation conditions inside the conduit	Flux per unit area [mW/cm2]
Pattern A	12	Uniform irradiation	4.8
Pattern B	8	Uniform irradiation	3.2
Pattern C	4	Uniform irradiation	1.6
Pattern D	2	Non-uniform irradiation	1.07
Pattern E	0	Non-irradiation	0

The virus-containing gelatin filter was suspended in 5 mL of phosphate-buffered saline (PBS), heated at 37 °C, and then centrifuged at 2500g for 15 min. The filtrate was obtained using a 0.45 μm pore size membrane filter. The inactivation characteristics of SARS-CoV-2 were evaluated by RNA and nucleocapsid protein quantification. The RNA of SARS-CoV-2 was extracted from the filtrate, using the QIAamp Viral RNA Mini Kit (QIAGEN, Hilden, Germany), following the manufacturer's protocol. A total of 140 µL of the filtrate was used for extraction, resulting in an elution volume of 60 µL. The extracted RNA (5 µL) was subjected to RNA quantification by reverse transcription-quantitative polymerase chain reaction (RT-qPCR) using the PrimeDirect Probe RT-qPCR Mix (Takara Bio, Shiga, Japan; cat#RR650A) and an N2 primer/probe set according to the laboratory diagnostic protocol for SARS-CoV-2 developed by the National Institute of Infectious Diseases in Japan (Shirato et al., 2020). The oligonucleotide sequences for the N2 primer set and N2 probe are 5́-AAATTTTGGGGACCAGGAAC (forward) / 5́-TGGCAGCTGTGTAGGTCAAC (reverse) and 6-FAM-ATGTCGCGCATTGGCATGGA-BHQ, respectively. The primer/probe set for RT-qPCR and positive control RNA for the standard curves were purchased from Applied Biosystems (Waltham, MA, USA) and Nihon Gene Research Laboratories, respectively. The reaction mixture (20 µL) comprised the N2 forward and reverse primers, along with the probe, at final concentrations of 0.5, 0.7, and 0.2 μM, respectively. The triplex assays were performed using the Thermal Cycler Dice Real Time System III (Takara Bio) under the following thermal cycling conditions: reverse transcription at 52 °C for 5 min, initial denaturalization at 95 °C for 10 s, followed by 45-cycle amplification at 95 °C for 5 s and 60 °C for 30 s. Copy number analysis and generation of linear regression curves were conducted using the Multiplate RQ (Takara Bio) software integrated into the Thermal Cycler Dice Real Time System III. The lower limit of detection for RT-qPCR was set at 10 copies per reaction. The copy numbers of viral RNA were determined using the Thermal Cycler Dice Real Time System III (Takara Bio). The nucleocapsid protein was quantified using an ELISA-based assay with the SARS-CoV-2 nucleocapsid Protein Titer Assay Kit (Product No. RAS-A010, Acro Biosystems). The linear dynamic range of this assay is approximately from 20 to 600 pg/mL, as it demonstrates a linear relationship between OD_450nm value and the concentration of nucleocapsid protein recognized by two monoclonal antibodies. In the present study, the protein concentrations within the linear dynamic range of the assay (20–600 pg/mL) were measured following the manufacturer's protocol. Exposing viruses to DUV radiation does not affect the quantity of nucleocapsid protein (Lo et al., 2021). Therefore, it should be noted that the measurement of nucleocapsid protein quantity using ELISA in this study was not conducted to examine virus damage, but to confirm the amount of virus captured by the gelatin filter through nebulization. The viral infection titer was determined by a plaque assay using VeroE6/TMPRSS2 cells. The measured viral load and infection titer were estimated as the concentration and infectivity titer of the 5 mL gelatin filter suspension.

2.3. Numerical simulation conditions

The time required for the virus to pass through the virus inactivation unit was estimated using numerical simulation (SCRYU/Tetra, Software Cradle Co., Ltd.). Similar simulation conditions were used by Takamure et al. (2022). Large eddy simulations (LES) were performed to simulate the airflow. The standard Smagorinsky model was applied to the turbulence model. The number of grid points in the computational domain was approximately 31,250,000. The boundary conditions for the inlet and outlet regions are the uniform flowrate condition of Q_inlet and the static pressure condition, respectively. In addition, a no-slip condition was applied to the inner wall surface of the virus activation unit.

The procedure of this simulation is as follows: After running the simulation for 5 sec in advance, particles with a diameter of 5 μm are injected at 500 particles/sec, and a total of 10,000 particles are injected into the virus inactivation unit for 20 sec. The total simulation time is 40 sec. Air temperature and particle density are set at 15 °C and 999.1 kg/m³, respectively.

3. Results and discussion

3.1. Numerical simulation results

In this study, numerical simulations were conducted at three flowrates to control the irradiation time for aerosolized viruses. The three flowrates were specified as Q_inlet = 0.03, 0.04, and 0.05 m³/min (corresponding to U_inlet = 0.2, 0.27, and 0.33 m/s, respectively, in terms of inlet flow velocity U_inlet). Fig. 4 shows the duration required for virus particles to pass through the virus inactivation unit. The virus particles that collided and were collected inside the virus inactivation unit were not counted. The virus particles passed through the virus inactivation unit in a shorter time as the flowrate increased. The ensemble average $\overline{t_p}$ of the time through the virus inactivation unit at Q_inlet = 0.03, 0.04, and 0.05 m³/min are $\overline{t_p}$ = 13.1, 10.3, and 7.6 s, respectively.

Fig. 4.

Passage time of particles passing through a rectangular conduit. (a) Q_inlet = 0.03 m³/min, (b) Q_inlet = 0.04 m³/min, and (c) Q_inlet = 0.05 m³/min. The particle diameter is 5 μm.

Table 3 summarizes the results of the numerical simulations for the dimensionless pressure drop 2Δp/{ρ(U_inlet)²} where Δp (=P_inlet - P_outlet) is the pressure difference between the inlet pressure P_inlet and outlet pressure P_outlet, and ρ is the density of air, time $\overline{t_p}$, and particle collision rate to the inner wall of virus inactivation unit at each Q_inlet. The dimensionless pressure drop has a relatively constant value of approximately 2Δp/{ρ(U_inlet)²} = 19, independent of Q_inlet. Moreover, the particle collision rates were in the range of 44–56% in all cases, with no significant differences due to changes in Q_inlet.

Table 3.

Results of various parameters when changing flowrate in numerical simulation.

Qinlet [m3/min]	Pressure drop 2Δp/{ρ(Uinlet)2} [-]	$\overline{{\mathit{t}}_{\mathit{p}}}$ [s]	Particle collision rate [%]
0.03	18.6	13.1	44
0.04	19.1	10.3	51
0.05	18.9	7.6	56

2Δp/{ρ(U_inlet)²} where Δp (=P_inlet - P_outlet) is the pressure difference between the inlet pressure P_inlet and outlet pressure P_outlet, and ρ is the density of air, time $\overline{t_p}$, and particle collision rate to the inner wall of virus inactivation unit at each Q_inlet.

3.2. Experimental results

The total irradiation dose of the virus particles passing through the virus inactivation unit was controlled by combining the inlet flowrate, Q_inlet and the irradiation pattern of the LEDs. Table 4 shows the inlet flowrate Q_inlet, the LED irradiation pattern, and the total irradiation dose in each case (Exp. #1–8). The total irradiation dose varied from 0 to 36.5 mJ/cm². Triplicate measurements were performed in each case.

Table 4.

Conditions and results in each experimental case (Exp #1–8).

Case	Qinlet [m3/min]	LED irradiation pattern	Total doses of LED energy [mJ/cm2]	Inactivation rate
Exp#1	0.05	Pattern A	36.5	LOD
Exp#2	0.04	Pattern B	33	LOD
Exp#3	0.05	Pattern B	24.3	LOD
Exp#4	0.03	Pattern C	20.1	LOD
Exp#5	0.04	Pattern C	16.5	LOD
Exp#6	0.05	Pattern C	12.2	LOD
Exp#7	0.05	Pattern D	8.1	99.16%
Exp#8	0.05	Pattern E	0	---

The limit of detection (LOD) corresponds to an inactivation rate of 99.89%.

Fig. 5(a) shows the RNA concentration in the N2 region following inactivation experiments in each case. The quantitative assay for viral RNA was performed using RT-qPCR, targeting 158 bases long in the N region. When the total irradiation dose of LED energy exceeded 16.5 mJ/cm² (Exp #1–5), the RNA concentration of SARS-CoV-2 converged to approximately 2 × 10⁷ copies/mL. This suggests that RNA damage may have occurred in regions below the detection threshold of RT-qPCR measurement. However, when the total irradiation dose was less than 16.5 mJ/cm², the RNA concentration monotonically increased with a decreasing LED irradiation dose.

Fig. 5.

Experimental results. (a) RNA concentration based on RT-qPCR targeted for a short amplicon, (b) nucleocapsid protein concentration, and (c) ratio of RNA to the nucleocapsid protein of SARS-CoV-2 virus under each experimental case.

Fig. 5(b) shows the nucleocapsid protein concentration measurement results. In Experiments #6–8, where the total irradiation dose of LED energy was small, the nucleocapsid protein concentration became slightly larger compared with the other cases, but there was no significant tendency for the total irradiation dose. Lo et al. (2021) irradiated a Petri dish with drops of SARS-CoV-2 solution using DUV light at a wavelength of 253.7 nm and investigated the amount of RNA and nucleocapsid proteins present when the infectivity decreased to 99.99%. The amount of RNA was reduced compared to that before DUV irradiation, but the amount of nucleocapsid protein was not. That is, the aerosolized SARS-CoV-2 in this experiment showed the same trend as that in the Lo et al. (2021) experiment.

Fig. 5(c) shows the ratio of the RNA concentration to the nucleocapsid protein concentration (expressed as the RNA/N ratio). The RNA/N ratio was similar to the distribution of the RNA concentration in Fig. 5(a). This finding indicates that DUV irradiation acts only on RNA.

Fig. 6 shows the plaque assay for each case, where each case shows the results of three experiments with solutions diluted 2, 20, and 200 times, respectively. In the control experiment without UV irradiation, plaques were confirmed at all dilution rates. (Exp #8). The average infectious titer of the control experiment without UV irradiation (Exp #8) was 4,408 PFU/mL. Experiments #1–6 showed no plaques at any dilution rate, indicating a virus reduction rate of 99.89%. However, in Exp #7, the presence of plaques was confirmed at dilution columns of 1/2 and 1/20. The average infectious titer of Exp #7 was 37 PFU/mL, which corresponds to a virus reduction rate of 99.16%.

Fig. 6.

Results of the plaque assay in each experimental case.

Compared with the experimental results of the inactivation of aerosolized SARS-CoV-2 conducted by Ruetalo et al. (2022) and Ueki et al. (2022)(Table 1), the total irradiation dose required for a virus inactivation rate of nearly 99.9% in this study was slightly higher. This result was expected as the emission wavelengths of the LEDs used in the previous studies were 254 nm (Ruetalo et al., 2022) and 265 nm (Ueki et al., 2022), shorter than those used in this study. Previous studies in which SARS-CoV-2 solution was dropped onto a Petri dish or well plate also showed that DUV light with a wavelength of 254 nm (Sabino et al., 2020) or 265 nm (Ueki et al., 2022) could inactivate the virus by up to 99.9% at less irradiation dose compared to that with a wavelength of 280 nm (Inagaki et al., 2020).

The irradiation dose required to inactivate aerosolized SARS-CoV-2 to 99.9% in DUV light at a wavelength of 280 nm is one-sixth that required to inactivate SARS-CoV-2 in a Petri dish (Inagaki et al., 2020). In addition, a previous study by the authors (Takamure et al., 2022) showed an inactivation rate of 99.38% at 35.36 mJ/cm² for aerosolized SARS-CoV-2, whereas this study achieved a virus inactivation rate of 99.89% (below the detection limit) at 12.2 mJ/cm² and 99.16% at 8.1 mJ/cm². This means that more than 99% of the virus can actually be inactivated at an irradiation dose of 23% of the maximum output of the LED mounted on the virus inactivation unit. These findings will contribute to space saving and a higher flowrate of the virus inactivation unit, which is expected to improve the versatility of the system.

4. Conclusions

A non-filter virus inactivation unit was developed to regulate the irradiation dose to aerosolized viruses by controlling the lighting pattern of a 280 nm DUV-LED and adjusting the air flowrate. The objective of this study was to quantitatively evaluate the inactivation properties of aerosolized SARS-CoV-2 by manipulating the irradiation dose inside the inactivation unit. The results obtained are summarized as follows:

• 1.The RNA concentration of SARS-CoV-2 was constant when the total irradiation dose of DUV irradiation of the virus exceeded 16.5 mJ/cm². However, when the total irradiation dose was less than 16.5 mJ/cm², the RNA concentration monotonically increased with a decreasing LED irradiation dose.
• 2.The nucleocapsid protein concentration did not show a significant trend, even with variations in the total irradiation dose.
• 3.The plaque assay revealed that at an irradiation dose of 8.1 mJ/cm², 99.16% of the virus was effectively inactivated. Furthermore, at a higher irradiation dose of 12.2 mJ/cm², no traces of the virus were detected, indicating a virus inactivation rate of 99.89%. In other words, more than 99% of SARS-CoV-2 can be inactivated at an irradiation dose of 23% of the maximum irradiation capacity of the virus inactivation unit.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Thanh Nguyen

Data availability

Data will be made available on request.